ZTWHHH commited on
Commit
5089912
·
verified ·
1 Parent(s): 84b1f54

Add files using upload-large-folder tool

Browse files
This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. .gitattributes +1 -0
  2. janus/share/man/fr/man1/lzcat.1 +1985 -0
  3. janus/share/man/fr/man1/lzless.1 +49 -0
  4. janus/share/man/fr/man1/lzma.1 +1985 -0
  5. janus/share/man/fr/man1/lzmadec.1 +90 -0
  6. janus/share/man/fr/man1/lzmainfo.1 +47 -0
  7. janus/share/man/fr/man1/unlzma.1 +1985 -0
  8. janus/share/man/fr/man1/unxz.1 +1985 -0
  9. janus/share/man/fr/man1/xz.1 +1985 -0
  10. janus/share/man/fr/man1/xzcat.1 +1985 -0
  11. janus/share/man/fr/man1/xzdec.1 +90 -0
  12. janus/share/man/fr/man1/xzless.1 +49 -0
  13. janus/share/man/man1/lzcat.1 +3198 -0
  14. janus/share/man/man1/lzmore.1 +65 -0
  15. janus/share/man/man1/sqlite3.1 +161 -0
  16. janus/share/man/man1/unxz.1 +3198 -0
  17. janus/share/man/man1/xzgrep.1 +144 -0
  18. janus/share/man/man1/xzless.1 +67 -0
  19. janus/share/man/man1/xzmore.1 +65 -0
  20. janus/share/man/pt_BR/man1/lzcat.1 +1951 -0
  21. janus/share/man/uk/man1/lzdiff.1 +55 -0
  22. janus/share/zoneinfo/Asia/Almaty +0 -0
  23. janus/share/zoneinfo/Asia/Barnaul +0 -0
  24. janus/share/zoneinfo/Asia/Bishkek +0 -0
  25. janus/share/zoneinfo/Asia/Calcutta +0 -0
  26. janus/share/zoneinfo/Asia/Chungking +0 -0
  27. janus/share/zoneinfo/Asia/Dhaka +0 -0
  28. janus/share/zoneinfo/Asia/Gaza +0 -0
  29. janus/share/zoneinfo/Asia/Hebron +0 -0
  30. janus/share/zoneinfo/Asia/Katmandu +0 -0
  31. janus/share/zoneinfo/Asia/Khandyga +0 -0
  32. janus/share/zoneinfo/Asia/Makassar +0 -0
  33. janus/share/zoneinfo/Asia/Novokuznetsk +0 -0
  34. janus/share/zoneinfo/Asia/Pontianak +0 -0
  35. janus/share/zoneinfo/Asia/Pyongyang +0 -0
  36. janus/share/zoneinfo/Asia/Rangoon +0 -0
  37. janus/share/zoneinfo/Asia/Samarkand +0 -0
  38. janus/share/zoneinfo/Asia/Seoul +0 -0
  39. janus/share/zoneinfo/Asia/Shanghai +0 -0
  40. janus/share/zoneinfo/Asia/Thimphu +0 -0
  41. janus/share/zoneinfo/Asia/Ulan_Bator +0 -0
  42. janus/share/zoneinfo/Asia/Ust-Nera +0 -0
  43. janus/share/zoneinfo/Asia/Yerevan +0 -0
  44. janus/share/zoneinfo/Cuba +0 -0
  45. janus/share/zoneinfo/EST +0 -0
  46. janus/share/zoneinfo/EST5EDT +0 -0
  47. janus/share/zoneinfo/Egypt +0 -0
  48. janus/share/zoneinfo/Factory +0 -0
  49. janus/share/zoneinfo/GB +0 -0
  50. janus/share/zoneinfo/GMT +0 -0
.gitattributes CHANGED
@@ -529,3 +529,4 @@ janus/lib/python3.10/lib-dynload/_decimal.cpython-310-x86_64-linux-gnu.so filter
529
  janus/lib/sqlite3.44.2/libsqlite3.44.2.so filter=lfs diff=lfs merge=lfs -text
530
  janus/lib/itcl4.2.4/libitcl4.2.4.so filter=lfs diff=lfs merge=lfs -text
531
  janus/x86_64-conda-linux-gnu/bin/ld filter=lfs diff=lfs merge=lfs -text
 
 
529
  janus/lib/sqlite3.44.2/libsqlite3.44.2.so filter=lfs diff=lfs merge=lfs -text
530
  janus/lib/itcl4.2.4/libitcl4.2.4.so filter=lfs diff=lfs merge=lfs -text
531
  janus/x86_64-conda-linux-gnu/bin/ld filter=lfs diff=lfs merge=lfs -text
532
+ janus/x86_64-conda_cos7-linux-gnu/bin/ld filter=lfs diff=lfs merge=lfs -text
janus/share/man/fr/man1/lzcat.1 ADDED
@@ -0,0 +1,1985 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ '\" t
2
+ .\" SPDX-License-Identifier: 0BSD
3
+ .\"
4
+ .\" Authors: Lasse Collin
5
+ .\" Jia Tan
6
+ .\"
7
+ .\" french translation of XZ Utils man
8
+ .\" Copyright (C) 2021 Debian French l10n team <debian-l10n-french@lists.debian.org>
9
+ .\" Translator
10
+ .\" bubu <bubub@no-log.org>, 2021.
11
+ .\"
12
+ .\"*******************************************************************
13
+ .\"
14
+ .\" This file was generated with po4a. Translate the source file.
15
+ .\"
16
+ .\"*******************************************************************
17
+ .TH XZ 1 2024\-12\-30 Tukaani "Utilitaires XZ"
18
+ .
19
+ .SH NOM
20
+ xz, unxz, xzcat, lzma, unlzma, lzcat \- Compresser ou décompresser des
21
+ fichiers .xz et .lzma
22
+ .
23
+ .SH SYNOPSIS
24
+ \fBxz\fP [\fIoption...\fP] [\fIfichier...\fP]
25
+ .
26
+ .SH "ALIAS DES COMMANDES"
27
+ \fBunxz\fP est équivalent à \fBxz \-\-decompress\fP.
28
+ .br
29
+ \fBxzcat\fP est équivalent à \fBxz \-\-decompress \-\-stdout\fP
30
+ .br
31
+ \fBlzma\fP est équivalent à \fBxz \-\-format=lzma\fP
32
+ .br
33
+ \fBunlzma\fP est équivalent à \fBxz \-\-format=lzma \-\-decompress\fP
34
+ .br
35
+ \fBlzcat\fP est équivalent à \fBxz \-\-format=lzma \-\-decompress \-\- stdout\fP
36
+ .PP
37
+ Lors de l'écriture de scripts qui nécessitent de décompresser des fichiers,
38
+ il est recommandé de toujours utiliser la commande \fBxz\fP avec les arguments
39
+ appropriés (\fBxz \-d\fP ou \fBxz \-dc\fP) au lieu des commandes \fBunxz\fP et
40
+ \fBxzcat\fP.
41
+ .
42
+ .SH DESCRIPTION
43
+ \fBxz\fP is a general\-purpose data compression tool with command line syntax
44
+ similar to \fBgzip\fP(1) and \fBbzip2\fP(1). The native file format is the
45
+ \&\fB.xz\fP format, but the legacy \fB.lzma\fP format used by LZMA Utils and raw
46
+ compressed streams with no container format headers are also supported. In
47
+ addition, decompression of the \fB.lz\fP format used by \fBlzip\fP is supported.
48
+ .PP
49
+ \fBxz\fP compresse ou décompresse chaque \fIfichier\fP en fonction du mode
50
+ d'opération choisi. Si aucun \fIfichier\fP n'est donné ou \fIfichier\fP est \fB\-\fP,
51
+ \fBxz\fP lit depuis l'entrée standard et écrit les données traitées sur la
52
+ sortie standard. \fBxz\fP refusera (affichera une erreur et ignorera le
53
+ \fIfichier\fP) d'écrire les données compressées sur la sortie standard si c'est
54
+ un terminal. De même, \fBxz\fP refusera de lire des données compressées depuis
55
+ l'entrée standard si c'est un terminal.
56
+ .PP
57
+ A moins que \fB\-\-sdout\fP ne soit indiqué, les \fIfichiers\fP autres que \fB\-\fP sont
58
+ écrits dans un nouveau fichier dont le nom est dérivé du nom de \fIfichier\fP
59
+ source :
60
+ .IP \(bu 3
61
+ Lors de la compression, le suffixe du format de fichier cible (\fB.xz\fP ou
62
+ \&\fB.lzma\fP) est ajouté au nom de fichier source pour obtenir le nom de fichier
63
+ cible.
64
+ .IP \(bu 3
65
+ When decompressing, the \fB.xz\fP, \fB.lzma\fP, or \fB.lz\fP suffix is removed from
66
+ the filename to get the target filename. \fBxz\fP also recognizes the suffixes
67
+ \&\fB.txz\fP and \fB.tlz\fP, and replaces them with the \fB.tar\fP suffix.
68
+ .PP
69
+ Si le fichier cible existe déjà, une erreur est affichée et le \fIfichier\fP
70
+ est ignoré.
71
+ .PP
72
+ Sauf s'il écrit dans la sortie standard, \fBxz\fP affichera un avertissement et
73
+ ignorera le \fIfichier\fP dans les cas suivants :
74
+ .IP \(bu 3
75
+ \fIfichier\fP n'est pas un fichier normal. Les liens symboliques ne sont pas
76
+ suivis et donc ne sont pas considérés comme des fichiers normaux.
77
+ .IP \(bu 3
78
+ \fIfichier\fP a plusieurs liens physiques.
79
+ .IP \(bu 3
80
+ \fIfichier\fP a un setuid, setgid ou sticky bit positionné.
81
+ .IP \(bu 3
82
+ Le mode d'opération est défini pour compresser et le \fIfichier\fP a déjà un
83
+ suffixe du format de fichier cible (\fB.xz\fP ou \fB.txz\fP lors d'une compression
84
+ en format \fB.xz\fP, et \fB.lzma\fP ou \fB.tlz\fP lors d'une compression en format
85
+ \&\fB.lzma\fP).
86
+ .IP \(bu 3
87
+ The operation mode is set to decompress and the \fIfile\fP doesn't have a
88
+ suffix of any of the supported file formats (\fB.xz\fP, \fB.txz\fP, \fB.lzma\fP,
89
+ \&\fB.tlz\fP, or \fB.lz\fP).
90
+ .PP
91
+ Après la compression ou la décompression réussie du \fIfichier\fP, \fBxz\fP copie
92
+ les permissions du propriétaire, du groupe, la date d'accès, et les
93
+ modifications d'heure depuis le \fIfichier\fP source du fichier cible. Si la
94
+ copie du groupe échoue, les permissions sont modifiées pour que le fichier
95
+ cible ne soit pas accessible aux utilisateurs qui n'ont pas les droits
96
+ d'accès au \fIfichier\fP source. \fBxz\fP ne prend actuellement pas en charge la
97
+ copie d'autres métadonnées telles que les listes de contrôle d'accès ou les
98
+ attributs étendus.
99
+ .PP
100
+ Once the target file has been successfully closed, the source \fIfile\fP is
101
+ removed unless \fB\-\-keep\fP was specified. The source \fIfile\fP is never removed
102
+ if the output is written to standard output or if an error occurs.
103
+ .PP
104
+ Envoyer \fBSIGINFO\fP ou \fBSIGURSR1\fP au processus \fBxz\fP, lui fait afficher
105
+ l'information de progression sur l'erreur standard. Cela a un intérêt limité
106
+ car lorsque l'erreur standard est un terminal, utiliser \fB\-\-verbose\fP
107
+ affichera automatiquement un indicateur de progression du processus.
108
+ .
109
+ .SS "Utilisation de la mémoire"
110
+ L'utilisation de la mémoire par \fBxz\fP varie de quelques centaines de
111
+ kilo\-octets à plusieurs gigaoctects en fonction des paramètres de
112
+ compression. Les réglages utilisés lors de la compression d'un fichier
113
+ déterminent les besoins en mémoire pour la décompression. Habituellement la
114
+ décompression nécessite 5\% à 20\% de la quantité de mémoire utilisée pour
115
+ la compression du fichier. Par exemple, décompresser un fichier créé avec
116
+ \fBxz\-9\fP recquiert habituellement 65\ Mio de mémoire. Bien qu'il soit
117
+ possible d'avoir des fichiers \fB.xz\fP nécessitant plusieurs gigaoctets de
118
+ mémoire pour être décompressés.
119
+ .PP
120
+ Especially users of older systems may find the possibility of very large
121
+ memory usage annoying. To prevent uncomfortable surprises, \fBxz\fP has a
122
+ built\-in memory usage limiter, which is disabled by default. While some
123
+ operating systems provide ways to limit the memory usage of processes,
124
+ relying on it wasn't deemed to be flexible enough (for example, using
125
+ \fBulimit\fP(1) to limit virtual memory tends to cripple \fBmmap\fP(2)).
126
+ .PP
127
+ The memory usage limiter can be enabled with the command line option
128
+ \fB\-\-memlimit=\fP\fIlimit\fP. Often it is more convenient to enable the limiter
129
+ by default by setting the environment variable \fBXZ_DEFAULTS\fP, for example,
130
+ \fBXZ_DEFAULTS=\-\-memlimit=150MiB\fP. It is possible to set the limits
131
+ separately for compression and decompression by using
132
+ \fB\-\-memlimit\-compress=\fP\fIlimit\fP and \fB\-\-memlimit\-decompress=\fP\fIlimit\fP.
133
+ Using these two options outside \fBXZ_DEFAULTS\fP is rarely useful because a
134
+ single run of \fBxz\fP cannot do both compression and decompression and
135
+ \fB\-\-memlimit=\fP\fIlimit\fP (or \fB\-M\fP \fIlimit\fP) is shorter to type on the
136
+ command line.
137
+ .PP
138
+ If the specified memory usage limit is exceeded when decompressing, \fBxz\fP
139
+ will display an error and decompressing the file will fail. If the limit is
140
+ exceeded when compressing, \fBxz\fP will try to scale the settings down so that
141
+ the limit is no longer exceeded (except when using \fB\-\-format=raw\fP or
142
+ \fB\-\-no\-adjust\fP). This way the operation won't fail unless the limit is very
143
+ small. The scaling of the settings is done in steps that don't match the
144
+ compression level presets, for example, if the limit is only slightly less
145
+ than the amount required for \fBxz \-9\fP, the settings will be scaled down only
146
+ a little, not all the way down to \fBxz \-8\fP.
147
+ .
148
+ .SS "Concaténation et remplissage avec des fichiers .xz"
149
+ Il est possible de concaténer les fichiers \fB.xz\fP tels quel. \fBxz\fP
150
+ décompressera de tels fichiers comme s'ils étaient un unique fichier \fB.xz\fP.
151
+ .PP
152
+ It is possible to insert padding between the concatenated parts or after the
153
+ last part. The padding must consist of null bytes and the size of the
154
+ padding must be a multiple of four bytes. This can be useful, for example,
155
+ if the \fB.xz\fP file is stored on a medium that measures file sizes in
156
+ 512\-byte blocks.
157
+ .PP
158
+ La concaténation et le remplissage ne sont pas autorisés avec les fichiers
159
+ \&\fB.lzma\fP ou les flux bruts.
160
+ .
161
+ .SH OPTIONS
162
+ .
163
+ .SS "Suffixes entiers et valeurs spéciales."
164
+ Dans la plupart des endroits où un argument entier est attendu, un suffixe
165
+ optionel permet d'indiquer facilement les grands entiers. Il ne doit pas y
166
+ avoir d'espace entre l'entier et le suffixe.
167
+ .TP
168
+ \fBKiB\fP
169
+ Multiplier l'entier par 1024 (2^10). \fBKi\fP, \fBk\fP, \fBkB\fP, \fBK\fP et \fBKB\fP sont
170
+ acceptés comme synonymes de \fBKiB\fP.
171
+ .TP
172
+ \fBMiB\fP
173
+ Multiplier l'entier par 1 048 576 (2^20). \fBMi\fP, \fBm\fP, \fBM\fP et \fBMB\fP sont
174
+ acceptés comme synonymes de \fBMiB\fP.
175
+ .TP
176
+ \fBGiB\fP
177
+ Multiplier l'entier par 1 073 741 824 (2^30). \fBGi\fP, \fBg\fP, \fBG\fP et \fBGB\fP
178
+ sont acceptés comme synonymes de \fBGiB\fP.
179
+ .PP
180
+ La valeur spéciale \fBmax\fP peut être utilisée pour indiquer la valeur
181
+ maximale de l'entier prise en charge par l'option.
182
+ .
183
+ .SS "Mode d'opération"
184
+ Si plusieurs options de mode d'opération sont données, la dernière prend
185
+ effet.
186
+ .TP
187
+ \fB\-z\fP, \fB\-\-compress\fP
188
+ Compresser. C'est le mode d'opération par défaut lorsque aucune option de
189
+ mode opératoire n'est spécifiée ou qu'aucun autre mode d'opération n'est
190
+ sous\-entendu par le nom de la commande (par exemple \fBunxz\fP sous\-entend
191
+ \fB\-\-decompress\fP).
192
+ .IP ""
193
+ .\" The DESCRIPTION section already says this but it's good to repeat it
194
+ .\" here because the default behavior is a bit dangerous and new users
195
+ .\" in a hurry may skip reading the DESCRIPTION section.
196
+ After successful compression, the source file is removed unless writing to
197
+ standard output or \fB\-\-keep\fP was specified.
198
+ .TP
199
+ \fB\-d\fP, \fB\-\-decompress\fP, \fB\-\-uncompress\fP
200
+ .\" The DESCRIPTION section already says this but it's good to repeat it
201
+ .\" here because the default behavior is a bit dangerous and new users
202
+ .\" in a hurry may skip reading the DESCRIPTION section.
203
+ Decompress. After successful decompression, the source file is removed
204
+ unless writing to standard output or \fB\-\-keep\fP was specified.
205
+ .TP
206
+ \fB\-t\fP, \fB\-\-test\fP
207
+ Tester l'intégrité des \fIfichiers\fP compressés. Cette option est équivalente
208
+ à \fB\-\-decompress \-\-stdout\fP sauf que les données décompressées sont rejetées
209
+ au lieu d'être écrites sur la sortie standard. Aucun fichier n'est créé ou
210
+ supprimé.
211
+ .TP
212
+ \fB\-l\fP, \fB\-\-list\fP
213
+ Afficher l'information sur les \fIfichiers\fP compressés. Aucune sortie
214
+ non\-compressée n'est produite et aucun fichier n'est créé ou supprimé. En
215
+ mode liste, le programme ne peut pas lire les données compressées depuis
216
+ l'entrée standard ou depuis d'autres sources non adressables.
217
+ .IP ""
218
+ The default listing shows basic information about \fIfiles\fP, one file per
219
+ line. To get more detailed information, use also the \fB\-\-verbose\fP option.
220
+ For even more information, use \fB\-\-verbose\fP twice, but note that this may be
221
+ slow, because getting all the extra information requires many seeks. The
222
+ width of verbose output exceeds 80 characters, so piping the output to, for
223
+ example, \fBless\ \-S\fP may be convenient if the terminal isn't wide enough.
224
+ .IP ""
225
+ La sortie exacte peut varier suivant les versions de \fBxz\fP et les différents
226
+ paramètres régionaux. Pour une sortie lisible par la machine, utiliser
227
+ \fB\-\-robot \-\-list\fP.
228
+ .
229
+ .SS "Modificateurs d'opération"
230
+ .TP
231
+ \fB\-k\fP, \fB\-\-keep\fP
232
+ Ne pas effacer les fichiers d'entrée.
233
+ .IP ""
234
+ Since \fBxz\fP 5.2.6, this option also makes \fBxz\fP compress or decompress even
235
+ if the input is a symbolic link to a regular file, has more than one hard
236
+ link, or has the setuid, setgid, or sticky bit set. The setuid, setgid, and
237
+ sticky bits are not copied to the target file. In earlier versions this was
238
+ only done with \fB\-\-force\fP.
239
+ .TP
240
+ \fB\-f\fP, \fB\-\-force\fP
241
+ Cette option a plusieurs effets :
242
+ .RS
243
+ .IP \(bu 3
244
+ Si le fichier cible existe déjà, l'effacer avant de compresser ou
245
+ décompresser.
246
+ .IP \(bu 3
247
+ Compresser ou décompresser même si l'entrée est un lien symbolique vers un
248
+ fichier normal, a plus qu'un lien physique, ou a le bit setuid, setgid ou
249
+ sticky défini. Les bits setuid, setgid et sticky bits ne sont pas copiés
250
+ dans le fichier cible.
251
+ .IP \(bu 3
252
+ Lorsque \fBxz\fP est utilisé avec \fB\-\-decompress\fP \fB\-\-stdout\fP et qu'il ne peut
253
+ pas reconnaitre le type du fichier source, copier le fichier source tel quel
254
+ dans la sortie standard. Celà permet à \fBxzcat\fP \fB\-\-force\fP d'être utilisé
255
+ comme \fBcat\fP(1) pour les fichiers qui n'ont pas été compressé avec
256
+ \fBxz\fP. Remarquez que dans le futur, \fBxz\fP devrait prendre en charge de
257
+ nouveaux formats de fichiers compressés, ce qui permettra à \fBxz\fP de
258
+ décompresser plus de types de fichiers au lieu de les copier tels quels dans
259
+ la sortie standard. \fB\-\-format=\fP\fIformat\fP peut être utilisé pour contraindre
260
+ \fBxz\fP à décompresser seulement un format de fichier.
261
+ .RE
262
+ .TP
263
+ \fB\-c\fP, \fB\-\-stdout\fP, \fB\-\-to\-stdout\fP
264
+ Écrire les données compressées ou décompressées sur la sortie standard
265
+ plutôt que dans un fichier. Cela necessite \fB\-\-keep\fP.
266
+ .TP
267
+ \fB\-\-single\-stream\fP
268
+ Décompresser seulement le premier flux \fB.xz\fP et ignorer silencieusement les
269
+ possibles données d'entrée résiduelles qui suivent le flux. Normalement ces
270
+ déchets excédentaires provoquent l'affichage d'une erreur par \fBxz\fP.
271
+ .IP ""
272
+ \fBxz\fP ne décompresse jamais plus d'un flux à partir de fichiers \fB.lzma\fP ou
273
+ de flux bruts, mais cette option fait aussi que \fBxz\fP ignorera les données
274
+ résiduelles après le fichier \fB.lzma\fP ou le flux brut.
275
+ .IP ""
276
+ Cette option n'a aucun effet si le mode d'opération n'est pas
277
+ \fB\-\-decompress\fP ou \fB\-\-test\fP.
278
+ .TP
279
+ \fB\-\-no\-sparse\fP
280
+ Désactiver la création de fichiers peu denses. Par défaut, lors de la
281
+ décompression en un fichier normal, \fBxz\fP essaie d'en faire un fichier creux
282
+ si les données décompressées contiennent de longues séquences de zéros
283
+ binaires. Cela fonctionne aussi lors de l'écriture sur la sortie standard
284
+ aussi longtemps que la sortie standard est connectée à un fichier normal et
285
+ que certaines conditions supplémentaires sont satisfaites pour le faire de
286
+ manière sécurisée. Créer des fichiers creux peut épargner de l'espace disque
287
+ et accélérer la décompression en réduisant la quantité d'entrées/sorties sur
288
+ le disque.
289
+ .TP
290
+ \fB\-S\fP \fI.suf\fP, \fB\-\-suffix=\fP\fI.suf\fP
291
+ Lors de la compression, utiliser \fB.suf\fP comme suffixe du fichier cible au
292
+ lieu de \fB.xz\fP ou \fB.lzma\fP. Si \fBxz\fP n'écrit pas sur la sortie standard et
293
+ si le fichier source a déja le suffixe \fB.suf\fP, un avertissement est affiché
294
+ et le fichier est ignoré.
295
+ .IP ""
296
+ When decompressing, recognize files with the suffix \fI.suf\fP in addition to
297
+ files with the \fB.xz\fP, \fB.txz\fP, \fB.lzma\fP, \fB.tlz\fP, or \fB.lz\fP suffix. If the
298
+ source file has the suffix \fI.suf\fP, the suffix is removed to get the target
299
+ filename.
300
+ .IP ""
301
+ Lors de la compression ou décompression de flux bruts (\fB\-\-fomat=raw\fP), le
302
+ suffixe doit toujours être spécifié à moins d'écrire sur la sortie standard,
303
+ car il n'y a pas de suffixe par défaut pour les flux bruts.
304
+ .TP
305
+ \fB\-\-files\fP[\fB=\fP\fIfichier\fP]
306
+ Lire les noms de fichier à traiter depuis \fIfichier\fP ; si \fIfichier\fP est
307
+ omis , les noms de fichier sont lus sur l'entrée standard. Les noms de
308
+ fichier doivent se terminer avec le caractère de nouvelle ligne. Un tiret
309
+ (\fB\-\fP) est considéré comme un nom de fichier normal ; ce qui ne signifie pas
310
+ entrée standard. Si les noms de fichier sont aussi donnés comme arguments de
311
+ ligne de commande, ils sont traités avant les noms de fichier lus depuis
312
+ \fIfichier\fP.
313
+ .TP
314
+ \fB\-\-files0\fP[\fB=\fP\fIfichier\fP]
315
+ Cela est identique à \fB\-\-files\fP[\fB=\fP\fIfichier\fP] sauf que chaque nom de
316
+ fichier doit se terminer par le caractère null.
317
+ .
318
+ .SS "Format de fichier basique et options de compression"
319
+ .TP
320
+ \fB\-F\fP \fIformat\fP, \fB\-\-format=\fP\fIformat\fP
321
+ Indiquer le \fIformat\fP de fichier à compresser ou décompresser :
322
+ .RS
323
+ .TP
324
+ \fBauto\fP
325
+ C'est celui par défaut. Lors de la compression, \fBauto\fP est équivalent à
326
+ \fBxz\fP. Lors de la décompression, le format du fichier en entrée est détecté
327
+ automatiquement. Notez que les flux bruts (créés avec \fB\-\-format=raw\fP) ne
328
+ peuvent pas être détectés automatiquement.
329
+ .TP
330
+ \fBxz\fP
331
+ Compresser dans le format de fichier \fB.xz\fP ou n'accepter que les fichiers
332
+ \&\fB.xz\fP à décompresser.
333
+ .TP
334
+ \fBlzma\fP, \fBalone\fP
335
+ Compresser au format de fichier \fB.lzma\fP historique, ou n'accepter que les
336
+ fichiers \fB.lzma\fP lors de la décompression. Le nom alternatif \fBalone\fP est
337
+ fourni pour la rétrocompatibilité avec les utilitaires LZMA.
338
+ .TP
339
+ \fBlzip\fP
340
+ Accept only \fB.lz\fP files when decompressing. Compression is not supported.
341
+ .IP ""
342
+ The \fB.lz\fP format version 0 and the unextended version 1 are supported.
343
+ Version 0 files were produced by \fBlzip\fP 1.3 and older. Such files aren't
344
+ common but may be found from file archives as a few source packages were
345
+ released in this format. People might have old personal files in this
346
+ format too. Decompression support for the format version 0 was removed in
347
+ \fBlzip\fP 1.18.
348
+ .IP ""
349
+ \fBlzip\fP 1.4 and later create files in the format version 1. The sync flush
350
+ marker extension to the format version 1 was added in \fBlzip\fP 1.6. This
351
+ extension is rarely used and isn't supported by \fBxz\fP (diagnosed as corrupt
352
+ input).
353
+ .TP
354
+ \fBraw\fP
355
+ Compresser ou décompresser un flux brut (sans en\-têtes). Cela est réservé
356
+ seulement aux utilisateurs aguerris. Pour décoder des flux bruts, vous devez
357
+ utiliser \fB\-\-format=raw\fP et spécifier explicitement la chaîne de filtre, qui
358
+ normalement aurait du être stockée dans les en\-têtes du conteneur.
359
+ .RE
360
+ .TP
361
+ \fB\-C\fP \fIvérif.\fP, \fB\-\-check=\fP\fIvérif.\fP
362
+ Spécifier le type d'intégrité à vérifier. La vérification est calculée à
363
+ partir des données non\-compressées et stockées dans le fichier \fB.xz\fP. Cette
364
+ option n'a effet que si la compression a été faite dans le format \fB.xz\fP ;
365
+ le format \fB.lzma\fP ne gère pas les vérifications d'intégrité. Le contrôle
366
+ d'intégrité (s'il y en a) est vérifié lorsque le fichier \fB.xz\fP est
367
+ décompressé.
368
+ .IP ""
369
+ Types de \fIvérification\fP pris en charge :
370
+ .RS
371
+ .TP
372
+ \fBnone\fP
373
+ Ne pas calculer de vérification d'intégrité du tout. C'est généralement une
374
+ mauvaise idée. Cela peut être utile lorsque l'intégrité des données est
375
+ vérifiée de toute façon par d'autres manières.
376
+ .TP
377
+ \fBcrc32\fP
378
+ Calculer CRC32 en utilisant le polynôme de IEEE\-802.3 (Ethernet).
379
+ .TP
380
+ \fBcrc64\fP
381
+ Calculer CRC64 en utilisant le polynôme de ECMA\-182. C'est la manière
382
+ utilisée par défaut, car c'est légèrement mieux que CRC32 pour détecter les
383
+ fichiers endommagés et la différence de vitesse est négligeable.
384
+ .TP
385
+ \fBsha256\fP
386
+ Calculer SHA\-256. C'est quelque peu plus lent que CRC32 et CRC64.
387
+ .RE
388
+ .IP ""
389
+ L'intégrité des en\-têtes \fB.xz\fP est toujours vérifiée avec CRC32. Il n'est
390
+ pas possible de le changer ou de le désactiver.
391
+ .TP
392
+ \fB\-\-ignore\-check\fP
393
+ Ne pas contrôler la vérification d'intégrité des données lors de la
394
+ décompression. Les valeurs CRC32 dans les en\-têtes \fB.xz\fP seront normalement
395
+ toujours vérifiées.
396
+ .IP ""
397
+ \fBN'utilisez pas cette option à moins de savoir ce que vous faites.\fP Les
398
+ raisons possibles pour utiliser cette option :
399
+ .RS
400
+ .IP \(bu 3
401
+ Essayer de récupérer des données d'un fichier .xz corrompu.
402
+ .IP \(bu 3
403
+ Accélérer la décompression. Cela importe surtout avec SHA\-256 ou avec les
404
+ fichiers qui ont été compressés extrêmement bien. Il est recommandé de ne
405
+ pas utiliser cette option dans ce but à moins que l'intégrité du fichier ne
406
+ soit vérifiée extérieurement d'une autre manière.
407
+ .RE
408
+ .TP
409
+ \fB\-0\fP ... \fB\-9\fP
410
+ Choisir un niveau de compression prédéfini. La valeur par défaut est
411
+ \fB6\fP. Si plusieurs niveaux de préréglage sont spécifi��s, c'est le dernier
412
+ qui sera pris en compte. Si une chaîne de filtres personnalisée a déjà été
413
+ choisie, définir un niveau de compression préréglé efface la chaîne de
414
+ filtres personnalisée.
415
+ .IP ""
416
+ Les différences entre les préréglages sont plus significatives qu'avec
417
+ \fBgzip\fP(1) et \fBbzip2\fP(1). les réglages de compression sélectionnés
418
+ déterminent les exigences en mémoire pour la décompression, ainsi, utiliser
419
+ un niveau de préréglage trop élevé peut rendre difficile à décompresser un
420
+ fichier sur un vieux système avec peu de RAM. Clairement, \fBce n'est pas une bonne idée d'utiliser \-9 aveuglément pour tout\fP comme ça l'est souvent avec
421
+ \fBgzip\fP(1) et \fBbzip2\fP(1).
422
+ .RS
423
+ .TP
424
+ \fB\-0\fP ... \fB\-3\fP
425
+ Ce sont des préréglages relativement rapides. \fB0\fP est parfois plus rapide
426
+ que \fBgzip \-9\fP tout en compressant bien mieux. Les réglages plus élevés ont
427
+ souvent une rapidité comparable à celle de \fBbzip2\fP(1) avec un taux de
428
+ compression comparable ou meilleur, même si les résultats dépendent beaucoup
429
+ du genre de données compressées.
430
+ .TP
431
+ \fB\-4\fP ... \fB\-6\fP
432
+ Good to very good compression while keeping decompressor memory usage
433
+ reasonable even for old systems. \fB\-6\fP is the default, which is usually a
434
+ good choice for distributing files that need to be decompressible even on
435
+ systems with only 16\ MiB RAM. (\fB\-5e\fP or \fB\-6e\fP may be worth considering
436
+ too. See \fB\-\-extreme\fP.)
437
+ .TP
438
+ \fB\-7 ... \-9\fP
439
+ C'est comme \fB\-6\fP mais avec des besoins en mémoire plus élevés pour la
440
+ compression et la décompression. Ce n'est utile que lorsque les fichiers
441
+ sont plus gros que 8\ Mio, 16\ Mio et 32\ Mio respectivement.
442
+ .RE
443
+ .IP ""
444
+ Sur le même matériel, la vitesse de décompression est sensiblement un nombre
445
+ constant d'octets de données compressées par seconde. En d'autres termes,
446
+ meilleure est la compression, plus rapide sera en général la
447
+ décompression. Cela signifie aussi que la quantité de sortie non compressée
448
+ produite par seconde peut varier beaucoup.
449
+ .IP ""
450
+ Le tableau suivant résume les caractéristiques des préréglages :
451
+ .RS
452
+ .RS
453
+ .PP
454
+ .TS
455
+ tab(;);
456
+ c c c c c
457
+ n n n n n.
458
+ Préréglage;DictSize;CompCPU;CompMem;DecMem
459
+ \-0;256 KiB;0;3 MiB;1 MiB
460
+ \-1;1 MiB;1;9 MiB;2 MiB
461
+ \-2;2 MiB;2;17 MiB;3 MiB
462
+ \-3;4 MiB;3;32 MiB;5 MiB
463
+ \-4;4 MiB;4;48 MiB;5 MiB
464
+ \-5;8 MiB;5;94 MiB;9 MiB
465
+ \-6;8 MiB;6;94 MiB;9 MiB
466
+ \-7;16 MiB;6;186 MiB;17 MiB
467
+ \-8;32 MiB;6;370 MiB;33 MiB
468
+ \-9;64 MiB;6;674 MiB;65 MiB
469
+ .TE
470
+ .RE
471
+ .RE
472
+ .IP ""
473
+ Descriptions des colonnes :
474
+ .RS
475
+ .IP \(bu 3
476
+ DictSize est la taille du dictionnaire de LZMA2. Utiliser un dictionnaire
477
+ plus gros que la taille du fichier non compressé est un gaspillage de
478
+ mémoire. C'est pourquoi il est bon d'éviter d'utiliser les préréglages de
479
+ \fB\-7\fP à \fB\-9\fP lorsqu'il n'y en a pas vraiment besoin. A \fB\-6\fP et plus bas,
480
+ la quantité de mémoire gaspillée est généralement assez basse pour ne pas
481
+ être un problème.
482
+ .IP \(bu 3
483
+ CompCPU est une représentation des préréglages de LZMA2 qui affectent la
484
+ vitesse de compression. La taille du dictionnaire aussi affecte la vitesse,
485
+ alors comme CompCPU est le même pour les niveaux de \fB\-6\fP à \fB\-9\fP, les plus
486
+ haut niveaux tendent à être un peu moins rapides. Pour être encore moins
487
+ rapide et du coup obtenir peut être une meilleure compression, consultez
488
+ \fB\-\-extreme\fP.
489
+ .IP \(bu 3
490
+ CompMem contains the compressor memory requirements in the single\-threaded
491
+ mode. It may vary slightly between \fBxz\fP versions.
492
+ .IP \(bu 3
493
+ DecMem contient les besoins en mémoire du décompresseur. Ce sont les
494
+ réglages de la compression qui déterminent les besoins en mémoire de la
495
+ décompression. L'exacte utilisation de la mémoire est légèrement supérieure
496
+ à la taille du dictionnaire LZMA2, mais les valeurs dans la table ont été
497
+ arrondies au prochain Mio supérieur.
498
+ .RE
499
+ .IP ""
500
+ Memory requirements of the multi\-threaded mode are significantly higher than
501
+ that of the single\-threaded mode. With the default value of
502
+ \fB\-\-block\-size\fP, each thread needs 3*3*DictSize plus CompMem or DecMem. For
503
+ example, four threads with preset \fB\-6\fP needs 660\(en670\ MiB of memory.
504
+ .TP
505
+ \fB\-e\fP, \fB\-\-extreme\fP
506
+ Utilisez un variant plus lent que les préréglages (\fB\-0\fP à \fB\-9\fP) pour
507
+ espérer avoir un taux de compression légèrement meilleur, mais en cas de
508
+ malchance cela peut être pire. L'utilisation mémoire du décompresseur n'est
509
+ pas affectée, mais l'utilisation mémoire du compresseur augmente un peu aux
510
+ niveaux de préréglages de \fB\-0\fP à \fB\-3\fP.
511
+ .IP ""
512
+ Depuis qu'il y a deux préréglages avec des tailles de dictionnaire de 4\ Mio
513
+ et 8 \Mio, les préréglages \fB\-3e\fP et \fB\-5e\fP utilisent des réglages
514
+ légèrement plus rapides que \fB\-4e\fP et \fB\-6e\fP, respectivement. De cette
515
+ manière, il n'y a pas deux préréglages identiques.
516
+ .RS
517
+ .RS
518
+ .PP
519
+ .TS
520
+ tab(;);
521
+ c c c c c
522
+ n n n n n.
523
+ Préréglage;DictSize;CompCPU;CompMem;DecMem
524
+ \-0e;256 KiB;8;4 MiB;1 MiB
525
+ \-1e;1 MiB;8;13 MiB;2 MiB
526
+ \-2e;2 MiB;8;25 MiB;3 MiB
527
+ \-3e;4 MiB;7;48 MiB;5 MiB
528
+ \-4e;4 MiB;8;48 MiB;5 MiB
529
+ \-5e;8 MiB;7;94 MiB;9 MiB
530
+ \-6e;8 MiB;8;94 MiB;9 MiB
531
+ \-7e;16 MiB;8;186 MiB;17 MiB
532
+ \-8e;32 MiB;8;370 MiB;33 MiB
533
+ \-9e;64 MiB;8;674 MiB;65 MiB
534
+ .TE
535
+ .RE
536
+ .RE
537
+ .IP ""
538
+ Par exemple, il y a un total de quatre préréglages qui utilisent un
539
+ dictionnaire de 8 Mio et qui sont dans l'ordre du plus rapide au plus lent :
540
+ \fB\-5\fP, \fB\-6\fP, \fB\-5e\fP et \fB\-6e\fP.
541
+ .TP
542
+ \fB\-\-fast\fP
543
+ .PD 0
544
+ .TP
545
+ \fB\-\-best\fP
546
+ .PD
547
+ Il y a néanmoins des alias trompeurs pour \fB\-0\fP et \fB\-9\fP,
548
+ respectivement. Ils ne sont fournis que pour des besoins de
549
+ rétro\-compatibilité avec les utilitaires LZMA. Evitez d'utiliser ces
550
+ options.
551
+ .TP
552
+ \fB\-\-block\-size=\fP\fItaille\fP
553
+ Lors de la compression dans le format \fB.xz\fP, les données de l'entrée sont
554
+ réparties en blocs de \fItaille\fP octets. Les blocs sont compressés
555
+ indépendamment les un des autres, ce qui aide avec le mode multithread
556
+ (multi\-threading) et rend possible la décompression à accès aléatoire
557
+ limité. Cette option est typiquement utilisée pour outrepasser la taille de
558
+ bloc en mode multithread, mais cette option peut aussi être utilisée en mode
559
+ single\-thread.
560
+ .IP ""
561
+ In multi\-threaded mode about three times \fIsize\fP bytes will be allocated in
562
+ each thread for buffering input and output. The default \fIsize\fP is three
563
+ times the LZMA2 dictionary size or 1 MiB, whichever is more. Typically a
564
+ good value is 2\(en4 times the size of the LZMA2 dictionary or at least 1
565
+ MiB. Using \fIsize\fP less than the LZMA2 dictionary size is waste of RAM
566
+ because then the LZMA2 dictionary buffer will never get fully used. In
567
+ multi\-threaded mode, the sizes of the blocks are stored in the block
568
+ headers. This size information is required for multi\-threaded
569
+ decompression.
570
+ .IP ""
571
+ In single\-threaded mode no block splitting is done by default. Setting this
572
+ option doesn't affect memory usage. No size information is stored in block
573
+ headers, thus files created in single\-threaded mode won't be identical to
574
+ files created in multi\-threaded mode. The lack of size information also
575
+ means that \fBxz\fP won't be able decompress the files in multi\-threaded mode.
576
+ .TP
577
+ \fB\-\-block\-list=\fP\fIitems\fP
578
+ When compressing to the \fB.xz\fP format, start a new block with an optional
579
+ custom filter chain after the given intervals of uncompressed data.
580
+ .IP ""
581
+ The \fIitems\fP are a comma\-separated list. Each item consists of an optional
582
+ filter chain number between 0 and 9 followed by a colon (\fB:\fP) and a
583
+ required size of uncompressed data. Omitting an item (two or more
584
+ consecutive commas) is a shorthand to use the size and filters of the
585
+ previous item.
586
+ .IP ""
587
+ If the input file is bigger than the sum of the sizes in \fIitems\fP, the last
588
+ item is repeated until the end of the file. A special value of \fB0\fP may be
589
+ used as the last size to indicate that the rest of the file should be
590
+ encoded as a single block.
591
+ .IP ""
592
+ An alternative filter chain for each block can be specified in combination
593
+ with the \fB\-\-filters1=\fP\fIfilters\fP \&...\& \fB\-\-filters9=\fP\fIfilters\fP options.
594
+ These options define filter chains with an identifier between 1\(en9.
595
+ Filter chain 0 can be used to refer to the default filter chain, which is
596
+ the same as not specifying a filter chain. The filter chain identifier can
597
+ be used before the uncompressed size, followed by a colon (\fB:\fP). For
598
+ example, if one specifies \fB\-\-block\-list=1:2MiB,3:2MiB,2:4MiB,,2MiB,0:4MiB\fP
599
+ then blocks will be created using:
600
+ .RS
601
+ .IP \(bu 3
602
+ The filter chain specified by \fB\-\-filters1\fP and 2 MiB input
603
+ .IP \(bu 3
604
+ The filter chain specified by \fB\-\-filters3\fP and 2 MiB input
605
+ .IP \(bu 3
606
+ The filter chain specified by \fB\-\-filters2\fP and 4 MiB input
607
+ .IP \(bu 3
608
+ The filter chain specified by \fB\-\-filters2\fP and 4 MiB input
609
+ .IP \(bu 3
610
+ The default filter chain and 2 MiB input
611
+ .IP \(bu 3
612
+ The default filter chain and 4 MiB input for every block until end of input.
613
+ .RE
614
+ .IP ""
615
+ If one specifies a size that exceeds the encoder's block size (either the
616
+ default value in threaded mode or the value specified with
617
+ \fB\-\-block\-size=\fP\fIsize\fP), the encoder will create additional blocks while
618
+ keeping the boundaries specified in \fIitems\fP. For example, if one specifies
619
+ \fB\-\-block\-size=10MiB\fP \fB\-\-block\-list=5MiB,10MiB,8MiB,12MiB,24MiB\fP and the
620
+ input file is 80 MiB, one will get 11 blocks: 5, 10, 8, 10, 2, 10, 10, 4,
621
+ 10, 10, and 1 MiB.
622
+ .IP ""
623
+ En mode multi\-threadé les tailles de blocs sont stockées dans les en\-têtes
624
+ du bloc. Cela ne se fait pas en mode mono\-threadé, la sortie encodée ne sera
625
+ donc pas identique à celle faite en mode multi\-threadé.
626
+ .TP
627
+ \fB\-\-flush\-timeout=\fP\fItemps_d'attente\fP
628
+ Lors de la compression, si plus que \fItemps_d'attente\fP millisecondes (un
629
+ entier positif) se sont écoulées depuis le précédent vidage et que lire plus
630
+ de données bloquerait, toutes les données d'entrée en attente sont vidées de
631
+ l'encodeur et mises à disposition dans le flux de sortie. Cela peut être
632
+ utile si \fBxz\fP est utilisé pour compresser les données qui sont diffusées
633
+ sur un réseau. Des petites valeurs de \fItemps_d'attente\fP rendent les données
634
+ disponibles à l'extrémité réceptrice avec un léger retard, mais les grandes
635
+ valeurs de \fItemps_d'attente\fP donnent un meilleur taux de compression.
636
+ .IP ""
637
+ Cette option est désactivée par défaut. Si cette option est indiquée plus
638
+ d'une fois, la dernière prend effet. La valeur spéciale de
639
+ \fItemps_d'attente\fP de \fB0\fP peut être utilisée pour explicitement désactiver
640
+ cette option.
641
+ .IP ""
642
+ Cette option n'est pas disponible sur les systèmes qui ne sont pas POSIX.
643
+ .IP ""
644
+ .\" FIXME
645
+ \fBCette option est encore expérimentale.\fP Actuellement, \fBxz\fP ne convient
646
+ pas pour décompresser le flux en temps réel en raison de la façon dont \fBxz\fP
647
+ effectue la mise en mémoire tampon.
648
+ .TP
649
+ \fB\-\-memlimit\-compress=\fP\fIlimite\fP
650
+ Indiquer une limite d'utilisation de la mémoire pour la compression. Si
651
+ cette option est indiquée plusieurs fois, c'est la dernière qui est prise en
652
+ compte.
653
+ .IP ""
654
+ If the compression settings exceed the \fIlimit\fP, \fBxz\fP will attempt to
655
+ adjust the settings downwards so that the limit is no longer exceeded and
656
+ display a notice that automatic adjustment was done. The adjustments are
657
+ done in this order: reducing the number of threads, switching to
658
+ single\-threaded mode if even one thread in multi\-threaded mode exceeds the
659
+ \fIlimit\fP, and finally reducing the LZMA2 dictionary size.
660
+ .IP ""
661
+ When compressing with \fB\-\-format=raw\fP or if \fB\-\-no\-adjust\fP has been
662
+ specified, only the number of threads may be reduced since it can be done
663
+ without affecting the compressed output.
664
+ .IP ""
665
+ If the \fIlimit\fP cannot be met even with the adjustments described above, an
666
+ error is displayed and \fBxz\fP will exit with exit status 1.
667
+ .IP ""
668
+ La \fIlimite\fP peut être indiquée de plusieurs façons :
669
+ .RS
670
+ .IP \(bu 3
671
+ La \fIlimite\fP peut être une valeur absolue en octets. Utiliser un suffixe
672
+ d'entier comme \fBMiB\fP peut être utile. Exemple :
673
+ \fB\-\-memlimit\-compress=80MiB\fP
674
+ .IP \(bu 3
675
+ La \fIlimite\fP peut être indiquée sous forme d'un pourcentage de la mémoire
676
+ physique totale (RAM). Cela peut être particulièrement utile quand la
677
+ variable d'environnement \fBXZ_DEFAULTS\fP est indiquée dans un script
678
+ d'initialisation de l'interpréteur partagé entre différents ordinateurs. De
679
+ cette façon la limite est automatiquement plus grande sur les systèmes avec
680
+ plus de mémoire. Exemple : \fB\-\-memlimit=70%\fP
681
+ .IP \(bu 3
682
+ The \fIlimit\fP can be reset back to its default value by setting it to \fB0\fP.
683
+ This is currently equivalent to setting the \fIlimit\fP to \fBmax\fP (no memory
684
+ usage limit).
685
+ .RE
686
+ .IP ""
687
+ For 32\-bit \fBxz\fP there is a special case: if the \fIlimit\fP would be over
688
+ \fB4020\ MiB\fP, the \fIlimit\fP is set to \fB4020\ MiB\fP. On MIPS32 \fB2000\ MiB\fP
689
+ is used instead. (The values \fB0\fP and \fBmax\fP aren't affected by this. A
690
+ similar feature doesn't exist for decompression.) This can be helpful when
691
+ a 32\-bit executable has access to 4\ GiB address space (2 GiB on MIPS32)
692
+ while hopefully doing no harm in other situations.
693
+ .IP ""
694
+ Voir aussi la section \fButilisation de la mémoire\fP.
695
+ .TP
696
+ \fB\-\-memlimit\-decompress=\fP\fIlimite\fP
697
+ Régler une limite d'utilisation de la mémoire pour la décompression. Cela a
698
+ un effet sur le mode \fB\-\-list\fP. Si l'opération n'est pas possible sans
699
+ dépasser la \fIlimite\fP, \fBxz\fP affichera une erreur et la décompression
700
+ échouera. Voir \fB\-\-memlimit\-compress=\fP\fIlimite\fP pour les manières possibles
701
+ d'indiquer la \fIlimite\fP.
702
+ .TP
703
+ \fB\-\-memlimit\-mt\-decompress=\fP\fIlimit\fP
704
+ Set a memory usage limit for multi\-threaded decompression. This can only
705
+ affect the number of threads; this will never make \fBxz\fP refuse to
706
+ decompress a file. If \fIlimit\fP is too low to allow any multi\-threading, the
707
+ \fIlimit\fP is ignored and \fBxz\fP will continue in single\-threaded mode. Note
708
+ that if also \fB\-\-memlimit\-decompress\fP is used, it will always apply to both
709
+ single\-threaded and multi\-threaded modes, and so the effective \fIlimit\fP for
710
+ multi\-threading will never be higher than the limit set with
711
+ \fB\-\-memlimit\-decompress\fP.
712
+ .IP ""
713
+ In contrast to the other memory usage limit options,
714
+ \fB\-\-memlimit\-mt\-decompress=\fP\fIlimit\fP has a system\-specific default
715
+ \fIlimit\fP. \fBxz \-\-info\-memory\fP can be used to see the current value.
716
+ .IP ""
717
+ This option and its default value exist because without any limit the
718
+ threaded decompressor could end up allocating an insane amount of memory
719
+ with some input files. If the default \fIlimit\fP is too low on your system,
720
+ feel free to increase the \fIlimit\fP but never set it to a value larger than
721
+ the amount of usable RAM as with appropriate input files \fBxz\fP will attempt
722
+ to use that amount of memory even with a low number of threads. Running out
723
+ of memory or swapping will not improve decompression performance.
724
+ .IP ""
725
+ See \fB\-\-memlimit\-compress=\fP\fIlimit\fP for possible ways to specify the
726
+ \fIlimit\fP. Setting \fIlimit\fP to \fB0\fP resets the \fIlimit\fP to the default
727
+ system\-specific value.
728
+ .TP
729
+ \fB\-M\fP \fIlimite\fP, \fB\-\-memlimit=\fP\fIlimite\fP, \fB\-\-memory=\fP\fIlimite\fP
730
+ This is equivalent to specifying \fB\-\-memlimit\-compress=\fP\fIlimit\fP
731
+ \fB\-\-memlimit\-decompress=\fP\fIlimit\fP \fB\-\-memlimit\-mt\-decompress=\fP\fIlimit\fP.
732
+ .TP
733
+ \fB\-\-no\-adjust\fP
734
+ Display an error and exit if the memory usage limit cannot be met without
735
+ adjusting settings that affect the compressed output. That is, this
736
+ prevents \fBxz\fP from switching the encoder from multi\-threaded mode to
737
+ single\-threaded mode and from reducing the LZMA2 dictionary size. Even when
738
+ this option is used the number of threads may be reduced to meet the memory
739
+ usage limit as that won't affect the compressed output.
740
+ .IP ""
741
+ Automatic adjusting is always disabled when creating raw streams
742
+ (\fB\-\-format=raw\fP).
743
+ .TP
744
+ \fB\-T\fP \fIthreads\fP, \fB\-\-threads=\fP\fIthreads\fP
745
+ Specify the number of worker threads to use. Setting \fIthreads\fP to a
746
+ special value \fB0\fP makes \fBxz\fP use up to as many threads as the processor(s)
747
+ on the system support. The actual number of threads can be fewer than
748
+ \fIthreads\fP if the input file is not big enough for threading with the given
749
+ settings or if using more threads would exceed the memory usage limit.
750
+ .IP ""
751
+ The single\-threaded and multi\-threaded compressors produce different
752
+ output. Single\-threaded compressor will give the smallest file size but
753
+ only the output from the multi\-threaded compressor can be decompressed using
754
+ multiple threads. Setting \fIthreads\fP to \fB1\fP will use the single\-threaded
755
+ mode. Setting \fIthreads\fP to any other value, including \fB0\fP, will use the
756
+ multi\-threaded compressor even if the system supports only one hardware
757
+ thread. (\fBxz\fP 5.2.x used single\-threaded mode in this situation.)
758
+ .IP ""
759
+ To use multi\-threaded mode with only one thread, set \fIthreads\fP to \fB+1\fP.
760
+ The \fB+\fP prefix has no effect with values other than \fB1\fP. A memory usage
761
+ limit can still make \fBxz\fP switch to single\-threaded mode unless
762
+ \fB\-\-no\-adjust\fP is used. Support for the \fB+\fP prefix was added in \fBxz\fP
763
+ 5.4.0.
764
+ .IP ""
765
+ If an automatic number of threads has been requested and no memory usage
766
+ limit has been specified, then a system\-specific default soft limit will be
767
+ used to possibly limit the number of threads. It is a soft limit in sense
768
+ that it is ignored if the number of threads becomes one, thus a soft limit
769
+ will never stop \fBxz\fP from compressing or decompressing. This default soft
770
+ limit will not make \fBxz\fP switch from multi\-threaded mode to single\-threaded
771
+ mode. The active limits can be seen with \fBxz \-\-info\-memory\fP.
772
+ .IP ""
773
+ Actuellement, la seule méthode de gestion avec des threads consiste à
774
+ séparer l'entrée en blocs et de les compresser indépendamment les uns des
775
+ autres. La taille par défaut des blocs dépend du niveau de compression et
776
+ peut\-être remplacée avec l'option \fB\-\-block\-size=\fP\fItaille\fP.
777
+ .IP ""
778
+ Threaded decompression only works on files that contain multiple blocks with
779
+ size information in block headers. All large enough files compressed in
780
+ multi\-threaded mode meet this condition, but files compressed in
781
+ single\-threaded mode don't even if \fB\-\-block\-size=\fP\fIsize\fP has been used.
782
+ .IP ""
783
+ The default value for \fIthreads\fP is \fB0\fP. In \fBxz\fP 5.4.x and older the
784
+ default is \fB1\fP.
785
+ .
786
+ .SS "Chaînes de filtres de compresseur personnalisées"
787
+ A custom filter chain allows specifying the compression settings in detail
788
+ instead of relying on the settings associated to the presets. When a custom
789
+ filter chain is specified, preset options (\fB\-0\fP \&...\& \fB\-9\fP and
790
+ \fB\-\-extreme\fP) earlier on the command line are forgotten. If a preset
791
+ option is specified after one or more custom filter chain options, the new
792
+ preset takes effect and the custom filter chain options specified earlier
793
+ are forgotten.
794
+ .PP
795
+ Une chaîne de filtre est comparable à une redirection (pipe) sur la ligne de
796
+ commande. Lors de la compression, les entrées non compressées vont au
797
+ premier filtre, dont la sortie va au prochain filtre (s'il y en a). La
798
+ sortie du dernier filtre est écrite sur le fichier compressé. Le nombre
799
+ maximal de filtres dans la chaîne est quatre, mais habituellement, un chaîne
800
+ de filtre n'a qu'un ou deux filtres.
801
+ .PP
802
+ Beaucoup de filtres ont des limitations sur l'endroit où ils peuvent se
803
+ placer dans la chaîne de filtre : quelques filtres ne peuvent fonctionner
804
+ qu'en tant que dernier filtre dans la chaîne, quelques uns en tant que non
805
+ dernier filtre, et d'autres à n'importe quelle position dans la
806
+ chaîne. Suivant le filtre, cette limitation est soit inhérente au profil du
807
+ filtre, soit existe pour des raisons de sécurité.
808
+ .PP
809
+ A custom filter chain can be specified in two different ways. The options
810
+ \fB\-\-filters=\fP\fIfilters\fP and \fB\-\-filters1=\fP\fIfilters\fP \&...\&
811
+ \fB\-\-filters9=\fP\fIfilters\fP allow specifying an entire filter chain in one
812
+ option using the liblzma filter string syntax. Alternatively, a filter
813
+ chain can be specified by using one or more individual filter options in the
814
+ order they are wanted in the filter chain. That is, the order of the
815
+ individual filter options is significant! When decoding raw streams
816
+ (\fB\-\-format=raw\fP), the filter chain must be specified in the same order as
817
+ it was specified when compressing. Any individual filter or preset options
818
+ specified before the full chain option (\fB\-\-filters=\fP\fIfilters\fP) will be
819
+ forgotten. Individual filters specified after the full chain option will
820
+ reset the filter chain.
821
+ .PP
822
+ Both the full and individual filter options take filter\-specific \fIoptions\fP
823
+ as a comma\-separated list. Extra commas in \fIoptions\fP are ignored. Every
824
+ option has a default value, so specify those you want to change.
825
+ .PP
826
+ Pour voir l'entièreté de la chaîne de filtres et ses \fIoptions\fP, utilisez
827
+ \fBxz \-vv\fP (ce qui est comme utiliser \fB\-\-verbose\fP deux fois). Cela
828
+ fonctionne aussi pour voir les options de chaîne de filtres utilisées par
829
+ les préréglages.
830
+ .TP
831
+ \fB\-\-filters=\fP\fIfilters\fP
832
+ Specify the full filter chain or a preset in a single option. Each filter
833
+ can be separated by spaces or two dashes (\fB\-\-\fP). \fIfilters\fP may need to be
834
+ quoted on the shell command line so it is parsed as a single option. To
835
+ denote \fIoptions\fP, use \fB:\fP or \fB=\fP. A preset can be prefixed with a \fB\-\fP
836
+ and followed with zero or more flags. The only supported flag is \fBe\fP to
837
+ apply the same options as \fB\-\-extreme\fP.
838
+ .TP
839
+ \fB\-\-filters1\fP=\fIfilters\fP ... \fB\-\-filters9\fP=\fIfilters\fP
840
+ Specify up to nine additional filter chains that can be used with
841
+ \fB\-\-block\-list\fP.
842
+ .IP ""
843
+ For example, when compressing an archive with executable files followed by
844
+ text files, the executable part could use a filter chain with a BCJ filter
845
+ and the text part only the LZMA2 filter.
846
+ .TP
847
+ \fB\-\-filters\-help\fP
848
+ Display a help message describing how to specify presets and custom filter
849
+ chains in the \fB\-\-filters\fP and \fB\-\-filters1=\fP\fIfilters\fP \&...\&
850
+ \fB\-\-filters9=\fP\fIfilters\fP options, and exit successfully.
851
+ .TP
852
+ \fB\-\-lzma1\fP[\fB=\fP\fIoptions\fP]
853
+ .PD 0
854
+ .TP
855
+ \fB\-\-lzma2\fP[\fB=\fP\fIoptions\fP]
856
+ .PD
857
+ Ajouter le filtre LZMA1 ou LZMA2 à la chaîne de filtres. Ces filtres ne
858
+ peuvent être utilisés que comme dernier filtre dans la chaîne.
859
+ .IP ""
860
+ LZMA1 est un filtre historique, qui n'est pris en charge presque uniquement
861
+ à cause de l'ancien format de fichier \fB.lzma\fP, qui ne prend en charge que
862
+ LZMA1. LZMA2 est une version mise à jour de LZMA1 pour régler certains
863
+ problèmes pratiques de LZMA1. Le format \fBxz\fP utilise LZMA2 et ne prend pas
864
+ du tout en charge LZMA1. Les taux et vitesses de compression de LZMA1 et
865
+ LZMA2 sont pratiquement identiques.
866
+ .IP ""
867
+ LZMA1 et LZMA2 partagent le même ensemble d'\fIoptions\fP :
868
+ .RS
869
+ .TP
870
+ \fBpreset=\fP\fIpréréglage\fP
871
+ Reset all LZMA1 or LZMA2 \fIoptions\fP to \fIpreset\fP. \fIPreset\fP consist of an
872
+ integer, which may be followed by single\-letter preset modifiers. The
873
+ integer can be from \fB0\fP to \fB9\fP, matching the command line options \fB\-0\fP
874
+ \&...\& \fB\-9\fP. The only supported modifier is currently \fBe\fP, which matches
875
+ \fB\-\-extreme\fP. If no \fBpreset\fP is specified, the default values of LZMA1 or
876
+ LZMA2 \fIoptions\fP are taken from the preset \fB6\fP.
877
+ .TP
878
+ \fBdict=\fP\fItaille\fP
879
+ La \fItaille\fP du dictionnaire (historique du tampon) indique combien d'octets
880
+ des données récement décompressées sont gardés en mémoire. L'algorithme
881
+ essaie de trouver les séquences d'octets répétées (identiques) dans les
882
+ données décompressées et les remplace par les données actuellement dans le
883
+ dictionnaire. Plus gros est le dictionnaire, plus grande est la chance de
884
+ trouver une correspondance. Ainsi, l'augmentation de la \fItaille\fP du
885
+ dictionnaire augmente habituellement le taux de compression, mais un
886
+ dictionnaire plus gros que le fichier non compressé est un gachis de
887
+ mémoire.
888
+ .IP ""
889
+ Généralement la \fItaille\fP du dictionnaire est entre 64\ Kio et 64\ Mio. Le
890
+ minimum étant 4\ Kio. La \fItaille\fP maximale pour la compression est
891
+ habituellement 1,5\ Gio (1536\ Mio). Le décompresseur prend en charge les
892
+ dictionnaires jusqu'à un octet de moins que 4\ Gio, ce qui est le maximum
893
+ pour les formats de flux LZMA1 et LZMA2.
894
+ .IP ""
895
+ La \fItaille\fP du dictionnaire et le chercheur de correspondance (match
896
+ finder) (\fImf\fP) déterminent ensemble l'utilisation de la mémoire de
897
+ l'encodeur LZMA1 ou LZMA2. La même (ou une plus grosse) \fItaille\fP de
898
+ dictionnaire est requise pour décompresser que ce qui a été utilisé pour la
899
+ compression, ainsi l'utilisation de la mémoire du décodeur est déterminée
900
+ par la taille du dictionnaire utilisée lors de la compression. Les en\-têtes
901
+ de \fB.xz\fP stockent la \fItaille\fP de dictionnaire sous la forme 2^\fIn\fP ou
902
+ 2^\fIn\fP + 2^(\fIn\fP\-1), de sorte que ces \fItailles\fP sont quelque peu préférées
903
+ pour la compression. Les autres \fItailles\fP seront arrondies à la hausse
904
+ lorsque stockées dans les en\-têtes de \fB.xz\fP.
905
+ .TP
906
+ \fBlc=\fP\fIlc\fP
907
+ Spécifiez le nombre d'octets de contexte littéraux. Le minimum est \fB0\fP et
908
+ le maximum est \fB4\fP. La valeur par défaut est \fB3\fP. En plus, la somme de
909
+ \fIlc\fP et \fIlp\fP ne doit pas excéder \fB4\fP.
910
+ .IP ""
911
+ Tous les octets qui ne peuvent pas être codés comme des correspondances sont
912
+ codés comme des littéraux. C'est à dire que les littéraux sont simplement
913
+ des octets 8 bits encodés un à la fois.
914
+ .IP ""
915
+ The literal coding makes an assumption that the highest \fIlc\fP bits of the
916
+ previous uncompressed byte correlate with the next byte. For example, in
917
+ typical English text, an upper\-case letter is often followed by a lower\-case
918
+ letter, and a lower\-case letter is usually followed by another lower\-case
919
+ letter. In the US\-ASCII character set, the highest three bits are 010 for
920
+ upper\-case letters and 011 for lower\-case letters. When \fIlc\fP is at least
921
+ 3, the literal coding can take advantage of this property in the
922
+ uncompressed data.
923
+ .IP ""
924
+ The default value (3) is usually good. If you want maximum compression,
925
+ test \fBlc=4\fP. Sometimes it helps a little, and sometimes it makes
926
+ compression worse. If it makes it worse, test \fBlc=2\fP too.
927
+ .TP
928
+ \fBlp=\fP\fIlp\fP
929
+ Indiquer le nombre de bits de position littérale. Le minimum est \fB0\fP et le
930
+ maximum \fB4\fP; par défaut c'est \fB0\fP.
931
+ .IP ""
932
+ \fILp\fP affecte le type d'alignement dans les données décompressées qui est
933
+ présumé lors de l'encodage des littéraux. Voir \fIpb\fP ci dessous pour plus
934
+ d'information sur l'alignement.
935
+ .TP
936
+ \fBpb=\fP\fIpb\fP
937
+ Indiquer le nombre de bits de position. Le minimum est \fB0\fP et le maximum
938
+ \fB4\fP; par défaut \fB2\fP.
939
+ .IP ""
940
+ \fIPb\fP affecte quel genre d'alignement est présumé en général dans les
941
+ données non compressées. Par défaut c'est un alignement de quatre octets
942
+ (2^\fIpb\fP=2^2=4), ce qui est généralement un bon choix lorsqu'il n'y a pas de
943
+ meilleure estimation.
944
+ .IP ""
945
+ When the alignment is known, setting \fIpb\fP accordingly may reduce the file
946
+ size a little. For example, with text files having one\-byte alignment
947
+ (US\-ASCII, ISO\-8859\-*, UTF\-8), setting \fBpb=0\fP can improve compression
948
+ slightly. For UTF\-16 text, \fBpb=1\fP is a good choice. If the alignment is
949
+ an odd number like 3 bytes, \fBpb=0\fP might be the best choice.
950
+ .IP ""
951
+ Même si l'alignement présumé peut être ajusté avec \fIpb\fP et \fIlp\fP, LZMA1 et
952
+ LZMA2 favorisent toujours légèrement l'alignement sur 16 octets. Il peut
953
+ être utile d'en tenir compte lors de la conception de formats de fichiers
954
+ susceptibles d'être souvent compressés avec LZMA1 ou LZMA2.
955
+ .TP
956
+ \fBmf=\fP\fImf\fP
957
+ Match finder has a major effect on encoder speed, memory usage, and
958
+ compression ratio. Usually Hash Chain match finders are faster than Binary
959
+ Tree match finders. The default depends on the \fIpreset\fP: 0 uses \fBhc3\fP,
960
+ 1\(en3 use \fBhc4\fP, and the rest use \fBbt4\fP.
961
+ .IP ""
962
+ Les chercheurs de correspondance suivants sont pris en charge. Les formules
963
+ d'utilisation de la mémoire ci\-dessous sont des approximations grossières
964
+ qui sont les plus proches de la réalité lorsque \fIdict\fP est une puissance de
965
+ deux.
966
+ .RS
967
+ .TP
968
+ \fBhc3\fP
969
+ Chaîne de hachage avec hachage de 2 et 3 octets
970
+ .br
971
+ Valeur minimale pour \fInice\fP : \fB3\fP
972
+ .br
973
+ Utilisation de la mémoire :
974
+ .br
975
+ \fIdict\fP * 7.5 (if \fIdict\fP <= 16 Mio);
976
+ .br
977
+ \fIdict\fP * 5.5 + 64 MiB (si \fIdict\fP > 16 Mio)
978
+ .TP
979
+ \fBhc4\fP
980
+ Chaîne de hachage avec hachage de 2, 3 et 4 octets
981
+ .br
982
+ Valeur minimale pour \fInice\fP : \fB4\fP
983
+ .br
984
+ Utilisation de la mémoire :
985
+ .br
986
+ \fIdict\fP * 7.5 (si \fIdict\fP <= 32 Mio);
987
+ .br
988
+ \fIdict\fP * 6.5 (si \fIdict\fP > 32 Mio)
989
+ .TP
990
+ \fBbt2\fP
991
+ Arbre binaire avec hachage de 2 octets
992
+ .br
993
+ Valeur minimale pour \fInice\fP : \fB2\fP
994
+ .br
995
+ Utilisation de la mémoire : \fIdict\fP * 9.5
996
+ .TP
997
+ \fBbt3\fP
998
+ Arbre binaire avec hachage de 2 et 3 octets
999
+ .br
1000
+ Valeur minimale pour \fInice\fP : \fB3\fP
1001
+ .br
1002
+ Utilisation de la mémoire :
1003
+ .br
1004
+ \fIdict\fP * 11.5 (si \fIdict\fP <= 16 Mio);
1005
+ .br
1006
+ \fIdict\fP * 9.5 + 64 MiB (si \fIdict\fP > 16 Mio)
1007
+ .TP
1008
+ \fBbt4\fP
1009
+ Arbre binaire avec hachage 2, 3 et 4 octets
1010
+ .br
1011
+ Valeur minimale pour \fInice\fP : \fB4\fP
1012
+ .br
1013
+ Utilisation de la mémoire :
1014
+ .br
1015
+ \fIdict\fP * 11.5 (si \fIdict\fP <= 32 Mio);
1016
+ .br
1017
+ \fIdict\fP * 10.5 (si \fIdict\fP > 32 Mio)
1018
+ .RE
1019
+ .TP
1020
+ \fBmode=\fP\fImode\fP
1021
+ Compression \fImode\fP specifies the method to analyze the data produced by the
1022
+ match finder. Supported \fImodes\fP are \fBfast\fP and \fBnormal\fP. The default is
1023
+ \fBfast\fP for \fIpresets\fP 0\(en3 and \fBnormal\fP for \fIpresets\fP 4\(en9.
1024
+ .IP ""
1025
+ Habituellement, \fBfast\fP est utilisé avec les chercheurs de correspondance de
1026
+ chaîne de hachage et \fBnormal\fP avec les chercheurs de correspondance d'arbre
1027
+ binaire. C'est aussi ce que font les \fIpréréglages\fP.
1028
+ .TP
1029
+ \fBnice=\fP\fInice\fP
1030
+ Spécifier ce qui est considéré comme une bonne longueur pour une
1031
+ correspondance. Une fois que la correspondance d'au moins \fInice\fP octets est
1032
+ trouvée, l'algorithme arrête de chercher de meilleures correspondances
1033
+ possibles.
1034
+ .IP ""
1035
+ \fINice\fP can be 2\(en273 bytes. Higher values tend to give better
1036
+ compression ratio at the expense of speed. The default depends on the
1037
+ \fIpreset\fP.
1038
+ .TP
1039
+ \fBdepth=\fP\fIprofondeur\fP
1040
+ Spécifier la profondeur de recherche maximale dans l'outil de recherche de
1041
+ correspondances. La valeur par défaut est \fB0\fP, ce qui fait que le
1042
+ compresseur détermine une \fIprofondeur\fP raisonnable en fonction de \fImf\fP et
1043
+ \fInice\fP.
1044
+ .IP ""
1045
+ Reasonable \fIdepth\fP for Hash Chains is 4\(en100 and 16\(en1000 for Binary
1046
+ Trees. Using very high values for \fIdepth\fP can make the encoder extremely
1047
+ slow with some files. Avoid setting the \fIdepth\fP over 1000 unless you are
1048
+ prepared to interrupt the compression in case it is taking far too long.
1049
+ .RE
1050
+ .IP ""
1051
+ Lors du décodage des flux bruts (\fB\-\-format=raw\fP), LZMA2 nécessite seulement
1052
+ la \fItaille\fP du dictionnaire. LZMA1 nécessite aussi \fIlc\fP, \fIlp\fP et \fIpb\fP.
1053
+ .TP
1054
+ \fB\-\-x86\fP[\fB=\fP\fIoptions\fP]
1055
+ .PD 0
1056
+ .TP
1057
+ \fB\-\-arm\fP[\fB=\fP\fIoptions\fP]
1058
+ .TP
1059
+ \fB\-\-armthumb\fP[\fB=\fP\fIoptions\fP]
1060
+ .TP
1061
+ \fB\-\-arm64\fP[\fB=\fP\fIoptions\fP]
1062
+ .TP
1063
+ \fB\-\-powerpc\fP[\fB=\fP\fIoptions\fP]
1064
+ .TP
1065
+ \fB\-\-ia64\fP[\fB=\fP\fIoptions\fP]
1066
+ .TP
1067
+ \fB\-\-sparc\fP[\fB=\fP\fIoptions\fP]
1068
+ .TP
1069
+ \fB\-\-riscv\fP[\fB=\fP\fIoptions\fP]
1070
+ .PD
1071
+ Ajouter un filtre branch/call/jump (BCJ) à la chaîne de filtres. Ces filtres
1072
+ ne peuvent être utilisés que s'ils ne sont pas le dernier filtre de la
1073
+ chaîne de filtrage.
1074
+ .IP ""
1075
+ A BCJ filter converts relative addresses in the machine code to their
1076
+ absolute counterparts. This doesn't change the size of the data but it
1077
+ increases redundancy, which can help LZMA2 to produce 0\(en15\ % smaller
1078
+ \&\fB.xz\fP file. The BCJ filters are always reversible, so using a BCJ filter
1079
+ for wrong type of data doesn't cause any data loss, although it may make the
1080
+ compression ratio slightly worse. The BCJ filters are very fast and use an
1081
+ insignificant amount of memory.
1082
+ .IP ""
1083
+ Ces filtres BCJ présentent des problèmes connus liés au taux de
1084
+ compression :
1085
+ .RS
1086
+ .IP \(bu 3
1087
+ Some types of files containing executable code (for example, object files,
1088
+ static libraries, and Linux kernel modules) have the addresses in the
1089
+ instructions filled with filler values. These BCJ filters will still do the
1090
+ address conversion, which will make the compression worse with these files.
1091
+ .IP \(bu 3
1092
+ If a BCJ filter is applied on an archive, it is possible that it makes the
1093
+ compression ratio worse than not using a BCJ filter. For example, if there
1094
+ are similar or even identical executables then filtering will likely make
1095
+ the files less similar and thus compression is worse. The contents of
1096
+ non\-executable files in the same archive can matter too. In practice one
1097
+ has to try with and without a BCJ filter to see which is better in each
1098
+ situation.
1099
+ .RE
1100
+ .IP ""
1101
+ Different instruction sets have different alignment: the executable file
1102
+ must be aligned to a multiple of this value in the input data to make the
1103
+ filter work.
1104
+ .RS
1105
+ .RS
1106
+ .PP
1107
+ .TS
1108
+ tab(;);
1109
+ l n l
1110
+ l n l.
1111
+ Filtre;Alignement;Notes
1112
+ x86;1;32 bits ou 64 bits x86
1113
+ ARM;4;
1114
+ ARM\-Thumb;2;
1115
+ ARM64;4;4096\-byte alignment is best
1116
+ PowerPC;4;Grand boutiste seulement
1117
+ IA\-64;16;Itanium
1118
+ SPARC;4;
1119
+ RISC\-V;2;
1120
+ .TE
1121
+ .RE
1122
+ .RE
1123
+ .IP ""
1124
+ Since the BCJ\-filtered data is usually compressed with LZMA2, the
1125
+ compression ratio may be improved slightly if the LZMA2 options are set to
1126
+ match the alignment of the selected BCJ filter. Examples:
1127
+ .RS
1128
+ .IP \(bu 3
1129
+ IA\-64 filter has 16\-byte alignment so \fBpb=4,lp=4,lc=0\fP is good with LZMA2
1130
+ (2^4=16).
1131
+ .IP \(bu 3
1132
+ RISC\-V code has 2\-byte or 4\-byte alignment depending on whether the file
1133
+ contains 16\-bit compressed instructions (the C extension). When 16\-bit
1134
+ instructions are used, \fBpb=2,lp=1,lc=3\fP or \fBpb=1,lp=1,lc=3\fP is good. When
1135
+ 16\-bit instructions aren't present, \fBpb=2,lp=2,lc=2\fP is the best.
1136
+ \fBreadelf \-h\fP can be used to check if "RVC" appears on the "Flags" line.
1137
+ .IP \(bu 3
1138
+ ARM64 is always 4\-byte aligned so \fBpb=2,lp=2,lc=2\fP is the best.
1139
+ .IP \(bu 3
1140
+ The x86 filter is an exception. It's usually good to stick to LZMA2's
1141
+ defaults (\fBpb=2,lp=0,lc=3\fP) when compressing x86 executables.
1142
+ .RE
1143
+ .IP ""
1144
+ Tous les filtres BCJ prennent en charge les mêmes \fIoptions\fP :
1145
+ .RS
1146
+ .TP
1147
+ \fBstart=\fP\fIdécalage\fP
1148
+ Spécifier le \fIdécalage\fP de départ qui est utilisé lors de la conversion
1149
+ entre les adresses relatives et absolues. Le \fIdécalage\fP doit être un
1150
+ multiple de l'alignement du filtre (voir la table ci\-dessus). Sa valeur par
1151
+ défaut est zéro. En pratique, cette dernière convient ; indiquer un
1152
+ \fIdécalage\fP personnalisé est la plupart du temps inutile.
1153
+ .RE
1154
+ .TP
1155
+ \fB\-\-delta\fP[\fB=\fP\fIoptions\fP]
1156
+ Ajouter le filtre Delta à la chaîne de filtres. Le filtre Delta ne peut être
1157
+ utilisé que s'il n'est pas le dernier filtre dans la chaîne.
1158
+ .IP ""
1159
+ Currently only simple byte\-wise delta calculation is supported. It can be
1160
+ useful when compressing, for example, uncompressed bitmap images or
1161
+ uncompressed PCM audio. However, special purpose algorithms may give
1162
+ significantly better results than Delta + LZMA2. This is true especially
1163
+ with audio, which compresses faster and better, for example, with
1164
+ \fBflac\fP(1).
1165
+ .IP ""
1166
+ \fIoptions\fP prises en charge :
1167
+ .RS
1168
+ .TP
1169
+ \fBdist=\fP\fIdistance\fP
1170
+ Specify the \fIdistance\fP of the delta calculation in bytes. \fIdistance\fP must
1171
+ be 1\(en256. The default is 1.
1172
+ .IP ""
1173
+ Par exemple, avec \fBdist=2\fP et une entrée huit octets A1 B1 A2 B3 A3 B5 A4
1174
+ B7, la sortie sera A1 B1 01 02 01 02 01 02.
1175
+ .RE
1176
+ .
1177
+ .SS "Autres options"
1178
+ .TP
1179
+ \fB\-q\fP, \fB\-\-quiet\fP
1180
+ Supprimer les avertissements et les notifications. Indiquer cela deux fois
1181
+ supprimera aussi les erreurs. Cette option n'a aucun effet sur le statut de
1182
+ sortie. Cela dit, même si un avertissement était supprimé, le statut de
1183
+ sortie indiquant un avertissement sera encore utilisé.
1184
+ .TP
1185
+ \fB\-v\fP, \fB\-\-verbose\fP
1186
+ Être bavard. Si l'erreur standard est connectée à un terminal, \fBxz\fP
1187
+ affichera une barre de progression. Indiquer \fB\-\-verbose\fP deux fois donnera
1188
+ une sortie encore plus bavarde.
1189
+ .IP ""
1190
+ La barre de progression montre l'information suivante :
1191
+ .RS
1192
+ .IP \(bu 3
1193
+ Le pourcentage de complétion est montré si la taille du fichier en entrée
1194
+ est connue. Néanmoins, le pourcentage ne peut pas être montré en cas de
1195
+ redirection.
1196
+ .IP \(bu 3
1197
+ Quantité de données compressées produites (compression) ou consommées
1198
+ (décompression).
1199
+ .IP \(bu 3
1200
+ Quantité de données non compressées consommées (compression) ou produites
1201
+ (décompression).
1202
+ .IP \(bu 3
1203
+ Le taux de compression, calculé en divisant la quantité de données
1204
+ compréssées déjà traitées par la quantité de données décompressées déjà
1205
+ traitées.
1206
+ .IP \(bu 3
1207
+ Vitesse de compression ou de décompression. Elle correspond à la quantité de
1208
+ données non compressées consommées (compression) ou produites
1209
+ (décompression) par seconde. Elle apparait quelques secondes après le début
1210
+ du traitement du fichier par \fBxz\fP.
1211
+ .IP \(bu 3
1212
+ Temps écoulé dans le format M:SS ou H:MM:SS.
1213
+ .IP \(bu 3
1214
+ Estimated remaining time is shown only when the size of the input file is
1215
+ known and a couple of seconds have already passed since \fBxz\fP started
1216
+ processing the file. The time is shown in a less precise format which never
1217
+ has any colons, for example, 2 min 30 s.
1218
+ .RE
1219
+ .IP ""
1220
+ When standard error is not a terminal, \fB\-\-verbose\fP will make \fBxz\fP print
1221
+ the filename, compressed size, uncompressed size, compression ratio, and
1222
+ possibly also the speed and elapsed time on a single line to standard error
1223
+ after compressing or decompressing the file. The speed and elapsed time are
1224
+ included only when the operation took at least a few seconds. If the
1225
+ operation didn't finish, for example, due to user interruption, also the
1226
+ completion percentage is printed if the size of the input file is known.
1227
+ .TP
1228
+ \fB\-Q\fP, \fB\-\-no\-warn\fP
1229
+ Ne pas mettre l'état de sortie à \fB2\fP même si une condition méritant un
1230
+ avertissement a été détectée. Cette option n'affecte pas le niveau de
1231
+ verbosité, néanmoins, les deux options \fB\-\-quiet\fP et \fB\-\-no\-warn\fP doivent
1232
+ être utilisées pour ne pas afficher d'avertissements, ni altérer le statut
1233
+ de sortie.
1234
+ .TP
1235
+ \fB\-\-robot\fP
1236
+ Afficher les messages dans un format analysable par une machine. Ceci est
1237
+ destiné à faciliter l'écriture des frontaux qui voudraient utiliser \fBxz\fP
1238
+ plutôt que liblzma, ce qui pourrait être le cas pour différents scripts. La
1239
+ sortie avec cette option activée est destinée à rester stable sur les
1240
+ différentes versions de \fBxz\fP. Consulter le paragraphe \fBROBOT MODE\fP pour
1241
+ les détails.
1242
+ .TP
1243
+ \fB\-\-info\-memory\fP
1244
+ Display, in human\-readable format, how much physical memory (RAM) and how
1245
+ many processor threads \fBxz\fP thinks the system has and the memory usage
1246
+ limits for compression and decompression, and exit successfully.
1247
+ .TP
1248
+ \fB\-h\fP, \fB\-\-help\fP
1249
+ Afficher un message d'aide décrivant les options les plus couramment
1250
+ utilisées et quitter.
1251
+ .TP
1252
+ \fB\-H\fP, \fB\-\-long\-help\fP
1253
+ Afficher un message d'aide décrivant toutes les options de \fBxz\fP et quitter.
1254
+ .TP
1255
+ \fB\-V\fP, \fB\-\-version\fP
1256
+ Afficher le numéro de version de \fBxz\fP et de liblzma dans un format lisible
1257
+ par un humain. Pour obtenir une sortie analysable par la machine, spécifiez
1258
+ \fB\-\-robot\fP avant \fB\-\-version\fP.
1259
+ .
1260
+ .SH "MODE ROBOT"
1261
+ The robot mode is activated with the \fB\-\-robot\fP option. It makes the output
1262
+ of \fBxz\fP easier to parse by other programs. Currently \fB\-\-robot\fP is
1263
+ supported only together with \fB\-\-list\fP, \fB\-\-filters\-help\fP, \fB\-\-info\-memory\fP,
1264
+ and \fB\-\-version\fP. It will be supported for compression and decompression in
1265
+ the future.
1266
+ .
1267
+ .SS "Mode liste"
1268
+ \fBxz \-\-robot \-\-list\fP utilise une sortie séparée par des tabulations. La
1269
+ première colonne de toutes les lignes possède une chaîne qui indique le type
1270
+ d'information trouvée sur cette ligne :
1271
+ .TP
1272
+ \fBname\fP
1273
+ C'est toujours la première ligne au début de la liste d'un fichier. La
1274
+ seconde colonne de la ligne est le nom de fichier.
1275
+ .TP
1276
+ \fBfile\fP
1277
+ Cette ligne contient l'information globale sur le fichier \fB.xz\fP. Cette
1278
+ ligne est toujours écrite après la ligne \fBname\fP.
1279
+ .TP
1280
+ \fBstream\fP
1281
+ Ce type de ligne n'est utilisée que lorsque \fB \-\-verbose\fP a été indiquée. Il
1282
+ y a autant de lignes \fBstream\fP qu'il y a de flux dans le fichier \fB.xz\fP.
1283
+ .TP
1284
+ \fBblock\fP
1285
+ Ce type de ligne n'est utilisé seulement lorsque \fB\-\-verbose\fP a été
1286
+ indiquée. Il y a autant de lignes \fBblock\fP qu'il y a de blocs dans le
1287
+ fichier \fB.xz\fP. Les lignes \fBblock\fP sont affichées après toutes les lignes
1288
+ \fBstream\fP ; les différents types de lignes ne sont pas imbriqués.
1289
+ .TP
1290
+ \fBsummary\fP
1291
+ Ce type de ligne n'est utilisé que lorsque \fB\-\-verbose\fP a été indiqué deux
1292
+ fois. Cette ligne est affichée après toutes les lignes \fBblock\fP. Comme la
1293
+ ligne \fBfile\fP, la ligne \fBsummary\fP contient l'information globale sur le
1294
+ fichier \fB.xz\fP.
1295
+ .TP
1296
+ \fBtotals\fP
1297
+ Cette ligne est toujours la toute dernière ligne de la sortie. Elle affiche
1298
+ les comptes et les tailles totaux.
1299
+ .PP
1300
+ Les colonnes des lignes \fBfile\fP :
1301
+ .PD 0
1302
+ .RS
1303
+ .IP 2. 4
1304
+ Nombre de flux dans le fichier
1305
+ .IP 3. 4
1306
+ Nombre total de blocs dans le ou les flux.
1307
+ .IP 4. 4
1308
+ Taille compressée du fichier
1309
+ .IP 5. 4
1310
+ Taille décompressée du fichier
1311
+ .IP 6. 4
1312
+ Compression ratio, for example, \fB0.123\fP. If ratio is over 9.999, three
1313
+ dashes (\fB\-\-\-\fP) are displayed instead of the ratio.
1314
+ .IP 7. 4
1315
+ Liste de noms de contrôles d'intégrité séparés par des virgules. Les chaînes
1316
+ suivantes sont utilisées pour les types de vérification connus : \fBNone\fP,
1317
+ \fBCRC32\fP, \fBCRC64\fP et \fBSHA256\fP. Pour le types de vérification inconnus,
1318
+ \fBUnknown\-\fP\fIN\fP est utilisé, où \fIN\fP est un identifiant de vérification sous
1319
+ la forme d'un nombre décimal (un ou deux chiffres).
1320
+ .IP 8. 4
1321
+ Taille totale du remplissage du flux dans le fichier
1322
+ .RE
1323
+ .PD
1324
+ .PP
1325
+ Les colonnes des lignes \fBstream\fP :
1326
+ .PD 0
1327
+ .RS
1328
+ .IP 2. 4
1329
+ Numéro de flux (le premier flux a le numéro 1)
1330
+ .IP 3. 4
1331
+ Nombre de blocs dans le flux
1332
+ .IP 4. 4
1333
+ Décalage de départ compressé
1334
+ .IP 5. 4
1335
+ Décalage de départ décompressé
1336
+ .IP 6. 4
1337
+ Taille compressée (ne comprend pas le remplissage du flux)
1338
+ .IP 7. 4
1339
+ Taille décompressée
1340
+ .IP 8. 4
1341
+ Taux de compression
1342
+ .IP 9. 4
1343
+ Nom de la vérification d'intégrité
1344
+ .IP 10. 4
1345
+ Taille du remplissage de flux
1346
+ .RE
1347
+ .PD
1348
+ .PP
1349
+ Les colonnes des lignes \fBblock\fP :
1350
+ .PD 0
1351
+ .RS
1352
+ .IP 2. 4
1353
+ Numéro du flux qui contient ce bloc
1354
+ .IP 3. 4
1355
+ Numéro du bloc relatif au commencement du flux (le premier bloc a pour
1356
+ numéro 1)
1357
+ .IP 4. 4
1358
+ Numéro du bloc relatif au début du fichier
1359
+ .IP 5. 4
1360
+ Décalage de départ compressé relatif au début du fichier
1361
+ .IP 6. 4
1362
+ Décalage de départ décompressé relatif au début du fichier
1363
+ .IP 7. 4
1364
+ Taille compressée totale du bloc (en\-têtes inclus)
1365
+ .IP 8. 4
1366
+ Taille décompressée
1367
+ .IP 9. 4
1368
+ Taux de compression
1369
+ .IP 10. 4
1370
+ Nom de la vérification d'intégrité
1371
+ .RE
1372
+ .PD
1373
+ .PP
1374
+ Si \fB\-\-verbose\fP a été indiqué deux fois, les colonnes additionnelles sont
1375
+ inclues sur les lignes \fBblock\fP. Elles ne sont pas affichées avec un seul
1376
+ \fB\-\-verbose\fP, car l'obtention de ces informations nécessite de nombreuses
1377
+ recherches et peut donc être lente :
1378
+ .PD 0
1379
+ .RS
1380
+ .IP 11. 4
1381
+ Valeur de la vérification d'intégrité en hexadécimal
1382
+ .IP 12. 4
1383
+ Taille d'en\-tête de bloc
1384
+ .IP 13. 4
1385
+ Drapeaux du bloc : \fBc\fP indique que la taille compressée est présente, et
1386
+ \fBu\fP indique que la taille décompréssée est présente. Si le drapeau n'est
1387
+ pas indiqué, un tiret (\fB\-\fP) est affiché à la place pour que la longueur de
1388
+ la chaîne reste fixe. De nouveaux drapeaux pourraient être ajoutés à la fin
1389
+ de la chaîne dans le futur.
1390
+ .IP 14. 4
1391
+ Taille des données effectivement compressées dans le bloc (en excluant
1392
+ l'en\-tête de bloc, le remplissage de bloc et les champs de vérification).
1393
+ .IP 15. 4
1394
+ Quantité de mémoire (en octets) nécessaire pour décompresser ce bloc avec
1395
+ cette version de \fBxz\fP.
1396
+ .IP 16. 4
1397
+ Chaîne de filtrage. Remarquez que la plupart des options utilisées au moment
1398
+ de la compression ne peuvent pas être connues, car seules les options
1399
+ nécessaires pour la décompression sont stockées dans les en\-têtes \fB.xz\fP.
1400
+ .RE
1401
+ .PD
1402
+ .PP
1403
+ Les colonnes des lignes \fBsummary\fP :
1404
+ .PD 0
1405
+ .RS
1406
+ .IP 2. 4
1407
+ Quantité de mémoire (en octets) nécessaire pour décompresser ce fichier avec
1408
+ cette version de \fBxz\fP.
1409
+ .IP 3. 4
1410
+ \fByes\fP ou \fBno\fP indique si tous les en\-têtes de bloc stockent à la fois la
1411
+ taille compressée et la taille décompressée.
1412
+ .PP
1413
+ \fIDepuis\fP \fBxz\fP \fI5.1.2alpha:\fP
1414
+ .IP 4. 4
1415
+ Version minimale de \fBxz\fP nécessaire pour décompresser le fichier.
1416
+ .RE
1417
+ .PD
1418
+ .PP
1419
+ Les colonnes de la ligne \fBtotals\fP :
1420
+ .PD 0
1421
+ .RS
1422
+ .IP 2. 4
1423
+ Nombre de flux
1424
+ .IP 3. 4
1425
+ Nombre de blocs
1426
+ .IP 4. 4
1427
+ Taille compressée
1428
+ .IP 5. 4
1429
+ Taille décompressée
1430
+ .IP 6. 4
1431
+ Taux de compression moyen
1432
+ .IP 7. 4
1433
+ Liste séparée par des virgules des noms de vérification d'intégrité qui
1434
+ étaient présents dans les fichiers
1435
+ .IP 8. 4
1436
+ Taille de remplissage de flux
1437
+ .IP 9. 4
1438
+ Nombre de fichiers. Permet de garder l'ordre des colonnes précédentes comme
1439
+ sur les lignes \fBfile\fP.
1440
+ .PD
1441
+ .RE
1442
+ .PP
1443
+ Si \fB\-\-verbose\fP a été indiqué deux fois, des colonnes supplémentaires sont
1444
+ incluses sur la ligne \fBtotals\fP :
1445
+ .PD 0
1446
+ .RS
1447
+ .IP 10. 4
1448
+ Quantité maximale de mémoire (en octets) nécessaire pour décompresser les
1449
+ fichiers avec cette version de \fBxz\fP.
1450
+ .IP 11. 4
1451
+ \fByes\fP ou \fBno\fP indique si tous les en\-têtes de bloc stockent à la fois la
1452
+ taille compressée et la taille décompressée.
1453
+ .PP
1454
+ \fIDepuis\fP \fBxz\fP \fI5.1.2alpha:\fP
1455
+ .IP 12. 4
1456
+ Version minimale de \fBxz\fP nécessaire pour décompresser le fichier.
1457
+ .RE
1458
+ .PD
1459
+ .PP
1460
+ Les versions futures pourront ajouter de nouveaux types de lignes et de
1461
+ nouvelles colonnes pourront être ajoutées aux types de lignes existants,
1462
+ mais les colonnes existantes ne seront pas modifiées.
1463
+ .
1464
+ .SS "Filters help"
1465
+ \fBxz \-\-robot \-\-filters\-help\fP prints the supported filters in the following
1466
+ format:
1467
+ .PP
1468
+ \fIfilter\fP\fB:\fP\fIoption\fP\fB=<\fP\fIvalue\fP\fB>,\fP\fIoption\fP\fB=<\fP\fIvalue\fP\fB>\fP...
1469
+ .TP
1470
+ \fIfilter\fP
1471
+ Name of the filter
1472
+ .TP
1473
+ \fIoption\fP
1474
+ Name of a filter specific option
1475
+ .TP
1476
+ \fIvalue\fP
1477
+ Numeric \fIvalue\fP ranges appear as \fB<\fP\fImin\fP\fB\-\fP\fImax\fP\fB>\fP. String
1478
+ \fIvalue\fP choices are shown within \fB< >\fP and separated by a \fB|\fP
1479
+ character.
1480
+ .PP
1481
+ Each filter is printed on its own line.
1482
+ .
1483
+ .SS "Information de limite de mémoire"
1484
+ \fBxz \-\-robot \-\-info\-memory\fP prints a single line with multiple tab\-separated
1485
+ columns:
1486
+ .IP 1. 4
1487
+ Total amount of physical memory (RAM) in bytes.
1488
+ .IP 2. 4
1489
+ Memory usage limit for compression in bytes (\fB\-\-memlimit\-compress\fP). A
1490
+ special value of \fB0\fP indicates the default setting which for
1491
+ single\-threaded mode is the same as no limit.
1492
+ .IP 3. 4
1493
+ Memory usage limit for decompression in bytes (\fB\-\-memlimit\-decompress\fP). A
1494
+ special value of \fB0\fP indicates the default setting which for
1495
+ single\-threaded mode is the same as no limit.
1496
+ .IP 4. 4
1497
+ Since \fBxz\fP 5.3.4alpha: Memory usage for multi\-threaded decompression in
1498
+ bytes (\fB\-\-memlimit\-mt\-decompress\fP). This is never zero because a
1499
+ system\-specific default value shown in the column 5 is used if no limit has
1500
+ been specified explicitly. This is also never greater than the value in the
1501
+ column 3 even if a larger value has been specified with
1502
+ \fB\-\-memlimit\-mt\-decompress\fP.
1503
+ .IP 5. 4
1504
+ Since \fBxz\fP 5.3.4alpha: A system\-specific default memory usage limit that is
1505
+ used to limit the number of threads when compressing with an automatic
1506
+ number of threads (\fB\-\-threads=0\fP) and no memory usage limit has been
1507
+ specified (\fB\-\-memlimit\-compress\fP). This is also used as the default value
1508
+ for \fB\-\-memlimit\-mt\-decompress\fP.
1509
+ .IP 6. 4
1510
+ Since \fBxz\fP 5.3.4alpha: Number of available processor threads.
1511
+ .PP
1512
+ Dans le futur, la sortie de \fBxz \-\-robot \-\-info\-memory\fP pourrait avoir plus
1513
+ de colonnes, mais jamais plus qu'une ligne unique.
1514
+ .
1515
+ .SS Version
1516
+ \fBxz \-\-robot \-\-version\fP prints the version number of \fBxz\fP and liblzma in
1517
+ the following format:
1518
+ .PP
1519
+ \fBXZ_VERSION=\fP\fIXYYYZZZS\fP
1520
+ .br
1521
+ \fBLIBLZMA_VERSION=\fP\fIXYYYZZZS\fP
1522
+ .TP
1523
+ \fIX\fP
1524
+ Version majeure.
1525
+ .TP
1526
+ \fIYYY\fP
1527
+ Version mineure. Les numéros pairs sont stables. Les numéros impairs sont
1528
+ des versions alpha ou beta.
1529
+ .TP
1530
+ \fIZZZ\fP
1531
+ Niveau de correctif pour les options stables ou juste un compteur pour les
1532
+ options de développement.
1533
+ .TP
1534
+ \fIS\fP
1535
+ Stabilité. 0 est alpha, 1 est bêta et 2 est stable. \fIS\fP devrait toujours
1536
+ être 2 quand \fIYYY\fP est pair.
1537
+ .PP
1538
+ \fIXYYYZZZS\fP sont identiques sur les deux lignes si \fBxz\fP et liblzma sont
1539
+ issus de la même version d'utilitaires XZ.
1540
+ .PP
1541
+ Exemples : 4.999.9beta est \fB49990091\fP et 5.0.0 est \fB50000002\fP.
1542
+ .
1543
+ .SH "STATUT DE SORTIE"
1544
+ .TP
1545
+ \fB0\fP
1546
+ Tout est bon.
1547
+ .TP
1548
+ \fB1\fP
1549
+ Une erreur est survenue.
1550
+ .TP
1551
+ \fB2\fP
1552
+ Quelquechose méritant un avertissement s'est produit, mais aucune erreur
1553
+ véritable n'est survenue.
1554
+ .PP
1555
+ Les notifications (pas les avertissements ou les erreurs) affichées sur
1556
+ l'erreur standard n'affectent pas le statut de sortie.
1557
+ .
1558
+ .SH ENVIRONNEMENT
1559
+ \fBxz\fP analyse les listes d'options séparées par des espaces à partir des
1560
+ variables d'environnement \fBXZ_DEFAULTS\fP et \fBXZ_OPT\fP, dans cet ordre, avant
1561
+ d'analyser les options de la ligne de commandes. Remarquez que seules les
1562
+ options sont analysées depuis l'environnement des variables ; toutes les
1563
+ non\-options sont ignorées silencieusement. L'analyse est faite avec
1564
+ \fBgetopt_long\fP(3) qui est aussi utilisé pour les arguments de la ligne de
1565
+ commandes.
1566
+ .TP
1567
+ \fBXZ_DEFAULTS\fP
1568
+ Options par défaut propres à l'utilisateur ou pour tout le système. Elles
1569
+ sont le plus souvent définies dans un script d'initialisation de
1570
+ l'interpréteur pour activer le limiteur d'utilisation de la mémoire de \fBxz\fP
1571
+ par défaut. A part pour les scripts d'initialisation de l'interpréteur ou
1572
+ des cas similaires, les sripts ne doivent jamais définir ou désactiver
1573
+ \fBXZ_DEFAULTS\fP.
1574
+ .TP
1575
+ \fBXZ_OPT\fP
1576
+ This is for passing options to \fBxz\fP when it is not possible to set the
1577
+ options directly on the \fBxz\fP command line. This is the case when \fBxz\fP is
1578
+ run by a script or tool, for example, GNU \fBtar\fP(1):
1579
+ .RS
1580
+ .RS
1581
+ .PP
1582
+ .nf
1583
+ \f(CRXZ_OPT=\-2v tar caf foo.tar.xz foo\fR
1584
+ .fi
1585
+ .RE
1586
+ .RE
1587
+ .IP ""
1588
+ Scripts may use \fBXZ_OPT\fP, for example, to set script\-specific default
1589
+ compression options. It is still recommended to allow users to override
1590
+ \fBXZ_OPT\fP if that is reasonable. For example, in \fBsh\fP(1) scripts one may
1591
+ use something like this:
1592
+ .RS
1593
+ .RS
1594
+ .PP
1595
+ .nf
1596
+ \f(CRXZ_OPT=${XZ_OPT\-"\-7e"}
1597
+ export XZ_OPT\fR
1598
+ .fi
1599
+ .RE
1600
+ .RE
1601
+ .
1602
+ .SH "Compatibilité des utilitaires LZMA"
1603
+ La syntaxe de la ligne de commande de \fBxz\fP est quasimment un sur\-ensemble
1604
+ de \fBlzma\fP, \fBunlzma\fP et \fBlzcat\fP comme ils sont trouvés dans les
1605
+ utilitaires LZMA 4.32.x . Dans la pluspart des cas, il est possible de
1606
+ remplacer les outils LZMA par les outils XZ sans casser les scripts
1607
+ existants. Il existe cependant certaines incompatibilités qui peuvent
1608
+ parfois poser des problèmes.
1609
+ .
1610
+ .SS "Niveaux de préréglage de la compression"
1611
+ La numérotation des préréglages de niveau de compression est différente
1612
+ entre les outils \fBxz\fP et LZMA. La différence la plus importante est la
1613
+ manière dont les tailles de dictionnaire sont affectées aux différents
1614
+ préréglages. La taille de dictionnaire est à peu près égale à celle
1615
+ d'utilisation de la mémoire de la décompression.
1616
+ .RS
1617
+ .PP
1618
+ .TS
1619
+ tab(;);
1620
+ c c c
1621
+ c n n.
1622
+ Niveau;xz;Utilitaires LZMA
1623
+ \-0;256 KiB;N/A
1624
+ \-1;1 MiB;64 KiB
1625
+ \-2;2 MiB;1 MiB
1626
+ \-3;4 MiB;512 KiB
1627
+ \-4;4 MiB;1 MiB
1628
+ \-5;8 MiB;2 MiB
1629
+ \-6;8 MiB;4 MiB
1630
+ \-7;16 MiB;8 MiB
1631
+ \-8;32 MiB;16 MiB
1632
+ \-9;64 MiB;32 MiB
1633
+ .TE
1634
+ .RE
1635
+ .PP
1636
+ Les différences de tailles des dictionnaires affectent aussi l'utilisation
1637
+ de la mémoire du compresseur, mais il y a quelques autres différences entre
1638
+ les outils LZMA et les outils XZ, qui rendent la différence encore plus
1639
+ grande :
1640
+ .RS
1641
+ .PP
1642
+ .TS
1643
+ tab(;);
1644
+ c c c
1645
+ c n n.
1646
+ Niveau;xz;Utilitaires LZMA 4.32.x
1647
+ \-0;3 MiB;N/A
1648
+ \-1;9 MiB;2 MiB
1649
+ \-2;17 MiB;12 MiB
1650
+ \-3;32 MiB;12 MiB
1651
+ \-4;48 MiB;16 MiB
1652
+ \-5;94 MiB;26 MiB
1653
+ \-6;94 MiB;45 MiB
1654
+ \-7;186 MiB;83 MiB
1655
+ \-8;370 MiB;159 MiB
1656
+ \-9;674 MiB;311 MiB
1657
+ .TE
1658
+ .RE
1659
+ .PP
1660
+ Le niveau de préréglage par défaut dans les outils LZMA est \fB\-7\fP alors que
1661
+ pour les outils XZ c'est \fB\-6\fP, les deux utilisent ainsi un dictionnaire de
1662
+ 8 Mio par défaut.
1663
+ .
1664
+ .SS "Fichiers .lzma en flux ou non"
1665
+ The uncompressed size of the file can be stored in the \fB.lzma\fP header.
1666
+ LZMA Utils does that when compressing regular files. The alternative is to
1667
+ mark that uncompressed size is unknown and use end\-of\-payload marker to
1668
+ indicate where the decompressor should stop. LZMA Utils uses this method
1669
+ when uncompressed size isn't known, which is the case, for example, in
1670
+ pipes.
1671
+ .PP
1672
+ \fBxz\fP prend en charge la décompression des fichiers \fB.lzma\fP avec ou sans
1673
+ marqueur de fin de charge utile, mais tous les fichiers \fB.lzma\fP créés par
1674
+ \fBxz\fP utiliseront un marqueur de fin de charge utile et ont la taille non
1675
+ compréssée marquée comme inconnue dans l'en\-tête \fB.lzma\fP. Cela peut être un
1676
+ problème dans quelques situations inhabituelles. Par exemple, un
1677
+ décompresseur \fB.lzma\fP dans un périphérique embarqué pourrait ne fonctionner
1678
+ qu'avec des fichiers dont la taille non comprimée est connue. Si vous vous
1679
+ heurtez à ce problème, vous devez utiliser les utilitaires LZMA ou LZMA SDK
1680
+ pour créer des fichiers \fB.lzma\fP avec une taille non compressée connue.
1681
+ .
1682
+ .SS "Fichiers .lzma non pris en charge"
1683
+ Le format \fB.lzma\fP autorise des valeurs \fIlc\fP jusqu'à 8, et des valeurs
1684
+ \fIlp\fP jusqu'à 4. Les outils LZMA peuvent décompresser des fichiers avec tous
1685
+ les \fIlc\fP et \fIlp\fP, mais créez toujours les fichiers avec \fBlc=3\fP et
1686
+ \fBlp=0\fP. Créer des fichiers avec d'autres valeurs \fIlc\fP et \fIlp\fP est
1687
+ possible avec \fBxz\fP et avec LZMA SDK.
1688
+ .PP
1689
+ L'implémentation du filtre LZMA1 dans liblzma nécessite que la somme de
1690
+ \fIlc\fP et \fIlp\fP soit inférieure ou égale à 4. Ainsi, les fichiers \fB.lzma\fP
1691
+ qui excèdent cette limitation ne peuvent pas être décompressés avec \fBxz\fP.
1692
+ .PP
1693
+ Les outils LZMA créent seulement des fichiers \fB.lzma\fP qui ont une taille de
1694
+ dictionnaire de 2^\fIn\fP (une puissance de 2) mais acceptent les fichiers avec
1695
+ toutes les tailles de dictionnaire. Libzlma n'accepte que les fichiers
1696
+ \&\fB.lzma\fP qui ont une taille dictionnaire de 2^\fIn\fP ou
1697
+ 2^\fIn\fP + 2^(\fIn\fP\-1). Cela afin de diminuer les faux positifs lors de la
1698
+ détection des fichiers \fB.lzma\fP.
1699
+ .PP
1700
+ Ces limitations ne devraient pas poser problème en pratique, car
1701
+ pratiquement tous les fichiers \fB.lzma\fP ont été compressés avec des réglages
1702
+ que liblzma accepte.
1703
+ .
1704
+ .SS "Déchets excédentaires"
1705
+ Lors de la décompession, l'utilitaire LZMA ignore silencieusement tout ce
1706
+ qui est après le premier flux \fB.lzma\fP. Dans la majorité des situations,
1707
+ c'est un bogue. Cela veut dire aussi que les outils LZMA ne gèrent pas la
1708
+ décompression de fichiers \fB.lzma\fP concaténés.
1709
+ .PP
1710
+ S'il reste des données après le premier flux \fB.lzma\fP, \fBxz\fP considère que
1711
+ le fichier est corrompu sauf si \fB\-\-single\-stream\fP a été utilisé. Cela peut
1712
+ casser des scripts obscurs qui ont supposé que les déchets de fin de ligne
1713
+ sont ignorés.
1714
+ .
1715
+ .SH NOTES
1716
+ .
1717
+ .SS "La sortie compressée peut varier"
1718
+ La sortie compressée exacte produite par les même fichiers non compressés en
1719
+ entrée peut varier en fonction des différentes versions de l'utilitaire XZ,
1720
+ même si les options de compression sont identiques. En effet, il est
1721
+ possible d'améliorer l'encodeur (compression plus rapide ou meilleure) sans
1722
+ affecter le format du fichier. La sortie peut même varier entre différentes
1723
+ compilations de la même version d'utilitaire XZ, si des options de
1724
+ construction différentes sont utilisées.
1725
+ .PP
1726
+ Cela signifie qu'une fois que \fB\-\-rsyncable\fP a été implémenté, les fichiers
1727
+ résultants ne seront pas nécessairement synchronisables avec rsync à moins
1728
+ que les nouveaux et anciens fichiers n'aient été compressés avec la même
1729
+ version de xz. Ce problème peut être résolu si une partie de
1730
+ l'implémentation est gelée pour garantir la stabilité de la sortie rsyncable
1731
+ à travers les versions de xz.
1732
+ .
1733
+ .SS "Décompresseurs .xz embarqués"
1734
+ Les implémentations de décompresseur embarqué comme XZ Embedded ne gèrent
1735
+ pas nécessairement les fichiers créés avec d'autres types de \fIvérification\fP
1736
+ d'intégrité que \fBnone\fP et \fBCRC32\fP. Comme la valeur par défaut est
1737
+ \fB\-\-check=crc64\fP, vous devez utiliser \fB\-\-check=none\fP ou \fB\-\-check=crc32\fP
1738
+ lors de la création de fichiers pour les systèmes embarqués.
1739
+ .PP
1740
+ En dehors des systèmes embarqués, tous les décompresseurs de format \fB.xz\fP
1741
+ gèrent tous les types de \fIvérification\fP ou sont au moins capables de
1742
+ décompresser le fichier sans effectuer la vérification d'intégrité si ce
1743
+ type de \fIvérification\fP particulière n'est pas pris en charge.
1744
+ .PP
1745
+ XZ Embedded prend en charge les filtres BCJ, mais seulement avec le décalage
1746
+ de départ par défaut.
1747
+ .
1748
+ .SH EXEMPLES
1749
+ .
1750
+ .SS Bases
1751
+ Compresser le fichier \fItoto\fP en \fItoto.xz\fP en utilisant le niveau de
1752
+ compression par défaut (\fB\-6\fP) et supprimer \fItoto\fP si la compression
1753
+ réussit :
1754
+ .RS
1755
+ .PP
1756
+ .nf
1757
+ \f(CRxz foo\fR
1758
+ .fi
1759
+ .RE
1760
+ .PP
1761
+ Décompresser \fIbidule.xz\fP en \fIbidule\fP et ne pas supprimer \fIbidule.xz\fP même
1762
+ si la compression réussit :
1763
+ .RS
1764
+ .PP
1765
+ .nf
1766
+ \f(CRxz \-dk bar.xz\fR
1767
+ .fi
1768
+ .RE
1769
+ .PP
1770
+ Create \fIbaz.tar.xz\fP with the preset \fB\-4e\fP (\fB\-4 \-\-extreme\fP), which is
1771
+ slower than the default \fB\-6\fP, but needs less memory for compression and
1772
+ decompression (48\ MiB and 5\ MiB, respectively):
1773
+ .RS
1774
+ .PP
1775
+ .nf
1776
+ \f(CRtar cf \- baz | xz \-4e > baz.tar.xz\fR
1777
+ .fi
1778
+ .RE
1779
+ .PP
1780
+ Un mélange de fichiers compressés et non compressés peuvent être
1781
+ décompressés vers la sortie standard avec une simple commande :
1782
+ .RS
1783
+ .PP
1784
+ .nf
1785
+ \f(CRxz \-dcf a.txt b.txt.xz c.txt d.txt.lzma > abcd.txt\fR
1786
+ .fi
1787
+ .RE
1788
+ .
1789
+ .SS "Compression en parallèle de plusieurs fichiers"
1790
+ Sur GNU et *BSD, \fBfind\fP(1) et \fBxargs\fP(1) peuvent être utilisés pour mettre
1791
+ en parallèle la compression de plusieurs fichiers :
1792
+ .RS
1793
+ .PP
1794
+ .nf
1795
+ \f(CRfind . \-type f \e! \-name '*.xz' \-print0 \e
1796
+ | xargs \-0r \-P4 \-n16 xz \-T1\fR
1797
+ .fi
1798
+ .RE
1799
+ .PP
1800
+ L'option \fBP\fP passée à \fBxargs\fP(1) fixe le nombre de processus \fBxz\fP en
1801
+ parallèles. La meilleure valeur pour l'option \fBn\fP dépend du nombre de
1802
+ fichiers à compresser. S\-il n'y a que quelques fichiers, la valeur sera
1803
+ probablement 1 ; avec des dizaines de milliers de fichiers, 100 ou même plus
1804
+ serait approprié pour réduire le nombre de processus \fBxz\fP que \fBxargs\fP(1)
1805
+ créera éventuellement.
1806
+ .PP
1807
+ L'option \fB\-T1\fP de \fBxz\fP est là pour le forcer en mode mono\-thread, car
1808
+ \fBxargs\fP(1) est utilisé pour contrôler la quantité de mise en parallèle.
1809
+ .
1810
+ .SS "Mode robot"
1811
+ Calculer combien d'octets ont été économisés au total après avoir compressé
1812
+ plusieurs fichiers :
1813
+ .RS
1814
+ .PP
1815
+ .nf
1816
+ \f(CRxz \-\-robot \-\-list *.xz | awk '/^totals/{print $5\-$4}'\fR
1817
+ .fi
1818
+ .RE
1819
+ .PP
1820
+ Un script peut vouloir savoir qu'il utilise une version suffisamment récente
1821
+ de \fBxz\fP. Le script \fBsh\fP(1) suivant vérifie que le numéro de version de
1822
+ l'outil \fBxz\fP soit au minimum 5.0.0. Cette méthode est compatible avec les
1823
+ vieilles versions bêta, qui ne gèrent pas l'option \fB\-\-robot\fP :
1824
+ .RS
1825
+ .PP
1826
+ .nf
1827
+ \f(CRif ! eval "$(xz \-\-robot \-\-version 2> /dev/null)" ||
1828
+ [ "$XZ_VERSION" \-lt 50000002 ]; then
1829
+ echo "Your xz is too old."
1830
+ fi
1831
+ unset XZ_VERSION LIBLZMA_VERSION\fR
1832
+ .fi
1833
+ .RE
1834
+ .PP
1835
+ Régler une limite d'utilisation de la mémoire pour la décompression en
1836
+ utilisant \fBXZ_OPT\fP, mais si une limite a déjà été définie, ne pas
1837
+ l'augmenter :
1838
+ .RS
1839
+ .PP
1840
+ .nf
1841
+ \f(CRNEWLIM=$((123 << 20))\ \ # 123 MiB
1842
+ OLDLIM=$(xz \-\-robot \-\-info\-memory | cut \-f3)
1843
+ if [ $OLDLIM \-eq 0 \-o $OLDLIM \-gt $NEWLIM ]; then
1844
+ XZ_OPT="$XZ_OPT \-\-memlimit\-decompress=$NEWLIM"
1845
+ export XZ_OPT
1846
+ fi\fR
1847
+ .fi
1848
+ .RE
1849
+ .
1850
+ .SS "Chaînes de filtres de compresseur personnalisées"
1851
+ L'utilisation la plus simple des chaînes de filtres personnalisées est la
1852
+ personnalisation d'un préréglage LZMA2. Cela peut être utile, car les
1853
+ préréglages ne couvrent qu'un sous\-ensemble des réglages de compression
1854
+ potentiellement utiles.
1855
+ .PP
1856
+ Les colonnes CompCPU des tableaux des descriptions des options \fB\-0\fP à \fB\-9\fP
1857
+ et \fB\-\-extreme\fP sont utiles lors de la personnalisation des préréglages
1858
+ LZMA2. Voici les parties pertinentes recueillies à partir de ces deux
1859
+ tableaux :
1860
+ .RS
1861
+ .PP
1862
+ .TS
1863
+ tab(;);
1864
+ c c
1865
+ n n.
1866
+ Préréglage;CompCPU
1867
+ \-0;0
1868
+ \-1;1
1869
+ \-2;2
1870
+ \-3;3
1871
+ \-4;4
1872
+ \-5;5
1873
+ \-6;6
1874
+ \-5e;7
1875
+ \-6e;8
1876
+ .TE
1877
+ .RE
1878
+ .PP
1879
+ If you know that a file requires somewhat big dictionary (for example, 32\ MiB) to compress well, but you want to compress it quicker than \fBxz \-8\fP
1880
+ would do, a preset with a low CompCPU value (for example, 1) can be
1881
+ modified to use a bigger dictionary:
1882
+ .RS
1883
+ .PP
1884
+ .nf
1885
+ \f(CRxz \-\-lzma2=preset=1,dict=32MiB foo.tar\fR
1886
+ .fi
1887
+ .RE
1888
+ .PP
1889
+ Avec certains fichiers, la commande ci\-dessus peut être plus rapide que
1890
+ \fBxz\-6\fP tout en compressant bien mieux. Cependant, il faut souligner que
1891
+ seuls certains fichiers bénéficient d'un grand dictionnaire tout en gardant
1892
+ la valeur de CompCPU faible. La siutation la plus évidente où un gros
1893
+ dictionnaire peut baucoup aider, est une archive contenant des fichiers très
1894
+ similaires de quelques megaoctets chacun. La taille de dictionnaire doit
1895
+ être significativement plus grosse que tout fichier individuel pour
1896
+ permettre à LZMA2 de tirer pleinement partie des similarités entre des
1897
+ fichiers consécutifs.
1898
+ .PP
1899
+ Si une utilisation de la mémoire élevée pour la compression et décompression
1900
+ convient, et que le fichier à compresser a une taille de plusieurs centaines
1901
+ de megaoctets, il peut être utile d'utiliser un plus gros dictionnaire que
1902
+ celui fourni par \fBxz\-9\fP (64 Mio) :
1903
+ .RS
1904
+ .PP
1905
+ .nf
1906
+ \f(CRxz \-vv \-\-lzma2=dict=192MiB big_foo.tar\fR
1907
+ .fi
1908
+ .RE
1909
+ .PP
1910
+ Utiliser \fB\-vv\fP (\fB\-\-verbose\-\-verbose\fP) comme dans l'exemple ci\-dessus peut
1911
+ être utile pour voir les besoins en mémoire du compresseur et du
1912
+ décompresseur. Rappelez\-vous qu'utiliser un dictionnaire plus gros que la
1913
+ taille du fichier non compressé est un gachis de mémoire, donc la commande
1914
+ ci\-dessus n'est pas utile pour les petits fichiers.
1915
+ .PP
1916
+ Sometimes the compression time doesn't matter, but the decompressor memory
1917
+ usage has to be kept low, for example, to make it possible to decompress the
1918
+ file on an embedded system. The following command uses \fB\-6e\fP (\fB\-6 \-\-extreme\fP) as a base and sets the dictionary to only 64\ KiB. The
1919
+ resulting file can be decompressed with XZ Embedded (that's why there is
1920
+ \fB\-\-check=crc32\fP) using about 100\ KiB of memory.
1921
+ .RS
1922
+ .PP
1923
+ .nf
1924
+ \f(CRxz \-\-check=crc32 \-\-lzma2=preset=6e,dict=64KiB foo\fR
1925
+ .fi
1926
+ .RE
1927
+ .PP
1928
+ If you want to squeeze out as many bytes as possible, adjusting the number
1929
+ of literal context bits (\fIlc\fP) and number of position bits (\fIpb\fP) can
1930
+ sometimes help. Adjusting the number of literal position bits (\fIlp\fP)
1931
+ might help too, but usually \fIlc\fP and \fIpb\fP are more important. For
1932
+ example, a source code archive contains mostly US\-ASCII text, so something
1933
+ like the following might give slightly (like 0.1\ %) smaller file than \fBxz \-6e\fP (try also without \fBlc=4\fP):
1934
+ .RS
1935
+ .PP
1936
+ .nf
1937
+ \f(CRxz \-\-lzma2=preset=6e,pb=0,lc=4 source_code.tar\fR
1938
+ .fi
1939
+ .RE
1940
+ .PP
1941
+ Using another filter together with LZMA2 can improve compression with
1942
+ certain file types. For example, to compress a x86\-32 or x86\-64 shared
1943
+ library using the x86 BCJ filter:
1944
+ .RS
1945
+ .PP
1946
+ .nf
1947
+ \f(CRxz \-\-x86 \-\-lzma2 libfoo.so\fR
1948
+ .fi
1949
+ .RE
1950
+ .PP
1951
+ Notez que l'ordre des options de filtre est significatif. Si \fB\-\-x86\fP est
1952
+ indiqué après \fB\-\-lzma2\fP, \fBxz\fP donnera une erreur, car il ne peut y avoir
1953
+ aucun filtre après LZMA2, et aussi parce que le filtre BCJ x86 ne peut pas
1954
+ être utilisé comme dernier filtre dans la chaîne.
1955
+ .PP
1956
+ Le filtre Delta associé à LZMA2 peut donner de bons résultats avec les
1957
+ images bitmap. Cela devrait habituellement battre PNG, qui a quelques
1958
+ filtres avancés supplémentaires qu'un simple delta, mais qui utilise Deflate
1959
+ pour la compression effective.
1960
+ .PP
1961
+ The image has to be saved in uncompressed format, for example, as
1962
+ uncompressed TIFF. The distance parameter of the Delta filter is set to
1963
+ match the number of bytes per pixel in the image. For example, 24\-bit RGB
1964
+ bitmap needs \fBdist=3\fP, and it is also good to pass \fBpb=0\fP to LZMA2 to
1965
+ accommodate the three\-byte alignment:
1966
+ .RS
1967
+ .PP
1968
+ .nf
1969
+ \f(CRxz \-\-delta=dist=3 \-\-lzma2=pb=0 foo.tiff\fR
1970
+ .fi
1971
+ .RE
1972
+ .PP
1973
+ If multiple images have been put into a single archive (for example,
1974
+ \&\fB.tar\fP), the Delta filter will work on that too as long as all images have
1975
+ the same number of bytes per pixel.
1976
+ .
1977
+ .SH "VOIR AUSSI"
1978
+ \fBxzdec\fP(1), \fBxzdiff\fP(1), \fBxzgrep\fP(1), \fBxzless\fP(1), \fBxzmore\fP(1),
1979
+ \fBgzip\fP(1), \fBbzip2\fP(1), \fB7z\fP(1)
1980
+ .PP
1981
+ XZ Utils: <https://tukaani.org/xz/>
1982
+ .br
1983
+ XZ Embedded: <https://tukaani.org/xz/embedded.html>
1984
+ .br
1985
+ LZMA SDK: <https://7\-zip.org/sdk.html>
janus/share/man/fr/man1/lzless.1 ADDED
@@ -0,0 +1,49 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ .\" SPDX-License-Identifier: 0BSD
2
+ .\"
3
+ .\" Authors: Andrew Dudman
4
+ .\" Lasse Collin
5
+ .\"
6
+ .\" french translation of XZ Utils man
7
+ .\" Copyright (C) 2021 Debian French l10n team <debian-l10n-french@lists.debian.org>
8
+ .\" Translator
9
+ .\" bubu <bubub@no-log.org>, 2021.
10
+ .\"
11
+ .\" (Note that this file is not based on gzip's zless.1.)
12
+ .\"
13
+ .\"*******************************************************************
14
+ .\"
15
+ .\" This file was generated with po4a. Translate the source file.
16
+ .\"
17
+ .\"*******************************************************************
18
+ .TH XZLESS 1 2024\-02\-12 Tukaani "Utilitaires XZ"
19
+ .SH NOM
20
+ xzless, lzless \- Voir le contenu des fichiers (texte) compressés xz ou lzma
21
+ .SH SYNOPSIS
22
+ \fBxzless\fP [\fIfichier\fP...]
23
+ .br
24
+ \fBlzless\fP [\fIfichier\fP...]
25
+ .SH DESCRIPTION
26
+ \fBxzless\fP is a filter that displays text from compressed files to a
27
+ terminal. Files supported by \fBxz\fP(1) are decompressed; other files are
28
+ assumed to be in uncompressed form already. If no \fIfiles\fP are given,
29
+ \fBxzless\fP reads from standard input.
30
+ .PP
31
+ \fBxzless\fP utilise \fBless\fP(1) pour afficher sa sortie. Contrairement à
32
+ \fBxzmore\fP, son choix d'afficheur ne peut pas être modifié en indiquant une
33
+ variable d'environnement. Les commandes sont basées sur \fBmore\fP(1) et
34
+ \fBvi\fP(1) et permettent des déplacements en avant et en arrière et des
35
+ recherches. Voir le manuel de \fBless\fP(1) pour plus d'information.
36
+ .PP
37
+ La commande nommée \fBlzless\fP est fournie pour la rétrocompatibilité avec les
38
+ utilitaires LZMA.
39
+ .SH ENVIRONNEMENT
40
+ .TP
41
+ \fBLESSMETACHARS\fP
42
+ Une liste de caractères spéciaux pour l'interpréteur. Définis par \fBxzless\fP
43
+ à moins qu'ils ne l'aient déjà été dans l'environnement.
44
+ .TP
45
+ \fBLESSOPEN\fP
46
+ Définir en ligne de commande le décompresseur \fBxz\fP(1) à invoquer pour
47
+ préparer les fichiers en entrée pour \fBless\fP(1).
48
+ .SH "VOIR AUSSI"
49
+ \fBless\fP(1), \fBxz\fP(1), \fBxzmore\fP(1), \fBzless\fP(1)
janus/share/man/fr/man1/lzma.1 ADDED
@@ -0,0 +1,1985 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ '\" t
2
+ .\" SPDX-License-Identifier: 0BSD
3
+ .\"
4
+ .\" Authors: Lasse Collin
5
+ .\" Jia Tan
6
+ .\"
7
+ .\" french translation of XZ Utils man
8
+ .\" Copyright (C) 2021 Debian French l10n team <debian-l10n-french@lists.debian.org>
9
+ .\" Translator
10
+ .\" bubu <bubub@no-log.org>, 2021.
11
+ .\"
12
+ .\"*******************************************************************
13
+ .\"
14
+ .\" This file was generated with po4a. Translate the source file.
15
+ .\"
16
+ .\"*******************************************************************
17
+ .TH XZ 1 2024\-12\-30 Tukaani "Utilitaires XZ"
18
+ .
19
+ .SH NOM
20
+ xz, unxz, xzcat, lzma, unlzma, lzcat \- Compresser ou décompresser des
21
+ fichiers .xz et .lzma
22
+ .
23
+ .SH SYNOPSIS
24
+ \fBxz\fP [\fIoption...\fP] [\fIfichier...\fP]
25
+ .
26
+ .SH "ALIAS DES COMMANDES"
27
+ \fBunxz\fP est équivalent à \fBxz \-\-decompress\fP.
28
+ .br
29
+ \fBxzcat\fP est équivalent à \fBxz \-\-decompress \-\-stdout\fP
30
+ .br
31
+ \fBlzma\fP est équivalent à \fBxz \-\-format=lzma\fP
32
+ .br
33
+ \fBunlzma\fP est équivalent à \fBxz \-\-format=lzma \-\-decompress\fP
34
+ .br
35
+ \fBlzcat\fP est équivalent à \fBxz \-\-format=lzma \-\-decompress \-\- stdout\fP
36
+ .PP
37
+ Lors de l'écriture de scripts qui nécessitent de décompresser des fichiers,
38
+ il est recommandé de toujours utiliser la commande \fBxz\fP avec les arguments
39
+ appropriés (\fBxz \-d\fP ou \fBxz \-dc\fP) au lieu des commandes \fBunxz\fP et
40
+ \fBxzcat\fP.
41
+ .
42
+ .SH DESCRIPTION
43
+ \fBxz\fP is a general\-purpose data compression tool with command line syntax
44
+ similar to \fBgzip\fP(1) and \fBbzip2\fP(1). The native file format is the
45
+ \&\fB.xz\fP format, but the legacy \fB.lzma\fP format used by LZMA Utils and raw
46
+ compressed streams with no container format headers are also supported. In
47
+ addition, decompression of the \fB.lz\fP format used by \fBlzip\fP is supported.
48
+ .PP
49
+ \fBxz\fP compresse ou décompresse chaque \fIfichier\fP en fonction du mode
50
+ d'opération choisi. Si aucun \fIfichier\fP n'est donné ou \fIfichier\fP est \fB\-\fP,
51
+ \fBxz\fP lit depuis l'entrée standard et écrit les données traitées sur la
52
+ sortie standard. \fBxz\fP refusera (affichera une erreur et ignorera le
53
+ \fIfichier\fP) d'écrire les données compressées sur la sortie standard si c'est
54
+ un terminal. De même, \fBxz\fP refusera de lire des données compressées depuis
55
+ l'entrée standard si c'est un terminal.
56
+ .PP
57
+ A moins que \fB\-\-sdout\fP ne soit indiqué, les \fIfichiers\fP autres que \fB\-\fP sont
58
+ écrits dans un nouveau fichier dont le nom est dérivé du nom de \fIfichier\fP
59
+ source :
60
+ .IP \(bu 3
61
+ Lors de la compression, le suffixe du format de fichier cible (\fB.xz\fP ou
62
+ \&\fB.lzma\fP) est ajouté au nom de fichier source pour obtenir le nom de fichier
63
+ cible.
64
+ .IP \(bu 3
65
+ When decompressing, the \fB.xz\fP, \fB.lzma\fP, or \fB.lz\fP suffix is removed from
66
+ the filename to get the target filename. \fBxz\fP also recognizes the suffixes
67
+ \&\fB.txz\fP and \fB.tlz\fP, and replaces them with the \fB.tar\fP suffix.
68
+ .PP
69
+ Si le fichier cible existe déjà, une erreur est affichée et le \fIfichier\fP
70
+ est ignoré.
71
+ .PP
72
+ Sauf s'il écrit dans la sortie standard, \fBxz\fP affichera un avertissement et
73
+ ignorera le \fIfichier\fP dans les cas suivants :
74
+ .IP \(bu 3
75
+ \fIfichier\fP n'est pas un fichier normal. Les liens symboliques ne sont pas
76
+ suivis et donc ne sont pas considérés comme des fichiers normaux.
77
+ .IP \(bu 3
78
+ \fIfichier\fP a plusieurs liens physiques.
79
+ .IP \(bu 3
80
+ \fIfichier\fP a un setuid, setgid ou sticky bit positionné.
81
+ .IP \(bu 3
82
+ Le mode d'opération est défini pour compresser et le \fIfichier\fP a déjà un
83
+ suffixe du format de fichier cible (\fB.xz\fP ou \fB.txz\fP lors d'une compression
84
+ en format \fB.xz\fP, et \fB.lzma\fP ou \fB.tlz\fP lors d'une compression en format
85
+ \&\fB.lzma\fP).
86
+ .IP \(bu 3
87
+ The operation mode is set to decompress and the \fIfile\fP doesn't have a
88
+ suffix of any of the supported file formats (\fB.xz\fP, \fB.txz\fP, \fB.lzma\fP,
89
+ \&\fB.tlz\fP, or \fB.lz\fP).
90
+ .PP
91
+ Après la compression ou la décompression réussie du \fIfichier\fP, \fBxz\fP copie
92
+ les permissions du propriétaire, du groupe, la date d'accès, et les
93
+ modifications d'heure depuis le \fIfichier\fP source du fichier cible. Si la
94
+ copie du groupe échoue, les permissions sont modifiées pour que le fichier
95
+ cible ne soit pas accessible aux utilisateurs qui n'ont pas les droits
96
+ d'accès au \fIfichier\fP source. \fBxz\fP ne prend actuellement pas en charge la
97
+ copie d'autres métadonnées telles que les listes de contrôle d'accès ou les
98
+ attributs étendus.
99
+ .PP
100
+ Once the target file has been successfully closed, the source \fIfile\fP is
101
+ removed unless \fB\-\-keep\fP was specified. The source \fIfile\fP is never removed
102
+ if the output is written to standard output or if an error occurs.
103
+ .PP
104
+ Envoyer \fBSIGINFO\fP ou \fBSIGURSR1\fP au processus \fBxz\fP, lui fait afficher
105
+ l'information de progression sur l'erreur standard. Cela a un intérêt limité
106
+ car lorsque l'erreur standard est un terminal, utiliser \fB\-\-verbose\fP
107
+ affichera automatiquement un indicateur de progression du processus.
108
+ .
109
+ .SS "Utilisation de la mémoire"
110
+ L'utilisation de la mémoire par \fBxz\fP varie de quelques centaines de
111
+ kilo\-octets à plusieurs gigaoctects en fonction des paramètres de
112
+ compression. Les réglages utilisés lors de la compression d'un fichier
113
+ déterminent les besoins en mémoire pour la décompression. Habituellement la
114
+ décompression nécessite 5\% à 20\% de la quantité de mémoire utilisée pour
115
+ la compression du fichier. Par exemple, décompresser un fichier créé avec
116
+ \fBxz\-9\fP recquiert habituellement 65\ Mio de mémoire. Bien qu'il soit
117
+ possible d'avoir des fichiers \fB.xz\fP nécessitant plusieurs gigaoctets de
118
+ mémoire pour être décompressés.
119
+ .PP
120
+ Especially users of older systems may find the possibility of very large
121
+ memory usage annoying. To prevent uncomfortable surprises, \fBxz\fP has a
122
+ built\-in memory usage limiter, which is disabled by default. While some
123
+ operating systems provide ways to limit the memory usage of processes,
124
+ relying on it wasn't deemed to be flexible enough (for example, using
125
+ \fBulimit\fP(1) to limit virtual memory tends to cripple \fBmmap\fP(2)).
126
+ .PP
127
+ The memory usage limiter can be enabled with the command line option
128
+ \fB\-\-memlimit=\fP\fIlimit\fP. Often it is more convenient to enable the limiter
129
+ by default by setting the environment variable \fBXZ_DEFAULTS\fP, for example,
130
+ \fBXZ_DEFAULTS=\-\-memlimit=150MiB\fP. It is possible to set the limits
131
+ separately for compression and decompression by using
132
+ \fB\-\-memlimit\-compress=\fP\fIlimit\fP and \fB\-\-memlimit\-decompress=\fP\fIlimit\fP.
133
+ Using these two options outside \fBXZ_DEFAULTS\fP is rarely useful because a
134
+ single run of \fBxz\fP cannot do both compression and decompression and
135
+ \fB\-\-memlimit=\fP\fIlimit\fP (or \fB\-M\fP \fIlimit\fP) is shorter to type on the
136
+ command line.
137
+ .PP
138
+ If the specified memory usage limit is exceeded when decompressing, \fBxz\fP
139
+ will display an error and decompressing the file will fail. If the limit is
140
+ exceeded when compressing, \fBxz\fP will try to scale the settings down so that
141
+ the limit is no longer exceeded (except when using \fB\-\-format=raw\fP or
142
+ \fB\-\-no\-adjust\fP). This way the operation won't fail unless the limit is very
143
+ small. The scaling of the settings is done in steps that don't match the
144
+ compression level presets, for example, if the limit is only slightly less
145
+ than the amount required for \fBxz \-9\fP, the settings will be scaled down only
146
+ a little, not all the way down to \fBxz \-8\fP.
147
+ .
148
+ .SS "Concaténation et remplissage avec des fichiers .xz"
149
+ Il est possible de concaténer les fichiers \fB.xz\fP tels quel. \fBxz\fP
150
+ décompressera de tels fichiers comme s'ils étaient un unique fichier \fB.xz\fP.
151
+ .PP
152
+ It is possible to insert padding between the concatenated parts or after the
153
+ last part. The padding must consist of null bytes and the size of the
154
+ padding must be a multiple of four bytes. This can be useful, for example,
155
+ if the \fB.xz\fP file is stored on a medium that measures file sizes in
156
+ 512\-byte blocks.
157
+ .PP
158
+ La concaténation et le remplissage ne sont pas autorisés avec les fichiers
159
+ \&\fB.lzma\fP ou les flux bruts.
160
+ .
161
+ .SH OPTIONS
162
+ .
163
+ .SS "Suffixes entiers et valeurs spéciales."
164
+ Dans la plupart des endroits où un argument entier est attendu, un suffixe
165
+ optionel permet d'indiquer facilement les grands entiers. Il ne doit pas y
166
+ avoir d'espace entre l'entier et le suffixe.
167
+ .TP
168
+ \fBKiB\fP
169
+ Multiplier l'entier par 1024 (2^10). \fBKi\fP, \fBk\fP, \fBkB\fP, \fBK\fP et \fBKB\fP sont
170
+ acceptés comme synonymes de \fBKiB\fP.
171
+ .TP
172
+ \fBMiB\fP
173
+ Multiplier l'entier par 1 048 576 (2^20). \fBMi\fP, \fBm\fP, \fBM\fP et \fBMB\fP sont
174
+ acceptés comme synonymes de \fBMiB\fP.
175
+ .TP
176
+ \fBGiB\fP
177
+ Multiplier l'entier par 1 073 741 824 (2^30). \fBGi\fP, \fBg\fP, \fBG\fP et \fBGB\fP
178
+ sont acceptés comme synonymes de \fBGiB\fP.
179
+ .PP
180
+ La valeur spéciale \fBmax\fP peut être utilisée pour indiquer la valeur
181
+ maximale de l'entier prise en charge par l'option.
182
+ .
183
+ .SS "Mode d'opération"
184
+ Si plusieurs options de mode d'opération sont données, la dernière prend
185
+ effet.
186
+ .TP
187
+ \fB\-z\fP, \fB\-\-compress\fP
188
+ Compresser. C'est le mode d'opération par défaut lorsque aucune option de
189
+ mode opératoire n'est spécifiée ou qu'aucun autre mode d'opération n'est
190
+ sous\-entendu par le nom de la commande (par exemple \fBunxz\fP sous\-entend
191
+ \fB\-\-decompress\fP).
192
+ .IP ""
193
+ .\" The DESCRIPTION section already says this but it's good to repeat it
194
+ .\" here because the default behavior is a bit dangerous and new users
195
+ .\" in a hurry may skip reading the DESCRIPTION section.
196
+ After successful compression, the source file is removed unless writing to
197
+ standard output or \fB\-\-keep\fP was specified.
198
+ .TP
199
+ \fB\-d\fP, \fB\-\-decompress\fP, \fB\-\-uncompress\fP
200
+ .\" The DESCRIPTION section already says this but it's good to repeat it
201
+ .\" here because the default behavior is a bit dangerous and new users
202
+ .\" in a hurry may skip reading the DESCRIPTION section.
203
+ Decompress. After successful decompression, the source file is removed
204
+ unless writing to standard output or \fB\-\-keep\fP was specified.
205
+ .TP
206
+ \fB\-t\fP, \fB\-\-test\fP
207
+ Tester l'intégrité des \fIfichiers\fP compressés. Cette option est équivalente
208
+ à \fB\-\-decompress \-\-stdout\fP sauf que les données décompressées sont rejetées
209
+ au lieu d'être écrites sur la sortie standard. Aucun fichier n'est créé ou
210
+ supprimé.
211
+ .TP
212
+ \fB\-l\fP, \fB\-\-list\fP
213
+ Afficher l'information sur les \fIfichiers\fP compressés. Aucune sortie
214
+ non\-compressée n'est produite et aucun fichier n'est créé ou supprimé. En
215
+ mode liste, le programme ne peut pas lire les données compressées depuis
216
+ l'entrée standard ou depuis d'autres sources non adressables.
217
+ .IP ""
218
+ The default listing shows basic information about \fIfiles\fP, one file per
219
+ line. To get more detailed information, use also the \fB\-\-verbose\fP option.
220
+ For even more information, use \fB\-\-verbose\fP twice, but note that this may be
221
+ slow, because getting all the extra information requires many seeks. The
222
+ width of verbose output exceeds 80 characters, so piping the output to, for
223
+ example, \fBless\ \-S\fP may be convenient if the terminal isn't wide enough.
224
+ .IP ""
225
+ La sortie exacte peut varier suivant les versions de \fBxz\fP et les différents
226
+ paramètres régionaux. Pour une sortie lisible par la machine, utiliser
227
+ \fB\-\-robot \-\-list\fP.
228
+ .
229
+ .SS "Modificateurs d'opération"
230
+ .TP
231
+ \fB\-k\fP, \fB\-\-keep\fP
232
+ Ne pas effacer les fichiers d'entrée.
233
+ .IP ""
234
+ Since \fBxz\fP 5.2.6, this option also makes \fBxz\fP compress or decompress even
235
+ if the input is a symbolic link to a regular file, has more than one hard
236
+ link, or has the setuid, setgid, or sticky bit set. The setuid, setgid, and
237
+ sticky bits are not copied to the target file. In earlier versions this was
238
+ only done with \fB\-\-force\fP.
239
+ .TP
240
+ \fB\-f\fP, \fB\-\-force\fP
241
+ Cette option a plusieurs effets :
242
+ .RS
243
+ .IP \(bu 3
244
+ Si le fichier cible existe déjà, l'effacer avant de compresser ou
245
+ décompresser.
246
+ .IP \(bu 3
247
+ Compresser ou décompresser même si l'entrée est un lien symbolique vers un
248
+ fichier normal, a plus qu'un lien physique, ou a le bit setuid, setgid ou
249
+ sticky défini. Les bits setuid, setgid et sticky bits ne sont pas copiés
250
+ dans le fichier cible.
251
+ .IP \(bu 3
252
+ Lorsque \fBxz\fP est utilisé avec \fB\-\-decompress\fP \fB\-\-stdout\fP et qu'il ne peut
253
+ pas reconnaitre le type du fichier source, copier le fichier source tel quel
254
+ dans la sortie standard. Celà permet à \fBxzcat\fP \fB\-\-force\fP d'être utilisé
255
+ comme \fBcat\fP(1) pour les fichiers qui n'ont pas été compressé avec
256
+ \fBxz\fP. Remarquez que dans le futur, \fBxz\fP devrait prendre en charge de
257
+ nouveaux formats de fichiers compressés, ce qui permettra à \fBxz\fP de
258
+ décompresser plus de types de fichiers au lieu de les copier tels quels dans
259
+ la sortie standard. \fB\-\-format=\fP\fIformat\fP peut être utilisé pour contraindre
260
+ \fBxz\fP à décompresser seulement un format de fichier.
261
+ .RE
262
+ .TP
263
+ \fB\-c\fP, \fB\-\-stdout\fP, \fB\-\-to\-stdout\fP
264
+ Écrire les données compressées ou décompressées sur la sortie standard
265
+ plutôt que dans un fichier. Cela necessite \fB\-\-keep\fP.
266
+ .TP
267
+ \fB\-\-single\-stream\fP
268
+ Décompresser seulement le premier flux \fB.xz\fP et ignorer silencieusement les
269
+ possibles données d'entrée résiduelles qui suivent le flux. Normalement ces
270
+ déchets excédentaires provoquent l'affichage d'une erreur par \fBxz\fP.
271
+ .IP ""
272
+ \fBxz\fP ne décompresse jamais plus d'un flux à partir de fichiers \fB.lzma\fP ou
273
+ de flux bruts, mais cette option fait aussi que \fBxz\fP ignorera les données
274
+ résiduelles après le fichier \fB.lzma\fP ou le flux brut.
275
+ .IP ""
276
+ Cette option n'a aucun effet si le mode d'opération n'est pas
277
+ \fB\-\-decompress\fP ou \fB\-\-test\fP.
278
+ .TP
279
+ \fB\-\-no\-sparse\fP
280
+ Désactiver la création de fichiers peu denses. Par défaut, lors de la
281
+ décompression en un fichier normal, \fBxz\fP essaie d'en faire un fichier creux
282
+ si les données décompressées contiennent de longues séquences de zéros
283
+ binaires. Cela fonctionne aussi lors de l'écriture sur la sortie standard
284
+ aussi longtemps que la sortie standard est connectée à un fichier normal et
285
+ que certaines conditions supplémentaires sont satisfaites pour le faire de
286
+ manière sécurisée. Créer des fichiers creux peut épargner de l'espace disque
287
+ et accélérer la décompression en réduisant la quantité d'entrées/sorties sur
288
+ le disque.
289
+ .TP
290
+ \fB\-S\fP \fI.suf\fP, \fB\-\-suffix=\fP\fI.suf\fP
291
+ Lors de la compression, utiliser \fB.suf\fP comme suffixe du fichier cible au
292
+ lieu de \fB.xz\fP ou \fB.lzma\fP. Si \fBxz\fP n'écrit pas sur la sortie standard et
293
+ si le fichier source a déja le suffixe \fB.suf\fP, un avertissement est affiché
294
+ et le fichier est ignoré.
295
+ .IP ""
296
+ When decompressing, recognize files with the suffix \fI.suf\fP in addition to
297
+ files with the \fB.xz\fP, \fB.txz\fP, \fB.lzma\fP, \fB.tlz\fP, or \fB.lz\fP suffix. If the
298
+ source file has the suffix \fI.suf\fP, the suffix is removed to get the target
299
+ filename.
300
+ .IP ""
301
+ Lors de la compression ou décompression de flux bruts (\fB\-\-fomat=raw\fP), le
302
+ suffixe doit toujours être spécifié à moins d'écrire sur la sortie standard,
303
+ car il n'y a pas de suffixe par défaut pour les flux bruts.
304
+ .TP
305
+ \fB\-\-files\fP[\fB=\fP\fIfichier\fP]
306
+ Lire les noms de fichier à traiter depuis \fIfichier\fP ; si \fIfichier\fP est
307
+ omis , les noms de fichier sont lus sur l'entrée standard. Les noms de
308
+ fichier doivent se terminer avec le caractère de nouvelle ligne. Un tiret
309
+ (\fB\-\fP) est considéré comme un nom de fichier normal ; ce qui ne signifie pas
310
+ entrée standard. Si les noms de fichier sont aussi donnés comme arguments de
311
+ ligne de commande, ils sont traités avant les noms de fichier lus depuis
312
+ \fIfichier\fP.
313
+ .TP
314
+ \fB\-\-files0\fP[\fB=\fP\fIfichier\fP]
315
+ Cela est identique à \fB\-\-files\fP[\fB=\fP\fIfichier\fP] sauf que chaque nom de
316
+ fichier doit se terminer par le caractère null.
317
+ .
318
+ .SS "Format de fichier basique et options de compression"
319
+ .TP
320
+ \fB\-F\fP \fIformat\fP, \fB\-\-format=\fP\fIformat\fP
321
+ Indiquer le \fIformat\fP de fichier à compresser ou décompresser :
322
+ .RS
323
+ .TP
324
+ \fBauto\fP
325
+ C'est celui par défaut. Lors de la compression, \fBauto\fP est équivalent à
326
+ \fBxz\fP. Lors de la décompression, le format du fichier en entrée est détecté
327
+ automatiquement. Notez que les flux bruts (créés avec \fB\-\-format=raw\fP) ne
328
+ peuvent pas être détectés automatiquement.
329
+ .TP
330
+ \fBxz\fP
331
+ Compresser dans le format de fichier \fB.xz\fP ou n'accepter que les fichiers
332
+ \&\fB.xz\fP à décompresser.
333
+ .TP
334
+ \fBlzma\fP, \fBalone\fP
335
+ Compresser au format de fichier \fB.lzma\fP historique, ou n'accepter que les
336
+ fichiers \fB.lzma\fP lors de la décompression. Le nom alternatif \fBalone\fP est
337
+ fourni pour la rétrocompatibilité avec les utilitaires LZMA.
338
+ .TP
339
+ \fBlzip\fP
340
+ Accept only \fB.lz\fP files when decompressing. Compression is not supported.
341
+ .IP ""
342
+ The \fB.lz\fP format version 0 and the unextended version 1 are supported.
343
+ Version 0 files were produced by \fBlzip\fP 1.3 and older. Such files aren't
344
+ common but may be found from file archives as a few source packages were
345
+ released in this format. People might have old personal files in this
346
+ format too. Decompression support for the format version 0 was removed in
347
+ \fBlzip\fP 1.18.
348
+ .IP ""
349
+ \fBlzip\fP 1.4 and later create files in the format version 1. The sync flush
350
+ marker extension to the format version 1 was added in \fBlzip\fP 1.6. This
351
+ extension is rarely used and isn't supported by \fBxz\fP (diagnosed as corrupt
352
+ input).
353
+ .TP
354
+ \fBraw\fP
355
+ Compresser ou décompresser un flux brut (sans en\-têtes). Cela est réservé
356
+ seulement aux utilisateurs aguerris. Pour décoder des flux bruts, vous devez
357
+ utiliser \fB\-\-format=raw\fP et spécifier explicitement la chaîne de filtre, qui
358
+ normalement aurait du être stockée dans les en\-têtes du conteneur.
359
+ .RE
360
+ .TP
361
+ \fB\-C\fP \fIvérif.\fP, \fB\-\-check=\fP\fIvérif.\fP
362
+ Spécifier le type d'intégrité à vérifier. La vérification est calculée à
363
+ partir des données non\-compressées et stockées dans le fichier \fB.xz\fP. Cette
364
+ option n'a effet que si la compression a été faite dans le format \fB.xz\fP ;
365
+ le format \fB.lzma\fP ne gère pas les vérifications d'intégrité. Le contrôle
366
+ d'intégrité (s'il y en a) est vérifié lorsque le fichier \fB.xz\fP est
367
+ décompressé.
368
+ .IP ""
369
+ Types de \fIvérification\fP pris en charge :
370
+ .RS
371
+ .TP
372
+ \fBnone\fP
373
+ Ne pas calculer de vérification d'intégrité du tout. C'est généralement une
374
+ mauvaise idée. Cela peut être utile lorsque l'intégrité des données est
375
+ vérifiée de toute façon par d'autres manières.
376
+ .TP
377
+ \fBcrc32\fP
378
+ Calculer CRC32 en utilisant le polynôme de IEEE\-802.3 (Ethernet).
379
+ .TP
380
+ \fBcrc64\fP
381
+ Calculer CRC64 en utilisant le polynôme de ECMA\-182. C'est la manière
382
+ utilisée par défaut, car c'est légèrement mieux que CRC32 pour détecter les
383
+ fichiers endommagés et la différence de vitesse est négligeable.
384
+ .TP
385
+ \fBsha256\fP
386
+ Calculer SHA\-256. C'est quelque peu plus lent que CRC32 et CRC64.
387
+ .RE
388
+ .IP ""
389
+ L'intégrité des en\-têtes \fB.xz\fP est toujours vérifiée avec CRC32. Il n'est
390
+ pas possible de le changer ou de le désactiver.
391
+ .TP
392
+ \fB\-\-ignore\-check\fP
393
+ Ne pas contrôler la vérification d'intégrité des données lors de la
394
+ décompression. Les valeurs CRC32 dans les en\-têtes \fB.xz\fP seront normalement
395
+ toujours vérifiées.
396
+ .IP ""
397
+ \fBN'utilisez pas cette option à moins de savoir ce que vous faites.\fP Les
398
+ raisons possibles pour utiliser cette option :
399
+ .RS
400
+ .IP \(bu 3
401
+ Essayer de récupérer des données d'un fichier .xz corrompu.
402
+ .IP \(bu 3
403
+ Accélérer la décompression. Cela importe surtout avec SHA\-256 ou avec les
404
+ fichiers qui ont été compressés extrêmement bien. Il est recommandé de ne
405
+ pas utiliser cette option dans ce but à moins que l'intégrité du fichier ne
406
+ soit vérifiée extérieurement d'une autre manière.
407
+ .RE
408
+ .TP
409
+ \fB\-0\fP ... \fB\-9\fP
410
+ Choisir un niveau de compression prédéfini. La valeur par défaut est
411
+ \fB6\fP. Si plusieurs niveaux de préréglage sont spécifi��s, c'est le dernier
412
+ qui sera pris en compte. Si une chaîne de filtres personnalisée a déjà été
413
+ choisie, définir un niveau de compression préréglé efface la chaîne de
414
+ filtres personnalisée.
415
+ .IP ""
416
+ Les différences entre les préréglages sont plus significatives qu'avec
417
+ \fBgzip\fP(1) et \fBbzip2\fP(1). les réglages de compression sélectionnés
418
+ déterminent les exigences en mémoire pour la décompression, ainsi, utiliser
419
+ un niveau de préréglage trop élevé peut rendre difficile à décompresser un
420
+ fichier sur un vieux système avec peu de RAM. Clairement, \fBce n'est pas une bonne idée d'utiliser \-9 aveuglément pour tout\fP comme ça l'est souvent avec
421
+ \fBgzip\fP(1) et \fBbzip2\fP(1).
422
+ .RS
423
+ .TP
424
+ \fB\-0\fP ... \fB\-3\fP
425
+ Ce sont des préréglages relativement rapides. \fB0\fP est parfois plus rapide
426
+ que \fBgzip \-9\fP tout en compressant bien mieux. Les réglages plus élevés ont
427
+ souvent une rapidité comparable à celle de \fBbzip2\fP(1) avec un taux de
428
+ compression comparable ou meilleur, même si les résultats dépendent beaucoup
429
+ du genre de données compressées.
430
+ .TP
431
+ \fB\-4\fP ... \fB\-6\fP
432
+ Good to very good compression while keeping decompressor memory usage
433
+ reasonable even for old systems. \fB\-6\fP is the default, which is usually a
434
+ good choice for distributing files that need to be decompressible even on
435
+ systems with only 16\ MiB RAM. (\fB\-5e\fP or \fB\-6e\fP may be worth considering
436
+ too. See \fB\-\-extreme\fP.)
437
+ .TP
438
+ \fB\-7 ... \-9\fP
439
+ C'est comme \fB\-6\fP mais avec des besoins en mémoire plus élevés pour la
440
+ compression et la décompression. Ce n'est utile que lorsque les fichiers
441
+ sont plus gros que 8\ Mio, 16\ Mio et 32\ Mio respectivement.
442
+ .RE
443
+ .IP ""
444
+ Sur le même matériel, la vitesse de décompression est sensiblement un nombre
445
+ constant d'octets de données compressées par seconde. En d'autres termes,
446
+ meilleure est la compression, plus rapide sera en général la
447
+ décompression. Cela signifie aussi que la quantité de sortie non compressée
448
+ produite par seconde peut varier beaucoup.
449
+ .IP ""
450
+ Le tableau suivant résume les caractéristiques des préréglages :
451
+ .RS
452
+ .RS
453
+ .PP
454
+ .TS
455
+ tab(;);
456
+ c c c c c
457
+ n n n n n.
458
+ Préréglage;DictSize;CompCPU;CompMem;DecMem
459
+ \-0;256 KiB;0;3 MiB;1 MiB
460
+ \-1;1 MiB;1;9 MiB;2 MiB
461
+ \-2;2 MiB;2;17 MiB;3 MiB
462
+ \-3;4 MiB;3;32 MiB;5 MiB
463
+ \-4;4 MiB;4;48 MiB;5 MiB
464
+ \-5;8 MiB;5;94 MiB;9 MiB
465
+ \-6;8 MiB;6;94 MiB;9 MiB
466
+ \-7;16 MiB;6;186 MiB;17 MiB
467
+ \-8;32 MiB;6;370 MiB;33 MiB
468
+ \-9;64 MiB;6;674 MiB;65 MiB
469
+ .TE
470
+ .RE
471
+ .RE
472
+ .IP ""
473
+ Descriptions des colonnes :
474
+ .RS
475
+ .IP \(bu 3
476
+ DictSize est la taille du dictionnaire de LZMA2. Utiliser un dictionnaire
477
+ plus gros que la taille du fichier non compressé est un gaspillage de
478
+ mémoire. C'est pourquoi il est bon d'éviter d'utiliser les préréglages de
479
+ \fB\-7\fP à \fB\-9\fP lorsqu'il n'y en a pas vraiment besoin. A \fB\-6\fP et plus bas,
480
+ la quantité de mémoire gaspillée est généralement assez basse pour ne pas
481
+ être un problème.
482
+ .IP \(bu 3
483
+ CompCPU est une représentation des préréglages de LZMA2 qui affectent la
484
+ vitesse de compression. La taille du dictionnaire aussi affecte la vitesse,
485
+ alors comme CompCPU est le même pour les niveaux de \fB\-6\fP à \fB\-9\fP, les plus
486
+ haut niveaux tendent à être un peu moins rapides. Pour être encore moins
487
+ rapide et du coup obtenir peut être une meilleure compression, consultez
488
+ \fB\-\-extreme\fP.
489
+ .IP \(bu 3
490
+ CompMem contains the compressor memory requirements in the single\-threaded
491
+ mode. It may vary slightly between \fBxz\fP versions.
492
+ .IP \(bu 3
493
+ DecMem contient les besoins en mémoire du décompresseur. Ce sont les
494
+ réglages de la compression qui déterminent les besoins en mémoire de la
495
+ décompression. L'exacte utilisation de la mémoire est légèrement supérieure
496
+ à la taille du dictionnaire LZMA2, mais les valeurs dans la table ont été
497
+ arrondies au prochain Mio supérieur.
498
+ .RE
499
+ .IP ""
500
+ Memory requirements of the multi\-threaded mode are significantly higher than
501
+ that of the single\-threaded mode. With the default value of
502
+ \fB\-\-block\-size\fP, each thread needs 3*3*DictSize plus CompMem or DecMem. For
503
+ example, four threads with preset \fB\-6\fP needs 660\(en670\ MiB of memory.
504
+ .TP
505
+ \fB\-e\fP, \fB\-\-extreme\fP
506
+ Utilisez un variant plus lent que les préréglages (\fB\-0\fP à \fB\-9\fP) pour
507
+ espérer avoir un taux de compression légèrement meilleur, mais en cas de
508
+ malchance cela peut être pire. L'utilisation mémoire du décompresseur n'est
509
+ pas affectée, mais l'utilisation mémoire du compresseur augmente un peu aux
510
+ niveaux de préréglages de \fB\-0\fP à \fB\-3\fP.
511
+ .IP ""
512
+ Depuis qu'il y a deux préréglages avec des tailles de dictionnaire de 4\ Mio
513
+ et 8 \Mio, les préréglages \fB\-3e\fP et \fB\-5e\fP utilisent des réglages
514
+ légèrement plus rapides que \fB\-4e\fP et \fB\-6e\fP, respectivement. De cette
515
+ manière, il n'y a pas deux préréglages identiques.
516
+ .RS
517
+ .RS
518
+ .PP
519
+ .TS
520
+ tab(;);
521
+ c c c c c
522
+ n n n n n.
523
+ Préréglage;DictSize;CompCPU;CompMem;DecMem
524
+ \-0e;256 KiB;8;4 MiB;1 MiB
525
+ \-1e;1 MiB;8;13 MiB;2 MiB
526
+ \-2e;2 MiB;8;25 MiB;3 MiB
527
+ \-3e;4 MiB;7;48 MiB;5 MiB
528
+ \-4e;4 MiB;8;48 MiB;5 MiB
529
+ \-5e;8 MiB;7;94 MiB;9 MiB
530
+ \-6e;8 MiB;8;94 MiB;9 MiB
531
+ \-7e;16 MiB;8;186 MiB;17 MiB
532
+ \-8e;32 MiB;8;370 MiB;33 MiB
533
+ \-9e;64 MiB;8;674 MiB;65 MiB
534
+ .TE
535
+ .RE
536
+ .RE
537
+ .IP ""
538
+ Par exemple, il y a un total de quatre préréglages qui utilisent un
539
+ dictionnaire de 8 Mio et qui sont dans l'ordre du plus rapide au plus lent :
540
+ \fB\-5\fP, \fB\-6\fP, \fB\-5e\fP et \fB\-6e\fP.
541
+ .TP
542
+ \fB\-\-fast\fP
543
+ .PD 0
544
+ .TP
545
+ \fB\-\-best\fP
546
+ .PD
547
+ Il y a néanmoins des alias trompeurs pour \fB\-0\fP et \fB\-9\fP,
548
+ respectivement. Ils ne sont fournis que pour des besoins de
549
+ rétro\-compatibilité avec les utilitaires LZMA. Evitez d'utiliser ces
550
+ options.
551
+ .TP
552
+ \fB\-\-block\-size=\fP\fItaille\fP
553
+ Lors de la compression dans le format \fB.xz\fP, les données de l'entrée sont
554
+ réparties en blocs de \fItaille\fP octets. Les blocs sont compressés
555
+ indépendamment les un des autres, ce qui aide avec le mode multithread
556
+ (multi\-threading) et rend possible la décompression à accès aléatoire
557
+ limité. Cette option est typiquement utilisée pour outrepasser la taille de
558
+ bloc en mode multithread, mais cette option peut aussi être utilisée en mode
559
+ single\-thread.
560
+ .IP ""
561
+ In multi\-threaded mode about three times \fIsize\fP bytes will be allocated in
562
+ each thread for buffering input and output. The default \fIsize\fP is three
563
+ times the LZMA2 dictionary size or 1 MiB, whichever is more. Typically a
564
+ good value is 2\(en4 times the size of the LZMA2 dictionary or at least 1
565
+ MiB. Using \fIsize\fP less than the LZMA2 dictionary size is waste of RAM
566
+ because then the LZMA2 dictionary buffer will never get fully used. In
567
+ multi\-threaded mode, the sizes of the blocks are stored in the block
568
+ headers. This size information is required for multi\-threaded
569
+ decompression.
570
+ .IP ""
571
+ In single\-threaded mode no block splitting is done by default. Setting this
572
+ option doesn't affect memory usage. No size information is stored in block
573
+ headers, thus files created in single\-threaded mode won't be identical to
574
+ files created in multi\-threaded mode. The lack of size information also
575
+ means that \fBxz\fP won't be able decompress the files in multi\-threaded mode.
576
+ .TP
577
+ \fB\-\-block\-list=\fP\fIitems\fP
578
+ When compressing to the \fB.xz\fP format, start a new block with an optional
579
+ custom filter chain after the given intervals of uncompressed data.
580
+ .IP ""
581
+ The \fIitems\fP are a comma\-separated list. Each item consists of an optional
582
+ filter chain number between 0 and 9 followed by a colon (\fB:\fP) and a
583
+ required size of uncompressed data. Omitting an item (two or more
584
+ consecutive commas) is a shorthand to use the size and filters of the
585
+ previous item.
586
+ .IP ""
587
+ If the input file is bigger than the sum of the sizes in \fIitems\fP, the last
588
+ item is repeated until the end of the file. A special value of \fB0\fP may be
589
+ used as the last size to indicate that the rest of the file should be
590
+ encoded as a single block.
591
+ .IP ""
592
+ An alternative filter chain for each block can be specified in combination
593
+ with the \fB\-\-filters1=\fP\fIfilters\fP \&...\& \fB\-\-filters9=\fP\fIfilters\fP options.
594
+ These options define filter chains with an identifier between 1\(en9.
595
+ Filter chain 0 can be used to refer to the default filter chain, which is
596
+ the same as not specifying a filter chain. The filter chain identifier can
597
+ be used before the uncompressed size, followed by a colon (\fB:\fP). For
598
+ example, if one specifies \fB\-\-block\-list=1:2MiB,3:2MiB,2:4MiB,,2MiB,0:4MiB\fP
599
+ then blocks will be created using:
600
+ .RS
601
+ .IP \(bu 3
602
+ The filter chain specified by \fB\-\-filters1\fP and 2 MiB input
603
+ .IP \(bu 3
604
+ The filter chain specified by \fB\-\-filters3\fP and 2 MiB input
605
+ .IP \(bu 3
606
+ The filter chain specified by \fB\-\-filters2\fP and 4 MiB input
607
+ .IP \(bu 3
608
+ The filter chain specified by \fB\-\-filters2\fP and 4 MiB input
609
+ .IP \(bu 3
610
+ The default filter chain and 2 MiB input
611
+ .IP \(bu 3
612
+ The default filter chain and 4 MiB input for every block until end of input.
613
+ .RE
614
+ .IP ""
615
+ If one specifies a size that exceeds the encoder's block size (either the
616
+ default value in threaded mode or the value specified with
617
+ \fB\-\-block\-size=\fP\fIsize\fP), the encoder will create additional blocks while
618
+ keeping the boundaries specified in \fIitems\fP. For example, if one specifies
619
+ \fB\-\-block\-size=10MiB\fP \fB\-\-block\-list=5MiB,10MiB,8MiB,12MiB,24MiB\fP and the
620
+ input file is 80 MiB, one will get 11 blocks: 5, 10, 8, 10, 2, 10, 10, 4,
621
+ 10, 10, and 1 MiB.
622
+ .IP ""
623
+ En mode multi\-threadé les tailles de blocs sont stockées dans les en\-têtes
624
+ du bloc. Cela ne se fait pas en mode mono\-threadé, la sortie encodée ne sera
625
+ donc pas identique à celle faite en mode multi\-threadé.
626
+ .TP
627
+ \fB\-\-flush\-timeout=\fP\fItemps_d'attente\fP
628
+ Lors de la compression, si plus que \fItemps_d'attente\fP millisecondes (un
629
+ entier positif) se sont écoulées depuis le précédent vidage et que lire plus
630
+ de données bloquerait, toutes les données d'entrée en attente sont vidées de
631
+ l'encodeur et mises à disposition dans le flux de sortie. Cela peut être
632
+ utile si \fBxz\fP est utilisé pour compresser les données qui sont diffusées
633
+ sur un réseau. Des petites valeurs de \fItemps_d'attente\fP rendent les données
634
+ disponibles à l'extrémité réceptrice avec un léger retard, mais les grandes
635
+ valeurs de \fItemps_d'attente\fP donnent un meilleur taux de compression.
636
+ .IP ""
637
+ Cette option est désactivée par défaut. Si cette option est indiquée plus
638
+ d'une fois, la dernière prend effet. La valeur spéciale de
639
+ \fItemps_d'attente\fP de \fB0\fP peut être utilisée pour explicitement désactiver
640
+ cette option.
641
+ .IP ""
642
+ Cette option n'est pas disponible sur les systèmes qui ne sont pas POSIX.
643
+ .IP ""
644
+ .\" FIXME
645
+ \fBCette option est encore expérimentale.\fP Actuellement, \fBxz\fP ne convient
646
+ pas pour décompresser le flux en temps réel en raison de la façon dont \fBxz\fP
647
+ effectue la mise en mémoire tampon.
648
+ .TP
649
+ \fB\-\-memlimit\-compress=\fP\fIlimite\fP
650
+ Indiquer une limite d'utilisation de la mémoire pour la compression. Si
651
+ cette option est indiquée plusieurs fois, c'est la dernière qui est prise en
652
+ compte.
653
+ .IP ""
654
+ If the compression settings exceed the \fIlimit\fP, \fBxz\fP will attempt to
655
+ adjust the settings downwards so that the limit is no longer exceeded and
656
+ display a notice that automatic adjustment was done. The adjustments are
657
+ done in this order: reducing the number of threads, switching to
658
+ single\-threaded mode if even one thread in multi\-threaded mode exceeds the
659
+ \fIlimit\fP, and finally reducing the LZMA2 dictionary size.
660
+ .IP ""
661
+ When compressing with \fB\-\-format=raw\fP or if \fB\-\-no\-adjust\fP has been
662
+ specified, only the number of threads may be reduced since it can be done
663
+ without affecting the compressed output.
664
+ .IP ""
665
+ If the \fIlimit\fP cannot be met even with the adjustments described above, an
666
+ error is displayed and \fBxz\fP will exit with exit status 1.
667
+ .IP ""
668
+ La \fIlimite\fP peut être indiquée de plusieurs façons :
669
+ .RS
670
+ .IP \(bu 3
671
+ La \fIlimite\fP peut être une valeur absolue en octets. Utiliser un suffixe
672
+ d'entier comme \fBMiB\fP peut être utile. Exemple :
673
+ \fB\-\-memlimit\-compress=80MiB\fP
674
+ .IP \(bu 3
675
+ La \fIlimite\fP peut être indiquée sous forme d'un pourcentage de la mémoire
676
+ physique totale (RAM). Cela peut être particulièrement utile quand la
677
+ variable d'environnement \fBXZ_DEFAULTS\fP est indiquée dans un script
678
+ d'initialisation de l'interpréteur partagé entre différents ordinateurs. De
679
+ cette façon la limite est automatiquement plus grande sur les systèmes avec
680
+ plus de mémoire. Exemple : \fB\-\-memlimit=70%\fP
681
+ .IP \(bu 3
682
+ The \fIlimit\fP can be reset back to its default value by setting it to \fB0\fP.
683
+ This is currently equivalent to setting the \fIlimit\fP to \fBmax\fP (no memory
684
+ usage limit).
685
+ .RE
686
+ .IP ""
687
+ For 32\-bit \fBxz\fP there is a special case: if the \fIlimit\fP would be over
688
+ \fB4020\ MiB\fP, the \fIlimit\fP is set to \fB4020\ MiB\fP. On MIPS32 \fB2000\ MiB\fP
689
+ is used instead. (The values \fB0\fP and \fBmax\fP aren't affected by this. A
690
+ similar feature doesn't exist for decompression.) This can be helpful when
691
+ a 32\-bit executable has access to 4\ GiB address space (2 GiB on MIPS32)
692
+ while hopefully doing no harm in other situations.
693
+ .IP ""
694
+ Voir aussi la section \fButilisation de la mémoire\fP.
695
+ .TP
696
+ \fB\-\-memlimit\-decompress=\fP\fIlimite\fP
697
+ Régler une limite d'utilisation de la mémoire pour la décompression. Cela a
698
+ un effet sur le mode \fB\-\-list\fP. Si l'opération n'est pas possible sans
699
+ dépasser la \fIlimite\fP, \fBxz\fP affichera une erreur et la décompression
700
+ échouera. Voir \fB\-\-memlimit\-compress=\fP\fIlimite\fP pour les manières possibles
701
+ d'indiquer la \fIlimite\fP.
702
+ .TP
703
+ \fB\-\-memlimit\-mt\-decompress=\fP\fIlimit\fP
704
+ Set a memory usage limit for multi\-threaded decompression. This can only
705
+ affect the number of threads; this will never make \fBxz\fP refuse to
706
+ decompress a file. If \fIlimit\fP is too low to allow any multi\-threading, the
707
+ \fIlimit\fP is ignored and \fBxz\fP will continue in single\-threaded mode. Note
708
+ that if also \fB\-\-memlimit\-decompress\fP is used, it will always apply to both
709
+ single\-threaded and multi\-threaded modes, and so the effective \fIlimit\fP for
710
+ multi\-threading will never be higher than the limit set with
711
+ \fB\-\-memlimit\-decompress\fP.
712
+ .IP ""
713
+ In contrast to the other memory usage limit options,
714
+ \fB\-\-memlimit\-mt\-decompress=\fP\fIlimit\fP has a system\-specific default
715
+ \fIlimit\fP. \fBxz \-\-info\-memory\fP can be used to see the current value.
716
+ .IP ""
717
+ This option and its default value exist because without any limit the
718
+ threaded decompressor could end up allocating an insane amount of memory
719
+ with some input files. If the default \fIlimit\fP is too low on your system,
720
+ feel free to increase the \fIlimit\fP but never set it to a value larger than
721
+ the amount of usable RAM as with appropriate input files \fBxz\fP will attempt
722
+ to use that amount of memory even with a low number of threads. Running out
723
+ of memory or swapping will not improve decompression performance.
724
+ .IP ""
725
+ See \fB\-\-memlimit\-compress=\fP\fIlimit\fP for possible ways to specify the
726
+ \fIlimit\fP. Setting \fIlimit\fP to \fB0\fP resets the \fIlimit\fP to the default
727
+ system\-specific value.
728
+ .TP
729
+ \fB\-M\fP \fIlimite\fP, \fB\-\-memlimit=\fP\fIlimite\fP, \fB\-\-memory=\fP\fIlimite\fP
730
+ This is equivalent to specifying \fB\-\-memlimit\-compress=\fP\fIlimit\fP
731
+ \fB\-\-memlimit\-decompress=\fP\fIlimit\fP \fB\-\-memlimit\-mt\-decompress=\fP\fIlimit\fP.
732
+ .TP
733
+ \fB\-\-no\-adjust\fP
734
+ Display an error and exit if the memory usage limit cannot be met without
735
+ adjusting settings that affect the compressed output. That is, this
736
+ prevents \fBxz\fP from switching the encoder from multi\-threaded mode to
737
+ single\-threaded mode and from reducing the LZMA2 dictionary size. Even when
738
+ this option is used the number of threads may be reduced to meet the memory
739
+ usage limit as that won't affect the compressed output.
740
+ .IP ""
741
+ Automatic adjusting is always disabled when creating raw streams
742
+ (\fB\-\-format=raw\fP).
743
+ .TP
744
+ \fB\-T\fP \fIthreads\fP, \fB\-\-threads=\fP\fIthreads\fP
745
+ Specify the number of worker threads to use. Setting \fIthreads\fP to a
746
+ special value \fB0\fP makes \fBxz\fP use up to as many threads as the processor(s)
747
+ on the system support. The actual number of threads can be fewer than
748
+ \fIthreads\fP if the input file is not big enough for threading with the given
749
+ settings or if using more threads would exceed the memory usage limit.
750
+ .IP ""
751
+ The single\-threaded and multi\-threaded compressors produce different
752
+ output. Single\-threaded compressor will give the smallest file size but
753
+ only the output from the multi\-threaded compressor can be decompressed using
754
+ multiple threads. Setting \fIthreads\fP to \fB1\fP will use the single\-threaded
755
+ mode. Setting \fIthreads\fP to any other value, including \fB0\fP, will use the
756
+ multi\-threaded compressor even if the system supports only one hardware
757
+ thread. (\fBxz\fP 5.2.x used single\-threaded mode in this situation.)
758
+ .IP ""
759
+ To use multi\-threaded mode with only one thread, set \fIthreads\fP to \fB+1\fP.
760
+ The \fB+\fP prefix has no effect with values other than \fB1\fP. A memory usage
761
+ limit can still make \fBxz\fP switch to single\-threaded mode unless
762
+ \fB\-\-no\-adjust\fP is used. Support for the \fB+\fP prefix was added in \fBxz\fP
763
+ 5.4.0.
764
+ .IP ""
765
+ If an automatic number of threads has been requested and no memory usage
766
+ limit has been specified, then a system\-specific default soft limit will be
767
+ used to possibly limit the number of threads. It is a soft limit in sense
768
+ that it is ignored if the number of threads becomes one, thus a soft limit
769
+ will never stop \fBxz\fP from compressing or decompressing. This default soft
770
+ limit will not make \fBxz\fP switch from multi\-threaded mode to single\-threaded
771
+ mode. The active limits can be seen with \fBxz \-\-info\-memory\fP.
772
+ .IP ""
773
+ Actuellement, la seule méthode de gestion avec des threads consiste à
774
+ séparer l'entrée en blocs et de les compresser indépendamment les uns des
775
+ autres. La taille par défaut des blocs dépend du niveau de compression et
776
+ peut\-être remplacée avec l'option \fB\-\-block\-size=\fP\fItaille\fP.
777
+ .IP ""
778
+ Threaded decompression only works on files that contain multiple blocks with
779
+ size information in block headers. All large enough files compressed in
780
+ multi\-threaded mode meet this condition, but files compressed in
781
+ single\-threaded mode don't even if \fB\-\-block\-size=\fP\fIsize\fP has been used.
782
+ .IP ""
783
+ The default value for \fIthreads\fP is \fB0\fP. In \fBxz\fP 5.4.x and older the
784
+ default is \fB1\fP.
785
+ .
786
+ .SS "Chaînes de filtres de compresseur personnalisées"
787
+ A custom filter chain allows specifying the compression settings in detail
788
+ instead of relying on the settings associated to the presets. When a custom
789
+ filter chain is specified, preset options (\fB\-0\fP \&...\& \fB\-9\fP and
790
+ \fB\-\-extreme\fP) earlier on the command line are forgotten. If a preset
791
+ option is specified after one or more custom filter chain options, the new
792
+ preset takes effect and the custom filter chain options specified earlier
793
+ are forgotten.
794
+ .PP
795
+ Une chaîne de filtre est comparable à une redirection (pipe) sur la ligne de
796
+ commande. Lors de la compression, les entrées non compressées vont au
797
+ premier filtre, dont la sortie va au prochain filtre (s'il y en a). La
798
+ sortie du dernier filtre est écrite sur le fichier compressé. Le nombre
799
+ maximal de filtres dans la chaîne est quatre, mais habituellement, un chaîne
800
+ de filtre n'a qu'un ou deux filtres.
801
+ .PP
802
+ Beaucoup de filtres ont des limitations sur l'endroit où ils peuvent se
803
+ placer dans la chaîne de filtre : quelques filtres ne peuvent fonctionner
804
+ qu'en tant que dernier filtre dans la chaîne, quelques uns en tant que non
805
+ dernier filtre, et d'autres à n'importe quelle position dans la
806
+ chaîne. Suivant le filtre, cette limitation est soit inhérente au profil du
807
+ filtre, soit existe pour des raisons de sécurité.
808
+ .PP
809
+ A custom filter chain can be specified in two different ways. The options
810
+ \fB\-\-filters=\fP\fIfilters\fP and \fB\-\-filters1=\fP\fIfilters\fP \&...\&
811
+ \fB\-\-filters9=\fP\fIfilters\fP allow specifying an entire filter chain in one
812
+ option using the liblzma filter string syntax. Alternatively, a filter
813
+ chain can be specified by using one or more individual filter options in the
814
+ order they are wanted in the filter chain. That is, the order of the
815
+ individual filter options is significant! When decoding raw streams
816
+ (\fB\-\-format=raw\fP), the filter chain must be specified in the same order as
817
+ it was specified when compressing. Any individual filter or preset options
818
+ specified before the full chain option (\fB\-\-filters=\fP\fIfilters\fP) will be
819
+ forgotten. Individual filters specified after the full chain option will
820
+ reset the filter chain.
821
+ .PP
822
+ Both the full and individual filter options take filter\-specific \fIoptions\fP
823
+ as a comma\-separated list. Extra commas in \fIoptions\fP are ignored. Every
824
+ option has a default value, so specify those you want to change.
825
+ .PP
826
+ Pour voir l'entièreté de la chaîne de filtres et ses \fIoptions\fP, utilisez
827
+ \fBxz \-vv\fP (ce qui est comme utiliser \fB\-\-verbose\fP deux fois). Cela
828
+ fonctionne aussi pour voir les options de chaîne de filtres utilisées par
829
+ les préréglages.
830
+ .TP
831
+ \fB\-\-filters=\fP\fIfilters\fP
832
+ Specify the full filter chain or a preset in a single option. Each filter
833
+ can be separated by spaces or two dashes (\fB\-\-\fP). \fIfilters\fP may need to be
834
+ quoted on the shell command line so it is parsed as a single option. To
835
+ denote \fIoptions\fP, use \fB:\fP or \fB=\fP. A preset can be prefixed with a \fB\-\fP
836
+ and followed with zero or more flags. The only supported flag is \fBe\fP to
837
+ apply the same options as \fB\-\-extreme\fP.
838
+ .TP
839
+ \fB\-\-filters1\fP=\fIfilters\fP ... \fB\-\-filters9\fP=\fIfilters\fP
840
+ Specify up to nine additional filter chains that can be used with
841
+ \fB\-\-block\-list\fP.
842
+ .IP ""
843
+ For example, when compressing an archive with executable files followed by
844
+ text files, the executable part could use a filter chain with a BCJ filter
845
+ and the text part only the LZMA2 filter.
846
+ .TP
847
+ \fB\-\-filters\-help\fP
848
+ Display a help message describing how to specify presets and custom filter
849
+ chains in the \fB\-\-filters\fP and \fB\-\-filters1=\fP\fIfilters\fP \&...\&
850
+ \fB\-\-filters9=\fP\fIfilters\fP options, and exit successfully.
851
+ .TP
852
+ \fB\-\-lzma1\fP[\fB=\fP\fIoptions\fP]
853
+ .PD 0
854
+ .TP
855
+ \fB\-\-lzma2\fP[\fB=\fP\fIoptions\fP]
856
+ .PD
857
+ Ajouter le filtre LZMA1 ou LZMA2 à la chaîne de filtres. Ces filtres ne
858
+ peuvent être utilisés que comme dernier filtre dans la chaîne.
859
+ .IP ""
860
+ LZMA1 est un filtre historique, qui n'est pris en charge presque uniquement
861
+ à cause de l'ancien format de fichier \fB.lzma\fP, qui ne prend en charge que
862
+ LZMA1. LZMA2 est une version mise à jour de LZMA1 pour régler certains
863
+ problèmes pratiques de LZMA1. Le format \fBxz\fP utilise LZMA2 et ne prend pas
864
+ du tout en charge LZMA1. Les taux et vitesses de compression de LZMA1 et
865
+ LZMA2 sont pratiquement identiques.
866
+ .IP ""
867
+ LZMA1 et LZMA2 partagent le même ensemble d'\fIoptions\fP :
868
+ .RS
869
+ .TP
870
+ \fBpreset=\fP\fIpréréglage\fP
871
+ Reset all LZMA1 or LZMA2 \fIoptions\fP to \fIpreset\fP. \fIPreset\fP consist of an
872
+ integer, which may be followed by single\-letter preset modifiers. The
873
+ integer can be from \fB0\fP to \fB9\fP, matching the command line options \fB\-0\fP
874
+ \&...\& \fB\-9\fP. The only supported modifier is currently \fBe\fP, which matches
875
+ \fB\-\-extreme\fP. If no \fBpreset\fP is specified, the default values of LZMA1 or
876
+ LZMA2 \fIoptions\fP are taken from the preset \fB6\fP.
877
+ .TP
878
+ \fBdict=\fP\fItaille\fP
879
+ La \fItaille\fP du dictionnaire (historique du tampon) indique combien d'octets
880
+ des données récement décompressées sont gardés en mémoire. L'algorithme
881
+ essaie de trouver les séquences d'octets répétées (identiques) dans les
882
+ données décompressées et les remplace par les données actuellement dans le
883
+ dictionnaire. Plus gros est le dictionnaire, plus grande est la chance de
884
+ trouver une correspondance. Ainsi, l'augmentation de la \fItaille\fP du
885
+ dictionnaire augmente habituellement le taux de compression, mais un
886
+ dictionnaire plus gros que le fichier non compressé est un gachis de
887
+ mémoire.
888
+ .IP ""
889
+ Généralement la \fItaille\fP du dictionnaire est entre 64\ Kio et 64\ Mio. Le
890
+ minimum étant 4\ Kio. La \fItaille\fP maximale pour la compression est
891
+ habituellement 1,5\ Gio (1536\ Mio). Le décompresseur prend en charge les
892
+ dictionnaires jusqu'à un octet de moins que 4\ Gio, ce qui est le maximum
893
+ pour les formats de flux LZMA1 et LZMA2.
894
+ .IP ""
895
+ La \fItaille\fP du dictionnaire et le chercheur de correspondance (match
896
+ finder) (\fImf\fP) déterminent ensemble l'utilisation de la mémoire de
897
+ l'encodeur LZMA1 ou LZMA2. La même (ou une plus grosse) \fItaille\fP de
898
+ dictionnaire est requise pour décompresser que ce qui a été utilisé pour la
899
+ compression, ainsi l'utilisation de la mémoire du décodeur est déterminée
900
+ par la taille du dictionnaire utilisée lors de la compression. Les en\-têtes
901
+ de \fB.xz\fP stockent la \fItaille\fP de dictionnaire sous la forme 2^\fIn\fP ou
902
+ 2^\fIn\fP + 2^(\fIn\fP\-1), de sorte que ces \fItailles\fP sont quelque peu préférées
903
+ pour la compression. Les autres \fItailles\fP seront arrondies à la hausse
904
+ lorsque stockées dans les en\-têtes de \fB.xz\fP.
905
+ .TP
906
+ \fBlc=\fP\fIlc\fP
907
+ Spécifiez le nombre d'octets de contexte littéraux. Le minimum est \fB0\fP et
908
+ le maximum est \fB4\fP. La valeur par défaut est \fB3\fP. En plus, la somme de
909
+ \fIlc\fP et \fIlp\fP ne doit pas excéder \fB4\fP.
910
+ .IP ""
911
+ Tous les octets qui ne peuvent pas être codés comme des correspondances sont
912
+ codés comme des littéraux. C'est à dire que les littéraux sont simplement
913
+ des octets 8 bits encodés un à la fois.
914
+ .IP ""
915
+ The literal coding makes an assumption that the highest \fIlc\fP bits of the
916
+ previous uncompressed byte correlate with the next byte. For example, in
917
+ typical English text, an upper\-case letter is often followed by a lower\-case
918
+ letter, and a lower\-case letter is usually followed by another lower\-case
919
+ letter. In the US\-ASCII character set, the highest three bits are 010 for
920
+ upper\-case letters and 011 for lower\-case letters. When \fIlc\fP is at least
921
+ 3, the literal coding can take advantage of this property in the
922
+ uncompressed data.
923
+ .IP ""
924
+ The default value (3) is usually good. If you want maximum compression,
925
+ test \fBlc=4\fP. Sometimes it helps a little, and sometimes it makes
926
+ compression worse. If it makes it worse, test \fBlc=2\fP too.
927
+ .TP
928
+ \fBlp=\fP\fIlp\fP
929
+ Indiquer le nombre de bits de position littérale. Le minimum est \fB0\fP et le
930
+ maximum \fB4\fP; par défaut c'est \fB0\fP.
931
+ .IP ""
932
+ \fILp\fP affecte le type d'alignement dans les données décompressées qui est
933
+ présumé lors de l'encodage des littéraux. Voir \fIpb\fP ci dessous pour plus
934
+ d'information sur l'alignement.
935
+ .TP
936
+ \fBpb=\fP\fIpb\fP
937
+ Indiquer le nombre de bits de position. Le minimum est \fB0\fP et le maximum
938
+ \fB4\fP; par défaut \fB2\fP.
939
+ .IP ""
940
+ \fIPb\fP affecte quel genre d'alignement est présumé en général dans les
941
+ données non compressées. Par défaut c'est un alignement de quatre octets
942
+ (2^\fIpb\fP=2^2=4), ce qui est généralement un bon choix lorsqu'il n'y a pas de
943
+ meilleure estimation.
944
+ .IP ""
945
+ When the alignment is known, setting \fIpb\fP accordingly may reduce the file
946
+ size a little. For example, with text files having one\-byte alignment
947
+ (US\-ASCII, ISO\-8859\-*, UTF\-8), setting \fBpb=0\fP can improve compression
948
+ slightly. For UTF\-16 text, \fBpb=1\fP is a good choice. If the alignment is
949
+ an odd number like 3 bytes, \fBpb=0\fP might be the best choice.
950
+ .IP ""
951
+ Même si l'alignement présumé peut être ajusté avec \fIpb\fP et \fIlp\fP, LZMA1 et
952
+ LZMA2 favorisent toujours légèrement l'alignement sur 16 octets. Il peut
953
+ être utile d'en tenir compte lors de la conception de formats de fichiers
954
+ susceptibles d'être souvent compressés avec LZMA1 ou LZMA2.
955
+ .TP
956
+ \fBmf=\fP\fImf\fP
957
+ Match finder has a major effect on encoder speed, memory usage, and
958
+ compression ratio. Usually Hash Chain match finders are faster than Binary
959
+ Tree match finders. The default depends on the \fIpreset\fP: 0 uses \fBhc3\fP,
960
+ 1\(en3 use \fBhc4\fP, and the rest use \fBbt4\fP.
961
+ .IP ""
962
+ Les chercheurs de correspondance suivants sont pris en charge. Les formules
963
+ d'utilisation de la mémoire ci\-dessous sont des approximations grossières
964
+ qui sont les plus proches de la réalité lorsque \fIdict\fP est une puissance de
965
+ deux.
966
+ .RS
967
+ .TP
968
+ \fBhc3\fP
969
+ Chaîne de hachage avec hachage de 2 et 3 octets
970
+ .br
971
+ Valeur minimale pour \fInice\fP : \fB3\fP
972
+ .br
973
+ Utilisation de la mémoire :
974
+ .br
975
+ \fIdict\fP * 7.5 (if \fIdict\fP <= 16 Mio);
976
+ .br
977
+ \fIdict\fP * 5.5 + 64 MiB (si \fIdict\fP > 16 Mio)
978
+ .TP
979
+ \fBhc4\fP
980
+ Chaîne de hachage avec hachage de 2, 3 et 4 octets
981
+ .br
982
+ Valeur minimale pour \fInice\fP : \fB4\fP
983
+ .br
984
+ Utilisation de la mémoire :
985
+ .br
986
+ \fIdict\fP * 7.5 (si \fIdict\fP <= 32 Mio);
987
+ .br
988
+ \fIdict\fP * 6.5 (si \fIdict\fP > 32 Mio)
989
+ .TP
990
+ \fBbt2\fP
991
+ Arbre binaire avec hachage de 2 octets
992
+ .br
993
+ Valeur minimale pour \fInice\fP : \fB2\fP
994
+ .br
995
+ Utilisation de la mémoire : \fIdict\fP * 9.5
996
+ .TP
997
+ \fBbt3\fP
998
+ Arbre binaire avec hachage de 2 et 3 octets
999
+ .br
1000
+ Valeur minimale pour \fInice\fP : \fB3\fP
1001
+ .br
1002
+ Utilisation de la mémoire :
1003
+ .br
1004
+ \fIdict\fP * 11.5 (si \fIdict\fP <= 16 Mio);
1005
+ .br
1006
+ \fIdict\fP * 9.5 + 64 MiB (si \fIdict\fP > 16 Mio)
1007
+ .TP
1008
+ \fBbt4\fP
1009
+ Arbre binaire avec hachage 2, 3 et 4 octets
1010
+ .br
1011
+ Valeur minimale pour \fInice\fP : \fB4\fP
1012
+ .br
1013
+ Utilisation de la mémoire :
1014
+ .br
1015
+ \fIdict\fP * 11.5 (si \fIdict\fP <= 32 Mio);
1016
+ .br
1017
+ \fIdict\fP * 10.5 (si \fIdict\fP > 32 Mio)
1018
+ .RE
1019
+ .TP
1020
+ \fBmode=\fP\fImode\fP
1021
+ Compression \fImode\fP specifies the method to analyze the data produced by the
1022
+ match finder. Supported \fImodes\fP are \fBfast\fP and \fBnormal\fP. The default is
1023
+ \fBfast\fP for \fIpresets\fP 0\(en3 and \fBnormal\fP for \fIpresets\fP 4\(en9.
1024
+ .IP ""
1025
+ Habituellement, \fBfast\fP est utilisé avec les chercheurs de correspondance de
1026
+ chaîne de hachage et \fBnormal\fP avec les chercheurs de correspondance d'arbre
1027
+ binaire. C'est aussi ce que font les \fIpréréglages\fP.
1028
+ .TP
1029
+ \fBnice=\fP\fInice\fP
1030
+ Spécifier ce qui est considéré comme une bonne longueur pour une
1031
+ correspondance. Une fois que la correspondance d'au moins \fInice\fP octets est
1032
+ trouvée, l'algorithme arrête de chercher de meilleures correspondances
1033
+ possibles.
1034
+ .IP ""
1035
+ \fINice\fP can be 2\(en273 bytes. Higher values tend to give better
1036
+ compression ratio at the expense of speed. The default depends on the
1037
+ \fIpreset\fP.
1038
+ .TP
1039
+ \fBdepth=\fP\fIprofondeur\fP
1040
+ Spécifier la profondeur de recherche maximale dans l'outil de recherche de
1041
+ correspondances. La valeur par défaut est \fB0\fP, ce qui fait que le
1042
+ compresseur détermine une \fIprofondeur\fP raisonnable en fonction de \fImf\fP et
1043
+ \fInice\fP.
1044
+ .IP ""
1045
+ Reasonable \fIdepth\fP for Hash Chains is 4\(en100 and 16\(en1000 for Binary
1046
+ Trees. Using very high values for \fIdepth\fP can make the encoder extremely
1047
+ slow with some files. Avoid setting the \fIdepth\fP over 1000 unless you are
1048
+ prepared to interrupt the compression in case it is taking far too long.
1049
+ .RE
1050
+ .IP ""
1051
+ Lors du décodage des flux bruts (\fB\-\-format=raw\fP), LZMA2 nécessite seulement
1052
+ la \fItaille\fP du dictionnaire. LZMA1 nécessite aussi \fIlc\fP, \fIlp\fP et \fIpb\fP.
1053
+ .TP
1054
+ \fB\-\-x86\fP[\fB=\fP\fIoptions\fP]
1055
+ .PD 0
1056
+ .TP
1057
+ \fB\-\-arm\fP[\fB=\fP\fIoptions\fP]
1058
+ .TP
1059
+ \fB\-\-armthumb\fP[\fB=\fP\fIoptions\fP]
1060
+ .TP
1061
+ \fB\-\-arm64\fP[\fB=\fP\fIoptions\fP]
1062
+ .TP
1063
+ \fB\-\-powerpc\fP[\fB=\fP\fIoptions\fP]
1064
+ .TP
1065
+ \fB\-\-ia64\fP[\fB=\fP\fIoptions\fP]
1066
+ .TP
1067
+ \fB\-\-sparc\fP[\fB=\fP\fIoptions\fP]
1068
+ .TP
1069
+ \fB\-\-riscv\fP[\fB=\fP\fIoptions\fP]
1070
+ .PD
1071
+ Ajouter un filtre branch/call/jump (BCJ) à la chaîne de filtres. Ces filtres
1072
+ ne peuvent être utilisés que s'ils ne sont pas le dernier filtre de la
1073
+ chaîne de filtrage.
1074
+ .IP ""
1075
+ A BCJ filter converts relative addresses in the machine code to their
1076
+ absolute counterparts. This doesn't change the size of the data but it
1077
+ increases redundancy, which can help LZMA2 to produce 0\(en15\ % smaller
1078
+ \&\fB.xz\fP file. The BCJ filters are always reversible, so using a BCJ filter
1079
+ for wrong type of data doesn't cause any data loss, although it may make the
1080
+ compression ratio slightly worse. The BCJ filters are very fast and use an
1081
+ insignificant amount of memory.
1082
+ .IP ""
1083
+ Ces filtres BCJ présentent des problèmes connus liés au taux de
1084
+ compression :
1085
+ .RS
1086
+ .IP \(bu 3
1087
+ Some types of files containing executable code (for example, object files,
1088
+ static libraries, and Linux kernel modules) have the addresses in the
1089
+ instructions filled with filler values. These BCJ filters will still do the
1090
+ address conversion, which will make the compression worse with these files.
1091
+ .IP \(bu 3
1092
+ If a BCJ filter is applied on an archive, it is possible that it makes the
1093
+ compression ratio worse than not using a BCJ filter. For example, if there
1094
+ are similar or even identical executables then filtering will likely make
1095
+ the files less similar and thus compression is worse. The contents of
1096
+ non\-executable files in the same archive can matter too. In practice one
1097
+ has to try with and without a BCJ filter to see which is better in each
1098
+ situation.
1099
+ .RE
1100
+ .IP ""
1101
+ Different instruction sets have different alignment: the executable file
1102
+ must be aligned to a multiple of this value in the input data to make the
1103
+ filter work.
1104
+ .RS
1105
+ .RS
1106
+ .PP
1107
+ .TS
1108
+ tab(;);
1109
+ l n l
1110
+ l n l.
1111
+ Filtre;Alignement;Notes
1112
+ x86;1;32 bits ou 64 bits x86
1113
+ ARM;4;
1114
+ ARM\-Thumb;2;
1115
+ ARM64;4;4096\-byte alignment is best
1116
+ PowerPC;4;Grand boutiste seulement
1117
+ IA\-64;16;Itanium
1118
+ SPARC;4;
1119
+ RISC\-V;2;
1120
+ .TE
1121
+ .RE
1122
+ .RE
1123
+ .IP ""
1124
+ Since the BCJ\-filtered data is usually compressed with LZMA2, the
1125
+ compression ratio may be improved slightly if the LZMA2 options are set to
1126
+ match the alignment of the selected BCJ filter. Examples:
1127
+ .RS
1128
+ .IP \(bu 3
1129
+ IA\-64 filter has 16\-byte alignment so \fBpb=4,lp=4,lc=0\fP is good with LZMA2
1130
+ (2^4=16).
1131
+ .IP \(bu 3
1132
+ RISC\-V code has 2\-byte or 4\-byte alignment depending on whether the file
1133
+ contains 16\-bit compressed instructions (the C extension). When 16\-bit
1134
+ instructions are used, \fBpb=2,lp=1,lc=3\fP or \fBpb=1,lp=1,lc=3\fP is good. When
1135
+ 16\-bit instructions aren't present, \fBpb=2,lp=2,lc=2\fP is the best.
1136
+ \fBreadelf \-h\fP can be used to check if "RVC" appears on the "Flags" line.
1137
+ .IP \(bu 3
1138
+ ARM64 is always 4\-byte aligned so \fBpb=2,lp=2,lc=2\fP is the best.
1139
+ .IP \(bu 3
1140
+ The x86 filter is an exception. It's usually good to stick to LZMA2's
1141
+ defaults (\fBpb=2,lp=0,lc=3\fP) when compressing x86 executables.
1142
+ .RE
1143
+ .IP ""
1144
+ Tous les filtres BCJ prennent en charge les mêmes \fIoptions\fP :
1145
+ .RS
1146
+ .TP
1147
+ \fBstart=\fP\fIdécalage\fP
1148
+ Spécifier le \fIdécalage\fP de départ qui est utilisé lors de la conversion
1149
+ entre les adresses relatives et absolues. Le \fIdécalage\fP doit être un
1150
+ multiple de l'alignement du filtre (voir la table ci\-dessus). Sa valeur par
1151
+ défaut est zéro. En pratique, cette dernière convient ; indiquer un
1152
+ \fIdécalage\fP personnalisé est la plupart du temps inutile.
1153
+ .RE
1154
+ .TP
1155
+ \fB\-\-delta\fP[\fB=\fP\fIoptions\fP]
1156
+ Ajouter le filtre Delta à la chaîne de filtres. Le filtre Delta ne peut être
1157
+ utilisé que s'il n'est pas le dernier filtre dans la chaîne.
1158
+ .IP ""
1159
+ Currently only simple byte\-wise delta calculation is supported. It can be
1160
+ useful when compressing, for example, uncompressed bitmap images or
1161
+ uncompressed PCM audio. However, special purpose algorithms may give
1162
+ significantly better results than Delta + LZMA2. This is true especially
1163
+ with audio, which compresses faster and better, for example, with
1164
+ \fBflac\fP(1).
1165
+ .IP ""
1166
+ \fIoptions\fP prises en charge :
1167
+ .RS
1168
+ .TP
1169
+ \fBdist=\fP\fIdistance\fP
1170
+ Specify the \fIdistance\fP of the delta calculation in bytes. \fIdistance\fP must
1171
+ be 1\(en256. The default is 1.
1172
+ .IP ""
1173
+ Par exemple, avec \fBdist=2\fP et une entrée huit octets A1 B1 A2 B3 A3 B5 A4
1174
+ B7, la sortie sera A1 B1 01 02 01 02 01 02.
1175
+ .RE
1176
+ .
1177
+ .SS "Autres options"
1178
+ .TP
1179
+ \fB\-q\fP, \fB\-\-quiet\fP
1180
+ Supprimer les avertissements et les notifications. Indiquer cela deux fois
1181
+ supprimera aussi les erreurs. Cette option n'a aucun effet sur le statut de
1182
+ sortie. Cela dit, même si un avertissement était supprimé, le statut de
1183
+ sortie indiquant un avertissement sera encore utilisé.
1184
+ .TP
1185
+ \fB\-v\fP, \fB\-\-verbose\fP
1186
+ Être bavard. Si l'erreur standard est connectée à un terminal, \fBxz\fP
1187
+ affichera une barre de progression. Indiquer \fB\-\-verbose\fP deux fois donnera
1188
+ une sortie encore plus bavarde.
1189
+ .IP ""
1190
+ La barre de progression montre l'information suivante :
1191
+ .RS
1192
+ .IP \(bu 3
1193
+ Le pourcentage de complétion est montré si la taille du fichier en entrée
1194
+ est connue. Néanmoins, le pourcentage ne peut pas être montré en cas de
1195
+ redirection.
1196
+ .IP \(bu 3
1197
+ Quantité de données compressées produites (compression) ou consommées
1198
+ (décompression).
1199
+ .IP \(bu 3
1200
+ Quantité de données non compressées consommées (compression) ou produites
1201
+ (décompression).
1202
+ .IP \(bu 3
1203
+ Le taux de compression, calculé en divisant la quantité de données
1204
+ compréssées déjà traitées par la quantité de données décompressées déjà
1205
+ traitées.
1206
+ .IP \(bu 3
1207
+ Vitesse de compression ou de décompression. Elle correspond à la quantité de
1208
+ données non compressées consommées (compression) ou produites
1209
+ (décompression) par seconde. Elle apparait quelques secondes après le début
1210
+ du traitement du fichier par \fBxz\fP.
1211
+ .IP \(bu 3
1212
+ Temps écoulé dans le format M:SS ou H:MM:SS.
1213
+ .IP \(bu 3
1214
+ Estimated remaining time is shown only when the size of the input file is
1215
+ known and a couple of seconds have already passed since \fBxz\fP started
1216
+ processing the file. The time is shown in a less precise format which never
1217
+ has any colons, for example, 2 min 30 s.
1218
+ .RE
1219
+ .IP ""
1220
+ When standard error is not a terminal, \fB\-\-verbose\fP will make \fBxz\fP print
1221
+ the filename, compressed size, uncompressed size, compression ratio, and
1222
+ possibly also the speed and elapsed time on a single line to standard error
1223
+ after compressing or decompressing the file. The speed and elapsed time are
1224
+ included only when the operation took at least a few seconds. If the
1225
+ operation didn't finish, for example, due to user interruption, also the
1226
+ completion percentage is printed if the size of the input file is known.
1227
+ .TP
1228
+ \fB\-Q\fP, \fB\-\-no\-warn\fP
1229
+ Ne pas mettre l'état de sortie à \fB2\fP même si une condition méritant un
1230
+ avertissement a été détectée. Cette option n'affecte pas le niveau de
1231
+ verbosité, néanmoins, les deux options \fB\-\-quiet\fP et \fB\-\-no\-warn\fP doivent
1232
+ être utilisées pour ne pas afficher d'avertissements, ni altérer le statut
1233
+ de sortie.
1234
+ .TP
1235
+ \fB\-\-robot\fP
1236
+ Afficher les messages dans un format analysable par une machine. Ceci est
1237
+ destiné à faciliter l'écriture des frontaux qui voudraient utiliser \fBxz\fP
1238
+ plutôt que liblzma, ce qui pourrait être le cas pour différents scripts. La
1239
+ sortie avec cette option activée est destinée à rester stable sur les
1240
+ différentes versions de \fBxz\fP. Consulter le paragraphe \fBROBOT MODE\fP pour
1241
+ les détails.
1242
+ .TP
1243
+ \fB\-\-info\-memory\fP
1244
+ Display, in human\-readable format, how much physical memory (RAM) and how
1245
+ many processor threads \fBxz\fP thinks the system has and the memory usage
1246
+ limits for compression and decompression, and exit successfully.
1247
+ .TP
1248
+ \fB\-h\fP, \fB\-\-help\fP
1249
+ Afficher un message d'aide décrivant les options les plus couramment
1250
+ utilisées et quitter.
1251
+ .TP
1252
+ \fB\-H\fP, \fB\-\-long\-help\fP
1253
+ Afficher un message d'aide décrivant toutes les options de \fBxz\fP et quitter.
1254
+ .TP
1255
+ \fB\-V\fP, \fB\-\-version\fP
1256
+ Afficher le numéro de version de \fBxz\fP et de liblzma dans un format lisible
1257
+ par un humain. Pour obtenir une sortie analysable par la machine, spécifiez
1258
+ \fB\-\-robot\fP avant \fB\-\-version\fP.
1259
+ .
1260
+ .SH "MODE ROBOT"
1261
+ The robot mode is activated with the \fB\-\-robot\fP option. It makes the output
1262
+ of \fBxz\fP easier to parse by other programs. Currently \fB\-\-robot\fP is
1263
+ supported only together with \fB\-\-list\fP, \fB\-\-filters\-help\fP, \fB\-\-info\-memory\fP,
1264
+ and \fB\-\-version\fP. It will be supported for compression and decompression in
1265
+ the future.
1266
+ .
1267
+ .SS "Mode liste"
1268
+ \fBxz \-\-robot \-\-list\fP utilise une sortie séparée par des tabulations. La
1269
+ première colonne de toutes les lignes possède une chaîne qui indique le type
1270
+ d'information trouvée sur cette ligne :
1271
+ .TP
1272
+ \fBname\fP
1273
+ C'est toujours la première ligne au début de la liste d'un fichier. La
1274
+ seconde colonne de la ligne est le nom de fichier.
1275
+ .TP
1276
+ \fBfile\fP
1277
+ Cette ligne contient l'information globale sur le fichier \fB.xz\fP. Cette
1278
+ ligne est toujours écrite après la ligne \fBname\fP.
1279
+ .TP
1280
+ \fBstream\fP
1281
+ Ce type de ligne n'est utilisée que lorsque \fB \-\-verbose\fP a été indiquée. Il
1282
+ y a autant de lignes \fBstream\fP qu'il y a de flux dans le fichier \fB.xz\fP.
1283
+ .TP
1284
+ \fBblock\fP
1285
+ Ce type de ligne n'est utilisé seulement lorsque \fB\-\-verbose\fP a été
1286
+ indiquée. Il y a autant de lignes \fBblock\fP qu'il y a de blocs dans le
1287
+ fichier \fB.xz\fP. Les lignes \fBblock\fP sont affichées après toutes les lignes
1288
+ \fBstream\fP ; les différents types de lignes ne sont pas imbriqués.
1289
+ .TP
1290
+ \fBsummary\fP
1291
+ Ce type de ligne n'est utilisé que lorsque \fB\-\-verbose\fP a été indiqué deux
1292
+ fois. Cette ligne est affichée après toutes les lignes \fBblock\fP. Comme la
1293
+ ligne \fBfile\fP, la ligne \fBsummary\fP contient l'information globale sur le
1294
+ fichier \fB.xz\fP.
1295
+ .TP
1296
+ \fBtotals\fP
1297
+ Cette ligne est toujours la toute dernière ligne de la sortie. Elle affiche
1298
+ les comptes et les tailles totaux.
1299
+ .PP
1300
+ Les colonnes des lignes \fBfile\fP :
1301
+ .PD 0
1302
+ .RS
1303
+ .IP 2. 4
1304
+ Nombre de flux dans le fichier
1305
+ .IP 3. 4
1306
+ Nombre total de blocs dans le ou les flux.
1307
+ .IP 4. 4
1308
+ Taille compressée du fichier
1309
+ .IP 5. 4
1310
+ Taille décompressée du fichier
1311
+ .IP 6. 4
1312
+ Compression ratio, for example, \fB0.123\fP. If ratio is over 9.999, three
1313
+ dashes (\fB\-\-\-\fP) are displayed instead of the ratio.
1314
+ .IP 7. 4
1315
+ Liste de noms de contrôles d'intégrité séparés par des virgules. Les chaînes
1316
+ suivantes sont utilisées pour les types de vérification connus : \fBNone\fP,
1317
+ \fBCRC32\fP, \fBCRC64\fP et \fBSHA256\fP. Pour le types de vérification inconnus,
1318
+ \fBUnknown\-\fP\fIN\fP est utilisé, où \fIN\fP est un identifiant de vérification sous
1319
+ la forme d'un nombre décimal (un ou deux chiffres).
1320
+ .IP 8. 4
1321
+ Taille totale du remplissage du flux dans le fichier
1322
+ .RE
1323
+ .PD
1324
+ .PP
1325
+ Les colonnes des lignes \fBstream\fP :
1326
+ .PD 0
1327
+ .RS
1328
+ .IP 2. 4
1329
+ Numéro de flux (le premier flux a le numéro 1)
1330
+ .IP 3. 4
1331
+ Nombre de blocs dans le flux
1332
+ .IP 4. 4
1333
+ Décalage de départ compressé
1334
+ .IP 5. 4
1335
+ Décalage de départ décompressé
1336
+ .IP 6. 4
1337
+ Taille compressée (ne comprend pas le remplissage du flux)
1338
+ .IP 7. 4
1339
+ Taille décompressée
1340
+ .IP 8. 4
1341
+ Taux de compression
1342
+ .IP 9. 4
1343
+ Nom de la vérification d'intégrité
1344
+ .IP 10. 4
1345
+ Taille du remplissage de flux
1346
+ .RE
1347
+ .PD
1348
+ .PP
1349
+ Les colonnes des lignes \fBblock\fP :
1350
+ .PD 0
1351
+ .RS
1352
+ .IP 2. 4
1353
+ Numéro du flux qui contient ce bloc
1354
+ .IP 3. 4
1355
+ Numéro du bloc relatif au commencement du flux (le premier bloc a pour
1356
+ numéro 1)
1357
+ .IP 4. 4
1358
+ Numéro du bloc relatif au début du fichier
1359
+ .IP 5. 4
1360
+ Décalage de départ compressé relatif au début du fichier
1361
+ .IP 6. 4
1362
+ Décalage de départ décompressé relatif au début du fichier
1363
+ .IP 7. 4
1364
+ Taille compressée totale du bloc (en\-têtes inclus)
1365
+ .IP 8. 4
1366
+ Taille décompressée
1367
+ .IP 9. 4
1368
+ Taux de compression
1369
+ .IP 10. 4
1370
+ Nom de la vérification d'intégrité
1371
+ .RE
1372
+ .PD
1373
+ .PP
1374
+ Si \fB\-\-verbose\fP a été indiqué deux fois, les colonnes additionnelles sont
1375
+ inclues sur les lignes \fBblock\fP. Elles ne sont pas affichées avec un seul
1376
+ \fB\-\-verbose\fP, car l'obtention de ces informations nécessite de nombreuses
1377
+ recherches et peut donc être lente :
1378
+ .PD 0
1379
+ .RS
1380
+ .IP 11. 4
1381
+ Valeur de la vérification d'intégrité en hexadécimal
1382
+ .IP 12. 4
1383
+ Taille d'en\-tête de bloc
1384
+ .IP 13. 4
1385
+ Drapeaux du bloc : \fBc\fP indique que la taille compressée est présente, et
1386
+ \fBu\fP indique que la taille décompréssée est présente. Si le drapeau n'est
1387
+ pas indiqué, un tiret (\fB\-\fP) est affiché à la place pour que la longueur de
1388
+ la chaîne reste fixe. De nouveaux drapeaux pourraient être ajoutés à la fin
1389
+ de la chaîne dans le futur.
1390
+ .IP 14. 4
1391
+ Taille des données effectivement compressées dans le bloc (en excluant
1392
+ l'en\-tête de bloc, le remplissage de bloc et les champs de vérification).
1393
+ .IP 15. 4
1394
+ Quantité de mémoire (en octets) nécessaire pour décompresser ce bloc avec
1395
+ cette version de \fBxz\fP.
1396
+ .IP 16. 4
1397
+ Chaîne de filtrage. Remarquez que la plupart des options utilisées au moment
1398
+ de la compression ne peuvent pas être connues, car seules les options
1399
+ nécessaires pour la décompression sont stockées dans les en\-têtes \fB.xz\fP.
1400
+ .RE
1401
+ .PD
1402
+ .PP
1403
+ Les colonnes des lignes \fBsummary\fP :
1404
+ .PD 0
1405
+ .RS
1406
+ .IP 2. 4
1407
+ Quantité de mémoire (en octets) nécessaire pour décompresser ce fichier avec
1408
+ cette version de \fBxz\fP.
1409
+ .IP 3. 4
1410
+ \fByes\fP ou \fBno\fP indique si tous les en\-têtes de bloc stockent à la fois la
1411
+ taille compressée et la taille décompressée.
1412
+ .PP
1413
+ \fIDepuis\fP \fBxz\fP \fI5.1.2alpha:\fP
1414
+ .IP 4. 4
1415
+ Version minimale de \fBxz\fP nécessaire pour décompresser le fichier.
1416
+ .RE
1417
+ .PD
1418
+ .PP
1419
+ Les colonnes de la ligne \fBtotals\fP :
1420
+ .PD 0
1421
+ .RS
1422
+ .IP 2. 4
1423
+ Nombre de flux
1424
+ .IP 3. 4
1425
+ Nombre de blocs
1426
+ .IP 4. 4
1427
+ Taille compressée
1428
+ .IP 5. 4
1429
+ Taille décompressée
1430
+ .IP 6. 4
1431
+ Taux de compression moyen
1432
+ .IP 7. 4
1433
+ Liste séparée par des virgules des noms de vérification d'intégrité qui
1434
+ étaient présents dans les fichiers
1435
+ .IP 8. 4
1436
+ Taille de remplissage de flux
1437
+ .IP 9. 4
1438
+ Nombre de fichiers. Permet de garder l'ordre des colonnes précédentes comme
1439
+ sur les lignes \fBfile\fP.
1440
+ .PD
1441
+ .RE
1442
+ .PP
1443
+ Si \fB\-\-verbose\fP a été indiqué deux fois, des colonnes supplémentaires sont
1444
+ incluses sur la ligne \fBtotals\fP :
1445
+ .PD 0
1446
+ .RS
1447
+ .IP 10. 4
1448
+ Quantité maximale de mémoire (en octets) nécessaire pour décompresser les
1449
+ fichiers avec cette version de \fBxz\fP.
1450
+ .IP 11. 4
1451
+ \fByes\fP ou \fBno\fP indique si tous les en\-têtes de bloc stockent à la fois la
1452
+ taille compressée et la taille décompressée.
1453
+ .PP
1454
+ \fIDepuis\fP \fBxz\fP \fI5.1.2alpha:\fP
1455
+ .IP 12. 4
1456
+ Version minimale de \fBxz\fP nécessaire pour décompresser le fichier.
1457
+ .RE
1458
+ .PD
1459
+ .PP
1460
+ Les versions futures pourront ajouter de nouveaux types de lignes et de
1461
+ nouvelles colonnes pourront être ajoutées aux types de lignes existants,
1462
+ mais les colonnes existantes ne seront pas modifiées.
1463
+ .
1464
+ .SS "Filters help"
1465
+ \fBxz \-\-robot \-\-filters\-help\fP prints the supported filters in the following
1466
+ format:
1467
+ .PP
1468
+ \fIfilter\fP\fB:\fP\fIoption\fP\fB=<\fP\fIvalue\fP\fB>,\fP\fIoption\fP\fB=<\fP\fIvalue\fP\fB>\fP...
1469
+ .TP
1470
+ \fIfilter\fP
1471
+ Name of the filter
1472
+ .TP
1473
+ \fIoption\fP
1474
+ Name of a filter specific option
1475
+ .TP
1476
+ \fIvalue\fP
1477
+ Numeric \fIvalue\fP ranges appear as \fB<\fP\fImin\fP\fB\-\fP\fImax\fP\fB>\fP. String
1478
+ \fIvalue\fP choices are shown within \fB< >\fP and separated by a \fB|\fP
1479
+ character.
1480
+ .PP
1481
+ Each filter is printed on its own line.
1482
+ .
1483
+ .SS "Information de limite de mémoire"
1484
+ \fBxz \-\-robot \-\-info\-memory\fP prints a single line with multiple tab\-separated
1485
+ columns:
1486
+ .IP 1. 4
1487
+ Total amount of physical memory (RAM) in bytes.
1488
+ .IP 2. 4
1489
+ Memory usage limit for compression in bytes (\fB\-\-memlimit\-compress\fP). A
1490
+ special value of \fB0\fP indicates the default setting which for
1491
+ single\-threaded mode is the same as no limit.
1492
+ .IP 3. 4
1493
+ Memory usage limit for decompression in bytes (\fB\-\-memlimit\-decompress\fP). A
1494
+ special value of \fB0\fP indicates the default setting which for
1495
+ single\-threaded mode is the same as no limit.
1496
+ .IP 4. 4
1497
+ Since \fBxz\fP 5.3.4alpha: Memory usage for multi\-threaded decompression in
1498
+ bytes (\fB\-\-memlimit\-mt\-decompress\fP). This is never zero because a
1499
+ system\-specific default value shown in the column 5 is used if no limit has
1500
+ been specified explicitly. This is also never greater than the value in the
1501
+ column 3 even if a larger value has been specified with
1502
+ \fB\-\-memlimit\-mt\-decompress\fP.
1503
+ .IP 5. 4
1504
+ Since \fBxz\fP 5.3.4alpha: A system\-specific default memory usage limit that is
1505
+ used to limit the number of threads when compressing with an automatic
1506
+ number of threads (\fB\-\-threads=0\fP) and no memory usage limit has been
1507
+ specified (\fB\-\-memlimit\-compress\fP). This is also used as the default value
1508
+ for \fB\-\-memlimit\-mt\-decompress\fP.
1509
+ .IP 6. 4
1510
+ Since \fBxz\fP 5.3.4alpha: Number of available processor threads.
1511
+ .PP
1512
+ Dans le futur, la sortie de \fBxz \-\-robot \-\-info\-memory\fP pourrait avoir plus
1513
+ de colonnes, mais jamais plus qu'une ligne unique.
1514
+ .
1515
+ .SS Version
1516
+ \fBxz \-\-robot \-\-version\fP prints the version number of \fBxz\fP and liblzma in
1517
+ the following format:
1518
+ .PP
1519
+ \fBXZ_VERSION=\fP\fIXYYYZZZS\fP
1520
+ .br
1521
+ \fBLIBLZMA_VERSION=\fP\fIXYYYZZZS\fP
1522
+ .TP
1523
+ \fIX\fP
1524
+ Version majeure.
1525
+ .TP
1526
+ \fIYYY\fP
1527
+ Version mineure. Les numéros pairs sont stables. Les numéros impairs sont
1528
+ des versions alpha ou beta.
1529
+ .TP
1530
+ \fIZZZ\fP
1531
+ Niveau de correctif pour les options stables ou juste un compteur pour les
1532
+ options de développement.
1533
+ .TP
1534
+ \fIS\fP
1535
+ Stabilité. 0 est alpha, 1 est bêta et 2 est stable. \fIS\fP devrait toujours
1536
+ être 2 quand \fIYYY\fP est pair.
1537
+ .PP
1538
+ \fIXYYYZZZS\fP sont identiques sur les deux lignes si \fBxz\fP et liblzma sont
1539
+ issus de la même version d'utilitaires XZ.
1540
+ .PP
1541
+ Exemples : 4.999.9beta est \fB49990091\fP et 5.0.0 est \fB50000002\fP.
1542
+ .
1543
+ .SH "STATUT DE SORTIE"
1544
+ .TP
1545
+ \fB0\fP
1546
+ Tout est bon.
1547
+ .TP
1548
+ \fB1\fP
1549
+ Une erreur est survenue.
1550
+ .TP
1551
+ \fB2\fP
1552
+ Quelquechose méritant un avertissement s'est produit, mais aucune erreur
1553
+ véritable n'est survenue.
1554
+ .PP
1555
+ Les notifications (pas les avertissements ou les erreurs) affichées sur
1556
+ l'erreur standard n'affectent pas le statut de sortie.
1557
+ .
1558
+ .SH ENVIRONNEMENT
1559
+ \fBxz\fP analyse les listes d'options séparées par des espaces à partir des
1560
+ variables d'environnement \fBXZ_DEFAULTS\fP et \fBXZ_OPT\fP, dans cet ordre, avant
1561
+ d'analyser les options de la ligne de commandes. Remarquez que seules les
1562
+ options sont analysées depuis l'environnement des variables ; toutes les
1563
+ non\-options sont ignorées silencieusement. L'analyse est faite avec
1564
+ \fBgetopt_long\fP(3) qui est aussi utilisé pour les arguments de la ligne de
1565
+ commandes.
1566
+ .TP
1567
+ \fBXZ_DEFAULTS\fP
1568
+ Options par défaut propres à l'utilisateur ou pour tout le système. Elles
1569
+ sont le plus souvent définies dans un script d'initialisation de
1570
+ l'interpréteur pour activer le limiteur d'utilisation de la mémoire de \fBxz\fP
1571
+ par défaut. A part pour les scripts d'initialisation de l'interpréteur ou
1572
+ des cas similaires, les sripts ne doivent jamais définir ou désactiver
1573
+ \fBXZ_DEFAULTS\fP.
1574
+ .TP
1575
+ \fBXZ_OPT\fP
1576
+ This is for passing options to \fBxz\fP when it is not possible to set the
1577
+ options directly on the \fBxz\fP command line. This is the case when \fBxz\fP is
1578
+ run by a script or tool, for example, GNU \fBtar\fP(1):
1579
+ .RS
1580
+ .RS
1581
+ .PP
1582
+ .nf
1583
+ \f(CRXZ_OPT=\-2v tar caf foo.tar.xz foo\fR
1584
+ .fi
1585
+ .RE
1586
+ .RE
1587
+ .IP ""
1588
+ Scripts may use \fBXZ_OPT\fP, for example, to set script\-specific default
1589
+ compression options. It is still recommended to allow users to override
1590
+ \fBXZ_OPT\fP if that is reasonable. For example, in \fBsh\fP(1) scripts one may
1591
+ use something like this:
1592
+ .RS
1593
+ .RS
1594
+ .PP
1595
+ .nf
1596
+ \f(CRXZ_OPT=${XZ_OPT\-"\-7e"}
1597
+ export XZ_OPT\fR
1598
+ .fi
1599
+ .RE
1600
+ .RE
1601
+ .
1602
+ .SH "Compatibilité des utilitaires LZMA"
1603
+ La syntaxe de la ligne de commande de \fBxz\fP est quasimment un sur\-ensemble
1604
+ de \fBlzma\fP, \fBunlzma\fP et \fBlzcat\fP comme ils sont trouvés dans les
1605
+ utilitaires LZMA 4.32.x . Dans la pluspart des cas, il est possible de
1606
+ remplacer les outils LZMA par les outils XZ sans casser les scripts
1607
+ existants. Il existe cependant certaines incompatibilités qui peuvent
1608
+ parfois poser des problèmes.
1609
+ .
1610
+ .SS "Niveaux de préréglage de la compression"
1611
+ La numérotation des préréglages de niveau de compression est différente
1612
+ entre les outils \fBxz\fP et LZMA. La différence la plus importante est la
1613
+ manière dont les tailles de dictionnaire sont affectées aux différents
1614
+ préréglages. La taille de dictionnaire est à peu près égale à celle
1615
+ d'utilisation de la mémoire de la décompression.
1616
+ .RS
1617
+ .PP
1618
+ .TS
1619
+ tab(;);
1620
+ c c c
1621
+ c n n.
1622
+ Niveau;xz;Utilitaires LZMA
1623
+ \-0;256 KiB;N/A
1624
+ \-1;1 MiB;64 KiB
1625
+ \-2;2 MiB;1 MiB
1626
+ \-3;4 MiB;512 KiB
1627
+ \-4;4 MiB;1 MiB
1628
+ \-5;8 MiB;2 MiB
1629
+ \-6;8 MiB;4 MiB
1630
+ \-7;16 MiB;8 MiB
1631
+ \-8;32 MiB;16 MiB
1632
+ \-9;64 MiB;32 MiB
1633
+ .TE
1634
+ .RE
1635
+ .PP
1636
+ Les différences de tailles des dictionnaires affectent aussi l'utilisation
1637
+ de la mémoire du compresseur, mais il y a quelques autres différences entre
1638
+ les outils LZMA et les outils XZ, qui rendent la différence encore plus
1639
+ grande :
1640
+ .RS
1641
+ .PP
1642
+ .TS
1643
+ tab(;);
1644
+ c c c
1645
+ c n n.
1646
+ Niveau;xz;Utilitaires LZMA 4.32.x
1647
+ \-0;3 MiB;N/A
1648
+ \-1;9 MiB;2 MiB
1649
+ \-2;17 MiB;12 MiB
1650
+ \-3;32 MiB;12 MiB
1651
+ \-4;48 MiB;16 MiB
1652
+ \-5;94 MiB;26 MiB
1653
+ \-6;94 MiB;45 MiB
1654
+ \-7;186 MiB;83 MiB
1655
+ \-8;370 MiB;159 MiB
1656
+ \-9;674 MiB;311 MiB
1657
+ .TE
1658
+ .RE
1659
+ .PP
1660
+ Le niveau de préréglage par défaut dans les outils LZMA est \fB\-7\fP alors que
1661
+ pour les outils XZ c'est \fB\-6\fP, les deux utilisent ainsi un dictionnaire de
1662
+ 8 Mio par défaut.
1663
+ .
1664
+ .SS "Fichiers .lzma en flux ou non"
1665
+ The uncompressed size of the file can be stored in the \fB.lzma\fP header.
1666
+ LZMA Utils does that when compressing regular files. The alternative is to
1667
+ mark that uncompressed size is unknown and use end\-of\-payload marker to
1668
+ indicate where the decompressor should stop. LZMA Utils uses this method
1669
+ when uncompressed size isn't known, which is the case, for example, in
1670
+ pipes.
1671
+ .PP
1672
+ \fBxz\fP prend en charge la décompression des fichiers \fB.lzma\fP avec ou sans
1673
+ marqueur de fin de charge utile, mais tous les fichiers \fB.lzma\fP créés par
1674
+ \fBxz\fP utiliseront un marqueur de fin de charge utile et ont la taille non
1675
+ compréssée marquée comme inconnue dans l'en\-tête \fB.lzma\fP. Cela peut être un
1676
+ problème dans quelques situations inhabituelles. Par exemple, un
1677
+ décompresseur \fB.lzma\fP dans un périphérique embarqué pourrait ne fonctionner
1678
+ qu'avec des fichiers dont la taille non comprimée est connue. Si vous vous
1679
+ heurtez à ce problème, vous devez utiliser les utilitaires LZMA ou LZMA SDK
1680
+ pour créer des fichiers \fB.lzma\fP avec une taille non compressée connue.
1681
+ .
1682
+ .SS "Fichiers .lzma non pris en charge"
1683
+ Le format \fB.lzma\fP autorise des valeurs \fIlc\fP jusqu'à 8, et des valeurs
1684
+ \fIlp\fP jusqu'à 4. Les outils LZMA peuvent décompresser des fichiers avec tous
1685
+ les \fIlc\fP et \fIlp\fP, mais créez toujours les fichiers avec \fBlc=3\fP et
1686
+ \fBlp=0\fP. Créer des fichiers avec d'autres valeurs \fIlc\fP et \fIlp\fP est
1687
+ possible avec \fBxz\fP et avec LZMA SDK.
1688
+ .PP
1689
+ L'implémentation du filtre LZMA1 dans liblzma nécessite que la somme de
1690
+ \fIlc\fP et \fIlp\fP soit inférieure ou égale à 4. Ainsi, les fichiers \fB.lzma\fP
1691
+ qui excèdent cette limitation ne peuvent pas être décompressés avec \fBxz\fP.
1692
+ .PP
1693
+ Les outils LZMA créent seulement des fichiers \fB.lzma\fP qui ont une taille de
1694
+ dictionnaire de 2^\fIn\fP (une puissance de 2) mais acceptent les fichiers avec
1695
+ toutes les tailles de dictionnaire. Libzlma n'accepte que les fichiers
1696
+ \&\fB.lzma\fP qui ont une taille dictionnaire de 2^\fIn\fP ou
1697
+ 2^\fIn\fP + 2^(\fIn\fP\-1). Cela afin de diminuer les faux positifs lors de la
1698
+ détection des fichiers \fB.lzma\fP.
1699
+ .PP
1700
+ Ces limitations ne devraient pas poser problème en pratique, car
1701
+ pratiquement tous les fichiers \fB.lzma\fP ont été compressés avec des réglages
1702
+ que liblzma accepte.
1703
+ .
1704
+ .SS "Déchets excédentaires"
1705
+ Lors de la décompession, l'utilitaire LZMA ignore silencieusement tout ce
1706
+ qui est après le premier flux \fB.lzma\fP. Dans la majorité des situations,
1707
+ c'est un bogue. Cela veut dire aussi que les outils LZMA ne gèrent pas la
1708
+ décompression de fichiers \fB.lzma\fP concaténés.
1709
+ .PP
1710
+ S'il reste des données après le premier flux \fB.lzma\fP, \fBxz\fP considère que
1711
+ le fichier est corrompu sauf si \fB\-\-single\-stream\fP a été utilisé. Cela peut
1712
+ casser des scripts obscurs qui ont supposé que les déchets de fin de ligne
1713
+ sont ignorés.
1714
+ .
1715
+ .SH NOTES
1716
+ .
1717
+ .SS "La sortie compressée peut varier"
1718
+ La sortie compressée exacte produite par les même fichiers non compressés en
1719
+ entrée peut varier en fonction des différentes versions de l'utilitaire XZ,
1720
+ même si les options de compression sont identiques. En effet, il est
1721
+ possible d'améliorer l'encodeur (compression plus rapide ou meilleure) sans
1722
+ affecter le format du fichier. La sortie peut même varier entre différentes
1723
+ compilations de la même version d'utilitaire XZ, si des options de
1724
+ construction différentes sont utilisées.
1725
+ .PP
1726
+ Cela signifie qu'une fois que \fB\-\-rsyncable\fP a été implémenté, les fichiers
1727
+ résultants ne seront pas nécessairement synchronisables avec rsync à moins
1728
+ que les nouveaux et anciens fichiers n'aient été compressés avec la même
1729
+ version de xz. Ce problème peut être résolu si une partie de
1730
+ l'implémentation est gelée pour garantir la stabilité de la sortie rsyncable
1731
+ à travers les versions de xz.
1732
+ .
1733
+ .SS "Décompresseurs .xz embarqués"
1734
+ Les implémentations de décompresseur embarqué comme XZ Embedded ne gèrent
1735
+ pas nécessairement les fichiers créés avec d'autres types de \fIvérification\fP
1736
+ d'intégrité que \fBnone\fP et \fBCRC32\fP. Comme la valeur par défaut est
1737
+ \fB\-\-check=crc64\fP, vous devez utiliser \fB\-\-check=none\fP ou \fB\-\-check=crc32\fP
1738
+ lors de la création de fichiers pour les systèmes embarqués.
1739
+ .PP
1740
+ En dehors des systèmes embarqués, tous les décompresseurs de format \fB.xz\fP
1741
+ gèrent tous les types de \fIvérification\fP ou sont au moins capables de
1742
+ décompresser le fichier sans effectuer la vérification d'intégrité si ce
1743
+ type de \fIvérification\fP particulière n'est pas pris en charge.
1744
+ .PP
1745
+ XZ Embedded prend en charge les filtres BCJ, mais seulement avec le décalage
1746
+ de départ par défaut.
1747
+ .
1748
+ .SH EXEMPLES
1749
+ .
1750
+ .SS Bases
1751
+ Compresser le fichier \fItoto\fP en \fItoto.xz\fP en utilisant le niveau de
1752
+ compression par défaut (\fB\-6\fP) et supprimer \fItoto\fP si la compression
1753
+ réussit :
1754
+ .RS
1755
+ .PP
1756
+ .nf
1757
+ \f(CRxz foo\fR
1758
+ .fi
1759
+ .RE
1760
+ .PP
1761
+ Décompresser \fIbidule.xz\fP en \fIbidule\fP et ne pas supprimer \fIbidule.xz\fP même
1762
+ si la compression réussit :
1763
+ .RS
1764
+ .PP
1765
+ .nf
1766
+ \f(CRxz \-dk bar.xz\fR
1767
+ .fi
1768
+ .RE
1769
+ .PP
1770
+ Create \fIbaz.tar.xz\fP with the preset \fB\-4e\fP (\fB\-4 \-\-extreme\fP), which is
1771
+ slower than the default \fB\-6\fP, but needs less memory for compression and
1772
+ decompression (48\ MiB and 5\ MiB, respectively):
1773
+ .RS
1774
+ .PP
1775
+ .nf
1776
+ \f(CRtar cf \- baz | xz \-4e > baz.tar.xz\fR
1777
+ .fi
1778
+ .RE
1779
+ .PP
1780
+ Un mélange de fichiers compressés et non compressés peuvent être
1781
+ décompressés vers la sortie standard avec une simple commande :
1782
+ .RS
1783
+ .PP
1784
+ .nf
1785
+ \f(CRxz \-dcf a.txt b.txt.xz c.txt d.txt.lzma > abcd.txt\fR
1786
+ .fi
1787
+ .RE
1788
+ .
1789
+ .SS "Compression en parallèle de plusieurs fichiers"
1790
+ Sur GNU et *BSD, \fBfind\fP(1) et \fBxargs\fP(1) peuvent être utilisés pour mettre
1791
+ en parallèle la compression de plusieurs fichiers :
1792
+ .RS
1793
+ .PP
1794
+ .nf
1795
+ \f(CRfind . \-type f \e! \-name '*.xz' \-print0 \e
1796
+ | xargs \-0r \-P4 \-n16 xz \-T1\fR
1797
+ .fi
1798
+ .RE
1799
+ .PP
1800
+ L'option \fBP\fP passée à \fBxargs\fP(1) fixe le nombre de processus \fBxz\fP en
1801
+ parallèles. La meilleure valeur pour l'option \fBn\fP dépend du nombre de
1802
+ fichiers à compresser. S\-il n'y a que quelques fichiers, la valeur sera
1803
+ probablement 1 ; avec des dizaines de milliers de fichiers, 100 ou même plus
1804
+ serait approprié pour réduire le nombre de processus \fBxz\fP que \fBxargs\fP(1)
1805
+ créera éventuellement.
1806
+ .PP
1807
+ L'option \fB\-T1\fP de \fBxz\fP est là pour le forcer en mode mono\-thread, car
1808
+ \fBxargs\fP(1) est utilisé pour contrôler la quantité de mise en parallèle.
1809
+ .
1810
+ .SS "Mode robot"
1811
+ Calculer combien d'octets ont été économisés au total après avoir compressé
1812
+ plusieurs fichiers :
1813
+ .RS
1814
+ .PP
1815
+ .nf
1816
+ \f(CRxz \-\-robot \-\-list *.xz | awk '/^totals/{print $5\-$4}'\fR
1817
+ .fi
1818
+ .RE
1819
+ .PP
1820
+ Un script peut vouloir savoir qu'il utilise une version suffisamment récente
1821
+ de \fBxz\fP. Le script \fBsh\fP(1) suivant vérifie que le numéro de version de
1822
+ l'outil \fBxz\fP soit au minimum 5.0.0. Cette méthode est compatible avec les
1823
+ vieilles versions bêta, qui ne gèrent pas l'option \fB\-\-robot\fP :
1824
+ .RS
1825
+ .PP
1826
+ .nf
1827
+ \f(CRif ! eval "$(xz \-\-robot \-\-version 2> /dev/null)" ||
1828
+ [ "$XZ_VERSION" \-lt 50000002 ]; then
1829
+ echo "Your xz is too old."
1830
+ fi
1831
+ unset XZ_VERSION LIBLZMA_VERSION\fR
1832
+ .fi
1833
+ .RE
1834
+ .PP
1835
+ Régler une limite d'utilisation de la mémoire pour la décompression en
1836
+ utilisant \fBXZ_OPT\fP, mais si une limite a déjà été définie, ne pas
1837
+ l'augmenter :
1838
+ .RS
1839
+ .PP
1840
+ .nf
1841
+ \f(CRNEWLIM=$((123 << 20))\ \ # 123 MiB
1842
+ OLDLIM=$(xz \-\-robot \-\-info\-memory | cut \-f3)
1843
+ if [ $OLDLIM \-eq 0 \-o $OLDLIM \-gt $NEWLIM ]; then
1844
+ XZ_OPT="$XZ_OPT \-\-memlimit\-decompress=$NEWLIM"
1845
+ export XZ_OPT
1846
+ fi\fR
1847
+ .fi
1848
+ .RE
1849
+ .
1850
+ .SS "Chaînes de filtres de compresseur personnalisées"
1851
+ L'utilisation la plus simple des chaînes de filtres personnalisées est la
1852
+ personnalisation d'un préréglage LZMA2. Cela peut être utile, car les
1853
+ préréglages ne couvrent qu'un sous\-ensemble des réglages de compression
1854
+ potentiellement utiles.
1855
+ .PP
1856
+ Les colonnes CompCPU des tableaux des descriptions des options \fB\-0\fP à \fB\-9\fP
1857
+ et \fB\-\-extreme\fP sont utiles lors de la personnalisation des préréglages
1858
+ LZMA2. Voici les parties pertinentes recueillies à partir de ces deux
1859
+ tableaux :
1860
+ .RS
1861
+ .PP
1862
+ .TS
1863
+ tab(;);
1864
+ c c
1865
+ n n.
1866
+ Préréglage;CompCPU
1867
+ \-0;0
1868
+ \-1;1
1869
+ \-2;2
1870
+ \-3;3
1871
+ \-4;4
1872
+ \-5;5
1873
+ \-6;6
1874
+ \-5e;7
1875
+ \-6e;8
1876
+ .TE
1877
+ .RE
1878
+ .PP
1879
+ If you know that a file requires somewhat big dictionary (for example, 32\ MiB) to compress well, but you want to compress it quicker than \fBxz \-8\fP
1880
+ would do, a preset with a low CompCPU value (for example, 1) can be
1881
+ modified to use a bigger dictionary:
1882
+ .RS
1883
+ .PP
1884
+ .nf
1885
+ \f(CRxz \-\-lzma2=preset=1,dict=32MiB foo.tar\fR
1886
+ .fi
1887
+ .RE
1888
+ .PP
1889
+ Avec certains fichiers, la commande ci\-dessus peut être plus rapide que
1890
+ \fBxz\-6\fP tout en compressant bien mieux. Cependant, il faut souligner que
1891
+ seuls certains fichiers bénéficient d'un grand dictionnaire tout en gardant
1892
+ la valeur de CompCPU faible. La siutation la plus évidente où un gros
1893
+ dictionnaire peut baucoup aider, est une archive contenant des fichiers très
1894
+ similaires de quelques megaoctets chacun. La taille de dictionnaire doit
1895
+ être significativement plus grosse que tout fichier individuel pour
1896
+ permettre à LZMA2 de tirer pleinement partie des similarités entre des
1897
+ fichiers consécutifs.
1898
+ .PP
1899
+ Si une utilisation de la mémoire élevée pour la compression et décompression
1900
+ convient, et que le fichier à compresser a une taille de plusieurs centaines
1901
+ de megaoctets, il peut être utile d'utiliser un plus gros dictionnaire que
1902
+ celui fourni par \fBxz\-9\fP (64 Mio) :
1903
+ .RS
1904
+ .PP
1905
+ .nf
1906
+ \f(CRxz \-vv \-\-lzma2=dict=192MiB big_foo.tar\fR
1907
+ .fi
1908
+ .RE
1909
+ .PP
1910
+ Utiliser \fB\-vv\fP (\fB\-\-verbose\-\-verbose\fP) comme dans l'exemple ci\-dessus peut
1911
+ être utile pour voir les besoins en mémoire du compresseur et du
1912
+ décompresseur. Rappelez\-vous qu'utiliser un dictionnaire plus gros que la
1913
+ taille du fichier non compressé est un gachis de mémoire, donc la commande
1914
+ ci\-dessus n'est pas utile pour les petits fichiers.
1915
+ .PP
1916
+ Sometimes the compression time doesn't matter, but the decompressor memory
1917
+ usage has to be kept low, for example, to make it possible to decompress the
1918
+ file on an embedded system. The following command uses \fB\-6e\fP (\fB\-6 \-\-extreme\fP) as a base and sets the dictionary to only 64\ KiB. The
1919
+ resulting file can be decompressed with XZ Embedded (that's why there is
1920
+ \fB\-\-check=crc32\fP) using about 100\ KiB of memory.
1921
+ .RS
1922
+ .PP
1923
+ .nf
1924
+ \f(CRxz \-\-check=crc32 \-\-lzma2=preset=6e,dict=64KiB foo\fR
1925
+ .fi
1926
+ .RE
1927
+ .PP
1928
+ If you want to squeeze out as many bytes as possible, adjusting the number
1929
+ of literal context bits (\fIlc\fP) and number of position bits (\fIpb\fP) can
1930
+ sometimes help. Adjusting the number of literal position bits (\fIlp\fP)
1931
+ might help too, but usually \fIlc\fP and \fIpb\fP are more important. For
1932
+ example, a source code archive contains mostly US\-ASCII text, so something
1933
+ like the following might give slightly (like 0.1\ %) smaller file than \fBxz \-6e\fP (try also without \fBlc=4\fP):
1934
+ .RS
1935
+ .PP
1936
+ .nf
1937
+ \f(CRxz \-\-lzma2=preset=6e,pb=0,lc=4 source_code.tar\fR
1938
+ .fi
1939
+ .RE
1940
+ .PP
1941
+ Using another filter together with LZMA2 can improve compression with
1942
+ certain file types. For example, to compress a x86\-32 or x86\-64 shared
1943
+ library using the x86 BCJ filter:
1944
+ .RS
1945
+ .PP
1946
+ .nf
1947
+ \f(CRxz \-\-x86 \-\-lzma2 libfoo.so\fR
1948
+ .fi
1949
+ .RE
1950
+ .PP
1951
+ Notez que l'ordre des options de filtre est significatif. Si \fB\-\-x86\fP est
1952
+ indiqué après \fB\-\-lzma2\fP, \fBxz\fP donnera une erreur, car il ne peut y avoir
1953
+ aucun filtre après LZMA2, et aussi parce que le filtre BCJ x86 ne peut pas
1954
+ être utilisé comme dernier filtre dans la chaîne.
1955
+ .PP
1956
+ Le filtre Delta associé à LZMA2 peut donner de bons résultats avec les
1957
+ images bitmap. Cela devrait habituellement battre PNG, qui a quelques
1958
+ filtres avancés supplémentaires qu'un simple delta, mais qui utilise Deflate
1959
+ pour la compression effective.
1960
+ .PP
1961
+ The image has to be saved in uncompressed format, for example, as
1962
+ uncompressed TIFF. The distance parameter of the Delta filter is set to
1963
+ match the number of bytes per pixel in the image. For example, 24\-bit RGB
1964
+ bitmap needs \fBdist=3\fP, and it is also good to pass \fBpb=0\fP to LZMA2 to
1965
+ accommodate the three\-byte alignment:
1966
+ .RS
1967
+ .PP
1968
+ .nf
1969
+ \f(CRxz \-\-delta=dist=3 \-\-lzma2=pb=0 foo.tiff\fR
1970
+ .fi
1971
+ .RE
1972
+ .PP
1973
+ If multiple images have been put into a single archive (for example,
1974
+ \&\fB.tar\fP), the Delta filter will work on that too as long as all images have
1975
+ the same number of bytes per pixel.
1976
+ .
1977
+ .SH "VOIR AUSSI"
1978
+ \fBxzdec\fP(1), \fBxzdiff\fP(1), \fBxzgrep\fP(1), \fBxzless\fP(1), \fBxzmore\fP(1),
1979
+ \fBgzip\fP(1), \fBbzip2\fP(1), \fB7z\fP(1)
1980
+ .PP
1981
+ XZ Utils: <https://tukaani.org/xz/>
1982
+ .br
1983
+ XZ Embedded: <https://tukaani.org/xz/embedded.html>
1984
+ .br
1985
+ LZMA SDK: <https://7\-zip.org/sdk.html>
janus/share/man/fr/man1/lzmadec.1 ADDED
@@ -0,0 +1,90 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ .\" SPDX-License-Identifier: 0BSD
2
+ .\"
3
+ .\" Author: Lasse Collin
4
+ .\"
5
+ .\" french translation of XZ Utils man
6
+ .\" Copyright (C) 2021 Debian French l10n team <debian-l10n-french@lists.debian.org>
7
+ .\" Translator
8
+ .\" bubu <bubub@no-log.org>, 2021.
9
+ .\"
10
+ .\"*******************************************************************
11
+ .\"
12
+ .\" This file was generated with po4a. Translate the source file.
13
+ .\"
14
+ .\"*******************************************************************
15
+ .TH XZDEC 1 2024\-04\-08 Tukaani "Utilitaires XZ"
16
+ .SH NOM
17
+ xzdec, lzmadec \- Small .xz et .lzma decompresseurs
18
+ .SH SYNOPSIS
19
+ \fBxzdec\fP [\fIoption...\fP] [\fIfichier...\fP]
20
+ .br
21
+ \fBlzmadec\fP [\fIoption...\fP] [\fIfichier...\fP]
22
+ .SH DESCRIPTION
23
+ \fBxzdec\fP est un outil uniquement de décompression, basé sur liblzma pour les
24
+ fichiers \fB.xz\fP (et seulement \fB.xz\fP). \fBxzdec\fP est destiné à remplacer
25
+ \fBxz\fP(1) dans les situations les plus courantes où un script a été écrit
26
+ pour utiliser \fBxz \-\-decompress \-\-stdout\fP (et possiblement quelques autres
27
+ options courantes) pour décompresser des fichiers \fB.xz\fP. \fBlzmadec\fP est
28
+ identique à \fBxzdec\fP, sauf que \fBlzmadec\fP prend en charge les fichiers
29
+ \&\fB.lzma\fP au lieu des fichiers \fB.xz\fP.
30
+ .PP
31
+ Pour réduire la taille de l'exécutable, \fBxzdec\fP ne prend en charge ni le
32
+ multithreading ni la localisation et ne lit pas les options des variables
33
+ d'environnement \fBXZ_DEFAULTS\fP et \fBXZ_OPT\fP. \fBxzdec\fP ne gère pas
34
+ l'affichage d'information sur la progression du traitement : envoyer
35
+ \fBSIGINFO\fP à \fBxzdec\fP ne fait rien, mais envoyer \fBSIGUSR1\fP termine le
36
+ processus au lieu d'afficher de l'information sur la progression.
37
+ .SH OPTIONS
38
+ .TP
39
+ \fB\-d\fP, \fB\-\-decompress\fP, \fB\-\-uncompress\fP
40
+ Ignoré pour la compatibilité avec \fBxz\fP(1), \fBxzdec\fP ne gère que la
41
+ décompression.
42
+ .TP
43
+ \fB\-k\fP, \fB\-\-keep\fP
44
+ Ignoré pour la compatibilité avec \fBxz\fP(1), \fBxzdec\fP ne crée ni ne supprime
45
+ jamais aucun fichier.
46
+ .TP
47
+ \fB\-c\fP, \fB\-\-stdout\fP, \fB\-\-to\-stdout\fP
48
+ Ignoré pour la compatibilité avec \fBxz\fP(1), \fBxzdec\fP écrit toujours les
49
+ données décompressées sur la sortie standard.
50
+ .TP
51
+ \fB\-q\fP, \fB\-\-quiet\fP
52
+ Spécifier cela une fois ne fait rien, car \fBxzdec\fP n'affiche jamais aucun
53
+ avertissement ou notification. Spécifier cela deux fois pour supprimer les
54
+ erreurs.
55
+ .TP
56
+ \fB\-Q\fP, \fB\-\-no\-warn\fP
57
+ Ignoré pour la compatibilité avec \fBxz\fP(1), \fBxzdec\fP n'utilise jamais le
58
+ satut de sortie 2.
59
+ .TP
60
+ \fB\-h\fP, \fB\-\-help\fP
61
+ Afficher un message d'aide et quitter.
62
+ .TP
63
+ \fB\-V\fP, \fB\-\-version\fP
64
+ Afficher le numéro de version de \fBxzdec\fP et liblzma.
65
+ .SH "STATUT DE SORTIE"
66
+ .TP
67
+ \fB0\fP
68
+ Tout s'est bien passé.
69
+ .TP
70
+ \fB1\fP
71
+ Une erreur est survenue.
72
+ .PP
73
+ A la différence de \fBxz\fP(1),\fBxzdec\fP n'a pas de messages d'avertissement, et
74
+ donc le statut de sortie 2 n'est pas utilisé par \fBxzdec\fP.
75
+ .SH NOTES
76
+ Utilisez \fBxz\fP(1) au lieu de \fBxzdec\fP ou \fBlzmadec\fP pour un usage normal de
77
+ tous les jours. \fBxzdec\fP ou \fBlzmadec\fP ne sont utiles que pour les
78
+ situations où il est important d'avoir un plus petit décompresseur que le
79
+ \fBxz\fP(1) complet.
80
+ .PP
81
+ \fBxzdec\fP et \fBlzmadec\fP ne sont en fait pas vraiment si petits. La taille
82
+ peut être encore réduite en abandonnant des fonctionnalités de liblzma au
83
+ moment de la compilation, mais cela ne devrait pas être fait pour des
84
+ exécutables distribués sur des systèmes d'exploitation classique non
85
+ embarqués. Si vous avez besoin d'un décompresseur vraiment petit, pensez à
86
+ utiliser XZ Embedded.
87
+ .SH "VOIR AUSSI"
88
+ \fBxz\fP(1)
89
+ .PP
90
+ XZ Embedded: <https://tukaani.org/xz/embedded.html>
janus/share/man/fr/man1/lzmainfo.1 ADDED
@@ -0,0 +1,47 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ .\" SPDX-License-Identifier: 0BSD
2
+ .\"
3
+ .\" Author: Lasse Collin
4
+ .\"
5
+ .\" french translation of XZ Utils man
6
+ .\" Copyright (C) 2021 Debian French l10n team <debian-l10n-french@lists.debian.org>
7
+ .\" Translator
8
+ .\" bubu <bubub@no-log.org>, 2021.
9
+ .\"
10
+ .\"*******************************************************************
11
+ .\"
12
+ .\" This file was generated with po4a. Translate the source file.
13
+ .\"
14
+ .\"*******************************************************************
15
+ .TH LZMAINFO 1 30\-06\-2013 Tukaani "Utilitaires XZ"
16
+ .SH NOM
17
+ lzmainfo \- Afficher l'information stockée dans l'en\-tête du fichier .lzma
18
+ .SH SYNOPSIS
19
+ \fBlzmainfo\fP [\fB\-\-help\fP] [\fB\-\-version\fP] [\fIfichier...\fP]
20
+ .SH DESCRIPTION
21
+ \fBlzmainfo\fP affiche l'information stockée dans l'en\-tête du fichier
22
+ \&\fB.lzma\fP. Il lit les 13 premiers octets du \fIfichier\fP indiqué, décode
23
+ l'en\-tête, et l'écrit sur la sortie standard dans un format lisible par un
24
+ humain. Si aucun \fIfichier\fP n'est spécifié ou si \fIfichier\fP est \fB\-\fP,
25
+ l'entrée standard est lue.
26
+ .PP
27
+ Habituellement l'information la plus interressante est la taille non
28
+ compressée et la taille du dictionnaire. La taille ne peut être affichée que
29
+ si le fichier est dans la variante du format \fB.lzma\fP qui n'est pas en
30
+ flux. La quantité de mémoire requise pour décompresser le fichier est de
31
+ quelques douzaines de kilooctets en plus de la taille du dictionnaire.
32
+ .PP
33
+ \fBlzmainfo\fP est inclus dans les utilitaires XZ essentiellement pour des
34
+ besoins de rétrocompatibilité avec les utilitaires LZMA.
35
+ .SH "STATUT DE SORTIE"
36
+ .TP
37
+ \fB0\fP
38
+ Tout est bon.
39
+ .TP
40
+ \fB1\fP
41
+ Une erreur est survenue.
42
+ .SH BOGUES
43
+ \fBlzmainfo\fP utilise \fBMB\fP alors que le suffixe correct devrait être
44
+ \fBMiB\fP(2^20 octets). Cela pour garder la sortie compatible avec les
45
+ utilitaires LZMA.
46
+ .SH "VOIR AUSSI"
47
+ \fBxz\fP(1)
janus/share/man/fr/man1/unlzma.1 ADDED
@@ -0,0 +1,1985 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ '\" t
2
+ .\" SPDX-License-Identifier: 0BSD
3
+ .\"
4
+ .\" Authors: Lasse Collin
5
+ .\" Jia Tan
6
+ .\"
7
+ .\" french translation of XZ Utils man
8
+ .\" Copyright (C) 2021 Debian French l10n team <debian-l10n-french@lists.debian.org>
9
+ .\" Translator
10
+ .\" bubu <bubub@no-log.org>, 2021.
11
+ .\"
12
+ .\"*******************************************************************
13
+ .\"
14
+ .\" This file was generated with po4a. Translate the source file.
15
+ .\"
16
+ .\"*******************************************************************
17
+ .TH XZ 1 2024\-12\-30 Tukaani "Utilitaires XZ"
18
+ .
19
+ .SH NOM
20
+ xz, unxz, xzcat, lzma, unlzma, lzcat \- Compresser ou décompresser des
21
+ fichiers .xz et .lzma
22
+ .
23
+ .SH SYNOPSIS
24
+ \fBxz\fP [\fIoption...\fP] [\fIfichier...\fP]
25
+ .
26
+ .SH "ALIAS DES COMMANDES"
27
+ \fBunxz\fP est équivalent à \fBxz \-\-decompress\fP.
28
+ .br
29
+ \fBxzcat\fP est équivalent à \fBxz \-\-decompress \-\-stdout\fP
30
+ .br
31
+ \fBlzma\fP est équivalent à \fBxz \-\-format=lzma\fP
32
+ .br
33
+ \fBunlzma\fP est équivalent à \fBxz \-\-format=lzma \-\-decompress\fP
34
+ .br
35
+ \fBlzcat\fP est équivalent à \fBxz \-\-format=lzma \-\-decompress \-\- stdout\fP
36
+ .PP
37
+ Lors de l'écriture de scripts qui nécessitent de décompresser des fichiers,
38
+ il est recommandé de toujours utiliser la commande \fBxz\fP avec les arguments
39
+ appropriés (\fBxz \-d\fP ou \fBxz \-dc\fP) au lieu des commandes \fBunxz\fP et
40
+ \fBxzcat\fP.
41
+ .
42
+ .SH DESCRIPTION
43
+ \fBxz\fP is a general\-purpose data compression tool with command line syntax
44
+ similar to \fBgzip\fP(1) and \fBbzip2\fP(1). The native file format is the
45
+ \&\fB.xz\fP format, but the legacy \fB.lzma\fP format used by LZMA Utils and raw
46
+ compressed streams with no container format headers are also supported. In
47
+ addition, decompression of the \fB.lz\fP format used by \fBlzip\fP is supported.
48
+ .PP
49
+ \fBxz\fP compresse ou décompresse chaque \fIfichier\fP en fonction du mode
50
+ d'opération choisi. Si aucun \fIfichier\fP n'est donné ou \fIfichier\fP est \fB\-\fP,
51
+ \fBxz\fP lit depuis l'entrée standard et écrit les données traitées sur la
52
+ sortie standard. \fBxz\fP refusera (affichera une erreur et ignorera le
53
+ \fIfichier\fP) d'écrire les données compressées sur la sortie standard si c'est
54
+ un terminal. De même, \fBxz\fP refusera de lire des données compressées depuis
55
+ l'entrée standard si c'est un terminal.
56
+ .PP
57
+ A moins que \fB\-\-sdout\fP ne soit indiqué, les \fIfichiers\fP autres que \fB\-\fP sont
58
+ écrits dans un nouveau fichier dont le nom est dérivé du nom de \fIfichier\fP
59
+ source :
60
+ .IP \(bu 3
61
+ Lors de la compression, le suffixe du format de fichier cible (\fB.xz\fP ou
62
+ \&\fB.lzma\fP) est ajouté au nom de fichier source pour obtenir le nom de fichier
63
+ cible.
64
+ .IP \(bu 3
65
+ When decompressing, the \fB.xz\fP, \fB.lzma\fP, or \fB.lz\fP suffix is removed from
66
+ the filename to get the target filename. \fBxz\fP also recognizes the suffixes
67
+ \&\fB.txz\fP and \fB.tlz\fP, and replaces them with the \fB.tar\fP suffix.
68
+ .PP
69
+ Si le fichier cible existe déjà, une erreur est affichée et le \fIfichier\fP
70
+ est ignoré.
71
+ .PP
72
+ Sauf s'il écrit dans la sortie standard, \fBxz\fP affichera un avertissement et
73
+ ignorera le \fIfichier\fP dans les cas suivants :
74
+ .IP \(bu 3
75
+ \fIfichier\fP n'est pas un fichier normal. Les liens symboliques ne sont pas
76
+ suivis et donc ne sont pas considérés comme des fichiers normaux.
77
+ .IP \(bu 3
78
+ \fIfichier\fP a plusieurs liens physiques.
79
+ .IP \(bu 3
80
+ \fIfichier\fP a un setuid, setgid ou sticky bit positionné.
81
+ .IP \(bu 3
82
+ Le mode d'opération est défini pour compresser et le \fIfichier\fP a déjà un
83
+ suffixe du format de fichier cible (\fB.xz\fP ou \fB.txz\fP lors d'une compression
84
+ en format \fB.xz\fP, et \fB.lzma\fP ou \fB.tlz\fP lors d'une compression en format
85
+ \&\fB.lzma\fP).
86
+ .IP \(bu 3
87
+ The operation mode is set to decompress and the \fIfile\fP doesn't have a
88
+ suffix of any of the supported file formats (\fB.xz\fP, \fB.txz\fP, \fB.lzma\fP,
89
+ \&\fB.tlz\fP, or \fB.lz\fP).
90
+ .PP
91
+ Après la compression ou la décompression réussie du \fIfichier\fP, \fBxz\fP copie
92
+ les permissions du propriétaire, du groupe, la date d'accès, et les
93
+ modifications d'heure depuis le \fIfichier\fP source du fichier cible. Si la
94
+ copie du groupe échoue, les permissions sont modifiées pour que le fichier
95
+ cible ne soit pas accessible aux utilisateurs qui n'ont pas les droits
96
+ d'accès au \fIfichier\fP source. \fBxz\fP ne prend actuellement pas en charge la
97
+ copie d'autres métadonnées telles que les listes de contrôle d'accès ou les
98
+ attributs étendus.
99
+ .PP
100
+ Once the target file has been successfully closed, the source \fIfile\fP is
101
+ removed unless \fB\-\-keep\fP was specified. The source \fIfile\fP is never removed
102
+ if the output is written to standard output or if an error occurs.
103
+ .PP
104
+ Envoyer \fBSIGINFO\fP ou \fBSIGURSR1\fP au processus \fBxz\fP, lui fait afficher
105
+ l'information de progression sur l'erreur standard. Cela a un intérêt limité
106
+ car lorsque l'erreur standard est un terminal, utiliser \fB\-\-verbose\fP
107
+ affichera automatiquement un indicateur de progression du processus.
108
+ .
109
+ .SS "Utilisation de la mémoire"
110
+ L'utilisation de la mémoire par \fBxz\fP varie de quelques centaines de
111
+ kilo\-octets à plusieurs gigaoctects en fonction des paramètres de
112
+ compression. Les réglages utilisés lors de la compression d'un fichier
113
+ déterminent les besoins en mémoire pour la décompression. Habituellement la
114
+ décompression nécessite 5\% à 20\% de la quantité de mémoire utilisée pour
115
+ la compression du fichier. Par exemple, décompresser un fichier créé avec
116
+ \fBxz\-9\fP recquiert habituellement 65\ Mio de mémoire. Bien qu'il soit
117
+ possible d'avoir des fichiers \fB.xz\fP nécessitant plusieurs gigaoctets de
118
+ mémoire pour être décompressés.
119
+ .PP
120
+ Especially users of older systems may find the possibility of very large
121
+ memory usage annoying. To prevent uncomfortable surprises, \fBxz\fP has a
122
+ built\-in memory usage limiter, which is disabled by default. While some
123
+ operating systems provide ways to limit the memory usage of processes,
124
+ relying on it wasn't deemed to be flexible enough (for example, using
125
+ \fBulimit\fP(1) to limit virtual memory tends to cripple \fBmmap\fP(2)).
126
+ .PP
127
+ The memory usage limiter can be enabled with the command line option
128
+ \fB\-\-memlimit=\fP\fIlimit\fP. Often it is more convenient to enable the limiter
129
+ by default by setting the environment variable \fBXZ_DEFAULTS\fP, for example,
130
+ \fBXZ_DEFAULTS=\-\-memlimit=150MiB\fP. It is possible to set the limits
131
+ separately for compression and decompression by using
132
+ \fB\-\-memlimit\-compress=\fP\fIlimit\fP and \fB\-\-memlimit\-decompress=\fP\fIlimit\fP.
133
+ Using these two options outside \fBXZ_DEFAULTS\fP is rarely useful because a
134
+ single run of \fBxz\fP cannot do both compression and decompression and
135
+ \fB\-\-memlimit=\fP\fIlimit\fP (or \fB\-M\fP \fIlimit\fP) is shorter to type on the
136
+ command line.
137
+ .PP
138
+ If the specified memory usage limit is exceeded when decompressing, \fBxz\fP
139
+ will display an error and decompressing the file will fail. If the limit is
140
+ exceeded when compressing, \fBxz\fP will try to scale the settings down so that
141
+ the limit is no longer exceeded (except when using \fB\-\-format=raw\fP or
142
+ \fB\-\-no\-adjust\fP). This way the operation won't fail unless the limit is very
143
+ small. The scaling of the settings is done in steps that don't match the
144
+ compression level presets, for example, if the limit is only slightly less
145
+ than the amount required for \fBxz \-9\fP, the settings will be scaled down only
146
+ a little, not all the way down to \fBxz \-8\fP.
147
+ .
148
+ .SS "Concaténation et remplissage avec des fichiers .xz"
149
+ Il est possible de concaténer les fichiers \fB.xz\fP tels quel. \fBxz\fP
150
+ décompressera de tels fichiers comme s'ils étaient un unique fichier \fB.xz\fP.
151
+ .PP
152
+ It is possible to insert padding between the concatenated parts or after the
153
+ last part. The padding must consist of null bytes and the size of the
154
+ padding must be a multiple of four bytes. This can be useful, for example,
155
+ if the \fB.xz\fP file is stored on a medium that measures file sizes in
156
+ 512\-byte blocks.
157
+ .PP
158
+ La concaténation et le remplissage ne sont pas autorisés avec les fichiers
159
+ \&\fB.lzma\fP ou les flux bruts.
160
+ .
161
+ .SH OPTIONS
162
+ .
163
+ .SS "Suffixes entiers et valeurs spéciales."
164
+ Dans la plupart des endroits où un argument entier est attendu, un suffixe
165
+ optionel permet d'indiquer facilement les grands entiers. Il ne doit pas y
166
+ avoir d'espace entre l'entier et le suffixe.
167
+ .TP
168
+ \fBKiB\fP
169
+ Multiplier l'entier par 1024 (2^10). \fBKi\fP, \fBk\fP, \fBkB\fP, \fBK\fP et \fBKB\fP sont
170
+ acceptés comme synonymes de \fBKiB\fP.
171
+ .TP
172
+ \fBMiB\fP
173
+ Multiplier l'entier par 1 048 576 (2^20). \fBMi\fP, \fBm\fP, \fBM\fP et \fBMB\fP sont
174
+ acceptés comme synonymes de \fBMiB\fP.
175
+ .TP
176
+ \fBGiB\fP
177
+ Multiplier l'entier par 1 073 741 824 (2^30). \fBGi\fP, \fBg\fP, \fBG\fP et \fBGB\fP
178
+ sont acceptés comme synonymes de \fBGiB\fP.
179
+ .PP
180
+ La valeur spéciale \fBmax\fP peut être utilisée pour indiquer la valeur
181
+ maximale de l'entier prise en charge par l'option.
182
+ .
183
+ .SS "Mode d'opération"
184
+ Si plusieurs options de mode d'opération sont données, la dernière prend
185
+ effet.
186
+ .TP
187
+ \fB\-z\fP, \fB\-\-compress\fP
188
+ Compresser. C'est le mode d'opération par défaut lorsque aucune option de
189
+ mode opératoire n'est spécifiée ou qu'aucun autre mode d'opération n'est
190
+ sous\-entendu par le nom de la commande (par exemple \fBunxz\fP sous\-entend
191
+ \fB\-\-decompress\fP).
192
+ .IP ""
193
+ .\" The DESCRIPTION section already says this but it's good to repeat it
194
+ .\" here because the default behavior is a bit dangerous and new users
195
+ .\" in a hurry may skip reading the DESCRIPTION section.
196
+ After successful compression, the source file is removed unless writing to
197
+ standard output or \fB\-\-keep\fP was specified.
198
+ .TP
199
+ \fB\-d\fP, \fB\-\-decompress\fP, \fB\-\-uncompress\fP
200
+ .\" The DESCRIPTION section already says this but it's good to repeat it
201
+ .\" here because the default behavior is a bit dangerous and new users
202
+ .\" in a hurry may skip reading the DESCRIPTION section.
203
+ Decompress. After successful decompression, the source file is removed
204
+ unless writing to standard output or \fB\-\-keep\fP was specified.
205
+ .TP
206
+ \fB\-t\fP, \fB\-\-test\fP
207
+ Tester l'intégrité des \fIfichiers\fP compressés. Cette option est équivalente
208
+ à \fB\-\-decompress \-\-stdout\fP sauf que les données décompressées sont rejetées
209
+ au lieu d'être écrites sur la sortie standard. Aucun fichier n'est créé ou
210
+ supprimé.
211
+ .TP
212
+ \fB\-l\fP, \fB\-\-list\fP
213
+ Afficher l'information sur les \fIfichiers\fP compressés. Aucune sortie
214
+ non\-compressée n'est produite et aucun fichier n'est créé ou supprimé. En
215
+ mode liste, le programme ne peut pas lire les données compressées depuis
216
+ l'entrée standard ou depuis d'autres sources non adressables.
217
+ .IP ""
218
+ The default listing shows basic information about \fIfiles\fP, one file per
219
+ line. To get more detailed information, use also the \fB\-\-verbose\fP option.
220
+ For even more information, use \fB\-\-verbose\fP twice, but note that this may be
221
+ slow, because getting all the extra information requires many seeks. The
222
+ width of verbose output exceeds 80 characters, so piping the output to, for
223
+ example, \fBless\ \-S\fP may be convenient if the terminal isn't wide enough.
224
+ .IP ""
225
+ La sortie exacte peut varier suivant les versions de \fBxz\fP et les différents
226
+ paramètres régionaux. Pour une sortie lisible par la machine, utiliser
227
+ \fB\-\-robot \-\-list\fP.
228
+ .
229
+ .SS "Modificateurs d'opération"
230
+ .TP
231
+ \fB\-k\fP, \fB\-\-keep\fP
232
+ Ne pas effacer les fichiers d'entrée.
233
+ .IP ""
234
+ Since \fBxz\fP 5.2.6, this option also makes \fBxz\fP compress or decompress even
235
+ if the input is a symbolic link to a regular file, has more than one hard
236
+ link, or has the setuid, setgid, or sticky bit set. The setuid, setgid, and
237
+ sticky bits are not copied to the target file. In earlier versions this was
238
+ only done with \fB\-\-force\fP.
239
+ .TP
240
+ \fB\-f\fP, \fB\-\-force\fP
241
+ Cette option a plusieurs effets :
242
+ .RS
243
+ .IP \(bu 3
244
+ Si le fichier cible existe déjà, l'effacer avant de compresser ou
245
+ décompresser.
246
+ .IP \(bu 3
247
+ Compresser ou décompresser même si l'entrée est un lien symbolique vers un
248
+ fichier normal, a plus qu'un lien physique, ou a le bit setuid, setgid ou
249
+ sticky défini. Les bits setuid, setgid et sticky bits ne sont pas copiés
250
+ dans le fichier cible.
251
+ .IP \(bu 3
252
+ Lorsque \fBxz\fP est utilisé avec \fB\-\-decompress\fP \fB\-\-stdout\fP et qu'il ne peut
253
+ pas reconnaitre le type du fichier source, copier le fichier source tel quel
254
+ dans la sortie standard. Celà permet à \fBxzcat\fP \fB\-\-force\fP d'être utilisé
255
+ comme \fBcat\fP(1) pour les fichiers qui n'ont pas été compressé avec
256
+ \fBxz\fP. Remarquez que dans le futur, \fBxz\fP devrait prendre en charge de
257
+ nouveaux formats de fichiers compressés, ce qui permettra à \fBxz\fP de
258
+ décompresser plus de types de fichiers au lieu de les copier tels quels dans
259
+ la sortie standard. \fB\-\-format=\fP\fIformat\fP peut être utilisé pour contraindre
260
+ \fBxz\fP à décompresser seulement un format de fichier.
261
+ .RE
262
+ .TP
263
+ \fB\-c\fP, \fB\-\-stdout\fP, \fB\-\-to\-stdout\fP
264
+ Écrire les données compressées ou décompressées sur la sortie standard
265
+ plutôt que dans un fichier. Cela necessite \fB\-\-keep\fP.
266
+ .TP
267
+ \fB\-\-single\-stream\fP
268
+ Décompresser seulement le premier flux \fB.xz\fP et ignorer silencieusement les
269
+ possibles données d'entrée résiduelles qui suivent le flux. Normalement ces
270
+ déchets excédentaires provoquent l'affichage d'une erreur par \fBxz\fP.
271
+ .IP ""
272
+ \fBxz\fP ne décompresse jamais plus d'un flux à partir de fichiers \fB.lzma\fP ou
273
+ de flux bruts, mais cette option fait aussi que \fBxz\fP ignorera les données
274
+ résiduelles après le fichier \fB.lzma\fP ou le flux brut.
275
+ .IP ""
276
+ Cette option n'a aucun effet si le mode d'opération n'est pas
277
+ \fB\-\-decompress\fP ou \fB\-\-test\fP.
278
+ .TP
279
+ \fB\-\-no\-sparse\fP
280
+ Désactiver la création de fichiers peu denses. Par défaut, lors de la
281
+ décompression en un fichier normal, \fBxz\fP essaie d'en faire un fichier creux
282
+ si les données décompressées contiennent de longues séquences de zéros
283
+ binaires. Cela fonctionne aussi lors de l'écriture sur la sortie standard
284
+ aussi longtemps que la sortie standard est connectée à un fichier normal et
285
+ que certaines conditions supplémentaires sont satisfaites pour le faire de
286
+ manière sécurisée. Créer des fichiers creux peut épargner de l'espace disque
287
+ et accélérer la décompression en réduisant la quantité d'entrées/sorties sur
288
+ le disque.
289
+ .TP
290
+ \fB\-S\fP \fI.suf\fP, \fB\-\-suffix=\fP\fI.suf\fP
291
+ Lors de la compression, utiliser \fB.suf\fP comme suffixe du fichier cible au
292
+ lieu de \fB.xz\fP ou \fB.lzma\fP. Si \fBxz\fP n'écrit pas sur la sortie standard et
293
+ si le fichier source a déja le suffixe \fB.suf\fP, un avertissement est affiché
294
+ et le fichier est ignoré.
295
+ .IP ""
296
+ When decompressing, recognize files with the suffix \fI.suf\fP in addition to
297
+ files with the \fB.xz\fP, \fB.txz\fP, \fB.lzma\fP, \fB.tlz\fP, or \fB.lz\fP suffix. If the
298
+ source file has the suffix \fI.suf\fP, the suffix is removed to get the target
299
+ filename.
300
+ .IP ""
301
+ Lors de la compression ou décompression de flux bruts (\fB\-\-fomat=raw\fP), le
302
+ suffixe doit toujours être spécifié à moins d'écrire sur la sortie standard,
303
+ car il n'y a pas de suffixe par défaut pour les flux bruts.
304
+ .TP
305
+ \fB\-\-files\fP[\fB=\fP\fIfichier\fP]
306
+ Lire les noms de fichier à traiter depuis \fIfichier\fP ; si \fIfichier\fP est
307
+ omis , les noms de fichier sont lus sur l'entrée standard. Les noms de
308
+ fichier doivent se terminer avec le caractère de nouvelle ligne. Un tiret
309
+ (\fB\-\fP) est considéré comme un nom de fichier normal ; ce qui ne signifie pas
310
+ entrée standard. Si les noms de fichier sont aussi donnés comme arguments de
311
+ ligne de commande, ils sont traités avant les noms de fichier lus depuis
312
+ \fIfichier\fP.
313
+ .TP
314
+ \fB\-\-files0\fP[\fB=\fP\fIfichier\fP]
315
+ Cela est identique à \fB\-\-files\fP[\fB=\fP\fIfichier\fP] sauf que chaque nom de
316
+ fichier doit se terminer par le caractère null.
317
+ .
318
+ .SS "Format de fichier basique et options de compression"
319
+ .TP
320
+ \fB\-F\fP \fIformat\fP, \fB\-\-format=\fP\fIformat\fP
321
+ Indiquer le \fIformat\fP de fichier à compresser ou décompresser :
322
+ .RS
323
+ .TP
324
+ \fBauto\fP
325
+ C'est celui par défaut. Lors de la compression, \fBauto\fP est équivalent à
326
+ \fBxz\fP. Lors de la décompression, le format du fichier en entrée est détecté
327
+ automatiquement. Notez que les flux bruts (créés avec \fB\-\-format=raw\fP) ne
328
+ peuvent pas être détectés automatiquement.
329
+ .TP
330
+ \fBxz\fP
331
+ Compresser dans le format de fichier \fB.xz\fP ou n'accepter que les fichiers
332
+ \&\fB.xz\fP à décompresser.
333
+ .TP
334
+ \fBlzma\fP, \fBalone\fP
335
+ Compresser au format de fichier \fB.lzma\fP historique, ou n'accepter que les
336
+ fichiers \fB.lzma\fP lors de la décompression. Le nom alternatif \fBalone\fP est
337
+ fourni pour la rétrocompatibilité avec les utilitaires LZMA.
338
+ .TP
339
+ \fBlzip\fP
340
+ Accept only \fB.lz\fP files when decompressing. Compression is not supported.
341
+ .IP ""
342
+ The \fB.lz\fP format version 0 and the unextended version 1 are supported.
343
+ Version 0 files were produced by \fBlzip\fP 1.3 and older. Such files aren't
344
+ common but may be found from file archives as a few source packages were
345
+ released in this format. People might have old personal files in this
346
+ format too. Decompression support for the format version 0 was removed in
347
+ \fBlzip\fP 1.18.
348
+ .IP ""
349
+ \fBlzip\fP 1.4 and later create files in the format version 1. The sync flush
350
+ marker extension to the format version 1 was added in \fBlzip\fP 1.6. This
351
+ extension is rarely used and isn't supported by \fBxz\fP (diagnosed as corrupt
352
+ input).
353
+ .TP
354
+ \fBraw\fP
355
+ Compresser ou décompresser un flux brut (sans en\-têtes). Cela est réservé
356
+ seulement aux utilisateurs aguerris. Pour décoder des flux bruts, vous devez
357
+ utiliser \fB\-\-format=raw\fP et spécifier explicitement la chaîne de filtre, qui
358
+ normalement aurait du être stockée dans les en\-têtes du conteneur.
359
+ .RE
360
+ .TP
361
+ \fB\-C\fP \fIvérif.\fP, \fB\-\-check=\fP\fIvérif.\fP
362
+ Spécifier le type d'intégrité à vérifier. La vérification est calculée à
363
+ partir des données non\-compressées et stockées dans le fichier \fB.xz\fP. Cette
364
+ option n'a effet que si la compression a été faite dans le format \fB.xz\fP ;
365
+ le format \fB.lzma\fP ne gère pas les vérifications d'intégrité. Le contrôle
366
+ d'intégrité (s'il y en a) est vérifié lorsque le fichier \fB.xz\fP est
367
+ décompressé.
368
+ .IP ""
369
+ Types de \fIvérification\fP pris en charge :
370
+ .RS
371
+ .TP
372
+ \fBnone\fP
373
+ Ne pas calculer de vérification d'intégrité du tout. C'est généralement une
374
+ mauvaise idée. Cela peut être utile lorsque l'intégrité des données est
375
+ vérifiée de toute façon par d'autres manières.
376
+ .TP
377
+ \fBcrc32\fP
378
+ Calculer CRC32 en utilisant le polynôme de IEEE\-802.3 (Ethernet).
379
+ .TP
380
+ \fBcrc64\fP
381
+ Calculer CRC64 en utilisant le polynôme de ECMA\-182. C'est la manière
382
+ utilisée par défaut, car c'est légèrement mieux que CRC32 pour détecter les
383
+ fichiers endommagés et la différence de vitesse est négligeable.
384
+ .TP
385
+ \fBsha256\fP
386
+ Calculer SHA\-256. C'est quelque peu plus lent que CRC32 et CRC64.
387
+ .RE
388
+ .IP ""
389
+ L'intégrité des en\-têtes \fB.xz\fP est toujours vérifiée avec CRC32. Il n'est
390
+ pas possible de le changer ou de le désactiver.
391
+ .TP
392
+ \fB\-\-ignore\-check\fP
393
+ Ne pas contrôler la vérification d'intégrité des données lors de la
394
+ décompression. Les valeurs CRC32 dans les en\-têtes \fB.xz\fP seront normalement
395
+ toujours vérifiées.
396
+ .IP ""
397
+ \fBN'utilisez pas cette option à moins de savoir ce que vous faites.\fP Les
398
+ raisons possibles pour utiliser cette option :
399
+ .RS
400
+ .IP \(bu 3
401
+ Essayer de récupérer des données d'un fichier .xz corrompu.
402
+ .IP \(bu 3
403
+ Accélérer la décompression. Cela importe surtout avec SHA\-256 ou avec les
404
+ fichiers qui ont été compressés extrêmement bien. Il est recommandé de ne
405
+ pas utiliser cette option dans ce but à moins que l'intégrité du fichier ne
406
+ soit vérifiée extérieurement d'une autre manière.
407
+ .RE
408
+ .TP
409
+ \fB\-0\fP ... \fB\-9\fP
410
+ Choisir un niveau de compression prédéfini. La valeur par défaut est
411
+ \fB6\fP. Si plusieurs niveaux de préréglage sont spécifi��s, c'est le dernier
412
+ qui sera pris en compte. Si une chaîne de filtres personnalisée a déjà été
413
+ choisie, définir un niveau de compression préréglé efface la chaîne de
414
+ filtres personnalisée.
415
+ .IP ""
416
+ Les différences entre les préréglages sont plus significatives qu'avec
417
+ \fBgzip\fP(1) et \fBbzip2\fP(1). les réglages de compression sélectionnés
418
+ déterminent les exigences en mémoire pour la décompression, ainsi, utiliser
419
+ un niveau de préréglage trop élevé peut rendre difficile à décompresser un
420
+ fichier sur un vieux système avec peu de RAM. Clairement, \fBce n'est pas une bonne idée d'utiliser \-9 aveuglément pour tout\fP comme ça l'est souvent avec
421
+ \fBgzip\fP(1) et \fBbzip2\fP(1).
422
+ .RS
423
+ .TP
424
+ \fB\-0\fP ... \fB\-3\fP
425
+ Ce sont des préréglages relativement rapides. \fB0\fP est parfois plus rapide
426
+ que \fBgzip \-9\fP tout en compressant bien mieux. Les réglages plus élevés ont
427
+ souvent une rapidité comparable à celle de \fBbzip2\fP(1) avec un taux de
428
+ compression comparable ou meilleur, même si les résultats dépendent beaucoup
429
+ du genre de données compressées.
430
+ .TP
431
+ \fB\-4\fP ... \fB\-6\fP
432
+ Good to very good compression while keeping decompressor memory usage
433
+ reasonable even for old systems. \fB\-6\fP is the default, which is usually a
434
+ good choice for distributing files that need to be decompressible even on
435
+ systems with only 16\ MiB RAM. (\fB\-5e\fP or \fB\-6e\fP may be worth considering
436
+ too. See \fB\-\-extreme\fP.)
437
+ .TP
438
+ \fB\-7 ... \-9\fP
439
+ C'est comme \fB\-6\fP mais avec des besoins en mémoire plus élevés pour la
440
+ compression et la décompression. Ce n'est utile que lorsque les fichiers
441
+ sont plus gros que 8\ Mio, 16\ Mio et 32\ Mio respectivement.
442
+ .RE
443
+ .IP ""
444
+ Sur le même matériel, la vitesse de décompression est sensiblement un nombre
445
+ constant d'octets de données compressées par seconde. En d'autres termes,
446
+ meilleure est la compression, plus rapide sera en général la
447
+ décompression. Cela signifie aussi que la quantité de sortie non compressée
448
+ produite par seconde peut varier beaucoup.
449
+ .IP ""
450
+ Le tableau suivant résume les caractéristiques des préréglages :
451
+ .RS
452
+ .RS
453
+ .PP
454
+ .TS
455
+ tab(;);
456
+ c c c c c
457
+ n n n n n.
458
+ Préréglage;DictSize;CompCPU;CompMem;DecMem
459
+ \-0;256 KiB;0;3 MiB;1 MiB
460
+ \-1;1 MiB;1;9 MiB;2 MiB
461
+ \-2;2 MiB;2;17 MiB;3 MiB
462
+ \-3;4 MiB;3;32 MiB;5 MiB
463
+ \-4;4 MiB;4;48 MiB;5 MiB
464
+ \-5;8 MiB;5;94 MiB;9 MiB
465
+ \-6;8 MiB;6;94 MiB;9 MiB
466
+ \-7;16 MiB;6;186 MiB;17 MiB
467
+ \-8;32 MiB;6;370 MiB;33 MiB
468
+ \-9;64 MiB;6;674 MiB;65 MiB
469
+ .TE
470
+ .RE
471
+ .RE
472
+ .IP ""
473
+ Descriptions des colonnes :
474
+ .RS
475
+ .IP \(bu 3
476
+ DictSize est la taille du dictionnaire de LZMA2. Utiliser un dictionnaire
477
+ plus gros que la taille du fichier non compressé est un gaspillage de
478
+ mémoire. C'est pourquoi il est bon d'éviter d'utiliser les préréglages de
479
+ \fB\-7\fP à \fB\-9\fP lorsqu'il n'y en a pas vraiment besoin. A \fB\-6\fP et plus bas,
480
+ la quantité de mémoire gaspillée est généralement assez basse pour ne pas
481
+ être un problème.
482
+ .IP \(bu 3
483
+ CompCPU est une représentation des préréglages de LZMA2 qui affectent la
484
+ vitesse de compression. La taille du dictionnaire aussi affecte la vitesse,
485
+ alors comme CompCPU est le même pour les niveaux de \fB\-6\fP à \fB\-9\fP, les plus
486
+ haut niveaux tendent à être un peu moins rapides. Pour être encore moins
487
+ rapide et du coup obtenir peut être une meilleure compression, consultez
488
+ \fB\-\-extreme\fP.
489
+ .IP \(bu 3
490
+ CompMem contains the compressor memory requirements in the single\-threaded
491
+ mode. It may vary slightly between \fBxz\fP versions.
492
+ .IP \(bu 3
493
+ DecMem contient les besoins en mémoire du décompresseur. Ce sont les
494
+ réglages de la compression qui déterminent les besoins en mémoire de la
495
+ décompression. L'exacte utilisation de la mémoire est légèrement supérieure
496
+ à la taille du dictionnaire LZMA2, mais les valeurs dans la table ont été
497
+ arrondies au prochain Mio supérieur.
498
+ .RE
499
+ .IP ""
500
+ Memory requirements of the multi\-threaded mode are significantly higher than
501
+ that of the single\-threaded mode. With the default value of
502
+ \fB\-\-block\-size\fP, each thread needs 3*3*DictSize plus CompMem or DecMem. For
503
+ example, four threads with preset \fB\-6\fP needs 660\(en670\ MiB of memory.
504
+ .TP
505
+ \fB\-e\fP, \fB\-\-extreme\fP
506
+ Utilisez un variant plus lent que les préréglages (\fB\-0\fP à \fB\-9\fP) pour
507
+ espérer avoir un taux de compression légèrement meilleur, mais en cas de
508
+ malchance cela peut être pire. L'utilisation mémoire du décompresseur n'est
509
+ pas affectée, mais l'utilisation mémoire du compresseur augmente un peu aux
510
+ niveaux de préréglages de \fB\-0\fP à \fB\-3\fP.
511
+ .IP ""
512
+ Depuis qu'il y a deux préréglages avec des tailles de dictionnaire de 4\ Mio
513
+ et 8 \Mio, les préréglages \fB\-3e\fP et \fB\-5e\fP utilisent des réglages
514
+ légèrement plus rapides que \fB\-4e\fP et \fB\-6e\fP, respectivement. De cette
515
+ manière, il n'y a pas deux préréglages identiques.
516
+ .RS
517
+ .RS
518
+ .PP
519
+ .TS
520
+ tab(;);
521
+ c c c c c
522
+ n n n n n.
523
+ Préréglage;DictSize;CompCPU;CompMem;DecMem
524
+ \-0e;256 KiB;8;4 MiB;1 MiB
525
+ \-1e;1 MiB;8;13 MiB;2 MiB
526
+ \-2e;2 MiB;8;25 MiB;3 MiB
527
+ \-3e;4 MiB;7;48 MiB;5 MiB
528
+ \-4e;4 MiB;8;48 MiB;5 MiB
529
+ \-5e;8 MiB;7;94 MiB;9 MiB
530
+ \-6e;8 MiB;8;94 MiB;9 MiB
531
+ \-7e;16 MiB;8;186 MiB;17 MiB
532
+ \-8e;32 MiB;8;370 MiB;33 MiB
533
+ \-9e;64 MiB;8;674 MiB;65 MiB
534
+ .TE
535
+ .RE
536
+ .RE
537
+ .IP ""
538
+ Par exemple, il y a un total de quatre préréglages qui utilisent un
539
+ dictionnaire de 8 Mio et qui sont dans l'ordre du plus rapide au plus lent :
540
+ \fB\-5\fP, \fB\-6\fP, \fB\-5e\fP et \fB\-6e\fP.
541
+ .TP
542
+ \fB\-\-fast\fP
543
+ .PD 0
544
+ .TP
545
+ \fB\-\-best\fP
546
+ .PD
547
+ Il y a néanmoins des alias trompeurs pour \fB\-0\fP et \fB\-9\fP,
548
+ respectivement. Ils ne sont fournis que pour des besoins de
549
+ rétro\-compatibilité avec les utilitaires LZMA. Evitez d'utiliser ces
550
+ options.
551
+ .TP
552
+ \fB\-\-block\-size=\fP\fItaille\fP
553
+ Lors de la compression dans le format \fB.xz\fP, les données de l'entrée sont
554
+ réparties en blocs de \fItaille\fP octets. Les blocs sont compressés
555
+ indépendamment les un des autres, ce qui aide avec le mode multithread
556
+ (multi\-threading) et rend possible la décompression à accès aléatoire
557
+ limité. Cette option est typiquement utilisée pour outrepasser la taille de
558
+ bloc en mode multithread, mais cette option peut aussi être utilisée en mode
559
+ single\-thread.
560
+ .IP ""
561
+ In multi\-threaded mode about three times \fIsize\fP bytes will be allocated in
562
+ each thread for buffering input and output. The default \fIsize\fP is three
563
+ times the LZMA2 dictionary size or 1 MiB, whichever is more. Typically a
564
+ good value is 2\(en4 times the size of the LZMA2 dictionary or at least 1
565
+ MiB. Using \fIsize\fP less than the LZMA2 dictionary size is waste of RAM
566
+ because then the LZMA2 dictionary buffer will never get fully used. In
567
+ multi\-threaded mode, the sizes of the blocks are stored in the block
568
+ headers. This size information is required for multi\-threaded
569
+ decompression.
570
+ .IP ""
571
+ In single\-threaded mode no block splitting is done by default. Setting this
572
+ option doesn't affect memory usage. No size information is stored in block
573
+ headers, thus files created in single\-threaded mode won't be identical to
574
+ files created in multi\-threaded mode. The lack of size information also
575
+ means that \fBxz\fP won't be able decompress the files in multi\-threaded mode.
576
+ .TP
577
+ \fB\-\-block\-list=\fP\fIitems\fP
578
+ When compressing to the \fB.xz\fP format, start a new block with an optional
579
+ custom filter chain after the given intervals of uncompressed data.
580
+ .IP ""
581
+ The \fIitems\fP are a comma\-separated list. Each item consists of an optional
582
+ filter chain number between 0 and 9 followed by a colon (\fB:\fP) and a
583
+ required size of uncompressed data. Omitting an item (two or more
584
+ consecutive commas) is a shorthand to use the size and filters of the
585
+ previous item.
586
+ .IP ""
587
+ If the input file is bigger than the sum of the sizes in \fIitems\fP, the last
588
+ item is repeated until the end of the file. A special value of \fB0\fP may be
589
+ used as the last size to indicate that the rest of the file should be
590
+ encoded as a single block.
591
+ .IP ""
592
+ An alternative filter chain for each block can be specified in combination
593
+ with the \fB\-\-filters1=\fP\fIfilters\fP \&...\& \fB\-\-filters9=\fP\fIfilters\fP options.
594
+ These options define filter chains with an identifier between 1\(en9.
595
+ Filter chain 0 can be used to refer to the default filter chain, which is
596
+ the same as not specifying a filter chain. The filter chain identifier can
597
+ be used before the uncompressed size, followed by a colon (\fB:\fP). For
598
+ example, if one specifies \fB\-\-block\-list=1:2MiB,3:2MiB,2:4MiB,,2MiB,0:4MiB\fP
599
+ then blocks will be created using:
600
+ .RS
601
+ .IP \(bu 3
602
+ The filter chain specified by \fB\-\-filters1\fP and 2 MiB input
603
+ .IP \(bu 3
604
+ The filter chain specified by \fB\-\-filters3\fP and 2 MiB input
605
+ .IP \(bu 3
606
+ The filter chain specified by \fB\-\-filters2\fP and 4 MiB input
607
+ .IP \(bu 3
608
+ The filter chain specified by \fB\-\-filters2\fP and 4 MiB input
609
+ .IP \(bu 3
610
+ The default filter chain and 2 MiB input
611
+ .IP \(bu 3
612
+ The default filter chain and 4 MiB input for every block until end of input.
613
+ .RE
614
+ .IP ""
615
+ If one specifies a size that exceeds the encoder's block size (either the
616
+ default value in threaded mode or the value specified with
617
+ \fB\-\-block\-size=\fP\fIsize\fP), the encoder will create additional blocks while
618
+ keeping the boundaries specified in \fIitems\fP. For example, if one specifies
619
+ \fB\-\-block\-size=10MiB\fP \fB\-\-block\-list=5MiB,10MiB,8MiB,12MiB,24MiB\fP and the
620
+ input file is 80 MiB, one will get 11 blocks: 5, 10, 8, 10, 2, 10, 10, 4,
621
+ 10, 10, and 1 MiB.
622
+ .IP ""
623
+ En mode multi\-threadé les tailles de blocs sont stockées dans les en\-têtes
624
+ du bloc. Cela ne se fait pas en mode mono\-threadé, la sortie encodée ne sera
625
+ donc pas identique à celle faite en mode multi\-threadé.
626
+ .TP
627
+ \fB\-\-flush\-timeout=\fP\fItemps_d'attente\fP
628
+ Lors de la compression, si plus que \fItemps_d'attente\fP millisecondes (un
629
+ entier positif) se sont écoulées depuis le précédent vidage et que lire plus
630
+ de données bloquerait, toutes les données d'entrée en attente sont vidées de
631
+ l'encodeur et mises à disposition dans le flux de sortie. Cela peut être
632
+ utile si \fBxz\fP est utilisé pour compresser les données qui sont diffusées
633
+ sur un réseau. Des petites valeurs de \fItemps_d'attente\fP rendent les données
634
+ disponibles à l'extrémité réceptrice avec un léger retard, mais les grandes
635
+ valeurs de \fItemps_d'attente\fP donnent un meilleur taux de compression.
636
+ .IP ""
637
+ Cette option est désactivée par défaut. Si cette option est indiquée plus
638
+ d'une fois, la dernière prend effet. La valeur spéciale de
639
+ \fItemps_d'attente\fP de \fB0\fP peut être utilisée pour explicitement désactiver
640
+ cette option.
641
+ .IP ""
642
+ Cette option n'est pas disponible sur les systèmes qui ne sont pas POSIX.
643
+ .IP ""
644
+ .\" FIXME
645
+ \fBCette option est encore expérimentale.\fP Actuellement, \fBxz\fP ne convient
646
+ pas pour décompresser le flux en temps réel en raison de la façon dont \fBxz\fP
647
+ effectue la mise en mémoire tampon.
648
+ .TP
649
+ \fB\-\-memlimit\-compress=\fP\fIlimite\fP
650
+ Indiquer une limite d'utilisation de la mémoire pour la compression. Si
651
+ cette option est indiquée plusieurs fois, c'est la dernière qui est prise en
652
+ compte.
653
+ .IP ""
654
+ If the compression settings exceed the \fIlimit\fP, \fBxz\fP will attempt to
655
+ adjust the settings downwards so that the limit is no longer exceeded and
656
+ display a notice that automatic adjustment was done. The adjustments are
657
+ done in this order: reducing the number of threads, switching to
658
+ single\-threaded mode if even one thread in multi\-threaded mode exceeds the
659
+ \fIlimit\fP, and finally reducing the LZMA2 dictionary size.
660
+ .IP ""
661
+ When compressing with \fB\-\-format=raw\fP or if \fB\-\-no\-adjust\fP has been
662
+ specified, only the number of threads may be reduced since it can be done
663
+ without affecting the compressed output.
664
+ .IP ""
665
+ If the \fIlimit\fP cannot be met even with the adjustments described above, an
666
+ error is displayed and \fBxz\fP will exit with exit status 1.
667
+ .IP ""
668
+ La \fIlimite\fP peut être indiquée de plusieurs façons :
669
+ .RS
670
+ .IP \(bu 3
671
+ La \fIlimite\fP peut être une valeur absolue en octets. Utiliser un suffixe
672
+ d'entier comme \fBMiB\fP peut être utile. Exemple :
673
+ \fB\-\-memlimit\-compress=80MiB\fP
674
+ .IP \(bu 3
675
+ La \fIlimite\fP peut être indiquée sous forme d'un pourcentage de la mémoire
676
+ physique totale (RAM). Cela peut être particulièrement utile quand la
677
+ variable d'environnement \fBXZ_DEFAULTS\fP est indiquée dans un script
678
+ d'initialisation de l'interpréteur partagé entre différents ordinateurs. De
679
+ cette façon la limite est automatiquement plus grande sur les systèmes avec
680
+ plus de mémoire. Exemple : \fB\-\-memlimit=70%\fP
681
+ .IP \(bu 3
682
+ The \fIlimit\fP can be reset back to its default value by setting it to \fB0\fP.
683
+ This is currently equivalent to setting the \fIlimit\fP to \fBmax\fP (no memory
684
+ usage limit).
685
+ .RE
686
+ .IP ""
687
+ For 32\-bit \fBxz\fP there is a special case: if the \fIlimit\fP would be over
688
+ \fB4020\ MiB\fP, the \fIlimit\fP is set to \fB4020\ MiB\fP. On MIPS32 \fB2000\ MiB\fP
689
+ is used instead. (The values \fB0\fP and \fBmax\fP aren't affected by this. A
690
+ similar feature doesn't exist for decompression.) This can be helpful when
691
+ a 32\-bit executable has access to 4\ GiB address space (2 GiB on MIPS32)
692
+ while hopefully doing no harm in other situations.
693
+ .IP ""
694
+ Voir aussi la section \fButilisation de la mémoire\fP.
695
+ .TP
696
+ \fB\-\-memlimit\-decompress=\fP\fIlimite\fP
697
+ Régler une limite d'utilisation de la mémoire pour la décompression. Cela a
698
+ un effet sur le mode \fB\-\-list\fP. Si l'opération n'est pas possible sans
699
+ dépasser la \fIlimite\fP, \fBxz\fP affichera une erreur et la décompression
700
+ échouera. Voir \fB\-\-memlimit\-compress=\fP\fIlimite\fP pour les manières possibles
701
+ d'indiquer la \fIlimite\fP.
702
+ .TP
703
+ \fB\-\-memlimit\-mt\-decompress=\fP\fIlimit\fP
704
+ Set a memory usage limit for multi\-threaded decompression. This can only
705
+ affect the number of threads; this will never make \fBxz\fP refuse to
706
+ decompress a file. If \fIlimit\fP is too low to allow any multi\-threading, the
707
+ \fIlimit\fP is ignored and \fBxz\fP will continue in single\-threaded mode. Note
708
+ that if also \fB\-\-memlimit\-decompress\fP is used, it will always apply to both
709
+ single\-threaded and multi\-threaded modes, and so the effective \fIlimit\fP for
710
+ multi\-threading will never be higher than the limit set with
711
+ \fB\-\-memlimit\-decompress\fP.
712
+ .IP ""
713
+ In contrast to the other memory usage limit options,
714
+ \fB\-\-memlimit\-mt\-decompress=\fP\fIlimit\fP has a system\-specific default
715
+ \fIlimit\fP. \fBxz \-\-info\-memory\fP can be used to see the current value.
716
+ .IP ""
717
+ This option and its default value exist because without any limit the
718
+ threaded decompressor could end up allocating an insane amount of memory
719
+ with some input files. If the default \fIlimit\fP is too low on your system,
720
+ feel free to increase the \fIlimit\fP but never set it to a value larger than
721
+ the amount of usable RAM as with appropriate input files \fBxz\fP will attempt
722
+ to use that amount of memory even with a low number of threads. Running out
723
+ of memory or swapping will not improve decompression performance.
724
+ .IP ""
725
+ See \fB\-\-memlimit\-compress=\fP\fIlimit\fP for possible ways to specify the
726
+ \fIlimit\fP. Setting \fIlimit\fP to \fB0\fP resets the \fIlimit\fP to the default
727
+ system\-specific value.
728
+ .TP
729
+ \fB\-M\fP \fIlimite\fP, \fB\-\-memlimit=\fP\fIlimite\fP, \fB\-\-memory=\fP\fIlimite\fP
730
+ This is equivalent to specifying \fB\-\-memlimit\-compress=\fP\fIlimit\fP
731
+ \fB\-\-memlimit\-decompress=\fP\fIlimit\fP \fB\-\-memlimit\-mt\-decompress=\fP\fIlimit\fP.
732
+ .TP
733
+ \fB\-\-no\-adjust\fP
734
+ Display an error and exit if the memory usage limit cannot be met without
735
+ adjusting settings that affect the compressed output. That is, this
736
+ prevents \fBxz\fP from switching the encoder from multi\-threaded mode to
737
+ single\-threaded mode and from reducing the LZMA2 dictionary size. Even when
738
+ this option is used the number of threads may be reduced to meet the memory
739
+ usage limit as that won't affect the compressed output.
740
+ .IP ""
741
+ Automatic adjusting is always disabled when creating raw streams
742
+ (\fB\-\-format=raw\fP).
743
+ .TP
744
+ \fB\-T\fP \fIthreads\fP, \fB\-\-threads=\fP\fIthreads\fP
745
+ Specify the number of worker threads to use. Setting \fIthreads\fP to a
746
+ special value \fB0\fP makes \fBxz\fP use up to as many threads as the processor(s)
747
+ on the system support. The actual number of threads can be fewer than
748
+ \fIthreads\fP if the input file is not big enough for threading with the given
749
+ settings or if using more threads would exceed the memory usage limit.
750
+ .IP ""
751
+ The single\-threaded and multi\-threaded compressors produce different
752
+ output. Single\-threaded compressor will give the smallest file size but
753
+ only the output from the multi\-threaded compressor can be decompressed using
754
+ multiple threads. Setting \fIthreads\fP to \fB1\fP will use the single\-threaded
755
+ mode. Setting \fIthreads\fP to any other value, including \fB0\fP, will use the
756
+ multi\-threaded compressor even if the system supports only one hardware
757
+ thread. (\fBxz\fP 5.2.x used single\-threaded mode in this situation.)
758
+ .IP ""
759
+ To use multi\-threaded mode with only one thread, set \fIthreads\fP to \fB+1\fP.
760
+ The \fB+\fP prefix has no effect with values other than \fB1\fP. A memory usage
761
+ limit can still make \fBxz\fP switch to single\-threaded mode unless
762
+ \fB\-\-no\-adjust\fP is used. Support for the \fB+\fP prefix was added in \fBxz\fP
763
+ 5.4.0.
764
+ .IP ""
765
+ If an automatic number of threads has been requested and no memory usage
766
+ limit has been specified, then a system\-specific default soft limit will be
767
+ used to possibly limit the number of threads. It is a soft limit in sense
768
+ that it is ignored if the number of threads becomes one, thus a soft limit
769
+ will never stop \fBxz\fP from compressing or decompressing. This default soft
770
+ limit will not make \fBxz\fP switch from multi\-threaded mode to single\-threaded
771
+ mode. The active limits can be seen with \fBxz \-\-info\-memory\fP.
772
+ .IP ""
773
+ Actuellement, la seule méthode de gestion avec des threads consiste à
774
+ séparer l'entrée en blocs et de les compresser indépendamment les uns des
775
+ autres. La taille par défaut des blocs dépend du niveau de compression et
776
+ peut\-être remplacée avec l'option \fB\-\-block\-size=\fP\fItaille\fP.
777
+ .IP ""
778
+ Threaded decompression only works on files that contain multiple blocks with
779
+ size information in block headers. All large enough files compressed in
780
+ multi\-threaded mode meet this condition, but files compressed in
781
+ single\-threaded mode don't even if \fB\-\-block\-size=\fP\fIsize\fP has been used.
782
+ .IP ""
783
+ The default value for \fIthreads\fP is \fB0\fP. In \fBxz\fP 5.4.x and older the
784
+ default is \fB1\fP.
785
+ .
786
+ .SS "Chaînes de filtres de compresseur personnalisées"
787
+ A custom filter chain allows specifying the compression settings in detail
788
+ instead of relying on the settings associated to the presets. When a custom
789
+ filter chain is specified, preset options (\fB\-0\fP \&...\& \fB\-9\fP and
790
+ \fB\-\-extreme\fP) earlier on the command line are forgotten. If a preset
791
+ option is specified after one or more custom filter chain options, the new
792
+ preset takes effect and the custom filter chain options specified earlier
793
+ are forgotten.
794
+ .PP
795
+ Une chaîne de filtre est comparable à une redirection (pipe) sur la ligne de
796
+ commande. Lors de la compression, les entrées non compressées vont au
797
+ premier filtre, dont la sortie va au prochain filtre (s'il y en a). La
798
+ sortie du dernier filtre est écrite sur le fichier compressé. Le nombre
799
+ maximal de filtres dans la chaîne est quatre, mais habituellement, un chaîne
800
+ de filtre n'a qu'un ou deux filtres.
801
+ .PP
802
+ Beaucoup de filtres ont des limitations sur l'endroit où ils peuvent se
803
+ placer dans la chaîne de filtre : quelques filtres ne peuvent fonctionner
804
+ qu'en tant que dernier filtre dans la chaîne, quelques uns en tant que non
805
+ dernier filtre, et d'autres à n'importe quelle position dans la
806
+ chaîne. Suivant le filtre, cette limitation est soit inhérente au profil du
807
+ filtre, soit existe pour des raisons de sécurité.
808
+ .PP
809
+ A custom filter chain can be specified in two different ways. The options
810
+ \fB\-\-filters=\fP\fIfilters\fP and \fB\-\-filters1=\fP\fIfilters\fP \&...\&
811
+ \fB\-\-filters9=\fP\fIfilters\fP allow specifying an entire filter chain in one
812
+ option using the liblzma filter string syntax. Alternatively, a filter
813
+ chain can be specified by using one or more individual filter options in the
814
+ order they are wanted in the filter chain. That is, the order of the
815
+ individual filter options is significant! When decoding raw streams
816
+ (\fB\-\-format=raw\fP), the filter chain must be specified in the same order as
817
+ it was specified when compressing. Any individual filter or preset options
818
+ specified before the full chain option (\fB\-\-filters=\fP\fIfilters\fP) will be
819
+ forgotten. Individual filters specified after the full chain option will
820
+ reset the filter chain.
821
+ .PP
822
+ Both the full and individual filter options take filter\-specific \fIoptions\fP
823
+ as a comma\-separated list. Extra commas in \fIoptions\fP are ignored. Every
824
+ option has a default value, so specify those you want to change.
825
+ .PP
826
+ Pour voir l'entièreté de la chaîne de filtres et ses \fIoptions\fP, utilisez
827
+ \fBxz \-vv\fP (ce qui est comme utiliser \fB\-\-verbose\fP deux fois). Cela
828
+ fonctionne aussi pour voir les options de chaîne de filtres utilisées par
829
+ les préréglages.
830
+ .TP
831
+ \fB\-\-filters=\fP\fIfilters\fP
832
+ Specify the full filter chain or a preset in a single option. Each filter
833
+ can be separated by spaces or two dashes (\fB\-\-\fP). \fIfilters\fP may need to be
834
+ quoted on the shell command line so it is parsed as a single option. To
835
+ denote \fIoptions\fP, use \fB:\fP or \fB=\fP. A preset can be prefixed with a \fB\-\fP
836
+ and followed with zero or more flags. The only supported flag is \fBe\fP to
837
+ apply the same options as \fB\-\-extreme\fP.
838
+ .TP
839
+ \fB\-\-filters1\fP=\fIfilters\fP ... \fB\-\-filters9\fP=\fIfilters\fP
840
+ Specify up to nine additional filter chains that can be used with
841
+ \fB\-\-block\-list\fP.
842
+ .IP ""
843
+ For example, when compressing an archive with executable files followed by
844
+ text files, the executable part could use a filter chain with a BCJ filter
845
+ and the text part only the LZMA2 filter.
846
+ .TP
847
+ \fB\-\-filters\-help\fP
848
+ Display a help message describing how to specify presets and custom filter
849
+ chains in the \fB\-\-filters\fP and \fB\-\-filters1=\fP\fIfilters\fP \&...\&
850
+ \fB\-\-filters9=\fP\fIfilters\fP options, and exit successfully.
851
+ .TP
852
+ \fB\-\-lzma1\fP[\fB=\fP\fIoptions\fP]
853
+ .PD 0
854
+ .TP
855
+ \fB\-\-lzma2\fP[\fB=\fP\fIoptions\fP]
856
+ .PD
857
+ Ajouter le filtre LZMA1 ou LZMA2 à la chaîne de filtres. Ces filtres ne
858
+ peuvent être utilisés que comme dernier filtre dans la chaîne.
859
+ .IP ""
860
+ LZMA1 est un filtre historique, qui n'est pris en charge presque uniquement
861
+ à cause de l'ancien format de fichier \fB.lzma\fP, qui ne prend en charge que
862
+ LZMA1. LZMA2 est une version mise à jour de LZMA1 pour régler certains
863
+ problèmes pratiques de LZMA1. Le format \fBxz\fP utilise LZMA2 et ne prend pas
864
+ du tout en charge LZMA1. Les taux et vitesses de compression de LZMA1 et
865
+ LZMA2 sont pratiquement identiques.
866
+ .IP ""
867
+ LZMA1 et LZMA2 partagent le même ensemble d'\fIoptions\fP :
868
+ .RS
869
+ .TP
870
+ \fBpreset=\fP\fIpréréglage\fP
871
+ Reset all LZMA1 or LZMA2 \fIoptions\fP to \fIpreset\fP. \fIPreset\fP consist of an
872
+ integer, which may be followed by single\-letter preset modifiers. The
873
+ integer can be from \fB0\fP to \fB9\fP, matching the command line options \fB\-0\fP
874
+ \&...\& \fB\-9\fP. The only supported modifier is currently \fBe\fP, which matches
875
+ \fB\-\-extreme\fP. If no \fBpreset\fP is specified, the default values of LZMA1 or
876
+ LZMA2 \fIoptions\fP are taken from the preset \fB6\fP.
877
+ .TP
878
+ \fBdict=\fP\fItaille\fP
879
+ La \fItaille\fP du dictionnaire (historique du tampon) indique combien d'octets
880
+ des données récement décompressées sont gardés en mémoire. L'algorithme
881
+ essaie de trouver les séquences d'octets répétées (identiques) dans les
882
+ données décompressées et les remplace par les données actuellement dans le
883
+ dictionnaire. Plus gros est le dictionnaire, plus grande est la chance de
884
+ trouver une correspondance. Ainsi, l'augmentation de la \fItaille\fP du
885
+ dictionnaire augmente habituellement le taux de compression, mais un
886
+ dictionnaire plus gros que le fichier non compressé est un gachis de
887
+ mémoire.
888
+ .IP ""
889
+ Généralement la \fItaille\fP du dictionnaire est entre 64\ Kio et 64\ Mio. Le
890
+ minimum étant 4\ Kio. La \fItaille\fP maximale pour la compression est
891
+ habituellement 1,5\ Gio (1536\ Mio). Le décompresseur prend en charge les
892
+ dictionnaires jusqu'à un octet de moins que 4\ Gio, ce qui est le maximum
893
+ pour les formats de flux LZMA1 et LZMA2.
894
+ .IP ""
895
+ La \fItaille\fP du dictionnaire et le chercheur de correspondance (match
896
+ finder) (\fImf\fP) déterminent ensemble l'utilisation de la mémoire de
897
+ l'encodeur LZMA1 ou LZMA2. La même (ou une plus grosse) \fItaille\fP de
898
+ dictionnaire est requise pour décompresser que ce qui a été utilisé pour la
899
+ compression, ainsi l'utilisation de la mémoire du décodeur est déterminée
900
+ par la taille du dictionnaire utilisée lors de la compression. Les en\-têtes
901
+ de \fB.xz\fP stockent la \fItaille\fP de dictionnaire sous la forme 2^\fIn\fP ou
902
+ 2^\fIn\fP + 2^(\fIn\fP\-1), de sorte que ces \fItailles\fP sont quelque peu préférées
903
+ pour la compression. Les autres \fItailles\fP seront arrondies à la hausse
904
+ lorsque stockées dans les en\-têtes de \fB.xz\fP.
905
+ .TP
906
+ \fBlc=\fP\fIlc\fP
907
+ Spécifiez le nombre d'octets de contexte littéraux. Le minimum est \fB0\fP et
908
+ le maximum est \fB4\fP. La valeur par défaut est \fB3\fP. En plus, la somme de
909
+ \fIlc\fP et \fIlp\fP ne doit pas excéder \fB4\fP.
910
+ .IP ""
911
+ Tous les octets qui ne peuvent pas être codés comme des correspondances sont
912
+ codés comme des littéraux. C'est à dire que les littéraux sont simplement
913
+ des octets 8 bits encodés un à la fois.
914
+ .IP ""
915
+ The literal coding makes an assumption that the highest \fIlc\fP bits of the
916
+ previous uncompressed byte correlate with the next byte. For example, in
917
+ typical English text, an upper\-case letter is often followed by a lower\-case
918
+ letter, and a lower\-case letter is usually followed by another lower\-case
919
+ letter. In the US\-ASCII character set, the highest three bits are 010 for
920
+ upper\-case letters and 011 for lower\-case letters. When \fIlc\fP is at least
921
+ 3, the literal coding can take advantage of this property in the
922
+ uncompressed data.
923
+ .IP ""
924
+ The default value (3) is usually good. If you want maximum compression,
925
+ test \fBlc=4\fP. Sometimes it helps a little, and sometimes it makes
926
+ compression worse. If it makes it worse, test \fBlc=2\fP too.
927
+ .TP
928
+ \fBlp=\fP\fIlp\fP
929
+ Indiquer le nombre de bits de position littérale. Le minimum est \fB0\fP et le
930
+ maximum \fB4\fP; par défaut c'est \fB0\fP.
931
+ .IP ""
932
+ \fILp\fP affecte le type d'alignement dans les données décompressées qui est
933
+ présumé lors de l'encodage des littéraux. Voir \fIpb\fP ci dessous pour plus
934
+ d'information sur l'alignement.
935
+ .TP
936
+ \fBpb=\fP\fIpb\fP
937
+ Indiquer le nombre de bits de position. Le minimum est \fB0\fP et le maximum
938
+ \fB4\fP; par défaut \fB2\fP.
939
+ .IP ""
940
+ \fIPb\fP affecte quel genre d'alignement est présumé en général dans les
941
+ données non compressées. Par défaut c'est un alignement de quatre octets
942
+ (2^\fIpb\fP=2^2=4), ce qui est généralement un bon choix lorsqu'il n'y a pas de
943
+ meilleure estimation.
944
+ .IP ""
945
+ When the alignment is known, setting \fIpb\fP accordingly may reduce the file
946
+ size a little. For example, with text files having one\-byte alignment
947
+ (US\-ASCII, ISO\-8859\-*, UTF\-8), setting \fBpb=0\fP can improve compression
948
+ slightly. For UTF\-16 text, \fBpb=1\fP is a good choice. If the alignment is
949
+ an odd number like 3 bytes, \fBpb=0\fP might be the best choice.
950
+ .IP ""
951
+ Même si l'alignement présumé peut être ajusté avec \fIpb\fP et \fIlp\fP, LZMA1 et
952
+ LZMA2 favorisent toujours légèrement l'alignement sur 16 octets. Il peut
953
+ être utile d'en tenir compte lors de la conception de formats de fichiers
954
+ susceptibles d'être souvent compressés avec LZMA1 ou LZMA2.
955
+ .TP
956
+ \fBmf=\fP\fImf\fP
957
+ Match finder has a major effect on encoder speed, memory usage, and
958
+ compression ratio. Usually Hash Chain match finders are faster than Binary
959
+ Tree match finders. The default depends on the \fIpreset\fP: 0 uses \fBhc3\fP,
960
+ 1\(en3 use \fBhc4\fP, and the rest use \fBbt4\fP.
961
+ .IP ""
962
+ Les chercheurs de correspondance suivants sont pris en charge. Les formules
963
+ d'utilisation de la mémoire ci\-dessous sont des approximations grossières
964
+ qui sont les plus proches de la réalité lorsque \fIdict\fP est une puissance de
965
+ deux.
966
+ .RS
967
+ .TP
968
+ \fBhc3\fP
969
+ Chaîne de hachage avec hachage de 2 et 3 octets
970
+ .br
971
+ Valeur minimale pour \fInice\fP : \fB3\fP
972
+ .br
973
+ Utilisation de la mémoire :
974
+ .br
975
+ \fIdict\fP * 7.5 (if \fIdict\fP <= 16 Mio);
976
+ .br
977
+ \fIdict\fP * 5.5 + 64 MiB (si \fIdict\fP > 16 Mio)
978
+ .TP
979
+ \fBhc4\fP
980
+ Chaîne de hachage avec hachage de 2, 3 et 4 octets
981
+ .br
982
+ Valeur minimale pour \fInice\fP : \fB4\fP
983
+ .br
984
+ Utilisation de la mémoire :
985
+ .br
986
+ \fIdict\fP * 7.5 (si \fIdict\fP <= 32 Mio);
987
+ .br
988
+ \fIdict\fP * 6.5 (si \fIdict\fP > 32 Mio)
989
+ .TP
990
+ \fBbt2\fP
991
+ Arbre binaire avec hachage de 2 octets
992
+ .br
993
+ Valeur minimale pour \fInice\fP : \fB2\fP
994
+ .br
995
+ Utilisation de la mémoire : \fIdict\fP * 9.5
996
+ .TP
997
+ \fBbt3\fP
998
+ Arbre binaire avec hachage de 2 et 3 octets
999
+ .br
1000
+ Valeur minimale pour \fInice\fP : \fB3\fP
1001
+ .br
1002
+ Utilisation de la mémoire :
1003
+ .br
1004
+ \fIdict\fP * 11.5 (si \fIdict\fP <= 16 Mio);
1005
+ .br
1006
+ \fIdict\fP * 9.5 + 64 MiB (si \fIdict\fP > 16 Mio)
1007
+ .TP
1008
+ \fBbt4\fP
1009
+ Arbre binaire avec hachage 2, 3 et 4 octets
1010
+ .br
1011
+ Valeur minimale pour \fInice\fP : \fB4\fP
1012
+ .br
1013
+ Utilisation de la mémoire :
1014
+ .br
1015
+ \fIdict\fP * 11.5 (si \fIdict\fP <= 32 Mio);
1016
+ .br
1017
+ \fIdict\fP * 10.5 (si \fIdict\fP > 32 Mio)
1018
+ .RE
1019
+ .TP
1020
+ \fBmode=\fP\fImode\fP
1021
+ Compression \fImode\fP specifies the method to analyze the data produced by the
1022
+ match finder. Supported \fImodes\fP are \fBfast\fP and \fBnormal\fP. The default is
1023
+ \fBfast\fP for \fIpresets\fP 0\(en3 and \fBnormal\fP for \fIpresets\fP 4\(en9.
1024
+ .IP ""
1025
+ Habituellement, \fBfast\fP est utilisé avec les chercheurs de correspondance de
1026
+ chaîne de hachage et \fBnormal\fP avec les chercheurs de correspondance d'arbre
1027
+ binaire. C'est aussi ce que font les \fIpréréglages\fP.
1028
+ .TP
1029
+ \fBnice=\fP\fInice\fP
1030
+ Spécifier ce qui est considéré comme une bonne longueur pour une
1031
+ correspondance. Une fois que la correspondance d'au moins \fInice\fP octets est
1032
+ trouvée, l'algorithme arrête de chercher de meilleures correspondances
1033
+ possibles.
1034
+ .IP ""
1035
+ \fINice\fP can be 2\(en273 bytes. Higher values tend to give better
1036
+ compression ratio at the expense of speed. The default depends on the
1037
+ \fIpreset\fP.
1038
+ .TP
1039
+ \fBdepth=\fP\fIprofondeur\fP
1040
+ Spécifier la profondeur de recherche maximale dans l'outil de recherche de
1041
+ correspondances. La valeur par défaut est \fB0\fP, ce qui fait que le
1042
+ compresseur détermine une \fIprofondeur\fP raisonnable en fonction de \fImf\fP et
1043
+ \fInice\fP.
1044
+ .IP ""
1045
+ Reasonable \fIdepth\fP for Hash Chains is 4\(en100 and 16\(en1000 for Binary
1046
+ Trees. Using very high values for \fIdepth\fP can make the encoder extremely
1047
+ slow with some files. Avoid setting the \fIdepth\fP over 1000 unless you are
1048
+ prepared to interrupt the compression in case it is taking far too long.
1049
+ .RE
1050
+ .IP ""
1051
+ Lors du décodage des flux bruts (\fB\-\-format=raw\fP), LZMA2 nécessite seulement
1052
+ la \fItaille\fP du dictionnaire. LZMA1 nécessite aussi \fIlc\fP, \fIlp\fP et \fIpb\fP.
1053
+ .TP
1054
+ \fB\-\-x86\fP[\fB=\fP\fIoptions\fP]
1055
+ .PD 0
1056
+ .TP
1057
+ \fB\-\-arm\fP[\fB=\fP\fIoptions\fP]
1058
+ .TP
1059
+ \fB\-\-armthumb\fP[\fB=\fP\fIoptions\fP]
1060
+ .TP
1061
+ \fB\-\-arm64\fP[\fB=\fP\fIoptions\fP]
1062
+ .TP
1063
+ \fB\-\-powerpc\fP[\fB=\fP\fIoptions\fP]
1064
+ .TP
1065
+ \fB\-\-ia64\fP[\fB=\fP\fIoptions\fP]
1066
+ .TP
1067
+ \fB\-\-sparc\fP[\fB=\fP\fIoptions\fP]
1068
+ .TP
1069
+ \fB\-\-riscv\fP[\fB=\fP\fIoptions\fP]
1070
+ .PD
1071
+ Ajouter un filtre branch/call/jump (BCJ) à la chaîne de filtres. Ces filtres
1072
+ ne peuvent être utilisés que s'ils ne sont pas le dernier filtre de la
1073
+ chaîne de filtrage.
1074
+ .IP ""
1075
+ A BCJ filter converts relative addresses in the machine code to their
1076
+ absolute counterparts. This doesn't change the size of the data but it
1077
+ increases redundancy, which can help LZMA2 to produce 0\(en15\ % smaller
1078
+ \&\fB.xz\fP file. The BCJ filters are always reversible, so using a BCJ filter
1079
+ for wrong type of data doesn't cause any data loss, although it may make the
1080
+ compression ratio slightly worse. The BCJ filters are very fast and use an
1081
+ insignificant amount of memory.
1082
+ .IP ""
1083
+ Ces filtres BCJ présentent des problèmes connus liés au taux de
1084
+ compression :
1085
+ .RS
1086
+ .IP \(bu 3
1087
+ Some types of files containing executable code (for example, object files,
1088
+ static libraries, and Linux kernel modules) have the addresses in the
1089
+ instructions filled with filler values. These BCJ filters will still do the
1090
+ address conversion, which will make the compression worse with these files.
1091
+ .IP \(bu 3
1092
+ If a BCJ filter is applied on an archive, it is possible that it makes the
1093
+ compression ratio worse than not using a BCJ filter. For example, if there
1094
+ are similar or even identical executables then filtering will likely make
1095
+ the files less similar and thus compression is worse. The contents of
1096
+ non\-executable files in the same archive can matter too. In practice one
1097
+ has to try with and without a BCJ filter to see which is better in each
1098
+ situation.
1099
+ .RE
1100
+ .IP ""
1101
+ Different instruction sets have different alignment: the executable file
1102
+ must be aligned to a multiple of this value in the input data to make the
1103
+ filter work.
1104
+ .RS
1105
+ .RS
1106
+ .PP
1107
+ .TS
1108
+ tab(;);
1109
+ l n l
1110
+ l n l.
1111
+ Filtre;Alignement;Notes
1112
+ x86;1;32 bits ou 64 bits x86
1113
+ ARM;4;
1114
+ ARM\-Thumb;2;
1115
+ ARM64;4;4096\-byte alignment is best
1116
+ PowerPC;4;Grand boutiste seulement
1117
+ IA\-64;16;Itanium
1118
+ SPARC;4;
1119
+ RISC\-V;2;
1120
+ .TE
1121
+ .RE
1122
+ .RE
1123
+ .IP ""
1124
+ Since the BCJ\-filtered data is usually compressed with LZMA2, the
1125
+ compression ratio may be improved slightly if the LZMA2 options are set to
1126
+ match the alignment of the selected BCJ filter. Examples:
1127
+ .RS
1128
+ .IP \(bu 3
1129
+ IA\-64 filter has 16\-byte alignment so \fBpb=4,lp=4,lc=0\fP is good with LZMA2
1130
+ (2^4=16).
1131
+ .IP \(bu 3
1132
+ RISC\-V code has 2\-byte or 4\-byte alignment depending on whether the file
1133
+ contains 16\-bit compressed instructions (the C extension). When 16\-bit
1134
+ instructions are used, \fBpb=2,lp=1,lc=3\fP or \fBpb=1,lp=1,lc=3\fP is good. When
1135
+ 16\-bit instructions aren't present, \fBpb=2,lp=2,lc=2\fP is the best.
1136
+ \fBreadelf \-h\fP can be used to check if "RVC" appears on the "Flags" line.
1137
+ .IP \(bu 3
1138
+ ARM64 is always 4\-byte aligned so \fBpb=2,lp=2,lc=2\fP is the best.
1139
+ .IP \(bu 3
1140
+ The x86 filter is an exception. It's usually good to stick to LZMA2's
1141
+ defaults (\fBpb=2,lp=0,lc=3\fP) when compressing x86 executables.
1142
+ .RE
1143
+ .IP ""
1144
+ Tous les filtres BCJ prennent en charge les mêmes \fIoptions\fP :
1145
+ .RS
1146
+ .TP
1147
+ \fBstart=\fP\fIdécalage\fP
1148
+ Spécifier le \fIdécalage\fP de départ qui est utilisé lors de la conversion
1149
+ entre les adresses relatives et absolues. Le \fIdécalage\fP doit être un
1150
+ multiple de l'alignement du filtre (voir la table ci\-dessus). Sa valeur par
1151
+ défaut est zéro. En pratique, cette dernière convient ; indiquer un
1152
+ \fIdécalage\fP personnalisé est la plupart du temps inutile.
1153
+ .RE
1154
+ .TP
1155
+ \fB\-\-delta\fP[\fB=\fP\fIoptions\fP]
1156
+ Ajouter le filtre Delta à la chaîne de filtres. Le filtre Delta ne peut être
1157
+ utilisé que s'il n'est pas le dernier filtre dans la chaîne.
1158
+ .IP ""
1159
+ Currently only simple byte\-wise delta calculation is supported. It can be
1160
+ useful when compressing, for example, uncompressed bitmap images or
1161
+ uncompressed PCM audio. However, special purpose algorithms may give
1162
+ significantly better results than Delta + LZMA2. This is true especially
1163
+ with audio, which compresses faster and better, for example, with
1164
+ \fBflac\fP(1).
1165
+ .IP ""
1166
+ \fIoptions\fP prises en charge :
1167
+ .RS
1168
+ .TP
1169
+ \fBdist=\fP\fIdistance\fP
1170
+ Specify the \fIdistance\fP of the delta calculation in bytes. \fIdistance\fP must
1171
+ be 1\(en256. The default is 1.
1172
+ .IP ""
1173
+ Par exemple, avec \fBdist=2\fP et une entrée huit octets A1 B1 A2 B3 A3 B5 A4
1174
+ B7, la sortie sera A1 B1 01 02 01 02 01 02.
1175
+ .RE
1176
+ .
1177
+ .SS "Autres options"
1178
+ .TP
1179
+ \fB\-q\fP, \fB\-\-quiet\fP
1180
+ Supprimer les avertissements et les notifications. Indiquer cela deux fois
1181
+ supprimera aussi les erreurs. Cette option n'a aucun effet sur le statut de
1182
+ sortie. Cela dit, même si un avertissement était supprimé, le statut de
1183
+ sortie indiquant un avertissement sera encore utilisé.
1184
+ .TP
1185
+ \fB\-v\fP, \fB\-\-verbose\fP
1186
+ Être bavard. Si l'erreur standard est connectée à un terminal, \fBxz\fP
1187
+ affichera une barre de progression. Indiquer \fB\-\-verbose\fP deux fois donnera
1188
+ une sortie encore plus bavarde.
1189
+ .IP ""
1190
+ La barre de progression montre l'information suivante :
1191
+ .RS
1192
+ .IP \(bu 3
1193
+ Le pourcentage de complétion est montré si la taille du fichier en entrée
1194
+ est connue. Néanmoins, le pourcentage ne peut pas être montré en cas de
1195
+ redirection.
1196
+ .IP \(bu 3
1197
+ Quantité de données compressées produites (compression) ou consommées
1198
+ (décompression).
1199
+ .IP \(bu 3
1200
+ Quantité de données non compressées consommées (compression) ou produites
1201
+ (décompression).
1202
+ .IP \(bu 3
1203
+ Le taux de compression, calculé en divisant la quantité de données
1204
+ compréssées déjà traitées par la quantité de données décompressées déjà
1205
+ traitées.
1206
+ .IP \(bu 3
1207
+ Vitesse de compression ou de décompression. Elle correspond à la quantité de
1208
+ données non compressées consommées (compression) ou produites
1209
+ (décompression) par seconde. Elle apparait quelques secondes après le début
1210
+ du traitement du fichier par \fBxz\fP.
1211
+ .IP \(bu 3
1212
+ Temps écoulé dans le format M:SS ou H:MM:SS.
1213
+ .IP \(bu 3
1214
+ Estimated remaining time is shown only when the size of the input file is
1215
+ known and a couple of seconds have already passed since \fBxz\fP started
1216
+ processing the file. The time is shown in a less precise format which never
1217
+ has any colons, for example, 2 min 30 s.
1218
+ .RE
1219
+ .IP ""
1220
+ When standard error is not a terminal, \fB\-\-verbose\fP will make \fBxz\fP print
1221
+ the filename, compressed size, uncompressed size, compression ratio, and
1222
+ possibly also the speed and elapsed time on a single line to standard error
1223
+ after compressing or decompressing the file. The speed and elapsed time are
1224
+ included only when the operation took at least a few seconds. If the
1225
+ operation didn't finish, for example, due to user interruption, also the
1226
+ completion percentage is printed if the size of the input file is known.
1227
+ .TP
1228
+ \fB\-Q\fP, \fB\-\-no\-warn\fP
1229
+ Ne pas mettre l'état de sortie à \fB2\fP même si une condition méritant un
1230
+ avertissement a été détectée. Cette option n'affecte pas le niveau de
1231
+ verbosité, néanmoins, les deux options \fB\-\-quiet\fP et \fB\-\-no\-warn\fP doivent
1232
+ être utilisées pour ne pas afficher d'avertissements, ni altérer le statut
1233
+ de sortie.
1234
+ .TP
1235
+ \fB\-\-robot\fP
1236
+ Afficher les messages dans un format analysable par une machine. Ceci est
1237
+ destiné à faciliter l'écriture des frontaux qui voudraient utiliser \fBxz\fP
1238
+ plutôt que liblzma, ce qui pourrait être le cas pour différents scripts. La
1239
+ sortie avec cette option activée est destinée à rester stable sur les
1240
+ différentes versions de \fBxz\fP. Consulter le paragraphe \fBROBOT MODE\fP pour
1241
+ les détails.
1242
+ .TP
1243
+ \fB\-\-info\-memory\fP
1244
+ Display, in human\-readable format, how much physical memory (RAM) and how
1245
+ many processor threads \fBxz\fP thinks the system has and the memory usage
1246
+ limits for compression and decompression, and exit successfully.
1247
+ .TP
1248
+ \fB\-h\fP, \fB\-\-help\fP
1249
+ Afficher un message d'aide décrivant les options les plus couramment
1250
+ utilisées et quitter.
1251
+ .TP
1252
+ \fB\-H\fP, \fB\-\-long\-help\fP
1253
+ Afficher un message d'aide décrivant toutes les options de \fBxz\fP et quitter.
1254
+ .TP
1255
+ \fB\-V\fP, \fB\-\-version\fP
1256
+ Afficher le numéro de version de \fBxz\fP et de liblzma dans un format lisible
1257
+ par un humain. Pour obtenir une sortie analysable par la machine, spécifiez
1258
+ \fB\-\-robot\fP avant \fB\-\-version\fP.
1259
+ .
1260
+ .SH "MODE ROBOT"
1261
+ The robot mode is activated with the \fB\-\-robot\fP option. It makes the output
1262
+ of \fBxz\fP easier to parse by other programs. Currently \fB\-\-robot\fP is
1263
+ supported only together with \fB\-\-list\fP, \fB\-\-filters\-help\fP, \fB\-\-info\-memory\fP,
1264
+ and \fB\-\-version\fP. It will be supported for compression and decompression in
1265
+ the future.
1266
+ .
1267
+ .SS "Mode liste"
1268
+ \fBxz \-\-robot \-\-list\fP utilise une sortie séparée par des tabulations. La
1269
+ première colonne de toutes les lignes possède une chaîne qui indique le type
1270
+ d'information trouvée sur cette ligne :
1271
+ .TP
1272
+ \fBname\fP
1273
+ C'est toujours la première ligne au début de la liste d'un fichier. La
1274
+ seconde colonne de la ligne est le nom de fichier.
1275
+ .TP
1276
+ \fBfile\fP
1277
+ Cette ligne contient l'information globale sur le fichier \fB.xz\fP. Cette
1278
+ ligne est toujours écrite après la ligne \fBname\fP.
1279
+ .TP
1280
+ \fBstream\fP
1281
+ Ce type de ligne n'est utilisée que lorsque \fB \-\-verbose\fP a été indiquée. Il
1282
+ y a autant de lignes \fBstream\fP qu'il y a de flux dans le fichier \fB.xz\fP.
1283
+ .TP
1284
+ \fBblock\fP
1285
+ Ce type de ligne n'est utilisé seulement lorsque \fB\-\-verbose\fP a été
1286
+ indiquée. Il y a autant de lignes \fBblock\fP qu'il y a de blocs dans le
1287
+ fichier \fB.xz\fP. Les lignes \fBblock\fP sont affichées après toutes les lignes
1288
+ \fBstream\fP ; les différents types de lignes ne sont pas imbriqués.
1289
+ .TP
1290
+ \fBsummary\fP
1291
+ Ce type de ligne n'est utilisé que lorsque \fB\-\-verbose\fP a été indiqué deux
1292
+ fois. Cette ligne est affichée après toutes les lignes \fBblock\fP. Comme la
1293
+ ligne \fBfile\fP, la ligne \fBsummary\fP contient l'information globale sur le
1294
+ fichier \fB.xz\fP.
1295
+ .TP
1296
+ \fBtotals\fP
1297
+ Cette ligne est toujours la toute dernière ligne de la sortie. Elle affiche
1298
+ les comptes et les tailles totaux.
1299
+ .PP
1300
+ Les colonnes des lignes \fBfile\fP :
1301
+ .PD 0
1302
+ .RS
1303
+ .IP 2. 4
1304
+ Nombre de flux dans le fichier
1305
+ .IP 3. 4
1306
+ Nombre total de blocs dans le ou les flux.
1307
+ .IP 4. 4
1308
+ Taille compressée du fichier
1309
+ .IP 5. 4
1310
+ Taille décompressée du fichier
1311
+ .IP 6. 4
1312
+ Compression ratio, for example, \fB0.123\fP. If ratio is over 9.999, three
1313
+ dashes (\fB\-\-\-\fP) are displayed instead of the ratio.
1314
+ .IP 7. 4
1315
+ Liste de noms de contrôles d'intégrité séparés par des virgules. Les chaînes
1316
+ suivantes sont utilisées pour les types de vérification connus : \fBNone\fP,
1317
+ \fBCRC32\fP, \fBCRC64\fP et \fBSHA256\fP. Pour le types de vérification inconnus,
1318
+ \fBUnknown\-\fP\fIN\fP est utilisé, où \fIN\fP est un identifiant de vérification sous
1319
+ la forme d'un nombre décimal (un ou deux chiffres).
1320
+ .IP 8. 4
1321
+ Taille totale du remplissage du flux dans le fichier
1322
+ .RE
1323
+ .PD
1324
+ .PP
1325
+ Les colonnes des lignes \fBstream\fP :
1326
+ .PD 0
1327
+ .RS
1328
+ .IP 2. 4
1329
+ Numéro de flux (le premier flux a le numéro 1)
1330
+ .IP 3. 4
1331
+ Nombre de blocs dans le flux
1332
+ .IP 4. 4
1333
+ Décalage de départ compressé
1334
+ .IP 5. 4
1335
+ Décalage de départ décompressé
1336
+ .IP 6. 4
1337
+ Taille compressée (ne comprend pas le remplissage du flux)
1338
+ .IP 7. 4
1339
+ Taille décompressée
1340
+ .IP 8. 4
1341
+ Taux de compression
1342
+ .IP 9. 4
1343
+ Nom de la vérification d'intégrité
1344
+ .IP 10. 4
1345
+ Taille du remplissage de flux
1346
+ .RE
1347
+ .PD
1348
+ .PP
1349
+ Les colonnes des lignes \fBblock\fP :
1350
+ .PD 0
1351
+ .RS
1352
+ .IP 2. 4
1353
+ Numéro du flux qui contient ce bloc
1354
+ .IP 3. 4
1355
+ Numéro du bloc relatif au commencement du flux (le premier bloc a pour
1356
+ numéro 1)
1357
+ .IP 4. 4
1358
+ Numéro du bloc relatif au début du fichier
1359
+ .IP 5. 4
1360
+ Décalage de départ compressé relatif au début du fichier
1361
+ .IP 6. 4
1362
+ Décalage de départ décompressé relatif au début du fichier
1363
+ .IP 7. 4
1364
+ Taille compressée totale du bloc (en\-têtes inclus)
1365
+ .IP 8. 4
1366
+ Taille décompressée
1367
+ .IP 9. 4
1368
+ Taux de compression
1369
+ .IP 10. 4
1370
+ Nom de la vérification d'intégrité
1371
+ .RE
1372
+ .PD
1373
+ .PP
1374
+ Si \fB\-\-verbose\fP a été indiqué deux fois, les colonnes additionnelles sont
1375
+ inclues sur les lignes \fBblock\fP. Elles ne sont pas affichées avec un seul
1376
+ \fB\-\-verbose\fP, car l'obtention de ces informations nécessite de nombreuses
1377
+ recherches et peut donc être lente :
1378
+ .PD 0
1379
+ .RS
1380
+ .IP 11. 4
1381
+ Valeur de la vérification d'intégrité en hexadécimal
1382
+ .IP 12. 4
1383
+ Taille d'en\-tête de bloc
1384
+ .IP 13. 4
1385
+ Drapeaux du bloc : \fBc\fP indique que la taille compressée est présente, et
1386
+ \fBu\fP indique que la taille décompréssée est présente. Si le drapeau n'est
1387
+ pas indiqué, un tiret (\fB\-\fP) est affiché à la place pour que la longueur de
1388
+ la chaîne reste fixe. De nouveaux drapeaux pourraient être ajoutés à la fin
1389
+ de la chaîne dans le futur.
1390
+ .IP 14. 4
1391
+ Taille des données effectivement compressées dans le bloc (en excluant
1392
+ l'en\-tête de bloc, le remplissage de bloc et les champs de vérification).
1393
+ .IP 15. 4
1394
+ Quantité de mémoire (en octets) nécessaire pour décompresser ce bloc avec
1395
+ cette version de \fBxz\fP.
1396
+ .IP 16. 4
1397
+ Chaîne de filtrage. Remarquez que la plupart des options utilisées au moment
1398
+ de la compression ne peuvent pas être connues, car seules les options
1399
+ nécessaires pour la décompression sont stockées dans les en\-têtes \fB.xz\fP.
1400
+ .RE
1401
+ .PD
1402
+ .PP
1403
+ Les colonnes des lignes \fBsummary\fP :
1404
+ .PD 0
1405
+ .RS
1406
+ .IP 2. 4
1407
+ Quantité de mémoire (en octets) nécessaire pour décompresser ce fichier avec
1408
+ cette version de \fBxz\fP.
1409
+ .IP 3. 4
1410
+ \fByes\fP ou \fBno\fP indique si tous les en\-têtes de bloc stockent à la fois la
1411
+ taille compressée et la taille décompressée.
1412
+ .PP
1413
+ \fIDepuis\fP \fBxz\fP \fI5.1.2alpha:\fP
1414
+ .IP 4. 4
1415
+ Version minimale de \fBxz\fP nécessaire pour décompresser le fichier.
1416
+ .RE
1417
+ .PD
1418
+ .PP
1419
+ Les colonnes de la ligne \fBtotals\fP :
1420
+ .PD 0
1421
+ .RS
1422
+ .IP 2. 4
1423
+ Nombre de flux
1424
+ .IP 3. 4
1425
+ Nombre de blocs
1426
+ .IP 4. 4
1427
+ Taille compressée
1428
+ .IP 5. 4
1429
+ Taille décompressée
1430
+ .IP 6. 4
1431
+ Taux de compression moyen
1432
+ .IP 7. 4
1433
+ Liste séparée par des virgules des noms de vérification d'intégrité qui
1434
+ étaient présents dans les fichiers
1435
+ .IP 8. 4
1436
+ Taille de remplissage de flux
1437
+ .IP 9. 4
1438
+ Nombre de fichiers. Permet de garder l'ordre des colonnes précédentes comme
1439
+ sur les lignes \fBfile\fP.
1440
+ .PD
1441
+ .RE
1442
+ .PP
1443
+ Si \fB\-\-verbose\fP a été indiqué deux fois, des colonnes supplémentaires sont
1444
+ incluses sur la ligne \fBtotals\fP :
1445
+ .PD 0
1446
+ .RS
1447
+ .IP 10. 4
1448
+ Quantité maximale de mémoire (en octets) nécessaire pour décompresser les
1449
+ fichiers avec cette version de \fBxz\fP.
1450
+ .IP 11. 4
1451
+ \fByes\fP ou \fBno\fP indique si tous les en\-têtes de bloc stockent à la fois la
1452
+ taille compressée et la taille décompressée.
1453
+ .PP
1454
+ \fIDepuis\fP \fBxz\fP \fI5.1.2alpha:\fP
1455
+ .IP 12. 4
1456
+ Version minimale de \fBxz\fP nécessaire pour décompresser le fichier.
1457
+ .RE
1458
+ .PD
1459
+ .PP
1460
+ Les versions futures pourront ajouter de nouveaux types de lignes et de
1461
+ nouvelles colonnes pourront être ajoutées aux types de lignes existants,
1462
+ mais les colonnes existantes ne seront pas modifiées.
1463
+ .
1464
+ .SS "Filters help"
1465
+ \fBxz \-\-robot \-\-filters\-help\fP prints the supported filters in the following
1466
+ format:
1467
+ .PP
1468
+ \fIfilter\fP\fB:\fP\fIoption\fP\fB=<\fP\fIvalue\fP\fB>,\fP\fIoption\fP\fB=<\fP\fIvalue\fP\fB>\fP...
1469
+ .TP
1470
+ \fIfilter\fP
1471
+ Name of the filter
1472
+ .TP
1473
+ \fIoption\fP
1474
+ Name of a filter specific option
1475
+ .TP
1476
+ \fIvalue\fP
1477
+ Numeric \fIvalue\fP ranges appear as \fB<\fP\fImin\fP\fB\-\fP\fImax\fP\fB>\fP. String
1478
+ \fIvalue\fP choices are shown within \fB< >\fP and separated by a \fB|\fP
1479
+ character.
1480
+ .PP
1481
+ Each filter is printed on its own line.
1482
+ .
1483
+ .SS "Information de limite de mémoire"
1484
+ \fBxz \-\-robot \-\-info\-memory\fP prints a single line with multiple tab\-separated
1485
+ columns:
1486
+ .IP 1. 4
1487
+ Total amount of physical memory (RAM) in bytes.
1488
+ .IP 2. 4
1489
+ Memory usage limit for compression in bytes (\fB\-\-memlimit\-compress\fP). A
1490
+ special value of \fB0\fP indicates the default setting which for
1491
+ single\-threaded mode is the same as no limit.
1492
+ .IP 3. 4
1493
+ Memory usage limit for decompression in bytes (\fB\-\-memlimit\-decompress\fP). A
1494
+ special value of \fB0\fP indicates the default setting which for
1495
+ single\-threaded mode is the same as no limit.
1496
+ .IP 4. 4
1497
+ Since \fBxz\fP 5.3.4alpha: Memory usage for multi\-threaded decompression in
1498
+ bytes (\fB\-\-memlimit\-mt\-decompress\fP). This is never zero because a
1499
+ system\-specific default value shown in the column 5 is used if no limit has
1500
+ been specified explicitly. This is also never greater than the value in the
1501
+ column 3 even if a larger value has been specified with
1502
+ \fB\-\-memlimit\-mt\-decompress\fP.
1503
+ .IP 5. 4
1504
+ Since \fBxz\fP 5.3.4alpha: A system\-specific default memory usage limit that is
1505
+ used to limit the number of threads when compressing with an automatic
1506
+ number of threads (\fB\-\-threads=0\fP) and no memory usage limit has been
1507
+ specified (\fB\-\-memlimit\-compress\fP). This is also used as the default value
1508
+ for \fB\-\-memlimit\-mt\-decompress\fP.
1509
+ .IP 6. 4
1510
+ Since \fBxz\fP 5.3.4alpha: Number of available processor threads.
1511
+ .PP
1512
+ Dans le futur, la sortie de \fBxz \-\-robot \-\-info\-memory\fP pourrait avoir plus
1513
+ de colonnes, mais jamais plus qu'une ligne unique.
1514
+ .
1515
+ .SS Version
1516
+ \fBxz \-\-robot \-\-version\fP prints the version number of \fBxz\fP and liblzma in
1517
+ the following format:
1518
+ .PP
1519
+ \fBXZ_VERSION=\fP\fIXYYYZZZS\fP
1520
+ .br
1521
+ \fBLIBLZMA_VERSION=\fP\fIXYYYZZZS\fP
1522
+ .TP
1523
+ \fIX\fP
1524
+ Version majeure.
1525
+ .TP
1526
+ \fIYYY\fP
1527
+ Version mineure. Les numéros pairs sont stables. Les numéros impairs sont
1528
+ des versions alpha ou beta.
1529
+ .TP
1530
+ \fIZZZ\fP
1531
+ Niveau de correctif pour les options stables ou juste un compteur pour les
1532
+ options de développement.
1533
+ .TP
1534
+ \fIS\fP
1535
+ Stabilité. 0 est alpha, 1 est bêta et 2 est stable. \fIS\fP devrait toujours
1536
+ être 2 quand \fIYYY\fP est pair.
1537
+ .PP
1538
+ \fIXYYYZZZS\fP sont identiques sur les deux lignes si \fBxz\fP et liblzma sont
1539
+ issus de la même version d'utilitaires XZ.
1540
+ .PP
1541
+ Exemples : 4.999.9beta est \fB49990091\fP et 5.0.0 est \fB50000002\fP.
1542
+ .
1543
+ .SH "STATUT DE SORTIE"
1544
+ .TP
1545
+ \fB0\fP
1546
+ Tout est bon.
1547
+ .TP
1548
+ \fB1\fP
1549
+ Une erreur est survenue.
1550
+ .TP
1551
+ \fB2\fP
1552
+ Quelquechose méritant un avertissement s'est produit, mais aucune erreur
1553
+ véritable n'est survenue.
1554
+ .PP
1555
+ Les notifications (pas les avertissements ou les erreurs) affichées sur
1556
+ l'erreur standard n'affectent pas le statut de sortie.
1557
+ .
1558
+ .SH ENVIRONNEMENT
1559
+ \fBxz\fP analyse les listes d'options séparées par des espaces à partir des
1560
+ variables d'environnement \fBXZ_DEFAULTS\fP et \fBXZ_OPT\fP, dans cet ordre, avant
1561
+ d'analyser les options de la ligne de commandes. Remarquez que seules les
1562
+ options sont analysées depuis l'environnement des variables ; toutes les
1563
+ non\-options sont ignorées silencieusement. L'analyse est faite avec
1564
+ \fBgetopt_long\fP(3) qui est aussi utilisé pour les arguments de la ligne de
1565
+ commandes.
1566
+ .TP
1567
+ \fBXZ_DEFAULTS\fP
1568
+ Options par défaut propres à l'utilisateur ou pour tout le système. Elles
1569
+ sont le plus souvent définies dans un script d'initialisation de
1570
+ l'interpréteur pour activer le limiteur d'utilisation de la mémoire de \fBxz\fP
1571
+ par défaut. A part pour les scripts d'initialisation de l'interpréteur ou
1572
+ des cas similaires, les sripts ne doivent jamais définir ou désactiver
1573
+ \fBXZ_DEFAULTS\fP.
1574
+ .TP
1575
+ \fBXZ_OPT\fP
1576
+ This is for passing options to \fBxz\fP when it is not possible to set the
1577
+ options directly on the \fBxz\fP command line. This is the case when \fBxz\fP is
1578
+ run by a script or tool, for example, GNU \fBtar\fP(1):
1579
+ .RS
1580
+ .RS
1581
+ .PP
1582
+ .nf
1583
+ \f(CRXZ_OPT=\-2v tar caf foo.tar.xz foo\fR
1584
+ .fi
1585
+ .RE
1586
+ .RE
1587
+ .IP ""
1588
+ Scripts may use \fBXZ_OPT\fP, for example, to set script\-specific default
1589
+ compression options. It is still recommended to allow users to override
1590
+ \fBXZ_OPT\fP if that is reasonable. For example, in \fBsh\fP(1) scripts one may
1591
+ use something like this:
1592
+ .RS
1593
+ .RS
1594
+ .PP
1595
+ .nf
1596
+ \f(CRXZ_OPT=${XZ_OPT\-"\-7e"}
1597
+ export XZ_OPT\fR
1598
+ .fi
1599
+ .RE
1600
+ .RE
1601
+ .
1602
+ .SH "Compatibilité des utilitaires LZMA"
1603
+ La syntaxe de la ligne de commande de \fBxz\fP est quasimment un sur\-ensemble
1604
+ de \fBlzma\fP, \fBunlzma\fP et \fBlzcat\fP comme ils sont trouvés dans les
1605
+ utilitaires LZMA 4.32.x . Dans la pluspart des cas, il est possible de
1606
+ remplacer les outils LZMA par les outils XZ sans casser les scripts
1607
+ existants. Il existe cependant certaines incompatibilités qui peuvent
1608
+ parfois poser des problèmes.
1609
+ .
1610
+ .SS "Niveaux de préréglage de la compression"
1611
+ La numérotation des préréglages de niveau de compression est différente
1612
+ entre les outils \fBxz\fP et LZMA. La différence la plus importante est la
1613
+ manière dont les tailles de dictionnaire sont affectées aux différents
1614
+ préréglages. La taille de dictionnaire est à peu près égale à celle
1615
+ d'utilisation de la mémoire de la décompression.
1616
+ .RS
1617
+ .PP
1618
+ .TS
1619
+ tab(;);
1620
+ c c c
1621
+ c n n.
1622
+ Niveau;xz;Utilitaires LZMA
1623
+ \-0;256 KiB;N/A
1624
+ \-1;1 MiB;64 KiB
1625
+ \-2;2 MiB;1 MiB
1626
+ \-3;4 MiB;512 KiB
1627
+ \-4;4 MiB;1 MiB
1628
+ \-5;8 MiB;2 MiB
1629
+ \-6;8 MiB;4 MiB
1630
+ \-7;16 MiB;8 MiB
1631
+ \-8;32 MiB;16 MiB
1632
+ \-9;64 MiB;32 MiB
1633
+ .TE
1634
+ .RE
1635
+ .PP
1636
+ Les différences de tailles des dictionnaires affectent aussi l'utilisation
1637
+ de la mémoire du compresseur, mais il y a quelques autres différences entre
1638
+ les outils LZMA et les outils XZ, qui rendent la différence encore plus
1639
+ grande :
1640
+ .RS
1641
+ .PP
1642
+ .TS
1643
+ tab(;);
1644
+ c c c
1645
+ c n n.
1646
+ Niveau;xz;Utilitaires LZMA 4.32.x
1647
+ \-0;3 MiB;N/A
1648
+ \-1;9 MiB;2 MiB
1649
+ \-2;17 MiB;12 MiB
1650
+ \-3;32 MiB;12 MiB
1651
+ \-4;48 MiB;16 MiB
1652
+ \-5;94 MiB;26 MiB
1653
+ \-6;94 MiB;45 MiB
1654
+ \-7;186 MiB;83 MiB
1655
+ \-8;370 MiB;159 MiB
1656
+ \-9;674 MiB;311 MiB
1657
+ .TE
1658
+ .RE
1659
+ .PP
1660
+ Le niveau de préréglage par défaut dans les outils LZMA est \fB\-7\fP alors que
1661
+ pour les outils XZ c'est \fB\-6\fP, les deux utilisent ainsi un dictionnaire de
1662
+ 8 Mio par défaut.
1663
+ .
1664
+ .SS "Fichiers .lzma en flux ou non"
1665
+ The uncompressed size of the file can be stored in the \fB.lzma\fP header.
1666
+ LZMA Utils does that when compressing regular files. The alternative is to
1667
+ mark that uncompressed size is unknown and use end\-of\-payload marker to
1668
+ indicate where the decompressor should stop. LZMA Utils uses this method
1669
+ when uncompressed size isn't known, which is the case, for example, in
1670
+ pipes.
1671
+ .PP
1672
+ \fBxz\fP prend en charge la décompression des fichiers \fB.lzma\fP avec ou sans
1673
+ marqueur de fin de charge utile, mais tous les fichiers \fB.lzma\fP créés par
1674
+ \fBxz\fP utiliseront un marqueur de fin de charge utile et ont la taille non
1675
+ compréssée marquée comme inconnue dans l'en\-tête \fB.lzma\fP. Cela peut être un
1676
+ problème dans quelques situations inhabituelles. Par exemple, un
1677
+ décompresseur \fB.lzma\fP dans un périphérique embarqué pourrait ne fonctionner
1678
+ qu'avec des fichiers dont la taille non comprimée est connue. Si vous vous
1679
+ heurtez à ce problème, vous devez utiliser les utilitaires LZMA ou LZMA SDK
1680
+ pour créer des fichiers \fB.lzma\fP avec une taille non compressée connue.
1681
+ .
1682
+ .SS "Fichiers .lzma non pris en charge"
1683
+ Le format \fB.lzma\fP autorise des valeurs \fIlc\fP jusqu'à 8, et des valeurs
1684
+ \fIlp\fP jusqu'à 4. Les outils LZMA peuvent décompresser des fichiers avec tous
1685
+ les \fIlc\fP et \fIlp\fP, mais créez toujours les fichiers avec \fBlc=3\fP et
1686
+ \fBlp=0\fP. Créer des fichiers avec d'autres valeurs \fIlc\fP et \fIlp\fP est
1687
+ possible avec \fBxz\fP et avec LZMA SDK.
1688
+ .PP
1689
+ L'implémentation du filtre LZMA1 dans liblzma nécessite que la somme de
1690
+ \fIlc\fP et \fIlp\fP soit inférieure ou égale à 4. Ainsi, les fichiers \fB.lzma\fP
1691
+ qui excèdent cette limitation ne peuvent pas être décompressés avec \fBxz\fP.
1692
+ .PP
1693
+ Les outils LZMA créent seulement des fichiers \fB.lzma\fP qui ont une taille de
1694
+ dictionnaire de 2^\fIn\fP (une puissance de 2) mais acceptent les fichiers avec
1695
+ toutes les tailles de dictionnaire. Libzlma n'accepte que les fichiers
1696
+ \&\fB.lzma\fP qui ont une taille dictionnaire de 2^\fIn\fP ou
1697
+ 2^\fIn\fP + 2^(\fIn\fP\-1). Cela afin de diminuer les faux positifs lors de la
1698
+ détection des fichiers \fB.lzma\fP.
1699
+ .PP
1700
+ Ces limitations ne devraient pas poser problème en pratique, car
1701
+ pratiquement tous les fichiers \fB.lzma\fP ont été compressés avec des réglages
1702
+ que liblzma accepte.
1703
+ .
1704
+ .SS "Déchets excédentaires"
1705
+ Lors de la décompession, l'utilitaire LZMA ignore silencieusement tout ce
1706
+ qui est après le premier flux \fB.lzma\fP. Dans la majorité des situations,
1707
+ c'est un bogue. Cela veut dire aussi que les outils LZMA ne gèrent pas la
1708
+ décompression de fichiers \fB.lzma\fP concaténés.
1709
+ .PP
1710
+ S'il reste des données après le premier flux \fB.lzma\fP, \fBxz\fP considère que
1711
+ le fichier est corrompu sauf si \fB\-\-single\-stream\fP a été utilisé. Cela peut
1712
+ casser des scripts obscurs qui ont supposé que les déchets de fin de ligne
1713
+ sont ignorés.
1714
+ .
1715
+ .SH NOTES
1716
+ .
1717
+ .SS "La sortie compressée peut varier"
1718
+ La sortie compressée exacte produite par les même fichiers non compressés en
1719
+ entrée peut varier en fonction des différentes versions de l'utilitaire XZ,
1720
+ même si les options de compression sont identiques. En effet, il est
1721
+ possible d'améliorer l'encodeur (compression plus rapide ou meilleure) sans
1722
+ affecter le format du fichier. La sortie peut même varier entre différentes
1723
+ compilations de la même version d'utilitaire XZ, si des options de
1724
+ construction différentes sont utilisées.
1725
+ .PP
1726
+ Cela signifie qu'une fois que \fB\-\-rsyncable\fP a été implémenté, les fichiers
1727
+ résultants ne seront pas nécessairement synchronisables avec rsync à moins
1728
+ que les nouveaux et anciens fichiers n'aient été compressés avec la même
1729
+ version de xz. Ce problème peut être résolu si une partie de
1730
+ l'implémentation est gelée pour garantir la stabilité de la sortie rsyncable
1731
+ à travers les versions de xz.
1732
+ .
1733
+ .SS "Décompresseurs .xz embarqués"
1734
+ Les implémentations de décompresseur embarqué comme XZ Embedded ne gèrent
1735
+ pas nécessairement les fichiers créés avec d'autres types de \fIvérification\fP
1736
+ d'intégrité que \fBnone\fP et \fBCRC32\fP. Comme la valeur par défaut est
1737
+ \fB\-\-check=crc64\fP, vous devez utiliser \fB\-\-check=none\fP ou \fB\-\-check=crc32\fP
1738
+ lors de la création de fichiers pour les systèmes embarqués.
1739
+ .PP
1740
+ En dehors des systèmes embarqués, tous les décompresseurs de format \fB.xz\fP
1741
+ gèrent tous les types de \fIvérification\fP ou sont au moins capables de
1742
+ décompresser le fichier sans effectuer la vérification d'intégrité si ce
1743
+ type de \fIvérification\fP particulière n'est pas pris en charge.
1744
+ .PP
1745
+ XZ Embedded prend en charge les filtres BCJ, mais seulement avec le décalage
1746
+ de départ par défaut.
1747
+ .
1748
+ .SH EXEMPLES
1749
+ .
1750
+ .SS Bases
1751
+ Compresser le fichier \fItoto\fP en \fItoto.xz\fP en utilisant le niveau de
1752
+ compression par défaut (\fB\-6\fP) et supprimer \fItoto\fP si la compression
1753
+ réussit :
1754
+ .RS
1755
+ .PP
1756
+ .nf
1757
+ \f(CRxz foo\fR
1758
+ .fi
1759
+ .RE
1760
+ .PP
1761
+ Décompresser \fIbidule.xz\fP en \fIbidule\fP et ne pas supprimer \fIbidule.xz\fP même
1762
+ si la compression réussit :
1763
+ .RS
1764
+ .PP
1765
+ .nf
1766
+ \f(CRxz \-dk bar.xz\fR
1767
+ .fi
1768
+ .RE
1769
+ .PP
1770
+ Create \fIbaz.tar.xz\fP with the preset \fB\-4e\fP (\fB\-4 \-\-extreme\fP), which is
1771
+ slower than the default \fB\-6\fP, but needs less memory for compression and
1772
+ decompression (48\ MiB and 5\ MiB, respectively):
1773
+ .RS
1774
+ .PP
1775
+ .nf
1776
+ \f(CRtar cf \- baz | xz \-4e > baz.tar.xz\fR
1777
+ .fi
1778
+ .RE
1779
+ .PP
1780
+ Un mélange de fichiers compressés et non compressés peuvent être
1781
+ décompressés vers la sortie standard avec une simple commande :
1782
+ .RS
1783
+ .PP
1784
+ .nf
1785
+ \f(CRxz \-dcf a.txt b.txt.xz c.txt d.txt.lzma > abcd.txt\fR
1786
+ .fi
1787
+ .RE
1788
+ .
1789
+ .SS "Compression en parallèle de plusieurs fichiers"
1790
+ Sur GNU et *BSD, \fBfind\fP(1) et \fBxargs\fP(1) peuvent être utilisés pour mettre
1791
+ en parallèle la compression de plusieurs fichiers :
1792
+ .RS
1793
+ .PP
1794
+ .nf
1795
+ \f(CRfind . \-type f \e! \-name '*.xz' \-print0 \e
1796
+ | xargs \-0r \-P4 \-n16 xz \-T1\fR
1797
+ .fi
1798
+ .RE
1799
+ .PP
1800
+ L'option \fBP\fP passée à \fBxargs\fP(1) fixe le nombre de processus \fBxz\fP en
1801
+ parallèles. La meilleure valeur pour l'option \fBn\fP dépend du nombre de
1802
+ fichiers à compresser. S\-il n'y a que quelques fichiers, la valeur sera
1803
+ probablement 1 ; avec des dizaines de milliers de fichiers, 100 ou même plus
1804
+ serait approprié pour réduire le nombre de processus \fBxz\fP que \fBxargs\fP(1)
1805
+ créera éventuellement.
1806
+ .PP
1807
+ L'option \fB\-T1\fP de \fBxz\fP est là pour le forcer en mode mono\-thread, car
1808
+ \fBxargs\fP(1) est utilisé pour contrôler la quantité de mise en parallèle.
1809
+ .
1810
+ .SS "Mode robot"
1811
+ Calculer combien d'octets ont été économisés au total après avoir compressé
1812
+ plusieurs fichiers :
1813
+ .RS
1814
+ .PP
1815
+ .nf
1816
+ \f(CRxz \-\-robot \-\-list *.xz | awk '/^totals/{print $5\-$4}'\fR
1817
+ .fi
1818
+ .RE
1819
+ .PP
1820
+ Un script peut vouloir savoir qu'il utilise une version suffisamment récente
1821
+ de \fBxz\fP. Le script \fBsh\fP(1) suivant vérifie que le numéro de version de
1822
+ l'outil \fBxz\fP soit au minimum 5.0.0. Cette méthode est compatible avec les
1823
+ vieilles versions bêta, qui ne gèrent pas l'option \fB\-\-robot\fP :
1824
+ .RS
1825
+ .PP
1826
+ .nf
1827
+ \f(CRif ! eval "$(xz \-\-robot \-\-version 2> /dev/null)" ||
1828
+ [ "$XZ_VERSION" \-lt 50000002 ]; then
1829
+ echo "Your xz is too old."
1830
+ fi
1831
+ unset XZ_VERSION LIBLZMA_VERSION\fR
1832
+ .fi
1833
+ .RE
1834
+ .PP
1835
+ Régler une limite d'utilisation de la mémoire pour la décompression en
1836
+ utilisant \fBXZ_OPT\fP, mais si une limite a déjà été définie, ne pas
1837
+ l'augmenter :
1838
+ .RS
1839
+ .PP
1840
+ .nf
1841
+ \f(CRNEWLIM=$((123 << 20))\ \ # 123 MiB
1842
+ OLDLIM=$(xz \-\-robot \-\-info\-memory | cut \-f3)
1843
+ if [ $OLDLIM \-eq 0 \-o $OLDLIM \-gt $NEWLIM ]; then
1844
+ XZ_OPT="$XZ_OPT \-\-memlimit\-decompress=$NEWLIM"
1845
+ export XZ_OPT
1846
+ fi\fR
1847
+ .fi
1848
+ .RE
1849
+ .
1850
+ .SS "Chaînes de filtres de compresseur personnalisées"
1851
+ L'utilisation la plus simple des chaînes de filtres personnalisées est la
1852
+ personnalisation d'un préréglage LZMA2. Cela peut être utile, car les
1853
+ préréglages ne couvrent qu'un sous\-ensemble des réglages de compression
1854
+ potentiellement utiles.
1855
+ .PP
1856
+ Les colonnes CompCPU des tableaux des descriptions des options \fB\-0\fP à \fB\-9\fP
1857
+ et \fB\-\-extreme\fP sont utiles lors de la personnalisation des préréglages
1858
+ LZMA2. Voici les parties pertinentes recueillies à partir de ces deux
1859
+ tableaux :
1860
+ .RS
1861
+ .PP
1862
+ .TS
1863
+ tab(;);
1864
+ c c
1865
+ n n.
1866
+ Préréglage;CompCPU
1867
+ \-0;0
1868
+ \-1;1
1869
+ \-2;2
1870
+ \-3;3
1871
+ \-4;4
1872
+ \-5;5
1873
+ \-6;6
1874
+ \-5e;7
1875
+ \-6e;8
1876
+ .TE
1877
+ .RE
1878
+ .PP
1879
+ If you know that a file requires somewhat big dictionary (for example, 32\ MiB) to compress well, but you want to compress it quicker than \fBxz \-8\fP
1880
+ would do, a preset with a low CompCPU value (for example, 1) can be
1881
+ modified to use a bigger dictionary:
1882
+ .RS
1883
+ .PP
1884
+ .nf
1885
+ \f(CRxz \-\-lzma2=preset=1,dict=32MiB foo.tar\fR
1886
+ .fi
1887
+ .RE
1888
+ .PP
1889
+ Avec certains fichiers, la commande ci\-dessus peut être plus rapide que
1890
+ \fBxz\-6\fP tout en compressant bien mieux. Cependant, il faut souligner que
1891
+ seuls certains fichiers bénéficient d'un grand dictionnaire tout en gardant
1892
+ la valeur de CompCPU faible. La siutation la plus évidente où un gros
1893
+ dictionnaire peut baucoup aider, est une archive contenant des fichiers très
1894
+ similaires de quelques megaoctets chacun. La taille de dictionnaire doit
1895
+ être significativement plus grosse que tout fichier individuel pour
1896
+ permettre à LZMA2 de tirer pleinement partie des similarités entre des
1897
+ fichiers consécutifs.
1898
+ .PP
1899
+ Si une utilisation de la mémoire élevée pour la compression et décompression
1900
+ convient, et que le fichier à compresser a une taille de plusieurs centaines
1901
+ de megaoctets, il peut être utile d'utiliser un plus gros dictionnaire que
1902
+ celui fourni par \fBxz\-9\fP (64 Mio) :
1903
+ .RS
1904
+ .PP
1905
+ .nf
1906
+ \f(CRxz \-vv \-\-lzma2=dict=192MiB big_foo.tar\fR
1907
+ .fi
1908
+ .RE
1909
+ .PP
1910
+ Utiliser \fB\-vv\fP (\fB\-\-verbose\-\-verbose\fP) comme dans l'exemple ci\-dessus peut
1911
+ être utile pour voir les besoins en mémoire du compresseur et du
1912
+ décompresseur. Rappelez\-vous qu'utiliser un dictionnaire plus gros que la
1913
+ taille du fichier non compressé est un gachis de mémoire, donc la commande
1914
+ ci\-dessus n'est pas utile pour les petits fichiers.
1915
+ .PP
1916
+ Sometimes the compression time doesn't matter, but the decompressor memory
1917
+ usage has to be kept low, for example, to make it possible to decompress the
1918
+ file on an embedded system. The following command uses \fB\-6e\fP (\fB\-6 \-\-extreme\fP) as a base and sets the dictionary to only 64\ KiB. The
1919
+ resulting file can be decompressed with XZ Embedded (that's why there is
1920
+ \fB\-\-check=crc32\fP) using about 100\ KiB of memory.
1921
+ .RS
1922
+ .PP
1923
+ .nf
1924
+ \f(CRxz \-\-check=crc32 \-\-lzma2=preset=6e,dict=64KiB foo\fR
1925
+ .fi
1926
+ .RE
1927
+ .PP
1928
+ If you want to squeeze out as many bytes as possible, adjusting the number
1929
+ of literal context bits (\fIlc\fP) and number of position bits (\fIpb\fP) can
1930
+ sometimes help. Adjusting the number of literal position bits (\fIlp\fP)
1931
+ might help too, but usually \fIlc\fP and \fIpb\fP are more important. For
1932
+ example, a source code archive contains mostly US\-ASCII text, so something
1933
+ like the following might give slightly (like 0.1\ %) smaller file than \fBxz \-6e\fP (try also without \fBlc=4\fP):
1934
+ .RS
1935
+ .PP
1936
+ .nf
1937
+ \f(CRxz \-\-lzma2=preset=6e,pb=0,lc=4 source_code.tar\fR
1938
+ .fi
1939
+ .RE
1940
+ .PP
1941
+ Using another filter together with LZMA2 can improve compression with
1942
+ certain file types. For example, to compress a x86\-32 or x86\-64 shared
1943
+ library using the x86 BCJ filter:
1944
+ .RS
1945
+ .PP
1946
+ .nf
1947
+ \f(CRxz \-\-x86 \-\-lzma2 libfoo.so\fR
1948
+ .fi
1949
+ .RE
1950
+ .PP
1951
+ Notez que l'ordre des options de filtre est significatif. Si \fB\-\-x86\fP est
1952
+ indiqué après \fB\-\-lzma2\fP, \fBxz\fP donnera une erreur, car il ne peut y avoir
1953
+ aucun filtre après LZMA2, et aussi parce que le filtre BCJ x86 ne peut pas
1954
+ être utilisé comme dernier filtre dans la chaîne.
1955
+ .PP
1956
+ Le filtre Delta associé à LZMA2 peut donner de bons résultats avec les
1957
+ images bitmap. Cela devrait habituellement battre PNG, qui a quelques
1958
+ filtres avancés supplémentaires qu'un simple delta, mais qui utilise Deflate
1959
+ pour la compression effective.
1960
+ .PP
1961
+ The image has to be saved in uncompressed format, for example, as
1962
+ uncompressed TIFF. The distance parameter of the Delta filter is set to
1963
+ match the number of bytes per pixel in the image. For example, 24\-bit RGB
1964
+ bitmap needs \fBdist=3\fP, and it is also good to pass \fBpb=0\fP to LZMA2 to
1965
+ accommodate the three\-byte alignment:
1966
+ .RS
1967
+ .PP
1968
+ .nf
1969
+ \f(CRxz \-\-delta=dist=3 \-\-lzma2=pb=0 foo.tiff\fR
1970
+ .fi
1971
+ .RE
1972
+ .PP
1973
+ If multiple images have been put into a single archive (for example,
1974
+ \&\fB.tar\fP), the Delta filter will work on that too as long as all images have
1975
+ the same number of bytes per pixel.
1976
+ .
1977
+ .SH "VOIR AUSSI"
1978
+ \fBxzdec\fP(1), \fBxzdiff\fP(1), \fBxzgrep\fP(1), \fBxzless\fP(1), \fBxzmore\fP(1),
1979
+ \fBgzip\fP(1), \fBbzip2\fP(1), \fB7z\fP(1)
1980
+ .PP
1981
+ XZ Utils: <https://tukaani.org/xz/>
1982
+ .br
1983
+ XZ Embedded: <https://tukaani.org/xz/embedded.html>
1984
+ .br
1985
+ LZMA SDK: <https://7\-zip.org/sdk.html>
janus/share/man/fr/man1/unxz.1 ADDED
@@ -0,0 +1,1985 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ '\" t
2
+ .\" SPDX-License-Identifier: 0BSD
3
+ .\"
4
+ .\" Authors: Lasse Collin
5
+ .\" Jia Tan
6
+ .\"
7
+ .\" french translation of XZ Utils man
8
+ .\" Copyright (C) 2021 Debian French l10n team <debian-l10n-french@lists.debian.org>
9
+ .\" Translator
10
+ .\" bubu <bubub@no-log.org>, 2021.
11
+ .\"
12
+ .\"*******************************************************************
13
+ .\"
14
+ .\" This file was generated with po4a. Translate the source file.
15
+ .\"
16
+ .\"*******************************************************************
17
+ .TH XZ 1 2024\-12\-30 Tukaani "Utilitaires XZ"
18
+ .
19
+ .SH NOM
20
+ xz, unxz, xzcat, lzma, unlzma, lzcat \- Compresser ou décompresser des
21
+ fichiers .xz et .lzma
22
+ .
23
+ .SH SYNOPSIS
24
+ \fBxz\fP [\fIoption...\fP] [\fIfichier...\fP]
25
+ .
26
+ .SH "ALIAS DES COMMANDES"
27
+ \fBunxz\fP est équivalent à \fBxz \-\-decompress\fP.
28
+ .br
29
+ \fBxzcat\fP est équivalent à \fBxz \-\-decompress \-\-stdout\fP
30
+ .br
31
+ \fBlzma\fP est équivalent à \fBxz \-\-format=lzma\fP
32
+ .br
33
+ \fBunlzma\fP est équivalent à \fBxz \-\-format=lzma \-\-decompress\fP
34
+ .br
35
+ \fBlzcat\fP est équivalent à \fBxz \-\-format=lzma \-\-decompress \-\- stdout\fP
36
+ .PP
37
+ Lors de l'écriture de scripts qui nécessitent de décompresser des fichiers,
38
+ il est recommandé de toujours utiliser la commande \fBxz\fP avec les arguments
39
+ appropriés (\fBxz \-d\fP ou \fBxz \-dc\fP) au lieu des commandes \fBunxz\fP et
40
+ \fBxzcat\fP.
41
+ .
42
+ .SH DESCRIPTION
43
+ \fBxz\fP is a general\-purpose data compression tool with command line syntax
44
+ similar to \fBgzip\fP(1) and \fBbzip2\fP(1). The native file format is the
45
+ \&\fB.xz\fP format, but the legacy \fB.lzma\fP format used by LZMA Utils and raw
46
+ compressed streams with no container format headers are also supported. In
47
+ addition, decompression of the \fB.lz\fP format used by \fBlzip\fP is supported.
48
+ .PP
49
+ \fBxz\fP compresse ou décompresse chaque \fIfichier\fP en fonction du mode
50
+ d'opération choisi. Si aucun \fIfichier\fP n'est donné ou \fIfichier\fP est \fB\-\fP,
51
+ \fBxz\fP lit depuis l'entrée standard et écrit les données traitées sur la
52
+ sortie standard. \fBxz\fP refusera (affichera une erreur et ignorera le
53
+ \fIfichier\fP) d'écrire les données compressées sur la sortie standard si c'est
54
+ un terminal. De même, \fBxz\fP refusera de lire des données compressées depuis
55
+ l'entrée standard si c'est un terminal.
56
+ .PP
57
+ A moins que \fB\-\-sdout\fP ne soit indiqué, les \fIfichiers\fP autres que \fB\-\fP sont
58
+ écrits dans un nouveau fichier dont le nom est dérivé du nom de \fIfichier\fP
59
+ source :
60
+ .IP \(bu 3
61
+ Lors de la compression, le suffixe du format de fichier cible (\fB.xz\fP ou
62
+ \&\fB.lzma\fP) est ajouté au nom de fichier source pour obtenir le nom de fichier
63
+ cible.
64
+ .IP \(bu 3
65
+ When decompressing, the \fB.xz\fP, \fB.lzma\fP, or \fB.lz\fP suffix is removed from
66
+ the filename to get the target filename. \fBxz\fP also recognizes the suffixes
67
+ \&\fB.txz\fP and \fB.tlz\fP, and replaces them with the \fB.tar\fP suffix.
68
+ .PP
69
+ Si le fichier cible existe déjà, une erreur est affichée et le \fIfichier\fP
70
+ est ignoré.
71
+ .PP
72
+ Sauf s'il écrit dans la sortie standard, \fBxz\fP affichera un avertissement et
73
+ ignorera le \fIfichier\fP dans les cas suivants :
74
+ .IP \(bu 3
75
+ \fIfichier\fP n'est pas un fichier normal. Les liens symboliques ne sont pas
76
+ suivis et donc ne sont pas considérés comme des fichiers normaux.
77
+ .IP \(bu 3
78
+ \fIfichier\fP a plusieurs liens physiques.
79
+ .IP \(bu 3
80
+ \fIfichier\fP a un setuid, setgid ou sticky bit positionné.
81
+ .IP \(bu 3
82
+ Le mode d'opération est défini pour compresser et le \fIfichier\fP a déjà un
83
+ suffixe du format de fichier cible (\fB.xz\fP ou \fB.txz\fP lors d'une compression
84
+ en format \fB.xz\fP, et \fB.lzma\fP ou \fB.tlz\fP lors d'une compression en format
85
+ \&\fB.lzma\fP).
86
+ .IP \(bu 3
87
+ The operation mode is set to decompress and the \fIfile\fP doesn't have a
88
+ suffix of any of the supported file formats (\fB.xz\fP, \fB.txz\fP, \fB.lzma\fP,
89
+ \&\fB.tlz\fP, or \fB.lz\fP).
90
+ .PP
91
+ Après la compression ou la décompression réussie du \fIfichier\fP, \fBxz\fP copie
92
+ les permissions du propriétaire, du groupe, la date d'accès, et les
93
+ modifications d'heure depuis le \fIfichier\fP source du fichier cible. Si la
94
+ copie du groupe échoue, les permissions sont modifiées pour que le fichier
95
+ cible ne soit pas accessible aux utilisateurs qui n'ont pas les droits
96
+ d'accès au \fIfichier\fP source. \fBxz\fP ne prend actuellement pas en charge la
97
+ copie d'autres métadonnées telles que les listes de contrôle d'accès ou les
98
+ attributs étendus.
99
+ .PP
100
+ Once the target file has been successfully closed, the source \fIfile\fP is
101
+ removed unless \fB\-\-keep\fP was specified. The source \fIfile\fP is never removed
102
+ if the output is written to standard output or if an error occurs.
103
+ .PP
104
+ Envoyer \fBSIGINFO\fP ou \fBSIGURSR1\fP au processus \fBxz\fP, lui fait afficher
105
+ l'information de progression sur l'erreur standard. Cela a un intérêt limité
106
+ car lorsque l'erreur standard est un terminal, utiliser \fB\-\-verbose\fP
107
+ affichera automatiquement un indicateur de progression du processus.
108
+ .
109
+ .SS "Utilisation de la mémoire"
110
+ L'utilisation de la mémoire par \fBxz\fP varie de quelques centaines de
111
+ kilo\-octets à plusieurs gigaoctects en fonction des paramètres de
112
+ compression. Les réglages utilisés lors de la compression d'un fichier
113
+ déterminent les besoins en mémoire pour la décompression. Habituellement la
114
+ décompression nécessite 5\% à 20\% de la quantité de mémoire utilisée pour
115
+ la compression du fichier. Par exemple, décompresser un fichier créé avec
116
+ \fBxz\-9\fP recquiert habituellement 65\ Mio de mémoire. Bien qu'il soit
117
+ possible d'avoir des fichiers \fB.xz\fP nécessitant plusieurs gigaoctets de
118
+ mémoire pour être décompressés.
119
+ .PP
120
+ Especially users of older systems may find the possibility of very large
121
+ memory usage annoying. To prevent uncomfortable surprises, \fBxz\fP has a
122
+ built\-in memory usage limiter, which is disabled by default. While some
123
+ operating systems provide ways to limit the memory usage of processes,
124
+ relying on it wasn't deemed to be flexible enough (for example, using
125
+ \fBulimit\fP(1) to limit virtual memory tends to cripple \fBmmap\fP(2)).
126
+ .PP
127
+ The memory usage limiter can be enabled with the command line option
128
+ \fB\-\-memlimit=\fP\fIlimit\fP. Often it is more convenient to enable the limiter
129
+ by default by setting the environment variable \fBXZ_DEFAULTS\fP, for example,
130
+ \fBXZ_DEFAULTS=\-\-memlimit=150MiB\fP. It is possible to set the limits
131
+ separately for compression and decompression by using
132
+ \fB\-\-memlimit\-compress=\fP\fIlimit\fP and \fB\-\-memlimit\-decompress=\fP\fIlimit\fP.
133
+ Using these two options outside \fBXZ_DEFAULTS\fP is rarely useful because a
134
+ single run of \fBxz\fP cannot do both compression and decompression and
135
+ \fB\-\-memlimit=\fP\fIlimit\fP (or \fB\-M\fP \fIlimit\fP) is shorter to type on the
136
+ command line.
137
+ .PP
138
+ If the specified memory usage limit is exceeded when decompressing, \fBxz\fP
139
+ will display an error and decompressing the file will fail. If the limit is
140
+ exceeded when compressing, \fBxz\fP will try to scale the settings down so that
141
+ the limit is no longer exceeded (except when using \fB\-\-format=raw\fP or
142
+ \fB\-\-no\-adjust\fP). This way the operation won't fail unless the limit is very
143
+ small. The scaling of the settings is done in steps that don't match the
144
+ compression level presets, for example, if the limit is only slightly less
145
+ than the amount required for \fBxz \-9\fP, the settings will be scaled down only
146
+ a little, not all the way down to \fBxz \-8\fP.
147
+ .
148
+ .SS "Concaténation et remplissage avec des fichiers .xz"
149
+ Il est possible de concaténer les fichiers \fB.xz\fP tels quel. \fBxz\fP
150
+ décompressera de tels fichiers comme s'ils étaient un unique fichier \fB.xz\fP.
151
+ .PP
152
+ It is possible to insert padding between the concatenated parts or after the
153
+ last part. The padding must consist of null bytes and the size of the
154
+ padding must be a multiple of four bytes. This can be useful, for example,
155
+ if the \fB.xz\fP file is stored on a medium that measures file sizes in
156
+ 512\-byte blocks.
157
+ .PP
158
+ La concaténation et le remplissage ne sont pas autorisés avec les fichiers
159
+ \&\fB.lzma\fP ou les flux bruts.
160
+ .
161
+ .SH OPTIONS
162
+ .
163
+ .SS "Suffixes entiers et valeurs spéciales."
164
+ Dans la plupart des endroits où un argument entier est attendu, un suffixe
165
+ optionel permet d'indiquer facilement les grands entiers. Il ne doit pas y
166
+ avoir d'espace entre l'entier et le suffixe.
167
+ .TP
168
+ \fBKiB\fP
169
+ Multiplier l'entier par 1024 (2^10). \fBKi\fP, \fBk\fP, \fBkB\fP, \fBK\fP et \fBKB\fP sont
170
+ acceptés comme synonymes de \fBKiB\fP.
171
+ .TP
172
+ \fBMiB\fP
173
+ Multiplier l'entier par 1 048 576 (2^20). \fBMi\fP, \fBm\fP, \fBM\fP et \fBMB\fP sont
174
+ acceptés comme synonymes de \fBMiB\fP.
175
+ .TP
176
+ \fBGiB\fP
177
+ Multiplier l'entier par 1 073 741 824 (2^30). \fBGi\fP, \fBg\fP, \fBG\fP et \fBGB\fP
178
+ sont acceptés comme synonymes de \fBGiB\fP.
179
+ .PP
180
+ La valeur spéciale \fBmax\fP peut être utilisée pour indiquer la valeur
181
+ maximale de l'entier prise en charge par l'option.
182
+ .
183
+ .SS "Mode d'opération"
184
+ Si plusieurs options de mode d'opération sont données, la dernière prend
185
+ effet.
186
+ .TP
187
+ \fB\-z\fP, \fB\-\-compress\fP
188
+ Compresser. C'est le mode d'opération par défaut lorsque aucune option de
189
+ mode opératoire n'est spécifiée ou qu'aucun autre mode d'opération n'est
190
+ sous\-entendu par le nom de la commande (par exemple \fBunxz\fP sous\-entend
191
+ \fB\-\-decompress\fP).
192
+ .IP ""
193
+ .\" The DESCRIPTION section already says this but it's good to repeat it
194
+ .\" here because the default behavior is a bit dangerous and new users
195
+ .\" in a hurry may skip reading the DESCRIPTION section.
196
+ After successful compression, the source file is removed unless writing to
197
+ standard output or \fB\-\-keep\fP was specified.
198
+ .TP
199
+ \fB\-d\fP, \fB\-\-decompress\fP, \fB\-\-uncompress\fP
200
+ .\" The DESCRIPTION section already says this but it's good to repeat it
201
+ .\" here because the default behavior is a bit dangerous and new users
202
+ .\" in a hurry may skip reading the DESCRIPTION section.
203
+ Decompress. After successful decompression, the source file is removed
204
+ unless writing to standard output or \fB\-\-keep\fP was specified.
205
+ .TP
206
+ \fB\-t\fP, \fB\-\-test\fP
207
+ Tester l'intégrité des \fIfichiers\fP compressés. Cette option est équivalente
208
+ à \fB\-\-decompress \-\-stdout\fP sauf que les données décompressées sont rejetées
209
+ au lieu d'être écrites sur la sortie standard. Aucun fichier n'est créé ou
210
+ supprimé.
211
+ .TP
212
+ \fB\-l\fP, \fB\-\-list\fP
213
+ Afficher l'information sur les \fIfichiers\fP compressés. Aucune sortie
214
+ non\-compressée n'est produite et aucun fichier n'est créé ou supprimé. En
215
+ mode liste, le programme ne peut pas lire les données compressées depuis
216
+ l'entrée standard ou depuis d'autres sources non adressables.
217
+ .IP ""
218
+ The default listing shows basic information about \fIfiles\fP, one file per
219
+ line. To get more detailed information, use also the \fB\-\-verbose\fP option.
220
+ For even more information, use \fB\-\-verbose\fP twice, but note that this may be
221
+ slow, because getting all the extra information requires many seeks. The
222
+ width of verbose output exceeds 80 characters, so piping the output to, for
223
+ example, \fBless\ \-S\fP may be convenient if the terminal isn't wide enough.
224
+ .IP ""
225
+ La sortie exacte peut varier suivant les versions de \fBxz\fP et les différents
226
+ paramètres régionaux. Pour une sortie lisible par la machine, utiliser
227
+ \fB\-\-robot \-\-list\fP.
228
+ .
229
+ .SS "Modificateurs d'opération"
230
+ .TP
231
+ \fB\-k\fP, \fB\-\-keep\fP
232
+ Ne pas effacer les fichiers d'entrée.
233
+ .IP ""
234
+ Since \fBxz\fP 5.2.6, this option also makes \fBxz\fP compress or decompress even
235
+ if the input is a symbolic link to a regular file, has more than one hard
236
+ link, or has the setuid, setgid, or sticky bit set. The setuid, setgid, and
237
+ sticky bits are not copied to the target file. In earlier versions this was
238
+ only done with \fB\-\-force\fP.
239
+ .TP
240
+ \fB\-f\fP, \fB\-\-force\fP
241
+ Cette option a plusieurs effets :
242
+ .RS
243
+ .IP \(bu 3
244
+ Si le fichier cible existe déjà, l'effacer avant de compresser ou
245
+ décompresser.
246
+ .IP \(bu 3
247
+ Compresser ou décompresser même si l'entrée est un lien symbolique vers un
248
+ fichier normal, a plus qu'un lien physique, ou a le bit setuid, setgid ou
249
+ sticky défini. Les bits setuid, setgid et sticky bits ne sont pas copiés
250
+ dans le fichier cible.
251
+ .IP \(bu 3
252
+ Lorsque \fBxz\fP est utilisé avec \fB\-\-decompress\fP \fB\-\-stdout\fP et qu'il ne peut
253
+ pas reconnaitre le type du fichier source, copier le fichier source tel quel
254
+ dans la sortie standard. Celà permet à \fBxzcat\fP \fB\-\-force\fP d'être utilisé
255
+ comme \fBcat\fP(1) pour les fichiers qui n'ont pas été compressé avec
256
+ \fBxz\fP. Remarquez que dans le futur, \fBxz\fP devrait prendre en charge de
257
+ nouveaux formats de fichiers compressés, ce qui permettra à \fBxz\fP de
258
+ décompresser plus de types de fichiers au lieu de les copier tels quels dans
259
+ la sortie standard. \fB\-\-format=\fP\fIformat\fP peut être utilisé pour contraindre
260
+ \fBxz\fP à décompresser seulement un format de fichier.
261
+ .RE
262
+ .TP
263
+ \fB\-c\fP, \fB\-\-stdout\fP, \fB\-\-to\-stdout\fP
264
+ Écrire les données compressées ou décompressées sur la sortie standard
265
+ plutôt que dans un fichier. Cela necessite \fB\-\-keep\fP.
266
+ .TP
267
+ \fB\-\-single\-stream\fP
268
+ Décompresser seulement le premier flux \fB.xz\fP et ignorer silencieusement les
269
+ possibles données d'entrée résiduelles qui suivent le flux. Normalement ces
270
+ déchets excédentaires provoquent l'affichage d'une erreur par \fBxz\fP.
271
+ .IP ""
272
+ \fBxz\fP ne décompresse jamais plus d'un flux à partir de fichiers \fB.lzma\fP ou
273
+ de flux bruts, mais cette option fait aussi que \fBxz\fP ignorera les données
274
+ résiduelles après le fichier \fB.lzma\fP ou le flux brut.
275
+ .IP ""
276
+ Cette option n'a aucun effet si le mode d'opération n'est pas
277
+ \fB\-\-decompress\fP ou \fB\-\-test\fP.
278
+ .TP
279
+ \fB\-\-no\-sparse\fP
280
+ Désactiver la création de fichiers peu denses. Par défaut, lors de la
281
+ décompression en un fichier normal, \fBxz\fP essaie d'en faire un fichier creux
282
+ si les données décompressées contiennent de longues séquences de zéros
283
+ binaires. Cela fonctionne aussi lors de l'écriture sur la sortie standard
284
+ aussi longtemps que la sortie standard est connectée à un fichier normal et
285
+ que certaines conditions supplémentaires sont satisfaites pour le faire de
286
+ manière sécurisée. Créer des fichiers creux peut épargner de l'espace disque
287
+ et accélérer la décompression en réduisant la quantité d'entrées/sorties sur
288
+ le disque.
289
+ .TP
290
+ \fB\-S\fP \fI.suf\fP, \fB\-\-suffix=\fP\fI.suf\fP
291
+ Lors de la compression, utiliser \fB.suf\fP comme suffixe du fichier cible au
292
+ lieu de \fB.xz\fP ou \fB.lzma\fP. Si \fBxz\fP n'écrit pas sur la sortie standard et
293
+ si le fichier source a déja le suffixe \fB.suf\fP, un avertissement est affiché
294
+ et le fichier est ignoré.
295
+ .IP ""
296
+ When decompressing, recognize files with the suffix \fI.suf\fP in addition to
297
+ files with the \fB.xz\fP, \fB.txz\fP, \fB.lzma\fP, \fB.tlz\fP, or \fB.lz\fP suffix. If the
298
+ source file has the suffix \fI.suf\fP, the suffix is removed to get the target
299
+ filename.
300
+ .IP ""
301
+ Lors de la compression ou décompression de flux bruts (\fB\-\-fomat=raw\fP), le
302
+ suffixe doit toujours être spécifié à moins d'écrire sur la sortie standard,
303
+ car il n'y a pas de suffixe par défaut pour les flux bruts.
304
+ .TP
305
+ \fB\-\-files\fP[\fB=\fP\fIfichier\fP]
306
+ Lire les noms de fichier à traiter depuis \fIfichier\fP ; si \fIfichier\fP est
307
+ omis , les noms de fichier sont lus sur l'entrée standard. Les noms de
308
+ fichier doivent se terminer avec le caractère de nouvelle ligne. Un tiret
309
+ (\fB\-\fP) est considéré comme un nom de fichier normal ; ce qui ne signifie pas
310
+ entrée standard. Si les noms de fichier sont aussi donnés comme arguments de
311
+ ligne de commande, ils sont traités avant les noms de fichier lus depuis
312
+ \fIfichier\fP.
313
+ .TP
314
+ \fB\-\-files0\fP[\fB=\fP\fIfichier\fP]
315
+ Cela est identique à \fB\-\-files\fP[\fB=\fP\fIfichier\fP] sauf que chaque nom de
316
+ fichier doit se terminer par le caractère null.
317
+ .
318
+ .SS "Format de fichier basique et options de compression"
319
+ .TP
320
+ \fB\-F\fP \fIformat\fP, \fB\-\-format=\fP\fIformat\fP
321
+ Indiquer le \fIformat\fP de fichier à compresser ou décompresser :
322
+ .RS
323
+ .TP
324
+ \fBauto\fP
325
+ C'est celui par défaut. Lors de la compression, \fBauto\fP est équivalent à
326
+ \fBxz\fP. Lors de la décompression, le format du fichier en entrée est détecté
327
+ automatiquement. Notez que les flux bruts (créés avec \fB\-\-format=raw\fP) ne
328
+ peuvent pas être détectés automatiquement.
329
+ .TP
330
+ \fBxz\fP
331
+ Compresser dans le format de fichier \fB.xz\fP ou n'accepter que les fichiers
332
+ \&\fB.xz\fP à décompresser.
333
+ .TP
334
+ \fBlzma\fP, \fBalone\fP
335
+ Compresser au format de fichier \fB.lzma\fP historique, ou n'accepter que les
336
+ fichiers \fB.lzma\fP lors de la décompression. Le nom alternatif \fBalone\fP est
337
+ fourni pour la rétrocompatibilité avec les utilitaires LZMA.
338
+ .TP
339
+ \fBlzip\fP
340
+ Accept only \fB.lz\fP files when decompressing. Compression is not supported.
341
+ .IP ""
342
+ The \fB.lz\fP format version 0 and the unextended version 1 are supported.
343
+ Version 0 files were produced by \fBlzip\fP 1.3 and older. Such files aren't
344
+ common but may be found from file archives as a few source packages were
345
+ released in this format. People might have old personal files in this
346
+ format too. Decompression support for the format version 0 was removed in
347
+ \fBlzip\fP 1.18.
348
+ .IP ""
349
+ \fBlzip\fP 1.4 and later create files in the format version 1. The sync flush
350
+ marker extension to the format version 1 was added in \fBlzip\fP 1.6. This
351
+ extension is rarely used and isn't supported by \fBxz\fP (diagnosed as corrupt
352
+ input).
353
+ .TP
354
+ \fBraw\fP
355
+ Compresser ou décompresser un flux brut (sans en\-têtes). Cela est réservé
356
+ seulement aux utilisateurs aguerris. Pour décoder des flux bruts, vous devez
357
+ utiliser \fB\-\-format=raw\fP et spécifier explicitement la chaîne de filtre, qui
358
+ normalement aurait du être stockée dans les en\-têtes du conteneur.
359
+ .RE
360
+ .TP
361
+ \fB\-C\fP \fIvérif.\fP, \fB\-\-check=\fP\fIvérif.\fP
362
+ Spécifier le type d'intégrité à vérifier. La vérification est calculée à
363
+ partir des données non\-compressées et stockées dans le fichier \fB.xz\fP. Cette
364
+ option n'a effet que si la compression a été faite dans le format \fB.xz\fP ;
365
+ le format \fB.lzma\fP ne gère pas les vérifications d'intégrité. Le contrôle
366
+ d'intégrité (s'il y en a) est vérifié lorsque le fichier \fB.xz\fP est
367
+ décompressé.
368
+ .IP ""
369
+ Types de \fIvérification\fP pris en charge :
370
+ .RS
371
+ .TP
372
+ \fBnone\fP
373
+ Ne pas calculer de vérification d'intégrité du tout. C'est généralement une
374
+ mauvaise idée. Cela peut être utile lorsque l'intégrité des données est
375
+ vérifiée de toute façon par d'autres manières.
376
+ .TP
377
+ \fBcrc32\fP
378
+ Calculer CRC32 en utilisant le polynôme de IEEE\-802.3 (Ethernet).
379
+ .TP
380
+ \fBcrc64\fP
381
+ Calculer CRC64 en utilisant le polynôme de ECMA\-182. C'est la manière
382
+ utilisée par défaut, car c'est légèrement mieux que CRC32 pour détecter les
383
+ fichiers endommagés et la différence de vitesse est négligeable.
384
+ .TP
385
+ \fBsha256\fP
386
+ Calculer SHA\-256. C'est quelque peu plus lent que CRC32 et CRC64.
387
+ .RE
388
+ .IP ""
389
+ L'intégrité des en\-têtes \fB.xz\fP est toujours vérifiée avec CRC32. Il n'est
390
+ pas possible de le changer ou de le désactiver.
391
+ .TP
392
+ \fB\-\-ignore\-check\fP
393
+ Ne pas contrôler la vérification d'intégrité des données lors de la
394
+ décompression. Les valeurs CRC32 dans les en\-têtes \fB.xz\fP seront normalement
395
+ toujours vérifiées.
396
+ .IP ""
397
+ \fBN'utilisez pas cette option à moins de savoir ce que vous faites.\fP Les
398
+ raisons possibles pour utiliser cette option :
399
+ .RS
400
+ .IP \(bu 3
401
+ Essayer de récupérer des données d'un fichier .xz corrompu.
402
+ .IP \(bu 3
403
+ Accélérer la décompression. Cela importe surtout avec SHA\-256 ou avec les
404
+ fichiers qui ont été compressés extrêmement bien. Il est recommandé de ne
405
+ pas utiliser cette option dans ce but à moins que l'intégrité du fichier ne
406
+ soit vérifiée extérieurement d'une autre manière.
407
+ .RE
408
+ .TP
409
+ \fB\-0\fP ... \fB\-9\fP
410
+ Choisir un niveau de compression prédéfini. La valeur par défaut est
411
+ \fB6\fP. Si plusieurs niveaux de préréglage sont spécifi��s, c'est le dernier
412
+ qui sera pris en compte. Si une chaîne de filtres personnalisée a déjà été
413
+ choisie, définir un niveau de compression préréglé efface la chaîne de
414
+ filtres personnalisée.
415
+ .IP ""
416
+ Les différences entre les préréglages sont plus significatives qu'avec
417
+ \fBgzip\fP(1) et \fBbzip2\fP(1). les réglages de compression sélectionnés
418
+ déterminent les exigences en mémoire pour la décompression, ainsi, utiliser
419
+ un niveau de préréglage trop élevé peut rendre difficile à décompresser un
420
+ fichier sur un vieux système avec peu de RAM. Clairement, \fBce n'est pas une bonne idée d'utiliser \-9 aveuglément pour tout\fP comme ça l'est souvent avec
421
+ \fBgzip\fP(1) et \fBbzip2\fP(1).
422
+ .RS
423
+ .TP
424
+ \fB\-0\fP ... \fB\-3\fP
425
+ Ce sont des préréglages relativement rapides. \fB0\fP est parfois plus rapide
426
+ que \fBgzip \-9\fP tout en compressant bien mieux. Les réglages plus élevés ont
427
+ souvent une rapidité comparable à celle de \fBbzip2\fP(1) avec un taux de
428
+ compression comparable ou meilleur, même si les résultats dépendent beaucoup
429
+ du genre de données compressées.
430
+ .TP
431
+ \fB\-4\fP ... \fB\-6\fP
432
+ Good to very good compression while keeping decompressor memory usage
433
+ reasonable even for old systems. \fB\-6\fP is the default, which is usually a
434
+ good choice for distributing files that need to be decompressible even on
435
+ systems with only 16\ MiB RAM. (\fB\-5e\fP or \fB\-6e\fP may be worth considering
436
+ too. See \fB\-\-extreme\fP.)
437
+ .TP
438
+ \fB\-7 ... \-9\fP
439
+ C'est comme \fB\-6\fP mais avec des besoins en mémoire plus élevés pour la
440
+ compression et la décompression. Ce n'est utile que lorsque les fichiers
441
+ sont plus gros que 8\ Mio, 16\ Mio et 32\ Mio respectivement.
442
+ .RE
443
+ .IP ""
444
+ Sur le même matériel, la vitesse de décompression est sensiblement un nombre
445
+ constant d'octets de données compressées par seconde. En d'autres termes,
446
+ meilleure est la compression, plus rapide sera en général la
447
+ décompression. Cela signifie aussi que la quantité de sortie non compressée
448
+ produite par seconde peut varier beaucoup.
449
+ .IP ""
450
+ Le tableau suivant résume les caractéristiques des préréglages :
451
+ .RS
452
+ .RS
453
+ .PP
454
+ .TS
455
+ tab(;);
456
+ c c c c c
457
+ n n n n n.
458
+ Préréglage;DictSize;CompCPU;CompMem;DecMem
459
+ \-0;256 KiB;0;3 MiB;1 MiB
460
+ \-1;1 MiB;1;9 MiB;2 MiB
461
+ \-2;2 MiB;2;17 MiB;3 MiB
462
+ \-3;4 MiB;3;32 MiB;5 MiB
463
+ \-4;4 MiB;4;48 MiB;5 MiB
464
+ \-5;8 MiB;5;94 MiB;9 MiB
465
+ \-6;8 MiB;6;94 MiB;9 MiB
466
+ \-7;16 MiB;6;186 MiB;17 MiB
467
+ \-8;32 MiB;6;370 MiB;33 MiB
468
+ \-9;64 MiB;6;674 MiB;65 MiB
469
+ .TE
470
+ .RE
471
+ .RE
472
+ .IP ""
473
+ Descriptions des colonnes :
474
+ .RS
475
+ .IP \(bu 3
476
+ DictSize est la taille du dictionnaire de LZMA2. Utiliser un dictionnaire
477
+ plus gros que la taille du fichier non compressé est un gaspillage de
478
+ mémoire. C'est pourquoi il est bon d'éviter d'utiliser les préréglages de
479
+ \fB\-7\fP à \fB\-9\fP lorsqu'il n'y en a pas vraiment besoin. A \fB\-6\fP et plus bas,
480
+ la quantité de mémoire gaspillée est généralement assez basse pour ne pas
481
+ être un problème.
482
+ .IP \(bu 3
483
+ CompCPU est une représentation des préréglages de LZMA2 qui affectent la
484
+ vitesse de compression. La taille du dictionnaire aussi affecte la vitesse,
485
+ alors comme CompCPU est le même pour les niveaux de \fB\-6\fP à \fB\-9\fP, les plus
486
+ haut niveaux tendent à être un peu moins rapides. Pour être encore moins
487
+ rapide et du coup obtenir peut être une meilleure compression, consultez
488
+ \fB\-\-extreme\fP.
489
+ .IP \(bu 3
490
+ CompMem contains the compressor memory requirements in the single\-threaded
491
+ mode. It may vary slightly between \fBxz\fP versions.
492
+ .IP \(bu 3
493
+ DecMem contient les besoins en mémoire du décompresseur. Ce sont les
494
+ réglages de la compression qui déterminent les besoins en mémoire de la
495
+ décompression. L'exacte utilisation de la mémoire est légèrement supérieure
496
+ à la taille du dictionnaire LZMA2, mais les valeurs dans la table ont été
497
+ arrondies au prochain Mio supérieur.
498
+ .RE
499
+ .IP ""
500
+ Memory requirements of the multi\-threaded mode are significantly higher than
501
+ that of the single\-threaded mode. With the default value of
502
+ \fB\-\-block\-size\fP, each thread needs 3*3*DictSize plus CompMem or DecMem. For
503
+ example, four threads with preset \fB\-6\fP needs 660\(en670\ MiB of memory.
504
+ .TP
505
+ \fB\-e\fP, \fB\-\-extreme\fP
506
+ Utilisez un variant plus lent que les préréglages (\fB\-0\fP à \fB\-9\fP) pour
507
+ espérer avoir un taux de compression légèrement meilleur, mais en cas de
508
+ malchance cela peut être pire. L'utilisation mémoire du décompresseur n'est
509
+ pas affectée, mais l'utilisation mémoire du compresseur augmente un peu aux
510
+ niveaux de préréglages de \fB\-0\fP à \fB\-3\fP.
511
+ .IP ""
512
+ Depuis qu'il y a deux préréglages avec des tailles de dictionnaire de 4\ Mio
513
+ et 8 \Mio, les préréglages \fB\-3e\fP et \fB\-5e\fP utilisent des réglages
514
+ légèrement plus rapides que \fB\-4e\fP et \fB\-6e\fP, respectivement. De cette
515
+ manière, il n'y a pas deux préréglages identiques.
516
+ .RS
517
+ .RS
518
+ .PP
519
+ .TS
520
+ tab(;);
521
+ c c c c c
522
+ n n n n n.
523
+ Préréglage;DictSize;CompCPU;CompMem;DecMem
524
+ \-0e;256 KiB;8;4 MiB;1 MiB
525
+ \-1e;1 MiB;8;13 MiB;2 MiB
526
+ \-2e;2 MiB;8;25 MiB;3 MiB
527
+ \-3e;4 MiB;7;48 MiB;5 MiB
528
+ \-4e;4 MiB;8;48 MiB;5 MiB
529
+ \-5e;8 MiB;7;94 MiB;9 MiB
530
+ \-6e;8 MiB;8;94 MiB;9 MiB
531
+ \-7e;16 MiB;8;186 MiB;17 MiB
532
+ \-8e;32 MiB;8;370 MiB;33 MiB
533
+ \-9e;64 MiB;8;674 MiB;65 MiB
534
+ .TE
535
+ .RE
536
+ .RE
537
+ .IP ""
538
+ Par exemple, il y a un total de quatre préréglages qui utilisent un
539
+ dictionnaire de 8 Mio et qui sont dans l'ordre du plus rapide au plus lent :
540
+ \fB\-5\fP, \fB\-6\fP, \fB\-5e\fP et \fB\-6e\fP.
541
+ .TP
542
+ \fB\-\-fast\fP
543
+ .PD 0
544
+ .TP
545
+ \fB\-\-best\fP
546
+ .PD
547
+ Il y a néanmoins des alias trompeurs pour \fB\-0\fP et \fB\-9\fP,
548
+ respectivement. Ils ne sont fournis que pour des besoins de
549
+ rétro\-compatibilité avec les utilitaires LZMA. Evitez d'utiliser ces
550
+ options.
551
+ .TP
552
+ \fB\-\-block\-size=\fP\fItaille\fP
553
+ Lors de la compression dans le format \fB.xz\fP, les données de l'entrée sont
554
+ réparties en blocs de \fItaille\fP octets. Les blocs sont compressés
555
+ indépendamment les un des autres, ce qui aide avec le mode multithread
556
+ (multi\-threading) et rend possible la décompression à accès aléatoire
557
+ limité. Cette option est typiquement utilisée pour outrepasser la taille de
558
+ bloc en mode multithread, mais cette option peut aussi être utilisée en mode
559
+ single\-thread.
560
+ .IP ""
561
+ In multi\-threaded mode about three times \fIsize\fP bytes will be allocated in
562
+ each thread for buffering input and output. The default \fIsize\fP is three
563
+ times the LZMA2 dictionary size or 1 MiB, whichever is more. Typically a
564
+ good value is 2\(en4 times the size of the LZMA2 dictionary or at least 1
565
+ MiB. Using \fIsize\fP less than the LZMA2 dictionary size is waste of RAM
566
+ because then the LZMA2 dictionary buffer will never get fully used. In
567
+ multi\-threaded mode, the sizes of the blocks are stored in the block
568
+ headers. This size information is required for multi\-threaded
569
+ decompression.
570
+ .IP ""
571
+ In single\-threaded mode no block splitting is done by default. Setting this
572
+ option doesn't affect memory usage. No size information is stored in block
573
+ headers, thus files created in single\-threaded mode won't be identical to
574
+ files created in multi\-threaded mode. The lack of size information also
575
+ means that \fBxz\fP won't be able decompress the files in multi\-threaded mode.
576
+ .TP
577
+ \fB\-\-block\-list=\fP\fIitems\fP
578
+ When compressing to the \fB.xz\fP format, start a new block with an optional
579
+ custom filter chain after the given intervals of uncompressed data.
580
+ .IP ""
581
+ The \fIitems\fP are a comma\-separated list. Each item consists of an optional
582
+ filter chain number between 0 and 9 followed by a colon (\fB:\fP) and a
583
+ required size of uncompressed data. Omitting an item (two or more
584
+ consecutive commas) is a shorthand to use the size and filters of the
585
+ previous item.
586
+ .IP ""
587
+ If the input file is bigger than the sum of the sizes in \fIitems\fP, the last
588
+ item is repeated until the end of the file. A special value of \fB0\fP may be
589
+ used as the last size to indicate that the rest of the file should be
590
+ encoded as a single block.
591
+ .IP ""
592
+ An alternative filter chain for each block can be specified in combination
593
+ with the \fB\-\-filters1=\fP\fIfilters\fP \&...\& \fB\-\-filters9=\fP\fIfilters\fP options.
594
+ These options define filter chains with an identifier between 1\(en9.
595
+ Filter chain 0 can be used to refer to the default filter chain, which is
596
+ the same as not specifying a filter chain. The filter chain identifier can
597
+ be used before the uncompressed size, followed by a colon (\fB:\fP). For
598
+ example, if one specifies \fB\-\-block\-list=1:2MiB,3:2MiB,2:4MiB,,2MiB,0:4MiB\fP
599
+ then blocks will be created using:
600
+ .RS
601
+ .IP \(bu 3
602
+ The filter chain specified by \fB\-\-filters1\fP and 2 MiB input
603
+ .IP \(bu 3
604
+ The filter chain specified by \fB\-\-filters3\fP and 2 MiB input
605
+ .IP \(bu 3
606
+ The filter chain specified by \fB\-\-filters2\fP and 4 MiB input
607
+ .IP \(bu 3
608
+ The filter chain specified by \fB\-\-filters2\fP and 4 MiB input
609
+ .IP \(bu 3
610
+ The default filter chain and 2 MiB input
611
+ .IP \(bu 3
612
+ The default filter chain and 4 MiB input for every block until end of input.
613
+ .RE
614
+ .IP ""
615
+ If one specifies a size that exceeds the encoder's block size (either the
616
+ default value in threaded mode or the value specified with
617
+ \fB\-\-block\-size=\fP\fIsize\fP), the encoder will create additional blocks while
618
+ keeping the boundaries specified in \fIitems\fP. For example, if one specifies
619
+ \fB\-\-block\-size=10MiB\fP \fB\-\-block\-list=5MiB,10MiB,8MiB,12MiB,24MiB\fP and the
620
+ input file is 80 MiB, one will get 11 blocks: 5, 10, 8, 10, 2, 10, 10, 4,
621
+ 10, 10, and 1 MiB.
622
+ .IP ""
623
+ En mode multi\-threadé les tailles de blocs sont stockées dans les en\-têtes
624
+ du bloc. Cela ne se fait pas en mode mono\-threadé, la sortie encodée ne sera
625
+ donc pas identique à celle faite en mode multi\-threadé.
626
+ .TP
627
+ \fB\-\-flush\-timeout=\fP\fItemps_d'attente\fP
628
+ Lors de la compression, si plus que \fItemps_d'attente\fP millisecondes (un
629
+ entier positif) se sont écoulées depuis le précédent vidage et que lire plus
630
+ de données bloquerait, toutes les données d'entrée en attente sont vidées de
631
+ l'encodeur et mises à disposition dans le flux de sortie. Cela peut être
632
+ utile si \fBxz\fP est utilisé pour compresser les données qui sont diffusées
633
+ sur un réseau. Des petites valeurs de \fItemps_d'attente\fP rendent les données
634
+ disponibles à l'extrémité réceptrice avec un léger retard, mais les grandes
635
+ valeurs de \fItemps_d'attente\fP donnent un meilleur taux de compression.
636
+ .IP ""
637
+ Cette option est désactivée par défaut. Si cette option est indiquée plus
638
+ d'une fois, la dernière prend effet. La valeur spéciale de
639
+ \fItemps_d'attente\fP de \fB0\fP peut être utilisée pour explicitement désactiver
640
+ cette option.
641
+ .IP ""
642
+ Cette option n'est pas disponible sur les systèmes qui ne sont pas POSIX.
643
+ .IP ""
644
+ .\" FIXME
645
+ \fBCette option est encore expérimentale.\fP Actuellement, \fBxz\fP ne convient
646
+ pas pour décompresser le flux en temps réel en raison de la façon dont \fBxz\fP
647
+ effectue la mise en mémoire tampon.
648
+ .TP
649
+ \fB\-\-memlimit\-compress=\fP\fIlimite\fP
650
+ Indiquer une limite d'utilisation de la mémoire pour la compression. Si
651
+ cette option est indiquée plusieurs fois, c'est la dernière qui est prise en
652
+ compte.
653
+ .IP ""
654
+ If the compression settings exceed the \fIlimit\fP, \fBxz\fP will attempt to
655
+ adjust the settings downwards so that the limit is no longer exceeded and
656
+ display a notice that automatic adjustment was done. The adjustments are
657
+ done in this order: reducing the number of threads, switching to
658
+ single\-threaded mode if even one thread in multi\-threaded mode exceeds the
659
+ \fIlimit\fP, and finally reducing the LZMA2 dictionary size.
660
+ .IP ""
661
+ When compressing with \fB\-\-format=raw\fP or if \fB\-\-no\-adjust\fP has been
662
+ specified, only the number of threads may be reduced since it can be done
663
+ without affecting the compressed output.
664
+ .IP ""
665
+ If the \fIlimit\fP cannot be met even with the adjustments described above, an
666
+ error is displayed and \fBxz\fP will exit with exit status 1.
667
+ .IP ""
668
+ La \fIlimite\fP peut être indiquée de plusieurs façons :
669
+ .RS
670
+ .IP \(bu 3
671
+ La \fIlimite\fP peut être une valeur absolue en octets. Utiliser un suffixe
672
+ d'entier comme \fBMiB\fP peut être utile. Exemple :
673
+ \fB\-\-memlimit\-compress=80MiB\fP
674
+ .IP \(bu 3
675
+ La \fIlimite\fP peut être indiquée sous forme d'un pourcentage de la mémoire
676
+ physique totale (RAM). Cela peut être particulièrement utile quand la
677
+ variable d'environnement \fBXZ_DEFAULTS\fP est indiquée dans un script
678
+ d'initialisation de l'interpréteur partagé entre différents ordinateurs. De
679
+ cette façon la limite est automatiquement plus grande sur les systèmes avec
680
+ plus de mémoire. Exemple : \fB\-\-memlimit=70%\fP
681
+ .IP \(bu 3
682
+ The \fIlimit\fP can be reset back to its default value by setting it to \fB0\fP.
683
+ This is currently equivalent to setting the \fIlimit\fP to \fBmax\fP (no memory
684
+ usage limit).
685
+ .RE
686
+ .IP ""
687
+ For 32\-bit \fBxz\fP there is a special case: if the \fIlimit\fP would be over
688
+ \fB4020\ MiB\fP, the \fIlimit\fP is set to \fB4020\ MiB\fP. On MIPS32 \fB2000\ MiB\fP
689
+ is used instead. (The values \fB0\fP and \fBmax\fP aren't affected by this. A
690
+ similar feature doesn't exist for decompression.) This can be helpful when
691
+ a 32\-bit executable has access to 4\ GiB address space (2 GiB on MIPS32)
692
+ while hopefully doing no harm in other situations.
693
+ .IP ""
694
+ Voir aussi la section \fButilisation de la mémoire\fP.
695
+ .TP
696
+ \fB\-\-memlimit\-decompress=\fP\fIlimite\fP
697
+ Régler une limite d'utilisation de la mémoire pour la décompression. Cela a
698
+ un effet sur le mode \fB\-\-list\fP. Si l'opération n'est pas possible sans
699
+ dépasser la \fIlimite\fP, \fBxz\fP affichera une erreur et la décompression
700
+ échouera. Voir \fB\-\-memlimit\-compress=\fP\fIlimite\fP pour les manières possibles
701
+ d'indiquer la \fIlimite\fP.
702
+ .TP
703
+ \fB\-\-memlimit\-mt\-decompress=\fP\fIlimit\fP
704
+ Set a memory usage limit for multi\-threaded decompression. This can only
705
+ affect the number of threads; this will never make \fBxz\fP refuse to
706
+ decompress a file. If \fIlimit\fP is too low to allow any multi\-threading, the
707
+ \fIlimit\fP is ignored and \fBxz\fP will continue in single\-threaded mode. Note
708
+ that if also \fB\-\-memlimit\-decompress\fP is used, it will always apply to both
709
+ single\-threaded and multi\-threaded modes, and so the effective \fIlimit\fP for
710
+ multi\-threading will never be higher than the limit set with
711
+ \fB\-\-memlimit\-decompress\fP.
712
+ .IP ""
713
+ In contrast to the other memory usage limit options,
714
+ \fB\-\-memlimit\-mt\-decompress=\fP\fIlimit\fP has a system\-specific default
715
+ \fIlimit\fP. \fBxz \-\-info\-memory\fP can be used to see the current value.
716
+ .IP ""
717
+ This option and its default value exist because without any limit the
718
+ threaded decompressor could end up allocating an insane amount of memory
719
+ with some input files. If the default \fIlimit\fP is too low on your system,
720
+ feel free to increase the \fIlimit\fP but never set it to a value larger than
721
+ the amount of usable RAM as with appropriate input files \fBxz\fP will attempt
722
+ to use that amount of memory even with a low number of threads. Running out
723
+ of memory or swapping will not improve decompression performance.
724
+ .IP ""
725
+ See \fB\-\-memlimit\-compress=\fP\fIlimit\fP for possible ways to specify the
726
+ \fIlimit\fP. Setting \fIlimit\fP to \fB0\fP resets the \fIlimit\fP to the default
727
+ system\-specific value.
728
+ .TP
729
+ \fB\-M\fP \fIlimite\fP, \fB\-\-memlimit=\fP\fIlimite\fP, \fB\-\-memory=\fP\fIlimite\fP
730
+ This is equivalent to specifying \fB\-\-memlimit\-compress=\fP\fIlimit\fP
731
+ \fB\-\-memlimit\-decompress=\fP\fIlimit\fP \fB\-\-memlimit\-mt\-decompress=\fP\fIlimit\fP.
732
+ .TP
733
+ \fB\-\-no\-adjust\fP
734
+ Display an error and exit if the memory usage limit cannot be met without
735
+ adjusting settings that affect the compressed output. That is, this
736
+ prevents \fBxz\fP from switching the encoder from multi\-threaded mode to
737
+ single\-threaded mode and from reducing the LZMA2 dictionary size. Even when
738
+ this option is used the number of threads may be reduced to meet the memory
739
+ usage limit as that won't affect the compressed output.
740
+ .IP ""
741
+ Automatic adjusting is always disabled when creating raw streams
742
+ (\fB\-\-format=raw\fP).
743
+ .TP
744
+ \fB\-T\fP \fIthreads\fP, \fB\-\-threads=\fP\fIthreads\fP
745
+ Specify the number of worker threads to use. Setting \fIthreads\fP to a
746
+ special value \fB0\fP makes \fBxz\fP use up to as many threads as the processor(s)
747
+ on the system support. The actual number of threads can be fewer than
748
+ \fIthreads\fP if the input file is not big enough for threading with the given
749
+ settings or if using more threads would exceed the memory usage limit.
750
+ .IP ""
751
+ The single\-threaded and multi\-threaded compressors produce different
752
+ output. Single\-threaded compressor will give the smallest file size but
753
+ only the output from the multi\-threaded compressor can be decompressed using
754
+ multiple threads. Setting \fIthreads\fP to \fB1\fP will use the single\-threaded
755
+ mode. Setting \fIthreads\fP to any other value, including \fB0\fP, will use the
756
+ multi\-threaded compressor even if the system supports only one hardware
757
+ thread. (\fBxz\fP 5.2.x used single\-threaded mode in this situation.)
758
+ .IP ""
759
+ To use multi\-threaded mode with only one thread, set \fIthreads\fP to \fB+1\fP.
760
+ The \fB+\fP prefix has no effect with values other than \fB1\fP. A memory usage
761
+ limit can still make \fBxz\fP switch to single\-threaded mode unless
762
+ \fB\-\-no\-adjust\fP is used. Support for the \fB+\fP prefix was added in \fBxz\fP
763
+ 5.4.0.
764
+ .IP ""
765
+ If an automatic number of threads has been requested and no memory usage
766
+ limit has been specified, then a system\-specific default soft limit will be
767
+ used to possibly limit the number of threads. It is a soft limit in sense
768
+ that it is ignored if the number of threads becomes one, thus a soft limit
769
+ will never stop \fBxz\fP from compressing or decompressing. This default soft
770
+ limit will not make \fBxz\fP switch from multi\-threaded mode to single\-threaded
771
+ mode. The active limits can be seen with \fBxz \-\-info\-memory\fP.
772
+ .IP ""
773
+ Actuellement, la seule méthode de gestion avec des threads consiste à
774
+ séparer l'entrée en blocs et de les compresser indépendamment les uns des
775
+ autres. La taille par défaut des blocs dépend du niveau de compression et
776
+ peut\-être remplacée avec l'option \fB\-\-block\-size=\fP\fItaille\fP.
777
+ .IP ""
778
+ Threaded decompression only works on files that contain multiple blocks with
779
+ size information in block headers. All large enough files compressed in
780
+ multi\-threaded mode meet this condition, but files compressed in
781
+ single\-threaded mode don't even if \fB\-\-block\-size=\fP\fIsize\fP has been used.
782
+ .IP ""
783
+ The default value for \fIthreads\fP is \fB0\fP. In \fBxz\fP 5.4.x and older the
784
+ default is \fB1\fP.
785
+ .
786
+ .SS "Chaînes de filtres de compresseur personnalisées"
787
+ A custom filter chain allows specifying the compression settings in detail
788
+ instead of relying on the settings associated to the presets. When a custom
789
+ filter chain is specified, preset options (\fB\-0\fP \&...\& \fB\-9\fP and
790
+ \fB\-\-extreme\fP) earlier on the command line are forgotten. If a preset
791
+ option is specified after one or more custom filter chain options, the new
792
+ preset takes effect and the custom filter chain options specified earlier
793
+ are forgotten.
794
+ .PP
795
+ Une chaîne de filtre est comparable à une redirection (pipe) sur la ligne de
796
+ commande. Lors de la compression, les entrées non compressées vont au
797
+ premier filtre, dont la sortie va au prochain filtre (s'il y en a). La
798
+ sortie du dernier filtre est écrite sur le fichier compressé. Le nombre
799
+ maximal de filtres dans la chaîne est quatre, mais habituellement, un chaîne
800
+ de filtre n'a qu'un ou deux filtres.
801
+ .PP
802
+ Beaucoup de filtres ont des limitations sur l'endroit où ils peuvent se
803
+ placer dans la chaîne de filtre : quelques filtres ne peuvent fonctionner
804
+ qu'en tant que dernier filtre dans la chaîne, quelques uns en tant que non
805
+ dernier filtre, et d'autres à n'importe quelle position dans la
806
+ chaîne. Suivant le filtre, cette limitation est soit inhérente au profil du
807
+ filtre, soit existe pour des raisons de sécurité.
808
+ .PP
809
+ A custom filter chain can be specified in two different ways. The options
810
+ \fB\-\-filters=\fP\fIfilters\fP and \fB\-\-filters1=\fP\fIfilters\fP \&...\&
811
+ \fB\-\-filters9=\fP\fIfilters\fP allow specifying an entire filter chain in one
812
+ option using the liblzma filter string syntax. Alternatively, a filter
813
+ chain can be specified by using one or more individual filter options in the
814
+ order they are wanted in the filter chain. That is, the order of the
815
+ individual filter options is significant! When decoding raw streams
816
+ (\fB\-\-format=raw\fP), the filter chain must be specified in the same order as
817
+ it was specified when compressing. Any individual filter or preset options
818
+ specified before the full chain option (\fB\-\-filters=\fP\fIfilters\fP) will be
819
+ forgotten. Individual filters specified after the full chain option will
820
+ reset the filter chain.
821
+ .PP
822
+ Both the full and individual filter options take filter\-specific \fIoptions\fP
823
+ as a comma\-separated list. Extra commas in \fIoptions\fP are ignored. Every
824
+ option has a default value, so specify those you want to change.
825
+ .PP
826
+ Pour voir l'entièreté de la chaîne de filtres et ses \fIoptions\fP, utilisez
827
+ \fBxz \-vv\fP (ce qui est comme utiliser \fB\-\-verbose\fP deux fois). Cela
828
+ fonctionne aussi pour voir les options de chaîne de filtres utilisées par
829
+ les préréglages.
830
+ .TP
831
+ \fB\-\-filters=\fP\fIfilters\fP
832
+ Specify the full filter chain or a preset in a single option. Each filter
833
+ can be separated by spaces or two dashes (\fB\-\-\fP). \fIfilters\fP may need to be
834
+ quoted on the shell command line so it is parsed as a single option. To
835
+ denote \fIoptions\fP, use \fB:\fP or \fB=\fP. A preset can be prefixed with a \fB\-\fP
836
+ and followed with zero or more flags. The only supported flag is \fBe\fP to
837
+ apply the same options as \fB\-\-extreme\fP.
838
+ .TP
839
+ \fB\-\-filters1\fP=\fIfilters\fP ... \fB\-\-filters9\fP=\fIfilters\fP
840
+ Specify up to nine additional filter chains that can be used with
841
+ \fB\-\-block\-list\fP.
842
+ .IP ""
843
+ For example, when compressing an archive with executable files followed by
844
+ text files, the executable part could use a filter chain with a BCJ filter
845
+ and the text part only the LZMA2 filter.
846
+ .TP
847
+ \fB\-\-filters\-help\fP
848
+ Display a help message describing how to specify presets and custom filter
849
+ chains in the \fB\-\-filters\fP and \fB\-\-filters1=\fP\fIfilters\fP \&...\&
850
+ \fB\-\-filters9=\fP\fIfilters\fP options, and exit successfully.
851
+ .TP
852
+ \fB\-\-lzma1\fP[\fB=\fP\fIoptions\fP]
853
+ .PD 0
854
+ .TP
855
+ \fB\-\-lzma2\fP[\fB=\fP\fIoptions\fP]
856
+ .PD
857
+ Ajouter le filtre LZMA1 ou LZMA2 à la chaîne de filtres. Ces filtres ne
858
+ peuvent être utilisés que comme dernier filtre dans la chaîne.
859
+ .IP ""
860
+ LZMA1 est un filtre historique, qui n'est pris en charge presque uniquement
861
+ à cause de l'ancien format de fichier \fB.lzma\fP, qui ne prend en charge que
862
+ LZMA1. LZMA2 est une version mise à jour de LZMA1 pour régler certains
863
+ problèmes pratiques de LZMA1. Le format \fBxz\fP utilise LZMA2 et ne prend pas
864
+ du tout en charge LZMA1. Les taux et vitesses de compression de LZMA1 et
865
+ LZMA2 sont pratiquement identiques.
866
+ .IP ""
867
+ LZMA1 et LZMA2 partagent le même ensemble d'\fIoptions\fP :
868
+ .RS
869
+ .TP
870
+ \fBpreset=\fP\fIpréréglage\fP
871
+ Reset all LZMA1 or LZMA2 \fIoptions\fP to \fIpreset\fP. \fIPreset\fP consist of an
872
+ integer, which may be followed by single\-letter preset modifiers. The
873
+ integer can be from \fB0\fP to \fB9\fP, matching the command line options \fB\-0\fP
874
+ \&...\& \fB\-9\fP. The only supported modifier is currently \fBe\fP, which matches
875
+ \fB\-\-extreme\fP. If no \fBpreset\fP is specified, the default values of LZMA1 or
876
+ LZMA2 \fIoptions\fP are taken from the preset \fB6\fP.
877
+ .TP
878
+ \fBdict=\fP\fItaille\fP
879
+ La \fItaille\fP du dictionnaire (historique du tampon) indique combien d'octets
880
+ des données récement décompressées sont gardés en mémoire. L'algorithme
881
+ essaie de trouver les séquences d'octets répétées (identiques) dans les
882
+ données décompressées et les remplace par les données actuellement dans le
883
+ dictionnaire. Plus gros est le dictionnaire, plus grande est la chance de
884
+ trouver une correspondance. Ainsi, l'augmentation de la \fItaille\fP du
885
+ dictionnaire augmente habituellement le taux de compression, mais un
886
+ dictionnaire plus gros que le fichier non compressé est un gachis de
887
+ mémoire.
888
+ .IP ""
889
+ Généralement la \fItaille\fP du dictionnaire est entre 64\ Kio et 64\ Mio. Le
890
+ minimum étant 4\ Kio. La \fItaille\fP maximale pour la compression est
891
+ habituellement 1,5\ Gio (1536\ Mio). Le décompresseur prend en charge les
892
+ dictionnaires jusqu'à un octet de moins que 4\ Gio, ce qui est le maximum
893
+ pour les formats de flux LZMA1 et LZMA2.
894
+ .IP ""
895
+ La \fItaille\fP du dictionnaire et le chercheur de correspondance (match
896
+ finder) (\fImf\fP) déterminent ensemble l'utilisation de la mémoire de
897
+ l'encodeur LZMA1 ou LZMA2. La même (ou une plus grosse) \fItaille\fP de
898
+ dictionnaire est requise pour décompresser que ce qui a été utilisé pour la
899
+ compression, ainsi l'utilisation de la mémoire du décodeur est déterminée
900
+ par la taille du dictionnaire utilisée lors de la compression. Les en\-têtes
901
+ de \fB.xz\fP stockent la \fItaille\fP de dictionnaire sous la forme 2^\fIn\fP ou
902
+ 2^\fIn\fP + 2^(\fIn\fP\-1), de sorte que ces \fItailles\fP sont quelque peu préférées
903
+ pour la compression. Les autres \fItailles\fP seront arrondies à la hausse
904
+ lorsque stockées dans les en\-têtes de \fB.xz\fP.
905
+ .TP
906
+ \fBlc=\fP\fIlc\fP
907
+ Spécifiez le nombre d'octets de contexte littéraux. Le minimum est \fB0\fP et
908
+ le maximum est \fB4\fP. La valeur par défaut est \fB3\fP. En plus, la somme de
909
+ \fIlc\fP et \fIlp\fP ne doit pas excéder \fB4\fP.
910
+ .IP ""
911
+ Tous les octets qui ne peuvent pas être codés comme des correspondances sont
912
+ codés comme des littéraux. C'est à dire que les littéraux sont simplement
913
+ des octets 8 bits encodés un à la fois.
914
+ .IP ""
915
+ The literal coding makes an assumption that the highest \fIlc\fP bits of the
916
+ previous uncompressed byte correlate with the next byte. For example, in
917
+ typical English text, an upper\-case letter is often followed by a lower\-case
918
+ letter, and a lower\-case letter is usually followed by another lower\-case
919
+ letter. In the US\-ASCII character set, the highest three bits are 010 for
920
+ upper\-case letters and 011 for lower\-case letters. When \fIlc\fP is at least
921
+ 3, the literal coding can take advantage of this property in the
922
+ uncompressed data.
923
+ .IP ""
924
+ The default value (3) is usually good. If you want maximum compression,
925
+ test \fBlc=4\fP. Sometimes it helps a little, and sometimes it makes
926
+ compression worse. If it makes it worse, test \fBlc=2\fP too.
927
+ .TP
928
+ \fBlp=\fP\fIlp\fP
929
+ Indiquer le nombre de bits de position littérale. Le minimum est \fB0\fP et le
930
+ maximum \fB4\fP; par défaut c'est \fB0\fP.
931
+ .IP ""
932
+ \fILp\fP affecte le type d'alignement dans les données décompressées qui est
933
+ présumé lors de l'encodage des littéraux. Voir \fIpb\fP ci dessous pour plus
934
+ d'information sur l'alignement.
935
+ .TP
936
+ \fBpb=\fP\fIpb\fP
937
+ Indiquer le nombre de bits de position. Le minimum est \fB0\fP et le maximum
938
+ \fB4\fP; par défaut \fB2\fP.
939
+ .IP ""
940
+ \fIPb\fP affecte quel genre d'alignement est présumé en général dans les
941
+ données non compressées. Par défaut c'est un alignement de quatre octets
942
+ (2^\fIpb\fP=2^2=4), ce qui est généralement un bon choix lorsqu'il n'y a pas de
943
+ meilleure estimation.
944
+ .IP ""
945
+ When the alignment is known, setting \fIpb\fP accordingly may reduce the file
946
+ size a little. For example, with text files having one\-byte alignment
947
+ (US\-ASCII, ISO\-8859\-*, UTF\-8), setting \fBpb=0\fP can improve compression
948
+ slightly. For UTF\-16 text, \fBpb=1\fP is a good choice. If the alignment is
949
+ an odd number like 3 bytes, \fBpb=0\fP might be the best choice.
950
+ .IP ""
951
+ Même si l'alignement présumé peut être ajusté avec \fIpb\fP et \fIlp\fP, LZMA1 et
952
+ LZMA2 favorisent toujours légèrement l'alignement sur 16 octets. Il peut
953
+ être utile d'en tenir compte lors de la conception de formats de fichiers
954
+ susceptibles d'être souvent compressés avec LZMA1 ou LZMA2.
955
+ .TP
956
+ \fBmf=\fP\fImf\fP
957
+ Match finder has a major effect on encoder speed, memory usage, and
958
+ compression ratio. Usually Hash Chain match finders are faster than Binary
959
+ Tree match finders. The default depends on the \fIpreset\fP: 0 uses \fBhc3\fP,
960
+ 1\(en3 use \fBhc4\fP, and the rest use \fBbt4\fP.
961
+ .IP ""
962
+ Les chercheurs de correspondance suivants sont pris en charge. Les formules
963
+ d'utilisation de la mémoire ci\-dessous sont des approximations grossières
964
+ qui sont les plus proches de la réalité lorsque \fIdict\fP est une puissance de
965
+ deux.
966
+ .RS
967
+ .TP
968
+ \fBhc3\fP
969
+ Chaîne de hachage avec hachage de 2 et 3 octets
970
+ .br
971
+ Valeur minimale pour \fInice\fP : \fB3\fP
972
+ .br
973
+ Utilisation de la mémoire :
974
+ .br
975
+ \fIdict\fP * 7.5 (if \fIdict\fP <= 16 Mio);
976
+ .br
977
+ \fIdict\fP * 5.5 + 64 MiB (si \fIdict\fP > 16 Mio)
978
+ .TP
979
+ \fBhc4\fP
980
+ Chaîne de hachage avec hachage de 2, 3 et 4 octets
981
+ .br
982
+ Valeur minimale pour \fInice\fP : \fB4\fP
983
+ .br
984
+ Utilisation de la mémoire :
985
+ .br
986
+ \fIdict\fP * 7.5 (si \fIdict\fP <= 32 Mio);
987
+ .br
988
+ \fIdict\fP * 6.5 (si \fIdict\fP > 32 Mio)
989
+ .TP
990
+ \fBbt2\fP
991
+ Arbre binaire avec hachage de 2 octets
992
+ .br
993
+ Valeur minimale pour \fInice\fP : \fB2\fP
994
+ .br
995
+ Utilisation de la mémoire : \fIdict\fP * 9.5
996
+ .TP
997
+ \fBbt3\fP
998
+ Arbre binaire avec hachage de 2 et 3 octets
999
+ .br
1000
+ Valeur minimale pour \fInice\fP : \fB3\fP
1001
+ .br
1002
+ Utilisation de la mémoire :
1003
+ .br
1004
+ \fIdict\fP * 11.5 (si \fIdict\fP <= 16 Mio);
1005
+ .br
1006
+ \fIdict\fP * 9.5 + 64 MiB (si \fIdict\fP > 16 Mio)
1007
+ .TP
1008
+ \fBbt4\fP
1009
+ Arbre binaire avec hachage 2, 3 et 4 octets
1010
+ .br
1011
+ Valeur minimale pour \fInice\fP : \fB4\fP
1012
+ .br
1013
+ Utilisation de la mémoire :
1014
+ .br
1015
+ \fIdict\fP * 11.5 (si \fIdict\fP <= 32 Mio);
1016
+ .br
1017
+ \fIdict\fP * 10.5 (si \fIdict\fP > 32 Mio)
1018
+ .RE
1019
+ .TP
1020
+ \fBmode=\fP\fImode\fP
1021
+ Compression \fImode\fP specifies the method to analyze the data produced by the
1022
+ match finder. Supported \fImodes\fP are \fBfast\fP and \fBnormal\fP. The default is
1023
+ \fBfast\fP for \fIpresets\fP 0\(en3 and \fBnormal\fP for \fIpresets\fP 4\(en9.
1024
+ .IP ""
1025
+ Habituellement, \fBfast\fP est utilisé avec les chercheurs de correspondance de
1026
+ chaîne de hachage et \fBnormal\fP avec les chercheurs de correspondance d'arbre
1027
+ binaire. C'est aussi ce que font les \fIpréréglages\fP.
1028
+ .TP
1029
+ \fBnice=\fP\fInice\fP
1030
+ Spécifier ce qui est considéré comme une bonne longueur pour une
1031
+ correspondance. Une fois que la correspondance d'au moins \fInice\fP octets est
1032
+ trouvée, l'algorithme arrête de chercher de meilleures correspondances
1033
+ possibles.
1034
+ .IP ""
1035
+ \fINice\fP can be 2\(en273 bytes. Higher values tend to give better
1036
+ compression ratio at the expense of speed. The default depends on the
1037
+ \fIpreset\fP.
1038
+ .TP
1039
+ \fBdepth=\fP\fIprofondeur\fP
1040
+ Spécifier la profondeur de recherche maximale dans l'outil de recherche de
1041
+ correspondances. La valeur par défaut est \fB0\fP, ce qui fait que le
1042
+ compresseur détermine une \fIprofondeur\fP raisonnable en fonction de \fImf\fP et
1043
+ \fInice\fP.
1044
+ .IP ""
1045
+ Reasonable \fIdepth\fP for Hash Chains is 4\(en100 and 16\(en1000 for Binary
1046
+ Trees. Using very high values for \fIdepth\fP can make the encoder extremely
1047
+ slow with some files. Avoid setting the \fIdepth\fP over 1000 unless you are
1048
+ prepared to interrupt the compression in case it is taking far too long.
1049
+ .RE
1050
+ .IP ""
1051
+ Lors du décodage des flux bruts (\fB\-\-format=raw\fP), LZMA2 nécessite seulement
1052
+ la \fItaille\fP du dictionnaire. LZMA1 nécessite aussi \fIlc\fP, \fIlp\fP et \fIpb\fP.
1053
+ .TP
1054
+ \fB\-\-x86\fP[\fB=\fP\fIoptions\fP]
1055
+ .PD 0
1056
+ .TP
1057
+ \fB\-\-arm\fP[\fB=\fP\fIoptions\fP]
1058
+ .TP
1059
+ \fB\-\-armthumb\fP[\fB=\fP\fIoptions\fP]
1060
+ .TP
1061
+ \fB\-\-arm64\fP[\fB=\fP\fIoptions\fP]
1062
+ .TP
1063
+ \fB\-\-powerpc\fP[\fB=\fP\fIoptions\fP]
1064
+ .TP
1065
+ \fB\-\-ia64\fP[\fB=\fP\fIoptions\fP]
1066
+ .TP
1067
+ \fB\-\-sparc\fP[\fB=\fP\fIoptions\fP]
1068
+ .TP
1069
+ \fB\-\-riscv\fP[\fB=\fP\fIoptions\fP]
1070
+ .PD
1071
+ Ajouter un filtre branch/call/jump (BCJ) à la chaîne de filtres. Ces filtres
1072
+ ne peuvent être utilisés que s'ils ne sont pas le dernier filtre de la
1073
+ chaîne de filtrage.
1074
+ .IP ""
1075
+ A BCJ filter converts relative addresses in the machine code to their
1076
+ absolute counterparts. This doesn't change the size of the data but it
1077
+ increases redundancy, which can help LZMA2 to produce 0\(en15\ % smaller
1078
+ \&\fB.xz\fP file. The BCJ filters are always reversible, so using a BCJ filter
1079
+ for wrong type of data doesn't cause any data loss, although it may make the
1080
+ compression ratio slightly worse. The BCJ filters are very fast and use an
1081
+ insignificant amount of memory.
1082
+ .IP ""
1083
+ Ces filtres BCJ présentent des problèmes connus liés au taux de
1084
+ compression :
1085
+ .RS
1086
+ .IP \(bu 3
1087
+ Some types of files containing executable code (for example, object files,
1088
+ static libraries, and Linux kernel modules) have the addresses in the
1089
+ instructions filled with filler values. These BCJ filters will still do the
1090
+ address conversion, which will make the compression worse with these files.
1091
+ .IP \(bu 3
1092
+ If a BCJ filter is applied on an archive, it is possible that it makes the
1093
+ compression ratio worse than not using a BCJ filter. For example, if there
1094
+ are similar or even identical executables then filtering will likely make
1095
+ the files less similar and thus compression is worse. The contents of
1096
+ non\-executable files in the same archive can matter too. In practice one
1097
+ has to try with and without a BCJ filter to see which is better in each
1098
+ situation.
1099
+ .RE
1100
+ .IP ""
1101
+ Different instruction sets have different alignment: the executable file
1102
+ must be aligned to a multiple of this value in the input data to make the
1103
+ filter work.
1104
+ .RS
1105
+ .RS
1106
+ .PP
1107
+ .TS
1108
+ tab(;);
1109
+ l n l
1110
+ l n l.
1111
+ Filtre;Alignement;Notes
1112
+ x86;1;32 bits ou 64 bits x86
1113
+ ARM;4;
1114
+ ARM\-Thumb;2;
1115
+ ARM64;4;4096\-byte alignment is best
1116
+ PowerPC;4;Grand boutiste seulement
1117
+ IA\-64;16;Itanium
1118
+ SPARC;4;
1119
+ RISC\-V;2;
1120
+ .TE
1121
+ .RE
1122
+ .RE
1123
+ .IP ""
1124
+ Since the BCJ\-filtered data is usually compressed with LZMA2, the
1125
+ compression ratio may be improved slightly if the LZMA2 options are set to
1126
+ match the alignment of the selected BCJ filter. Examples:
1127
+ .RS
1128
+ .IP \(bu 3
1129
+ IA\-64 filter has 16\-byte alignment so \fBpb=4,lp=4,lc=0\fP is good with LZMA2
1130
+ (2^4=16).
1131
+ .IP \(bu 3
1132
+ RISC\-V code has 2\-byte or 4\-byte alignment depending on whether the file
1133
+ contains 16\-bit compressed instructions (the C extension). When 16\-bit
1134
+ instructions are used, \fBpb=2,lp=1,lc=3\fP or \fBpb=1,lp=1,lc=3\fP is good. When
1135
+ 16\-bit instructions aren't present, \fBpb=2,lp=2,lc=2\fP is the best.
1136
+ \fBreadelf \-h\fP can be used to check if "RVC" appears on the "Flags" line.
1137
+ .IP \(bu 3
1138
+ ARM64 is always 4\-byte aligned so \fBpb=2,lp=2,lc=2\fP is the best.
1139
+ .IP \(bu 3
1140
+ The x86 filter is an exception. It's usually good to stick to LZMA2's
1141
+ defaults (\fBpb=2,lp=0,lc=3\fP) when compressing x86 executables.
1142
+ .RE
1143
+ .IP ""
1144
+ Tous les filtres BCJ prennent en charge les mêmes \fIoptions\fP :
1145
+ .RS
1146
+ .TP
1147
+ \fBstart=\fP\fIdécalage\fP
1148
+ Spécifier le \fIdécalage\fP de départ qui est utilisé lors de la conversion
1149
+ entre les adresses relatives et absolues. Le \fIdécalage\fP doit être un
1150
+ multiple de l'alignement du filtre (voir la table ci\-dessus). Sa valeur par
1151
+ défaut est zéro. En pratique, cette dernière convient ; indiquer un
1152
+ \fIdécalage\fP personnalisé est la plupart du temps inutile.
1153
+ .RE
1154
+ .TP
1155
+ \fB\-\-delta\fP[\fB=\fP\fIoptions\fP]
1156
+ Ajouter le filtre Delta à la chaîne de filtres. Le filtre Delta ne peut être
1157
+ utilisé que s'il n'est pas le dernier filtre dans la chaîne.
1158
+ .IP ""
1159
+ Currently only simple byte\-wise delta calculation is supported. It can be
1160
+ useful when compressing, for example, uncompressed bitmap images or
1161
+ uncompressed PCM audio. However, special purpose algorithms may give
1162
+ significantly better results than Delta + LZMA2. This is true especially
1163
+ with audio, which compresses faster and better, for example, with
1164
+ \fBflac\fP(1).
1165
+ .IP ""
1166
+ \fIoptions\fP prises en charge :
1167
+ .RS
1168
+ .TP
1169
+ \fBdist=\fP\fIdistance\fP
1170
+ Specify the \fIdistance\fP of the delta calculation in bytes. \fIdistance\fP must
1171
+ be 1\(en256. The default is 1.
1172
+ .IP ""
1173
+ Par exemple, avec \fBdist=2\fP et une entrée huit octets A1 B1 A2 B3 A3 B5 A4
1174
+ B7, la sortie sera A1 B1 01 02 01 02 01 02.
1175
+ .RE
1176
+ .
1177
+ .SS "Autres options"
1178
+ .TP
1179
+ \fB\-q\fP, \fB\-\-quiet\fP
1180
+ Supprimer les avertissements et les notifications. Indiquer cela deux fois
1181
+ supprimera aussi les erreurs. Cette option n'a aucun effet sur le statut de
1182
+ sortie. Cela dit, même si un avertissement était supprimé, le statut de
1183
+ sortie indiquant un avertissement sera encore utilisé.
1184
+ .TP
1185
+ \fB\-v\fP, \fB\-\-verbose\fP
1186
+ Être bavard. Si l'erreur standard est connectée à un terminal, \fBxz\fP
1187
+ affichera une barre de progression. Indiquer \fB\-\-verbose\fP deux fois donnera
1188
+ une sortie encore plus bavarde.
1189
+ .IP ""
1190
+ La barre de progression montre l'information suivante :
1191
+ .RS
1192
+ .IP \(bu 3
1193
+ Le pourcentage de complétion est montré si la taille du fichier en entrée
1194
+ est connue. Néanmoins, le pourcentage ne peut pas être montré en cas de
1195
+ redirection.
1196
+ .IP \(bu 3
1197
+ Quantité de données compressées produites (compression) ou consommées
1198
+ (décompression).
1199
+ .IP \(bu 3
1200
+ Quantité de données non compressées consommées (compression) ou produites
1201
+ (décompression).
1202
+ .IP \(bu 3
1203
+ Le taux de compression, calculé en divisant la quantité de données
1204
+ compréssées déjà traitées par la quantité de données décompressées déjà
1205
+ traitées.
1206
+ .IP \(bu 3
1207
+ Vitesse de compression ou de décompression. Elle correspond à la quantité de
1208
+ données non compressées consommées (compression) ou produites
1209
+ (décompression) par seconde. Elle apparait quelques secondes après le début
1210
+ du traitement du fichier par \fBxz\fP.
1211
+ .IP \(bu 3
1212
+ Temps écoulé dans le format M:SS ou H:MM:SS.
1213
+ .IP \(bu 3
1214
+ Estimated remaining time is shown only when the size of the input file is
1215
+ known and a couple of seconds have already passed since \fBxz\fP started
1216
+ processing the file. The time is shown in a less precise format which never
1217
+ has any colons, for example, 2 min 30 s.
1218
+ .RE
1219
+ .IP ""
1220
+ When standard error is not a terminal, \fB\-\-verbose\fP will make \fBxz\fP print
1221
+ the filename, compressed size, uncompressed size, compression ratio, and
1222
+ possibly also the speed and elapsed time on a single line to standard error
1223
+ after compressing or decompressing the file. The speed and elapsed time are
1224
+ included only when the operation took at least a few seconds. If the
1225
+ operation didn't finish, for example, due to user interruption, also the
1226
+ completion percentage is printed if the size of the input file is known.
1227
+ .TP
1228
+ \fB\-Q\fP, \fB\-\-no\-warn\fP
1229
+ Ne pas mettre l'état de sortie à \fB2\fP même si une condition méritant un
1230
+ avertissement a été détectée. Cette option n'affecte pas le niveau de
1231
+ verbosité, néanmoins, les deux options \fB\-\-quiet\fP et \fB\-\-no\-warn\fP doivent
1232
+ être utilisées pour ne pas afficher d'avertissements, ni altérer le statut
1233
+ de sortie.
1234
+ .TP
1235
+ \fB\-\-robot\fP
1236
+ Afficher les messages dans un format analysable par une machine. Ceci est
1237
+ destiné à faciliter l'écriture des frontaux qui voudraient utiliser \fBxz\fP
1238
+ plutôt que liblzma, ce qui pourrait être le cas pour différents scripts. La
1239
+ sortie avec cette option activée est destinée à rester stable sur les
1240
+ différentes versions de \fBxz\fP. Consulter le paragraphe \fBROBOT MODE\fP pour
1241
+ les détails.
1242
+ .TP
1243
+ \fB\-\-info\-memory\fP
1244
+ Display, in human\-readable format, how much physical memory (RAM) and how
1245
+ many processor threads \fBxz\fP thinks the system has and the memory usage
1246
+ limits for compression and decompression, and exit successfully.
1247
+ .TP
1248
+ \fB\-h\fP, \fB\-\-help\fP
1249
+ Afficher un message d'aide décrivant les options les plus couramment
1250
+ utilisées et quitter.
1251
+ .TP
1252
+ \fB\-H\fP, \fB\-\-long\-help\fP
1253
+ Afficher un message d'aide décrivant toutes les options de \fBxz\fP et quitter.
1254
+ .TP
1255
+ \fB\-V\fP, \fB\-\-version\fP
1256
+ Afficher le numéro de version de \fBxz\fP et de liblzma dans un format lisible
1257
+ par un humain. Pour obtenir une sortie analysable par la machine, spécifiez
1258
+ \fB\-\-robot\fP avant \fB\-\-version\fP.
1259
+ .
1260
+ .SH "MODE ROBOT"
1261
+ The robot mode is activated with the \fB\-\-robot\fP option. It makes the output
1262
+ of \fBxz\fP easier to parse by other programs. Currently \fB\-\-robot\fP is
1263
+ supported only together with \fB\-\-list\fP, \fB\-\-filters\-help\fP, \fB\-\-info\-memory\fP,
1264
+ and \fB\-\-version\fP. It will be supported for compression and decompression in
1265
+ the future.
1266
+ .
1267
+ .SS "Mode liste"
1268
+ \fBxz \-\-robot \-\-list\fP utilise une sortie séparée par des tabulations. La
1269
+ première colonne de toutes les lignes possède une chaîne qui indique le type
1270
+ d'information trouvée sur cette ligne :
1271
+ .TP
1272
+ \fBname\fP
1273
+ C'est toujours la première ligne au début de la liste d'un fichier. La
1274
+ seconde colonne de la ligne est le nom de fichier.
1275
+ .TP
1276
+ \fBfile\fP
1277
+ Cette ligne contient l'information globale sur le fichier \fB.xz\fP. Cette
1278
+ ligne est toujours écrite après la ligne \fBname\fP.
1279
+ .TP
1280
+ \fBstream\fP
1281
+ Ce type de ligne n'est utilisée que lorsque \fB \-\-verbose\fP a été indiquée. Il
1282
+ y a autant de lignes \fBstream\fP qu'il y a de flux dans le fichier \fB.xz\fP.
1283
+ .TP
1284
+ \fBblock\fP
1285
+ Ce type de ligne n'est utilisé seulement lorsque \fB\-\-verbose\fP a été
1286
+ indiquée. Il y a autant de lignes \fBblock\fP qu'il y a de blocs dans le
1287
+ fichier \fB.xz\fP. Les lignes \fBblock\fP sont affichées après toutes les lignes
1288
+ \fBstream\fP ; les différents types de lignes ne sont pas imbriqués.
1289
+ .TP
1290
+ \fBsummary\fP
1291
+ Ce type de ligne n'est utilisé que lorsque \fB\-\-verbose\fP a été indiqué deux
1292
+ fois. Cette ligne est affichée après toutes les lignes \fBblock\fP. Comme la
1293
+ ligne \fBfile\fP, la ligne \fBsummary\fP contient l'information globale sur le
1294
+ fichier \fB.xz\fP.
1295
+ .TP
1296
+ \fBtotals\fP
1297
+ Cette ligne est toujours la toute dernière ligne de la sortie. Elle affiche
1298
+ les comptes et les tailles totaux.
1299
+ .PP
1300
+ Les colonnes des lignes \fBfile\fP :
1301
+ .PD 0
1302
+ .RS
1303
+ .IP 2. 4
1304
+ Nombre de flux dans le fichier
1305
+ .IP 3. 4
1306
+ Nombre total de blocs dans le ou les flux.
1307
+ .IP 4. 4
1308
+ Taille compressée du fichier
1309
+ .IP 5. 4
1310
+ Taille décompressée du fichier
1311
+ .IP 6. 4
1312
+ Compression ratio, for example, \fB0.123\fP. If ratio is over 9.999, three
1313
+ dashes (\fB\-\-\-\fP) are displayed instead of the ratio.
1314
+ .IP 7. 4
1315
+ Liste de noms de contrôles d'intégrité séparés par des virgules. Les chaînes
1316
+ suivantes sont utilisées pour les types de vérification connus : \fBNone\fP,
1317
+ \fBCRC32\fP, \fBCRC64\fP et \fBSHA256\fP. Pour le types de vérification inconnus,
1318
+ \fBUnknown\-\fP\fIN\fP est utilisé, où \fIN\fP est un identifiant de vérification sous
1319
+ la forme d'un nombre décimal (un ou deux chiffres).
1320
+ .IP 8. 4
1321
+ Taille totale du remplissage du flux dans le fichier
1322
+ .RE
1323
+ .PD
1324
+ .PP
1325
+ Les colonnes des lignes \fBstream\fP :
1326
+ .PD 0
1327
+ .RS
1328
+ .IP 2. 4
1329
+ Numéro de flux (le premier flux a le numéro 1)
1330
+ .IP 3. 4
1331
+ Nombre de blocs dans le flux
1332
+ .IP 4. 4
1333
+ Décalage de départ compressé
1334
+ .IP 5. 4
1335
+ Décalage de départ décompressé
1336
+ .IP 6. 4
1337
+ Taille compressée (ne comprend pas le remplissage du flux)
1338
+ .IP 7. 4
1339
+ Taille décompressée
1340
+ .IP 8. 4
1341
+ Taux de compression
1342
+ .IP 9. 4
1343
+ Nom de la vérification d'intégrité
1344
+ .IP 10. 4
1345
+ Taille du remplissage de flux
1346
+ .RE
1347
+ .PD
1348
+ .PP
1349
+ Les colonnes des lignes \fBblock\fP :
1350
+ .PD 0
1351
+ .RS
1352
+ .IP 2. 4
1353
+ Numéro du flux qui contient ce bloc
1354
+ .IP 3. 4
1355
+ Numéro du bloc relatif au commencement du flux (le premier bloc a pour
1356
+ numéro 1)
1357
+ .IP 4. 4
1358
+ Numéro du bloc relatif au début du fichier
1359
+ .IP 5. 4
1360
+ Décalage de départ compressé relatif au début du fichier
1361
+ .IP 6. 4
1362
+ Décalage de départ décompressé relatif au début du fichier
1363
+ .IP 7. 4
1364
+ Taille compressée totale du bloc (en\-têtes inclus)
1365
+ .IP 8. 4
1366
+ Taille décompressée
1367
+ .IP 9. 4
1368
+ Taux de compression
1369
+ .IP 10. 4
1370
+ Nom de la vérification d'intégrité
1371
+ .RE
1372
+ .PD
1373
+ .PP
1374
+ Si \fB\-\-verbose\fP a été indiqué deux fois, les colonnes additionnelles sont
1375
+ inclues sur les lignes \fBblock\fP. Elles ne sont pas affichées avec un seul
1376
+ \fB\-\-verbose\fP, car l'obtention de ces informations nécessite de nombreuses
1377
+ recherches et peut donc être lente :
1378
+ .PD 0
1379
+ .RS
1380
+ .IP 11. 4
1381
+ Valeur de la vérification d'intégrité en hexadécimal
1382
+ .IP 12. 4
1383
+ Taille d'en\-tête de bloc
1384
+ .IP 13. 4
1385
+ Drapeaux du bloc : \fBc\fP indique que la taille compressée est présente, et
1386
+ \fBu\fP indique que la taille décompréssée est présente. Si le drapeau n'est
1387
+ pas indiqué, un tiret (\fB\-\fP) est affiché à la place pour que la longueur de
1388
+ la chaîne reste fixe. De nouveaux drapeaux pourraient être ajoutés à la fin
1389
+ de la chaîne dans le futur.
1390
+ .IP 14. 4
1391
+ Taille des données effectivement compressées dans le bloc (en excluant
1392
+ l'en\-tête de bloc, le remplissage de bloc et les champs de vérification).
1393
+ .IP 15. 4
1394
+ Quantité de mémoire (en octets) nécessaire pour décompresser ce bloc avec
1395
+ cette version de \fBxz\fP.
1396
+ .IP 16. 4
1397
+ Chaîne de filtrage. Remarquez que la plupart des options utilisées au moment
1398
+ de la compression ne peuvent pas être connues, car seules les options
1399
+ nécessaires pour la décompression sont stockées dans les en\-têtes \fB.xz\fP.
1400
+ .RE
1401
+ .PD
1402
+ .PP
1403
+ Les colonnes des lignes \fBsummary\fP :
1404
+ .PD 0
1405
+ .RS
1406
+ .IP 2. 4
1407
+ Quantité de mémoire (en octets) nécessaire pour décompresser ce fichier avec
1408
+ cette version de \fBxz\fP.
1409
+ .IP 3. 4
1410
+ \fByes\fP ou \fBno\fP indique si tous les en\-têtes de bloc stockent à la fois la
1411
+ taille compressée et la taille décompressée.
1412
+ .PP
1413
+ \fIDepuis\fP \fBxz\fP \fI5.1.2alpha:\fP
1414
+ .IP 4. 4
1415
+ Version minimale de \fBxz\fP nécessaire pour décompresser le fichier.
1416
+ .RE
1417
+ .PD
1418
+ .PP
1419
+ Les colonnes de la ligne \fBtotals\fP :
1420
+ .PD 0
1421
+ .RS
1422
+ .IP 2. 4
1423
+ Nombre de flux
1424
+ .IP 3. 4
1425
+ Nombre de blocs
1426
+ .IP 4. 4
1427
+ Taille compressée
1428
+ .IP 5. 4
1429
+ Taille décompressée
1430
+ .IP 6. 4
1431
+ Taux de compression moyen
1432
+ .IP 7. 4
1433
+ Liste séparée par des virgules des noms de vérification d'intégrité qui
1434
+ étaient présents dans les fichiers
1435
+ .IP 8. 4
1436
+ Taille de remplissage de flux
1437
+ .IP 9. 4
1438
+ Nombre de fichiers. Permet de garder l'ordre des colonnes précédentes comme
1439
+ sur les lignes \fBfile\fP.
1440
+ .PD
1441
+ .RE
1442
+ .PP
1443
+ Si \fB\-\-verbose\fP a été indiqué deux fois, des colonnes supplémentaires sont
1444
+ incluses sur la ligne \fBtotals\fP :
1445
+ .PD 0
1446
+ .RS
1447
+ .IP 10. 4
1448
+ Quantité maximale de mémoire (en octets) nécessaire pour décompresser les
1449
+ fichiers avec cette version de \fBxz\fP.
1450
+ .IP 11. 4
1451
+ \fByes\fP ou \fBno\fP indique si tous les en\-têtes de bloc stockent à la fois la
1452
+ taille compressée et la taille décompressée.
1453
+ .PP
1454
+ \fIDepuis\fP \fBxz\fP \fI5.1.2alpha:\fP
1455
+ .IP 12. 4
1456
+ Version minimale de \fBxz\fP nécessaire pour décompresser le fichier.
1457
+ .RE
1458
+ .PD
1459
+ .PP
1460
+ Les versions futures pourront ajouter de nouveaux types de lignes et de
1461
+ nouvelles colonnes pourront être ajoutées aux types de lignes existants,
1462
+ mais les colonnes existantes ne seront pas modifiées.
1463
+ .
1464
+ .SS "Filters help"
1465
+ \fBxz \-\-robot \-\-filters\-help\fP prints the supported filters in the following
1466
+ format:
1467
+ .PP
1468
+ \fIfilter\fP\fB:\fP\fIoption\fP\fB=<\fP\fIvalue\fP\fB>,\fP\fIoption\fP\fB=<\fP\fIvalue\fP\fB>\fP...
1469
+ .TP
1470
+ \fIfilter\fP
1471
+ Name of the filter
1472
+ .TP
1473
+ \fIoption\fP
1474
+ Name of a filter specific option
1475
+ .TP
1476
+ \fIvalue\fP
1477
+ Numeric \fIvalue\fP ranges appear as \fB<\fP\fImin\fP\fB\-\fP\fImax\fP\fB>\fP. String
1478
+ \fIvalue\fP choices are shown within \fB< >\fP and separated by a \fB|\fP
1479
+ character.
1480
+ .PP
1481
+ Each filter is printed on its own line.
1482
+ .
1483
+ .SS "Information de limite de mémoire"
1484
+ \fBxz \-\-robot \-\-info\-memory\fP prints a single line with multiple tab\-separated
1485
+ columns:
1486
+ .IP 1. 4
1487
+ Total amount of physical memory (RAM) in bytes.
1488
+ .IP 2. 4
1489
+ Memory usage limit for compression in bytes (\fB\-\-memlimit\-compress\fP). A
1490
+ special value of \fB0\fP indicates the default setting which for
1491
+ single\-threaded mode is the same as no limit.
1492
+ .IP 3. 4
1493
+ Memory usage limit for decompression in bytes (\fB\-\-memlimit\-decompress\fP). A
1494
+ special value of \fB0\fP indicates the default setting which for
1495
+ single\-threaded mode is the same as no limit.
1496
+ .IP 4. 4
1497
+ Since \fBxz\fP 5.3.4alpha: Memory usage for multi\-threaded decompression in
1498
+ bytes (\fB\-\-memlimit\-mt\-decompress\fP). This is never zero because a
1499
+ system\-specific default value shown in the column 5 is used if no limit has
1500
+ been specified explicitly. This is also never greater than the value in the
1501
+ column 3 even if a larger value has been specified with
1502
+ \fB\-\-memlimit\-mt\-decompress\fP.
1503
+ .IP 5. 4
1504
+ Since \fBxz\fP 5.3.4alpha: A system\-specific default memory usage limit that is
1505
+ used to limit the number of threads when compressing with an automatic
1506
+ number of threads (\fB\-\-threads=0\fP) and no memory usage limit has been
1507
+ specified (\fB\-\-memlimit\-compress\fP). This is also used as the default value
1508
+ for \fB\-\-memlimit\-mt\-decompress\fP.
1509
+ .IP 6. 4
1510
+ Since \fBxz\fP 5.3.4alpha: Number of available processor threads.
1511
+ .PP
1512
+ Dans le futur, la sortie de \fBxz \-\-robot \-\-info\-memory\fP pourrait avoir plus
1513
+ de colonnes, mais jamais plus qu'une ligne unique.
1514
+ .
1515
+ .SS Version
1516
+ \fBxz \-\-robot \-\-version\fP prints the version number of \fBxz\fP and liblzma in
1517
+ the following format:
1518
+ .PP
1519
+ \fBXZ_VERSION=\fP\fIXYYYZZZS\fP
1520
+ .br
1521
+ \fBLIBLZMA_VERSION=\fP\fIXYYYZZZS\fP
1522
+ .TP
1523
+ \fIX\fP
1524
+ Version majeure.
1525
+ .TP
1526
+ \fIYYY\fP
1527
+ Version mineure. Les numéros pairs sont stables. Les numéros impairs sont
1528
+ des versions alpha ou beta.
1529
+ .TP
1530
+ \fIZZZ\fP
1531
+ Niveau de correctif pour les options stables ou juste un compteur pour les
1532
+ options de développement.
1533
+ .TP
1534
+ \fIS\fP
1535
+ Stabilité. 0 est alpha, 1 est bêta et 2 est stable. \fIS\fP devrait toujours
1536
+ être 2 quand \fIYYY\fP est pair.
1537
+ .PP
1538
+ \fIXYYYZZZS\fP sont identiques sur les deux lignes si \fBxz\fP et liblzma sont
1539
+ issus de la même version d'utilitaires XZ.
1540
+ .PP
1541
+ Exemples : 4.999.9beta est \fB49990091\fP et 5.0.0 est \fB50000002\fP.
1542
+ .
1543
+ .SH "STATUT DE SORTIE"
1544
+ .TP
1545
+ \fB0\fP
1546
+ Tout est bon.
1547
+ .TP
1548
+ \fB1\fP
1549
+ Une erreur est survenue.
1550
+ .TP
1551
+ \fB2\fP
1552
+ Quelquechose méritant un avertissement s'est produit, mais aucune erreur
1553
+ véritable n'est survenue.
1554
+ .PP
1555
+ Les notifications (pas les avertissements ou les erreurs) affichées sur
1556
+ l'erreur standard n'affectent pas le statut de sortie.
1557
+ .
1558
+ .SH ENVIRONNEMENT
1559
+ \fBxz\fP analyse les listes d'options séparées par des espaces à partir des
1560
+ variables d'environnement \fBXZ_DEFAULTS\fP et \fBXZ_OPT\fP, dans cet ordre, avant
1561
+ d'analyser les options de la ligne de commandes. Remarquez que seules les
1562
+ options sont analysées depuis l'environnement des variables ; toutes les
1563
+ non\-options sont ignorées silencieusement. L'analyse est faite avec
1564
+ \fBgetopt_long\fP(3) qui est aussi utilisé pour les arguments de la ligne de
1565
+ commandes.
1566
+ .TP
1567
+ \fBXZ_DEFAULTS\fP
1568
+ Options par défaut propres à l'utilisateur ou pour tout le système. Elles
1569
+ sont le plus souvent définies dans un script d'initialisation de
1570
+ l'interpréteur pour activer le limiteur d'utilisation de la mémoire de \fBxz\fP
1571
+ par défaut. A part pour les scripts d'initialisation de l'interpréteur ou
1572
+ des cas similaires, les sripts ne doivent jamais définir ou désactiver
1573
+ \fBXZ_DEFAULTS\fP.
1574
+ .TP
1575
+ \fBXZ_OPT\fP
1576
+ This is for passing options to \fBxz\fP when it is not possible to set the
1577
+ options directly on the \fBxz\fP command line. This is the case when \fBxz\fP is
1578
+ run by a script or tool, for example, GNU \fBtar\fP(1):
1579
+ .RS
1580
+ .RS
1581
+ .PP
1582
+ .nf
1583
+ \f(CRXZ_OPT=\-2v tar caf foo.tar.xz foo\fR
1584
+ .fi
1585
+ .RE
1586
+ .RE
1587
+ .IP ""
1588
+ Scripts may use \fBXZ_OPT\fP, for example, to set script\-specific default
1589
+ compression options. It is still recommended to allow users to override
1590
+ \fBXZ_OPT\fP if that is reasonable. For example, in \fBsh\fP(1) scripts one may
1591
+ use something like this:
1592
+ .RS
1593
+ .RS
1594
+ .PP
1595
+ .nf
1596
+ \f(CRXZ_OPT=${XZ_OPT\-"\-7e"}
1597
+ export XZ_OPT\fR
1598
+ .fi
1599
+ .RE
1600
+ .RE
1601
+ .
1602
+ .SH "Compatibilité des utilitaires LZMA"
1603
+ La syntaxe de la ligne de commande de \fBxz\fP est quasimment un sur\-ensemble
1604
+ de \fBlzma\fP, \fBunlzma\fP et \fBlzcat\fP comme ils sont trouvés dans les
1605
+ utilitaires LZMA 4.32.x . Dans la pluspart des cas, il est possible de
1606
+ remplacer les outils LZMA par les outils XZ sans casser les scripts
1607
+ existants. Il existe cependant certaines incompatibilités qui peuvent
1608
+ parfois poser des problèmes.
1609
+ .
1610
+ .SS "Niveaux de préréglage de la compression"
1611
+ La numérotation des préréglages de niveau de compression est différente
1612
+ entre les outils \fBxz\fP et LZMA. La différence la plus importante est la
1613
+ manière dont les tailles de dictionnaire sont affectées aux différents
1614
+ préréglages. La taille de dictionnaire est à peu près égale à celle
1615
+ d'utilisation de la mémoire de la décompression.
1616
+ .RS
1617
+ .PP
1618
+ .TS
1619
+ tab(;);
1620
+ c c c
1621
+ c n n.
1622
+ Niveau;xz;Utilitaires LZMA
1623
+ \-0;256 KiB;N/A
1624
+ \-1;1 MiB;64 KiB
1625
+ \-2;2 MiB;1 MiB
1626
+ \-3;4 MiB;512 KiB
1627
+ \-4;4 MiB;1 MiB
1628
+ \-5;8 MiB;2 MiB
1629
+ \-6;8 MiB;4 MiB
1630
+ \-7;16 MiB;8 MiB
1631
+ \-8;32 MiB;16 MiB
1632
+ \-9;64 MiB;32 MiB
1633
+ .TE
1634
+ .RE
1635
+ .PP
1636
+ Les différences de tailles des dictionnaires affectent aussi l'utilisation
1637
+ de la mémoire du compresseur, mais il y a quelques autres différences entre
1638
+ les outils LZMA et les outils XZ, qui rendent la différence encore plus
1639
+ grande :
1640
+ .RS
1641
+ .PP
1642
+ .TS
1643
+ tab(;);
1644
+ c c c
1645
+ c n n.
1646
+ Niveau;xz;Utilitaires LZMA 4.32.x
1647
+ \-0;3 MiB;N/A
1648
+ \-1;9 MiB;2 MiB
1649
+ \-2;17 MiB;12 MiB
1650
+ \-3;32 MiB;12 MiB
1651
+ \-4;48 MiB;16 MiB
1652
+ \-5;94 MiB;26 MiB
1653
+ \-6;94 MiB;45 MiB
1654
+ \-7;186 MiB;83 MiB
1655
+ \-8;370 MiB;159 MiB
1656
+ \-9;674 MiB;311 MiB
1657
+ .TE
1658
+ .RE
1659
+ .PP
1660
+ Le niveau de préréglage par défaut dans les outils LZMA est \fB\-7\fP alors que
1661
+ pour les outils XZ c'est \fB\-6\fP, les deux utilisent ainsi un dictionnaire de
1662
+ 8 Mio par défaut.
1663
+ .
1664
+ .SS "Fichiers .lzma en flux ou non"
1665
+ The uncompressed size of the file can be stored in the \fB.lzma\fP header.
1666
+ LZMA Utils does that when compressing regular files. The alternative is to
1667
+ mark that uncompressed size is unknown and use end\-of\-payload marker to
1668
+ indicate where the decompressor should stop. LZMA Utils uses this method
1669
+ when uncompressed size isn't known, which is the case, for example, in
1670
+ pipes.
1671
+ .PP
1672
+ \fBxz\fP prend en charge la décompression des fichiers \fB.lzma\fP avec ou sans
1673
+ marqueur de fin de charge utile, mais tous les fichiers \fB.lzma\fP créés par
1674
+ \fBxz\fP utiliseront un marqueur de fin de charge utile et ont la taille non
1675
+ compréssée marquée comme inconnue dans l'en\-tête \fB.lzma\fP. Cela peut être un
1676
+ problème dans quelques situations inhabituelles. Par exemple, un
1677
+ décompresseur \fB.lzma\fP dans un périphérique embarqué pourrait ne fonctionner
1678
+ qu'avec des fichiers dont la taille non comprimée est connue. Si vous vous
1679
+ heurtez à ce problème, vous devez utiliser les utilitaires LZMA ou LZMA SDK
1680
+ pour créer des fichiers \fB.lzma\fP avec une taille non compressée connue.
1681
+ .
1682
+ .SS "Fichiers .lzma non pris en charge"
1683
+ Le format \fB.lzma\fP autorise des valeurs \fIlc\fP jusqu'à 8, et des valeurs
1684
+ \fIlp\fP jusqu'à 4. Les outils LZMA peuvent décompresser des fichiers avec tous
1685
+ les \fIlc\fP et \fIlp\fP, mais créez toujours les fichiers avec \fBlc=3\fP et
1686
+ \fBlp=0\fP. Créer des fichiers avec d'autres valeurs \fIlc\fP et \fIlp\fP est
1687
+ possible avec \fBxz\fP et avec LZMA SDK.
1688
+ .PP
1689
+ L'implémentation du filtre LZMA1 dans liblzma nécessite que la somme de
1690
+ \fIlc\fP et \fIlp\fP soit inférieure ou égale à 4. Ainsi, les fichiers \fB.lzma\fP
1691
+ qui excèdent cette limitation ne peuvent pas être décompressés avec \fBxz\fP.
1692
+ .PP
1693
+ Les outils LZMA créent seulement des fichiers \fB.lzma\fP qui ont une taille de
1694
+ dictionnaire de 2^\fIn\fP (une puissance de 2) mais acceptent les fichiers avec
1695
+ toutes les tailles de dictionnaire. Libzlma n'accepte que les fichiers
1696
+ \&\fB.lzma\fP qui ont une taille dictionnaire de 2^\fIn\fP ou
1697
+ 2^\fIn\fP + 2^(\fIn\fP\-1). Cela afin de diminuer les faux positifs lors de la
1698
+ détection des fichiers \fB.lzma\fP.
1699
+ .PP
1700
+ Ces limitations ne devraient pas poser problème en pratique, car
1701
+ pratiquement tous les fichiers \fB.lzma\fP ont été compressés avec des réglages
1702
+ que liblzma accepte.
1703
+ .
1704
+ .SS "Déchets excédentaires"
1705
+ Lors de la décompession, l'utilitaire LZMA ignore silencieusement tout ce
1706
+ qui est après le premier flux \fB.lzma\fP. Dans la majorité des situations,
1707
+ c'est un bogue. Cela veut dire aussi que les outils LZMA ne gèrent pas la
1708
+ décompression de fichiers \fB.lzma\fP concaténés.
1709
+ .PP
1710
+ S'il reste des données après le premier flux \fB.lzma\fP, \fBxz\fP considère que
1711
+ le fichier est corrompu sauf si \fB\-\-single\-stream\fP a été utilisé. Cela peut
1712
+ casser des scripts obscurs qui ont supposé que les déchets de fin de ligne
1713
+ sont ignorés.
1714
+ .
1715
+ .SH NOTES
1716
+ .
1717
+ .SS "La sortie compressée peut varier"
1718
+ La sortie compressée exacte produite par les même fichiers non compressés en
1719
+ entrée peut varier en fonction des différentes versions de l'utilitaire XZ,
1720
+ même si les options de compression sont identiques. En effet, il est
1721
+ possible d'améliorer l'encodeur (compression plus rapide ou meilleure) sans
1722
+ affecter le format du fichier. La sortie peut même varier entre différentes
1723
+ compilations de la même version d'utilitaire XZ, si des options de
1724
+ construction différentes sont utilisées.
1725
+ .PP
1726
+ Cela signifie qu'une fois que \fB\-\-rsyncable\fP a été implémenté, les fichiers
1727
+ résultants ne seront pas nécessairement synchronisables avec rsync à moins
1728
+ que les nouveaux et anciens fichiers n'aient été compressés avec la même
1729
+ version de xz. Ce problème peut être résolu si une partie de
1730
+ l'implémentation est gelée pour garantir la stabilité de la sortie rsyncable
1731
+ à travers les versions de xz.
1732
+ .
1733
+ .SS "Décompresseurs .xz embarqués"
1734
+ Les implémentations de décompresseur embarqué comme XZ Embedded ne gèrent
1735
+ pas nécessairement les fichiers créés avec d'autres types de \fIvérification\fP
1736
+ d'intégrité que \fBnone\fP et \fBCRC32\fP. Comme la valeur par défaut est
1737
+ \fB\-\-check=crc64\fP, vous devez utiliser \fB\-\-check=none\fP ou \fB\-\-check=crc32\fP
1738
+ lors de la création de fichiers pour les systèmes embarqués.
1739
+ .PP
1740
+ En dehors des systèmes embarqués, tous les décompresseurs de format \fB.xz\fP
1741
+ gèrent tous les types de \fIvérification\fP ou sont au moins capables de
1742
+ décompresser le fichier sans effectuer la vérification d'intégrité si ce
1743
+ type de \fIvérification\fP particulière n'est pas pris en charge.
1744
+ .PP
1745
+ XZ Embedded prend en charge les filtres BCJ, mais seulement avec le décalage
1746
+ de départ par défaut.
1747
+ .
1748
+ .SH EXEMPLES
1749
+ .
1750
+ .SS Bases
1751
+ Compresser le fichier \fItoto\fP en \fItoto.xz\fP en utilisant le niveau de
1752
+ compression par défaut (\fB\-6\fP) et supprimer \fItoto\fP si la compression
1753
+ réussit :
1754
+ .RS
1755
+ .PP
1756
+ .nf
1757
+ \f(CRxz foo\fR
1758
+ .fi
1759
+ .RE
1760
+ .PP
1761
+ Décompresser \fIbidule.xz\fP en \fIbidule\fP et ne pas supprimer \fIbidule.xz\fP même
1762
+ si la compression réussit :
1763
+ .RS
1764
+ .PP
1765
+ .nf
1766
+ \f(CRxz \-dk bar.xz\fR
1767
+ .fi
1768
+ .RE
1769
+ .PP
1770
+ Create \fIbaz.tar.xz\fP with the preset \fB\-4e\fP (\fB\-4 \-\-extreme\fP), which is
1771
+ slower than the default \fB\-6\fP, but needs less memory for compression and
1772
+ decompression (48\ MiB and 5\ MiB, respectively):
1773
+ .RS
1774
+ .PP
1775
+ .nf
1776
+ \f(CRtar cf \- baz | xz \-4e > baz.tar.xz\fR
1777
+ .fi
1778
+ .RE
1779
+ .PP
1780
+ Un mélange de fichiers compressés et non compressés peuvent être
1781
+ décompressés vers la sortie standard avec une simple commande :
1782
+ .RS
1783
+ .PP
1784
+ .nf
1785
+ \f(CRxz \-dcf a.txt b.txt.xz c.txt d.txt.lzma > abcd.txt\fR
1786
+ .fi
1787
+ .RE
1788
+ .
1789
+ .SS "Compression en parallèle de plusieurs fichiers"
1790
+ Sur GNU et *BSD, \fBfind\fP(1) et \fBxargs\fP(1) peuvent être utilisés pour mettre
1791
+ en parallèle la compression de plusieurs fichiers :
1792
+ .RS
1793
+ .PP
1794
+ .nf
1795
+ \f(CRfind . \-type f \e! \-name '*.xz' \-print0 \e
1796
+ | xargs \-0r \-P4 \-n16 xz \-T1\fR
1797
+ .fi
1798
+ .RE
1799
+ .PP
1800
+ L'option \fBP\fP passée à \fBxargs\fP(1) fixe le nombre de processus \fBxz\fP en
1801
+ parallèles. La meilleure valeur pour l'option \fBn\fP dépend du nombre de
1802
+ fichiers à compresser. S\-il n'y a que quelques fichiers, la valeur sera
1803
+ probablement 1 ; avec des dizaines de milliers de fichiers, 100 ou même plus
1804
+ serait approprié pour réduire le nombre de processus \fBxz\fP que \fBxargs\fP(1)
1805
+ créera éventuellement.
1806
+ .PP
1807
+ L'option \fB\-T1\fP de \fBxz\fP est là pour le forcer en mode mono\-thread, car
1808
+ \fBxargs\fP(1) est utilisé pour contrôler la quantité de mise en parallèle.
1809
+ .
1810
+ .SS "Mode robot"
1811
+ Calculer combien d'octets ont été économisés au total après avoir compressé
1812
+ plusieurs fichiers :
1813
+ .RS
1814
+ .PP
1815
+ .nf
1816
+ \f(CRxz \-\-robot \-\-list *.xz | awk '/^totals/{print $5\-$4}'\fR
1817
+ .fi
1818
+ .RE
1819
+ .PP
1820
+ Un script peut vouloir savoir qu'il utilise une version suffisamment récente
1821
+ de \fBxz\fP. Le script \fBsh\fP(1) suivant vérifie que le numéro de version de
1822
+ l'outil \fBxz\fP soit au minimum 5.0.0. Cette méthode est compatible avec les
1823
+ vieilles versions bêta, qui ne gèrent pas l'option \fB\-\-robot\fP :
1824
+ .RS
1825
+ .PP
1826
+ .nf
1827
+ \f(CRif ! eval "$(xz \-\-robot \-\-version 2> /dev/null)" ||
1828
+ [ "$XZ_VERSION" \-lt 50000002 ]; then
1829
+ echo "Your xz is too old."
1830
+ fi
1831
+ unset XZ_VERSION LIBLZMA_VERSION\fR
1832
+ .fi
1833
+ .RE
1834
+ .PP
1835
+ Régler une limite d'utilisation de la mémoire pour la décompression en
1836
+ utilisant \fBXZ_OPT\fP, mais si une limite a déjà été définie, ne pas
1837
+ l'augmenter :
1838
+ .RS
1839
+ .PP
1840
+ .nf
1841
+ \f(CRNEWLIM=$((123 << 20))\ \ # 123 MiB
1842
+ OLDLIM=$(xz \-\-robot \-\-info\-memory | cut \-f3)
1843
+ if [ $OLDLIM \-eq 0 \-o $OLDLIM \-gt $NEWLIM ]; then
1844
+ XZ_OPT="$XZ_OPT \-\-memlimit\-decompress=$NEWLIM"
1845
+ export XZ_OPT
1846
+ fi\fR
1847
+ .fi
1848
+ .RE
1849
+ .
1850
+ .SS "Chaînes de filtres de compresseur personnalisées"
1851
+ L'utilisation la plus simple des chaînes de filtres personnalisées est la
1852
+ personnalisation d'un préréglage LZMA2. Cela peut être utile, car les
1853
+ préréglages ne couvrent qu'un sous\-ensemble des réglages de compression
1854
+ potentiellement utiles.
1855
+ .PP
1856
+ Les colonnes CompCPU des tableaux des descriptions des options \fB\-0\fP à \fB\-9\fP
1857
+ et \fB\-\-extreme\fP sont utiles lors de la personnalisation des préréglages
1858
+ LZMA2. Voici les parties pertinentes recueillies à partir de ces deux
1859
+ tableaux :
1860
+ .RS
1861
+ .PP
1862
+ .TS
1863
+ tab(;);
1864
+ c c
1865
+ n n.
1866
+ Préréglage;CompCPU
1867
+ \-0;0
1868
+ \-1;1
1869
+ \-2;2
1870
+ \-3;3
1871
+ \-4;4
1872
+ \-5;5
1873
+ \-6;6
1874
+ \-5e;7
1875
+ \-6e;8
1876
+ .TE
1877
+ .RE
1878
+ .PP
1879
+ If you know that a file requires somewhat big dictionary (for example, 32\ MiB) to compress well, but you want to compress it quicker than \fBxz \-8\fP
1880
+ would do, a preset with a low CompCPU value (for example, 1) can be
1881
+ modified to use a bigger dictionary:
1882
+ .RS
1883
+ .PP
1884
+ .nf
1885
+ \f(CRxz \-\-lzma2=preset=1,dict=32MiB foo.tar\fR
1886
+ .fi
1887
+ .RE
1888
+ .PP
1889
+ Avec certains fichiers, la commande ci\-dessus peut être plus rapide que
1890
+ \fBxz\-6\fP tout en compressant bien mieux. Cependant, il faut souligner que
1891
+ seuls certains fichiers bénéficient d'un grand dictionnaire tout en gardant
1892
+ la valeur de CompCPU faible. La siutation la plus évidente où un gros
1893
+ dictionnaire peut baucoup aider, est une archive contenant des fichiers très
1894
+ similaires de quelques megaoctets chacun. La taille de dictionnaire doit
1895
+ être significativement plus grosse que tout fichier individuel pour
1896
+ permettre à LZMA2 de tirer pleinement partie des similarités entre des
1897
+ fichiers consécutifs.
1898
+ .PP
1899
+ Si une utilisation de la mémoire élevée pour la compression et décompression
1900
+ convient, et que le fichier à compresser a une taille de plusieurs centaines
1901
+ de megaoctets, il peut être utile d'utiliser un plus gros dictionnaire que
1902
+ celui fourni par \fBxz\-9\fP (64 Mio) :
1903
+ .RS
1904
+ .PP
1905
+ .nf
1906
+ \f(CRxz \-vv \-\-lzma2=dict=192MiB big_foo.tar\fR
1907
+ .fi
1908
+ .RE
1909
+ .PP
1910
+ Utiliser \fB\-vv\fP (\fB\-\-verbose\-\-verbose\fP) comme dans l'exemple ci\-dessus peut
1911
+ être utile pour voir les besoins en mémoire du compresseur et du
1912
+ décompresseur. Rappelez\-vous qu'utiliser un dictionnaire plus gros que la
1913
+ taille du fichier non compressé est un gachis de mémoire, donc la commande
1914
+ ci\-dessus n'est pas utile pour les petits fichiers.
1915
+ .PP
1916
+ Sometimes the compression time doesn't matter, but the decompressor memory
1917
+ usage has to be kept low, for example, to make it possible to decompress the
1918
+ file on an embedded system. The following command uses \fB\-6e\fP (\fB\-6 \-\-extreme\fP) as a base and sets the dictionary to only 64\ KiB. The
1919
+ resulting file can be decompressed with XZ Embedded (that's why there is
1920
+ \fB\-\-check=crc32\fP) using about 100\ KiB of memory.
1921
+ .RS
1922
+ .PP
1923
+ .nf
1924
+ \f(CRxz \-\-check=crc32 \-\-lzma2=preset=6e,dict=64KiB foo\fR
1925
+ .fi
1926
+ .RE
1927
+ .PP
1928
+ If you want to squeeze out as many bytes as possible, adjusting the number
1929
+ of literal context bits (\fIlc\fP) and number of position bits (\fIpb\fP) can
1930
+ sometimes help. Adjusting the number of literal position bits (\fIlp\fP)
1931
+ might help too, but usually \fIlc\fP and \fIpb\fP are more important. For
1932
+ example, a source code archive contains mostly US\-ASCII text, so something
1933
+ like the following might give slightly (like 0.1\ %) smaller file than \fBxz \-6e\fP (try also without \fBlc=4\fP):
1934
+ .RS
1935
+ .PP
1936
+ .nf
1937
+ \f(CRxz \-\-lzma2=preset=6e,pb=0,lc=4 source_code.tar\fR
1938
+ .fi
1939
+ .RE
1940
+ .PP
1941
+ Using another filter together with LZMA2 can improve compression with
1942
+ certain file types. For example, to compress a x86\-32 or x86\-64 shared
1943
+ library using the x86 BCJ filter:
1944
+ .RS
1945
+ .PP
1946
+ .nf
1947
+ \f(CRxz \-\-x86 \-\-lzma2 libfoo.so\fR
1948
+ .fi
1949
+ .RE
1950
+ .PP
1951
+ Notez que l'ordre des options de filtre est significatif. Si \fB\-\-x86\fP est
1952
+ indiqué après \fB\-\-lzma2\fP, \fBxz\fP donnera une erreur, car il ne peut y avoir
1953
+ aucun filtre après LZMA2, et aussi parce que le filtre BCJ x86 ne peut pas
1954
+ être utilisé comme dernier filtre dans la chaîne.
1955
+ .PP
1956
+ Le filtre Delta associé à LZMA2 peut donner de bons résultats avec les
1957
+ images bitmap. Cela devrait habituellement battre PNG, qui a quelques
1958
+ filtres avancés supplémentaires qu'un simple delta, mais qui utilise Deflate
1959
+ pour la compression effective.
1960
+ .PP
1961
+ The image has to be saved in uncompressed format, for example, as
1962
+ uncompressed TIFF. The distance parameter of the Delta filter is set to
1963
+ match the number of bytes per pixel in the image. For example, 24\-bit RGB
1964
+ bitmap needs \fBdist=3\fP, and it is also good to pass \fBpb=0\fP to LZMA2 to
1965
+ accommodate the three\-byte alignment:
1966
+ .RS
1967
+ .PP
1968
+ .nf
1969
+ \f(CRxz \-\-delta=dist=3 \-\-lzma2=pb=0 foo.tiff\fR
1970
+ .fi
1971
+ .RE
1972
+ .PP
1973
+ If multiple images have been put into a single archive (for example,
1974
+ \&\fB.tar\fP), the Delta filter will work on that too as long as all images have
1975
+ the same number of bytes per pixel.
1976
+ .
1977
+ .SH "VOIR AUSSI"
1978
+ \fBxzdec\fP(1), \fBxzdiff\fP(1), \fBxzgrep\fP(1), \fBxzless\fP(1), \fBxzmore\fP(1),
1979
+ \fBgzip\fP(1), \fBbzip2\fP(1), \fB7z\fP(1)
1980
+ .PP
1981
+ XZ Utils: <https://tukaani.org/xz/>
1982
+ .br
1983
+ XZ Embedded: <https://tukaani.org/xz/embedded.html>
1984
+ .br
1985
+ LZMA SDK: <https://7\-zip.org/sdk.html>
janus/share/man/fr/man1/xz.1 ADDED
@@ -0,0 +1,1985 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ '\" t
2
+ .\" SPDX-License-Identifier: 0BSD
3
+ .\"
4
+ .\" Authors: Lasse Collin
5
+ .\" Jia Tan
6
+ .\"
7
+ .\" french translation of XZ Utils man
8
+ .\" Copyright (C) 2021 Debian French l10n team <debian-l10n-french@lists.debian.org>
9
+ .\" Translator
10
+ .\" bubu <bubub@no-log.org>, 2021.
11
+ .\"
12
+ .\"*******************************************************************
13
+ .\"
14
+ .\" This file was generated with po4a. Translate the source file.
15
+ .\"
16
+ .\"*******************************************************************
17
+ .TH XZ 1 2024\-12\-30 Tukaani "Utilitaires XZ"
18
+ .
19
+ .SH NOM
20
+ xz, unxz, xzcat, lzma, unlzma, lzcat \- Compresser ou décompresser des
21
+ fichiers .xz et .lzma
22
+ .
23
+ .SH SYNOPSIS
24
+ \fBxz\fP [\fIoption...\fP] [\fIfichier...\fP]
25
+ .
26
+ .SH "ALIAS DES COMMANDES"
27
+ \fBunxz\fP est équivalent à \fBxz \-\-decompress\fP.
28
+ .br
29
+ \fBxzcat\fP est équivalent à \fBxz \-\-decompress \-\-stdout\fP
30
+ .br
31
+ \fBlzma\fP est équivalent à \fBxz \-\-format=lzma\fP
32
+ .br
33
+ \fBunlzma\fP est équivalent à \fBxz \-\-format=lzma \-\-decompress\fP
34
+ .br
35
+ \fBlzcat\fP est équivalent à \fBxz \-\-format=lzma \-\-decompress \-\- stdout\fP
36
+ .PP
37
+ Lors de l'écriture de scripts qui nécessitent de décompresser des fichiers,
38
+ il est recommandé de toujours utiliser la commande \fBxz\fP avec les arguments
39
+ appropriés (\fBxz \-d\fP ou \fBxz \-dc\fP) au lieu des commandes \fBunxz\fP et
40
+ \fBxzcat\fP.
41
+ .
42
+ .SH DESCRIPTION
43
+ \fBxz\fP is a general\-purpose data compression tool with command line syntax
44
+ similar to \fBgzip\fP(1) and \fBbzip2\fP(1). The native file format is the
45
+ \&\fB.xz\fP format, but the legacy \fB.lzma\fP format used by LZMA Utils and raw
46
+ compressed streams with no container format headers are also supported. In
47
+ addition, decompression of the \fB.lz\fP format used by \fBlzip\fP is supported.
48
+ .PP
49
+ \fBxz\fP compresse ou décompresse chaque \fIfichier\fP en fonction du mode
50
+ d'opération choisi. Si aucun \fIfichier\fP n'est donné ou \fIfichier\fP est \fB\-\fP,
51
+ \fBxz\fP lit depuis l'entrée standard et écrit les données traitées sur la
52
+ sortie standard. \fBxz\fP refusera (affichera une erreur et ignorera le
53
+ \fIfichier\fP) d'écrire les données compressées sur la sortie standard si c'est
54
+ un terminal. De même, \fBxz\fP refusera de lire des données compressées depuis
55
+ l'entrée standard si c'est un terminal.
56
+ .PP
57
+ A moins que \fB\-\-sdout\fP ne soit indiqué, les \fIfichiers\fP autres que \fB\-\fP sont
58
+ écrits dans un nouveau fichier dont le nom est dérivé du nom de \fIfichier\fP
59
+ source :
60
+ .IP \(bu 3
61
+ Lors de la compression, le suffixe du format de fichier cible (\fB.xz\fP ou
62
+ \&\fB.lzma\fP) est ajouté au nom de fichier source pour obtenir le nom de fichier
63
+ cible.
64
+ .IP \(bu 3
65
+ When decompressing, the \fB.xz\fP, \fB.lzma\fP, or \fB.lz\fP suffix is removed from
66
+ the filename to get the target filename. \fBxz\fP also recognizes the suffixes
67
+ \&\fB.txz\fP and \fB.tlz\fP, and replaces them with the \fB.tar\fP suffix.
68
+ .PP
69
+ Si le fichier cible existe déjà, une erreur est affichée et le \fIfichier\fP
70
+ est ignoré.
71
+ .PP
72
+ Sauf s'il écrit dans la sortie standard, \fBxz\fP affichera un avertissement et
73
+ ignorera le \fIfichier\fP dans les cas suivants :
74
+ .IP \(bu 3
75
+ \fIfichier\fP n'est pas un fichier normal. Les liens symboliques ne sont pas
76
+ suivis et donc ne sont pas considérés comme des fichiers normaux.
77
+ .IP \(bu 3
78
+ \fIfichier\fP a plusieurs liens physiques.
79
+ .IP \(bu 3
80
+ \fIfichier\fP a un setuid, setgid ou sticky bit positionné.
81
+ .IP \(bu 3
82
+ Le mode d'opération est défini pour compresser et le \fIfichier\fP a déjà un
83
+ suffixe du format de fichier cible (\fB.xz\fP ou \fB.txz\fP lors d'une compression
84
+ en format \fB.xz\fP, et \fB.lzma\fP ou \fB.tlz\fP lors d'une compression en format
85
+ \&\fB.lzma\fP).
86
+ .IP \(bu 3
87
+ The operation mode is set to decompress and the \fIfile\fP doesn't have a
88
+ suffix of any of the supported file formats (\fB.xz\fP, \fB.txz\fP, \fB.lzma\fP,
89
+ \&\fB.tlz\fP, or \fB.lz\fP).
90
+ .PP
91
+ Après la compression ou la décompression réussie du \fIfichier\fP, \fBxz\fP copie
92
+ les permissions du propriétaire, du groupe, la date d'accès, et les
93
+ modifications d'heure depuis le \fIfichier\fP source du fichier cible. Si la
94
+ copie du groupe échoue, les permissions sont modifiées pour que le fichier
95
+ cible ne soit pas accessible aux utilisateurs qui n'ont pas les droits
96
+ d'accès au \fIfichier\fP source. \fBxz\fP ne prend actuellement pas en charge la
97
+ copie d'autres métadonnées telles que les listes de contrôle d'accès ou les
98
+ attributs étendus.
99
+ .PP
100
+ Once the target file has been successfully closed, the source \fIfile\fP is
101
+ removed unless \fB\-\-keep\fP was specified. The source \fIfile\fP is never removed
102
+ if the output is written to standard output or if an error occurs.
103
+ .PP
104
+ Envoyer \fBSIGINFO\fP ou \fBSIGURSR1\fP au processus \fBxz\fP, lui fait afficher
105
+ l'information de progression sur l'erreur standard. Cela a un intérêt limité
106
+ car lorsque l'erreur standard est un terminal, utiliser \fB\-\-verbose\fP
107
+ affichera automatiquement un indicateur de progression du processus.
108
+ .
109
+ .SS "Utilisation de la mémoire"
110
+ L'utilisation de la mémoire par \fBxz\fP varie de quelques centaines de
111
+ kilo\-octets à plusieurs gigaoctects en fonction des paramètres de
112
+ compression. Les réglages utilisés lors de la compression d'un fichier
113
+ déterminent les besoins en mémoire pour la décompression. Habituellement la
114
+ décompression nécessite 5\% à 20\% de la quantité de mémoire utilisée pour
115
+ la compression du fichier. Par exemple, décompresser un fichier créé avec
116
+ \fBxz\-9\fP recquiert habituellement 65\ Mio de mémoire. Bien qu'il soit
117
+ possible d'avoir des fichiers \fB.xz\fP nécessitant plusieurs gigaoctets de
118
+ mémoire pour être décompressés.
119
+ .PP
120
+ Especially users of older systems may find the possibility of very large
121
+ memory usage annoying. To prevent uncomfortable surprises, \fBxz\fP has a
122
+ built\-in memory usage limiter, which is disabled by default. While some
123
+ operating systems provide ways to limit the memory usage of processes,
124
+ relying on it wasn't deemed to be flexible enough (for example, using
125
+ \fBulimit\fP(1) to limit virtual memory tends to cripple \fBmmap\fP(2)).
126
+ .PP
127
+ The memory usage limiter can be enabled with the command line option
128
+ \fB\-\-memlimit=\fP\fIlimit\fP. Often it is more convenient to enable the limiter
129
+ by default by setting the environment variable \fBXZ_DEFAULTS\fP, for example,
130
+ \fBXZ_DEFAULTS=\-\-memlimit=150MiB\fP. It is possible to set the limits
131
+ separately for compression and decompression by using
132
+ \fB\-\-memlimit\-compress=\fP\fIlimit\fP and \fB\-\-memlimit\-decompress=\fP\fIlimit\fP.
133
+ Using these two options outside \fBXZ_DEFAULTS\fP is rarely useful because a
134
+ single run of \fBxz\fP cannot do both compression and decompression and
135
+ \fB\-\-memlimit=\fP\fIlimit\fP (or \fB\-M\fP \fIlimit\fP) is shorter to type on the
136
+ command line.
137
+ .PP
138
+ If the specified memory usage limit is exceeded when decompressing, \fBxz\fP
139
+ will display an error and decompressing the file will fail. If the limit is
140
+ exceeded when compressing, \fBxz\fP will try to scale the settings down so that
141
+ the limit is no longer exceeded (except when using \fB\-\-format=raw\fP or
142
+ \fB\-\-no\-adjust\fP). This way the operation won't fail unless the limit is very
143
+ small. The scaling of the settings is done in steps that don't match the
144
+ compression level presets, for example, if the limit is only slightly less
145
+ than the amount required for \fBxz \-9\fP, the settings will be scaled down only
146
+ a little, not all the way down to \fBxz \-8\fP.
147
+ .
148
+ .SS "Concaténation et remplissage avec des fichiers .xz"
149
+ Il est possible de concaténer les fichiers \fB.xz\fP tels quel. \fBxz\fP
150
+ décompressera de tels fichiers comme s'ils étaient un unique fichier \fB.xz\fP.
151
+ .PP
152
+ It is possible to insert padding between the concatenated parts or after the
153
+ last part. The padding must consist of null bytes and the size of the
154
+ padding must be a multiple of four bytes. This can be useful, for example,
155
+ if the \fB.xz\fP file is stored on a medium that measures file sizes in
156
+ 512\-byte blocks.
157
+ .PP
158
+ La concaténation et le remplissage ne sont pas autorisés avec les fichiers
159
+ \&\fB.lzma\fP ou les flux bruts.
160
+ .
161
+ .SH OPTIONS
162
+ .
163
+ .SS "Suffixes entiers et valeurs spéciales."
164
+ Dans la plupart des endroits où un argument entier est attendu, un suffixe
165
+ optionel permet d'indiquer facilement les grands entiers. Il ne doit pas y
166
+ avoir d'espace entre l'entier et le suffixe.
167
+ .TP
168
+ \fBKiB\fP
169
+ Multiplier l'entier par 1024 (2^10). \fBKi\fP, \fBk\fP, \fBkB\fP, \fBK\fP et \fBKB\fP sont
170
+ acceptés comme synonymes de \fBKiB\fP.
171
+ .TP
172
+ \fBMiB\fP
173
+ Multiplier l'entier par 1 048 576 (2^20). \fBMi\fP, \fBm\fP, \fBM\fP et \fBMB\fP sont
174
+ acceptés comme synonymes de \fBMiB\fP.
175
+ .TP
176
+ \fBGiB\fP
177
+ Multiplier l'entier par 1 073 741 824 (2^30). \fBGi\fP, \fBg\fP, \fBG\fP et \fBGB\fP
178
+ sont acceptés comme synonymes de \fBGiB\fP.
179
+ .PP
180
+ La valeur spéciale \fBmax\fP peut être utilisée pour indiquer la valeur
181
+ maximale de l'entier prise en charge par l'option.
182
+ .
183
+ .SS "Mode d'opération"
184
+ Si plusieurs options de mode d'opération sont données, la dernière prend
185
+ effet.
186
+ .TP
187
+ \fB\-z\fP, \fB\-\-compress\fP
188
+ Compresser. C'est le mode d'opération par défaut lorsque aucune option de
189
+ mode opératoire n'est spécifiée ou qu'aucun autre mode d'opération n'est
190
+ sous\-entendu par le nom de la commande (par exemple \fBunxz\fP sous\-entend
191
+ \fB\-\-decompress\fP).
192
+ .IP ""
193
+ .\" The DESCRIPTION section already says this but it's good to repeat it
194
+ .\" here because the default behavior is a bit dangerous and new users
195
+ .\" in a hurry may skip reading the DESCRIPTION section.
196
+ After successful compression, the source file is removed unless writing to
197
+ standard output or \fB\-\-keep\fP was specified.
198
+ .TP
199
+ \fB\-d\fP, \fB\-\-decompress\fP, \fB\-\-uncompress\fP
200
+ .\" The DESCRIPTION section already says this but it's good to repeat it
201
+ .\" here because the default behavior is a bit dangerous and new users
202
+ .\" in a hurry may skip reading the DESCRIPTION section.
203
+ Decompress. After successful decompression, the source file is removed
204
+ unless writing to standard output or \fB\-\-keep\fP was specified.
205
+ .TP
206
+ \fB\-t\fP, \fB\-\-test\fP
207
+ Tester l'intégrité des \fIfichiers\fP compressés. Cette option est équivalente
208
+ à \fB\-\-decompress \-\-stdout\fP sauf que les données décompressées sont rejetées
209
+ au lieu d'être écrites sur la sortie standard. Aucun fichier n'est créé ou
210
+ supprimé.
211
+ .TP
212
+ \fB\-l\fP, \fB\-\-list\fP
213
+ Afficher l'information sur les \fIfichiers\fP compressés. Aucune sortie
214
+ non\-compressée n'est produite et aucun fichier n'est créé ou supprimé. En
215
+ mode liste, le programme ne peut pas lire les données compressées depuis
216
+ l'entrée standard ou depuis d'autres sources non adressables.
217
+ .IP ""
218
+ The default listing shows basic information about \fIfiles\fP, one file per
219
+ line. To get more detailed information, use also the \fB\-\-verbose\fP option.
220
+ For even more information, use \fB\-\-verbose\fP twice, but note that this may be
221
+ slow, because getting all the extra information requires many seeks. The
222
+ width of verbose output exceeds 80 characters, so piping the output to, for
223
+ example, \fBless\ \-S\fP may be convenient if the terminal isn't wide enough.
224
+ .IP ""
225
+ La sortie exacte peut varier suivant les versions de \fBxz\fP et les différents
226
+ paramètres régionaux. Pour une sortie lisible par la machine, utiliser
227
+ \fB\-\-robot \-\-list\fP.
228
+ .
229
+ .SS "Modificateurs d'opération"
230
+ .TP
231
+ \fB\-k\fP, \fB\-\-keep\fP
232
+ Ne pas effacer les fichiers d'entrée.
233
+ .IP ""
234
+ Since \fBxz\fP 5.2.6, this option also makes \fBxz\fP compress or decompress even
235
+ if the input is a symbolic link to a regular file, has more than one hard
236
+ link, or has the setuid, setgid, or sticky bit set. The setuid, setgid, and
237
+ sticky bits are not copied to the target file. In earlier versions this was
238
+ only done with \fB\-\-force\fP.
239
+ .TP
240
+ \fB\-f\fP, \fB\-\-force\fP
241
+ Cette option a plusieurs effets :
242
+ .RS
243
+ .IP \(bu 3
244
+ Si le fichier cible existe déjà, l'effacer avant de compresser ou
245
+ décompresser.
246
+ .IP \(bu 3
247
+ Compresser ou décompresser même si l'entrée est un lien symbolique vers un
248
+ fichier normal, a plus qu'un lien physique, ou a le bit setuid, setgid ou
249
+ sticky défini. Les bits setuid, setgid et sticky bits ne sont pas copiés
250
+ dans le fichier cible.
251
+ .IP \(bu 3
252
+ Lorsque \fBxz\fP est utilisé avec \fB\-\-decompress\fP \fB\-\-stdout\fP et qu'il ne peut
253
+ pas reconnaitre le type du fichier source, copier le fichier source tel quel
254
+ dans la sortie standard. Celà permet à \fBxzcat\fP \fB\-\-force\fP d'être utilisé
255
+ comme \fBcat\fP(1) pour les fichiers qui n'ont pas été compressé avec
256
+ \fBxz\fP. Remarquez que dans le futur, \fBxz\fP devrait prendre en charge de
257
+ nouveaux formats de fichiers compressés, ce qui permettra à \fBxz\fP de
258
+ décompresser plus de types de fichiers au lieu de les copier tels quels dans
259
+ la sortie standard. \fB\-\-format=\fP\fIformat\fP peut être utilisé pour contraindre
260
+ \fBxz\fP à décompresser seulement un format de fichier.
261
+ .RE
262
+ .TP
263
+ \fB\-c\fP, \fB\-\-stdout\fP, \fB\-\-to\-stdout\fP
264
+ Écrire les données compressées ou décompressées sur la sortie standard
265
+ plutôt que dans un fichier. Cela necessite \fB\-\-keep\fP.
266
+ .TP
267
+ \fB\-\-single\-stream\fP
268
+ Décompresser seulement le premier flux \fB.xz\fP et ignorer silencieusement les
269
+ possibles données d'entrée résiduelles qui suivent le flux. Normalement ces
270
+ déchets excédentaires provoquent l'affichage d'une erreur par \fBxz\fP.
271
+ .IP ""
272
+ \fBxz\fP ne décompresse jamais plus d'un flux à partir de fichiers \fB.lzma\fP ou
273
+ de flux bruts, mais cette option fait aussi que \fBxz\fP ignorera les données
274
+ résiduelles après le fichier \fB.lzma\fP ou le flux brut.
275
+ .IP ""
276
+ Cette option n'a aucun effet si le mode d'opération n'est pas
277
+ \fB\-\-decompress\fP ou \fB\-\-test\fP.
278
+ .TP
279
+ \fB\-\-no\-sparse\fP
280
+ Désactiver la création de fichiers peu denses. Par défaut, lors de la
281
+ décompression en un fichier normal, \fBxz\fP essaie d'en faire un fichier creux
282
+ si les données décompressées contiennent de longues séquences de zéros
283
+ binaires. Cela fonctionne aussi lors de l'écriture sur la sortie standard
284
+ aussi longtemps que la sortie standard est connectée à un fichier normal et
285
+ que certaines conditions supplémentaires sont satisfaites pour le faire de
286
+ manière sécurisée. Créer des fichiers creux peut épargner de l'espace disque
287
+ et accélérer la décompression en réduisant la quantité d'entrées/sorties sur
288
+ le disque.
289
+ .TP
290
+ \fB\-S\fP \fI.suf\fP, \fB\-\-suffix=\fP\fI.suf\fP
291
+ Lors de la compression, utiliser \fB.suf\fP comme suffixe du fichier cible au
292
+ lieu de \fB.xz\fP ou \fB.lzma\fP. Si \fBxz\fP n'écrit pas sur la sortie standard et
293
+ si le fichier source a déja le suffixe \fB.suf\fP, un avertissement est affiché
294
+ et le fichier est ignoré.
295
+ .IP ""
296
+ When decompressing, recognize files with the suffix \fI.suf\fP in addition to
297
+ files with the \fB.xz\fP, \fB.txz\fP, \fB.lzma\fP, \fB.tlz\fP, or \fB.lz\fP suffix. If the
298
+ source file has the suffix \fI.suf\fP, the suffix is removed to get the target
299
+ filename.
300
+ .IP ""
301
+ Lors de la compression ou décompression de flux bruts (\fB\-\-fomat=raw\fP), le
302
+ suffixe doit toujours être spécifié à moins d'écrire sur la sortie standard,
303
+ car il n'y a pas de suffixe par défaut pour les flux bruts.
304
+ .TP
305
+ \fB\-\-files\fP[\fB=\fP\fIfichier\fP]
306
+ Lire les noms de fichier à traiter depuis \fIfichier\fP ; si \fIfichier\fP est
307
+ omis , les noms de fichier sont lus sur l'entrée standard. Les noms de
308
+ fichier doivent se terminer avec le caractère de nouvelle ligne. Un tiret
309
+ (\fB\-\fP) est considéré comme un nom de fichier normal ; ce qui ne signifie pas
310
+ entrée standard. Si les noms de fichier sont aussi donnés comme arguments de
311
+ ligne de commande, ils sont traités avant les noms de fichier lus depuis
312
+ \fIfichier\fP.
313
+ .TP
314
+ \fB\-\-files0\fP[\fB=\fP\fIfichier\fP]
315
+ Cela est identique à \fB\-\-files\fP[\fB=\fP\fIfichier\fP] sauf que chaque nom de
316
+ fichier doit se terminer par le caractère null.
317
+ .
318
+ .SS "Format de fichier basique et options de compression"
319
+ .TP
320
+ \fB\-F\fP \fIformat\fP, \fB\-\-format=\fP\fIformat\fP
321
+ Indiquer le \fIformat\fP de fichier à compresser ou décompresser :
322
+ .RS
323
+ .TP
324
+ \fBauto\fP
325
+ C'est celui par défaut. Lors de la compression, \fBauto\fP est équivalent à
326
+ \fBxz\fP. Lors de la décompression, le format du fichier en entrée est détecté
327
+ automatiquement. Notez que les flux bruts (créés avec \fB\-\-format=raw\fP) ne
328
+ peuvent pas être détectés automatiquement.
329
+ .TP
330
+ \fBxz\fP
331
+ Compresser dans le format de fichier \fB.xz\fP ou n'accepter que les fichiers
332
+ \&\fB.xz\fP à décompresser.
333
+ .TP
334
+ \fBlzma\fP, \fBalone\fP
335
+ Compresser au format de fichier \fB.lzma\fP historique, ou n'accepter que les
336
+ fichiers \fB.lzma\fP lors de la décompression. Le nom alternatif \fBalone\fP est
337
+ fourni pour la rétrocompatibilité avec les utilitaires LZMA.
338
+ .TP
339
+ \fBlzip\fP
340
+ Accept only \fB.lz\fP files when decompressing. Compression is not supported.
341
+ .IP ""
342
+ The \fB.lz\fP format version 0 and the unextended version 1 are supported.
343
+ Version 0 files were produced by \fBlzip\fP 1.3 and older. Such files aren't
344
+ common but may be found from file archives as a few source packages were
345
+ released in this format. People might have old personal files in this
346
+ format too. Decompression support for the format version 0 was removed in
347
+ \fBlzip\fP 1.18.
348
+ .IP ""
349
+ \fBlzip\fP 1.4 and later create files in the format version 1. The sync flush
350
+ marker extension to the format version 1 was added in \fBlzip\fP 1.6. This
351
+ extension is rarely used and isn't supported by \fBxz\fP (diagnosed as corrupt
352
+ input).
353
+ .TP
354
+ \fBraw\fP
355
+ Compresser ou décompresser un flux brut (sans en\-têtes). Cela est réservé
356
+ seulement aux utilisateurs aguerris. Pour décoder des flux bruts, vous devez
357
+ utiliser \fB\-\-format=raw\fP et spécifier explicitement la chaîne de filtre, qui
358
+ normalement aurait du être stockée dans les en\-têtes du conteneur.
359
+ .RE
360
+ .TP
361
+ \fB\-C\fP \fIvérif.\fP, \fB\-\-check=\fP\fIvérif.\fP
362
+ Spécifier le type d'intégrité à vérifier. La vérification est calculée à
363
+ partir des données non\-compressées et stockées dans le fichier \fB.xz\fP. Cette
364
+ option n'a effet que si la compression a été faite dans le format \fB.xz\fP ;
365
+ le format \fB.lzma\fP ne gère pas les vérifications d'intégrité. Le contrôle
366
+ d'intégrité (s'il y en a) est vérifié lorsque le fichier \fB.xz\fP est
367
+ décompressé.
368
+ .IP ""
369
+ Types de \fIvérification\fP pris en charge :
370
+ .RS
371
+ .TP
372
+ \fBnone\fP
373
+ Ne pas calculer de vérification d'intégrité du tout. C'est généralement une
374
+ mauvaise idée. Cela peut être utile lorsque l'intégrité des données est
375
+ vérifiée de toute façon par d'autres manières.
376
+ .TP
377
+ \fBcrc32\fP
378
+ Calculer CRC32 en utilisant le polynôme de IEEE\-802.3 (Ethernet).
379
+ .TP
380
+ \fBcrc64\fP
381
+ Calculer CRC64 en utilisant le polynôme de ECMA\-182. C'est la manière
382
+ utilisée par défaut, car c'est légèrement mieux que CRC32 pour détecter les
383
+ fichiers endommagés et la différence de vitesse est négligeable.
384
+ .TP
385
+ \fBsha256\fP
386
+ Calculer SHA\-256. C'est quelque peu plus lent que CRC32 et CRC64.
387
+ .RE
388
+ .IP ""
389
+ L'intégrité des en\-têtes \fB.xz\fP est toujours vérifiée avec CRC32. Il n'est
390
+ pas possible de le changer ou de le désactiver.
391
+ .TP
392
+ \fB\-\-ignore\-check\fP
393
+ Ne pas contrôler la vérification d'intégrité des données lors de la
394
+ décompression. Les valeurs CRC32 dans les en\-têtes \fB.xz\fP seront normalement
395
+ toujours vérifiées.
396
+ .IP ""
397
+ \fBN'utilisez pas cette option à moins de savoir ce que vous faites.\fP Les
398
+ raisons possibles pour utiliser cette option :
399
+ .RS
400
+ .IP \(bu 3
401
+ Essayer de récupérer des données d'un fichier .xz corrompu.
402
+ .IP \(bu 3
403
+ Accélérer la décompression. Cela importe surtout avec SHA\-256 ou avec les
404
+ fichiers qui ont été compressés extrêmement bien. Il est recommandé de ne
405
+ pas utiliser cette option dans ce but à moins que l'intégrité du fichier ne
406
+ soit vérifiée extérieurement d'une autre manière.
407
+ .RE
408
+ .TP
409
+ \fB\-0\fP ... \fB\-9\fP
410
+ Choisir un niveau de compression prédéfini. La valeur par défaut est
411
+ \fB6\fP. Si plusieurs niveaux de préréglage sont spécifi��s, c'est le dernier
412
+ qui sera pris en compte. Si une chaîne de filtres personnalisée a déjà été
413
+ choisie, définir un niveau de compression préréglé efface la chaîne de
414
+ filtres personnalisée.
415
+ .IP ""
416
+ Les différences entre les préréglages sont plus significatives qu'avec
417
+ \fBgzip\fP(1) et \fBbzip2\fP(1). les réglages de compression sélectionnés
418
+ déterminent les exigences en mémoire pour la décompression, ainsi, utiliser
419
+ un niveau de préréglage trop élevé peut rendre difficile à décompresser un
420
+ fichier sur un vieux système avec peu de RAM. Clairement, \fBce n'est pas une bonne idée d'utiliser \-9 aveuglément pour tout\fP comme ça l'est souvent avec
421
+ \fBgzip\fP(1) et \fBbzip2\fP(1).
422
+ .RS
423
+ .TP
424
+ \fB\-0\fP ... \fB\-3\fP
425
+ Ce sont des préréglages relativement rapides. \fB0\fP est parfois plus rapide
426
+ que \fBgzip \-9\fP tout en compressant bien mieux. Les réglages plus élevés ont
427
+ souvent une rapidité comparable à celle de \fBbzip2\fP(1) avec un taux de
428
+ compression comparable ou meilleur, même si les résultats dépendent beaucoup
429
+ du genre de données compressées.
430
+ .TP
431
+ \fB\-4\fP ... \fB\-6\fP
432
+ Good to very good compression while keeping decompressor memory usage
433
+ reasonable even for old systems. \fB\-6\fP is the default, which is usually a
434
+ good choice for distributing files that need to be decompressible even on
435
+ systems with only 16\ MiB RAM. (\fB\-5e\fP or \fB\-6e\fP may be worth considering
436
+ too. See \fB\-\-extreme\fP.)
437
+ .TP
438
+ \fB\-7 ... \-9\fP
439
+ C'est comme \fB\-6\fP mais avec des besoins en mémoire plus élevés pour la
440
+ compression et la décompression. Ce n'est utile que lorsque les fichiers
441
+ sont plus gros que 8\ Mio, 16\ Mio et 32\ Mio respectivement.
442
+ .RE
443
+ .IP ""
444
+ Sur le même matériel, la vitesse de décompression est sensiblement un nombre
445
+ constant d'octets de données compressées par seconde. En d'autres termes,
446
+ meilleure est la compression, plus rapide sera en général la
447
+ décompression. Cela signifie aussi que la quantité de sortie non compressée
448
+ produite par seconde peut varier beaucoup.
449
+ .IP ""
450
+ Le tableau suivant résume les caractéristiques des préréglages :
451
+ .RS
452
+ .RS
453
+ .PP
454
+ .TS
455
+ tab(;);
456
+ c c c c c
457
+ n n n n n.
458
+ Préréglage;DictSize;CompCPU;CompMem;DecMem
459
+ \-0;256 KiB;0;3 MiB;1 MiB
460
+ \-1;1 MiB;1;9 MiB;2 MiB
461
+ \-2;2 MiB;2;17 MiB;3 MiB
462
+ \-3;4 MiB;3;32 MiB;5 MiB
463
+ \-4;4 MiB;4;48 MiB;5 MiB
464
+ \-5;8 MiB;5;94 MiB;9 MiB
465
+ \-6;8 MiB;6;94 MiB;9 MiB
466
+ \-7;16 MiB;6;186 MiB;17 MiB
467
+ \-8;32 MiB;6;370 MiB;33 MiB
468
+ \-9;64 MiB;6;674 MiB;65 MiB
469
+ .TE
470
+ .RE
471
+ .RE
472
+ .IP ""
473
+ Descriptions des colonnes :
474
+ .RS
475
+ .IP \(bu 3
476
+ DictSize est la taille du dictionnaire de LZMA2. Utiliser un dictionnaire
477
+ plus gros que la taille du fichier non compressé est un gaspillage de
478
+ mémoire. C'est pourquoi il est bon d'éviter d'utiliser les préréglages de
479
+ \fB\-7\fP à \fB\-9\fP lorsqu'il n'y en a pas vraiment besoin. A \fB\-6\fP et plus bas,
480
+ la quantité de mémoire gaspillée est généralement assez basse pour ne pas
481
+ être un problème.
482
+ .IP \(bu 3
483
+ CompCPU est une représentation des préréglages de LZMA2 qui affectent la
484
+ vitesse de compression. La taille du dictionnaire aussi affecte la vitesse,
485
+ alors comme CompCPU est le même pour les niveaux de \fB\-6\fP à \fB\-9\fP, les plus
486
+ haut niveaux tendent à être un peu moins rapides. Pour être encore moins
487
+ rapide et du coup obtenir peut être une meilleure compression, consultez
488
+ \fB\-\-extreme\fP.
489
+ .IP \(bu 3
490
+ CompMem contains the compressor memory requirements in the single\-threaded
491
+ mode. It may vary slightly between \fBxz\fP versions.
492
+ .IP \(bu 3
493
+ DecMem contient les besoins en mémoire du décompresseur. Ce sont les
494
+ réglages de la compression qui déterminent les besoins en mémoire de la
495
+ décompression. L'exacte utilisation de la mémoire est légèrement supérieure
496
+ à la taille du dictionnaire LZMA2, mais les valeurs dans la table ont été
497
+ arrondies au prochain Mio supérieur.
498
+ .RE
499
+ .IP ""
500
+ Memory requirements of the multi\-threaded mode are significantly higher than
501
+ that of the single\-threaded mode. With the default value of
502
+ \fB\-\-block\-size\fP, each thread needs 3*3*DictSize plus CompMem or DecMem. For
503
+ example, four threads with preset \fB\-6\fP needs 660\(en670\ MiB of memory.
504
+ .TP
505
+ \fB\-e\fP, \fB\-\-extreme\fP
506
+ Utilisez un variant plus lent que les préréglages (\fB\-0\fP à \fB\-9\fP) pour
507
+ espérer avoir un taux de compression légèrement meilleur, mais en cas de
508
+ malchance cela peut être pire. L'utilisation mémoire du décompresseur n'est
509
+ pas affectée, mais l'utilisation mémoire du compresseur augmente un peu aux
510
+ niveaux de préréglages de \fB\-0\fP à \fB\-3\fP.
511
+ .IP ""
512
+ Depuis qu'il y a deux préréglages avec des tailles de dictionnaire de 4\ Mio
513
+ et 8 \Mio, les préréglages \fB\-3e\fP et \fB\-5e\fP utilisent des réglages
514
+ légèrement plus rapides que \fB\-4e\fP et \fB\-6e\fP, respectivement. De cette
515
+ manière, il n'y a pas deux préréglages identiques.
516
+ .RS
517
+ .RS
518
+ .PP
519
+ .TS
520
+ tab(;);
521
+ c c c c c
522
+ n n n n n.
523
+ Préréglage;DictSize;CompCPU;CompMem;DecMem
524
+ \-0e;256 KiB;8;4 MiB;1 MiB
525
+ \-1e;1 MiB;8;13 MiB;2 MiB
526
+ \-2e;2 MiB;8;25 MiB;3 MiB
527
+ \-3e;4 MiB;7;48 MiB;5 MiB
528
+ \-4e;4 MiB;8;48 MiB;5 MiB
529
+ \-5e;8 MiB;7;94 MiB;9 MiB
530
+ \-6e;8 MiB;8;94 MiB;9 MiB
531
+ \-7e;16 MiB;8;186 MiB;17 MiB
532
+ \-8e;32 MiB;8;370 MiB;33 MiB
533
+ \-9e;64 MiB;8;674 MiB;65 MiB
534
+ .TE
535
+ .RE
536
+ .RE
537
+ .IP ""
538
+ Par exemple, il y a un total de quatre préréglages qui utilisent un
539
+ dictionnaire de 8 Mio et qui sont dans l'ordre du plus rapide au plus lent :
540
+ \fB\-5\fP, \fB\-6\fP, \fB\-5e\fP et \fB\-6e\fP.
541
+ .TP
542
+ \fB\-\-fast\fP
543
+ .PD 0
544
+ .TP
545
+ \fB\-\-best\fP
546
+ .PD
547
+ Il y a néanmoins des alias trompeurs pour \fB\-0\fP et \fB\-9\fP,
548
+ respectivement. Ils ne sont fournis que pour des besoins de
549
+ rétro\-compatibilité avec les utilitaires LZMA. Evitez d'utiliser ces
550
+ options.
551
+ .TP
552
+ \fB\-\-block\-size=\fP\fItaille\fP
553
+ Lors de la compression dans le format \fB.xz\fP, les données de l'entrée sont
554
+ réparties en blocs de \fItaille\fP octets. Les blocs sont compressés
555
+ indépendamment les un des autres, ce qui aide avec le mode multithread
556
+ (multi\-threading) et rend possible la décompression à accès aléatoire
557
+ limité. Cette option est typiquement utilisée pour outrepasser la taille de
558
+ bloc en mode multithread, mais cette option peut aussi être utilisée en mode
559
+ single\-thread.
560
+ .IP ""
561
+ In multi\-threaded mode about three times \fIsize\fP bytes will be allocated in
562
+ each thread for buffering input and output. The default \fIsize\fP is three
563
+ times the LZMA2 dictionary size or 1 MiB, whichever is more. Typically a
564
+ good value is 2\(en4 times the size of the LZMA2 dictionary or at least 1
565
+ MiB. Using \fIsize\fP less than the LZMA2 dictionary size is waste of RAM
566
+ because then the LZMA2 dictionary buffer will never get fully used. In
567
+ multi\-threaded mode, the sizes of the blocks are stored in the block
568
+ headers. This size information is required for multi\-threaded
569
+ decompression.
570
+ .IP ""
571
+ In single\-threaded mode no block splitting is done by default. Setting this
572
+ option doesn't affect memory usage. No size information is stored in block
573
+ headers, thus files created in single\-threaded mode won't be identical to
574
+ files created in multi\-threaded mode. The lack of size information also
575
+ means that \fBxz\fP won't be able decompress the files in multi\-threaded mode.
576
+ .TP
577
+ \fB\-\-block\-list=\fP\fIitems\fP
578
+ When compressing to the \fB.xz\fP format, start a new block with an optional
579
+ custom filter chain after the given intervals of uncompressed data.
580
+ .IP ""
581
+ The \fIitems\fP are a comma\-separated list. Each item consists of an optional
582
+ filter chain number between 0 and 9 followed by a colon (\fB:\fP) and a
583
+ required size of uncompressed data. Omitting an item (two or more
584
+ consecutive commas) is a shorthand to use the size and filters of the
585
+ previous item.
586
+ .IP ""
587
+ If the input file is bigger than the sum of the sizes in \fIitems\fP, the last
588
+ item is repeated until the end of the file. A special value of \fB0\fP may be
589
+ used as the last size to indicate that the rest of the file should be
590
+ encoded as a single block.
591
+ .IP ""
592
+ An alternative filter chain for each block can be specified in combination
593
+ with the \fB\-\-filters1=\fP\fIfilters\fP \&...\& \fB\-\-filters9=\fP\fIfilters\fP options.
594
+ These options define filter chains with an identifier between 1\(en9.
595
+ Filter chain 0 can be used to refer to the default filter chain, which is
596
+ the same as not specifying a filter chain. The filter chain identifier can
597
+ be used before the uncompressed size, followed by a colon (\fB:\fP). For
598
+ example, if one specifies \fB\-\-block\-list=1:2MiB,3:2MiB,2:4MiB,,2MiB,0:4MiB\fP
599
+ then blocks will be created using:
600
+ .RS
601
+ .IP \(bu 3
602
+ The filter chain specified by \fB\-\-filters1\fP and 2 MiB input
603
+ .IP \(bu 3
604
+ The filter chain specified by \fB\-\-filters3\fP and 2 MiB input
605
+ .IP \(bu 3
606
+ The filter chain specified by \fB\-\-filters2\fP and 4 MiB input
607
+ .IP \(bu 3
608
+ The filter chain specified by \fB\-\-filters2\fP and 4 MiB input
609
+ .IP \(bu 3
610
+ The default filter chain and 2 MiB input
611
+ .IP \(bu 3
612
+ The default filter chain and 4 MiB input for every block until end of input.
613
+ .RE
614
+ .IP ""
615
+ If one specifies a size that exceeds the encoder's block size (either the
616
+ default value in threaded mode or the value specified with
617
+ \fB\-\-block\-size=\fP\fIsize\fP), the encoder will create additional blocks while
618
+ keeping the boundaries specified in \fIitems\fP. For example, if one specifies
619
+ \fB\-\-block\-size=10MiB\fP \fB\-\-block\-list=5MiB,10MiB,8MiB,12MiB,24MiB\fP and the
620
+ input file is 80 MiB, one will get 11 blocks: 5, 10, 8, 10, 2, 10, 10, 4,
621
+ 10, 10, and 1 MiB.
622
+ .IP ""
623
+ En mode multi\-threadé les tailles de blocs sont stockées dans les en\-têtes
624
+ du bloc. Cela ne se fait pas en mode mono\-threadé, la sortie encodée ne sera
625
+ donc pas identique à celle faite en mode multi\-threadé.
626
+ .TP
627
+ \fB\-\-flush\-timeout=\fP\fItemps_d'attente\fP
628
+ Lors de la compression, si plus que \fItemps_d'attente\fP millisecondes (un
629
+ entier positif) se sont écoulées depuis le précédent vidage et que lire plus
630
+ de données bloquerait, toutes les données d'entrée en attente sont vidées de
631
+ l'encodeur et mises à disposition dans le flux de sortie. Cela peut être
632
+ utile si \fBxz\fP est utilisé pour compresser les données qui sont diffusées
633
+ sur un réseau. Des petites valeurs de \fItemps_d'attente\fP rendent les données
634
+ disponibles à l'extrémité réceptrice avec un léger retard, mais les grandes
635
+ valeurs de \fItemps_d'attente\fP donnent un meilleur taux de compression.
636
+ .IP ""
637
+ Cette option est désactivée par défaut. Si cette option est indiquée plus
638
+ d'une fois, la dernière prend effet. La valeur spéciale de
639
+ \fItemps_d'attente\fP de \fB0\fP peut être utilisée pour explicitement désactiver
640
+ cette option.
641
+ .IP ""
642
+ Cette option n'est pas disponible sur les systèmes qui ne sont pas POSIX.
643
+ .IP ""
644
+ .\" FIXME
645
+ \fBCette option est encore expérimentale.\fP Actuellement, \fBxz\fP ne convient
646
+ pas pour décompresser le flux en temps réel en raison de la façon dont \fBxz\fP
647
+ effectue la mise en mémoire tampon.
648
+ .TP
649
+ \fB\-\-memlimit\-compress=\fP\fIlimite\fP
650
+ Indiquer une limite d'utilisation de la mémoire pour la compression. Si
651
+ cette option est indiquée plusieurs fois, c'est la dernière qui est prise en
652
+ compte.
653
+ .IP ""
654
+ If the compression settings exceed the \fIlimit\fP, \fBxz\fP will attempt to
655
+ adjust the settings downwards so that the limit is no longer exceeded and
656
+ display a notice that automatic adjustment was done. The adjustments are
657
+ done in this order: reducing the number of threads, switching to
658
+ single\-threaded mode if even one thread in multi\-threaded mode exceeds the
659
+ \fIlimit\fP, and finally reducing the LZMA2 dictionary size.
660
+ .IP ""
661
+ When compressing with \fB\-\-format=raw\fP or if \fB\-\-no\-adjust\fP has been
662
+ specified, only the number of threads may be reduced since it can be done
663
+ without affecting the compressed output.
664
+ .IP ""
665
+ If the \fIlimit\fP cannot be met even with the adjustments described above, an
666
+ error is displayed and \fBxz\fP will exit with exit status 1.
667
+ .IP ""
668
+ La \fIlimite\fP peut être indiquée de plusieurs façons :
669
+ .RS
670
+ .IP \(bu 3
671
+ La \fIlimite\fP peut être une valeur absolue en octets. Utiliser un suffixe
672
+ d'entier comme \fBMiB\fP peut être utile. Exemple :
673
+ \fB\-\-memlimit\-compress=80MiB\fP
674
+ .IP \(bu 3
675
+ La \fIlimite\fP peut être indiquée sous forme d'un pourcentage de la mémoire
676
+ physique totale (RAM). Cela peut être particulièrement utile quand la
677
+ variable d'environnement \fBXZ_DEFAULTS\fP est indiquée dans un script
678
+ d'initialisation de l'interpréteur partagé entre différents ordinateurs. De
679
+ cette façon la limite est automatiquement plus grande sur les systèmes avec
680
+ plus de mémoire. Exemple : \fB\-\-memlimit=70%\fP
681
+ .IP \(bu 3
682
+ The \fIlimit\fP can be reset back to its default value by setting it to \fB0\fP.
683
+ This is currently equivalent to setting the \fIlimit\fP to \fBmax\fP (no memory
684
+ usage limit).
685
+ .RE
686
+ .IP ""
687
+ For 32\-bit \fBxz\fP there is a special case: if the \fIlimit\fP would be over
688
+ \fB4020\ MiB\fP, the \fIlimit\fP is set to \fB4020\ MiB\fP. On MIPS32 \fB2000\ MiB\fP
689
+ is used instead. (The values \fB0\fP and \fBmax\fP aren't affected by this. A
690
+ similar feature doesn't exist for decompression.) This can be helpful when
691
+ a 32\-bit executable has access to 4\ GiB address space (2 GiB on MIPS32)
692
+ while hopefully doing no harm in other situations.
693
+ .IP ""
694
+ Voir aussi la section \fButilisation de la mémoire\fP.
695
+ .TP
696
+ \fB\-\-memlimit\-decompress=\fP\fIlimite\fP
697
+ Régler une limite d'utilisation de la mémoire pour la décompression. Cela a
698
+ un effet sur le mode \fB\-\-list\fP. Si l'opération n'est pas possible sans
699
+ dépasser la \fIlimite\fP, \fBxz\fP affichera une erreur et la décompression
700
+ échouera. Voir \fB\-\-memlimit\-compress=\fP\fIlimite\fP pour les manières possibles
701
+ d'indiquer la \fIlimite\fP.
702
+ .TP
703
+ \fB\-\-memlimit\-mt\-decompress=\fP\fIlimit\fP
704
+ Set a memory usage limit for multi\-threaded decompression. This can only
705
+ affect the number of threads; this will never make \fBxz\fP refuse to
706
+ decompress a file. If \fIlimit\fP is too low to allow any multi\-threading, the
707
+ \fIlimit\fP is ignored and \fBxz\fP will continue in single\-threaded mode. Note
708
+ that if also \fB\-\-memlimit\-decompress\fP is used, it will always apply to both
709
+ single\-threaded and multi\-threaded modes, and so the effective \fIlimit\fP for
710
+ multi\-threading will never be higher than the limit set with
711
+ \fB\-\-memlimit\-decompress\fP.
712
+ .IP ""
713
+ In contrast to the other memory usage limit options,
714
+ \fB\-\-memlimit\-mt\-decompress=\fP\fIlimit\fP has a system\-specific default
715
+ \fIlimit\fP. \fBxz \-\-info\-memory\fP can be used to see the current value.
716
+ .IP ""
717
+ This option and its default value exist because without any limit the
718
+ threaded decompressor could end up allocating an insane amount of memory
719
+ with some input files. If the default \fIlimit\fP is too low on your system,
720
+ feel free to increase the \fIlimit\fP but never set it to a value larger than
721
+ the amount of usable RAM as with appropriate input files \fBxz\fP will attempt
722
+ to use that amount of memory even with a low number of threads. Running out
723
+ of memory or swapping will not improve decompression performance.
724
+ .IP ""
725
+ See \fB\-\-memlimit\-compress=\fP\fIlimit\fP for possible ways to specify the
726
+ \fIlimit\fP. Setting \fIlimit\fP to \fB0\fP resets the \fIlimit\fP to the default
727
+ system\-specific value.
728
+ .TP
729
+ \fB\-M\fP \fIlimite\fP, \fB\-\-memlimit=\fP\fIlimite\fP, \fB\-\-memory=\fP\fIlimite\fP
730
+ This is equivalent to specifying \fB\-\-memlimit\-compress=\fP\fIlimit\fP
731
+ \fB\-\-memlimit\-decompress=\fP\fIlimit\fP \fB\-\-memlimit\-mt\-decompress=\fP\fIlimit\fP.
732
+ .TP
733
+ \fB\-\-no\-adjust\fP
734
+ Display an error and exit if the memory usage limit cannot be met without
735
+ adjusting settings that affect the compressed output. That is, this
736
+ prevents \fBxz\fP from switching the encoder from multi\-threaded mode to
737
+ single\-threaded mode and from reducing the LZMA2 dictionary size. Even when
738
+ this option is used the number of threads may be reduced to meet the memory
739
+ usage limit as that won't affect the compressed output.
740
+ .IP ""
741
+ Automatic adjusting is always disabled when creating raw streams
742
+ (\fB\-\-format=raw\fP).
743
+ .TP
744
+ \fB\-T\fP \fIthreads\fP, \fB\-\-threads=\fP\fIthreads\fP
745
+ Specify the number of worker threads to use. Setting \fIthreads\fP to a
746
+ special value \fB0\fP makes \fBxz\fP use up to as many threads as the processor(s)
747
+ on the system support. The actual number of threads can be fewer than
748
+ \fIthreads\fP if the input file is not big enough for threading with the given
749
+ settings or if using more threads would exceed the memory usage limit.
750
+ .IP ""
751
+ The single\-threaded and multi\-threaded compressors produce different
752
+ output. Single\-threaded compressor will give the smallest file size but
753
+ only the output from the multi\-threaded compressor can be decompressed using
754
+ multiple threads. Setting \fIthreads\fP to \fB1\fP will use the single\-threaded
755
+ mode. Setting \fIthreads\fP to any other value, including \fB0\fP, will use the
756
+ multi\-threaded compressor even if the system supports only one hardware
757
+ thread. (\fBxz\fP 5.2.x used single\-threaded mode in this situation.)
758
+ .IP ""
759
+ To use multi\-threaded mode with only one thread, set \fIthreads\fP to \fB+1\fP.
760
+ The \fB+\fP prefix has no effect with values other than \fB1\fP. A memory usage
761
+ limit can still make \fBxz\fP switch to single\-threaded mode unless
762
+ \fB\-\-no\-adjust\fP is used. Support for the \fB+\fP prefix was added in \fBxz\fP
763
+ 5.4.0.
764
+ .IP ""
765
+ If an automatic number of threads has been requested and no memory usage
766
+ limit has been specified, then a system\-specific default soft limit will be
767
+ used to possibly limit the number of threads. It is a soft limit in sense
768
+ that it is ignored if the number of threads becomes one, thus a soft limit
769
+ will never stop \fBxz\fP from compressing or decompressing. This default soft
770
+ limit will not make \fBxz\fP switch from multi\-threaded mode to single\-threaded
771
+ mode. The active limits can be seen with \fBxz \-\-info\-memory\fP.
772
+ .IP ""
773
+ Actuellement, la seule méthode de gestion avec des threads consiste à
774
+ séparer l'entrée en blocs et de les compresser indépendamment les uns des
775
+ autres. La taille par défaut des blocs dépend du niveau de compression et
776
+ peut\-être remplacée avec l'option \fB\-\-block\-size=\fP\fItaille\fP.
777
+ .IP ""
778
+ Threaded decompression only works on files that contain multiple blocks with
779
+ size information in block headers. All large enough files compressed in
780
+ multi\-threaded mode meet this condition, but files compressed in
781
+ single\-threaded mode don't even if \fB\-\-block\-size=\fP\fIsize\fP has been used.
782
+ .IP ""
783
+ The default value for \fIthreads\fP is \fB0\fP. In \fBxz\fP 5.4.x and older the
784
+ default is \fB1\fP.
785
+ .
786
+ .SS "Chaînes de filtres de compresseur personnalisées"
787
+ A custom filter chain allows specifying the compression settings in detail
788
+ instead of relying on the settings associated to the presets. When a custom
789
+ filter chain is specified, preset options (\fB\-0\fP \&...\& \fB\-9\fP and
790
+ \fB\-\-extreme\fP) earlier on the command line are forgotten. If a preset
791
+ option is specified after one or more custom filter chain options, the new
792
+ preset takes effect and the custom filter chain options specified earlier
793
+ are forgotten.
794
+ .PP
795
+ Une chaîne de filtre est comparable à une redirection (pipe) sur la ligne de
796
+ commande. Lors de la compression, les entrées non compressées vont au
797
+ premier filtre, dont la sortie va au prochain filtre (s'il y en a). La
798
+ sortie du dernier filtre est écrite sur le fichier compressé. Le nombre
799
+ maximal de filtres dans la chaîne est quatre, mais habituellement, un chaîne
800
+ de filtre n'a qu'un ou deux filtres.
801
+ .PP
802
+ Beaucoup de filtres ont des limitations sur l'endroit où ils peuvent se
803
+ placer dans la chaîne de filtre : quelques filtres ne peuvent fonctionner
804
+ qu'en tant que dernier filtre dans la chaîne, quelques uns en tant que non
805
+ dernier filtre, et d'autres à n'importe quelle position dans la
806
+ chaîne. Suivant le filtre, cette limitation est soit inhérente au profil du
807
+ filtre, soit existe pour des raisons de sécurité.
808
+ .PP
809
+ A custom filter chain can be specified in two different ways. The options
810
+ \fB\-\-filters=\fP\fIfilters\fP and \fB\-\-filters1=\fP\fIfilters\fP \&...\&
811
+ \fB\-\-filters9=\fP\fIfilters\fP allow specifying an entire filter chain in one
812
+ option using the liblzma filter string syntax. Alternatively, a filter
813
+ chain can be specified by using one or more individual filter options in the
814
+ order they are wanted in the filter chain. That is, the order of the
815
+ individual filter options is significant! When decoding raw streams
816
+ (\fB\-\-format=raw\fP), the filter chain must be specified in the same order as
817
+ it was specified when compressing. Any individual filter or preset options
818
+ specified before the full chain option (\fB\-\-filters=\fP\fIfilters\fP) will be
819
+ forgotten. Individual filters specified after the full chain option will
820
+ reset the filter chain.
821
+ .PP
822
+ Both the full and individual filter options take filter\-specific \fIoptions\fP
823
+ as a comma\-separated list. Extra commas in \fIoptions\fP are ignored. Every
824
+ option has a default value, so specify those you want to change.
825
+ .PP
826
+ Pour voir l'entièreté de la chaîne de filtres et ses \fIoptions\fP, utilisez
827
+ \fBxz \-vv\fP (ce qui est comme utiliser \fB\-\-verbose\fP deux fois). Cela
828
+ fonctionne aussi pour voir les options de chaîne de filtres utilisées par
829
+ les préréglages.
830
+ .TP
831
+ \fB\-\-filters=\fP\fIfilters\fP
832
+ Specify the full filter chain or a preset in a single option. Each filter
833
+ can be separated by spaces or two dashes (\fB\-\-\fP). \fIfilters\fP may need to be
834
+ quoted on the shell command line so it is parsed as a single option. To
835
+ denote \fIoptions\fP, use \fB:\fP or \fB=\fP. A preset can be prefixed with a \fB\-\fP
836
+ and followed with zero or more flags. The only supported flag is \fBe\fP to
837
+ apply the same options as \fB\-\-extreme\fP.
838
+ .TP
839
+ \fB\-\-filters1\fP=\fIfilters\fP ... \fB\-\-filters9\fP=\fIfilters\fP
840
+ Specify up to nine additional filter chains that can be used with
841
+ \fB\-\-block\-list\fP.
842
+ .IP ""
843
+ For example, when compressing an archive with executable files followed by
844
+ text files, the executable part could use a filter chain with a BCJ filter
845
+ and the text part only the LZMA2 filter.
846
+ .TP
847
+ \fB\-\-filters\-help\fP
848
+ Display a help message describing how to specify presets and custom filter
849
+ chains in the \fB\-\-filters\fP and \fB\-\-filters1=\fP\fIfilters\fP \&...\&
850
+ \fB\-\-filters9=\fP\fIfilters\fP options, and exit successfully.
851
+ .TP
852
+ \fB\-\-lzma1\fP[\fB=\fP\fIoptions\fP]
853
+ .PD 0
854
+ .TP
855
+ \fB\-\-lzma2\fP[\fB=\fP\fIoptions\fP]
856
+ .PD
857
+ Ajouter le filtre LZMA1 ou LZMA2 à la chaîne de filtres. Ces filtres ne
858
+ peuvent être utilisés que comme dernier filtre dans la chaîne.
859
+ .IP ""
860
+ LZMA1 est un filtre historique, qui n'est pris en charge presque uniquement
861
+ à cause de l'ancien format de fichier \fB.lzma\fP, qui ne prend en charge que
862
+ LZMA1. LZMA2 est une version mise à jour de LZMA1 pour régler certains
863
+ problèmes pratiques de LZMA1. Le format \fBxz\fP utilise LZMA2 et ne prend pas
864
+ du tout en charge LZMA1. Les taux et vitesses de compression de LZMA1 et
865
+ LZMA2 sont pratiquement identiques.
866
+ .IP ""
867
+ LZMA1 et LZMA2 partagent le même ensemble d'\fIoptions\fP :
868
+ .RS
869
+ .TP
870
+ \fBpreset=\fP\fIpréréglage\fP
871
+ Reset all LZMA1 or LZMA2 \fIoptions\fP to \fIpreset\fP. \fIPreset\fP consist of an
872
+ integer, which may be followed by single\-letter preset modifiers. The
873
+ integer can be from \fB0\fP to \fB9\fP, matching the command line options \fB\-0\fP
874
+ \&...\& \fB\-9\fP. The only supported modifier is currently \fBe\fP, which matches
875
+ \fB\-\-extreme\fP. If no \fBpreset\fP is specified, the default values of LZMA1 or
876
+ LZMA2 \fIoptions\fP are taken from the preset \fB6\fP.
877
+ .TP
878
+ \fBdict=\fP\fItaille\fP
879
+ La \fItaille\fP du dictionnaire (historique du tampon) indique combien d'octets
880
+ des données récement décompressées sont gardés en mémoire. L'algorithme
881
+ essaie de trouver les séquences d'octets répétées (identiques) dans les
882
+ données décompressées et les remplace par les données actuellement dans le
883
+ dictionnaire. Plus gros est le dictionnaire, plus grande est la chance de
884
+ trouver une correspondance. Ainsi, l'augmentation de la \fItaille\fP du
885
+ dictionnaire augmente habituellement le taux de compression, mais un
886
+ dictionnaire plus gros que le fichier non compressé est un gachis de
887
+ mémoire.
888
+ .IP ""
889
+ Généralement la \fItaille\fP du dictionnaire est entre 64\ Kio et 64\ Mio. Le
890
+ minimum étant 4\ Kio. La \fItaille\fP maximale pour la compression est
891
+ habituellement 1,5\ Gio (1536\ Mio). Le décompresseur prend en charge les
892
+ dictionnaires jusqu'à un octet de moins que 4\ Gio, ce qui est le maximum
893
+ pour les formats de flux LZMA1 et LZMA2.
894
+ .IP ""
895
+ La \fItaille\fP du dictionnaire et le chercheur de correspondance (match
896
+ finder) (\fImf\fP) déterminent ensemble l'utilisation de la mémoire de
897
+ l'encodeur LZMA1 ou LZMA2. La même (ou une plus grosse) \fItaille\fP de
898
+ dictionnaire est requise pour décompresser que ce qui a été utilisé pour la
899
+ compression, ainsi l'utilisation de la mémoire du décodeur est déterminée
900
+ par la taille du dictionnaire utilisée lors de la compression. Les en\-têtes
901
+ de \fB.xz\fP stockent la \fItaille\fP de dictionnaire sous la forme 2^\fIn\fP ou
902
+ 2^\fIn\fP + 2^(\fIn\fP\-1), de sorte que ces \fItailles\fP sont quelque peu préférées
903
+ pour la compression. Les autres \fItailles\fP seront arrondies à la hausse
904
+ lorsque stockées dans les en\-têtes de \fB.xz\fP.
905
+ .TP
906
+ \fBlc=\fP\fIlc\fP
907
+ Spécifiez le nombre d'octets de contexte littéraux. Le minimum est \fB0\fP et
908
+ le maximum est \fB4\fP. La valeur par défaut est \fB3\fP. En plus, la somme de
909
+ \fIlc\fP et \fIlp\fP ne doit pas excéder \fB4\fP.
910
+ .IP ""
911
+ Tous les octets qui ne peuvent pas être codés comme des correspondances sont
912
+ codés comme des littéraux. C'est à dire que les littéraux sont simplement
913
+ des octets 8 bits encodés un à la fois.
914
+ .IP ""
915
+ The literal coding makes an assumption that the highest \fIlc\fP bits of the
916
+ previous uncompressed byte correlate with the next byte. For example, in
917
+ typical English text, an upper\-case letter is often followed by a lower\-case
918
+ letter, and a lower\-case letter is usually followed by another lower\-case
919
+ letter. In the US\-ASCII character set, the highest three bits are 010 for
920
+ upper\-case letters and 011 for lower\-case letters. When \fIlc\fP is at least
921
+ 3, the literal coding can take advantage of this property in the
922
+ uncompressed data.
923
+ .IP ""
924
+ The default value (3) is usually good. If you want maximum compression,
925
+ test \fBlc=4\fP. Sometimes it helps a little, and sometimes it makes
926
+ compression worse. If it makes it worse, test \fBlc=2\fP too.
927
+ .TP
928
+ \fBlp=\fP\fIlp\fP
929
+ Indiquer le nombre de bits de position littérale. Le minimum est \fB0\fP et le
930
+ maximum \fB4\fP; par défaut c'est \fB0\fP.
931
+ .IP ""
932
+ \fILp\fP affecte le type d'alignement dans les données décompressées qui est
933
+ présumé lors de l'encodage des littéraux. Voir \fIpb\fP ci dessous pour plus
934
+ d'information sur l'alignement.
935
+ .TP
936
+ \fBpb=\fP\fIpb\fP
937
+ Indiquer le nombre de bits de position. Le minimum est \fB0\fP et le maximum
938
+ \fB4\fP; par défaut \fB2\fP.
939
+ .IP ""
940
+ \fIPb\fP affecte quel genre d'alignement est présumé en général dans les
941
+ données non compressées. Par défaut c'est un alignement de quatre octets
942
+ (2^\fIpb\fP=2^2=4), ce qui est généralement un bon choix lorsqu'il n'y a pas de
943
+ meilleure estimation.
944
+ .IP ""
945
+ When the alignment is known, setting \fIpb\fP accordingly may reduce the file
946
+ size a little. For example, with text files having one\-byte alignment
947
+ (US\-ASCII, ISO\-8859\-*, UTF\-8), setting \fBpb=0\fP can improve compression
948
+ slightly. For UTF\-16 text, \fBpb=1\fP is a good choice. If the alignment is
949
+ an odd number like 3 bytes, \fBpb=0\fP might be the best choice.
950
+ .IP ""
951
+ Même si l'alignement présumé peut être ajusté avec \fIpb\fP et \fIlp\fP, LZMA1 et
952
+ LZMA2 favorisent toujours légèrement l'alignement sur 16 octets. Il peut
953
+ être utile d'en tenir compte lors de la conception de formats de fichiers
954
+ susceptibles d'être souvent compressés avec LZMA1 ou LZMA2.
955
+ .TP
956
+ \fBmf=\fP\fImf\fP
957
+ Match finder has a major effect on encoder speed, memory usage, and
958
+ compression ratio. Usually Hash Chain match finders are faster than Binary
959
+ Tree match finders. The default depends on the \fIpreset\fP: 0 uses \fBhc3\fP,
960
+ 1\(en3 use \fBhc4\fP, and the rest use \fBbt4\fP.
961
+ .IP ""
962
+ Les chercheurs de correspondance suivants sont pris en charge. Les formules
963
+ d'utilisation de la mémoire ci\-dessous sont des approximations grossières
964
+ qui sont les plus proches de la réalité lorsque \fIdict\fP est une puissance de
965
+ deux.
966
+ .RS
967
+ .TP
968
+ \fBhc3\fP
969
+ Chaîne de hachage avec hachage de 2 et 3 octets
970
+ .br
971
+ Valeur minimale pour \fInice\fP : \fB3\fP
972
+ .br
973
+ Utilisation de la mémoire :
974
+ .br
975
+ \fIdict\fP * 7.5 (if \fIdict\fP <= 16 Mio);
976
+ .br
977
+ \fIdict\fP * 5.5 + 64 MiB (si \fIdict\fP > 16 Mio)
978
+ .TP
979
+ \fBhc4\fP
980
+ Chaîne de hachage avec hachage de 2, 3 et 4 octets
981
+ .br
982
+ Valeur minimale pour \fInice\fP : \fB4\fP
983
+ .br
984
+ Utilisation de la mémoire :
985
+ .br
986
+ \fIdict\fP * 7.5 (si \fIdict\fP <= 32 Mio);
987
+ .br
988
+ \fIdict\fP * 6.5 (si \fIdict\fP > 32 Mio)
989
+ .TP
990
+ \fBbt2\fP
991
+ Arbre binaire avec hachage de 2 octets
992
+ .br
993
+ Valeur minimale pour \fInice\fP : \fB2\fP
994
+ .br
995
+ Utilisation de la mémoire : \fIdict\fP * 9.5
996
+ .TP
997
+ \fBbt3\fP
998
+ Arbre binaire avec hachage de 2 et 3 octets
999
+ .br
1000
+ Valeur minimale pour \fInice\fP : \fB3\fP
1001
+ .br
1002
+ Utilisation de la mémoire :
1003
+ .br
1004
+ \fIdict\fP * 11.5 (si \fIdict\fP <= 16 Mio);
1005
+ .br
1006
+ \fIdict\fP * 9.5 + 64 MiB (si \fIdict\fP > 16 Mio)
1007
+ .TP
1008
+ \fBbt4\fP
1009
+ Arbre binaire avec hachage 2, 3 et 4 octets
1010
+ .br
1011
+ Valeur minimale pour \fInice\fP : \fB4\fP
1012
+ .br
1013
+ Utilisation de la mémoire :
1014
+ .br
1015
+ \fIdict\fP * 11.5 (si \fIdict\fP <= 32 Mio);
1016
+ .br
1017
+ \fIdict\fP * 10.5 (si \fIdict\fP > 32 Mio)
1018
+ .RE
1019
+ .TP
1020
+ \fBmode=\fP\fImode\fP
1021
+ Compression \fImode\fP specifies the method to analyze the data produced by the
1022
+ match finder. Supported \fImodes\fP are \fBfast\fP and \fBnormal\fP. The default is
1023
+ \fBfast\fP for \fIpresets\fP 0\(en3 and \fBnormal\fP for \fIpresets\fP 4\(en9.
1024
+ .IP ""
1025
+ Habituellement, \fBfast\fP est utilisé avec les chercheurs de correspondance de
1026
+ chaîne de hachage et \fBnormal\fP avec les chercheurs de correspondance d'arbre
1027
+ binaire. C'est aussi ce que font les \fIpréréglages\fP.
1028
+ .TP
1029
+ \fBnice=\fP\fInice\fP
1030
+ Spécifier ce qui est considéré comme une bonne longueur pour une
1031
+ correspondance. Une fois que la correspondance d'au moins \fInice\fP octets est
1032
+ trouvée, l'algorithme arrête de chercher de meilleures correspondances
1033
+ possibles.
1034
+ .IP ""
1035
+ \fINice\fP can be 2\(en273 bytes. Higher values tend to give better
1036
+ compression ratio at the expense of speed. The default depends on the
1037
+ \fIpreset\fP.
1038
+ .TP
1039
+ \fBdepth=\fP\fIprofondeur\fP
1040
+ Spécifier la profondeur de recherche maximale dans l'outil de recherche de
1041
+ correspondances. La valeur par défaut est \fB0\fP, ce qui fait que le
1042
+ compresseur détermine une \fIprofondeur\fP raisonnable en fonction de \fImf\fP et
1043
+ \fInice\fP.
1044
+ .IP ""
1045
+ Reasonable \fIdepth\fP for Hash Chains is 4\(en100 and 16\(en1000 for Binary
1046
+ Trees. Using very high values for \fIdepth\fP can make the encoder extremely
1047
+ slow with some files. Avoid setting the \fIdepth\fP over 1000 unless you are
1048
+ prepared to interrupt the compression in case it is taking far too long.
1049
+ .RE
1050
+ .IP ""
1051
+ Lors du décodage des flux bruts (\fB\-\-format=raw\fP), LZMA2 nécessite seulement
1052
+ la \fItaille\fP du dictionnaire. LZMA1 nécessite aussi \fIlc\fP, \fIlp\fP et \fIpb\fP.
1053
+ .TP
1054
+ \fB\-\-x86\fP[\fB=\fP\fIoptions\fP]
1055
+ .PD 0
1056
+ .TP
1057
+ \fB\-\-arm\fP[\fB=\fP\fIoptions\fP]
1058
+ .TP
1059
+ \fB\-\-armthumb\fP[\fB=\fP\fIoptions\fP]
1060
+ .TP
1061
+ \fB\-\-arm64\fP[\fB=\fP\fIoptions\fP]
1062
+ .TP
1063
+ \fB\-\-powerpc\fP[\fB=\fP\fIoptions\fP]
1064
+ .TP
1065
+ \fB\-\-ia64\fP[\fB=\fP\fIoptions\fP]
1066
+ .TP
1067
+ \fB\-\-sparc\fP[\fB=\fP\fIoptions\fP]
1068
+ .TP
1069
+ \fB\-\-riscv\fP[\fB=\fP\fIoptions\fP]
1070
+ .PD
1071
+ Ajouter un filtre branch/call/jump (BCJ) à la chaîne de filtres. Ces filtres
1072
+ ne peuvent être utilisés que s'ils ne sont pas le dernier filtre de la
1073
+ chaîne de filtrage.
1074
+ .IP ""
1075
+ A BCJ filter converts relative addresses in the machine code to their
1076
+ absolute counterparts. This doesn't change the size of the data but it
1077
+ increases redundancy, which can help LZMA2 to produce 0\(en15\ % smaller
1078
+ \&\fB.xz\fP file. The BCJ filters are always reversible, so using a BCJ filter
1079
+ for wrong type of data doesn't cause any data loss, although it may make the
1080
+ compression ratio slightly worse. The BCJ filters are very fast and use an
1081
+ insignificant amount of memory.
1082
+ .IP ""
1083
+ Ces filtres BCJ présentent des problèmes connus liés au taux de
1084
+ compression :
1085
+ .RS
1086
+ .IP \(bu 3
1087
+ Some types of files containing executable code (for example, object files,
1088
+ static libraries, and Linux kernel modules) have the addresses in the
1089
+ instructions filled with filler values. These BCJ filters will still do the
1090
+ address conversion, which will make the compression worse with these files.
1091
+ .IP \(bu 3
1092
+ If a BCJ filter is applied on an archive, it is possible that it makes the
1093
+ compression ratio worse than not using a BCJ filter. For example, if there
1094
+ are similar or even identical executables then filtering will likely make
1095
+ the files less similar and thus compression is worse. The contents of
1096
+ non\-executable files in the same archive can matter too. In practice one
1097
+ has to try with and without a BCJ filter to see which is better in each
1098
+ situation.
1099
+ .RE
1100
+ .IP ""
1101
+ Different instruction sets have different alignment: the executable file
1102
+ must be aligned to a multiple of this value in the input data to make the
1103
+ filter work.
1104
+ .RS
1105
+ .RS
1106
+ .PP
1107
+ .TS
1108
+ tab(;);
1109
+ l n l
1110
+ l n l.
1111
+ Filtre;Alignement;Notes
1112
+ x86;1;32 bits ou 64 bits x86
1113
+ ARM;4;
1114
+ ARM\-Thumb;2;
1115
+ ARM64;4;4096\-byte alignment is best
1116
+ PowerPC;4;Grand boutiste seulement
1117
+ IA\-64;16;Itanium
1118
+ SPARC;4;
1119
+ RISC\-V;2;
1120
+ .TE
1121
+ .RE
1122
+ .RE
1123
+ .IP ""
1124
+ Since the BCJ\-filtered data is usually compressed with LZMA2, the
1125
+ compression ratio may be improved slightly if the LZMA2 options are set to
1126
+ match the alignment of the selected BCJ filter. Examples:
1127
+ .RS
1128
+ .IP \(bu 3
1129
+ IA\-64 filter has 16\-byte alignment so \fBpb=4,lp=4,lc=0\fP is good with LZMA2
1130
+ (2^4=16).
1131
+ .IP \(bu 3
1132
+ RISC\-V code has 2\-byte or 4\-byte alignment depending on whether the file
1133
+ contains 16\-bit compressed instructions (the C extension). When 16\-bit
1134
+ instructions are used, \fBpb=2,lp=1,lc=3\fP or \fBpb=1,lp=1,lc=3\fP is good. When
1135
+ 16\-bit instructions aren't present, \fBpb=2,lp=2,lc=2\fP is the best.
1136
+ \fBreadelf \-h\fP can be used to check if "RVC" appears on the "Flags" line.
1137
+ .IP \(bu 3
1138
+ ARM64 is always 4\-byte aligned so \fBpb=2,lp=2,lc=2\fP is the best.
1139
+ .IP \(bu 3
1140
+ The x86 filter is an exception. It's usually good to stick to LZMA2's
1141
+ defaults (\fBpb=2,lp=0,lc=3\fP) when compressing x86 executables.
1142
+ .RE
1143
+ .IP ""
1144
+ Tous les filtres BCJ prennent en charge les mêmes \fIoptions\fP :
1145
+ .RS
1146
+ .TP
1147
+ \fBstart=\fP\fIdécalage\fP
1148
+ Spécifier le \fIdécalage\fP de départ qui est utilisé lors de la conversion
1149
+ entre les adresses relatives et absolues. Le \fIdécalage\fP doit être un
1150
+ multiple de l'alignement du filtre (voir la table ci\-dessus). Sa valeur par
1151
+ défaut est zéro. En pratique, cette dernière convient ; indiquer un
1152
+ \fIdécalage\fP personnalisé est la plupart du temps inutile.
1153
+ .RE
1154
+ .TP
1155
+ \fB\-\-delta\fP[\fB=\fP\fIoptions\fP]
1156
+ Ajouter le filtre Delta à la chaîne de filtres. Le filtre Delta ne peut être
1157
+ utilisé que s'il n'est pas le dernier filtre dans la chaîne.
1158
+ .IP ""
1159
+ Currently only simple byte\-wise delta calculation is supported. It can be
1160
+ useful when compressing, for example, uncompressed bitmap images or
1161
+ uncompressed PCM audio. However, special purpose algorithms may give
1162
+ significantly better results than Delta + LZMA2. This is true especially
1163
+ with audio, which compresses faster and better, for example, with
1164
+ \fBflac\fP(1).
1165
+ .IP ""
1166
+ \fIoptions\fP prises en charge :
1167
+ .RS
1168
+ .TP
1169
+ \fBdist=\fP\fIdistance\fP
1170
+ Specify the \fIdistance\fP of the delta calculation in bytes. \fIdistance\fP must
1171
+ be 1\(en256. The default is 1.
1172
+ .IP ""
1173
+ Par exemple, avec \fBdist=2\fP et une entrée huit octets A1 B1 A2 B3 A3 B5 A4
1174
+ B7, la sortie sera A1 B1 01 02 01 02 01 02.
1175
+ .RE
1176
+ .
1177
+ .SS "Autres options"
1178
+ .TP
1179
+ \fB\-q\fP, \fB\-\-quiet\fP
1180
+ Supprimer les avertissements et les notifications. Indiquer cela deux fois
1181
+ supprimera aussi les erreurs. Cette option n'a aucun effet sur le statut de
1182
+ sortie. Cela dit, même si un avertissement était supprimé, le statut de
1183
+ sortie indiquant un avertissement sera encore utilisé.
1184
+ .TP
1185
+ \fB\-v\fP, \fB\-\-verbose\fP
1186
+ Être bavard. Si l'erreur standard est connectée à un terminal, \fBxz\fP
1187
+ affichera une barre de progression. Indiquer \fB\-\-verbose\fP deux fois donnera
1188
+ une sortie encore plus bavarde.
1189
+ .IP ""
1190
+ La barre de progression montre l'information suivante :
1191
+ .RS
1192
+ .IP \(bu 3
1193
+ Le pourcentage de complétion est montré si la taille du fichier en entrée
1194
+ est connue. Néanmoins, le pourcentage ne peut pas être montré en cas de
1195
+ redirection.
1196
+ .IP \(bu 3
1197
+ Quantité de données compressées produites (compression) ou consommées
1198
+ (décompression).
1199
+ .IP \(bu 3
1200
+ Quantité de données non compressées consommées (compression) ou produites
1201
+ (décompression).
1202
+ .IP \(bu 3
1203
+ Le taux de compression, calculé en divisant la quantité de données
1204
+ compréssées déjà traitées par la quantité de données décompressées déjà
1205
+ traitées.
1206
+ .IP \(bu 3
1207
+ Vitesse de compression ou de décompression. Elle correspond à la quantité de
1208
+ données non compressées consommées (compression) ou produites
1209
+ (décompression) par seconde. Elle apparait quelques secondes après le début
1210
+ du traitement du fichier par \fBxz\fP.
1211
+ .IP \(bu 3
1212
+ Temps écoulé dans le format M:SS ou H:MM:SS.
1213
+ .IP \(bu 3
1214
+ Estimated remaining time is shown only when the size of the input file is
1215
+ known and a couple of seconds have already passed since \fBxz\fP started
1216
+ processing the file. The time is shown in a less precise format which never
1217
+ has any colons, for example, 2 min 30 s.
1218
+ .RE
1219
+ .IP ""
1220
+ When standard error is not a terminal, \fB\-\-verbose\fP will make \fBxz\fP print
1221
+ the filename, compressed size, uncompressed size, compression ratio, and
1222
+ possibly also the speed and elapsed time on a single line to standard error
1223
+ after compressing or decompressing the file. The speed and elapsed time are
1224
+ included only when the operation took at least a few seconds. If the
1225
+ operation didn't finish, for example, due to user interruption, also the
1226
+ completion percentage is printed if the size of the input file is known.
1227
+ .TP
1228
+ \fB\-Q\fP, \fB\-\-no\-warn\fP
1229
+ Ne pas mettre l'état de sortie à \fB2\fP même si une condition méritant un
1230
+ avertissement a été détectée. Cette option n'affecte pas le niveau de
1231
+ verbosité, néanmoins, les deux options \fB\-\-quiet\fP et \fB\-\-no\-warn\fP doivent
1232
+ être utilisées pour ne pas afficher d'avertissements, ni altérer le statut
1233
+ de sortie.
1234
+ .TP
1235
+ \fB\-\-robot\fP
1236
+ Afficher les messages dans un format analysable par une machine. Ceci est
1237
+ destiné à faciliter l'écriture des frontaux qui voudraient utiliser \fBxz\fP
1238
+ plutôt que liblzma, ce qui pourrait être le cas pour différents scripts. La
1239
+ sortie avec cette option activée est destinée à rester stable sur les
1240
+ différentes versions de \fBxz\fP. Consulter le paragraphe \fBROBOT MODE\fP pour
1241
+ les détails.
1242
+ .TP
1243
+ \fB\-\-info\-memory\fP
1244
+ Display, in human\-readable format, how much physical memory (RAM) and how
1245
+ many processor threads \fBxz\fP thinks the system has and the memory usage
1246
+ limits for compression and decompression, and exit successfully.
1247
+ .TP
1248
+ \fB\-h\fP, \fB\-\-help\fP
1249
+ Afficher un message d'aide décrivant les options les plus couramment
1250
+ utilisées et quitter.
1251
+ .TP
1252
+ \fB\-H\fP, \fB\-\-long\-help\fP
1253
+ Afficher un message d'aide décrivant toutes les options de \fBxz\fP et quitter.
1254
+ .TP
1255
+ \fB\-V\fP, \fB\-\-version\fP
1256
+ Afficher le numéro de version de \fBxz\fP et de liblzma dans un format lisible
1257
+ par un humain. Pour obtenir une sortie analysable par la machine, spécifiez
1258
+ \fB\-\-robot\fP avant \fB\-\-version\fP.
1259
+ .
1260
+ .SH "MODE ROBOT"
1261
+ The robot mode is activated with the \fB\-\-robot\fP option. It makes the output
1262
+ of \fBxz\fP easier to parse by other programs. Currently \fB\-\-robot\fP is
1263
+ supported only together with \fB\-\-list\fP, \fB\-\-filters\-help\fP, \fB\-\-info\-memory\fP,
1264
+ and \fB\-\-version\fP. It will be supported for compression and decompression in
1265
+ the future.
1266
+ .
1267
+ .SS "Mode liste"
1268
+ \fBxz \-\-robot \-\-list\fP utilise une sortie séparée par des tabulations. La
1269
+ première colonne de toutes les lignes possède une chaîne qui indique le type
1270
+ d'information trouvée sur cette ligne :
1271
+ .TP
1272
+ \fBname\fP
1273
+ C'est toujours la première ligne au début de la liste d'un fichier. La
1274
+ seconde colonne de la ligne est le nom de fichier.
1275
+ .TP
1276
+ \fBfile\fP
1277
+ Cette ligne contient l'information globale sur le fichier \fB.xz\fP. Cette
1278
+ ligne est toujours écrite après la ligne \fBname\fP.
1279
+ .TP
1280
+ \fBstream\fP
1281
+ Ce type de ligne n'est utilisée que lorsque \fB \-\-verbose\fP a été indiquée. Il
1282
+ y a autant de lignes \fBstream\fP qu'il y a de flux dans le fichier \fB.xz\fP.
1283
+ .TP
1284
+ \fBblock\fP
1285
+ Ce type de ligne n'est utilisé seulement lorsque \fB\-\-verbose\fP a été
1286
+ indiquée. Il y a autant de lignes \fBblock\fP qu'il y a de blocs dans le
1287
+ fichier \fB.xz\fP. Les lignes \fBblock\fP sont affichées après toutes les lignes
1288
+ \fBstream\fP ; les différents types de lignes ne sont pas imbriqués.
1289
+ .TP
1290
+ \fBsummary\fP
1291
+ Ce type de ligne n'est utilisé que lorsque \fB\-\-verbose\fP a été indiqué deux
1292
+ fois. Cette ligne est affichée après toutes les lignes \fBblock\fP. Comme la
1293
+ ligne \fBfile\fP, la ligne \fBsummary\fP contient l'information globale sur le
1294
+ fichier \fB.xz\fP.
1295
+ .TP
1296
+ \fBtotals\fP
1297
+ Cette ligne est toujours la toute dernière ligne de la sortie. Elle affiche
1298
+ les comptes et les tailles totaux.
1299
+ .PP
1300
+ Les colonnes des lignes \fBfile\fP :
1301
+ .PD 0
1302
+ .RS
1303
+ .IP 2. 4
1304
+ Nombre de flux dans le fichier
1305
+ .IP 3. 4
1306
+ Nombre total de blocs dans le ou les flux.
1307
+ .IP 4. 4
1308
+ Taille compressée du fichier
1309
+ .IP 5. 4
1310
+ Taille décompressée du fichier
1311
+ .IP 6. 4
1312
+ Compression ratio, for example, \fB0.123\fP. If ratio is over 9.999, three
1313
+ dashes (\fB\-\-\-\fP) are displayed instead of the ratio.
1314
+ .IP 7. 4
1315
+ Liste de noms de contrôles d'intégrité séparés par des virgules. Les chaînes
1316
+ suivantes sont utilisées pour les types de vérification connus : \fBNone\fP,
1317
+ \fBCRC32\fP, \fBCRC64\fP et \fBSHA256\fP. Pour le types de vérification inconnus,
1318
+ \fBUnknown\-\fP\fIN\fP est utilisé, où \fIN\fP est un identifiant de vérification sous
1319
+ la forme d'un nombre décimal (un ou deux chiffres).
1320
+ .IP 8. 4
1321
+ Taille totale du remplissage du flux dans le fichier
1322
+ .RE
1323
+ .PD
1324
+ .PP
1325
+ Les colonnes des lignes \fBstream\fP :
1326
+ .PD 0
1327
+ .RS
1328
+ .IP 2. 4
1329
+ Numéro de flux (le premier flux a le numéro 1)
1330
+ .IP 3. 4
1331
+ Nombre de blocs dans le flux
1332
+ .IP 4. 4
1333
+ Décalage de départ compressé
1334
+ .IP 5. 4
1335
+ Décalage de départ décompressé
1336
+ .IP 6. 4
1337
+ Taille compressée (ne comprend pas le remplissage du flux)
1338
+ .IP 7. 4
1339
+ Taille décompressée
1340
+ .IP 8. 4
1341
+ Taux de compression
1342
+ .IP 9. 4
1343
+ Nom de la vérification d'intégrité
1344
+ .IP 10. 4
1345
+ Taille du remplissage de flux
1346
+ .RE
1347
+ .PD
1348
+ .PP
1349
+ Les colonnes des lignes \fBblock\fP :
1350
+ .PD 0
1351
+ .RS
1352
+ .IP 2. 4
1353
+ Numéro du flux qui contient ce bloc
1354
+ .IP 3. 4
1355
+ Numéro du bloc relatif au commencement du flux (le premier bloc a pour
1356
+ numéro 1)
1357
+ .IP 4. 4
1358
+ Numéro du bloc relatif au début du fichier
1359
+ .IP 5. 4
1360
+ Décalage de départ compressé relatif au début du fichier
1361
+ .IP 6. 4
1362
+ Décalage de départ décompressé relatif au début du fichier
1363
+ .IP 7. 4
1364
+ Taille compressée totale du bloc (en\-têtes inclus)
1365
+ .IP 8. 4
1366
+ Taille décompressée
1367
+ .IP 9. 4
1368
+ Taux de compression
1369
+ .IP 10. 4
1370
+ Nom de la vérification d'intégrité
1371
+ .RE
1372
+ .PD
1373
+ .PP
1374
+ Si \fB\-\-verbose\fP a été indiqué deux fois, les colonnes additionnelles sont
1375
+ inclues sur les lignes \fBblock\fP. Elles ne sont pas affichées avec un seul
1376
+ \fB\-\-verbose\fP, car l'obtention de ces informations nécessite de nombreuses
1377
+ recherches et peut donc être lente :
1378
+ .PD 0
1379
+ .RS
1380
+ .IP 11. 4
1381
+ Valeur de la vérification d'intégrité en hexadécimal
1382
+ .IP 12. 4
1383
+ Taille d'en\-tête de bloc
1384
+ .IP 13. 4
1385
+ Drapeaux du bloc : \fBc\fP indique que la taille compressée est présente, et
1386
+ \fBu\fP indique que la taille décompréssée est présente. Si le drapeau n'est
1387
+ pas indiqué, un tiret (\fB\-\fP) est affiché à la place pour que la longueur de
1388
+ la chaîne reste fixe. De nouveaux drapeaux pourraient être ajoutés à la fin
1389
+ de la chaîne dans le futur.
1390
+ .IP 14. 4
1391
+ Taille des données effectivement compressées dans le bloc (en excluant
1392
+ l'en\-tête de bloc, le remplissage de bloc et les champs de vérification).
1393
+ .IP 15. 4
1394
+ Quantité de mémoire (en octets) nécessaire pour décompresser ce bloc avec
1395
+ cette version de \fBxz\fP.
1396
+ .IP 16. 4
1397
+ Chaîne de filtrage. Remarquez que la plupart des options utilisées au moment
1398
+ de la compression ne peuvent pas être connues, car seules les options
1399
+ nécessaires pour la décompression sont stockées dans les en\-têtes \fB.xz\fP.
1400
+ .RE
1401
+ .PD
1402
+ .PP
1403
+ Les colonnes des lignes \fBsummary\fP :
1404
+ .PD 0
1405
+ .RS
1406
+ .IP 2. 4
1407
+ Quantité de mémoire (en octets) nécessaire pour décompresser ce fichier avec
1408
+ cette version de \fBxz\fP.
1409
+ .IP 3. 4
1410
+ \fByes\fP ou \fBno\fP indique si tous les en\-têtes de bloc stockent à la fois la
1411
+ taille compressée et la taille décompressée.
1412
+ .PP
1413
+ \fIDepuis\fP \fBxz\fP \fI5.1.2alpha:\fP
1414
+ .IP 4. 4
1415
+ Version minimale de \fBxz\fP nécessaire pour décompresser le fichier.
1416
+ .RE
1417
+ .PD
1418
+ .PP
1419
+ Les colonnes de la ligne \fBtotals\fP :
1420
+ .PD 0
1421
+ .RS
1422
+ .IP 2. 4
1423
+ Nombre de flux
1424
+ .IP 3. 4
1425
+ Nombre de blocs
1426
+ .IP 4. 4
1427
+ Taille compressée
1428
+ .IP 5. 4
1429
+ Taille décompressée
1430
+ .IP 6. 4
1431
+ Taux de compression moyen
1432
+ .IP 7. 4
1433
+ Liste séparée par des virgules des noms de vérification d'intégrité qui
1434
+ étaient présents dans les fichiers
1435
+ .IP 8. 4
1436
+ Taille de remplissage de flux
1437
+ .IP 9. 4
1438
+ Nombre de fichiers. Permet de garder l'ordre des colonnes précédentes comme
1439
+ sur les lignes \fBfile\fP.
1440
+ .PD
1441
+ .RE
1442
+ .PP
1443
+ Si \fB\-\-verbose\fP a été indiqué deux fois, des colonnes supplémentaires sont
1444
+ incluses sur la ligne \fBtotals\fP :
1445
+ .PD 0
1446
+ .RS
1447
+ .IP 10. 4
1448
+ Quantité maximale de mémoire (en octets) nécessaire pour décompresser les
1449
+ fichiers avec cette version de \fBxz\fP.
1450
+ .IP 11. 4
1451
+ \fByes\fP ou \fBno\fP indique si tous les en\-têtes de bloc stockent à la fois la
1452
+ taille compressée et la taille décompressée.
1453
+ .PP
1454
+ \fIDepuis\fP \fBxz\fP \fI5.1.2alpha:\fP
1455
+ .IP 12. 4
1456
+ Version minimale de \fBxz\fP nécessaire pour décompresser le fichier.
1457
+ .RE
1458
+ .PD
1459
+ .PP
1460
+ Les versions futures pourront ajouter de nouveaux types de lignes et de
1461
+ nouvelles colonnes pourront être ajoutées aux types de lignes existants,
1462
+ mais les colonnes existantes ne seront pas modifiées.
1463
+ .
1464
+ .SS "Filters help"
1465
+ \fBxz \-\-robot \-\-filters\-help\fP prints the supported filters in the following
1466
+ format:
1467
+ .PP
1468
+ \fIfilter\fP\fB:\fP\fIoption\fP\fB=<\fP\fIvalue\fP\fB>,\fP\fIoption\fP\fB=<\fP\fIvalue\fP\fB>\fP...
1469
+ .TP
1470
+ \fIfilter\fP
1471
+ Name of the filter
1472
+ .TP
1473
+ \fIoption\fP
1474
+ Name of a filter specific option
1475
+ .TP
1476
+ \fIvalue\fP
1477
+ Numeric \fIvalue\fP ranges appear as \fB<\fP\fImin\fP\fB\-\fP\fImax\fP\fB>\fP. String
1478
+ \fIvalue\fP choices are shown within \fB< >\fP and separated by a \fB|\fP
1479
+ character.
1480
+ .PP
1481
+ Each filter is printed on its own line.
1482
+ .
1483
+ .SS "Information de limite de mémoire"
1484
+ \fBxz \-\-robot \-\-info\-memory\fP prints a single line with multiple tab\-separated
1485
+ columns:
1486
+ .IP 1. 4
1487
+ Total amount of physical memory (RAM) in bytes.
1488
+ .IP 2. 4
1489
+ Memory usage limit for compression in bytes (\fB\-\-memlimit\-compress\fP). A
1490
+ special value of \fB0\fP indicates the default setting which for
1491
+ single\-threaded mode is the same as no limit.
1492
+ .IP 3. 4
1493
+ Memory usage limit for decompression in bytes (\fB\-\-memlimit\-decompress\fP). A
1494
+ special value of \fB0\fP indicates the default setting which for
1495
+ single\-threaded mode is the same as no limit.
1496
+ .IP 4. 4
1497
+ Since \fBxz\fP 5.3.4alpha: Memory usage for multi\-threaded decompression in
1498
+ bytes (\fB\-\-memlimit\-mt\-decompress\fP). This is never zero because a
1499
+ system\-specific default value shown in the column 5 is used if no limit has
1500
+ been specified explicitly. This is also never greater than the value in the
1501
+ column 3 even if a larger value has been specified with
1502
+ \fB\-\-memlimit\-mt\-decompress\fP.
1503
+ .IP 5. 4
1504
+ Since \fBxz\fP 5.3.4alpha: A system\-specific default memory usage limit that is
1505
+ used to limit the number of threads when compressing with an automatic
1506
+ number of threads (\fB\-\-threads=0\fP) and no memory usage limit has been
1507
+ specified (\fB\-\-memlimit\-compress\fP). This is also used as the default value
1508
+ for \fB\-\-memlimit\-mt\-decompress\fP.
1509
+ .IP 6. 4
1510
+ Since \fBxz\fP 5.3.4alpha: Number of available processor threads.
1511
+ .PP
1512
+ Dans le futur, la sortie de \fBxz \-\-robot \-\-info\-memory\fP pourrait avoir plus
1513
+ de colonnes, mais jamais plus qu'une ligne unique.
1514
+ .
1515
+ .SS Version
1516
+ \fBxz \-\-robot \-\-version\fP prints the version number of \fBxz\fP and liblzma in
1517
+ the following format:
1518
+ .PP
1519
+ \fBXZ_VERSION=\fP\fIXYYYZZZS\fP
1520
+ .br
1521
+ \fBLIBLZMA_VERSION=\fP\fIXYYYZZZS\fP
1522
+ .TP
1523
+ \fIX\fP
1524
+ Version majeure.
1525
+ .TP
1526
+ \fIYYY\fP
1527
+ Version mineure. Les numéros pairs sont stables. Les numéros impairs sont
1528
+ des versions alpha ou beta.
1529
+ .TP
1530
+ \fIZZZ\fP
1531
+ Niveau de correctif pour les options stables ou juste un compteur pour les
1532
+ options de développement.
1533
+ .TP
1534
+ \fIS\fP
1535
+ Stabilité. 0 est alpha, 1 est bêta et 2 est stable. \fIS\fP devrait toujours
1536
+ être 2 quand \fIYYY\fP est pair.
1537
+ .PP
1538
+ \fIXYYYZZZS\fP sont identiques sur les deux lignes si \fBxz\fP et liblzma sont
1539
+ issus de la même version d'utilitaires XZ.
1540
+ .PP
1541
+ Exemples : 4.999.9beta est \fB49990091\fP et 5.0.0 est \fB50000002\fP.
1542
+ .
1543
+ .SH "STATUT DE SORTIE"
1544
+ .TP
1545
+ \fB0\fP
1546
+ Tout est bon.
1547
+ .TP
1548
+ \fB1\fP
1549
+ Une erreur est survenue.
1550
+ .TP
1551
+ \fB2\fP
1552
+ Quelquechose méritant un avertissement s'est produit, mais aucune erreur
1553
+ véritable n'est survenue.
1554
+ .PP
1555
+ Les notifications (pas les avertissements ou les erreurs) affichées sur
1556
+ l'erreur standard n'affectent pas le statut de sortie.
1557
+ .
1558
+ .SH ENVIRONNEMENT
1559
+ \fBxz\fP analyse les listes d'options séparées par des espaces à partir des
1560
+ variables d'environnement \fBXZ_DEFAULTS\fP et \fBXZ_OPT\fP, dans cet ordre, avant
1561
+ d'analyser les options de la ligne de commandes. Remarquez que seules les
1562
+ options sont analysées depuis l'environnement des variables ; toutes les
1563
+ non\-options sont ignorées silencieusement. L'analyse est faite avec
1564
+ \fBgetopt_long\fP(3) qui est aussi utilisé pour les arguments de la ligne de
1565
+ commandes.
1566
+ .TP
1567
+ \fBXZ_DEFAULTS\fP
1568
+ Options par défaut propres à l'utilisateur ou pour tout le système. Elles
1569
+ sont le plus souvent définies dans un script d'initialisation de
1570
+ l'interpréteur pour activer le limiteur d'utilisation de la mémoire de \fBxz\fP
1571
+ par défaut. A part pour les scripts d'initialisation de l'interpréteur ou
1572
+ des cas similaires, les sripts ne doivent jamais définir ou désactiver
1573
+ \fBXZ_DEFAULTS\fP.
1574
+ .TP
1575
+ \fBXZ_OPT\fP
1576
+ This is for passing options to \fBxz\fP when it is not possible to set the
1577
+ options directly on the \fBxz\fP command line. This is the case when \fBxz\fP is
1578
+ run by a script or tool, for example, GNU \fBtar\fP(1):
1579
+ .RS
1580
+ .RS
1581
+ .PP
1582
+ .nf
1583
+ \f(CRXZ_OPT=\-2v tar caf foo.tar.xz foo\fR
1584
+ .fi
1585
+ .RE
1586
+ .RE
1587
+ .IP ""
1588
+ Scripts may use \fBXZ_OPT\fP, for example, to set script\-specific default
1589
+ compression options. It is still recommended to allow users to override
1590
+ \fBXZ_OPT\fP if that is reasonable. For example, in \fBsh\fP(1) scripts one may
1591
+ use something like this:
1592
+ .RS
1593
+ .RS
1594
+ .PP
1595
+ .nf
1596
+ \f(CRXZ_OPT=${XZ_OPT\-"\-7e"}
1597
+ export XZ_OPT\fR
1598
+ .fi
1599
+ .RE
1600
+ .RE
1601
+ .
1602
+ .SH "Compatibilité des utilitaires LZMA"
1603
+ La syntaxe de la ligne de commande de \fBxz\fP est quasimment un sur\-ensemble
1604
+ de \fBlzma\fP, \fBunlzma\fP et \fBlzcat\fP comme ils sont trouvés dans les
1605
+ utilitaires LZMA 4.32.x . Dans la pluspart des cas, il est possible de
1606
+ remplacer les outils LZMA par les outils XZ sans casser les scripts
1607
+ existants. Il existe cependant certaines incompatibilités qui peuvent
1608
+ parfois poser des problèmes.
1609
+ .
1610
+ .SS "Niveaux de préréglage de la compression"
1611
+ La numérotation des préréglages de niveau de compression est différente
1612
+ entre les outils \fBxz\fP et LZMA. La différence la plus importante est la
1613
+ manière dont les tailles de dictionnaire sont affectées aux différents
1614
+ préréglages. La taille de dictionnaire est à peu près égale à celle
1615
+ d'utilisation de la mémoire de la décompression.
1616
+ .RS
1617
+ .PP
1618
+ .TS
1619
+ tab(;);
1620
+ c c c
1621
+ c n n.
1622
+ Niveau;xz;Utilitaires LZMA
1623
+ \-0;256 KiB;N/A
1624
+ \-1;1 MiB;64 KiB
1625
+ \-2;2 MiB;1 MiB
1626
+ \-3;4 MiB;512 KiB
1627
+ \-4;4 MiB;1 MiB
1628
+ \-5;8 MiB;2 MiB
1629
+ \-6;8 MiB;4 MiB
1630
+ \-7;16 MiB;8 MiB
1631
+ \-8;32 MiB;16 MiB
1632
+ \-9;64 MiB;32 MiB
1633
+ .TE
1634
+ .RE
1635
+ .PP
1636
+ Les différences de tailles des dictionnaires affectent aussi l'utilisation
1637
+ de la mémoire du compresseur, mais il y a quelques autres différences entre
1638
+ les outils LZMA et les outils XZ, qui rendent la différence encore plus
1639
+ grande :
1640
+ .RS
1641
+ .PP
1642
+ .TS
1643
+ tab(;);
1644
+ c c c
1645
+ c n n.
1646
+ Niveau;xz;Utilitaires LZMA 4.32.x
1647
+ \-0;3 MiB;N/A
1648
+ \-1;9 MiB;2 MiB
1649
+ \-2;17 MiB;12 MiB
1650
+ \-3;32 MiB;12 MiB
1651
+ \-4;48 MiB;16 MiB
1652
+ \-5;94 MiB;26 MiB
1653
+ \-6;94 MiB;45 MiB
1654
+ \-7;186 MiB;83 MiB
1655
+ \-8;370 MiB;159 MiB
1656
+ \-9;674 MiB;311 MiB
1657
+ .TE
1658
+ .RE
1659
+ .PP
1660
+ Le niveau de préréglage par défaut dans les outils LZMA est \fB\-7\fP alors que
1661
+ pour les outils XZ c'est \fB\-6\fP, les deux utilisent ainsi un dictionnaire de
1662
+ 8 Mio par défaut.
1663
+ .
1664
+ .SS "Fichiers .lzma en flux ou non"
1665
+ The uncompressed size of the file can be stored in the \fB.lzma\fP header.
1666
+ LZMA Utils does that when compressing regular files. The alternative is to
1667
+ mark that uncompressed size is unknown and use end\-of\-payload marker to
1668
+ indicate where the decompressor should stop. LZMA Utils uses this method
1669
+ when uncompressed size isn't known, which is the case, for example, in
1670
+ pipes.
1671
+ .PP
1672
+ \fBxz\fP prend en charge la décompression des fichiers \fB.lzma\fP avec ou sans
1673
+ marqueur de fin de charge utile, mais tous les fichiers \fB.lzma\fP créés par
1674
+ \fBxz\fP utiliseront un marqueur de fin de charge utile et ont la taille non
1675
+ compréssée marquée comme inconnue dans l'en\-tête \fB.lzma\fP. Cela peut être un
1676
+ problème dans quelques situations inhabituelles. Par exemple, un
1677
+ décompresseur \fB.lzma\fP dans un périphérique embarqué pourrait ne fonctionner
1678
+ qu'avec des fichiers dont la taille non comprimée est connue. Si vous vous
1679
+ heurtez à ce problème, vous devez utiliser les utilitaires LZMA ou LZMA SDK
1680
+ pour créer des fichiers \fB.lzma\fP avec une taille non compressée connue.
1681
+ .
1682
+ .SS "Fichiers .lzma non pris en charge"
1683
+ Le format \fB.lzma\fP autorise des valeurs \fIlc\fP jusqu'à 8, et des valeurs
1684
+ \fIlp\fP jusqu'à 4. Les outils LZMA peuvent décompresser des fichiers avec tous
1685
+ les \fIlc\fP et \fIlp\fP, mais créez toujours les fichiers avec \fBlc=3\fP et
1686
+ \fBlp=0\fP. Créer des fichiers avec d'autres valeurs \fIlc\fP et \fIlp\fP est
1687
+ possible avec \fBxz\fP et avec LZMA SDK.
1688
+ .PP
1689
+ L'implémentation du filtre LZMA1 dans liblzma nécessite que la somme de
1690
+ \fIlc\fP et \fIlp\fP soit inférieure ou égale à 4. Ainsi, les fichiers \fB.lzma\fP
1691
+ qui excèdent cette limitation ne peuvent pas être décompressés avec \fBxz\fP.
1692
+ .PP
1693
+ Les outils LZMA créent seulement des fichiers \fB.lzma\fP qui ont une taille de
1694
+ dictionnaire de 2^\fIn\fP (une puissance de 2) mais acceptent les fichiers avec
1695
+ toutes les tailles de dictionnaire. Libzlma n'accepte que les fichiers
1696
+ \&\fB.lzma\fP qui ont une taille dictionnaire de 2^\fIn\fP ou
1697
+ 2^\fIn\fP + 2^(\fIn\fP\-1). Cela afin de diminuer les faux positifs lors de la
1698
+ détection des fichiers \fB.lzma\fP.
1699
+ .PP
1700
+ Ces limitations ne devraient pas poser problème en pratique, car
1701
+ pratiquement tous les fichiers \fB.lzma\fP ont été compressés avec des réglages
1702
+ que liblzma accepte.
1703
+ .
1704
+ .SS "Déchets excédentaires"
1705
+ Lors de la décompession, l'utilitaire LZMA ignore silencieusement tout ce
1706
+ qui est après le premier flux \fB.lzma\fP. Dans la majorité des situations,
1707
+ c'est un bogue. Cela veut dire aussi que les outils LZMA ne gèrent pas la
1708
+ décompression de fichiers \fB.lzma\fP concaténés.
1709
+ .PP
1710
+ S'il reste des données après le premier flux \fB.lzma\fP, \fBxz\fP considère que
1711
+ le fichier est corrompu sauf si \fB\-\-single\-stream\fP a été utilisé. Cela peut
1712
+ casser des scripts obscurs qui ont supposé que les déchets de fin de ligne
1713
+ sont ignorés.
1714
+ .
1715
+ .SH NOTES
1716
+ .
1717
+ .SS "La sortie compressée peut varier"
1718
+ La sortie compressée exacte produite par les même fichiers non compressés en
1719
+ entrée peut varier en fonction des différentes versions de l'utilitaire XZ,
1720
+ même si les options de compression sont identiques. En effet, il est
1721
+ possible d'améliorer l'encodeur (compression plus rapide ou meilleure) sans
1722
+ affecter le format du fichier. La sortie peut même varier entre différentes
1723
+ compilations de la même version d'utilitaire XZ, si des options de
1724
+ construction différentes sont utilisées.
1725
+ .PP
1726
+ Cela signifie qu'une fois que \fB\-\-rsyncable\fP a été implémenté, les fichiers
1727
+ résultants ne seront pas nécessairement synchronisables avec rsync à moins
1728
+ que les nouveaux et anciens fichiers n'aient été compressés avec la même
1729
+ version de xz. Ce problème peut être résolu si une partie de
1730
+ l'implémentation est gelée pour garantir la stabilité de la sortie rsyncable
1731
+ à travers les versions de xz.
1732
+ .
1733
+ .SS "Décompresseurs .xz embarqués"
1734
+ Les implémentations de décompresseur embarqué comme XZ Embedded ne gèrent
1735
+ pas nécessairement les fichiers créés avec d'autres types de \fIvérification\fP
1736
+ d'intégrité que \fBnone\fP et \fBCRC32\fP. Comme la valeur par défaut est
1737
+ \fB\-\-check=crc64\fP, vous devez utiliser \fB\-\-check=none\fP ou \fB\-\-check=crc32\fP
1738
+ lors de la création de fichiers pour les systèmes embarqués.
1739
+ .PP
1740
+ En dehors des systèmes embarqués, tous les décompresseurs de format \fB.xz\fP
1741
+ gèrent tous les types de \fIvérification\fP ou sont au moins capables de
1742
+ décompresser le fichier sans effectuer la vérification d'intégrité si ce
1743
+ type de \fIvérification\fP particulière n'est pas pris en charge.
1744
+ .PP
1745
+ XZ Embedded prend en charge les filtres BCJ, mais seulement avec le décalage
1746
+ de départ par défaut.
1747
+ .
1748
+ .SH EXEMPLES
1749
+ .
1750
+ .SS Bases
1751
+ Compresser le fichier \fItoto\fP en \fItoto.xz\fP en utilisant le niveau de
1752
+ compression par défaut (\fB\-6\fP) et supprimer \fItoto\fP si la compression
1753
+ réussit :
1754
+ .RS
1755
+ .PP
1756
+ .nf
1757
+ \f(CRxz foo\fR
1758
+ .fi
1759
+ .RE
1760
+ .PP
1761
+ Décompresser \fIbidule.xz\fP en \fIbidule\fP et ne pas supprimer \fIbidule.xz\fP même
1762
+ si la compression réussit :
1763
+ .RS
1764
+ .PP
1765
+ .nf
1766
+ \f(CRxz \-dk bar.xz\fR
1767
+ .fi
1768
+ .RE
1769
+ .PP
1770
+ Create \fIbaz.tar.xz\fP with the preset \fB\-4e\fP (\fB\-4 \-\-extreme\fP), which is
1771
+ slower than the default \fB\-6\fP, but needs less memory for compression and
1772
+ decompression (48\ MiB and 5\ MiB, respectively):
1773
+ .RS
1774
+ .PP
1775
+ .nf
1776
+ \f(CRtar cf \- baz | xz \-4e > baz.tar.xz\fR
1777
+ .fi
1778
+ .RE
1779
+ .PP
1780
+ Un mélange de fichiers compressés et non compressés peuvent être
1781
+ décompressés vers la sortie standard avec une simple commande :
1782
+ .RS
1783
+ .PP
1784
+ .nf
1785
+ \f(CRxz \-dcf a.txt b.txt.xz c.txt d.txt.lzma > abcd.txt\fR
1786
+ .fi
1787
+ .RE
1788
+ .
1789
+ .SS "Compression en parallèle de plusieurs fichiers"
1790
+ Sur GNU et *BSD, \fBfind\fP(1) et \fBxargs\fP(1) peuvent être utilisés pour mettre
1791
+ en parallèle la compression de plusieurs fichiers :
1792
+ .RS
1793
+ .PP
1794
+ .nf
1795
+ \f(CRfind . \-type f \e! \-name '*.xz' \-print0 \e
1796
+ | xargs \-0r \-P4 \-n16 xz \-T1\fR
1797
+ .fi
1798
+ .RE
1799
+ .PP
1800
+ L'option \fBP\fP passée à \fBxargs\fP(1) fixe le nombre de processus \fBxz\fP en
1801
+ parallèles. La meilleure valeur pour l'option \fBn\fP dépend du nombre de
1802
+ fichiers à compresser. S\-il n'y a que quelques fichiers, la valeur sera
1803
+ probablement 1 ; avec des dizaines de milliers de fichiers, 100 ou même plus
1804
+ serait approprié pour réduire le nombre de processus \fBxz\fP que \fBxargs\fP(1)
1805
+ créera éventuellement.
1806
+ .PP
1807
+ L'option \fB\-T1\fP de \fBxz\fP est là pour le forcer en mode mono\-thread, car
1808
+ \fBxargs\fP(1) est utilisé pour contrôler la quantité de mise en parallèle.
1809
+ .
1810
+ .SS "Mode robot"
1811
+ Calculer combien d'octets ont été économisés au total après avoir compressé
1812
+ plusieurs fichiers :
1813
+ .RS
1814
+ .PP
1815
+ .nf
1816
+ \f(CRxz \-\-robot \-\-list *.xz | awk '/^totals/{print $5\-$4}'\fR
1817
+ .fi
1818
+ .RE
1819
+ .PP
1820
+ Un script peut vouloir savoir qu'il utilise une version suffisamment récente
1821
+ de \fBxz\fP. Le script \fBsh\fP(1) suivant vérifie que le numéro de version de
1822
+ l'outil \fBxz\fP soit au minimum 5.0.0. Cette méthode est compatible avec les
1823
+ vieilles versions bêta, qui ne gèrent pas l'option \fB\-\-robot\fP :
1824
+ .RS
1825
+ .PP
1826
+ .nf
1827
+ \f(CRif ! eval "$(xz \-\-robot \-\-version 2> /dev/null)" ||
1828
+ [ "$XZ_VERSION" \-lt 50000002 ]; then
1829
+ echo "Your xz is too old."
1830
+ fi
1831
+ unset XZ_VERSION LIBLZMA_VERSION\fR
1832
+ .fi
1833
+ .RE
1834
+ .PP
1835
+ Régler une limite d'utilisation de la mémoire pour la décompression en
1836
+ utilisant \fBXZ_OPT\fP, mais si une limite a déjà été définie, ne pas
1837
+ l'augmenter :
1838
+ .RS
1839
+ .PP
1840
+ .nf
1841
+ \f(CRNEWLIM=$((123 << 20))\ \ # 123 MiB
1842
+ OLDLIM=$(xz \-\-robot \-\-info\-memory | cut \-f3)
1843
+ if [ $OLDLIM \-eq 0 \-o $OLDLIM \-gt $NEWLIM ]; then
1844
+ XZ_OPT="$XZ_OPT \-\-memlimit\-decompress=$NEWLIM"
1845
+ export XZ_OPT
1846
+ fi\fR
1847
+ .fi
1848
+ .RE
1849
+ .
1850
+ .SS "Chaînes de filtres de compresseur personnalisées"
1851
+ L'utilisation la plus simple des chaînes de filtres personnalisées est la
1852
+ personnalisation d'un préréglage LZMA2. Cela peut être utile, car les
1853
+ préréglages ne couvrent qu'un sous\-ensemble des réglages de compression
1854
+ potentiellement utiles.
1855
+ .PP
1856
+ Les colonnes CompCPU des tableaux des descriptions des options \fB\-0\fP à \fB\-9\fP
1857
+ et \fB\-\-extreme\fP sont utiles lors de la personnalisation des préréglages
1858
+ LZMA2. Voici les parties pertinentes recueillies à partir de ces deux
1859
+ tableaux :
1860
+ .RS
1861
+ .PP
1862
+ .TS
1863
+ tab(;);
1864
+ c c
1865
+ n n.
1866
+ Préréglage;CompCPU
1867
+ \-0;0
1868
+ \-1;1
1869
+ \-2;2
1870
+ \-3;3
1871
+ \-4;4
1872
+ \-5;5
1873
+ \-6;6
1874
+ \-5e;7
1875
+ \-6e;8
1876
+ .TE
1877
+ .RE
1878
+ .PP
1879
+ If you know that a file requires somewhat big dictionary (for example, 32\ MiB) to compress well, but you want to compress it quicker than \fBxz \-8\fP
1880
+ would do, a preset with a low CompCPU value (for example, 1) can be
1881
+ modified to use a bigger dictionary:
1882
+ .RS
1883
+ .PP
1884
+ .nf
1885
+ \f(CRxz \-\-lzma2=preset=1,dict=32MiB foo.tar\fR
1886
+ .fi
1887
+ .RE
1888
+ .PP
1889
+ Avec certains fichiers, la commande ci\-dessus peut être plus rapide que
1890
+ \fBxz\-6\fP tout en compressant bien mieux. Cependant, il faut souligner que
1891
+ seuls certains fichiers bénéficient d'un grand dictionnaire tout en gardant
1892
+ la valeur de CompCPU faible. La siutation la plus évidente où un gros
1893
+ dictionnaire peut baucoup aider, est une archive contenant des fichiers très
1894
+ similaires de quelques megaoctets chacun. La taille de dictionnaire doit
1895
+ être significativement plus grosse que tout fichier individuel pour
1896
+ permettre à LZMA2 de tirer pleinement partie des similarités entre des
1897
+ fichiers consécutifs.
1898
+ .PP
1899
+ Si une utilisation de la mémoire élevée pour la compression et décompression
1900
+ convient, et que le fichier à compresser a une taille de plusieurs centaines
1901
+ de megaoctets, il peut être utile d'utiliser un plus gros dictionnaire que
1902
+ celui fourni par \fBxz\-9\fP (64 Mio) :
1903
+ .RS
1904
+ .PP
1905
+ .nf
1906
+ \f(CRxz \-vv \-\-lzma2=dict=192MiB big_foo.tar\fR
1907
+ .fi
1908
+ .RE
1909
+ .PP
1910
+ Utiliser \fB\-vv\fP (\fB\-\-verbose\-\-verbose\fP) comme dans l'exemple ci\-dessus peut
1911
+ être utile pour voir les besoins en mémoire du compresseur et du
1912
+ décompresseur. Rappelez\-vous qu'utiliser un dictionnaire plus gros que la
1913
+ taille du fichier non compressé est un gachis de mémoire, donc la commande
1914
+ ci\-dessus n'est pas utile pour les petits fichiers.
1915
+ .PP
1916
+ Sometimes the compression time doesn't matter, but the decompressor memory
1917
+ usage has to be kept low, for example, to make it possible to decompress the
1918
+ file on an embedded system. The following command uses \fB\-6e\fP (\fB\-6 \-\-extreme\fP) as a base and sets the dictionary to only 64\ KiB. The
1919
+ resulting file can be decompressed with XZ Embedded (that's why there is
1920
+ \fB\-\-check=crc32\fP) using about 100\ KiB of memory.
1921
+ .RS
1922
+ .PP
1923
+ .nf
1924
+ \f(CRxz \-\-check=crc32 \-\-lzma2=preset=6e,dict=64KiB foo\fR
1925
+ .fi
1926
+ .RE
1927
+ .PP
1928
+ If you want to squeeze out as many bytes as possible, adjusting the number
1929
+ of literal context bits (\fIlc\fP) and number of position bits (\fIpb\fP) can
1930
+ sometimes help. Adjusting the number of literal position bits (\fIlp\fP)
1931
+ might help too, but usually \fIlc\fP and \fIpb\fP are more important. For
1932
+ example, a source code archive contains mostly US\-ASCII text, so something
1933
+ like the following might give slightly (like 0.1\ %) smaller file than \fBxz \-6e\fP (try also without \fBlc=4\fP):
1934
+ .RS
1935
+ .PP
1936
+ .nf
1937
+ \f(CRxz \-\-lzma2=preset=6e,pb=0,lc=4 source_code.tar\fR
1938
+ .fi
1939
+ .RE
1940
+ .PP
1941
+ Using another filter together with LZMA2 can improve compression with
1942
+ certain file types. For example, to compress a x86\-32 or x86\-64 shared
1943
+ library using the x86 BCJ filter:
1944
+ .RS
1945
+ .PP
1946
+ .nf
1947
+ \f(CRxz \-\-x86 \-\-lzma2 libfoo.so\fR
1948
+ .fi
1949
+ .RE
1950
+ .PP
1951
+ Notez que l'ordre des options de filtre est significatif. Si \fB\-\-x86\fP est
1952
+ indiqué après \fB\-\-lzma2\fP, \fBxz\fP donnera une erreur, car il ne peut y avoir
1953
+ aucun filtre après LZMA2, et aussi parce que le filtre BCJ x86 ne peut pas
1954
+ être utilisé comme dernier filtre dans la chaîne.
1955
+ .PP
1956
+ Le filtre Delta associé à LZMA2 peut donner de bons résultats avec les
1957
+ images bitmap. Cela devrait habituellement battre PNG, qui a quelques
1958
+ filtres avancés supplémentaires qu'un simple delta, mais qui utilise Deflate
1959
+ pour la compression effective.
1960
+ .PP
1961
+ The image has to be saved in uncompressed format, for example, as
1962
+ uncompressed TIFF. The distance parameter of the Delta filter is set to
1963
+ match the number of bytes per pixel in the image. For example, 24\-bit RGB
1964
+ bitmap needs \fBdist=3\fP, and it is also good to pass \fBpb=0\fP to LZMA2 to
1965
+ accommodate the three\-byte alignment:
1966
+ .RS
1967
+ .PP
1968
+ .nf
1969
+ \f(CRxz \-\-delta=dist=3 \-\-lzma2=pb=0 foo.tiff\fR
1970
+ .fi
1971
+ .RE
1972
+ .PP
1973
+ If multiple images have been put into a single archive (for example,
1974
+ \&\fB.tar\fP), the Delta filter will work on that too as long as all images have
1975
+ the same number of bytes per pixel.
1976
+ .
1977
+ .SH "VOIR AUSSI"
1978
+ \fBxzdec\fP(1), \fBxzdiff\fP(1), \fBxzgrep\fP(1), \fBxzless\fP(1), \fBxzmore\fP(1),
1979
+ \fBgzip\fP(1), \fBbzip2\fP(1), \fB7z\fP(1)
1980
+ .PP
1981
+ XZ Utils: <https://tukaani.org/xz/>
1982
+ .br
1983
+ XZ Embedded: <https://tukaani.org/xz/embedded.html>
1984
+ .br
1985
+ LZMA SDK: <https://7\-zip.org/sdk.html>
janus/share/man/fr/man1/xzcat.1 ADDED
@@ -0,0 +1,1985 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ '\" t
2
+ .\" SPDX-License-Identifier: 0BSD
3
+ .\"
4
+ .\" Authors: Lasse Collin
5
+ .\" Jia Tan
6
+ .\"
7
+ .\" french translation of XZ Utils man
8
+ .\" Copyright (C) 2021 Debian French l10n team <debian-l10n-french@lists.debian.org>
9
+ .\" Translator
10
+ .\" bubu <bubub@no-log.org>, 2021.
11
+ .\"
12
+ .\"*******************************************************************
13
+ .\"
14
+ .\" This file was generated with po4a. Translate the source file.
15
+ .\"
16
+ .\"*******************************************************************
17
+ .TH XZ 1 2024\-12\-30 Tukaani "Utilitaires XZ"
18
+ .
19
+ .SH NOM
20
+ xz, unxz, xzcat, lzma, unlzma, lzcat \- Compresser ou décompresser des
21
+ fichiers .xz et .lzma
22
+ .
23
+ .SH SYNOPSIS
24
+ \fBxz\fP [\fIoption...\fP] [\fIfichier...\fP]
25
+ .
26
+ .SH "ALIAS DES COMMANDES"
27
+ \fBunxz\fP est équivalent à \fBxz \-\-decompress\fP.
28
+ .br
29
+ \fBxzcat\fP est équivalent à \fBxz \-\-decompress \-\-stdout\fP
30
+ .br
31
+ \fBlzma\fP est équivalent à \fBxz \-\-format=lzma\fP
32
+ .br
33
+ \fBunlzma\fP est équivalent à \fBxz \-\-format=lzma \-\-decompress\fP
34
+ .br
35
+ \fBlzcat\fP est équivalent à \fBxz \-\-format=lzma \-\-decompress \-\- stdout\fP
36
+ .PP
37
+ Lors de l'écriture de scripts qui nécessitent de décompresser des fichiers,
38
+ il est recommandé de toujours utiliser la commande \fBxz\fP avec les arguments
39
+ appropriés (\fBxz \-d\fP ou \fBxz \-dc\fP) au lieu des commandes \fBunxz\fP et
40
+ \fBxzcat\fP.
41
+ .
42
+ .SH DESCRIPTION
43
+ \fBxz\fP is a general\-purpose data compression tool with command line syntax
44
+ similar to \fBgzip\fP(1) and \fBbzip2\fP(1). The native file format is the
45
+ \&\fB.xz\fP format, but the legacy \fB.lzma\fP format used by LZMA Utils and raw
46
+ compressed streams with no container format headers are also supported. In
47
+ addition, decompression of the \fB.lz\fP format used by \fBlzip\fP is supported.
48
+ .PP
49
+ \fBxz\fP compresse ou décompresse chaque \fIfichier\fP en fonction du mode
50
+ d'opération choisi. Si aucun \fIfichier\fP n'est donné ou \fIfichier\fP est \fB\-\fP,
51
+ \fBxz\fP lit depuis l'entrée standard et écrit les données traitées sur la
52
+ sortie standard. \fBxz\fP refusera (affichera une erreur et ignorera le
53
+ \fIfichier\fP) d'écrire les données compressées sur la sortie standard si c'est
54
+ un terminal. De même, \fBxz\fP refusera de lire des données compressées depuis
55
+ l'entrée standard si c'est un terminal.
56
+ .PP
57
+ A moins que \fB\-\-sdout\fP ne soit indiqué, les \fIfichiers\fP autres que \fB\-\fP sont
58
+ écrits dans un nouveau fichier dont le nom est dérivé du nom de \fIfichier\fP
59
+ source :
60
+ .IP \(bu 3
61
+ Lors de la compression, le suffixe du format de fichier cible (\fB.xz\fP ou
62
+ \&\fB.lzma\fP) est ajouté au nom de fichier source pour obtenir le nom de fichier
63
+ cible.
64
+ .IP \(bu 3
65
+ When decompressing, the \fB.xz\fP, \fB.lzma\fP, or \fB.lz\fP suffix is removed from
66
+ the filename to get the target filename. \fBxz\fP also recognizes the suffixes
67
+ \&\fB.txz\fP and \fB.tlz\fP, and replaces them with the \fB.tar\fP suffix.
68
+ .PP
69
+ Si le fichier cible existe déjà, une erreur est affichée et le \fIfichier\fP
70
+ est ignoré.
71
+ .PP
72
+ Sauf s'il écrit dans la sortie standard, \fBxz\fP affichera un avertissement et
73
+ ignorera le \fIfichier\fP dans les cas suivants :
74
+ .IP \(bu 3
75
+ \fIfichier\fP n'est pas un fichier normal. Les liens symboliques ne sont pas
76
+ suivis et donc ne sont pas considérés comme des fichiers normaux.
77
+ .IP \(bu 3
78
+ \fIfichier\fP a plusieurs liens physiques.
79
+ .IP \(bu 3
80
+ \fIfichier\fP a un setuid, setgid ou sticky bit positionné.
81
+ .IP \(bu 3
82
+ Le mode d'opération est défini pour compresser et le \fIfichier\fP a déjà un
83
+ suffixe du format de fichier cible (\fB.xz\fP ou \fB.txz\fP lors d'une compression
84
+ en format \fB.xz\fP, et \fB.lzma\fP ou \fB.tlz\fP lors d'une compression en format
85
+ \&\fB.lzma\fP).
86
+ .IP \(bu 3
87
+ The operation mode is set to decompress and the \fIfile\fP doesn't have a
88
+ suffix of any of the supported file formats (\fB.xz\fP, \fB.txz\fP, \fB.lzma\fP,
89
+ \&\fB.tlz\fP, or \fB.lz\fP).
90
+ .PP
91
+ Après la compression ou la décompression réussie du \fIfichier\fP, \fBxz\fP copie
92
+ les permissions du propriétaire, du groupe, la date d'accès, et les
93
+ modifications d'heure depuis le \fIfichier\fP source du fichier cible. Si la
94
+ copie du groupe échoue, les permissions sont modifiées pour que le fichier
95
+ cible ne soit pas accessible aux utilisateurs qui n'ont pas les droits
96
+ d'accès au \fIfichier\fP source. \fBxz\fP ne prend actuellement pas en charge la
97
+ copie d'autres métadonnées telles que les listes de contrôle d'accès ou les
98
+ attributs étendus.
99
+ .PP
100
+ Once the target file has been successfully closed, the source \fIfile\fP is
101
+ removed unless \fB\-\-keep\fP was specified. The source \fIfile\fP is never removed
102
+ if the output is written to standard output or if an error occurs.
103
+ .PP
104
+ Envoyer \fBSIGINFO\fP ou \fBSIGURSR1\fP au processus \fBxz\fP, lui fait afficher
105
+ l'information de progression sur l'erreur standard. Cela a un intérêt limité
106
+ car lorsque l'erreur standard est un terminal, utiliser \fB\-\-verbose\fP
107
+ affichera automatiquement un indicateur de progression du processus.
108
+ .
109
+ .SS "Utilisation de la mémoire"
110
+ L'utilisation de la mémoire par \fBxz\fP varie de quelques centaines de
111
+ kilo\-octets à plusieurs gigaoctects en fonction des paramètres de
112
+ compression. Les réglages utilisés lors de la compression d'un fichier
113
+ déterminent les besoins en mémoire pour la décompression. Habituellement la
114
+ décompression nécessite 5\% à 20\% de la quantité de mémoire utilisée pour
115
+ la compression du fichier. Par exemple, décompresser un fichier créé avec
116
+ \fBxz\-9\fP recquiert habituellement 65\ Mio de mémoire. Bien qu'il soit
117
+ possible d'avoir des fichiers \fB.xz\fP nécessitant plusieurs gigaoctets de
118
+ mémoire pour être décompressés.
119
+ .PP
120
+ Especially users of older systems may find the possibility of very large
121
+ memory usage annoying. To prevent uncomfortable surprises, \fBxz\fP has a
122
+ built\-in memory usage limiter, which is disabled by default. While some
123
+ operating systems provide ways to limit the memory usage of processes,
124
+ relying on it wasn't deemed to be flexible enough (for example, using
125
+ \fBulimit\fP(1) to limit virtual memory tends to cripple \fBmmap\fP(2)).
126
+ .PP
127
+ The memory usage limiter can be enabled with the command line option
128
+ \fB\-\-memlimit=\fP\fIlimit\fP. Often it is more convenient to enable the limiter
129
+ by default by setting the environment variable \fBXZ_DEFAULTS\fP, for example,
130
+ \fBXZ_DEFAULTS=\-\-memlimit=150MiB\fP. It is possible to set the limits
131
+ separately for compression and decompression by using
132
+ \fB\-\-memlimit\-compress=\fP\fIlimit\fP and \fB\-\-memlimit\-decompress=\fP\fIlimit\fP.
133
+ Using these two options outside \fBXZ_DEFAULTS\fP is rarely useful because a
134
+ single run of \fBxz\fP cannot do both compression and decompression and
135
+ \fB\-\-memlimit=\fP\fIlimit\fP (or \fB\-M\fP \fIlimit\fP) is shorter to type on the
136
+ command line.
137
+ .PP
138
+ If the specified memory usage limit is exceeded when decompressing, \fBxz\fP
139
+ will display an error and decompressing the file will fail. If the limit is
140
+ exceeded when compressing, \fBxz\fP will try to scale the settings down so that
141
+ the limit is no longer exceeded (except when using \fB\-\-format=raw\fP or
142
+ \fB\-\-no\-adjust\fP). This way the operation won't fail unless the limit is very
143
+ small. The scaling of the settings is done in steps that don't match the
144
+ compression level presets, for example, if the limit is only slightly less
145
+ than the amount required for \fBxz \-9\fP, the settings will be scaled down only
146
+ a little, not all the way down to \fBxz \-8\fP.
147
+ .
148
+ .SS "Concaténation et remplissage avec des fichiers .xz"
149
+ Il est possible de concaténer les fichiers \fB.xz\fP tels quel. \fBxz\fP
150
+ décompressera de tels fichiers comme s'ils étaient un unique fichier \fB.xz\fP.
151
+ .PP
152
+ It is possible to insert padding between the concatenated parts or after the
153
+ last part. The padding must consist of null bytes and the size of the
154
+ padding must be a multiple of four bytes. This can be useful, for example,
155
+ if the \fB.xz\fP file is stored on a medium that measures file sizes in
156
+ 512\-byte blocks.
157
+ .PP
158
+ La concaténation et le remplissage ne sont pas autorisés avec les fichiers
159
+ \&\fB.lzma\fP ou les flux bruts.
160
+ .
161
+ .SH OPTIONS
162
+ .
163
+ .SS "Suffixes entiers et valeurs spéciales."
164
+ Dans la plupart des endroits où un argument entier est attendu, un suffixe
165
+ optionel permet d'indiquer facilement les grands entiers. Il ne doit pas y
166
+ avoir d'espace entre l'entier et le suffixe.
167
+ .TP
168
+ \fBKiB\fP
169
+ Multiplier l'entier par 1024 (2^10). \fBKi\fP, \fBk\fP, \fBkB\fP, \fBK\fP et \fBKB\fP sont
170
+ acceptés comme synonymes de \fBKiB\fP.
171
+ .TP
172
+ \fBMiB\fP
173
+ Multiplier l'entier par 1 048 576 (2^20). \fBMi\fP, \fBm\fP, \fBM\fP et \fBMB\fP sont
174
+ acceptés comme synonymes de \fBMiB\fP.
175
+ .TP
176
+ \fBGiB\fP
177
+ Multiplier l'entier par 1 073 741 824 (2^30). \fBGi\fP, \fBg\fP, \fBG\fP et \fBGB\fP
178
+ sont acceptés comme synonymes de \fBGiB\fP.
179
+ .PP
180
+ La valeur spéciale \fBmax\fP peut être utilisée pour indiquer la valeur
181
+ maximale de l'entier prise en charge par l'option.
182
+ .
183
+ .SS "Mode d'opération"
184
+ Si plusieurs options de mode d'opération sont données, la dernière prend
185
+ effet.
186
+ .TP
187
+ \fB\-z\fP, \fB\-\-compress\fP
188
+ Compresser. C'est le mode d'opération par défaut lorsque aucune option de
189
+ mode opératoire n'est spécifiée ou qu'aucun autre mode d'opération n'est
190
+ sous\-entendu par le nom de la commande (par exemple \fBunxz\fP sous\-entend
191
+ \fB\-\-decompress\fP).
192
+ .IP ""
193
+ .\" The DESCRIPTION section already says this but it's good to repeat it
194
+ .\" here because the default behavior is a bit dangerous and new users
195
+ .\" in a hurry may skip reading the DESCRIPTION section.
196
+ After successful compression, the source file is removed unless writing to
197
+ standard output or \fB\-\-keep\fP was specified.
198
+ .TP
199
+ \fB\-d\fP, \fB\-\-decompress\fP, \fB\-\-uncompress\fP
200
+ .\" The DESCRIPTION section already says this but it's good to repeat it
201
+ .\" here because the default behavior is a bit dangerous and new users
202
+ .\" in a hurry may skip reading the DESCRIPTION section.
203
+ Decompress. After successful decompression, the source file is removed
204
+ unless writing to standard output or \fB\-\-keep\fP was specified.
205
+ .TP
206
+ \fB\-t\fP, \fB\-\-test\fP
207
+ Tester l'intégrité des \fIfichiers\fP compressés. Cette option est équivalente
208
+ à \fB\-\-decompress \-\-stdout\fP sauf que les données décompressées sont rejetées
209
+ au lieu d'être écrites sur la sortie standard. Aucun fichier n'est créé ou
210
+ supprimé.
211
+ .TP
212
+ \fB\-l\fP, \fB\-\-list\fP
213
+ Afficher l'information sur les \fIfichiers\fP compressés. Aucune sortie
214
+ non\-compressée n'est produite et aucun fichier n'est créé ou supprimé. En
215
+ mode liste, le programme ne peut pas lire les données compressées depuis
216
+ l'entrée standard ou depuis d'autres sources non adressables.
217
+ .IP ""
218
+ The default listing shows basic information about \fIfiles\fP, one file per
219
+ line. To get more detailed information, use also the \fB\-\-verbose\fP option.
220
+ For even more information, use \fB\-\-verbose\fP twice, but note that this may be
221
+ slow, because getting all the extra information requires many seeks. The
222
+ width of verbose output exceeds 80 characters, so piping the output to, for
223
+ example, \fBless\ \-S\fP may be convenient if the terminal isn't wide enough.
224
+ .IP ""
225
+ La sortie exacte peut varier suivant les versions de \fBxz\fP et les différents
226
+ paramètres régionaux. Pour une sortie lisible par la machine, utiliser
227
+ \fB\-\-robot \-\-list\fP.
228
+ .
229
+ .SS "Modificateurs d'opération"
230
+ .TP
231
+ \fB\-k\fP, \fB\-\-keep\fP
232
+ Ne pas effacer les fichiers d'entrée.
233
+ .IP ""
234
+ Since \fBxz\fP 5.2.6, this option also makes \fBxz\fP compress or decompress even
235
+ if the input is a symbolic link to a regular file, has more than one hard
236
+ link, or has the setuid, setgid, or sticky bit set. The setuid, setgid, and
237
+ sticky bits are not copied to the target file. In earlier versions this was
238
+ only done with \fB\-\-force\fP.
239
+ .TP
240
+ \fB\-f\fP, \fB\-\-force\fP
241
+ Cette option a plusieurs effets :
242
+ .RS
243
+ .IP \(bu 3
244
+ Si le fichier cible existe déjà, l'effacer avant de compresser ou
245
+ décompresser.
246
+ .IP \(bu 3
247
+ Compresser ou décompresser même si l'entrée est un lien symbolique vers un
248
+ fichier normal, a plus qu'un lien physique, ou a le bit setuid, setgid ou
249
+ sticky défini. Les bits setuid, setgid et sticky bits ne sont pas copiés
250
+ dans le fichier cible.
251
+ .IP \(bu 3
252
+ Lorsque \fBxz\fP est utilisé avec \fB\-\-decompress\fP \fB\-\-stdout\fP et qu'il ne peut
253
+ pas reconnaitre le type du fichier source, copier le fichier source tel quel
254
+ dans la sortie standard. Celà permet à \fBxzcat\fP \fB\-\-force\fP d'être utilisé
255
+ comme \fBcat\fP(1) pour les fichiers qui n'ont pas été compressé avec
256
+ \fBxz\fP. Remarquez que dans le futur, \fBxz\fP devrait prendre en charge de
257
+ nouveaux formats de fichiers compressés, ce qui permettra à \fBxz\fP de
258
+ décompresser plus de types de fichiers au lieu de les copier tels quels dans
259
+ la sortie standard. \fB\-\-format=\fP\fIformat\fP peut être utilisé pour contraindre
260
+ \fBxz\fP à décompresser seulement un format de fichier.
261
+ .RE
262
+ .TP
263
+ \fB\-c\fP, \fB\-\-stdout\fP, \fB\-\-to\-stdout\fP
264
+ Écrire les données compressées ou décompressées sur la sortie standard
265
+ plutôt que dans un fichier. Cela necessite \fB\-\-keep\fP.
266
+ .TP
267
+ \fB\-\-single\-stream\fP
268
+ Décompresser seulement le premier flux \fB.xz\fP et ignorer silencieusement les
269
+ possibles données d'entrée résiduelles qui suivent le flux. Normalement ces
270
+ déchets excédentaires provoquent l'affichage d'une erreur par \fBxz\fP.
271
+ .IP ""
272
+ \fBxz\fP ne décompresse jamais plus d'un flux à partir de fichiers \fB.lzma\fP ou
273
+ de flux bruts, mais cette option fait aussi que \fBxz\fP ignorera les données
274
+ résiduelles après le fichier \fB.lzma\fP ou le flux brut.
275
+ .IP ""
276
+ Cette option n'a aucun effet si le mode d'opération n'est pas
277
+ \fB\-\-decompress\fP ou \fB\-\-test\fP.
278
+ .TP
279
+ \fB\-\-no\-sparse\fP
280
+ Désactiver la création de fichiers peu denses. Par défaut, lors de la
281
+ décompression en un fichier normal, \fBxz\fP essaie d'en faire un fichier creux
282
+ si les données décompressées contiennent de longues séquences de zéros
283
+ binaires. Cela fonctionne aussi lors de l'écriture sur la sortie standard
284
+ aussi longtemps que la sortie standard est connectée à un fichier normal et
285
+ que certaines conditions supplémentaires sont satisfaites pour le faire de
286
+ manière sécurisée. Créer des fichiers creux peut épargner de l'espace disque
287
+ et accélérer la décompression en réduisant la quantité d'entrées/sorties sur
288
+ le disque.
289
+ .TP
290
+ \fB\-S\fP \fI.suf\fP, \fB\-\-suffix=\fP\fI.suf\fP
291
+ Lors de la compression, utiliser \fB.suf\fP comme suffixe du fichier cible au
292
+ lieu de \fB.xz\fP ou \fB.lzma\fP. Si \fBxz\fP n'écrit pas sur la sortie standard et
293
+ si le fichier source a déja le suffixe \fB.suf\fP, un avertissement est affiché
294
+ et le fichier est ignoré.
295
+ .IP ""
296
+ When decompressing, recognize files with the suffix \fI.suf\fP in addition to
297
+ files with the \fB.xz\fP, \fB.txz\fP, \fB.lzma\fP, \fB.tlz\fP, or \fB.lz\fP suffix. If the
298
+ source file has the suffix \fI.suf\fP, the suffix is removed to get the target
299
+ filename.
300
+ .IP ""
301
+ Lors de la compression ou décompression de flux bruts (\fB\-\-fomat=raw\fP), le
302
+ suffixe doit toujours être spécifié à moins d'écrire sur la sortie standard,
303
+ car il n'y a pas de suffixe par défaut pour les flux bruts.
304
+ .TP
305
+ \fB\-\-files\fP[\fB=\fP\fIfichier\fP]
306
+ Lire les noms de fichier à traiter depuis \fIfichier\fP ; si \fIfichier\fP est
307
+ omis , les noms de fichier sont lus sur l'entrée standard. Les noms de
308
+ fichier doivent se terminer avec le caractère de nouvelle ligne. Un tiret
309
+ (\fB\-\fP) est considéré comme un nom de fichier normal ; ce qui ne signifie pas
310
+ entrée standard. Si les noms de fichier sont aussi donnés comme arguments de
311
+ ligne de commande, ils sont traités avant les noms de fichier lus depuis
312
+ \fIfichier\fP.
313
+ .TP
314
+ \fB\-\-files0\fP[\fB=\fP\fIfichier\fP]
315
+ Cela est identique à \fB\-\-files\fP[\fB=\fP\fIfichier\fP] sauf que chaque nom de
316
+ fichier doit se terminer par le caractère null.
317
+ .
318
+ .SS "Format de fichier basique et options de compression"
319
+ .TP
320
+ \fB\-F\fP \fIformat\fP, \fB\-\-format=\fP\fIformat\fP
321
+ Indiquer le \fIformat\fP de fichier à compresser ou décompresser :
322
+ .RS
323
+ .TP
324
+ \fBauto\fP
325
+ C'est celui par défaut. Lors de la compression, \fBauto\fP est équivalent à
326
+ \fBxz\fP. Lors de la décompression, le format du fichier en entrée est détecté
327
+ automatiquement. Notez que les flux bruts (créés avec \fB\-\-format=raw\fP) ne
328
+ peuvent pas être détectés automatiquement.
329
+ .TP
330
+ \fBxz\fP
331
+ Compresser dans le format de fichier \fB.xz\fP ou n'accepter que les fichiers
332
+ \&\fB.xz\fP à décompresser.
333
+ .TP
334
+ \fBlzma\fP, \fBalone\fP
335
+ Compresser au format de fichier \fB.lzma\fP historique, ou n'accepter que les
336
+ fichiers \fB.lzma\fP lors de la décompression. Le nom alternatif \fBalone\fP est
337
+ fourni pour la rétrocompatibilité avec les utilitaires LZMA.
338
+ .TP
339
+ \fBlzip\fP
340
+ Accept only \fB.lz\fP files when decompressing. Compression is not supported.
341
+ .IP ""
342
+ The \fB.lz\fP format version 0 and the unextended version 1 are supported.
343
+ Version 0 files were produced by \fBlzip\fP 1.3 and older. Such files aren't
344
+ common but may be found from file archives as a few source packages were
345
+ released in this format. People might have old personal files in this
346
+ format too. Decompression support for the format version 0 was removed in
347
+ \fBlzip\fP 1.18.
348
+ .IP ""
349
+ \fBlzip\fP 1.4 and later create files in the format version 1. The sync flush
350
+ marker extension to the format version 1 was added in \fBlzip\fP 1.6. This
351
+ extension is rarely used and isn't supported by \fBxz\fP (diagnosed as corrupt
352
+ input).
353
+ .TP
354
+ \fBraw\fP
355
+ Compresser ou décompresser un flux brut (sans en\-têtes). Cela est réservé
356
+ seulement aux utilisateurs aguerris. Pour décoder des flux bruts, vous devez
357
+ utiliser \fB\-\-format=raw\fP et spécifier explicitement la chaîne de filtre, qui
358
+ normalement aurait du être stockée dans les en\-têtes du conteneur.
359
+ .RE
360
+ .TP
361
+ \fB\-C\fP \fIvérif.\fP, \fB\-\-check=\fP\fIvérif.\fP
362
+ Spécifier le type d'intégrité à vérifier. La vérification est calculée à
363
+ partir des données non\-compressées et stockées dans le fichier \fB.xz\fP. Cette
364
+ option n'a effet que si la compression a été faite dans le format \fB.xz\fP ;
365
+ le format \fB.lzma\fP ne gère pas les vérifications d'intégrité. Le contrôle
366
+ d'intégrité (s'il y en a) est vérifié lorsque le fichier \fB.xz\fP est
367
+ décompressé.
368
+ .IP ""
369
+ Types de \fIvérification\fP pris en charge :
370
+ .RS
371
+ .TP
372
+ \fBnone\fP
373
+ Ne pas calculer de vérification d'intégrité du tout. C'est généralement une
374
+ mauvaise idée. Cela peut être utile lorsque l'intégrité des données est
375
+ vérifiée de toute façon par d'autres manières.
376
+ .TP
377
+ \fBcrc32\fP
378
+ Calculer CRC32 en utilisant le polynôme de IEEE\-802.3 (Ethernet).
379
+ .TP
380
+ \fBcrc64\fP
381
+ Calculer CRC64 en utilisant le polynôme de ECMA\-182. C'est la manière
382
+ utilisée par défaut, car c'est légèrement mieux que CRC32 pour détecter les
383
+ fichiers endommagés et la différence de vitesse est négligeable.
384
+ .TP
385
+ \fBsha256\fP
386
+ Calculer SHA\-256. C'est quelque peu plus lent que CRC32 et CRC64.
387
+ .RE
388
+ .IP ""
389
+ L'intégrité des en\-têtes \fB.xz\fP est toujours vérifiée avec CRC32. Il n'est
390
+ pas possible de le changer ou de le désactiver.
391
+ .TP
392
+ \fB\-\-ignore\-check\fP
393
+ Ne pas contrôler la vérification d'intégrité des données lors de la
394
+ décompression. Les valeurs CRC32 dans les en\-têtes \fB.xz\fP seront normalement
395
+ toujours vérifiées.
396
+ .IP ""
397
+ \fBN'utilisez pas cette option à moins de savoir ce que vous faites.\fP Les
398
+ raisons possibles pour utiliser cette option :
399
+ .RS
400
+ .IP \(bu 3
401
+ Essayer de récupérer des données d'un fichier .xz corrompu.
402
+ .IP \(bu 3
403
+ Accélérer la décompression. Cela importe surtout avec SHA\-256 ou avec les
404
+ fichiers qui ont été compressés extrêmement bien. Il est recommandé de ne
405
+ pas utiliser cette option dans ce but à moins que l'intégrité du fichier ne
406
+ soit vérifiée extérieurement d'une autre manière.
407
+ .RE
408
+ .TP
409
+ \fB\-0\fP ... \fB\-9\fP
410
+ Choisir un niveau de compression prédéfini. La valeur par défaut est
411
+ \fB6\fP. Si plusieurs niveaux de préréglage sont spécifi��s, c'est le dernier
412
+ qui sera pris en compte. Si une chaîne de filtres personnalisée a déjà été
413
+ choisie, définir un niveau de compression préréglé efface la chaîne de
414
+ filtres personnalisée.
415
+ .IP ""
416
+ Les différences entre les préréglages sont plus significatives qu'avec
417
+ \fBgzip\fP(1) et \fBbzip2\fP(1). les réglages de compression sélectionnés
418
+ déterminent les exigences en mémoire pour la décompression, ainsi, utiliser
419
+ un niveau de préréglage trop élevé peut rendre difficile à décompresser un
420
+ fichier sur un vieux système avec peu de RAM. Clairement, \fBce n'est pas une bonne idée d'utiliser \-9 aveuglément pour tout\fP comme ça l'est souvent avec
421
+ \fBgzip\fP(1) et \fBbzip2\fP(1).
422
+ .RS
423
+ .TP
424
+ \fB\-0\fP ... \fB\-3\fP
425
+ Ce sont des préréglages relativement rapides. \fB0\fP est parfois plus rapide
426
+ que \fBgzip \-9\fP tout en compressant bien mieux. Les réglages plus élevés ont
427
+ souvent une rapidité comparable à celle de \fBbzip2\fP(1) avec un taux de
428
+ compression comparable ou meilleur, même si les résultats dépendent beaucoup
429
+ du genre de données compressées.
430
+ .TP
431
+ \fB\-4\fP ... \fB\-6\fP
432
+ Good to very good compression while keeping decompressor memory usage
433
+ reasonable even for old systems. \fB\-6\fP is the default, which is usually a
434
+ good choice for distributing files that need to be decompressible even on
435
+ systems with only 16\ MiB RAM. (\fB\-5e\fP or \fB\-6e\fP may be worth considering
436
+ too. See \fB\-\-extreme\fP.)
437
+ .TP
438
+ \fB\-7 ... \-9\fP
439
+ C'est comme \fB\-6\fP mais avec des besoins en mémoire plus élevés pour la
440
+ compression et la décompression. Ce n'est utile que lorsque les fichiers
441
+ sont plus gros que 8\ Mio, 16\ Mio et 32\ Mio respectivement.
442
+ .RE
443
+ .IP ""
444
+ Sur le même matériel, la vitesse de décompression est sensiblement un nombre
445
+ constant d'octets de données compressées par seconde. En d'autres termes,
446
+ meilleure est la compression, plus rapide sera en général la
447
+ décompression. Cela signifie aussi que la quantité de sortie non compressée
448
+ produite par seconde peut varier beaucoup.
449
+ .IP ""
450
+ Le tableau suivant résume les caractéristiques des préréglages :
451
+ .RS
452
+ .RS
453
+ .PP
454
+ .TS
455
+ tab(;);
456
+ c c c c c
457
+ n n n n n.
458
+ Préréglage;DictSize;CompCPU;CompMem;DecMem
459
+ \-0;256 KiB;0;3 MiB;1 MiB
460
+ \-1;1 MiB;1;9 MiB;2 MiB
461
+ \-2;2 MiB;2;17 MiB;3 MiB
462
+ \-3;4 MiB;3;32 MiB;5 MiB
463
+ \-4;4 MiB;4;48 MiB;5 MiB
464
+ \-5;8 MiB;5;94 MiB;9 MiB
465
+ \-6;8 MiB;6;94 MiB;9 MiB
466
+ \-7;16 MiB;6;186 MiB;17 MiB
467
+ \-8;32 MiB;6;370 MiB;33 MiB
468
+ \-9;64 MiB;6;674 MiB;65 MiB
469
+ .TE
470
+ .RE
471
+ .RE
472
+ .IP ""
473
+ Descriptions des colonnes :
474
+ .RS
475
+ .IP \(bu 3
476
+ DictSize est la taille du dictionnaire de LZMA2. Utiliser un dictionnaire
477
+ plus gros que la taille du fichier non compressé est un gaspillage de
478
+ mémoire. C'est pourquoi il est bon d'éviter d'utiliser les préréglages de
479
+ \fB\-7\fP à \fB\-9\fP lorsqu'il n'y en a pas vraiment besoin. A \fB\-6\fP et plus bas,
480
+ la quantité de mémoire gaspillée est généralement assez basse pour ne pas
481
+ être un problème.
482
+ .IP \(bu 3
483
+ CompCPU est une représentation des préréglages de LZMA2 qui affectent la
484
+ vitesse de compression. La taille du dictionnaire aussi affecte la vitesse,
485
+ alors comme CompCPU est le même pour les niveaux de \fB\-6\fP à \fB\-9\fP, les plus
486
+ haut niveaux tendent à être un peu moins rapides. Pour être encore moins
487
+ rapide et du coup obtenir peut être une meilleure compression, consultez
488
+ \fB\-\-extreme\fP.
489
+ .IP \(bu 3
490
+ CompMem contains the compressor memory requirements in the single\-threaded
491
+ mode. It may vary slightly between \fBxz\fP versions.
492
+ .IP \(bu 3
493
+ DecMem contient les besoins en mémoire du décompresseur. Ce sont les
494
+ réglages de la compression qui déterminent les besoins en mémoire de la
495
+ décompression. L'exacte utilisation de la mémoire est légèrement supérieure
496
+ à la taille du dictionnaire LZMA2, mais les valeurs dans la table ont été
497
+ arrondies au prochain Mio supérieur.
498
+ .RE
499
+ .IP ""
500
+ Memory requirements of the multi\-threaded mode are significantly higher than
501
+ that of the single\-threaded mode. With the default value of
502
+ \fB\-\-block\-size\fP, each thread needs 3*3*DictSize plus CompMem or DecMem. For
503
+ example, four threads with preset \fB\-6\fP needs 660\(en670\ MiB of memory.
504
+ .TP
505
+ \fB\-e\fP, \fB\-\-extreme\fP
506
+ Utilisez un variant plus lent que les préréglages (\fB\-0\fP à \fB\-9\fP) pour
507
+ espérer avoir un taux de compression légèrement meilleur, mais en cas de
508
+ malchance cela peut être pire. L'utilisation mémoire du décompresseur n'est
509
+ pas affectée, mais l'utilisation mémoire du compresseur augmente un peu aux
510
+ niveaux de préréglages de \fB\-0\fP à \fB\-3\fP.
511
+ .IP ""
512
+ Depuis qu'il y a deux préréglages avec des tailles de dictionnaire de 4\ Mio
513
+ et 8 \Mio, les préréglages \fB\-3e\fP et \fB\-5e\fP utilisent des réglages
514
+ légèrement plus rapides que \fB\-4e\fP et \fB\-6e\fP, respectivement. De cette
515
+ manière, il n'y a pas deux préréglages identiques.
516
+ .RS
517
+ .RS
518
+ .PP
519
+ .TS
520
+ tab(;);
521
+ c c c c c
522
+ n n n n n.
523
+ Préréglage;DictSize;CompCPU;CompMem;DecMem
524
+ \-0e;256 KiB;8;4 MiB;1 MiB
525
+ \-1e;1 MiB;8;13 MiB;2 MiB
526
+ \-2e;2 MiB;8;25 MiB;3 MiB
527
+ \-3e;4 MiB;7;48 MiB;5 MiB
528
+ \-4e;4 MiB;8;48 MiB;5 MiB
529
+ \-5e;8 MiB;7;94 MiB;9 MiB
530
+ \-6e;8 MiB;8;94 MiB;9 MiB
531
+ \-7e;16 MiB;8;186 MiB;17 MiB
532
+ \-8e;32 MiB;8;370 MiB;33 MiB
533
+ \-9e;64 MiB;8;674 MiB;65 MiB
534
+ .TE
535
+ .RE
536
+ .RE
537
+ .IP ""
538
+ Par exemple, il y a un total de quatre préréglages qui utilisent un
539
+ dictionnaire de 8 Mio et qui sont dans l'ordre du plus rapide au plus lent :
540
+ \fB\-5\fP, \fB\-6\fP, \fB\-5e\fP et \fB\-6e\fP.
541
+ .TP
542
+ \fB\-\-fast\fP
543
+ .PD 0
544
+ .TP
545
+ \fB\-\-best\fP
546
+ .PD
547
+ Il y a néanmoins des alias trompeurs pour \fB\-0\fP et \fB\-9\fP,
548
+ respectivement. Ils ne sont fournis que pour des besoins de
549
+ rétro\-compatibilité avec les utilitaires LZMA. Evitez d'utiliser ces
550
+ options.
551
+ .TP
552
+ \fB\-\-block\-size=\fP\fItaille\fP
553
+ Lors de la compression dans le format \fB.xz\fP, les données de l'entrée sont
554
+ réparties en blocs de \fItaille\fP octets. Les blocs sont compressés
555
+ indépendamment les un des autres, ce qui aide avec le mode multithread
556
+ (multi\-threading) et rend possible la décompression à accès aléatoire
557
+ limité. Cette option est typiquement utilisée pour outrepasser la taille de
558
+ bloc en mode multithread, mais cette option peut aussi être utilisée en mode
559
+ single\-thread.
560
+ .IP ""
561
+ In multi\-threaded mode about three times \fIsize\fP bytes will be allocated in
562
+ each thread for buffering input and output. The default \fIsize\fP is three
563
+ times the LZMA2 dictionary size or 1 MiB, whichever is more. Typically a
564
+ good value is 2\(en4 times the size of the LZMA2 dictionary or at least 1
565
+ MiB. Using \fIsize\fP less than the LZMA2 dictionary size is waste of RAM
566
+ because then the LZMA2 dictionary buffer will never get fully used. In
567
+ multi\-threaded mode, the sizes of the blocks are stored in the block
568
+ headers. This size information is required for multi\-threaded
569
+ decompression.
570
+ .IP ""
571
+ In single\-threaded mode no block splitting is done by default. Setting this
572
+ option doesn't affect memory usage. No size information is stored in block
573
+ headers, thus files created in single\-threaded mode won't be identical to
574
+ files created in multi\-threaded mode. The lack of size information also
575
+ means that \fBxz\fP won't be able decompress the files in multi\-threaded mode.
576
+ .TP
577
+ \fB\-\-block\-list=\fP\fIitems\fP
578
+ When compressing to the \fB.xz\fP format, start a new block with an optional
579
+ custom filter chain after the given intervals of uncompressed data.
580
+ .IP ""
581
+ The \fIitems\fP are a comma\-separated list. Each item consists of an optional
582
+ filter chain number between 0 and 9 followed by a colon (\fB:\fP) and a
583
+ required size of uncompressed data. Omitting an item (two or more
584
+ consecutive commas) is a shorthand to use the size and filters of the
585
+ previous item.
586
+ .IP ""
587
+ If the input file is bigger than the sum of the sizes in \fIitems\fP, the last
588
+ item is repeated until the end of the file. A special value of \fB0\fP may be
589
+ used as the last size to indicate that the rest of the file should be
590
+ encoded as a single block.
591
+ .IP ""
592
+ An alternative filter chain for each block can be specified in combination
593
+ with the \fB\-\-filters1=\fP\fIfilters\fP \&...\& \fB\-\-filters9=\fP\fIfilters\fP options.
594
+ These options define filter chains with an identifier between 1\(en9.
595
+ Filter chain 0 can be used to refer to the default filter chain, which is
596
+ the same as not specifying a filter chain. The filter chain identifier can
597
+ be used before the uncompressed size, followed by a colon (\fB:\fP). For
598
+ example, if one specifies \fB\-\-block\-list=1:2MiB,3:2MiB,2:4MiB,,2MiB,0:4MiB\fP
599
+ then blocks will be created using:
600
+ .RS
601
+ .IP \(bu 3
602
+ The filter chain specified by \fB\-\-filters1\fP and 2 MiB input
603
+ .IP \(bu 3
604
+ The filter chain specified by \fB\-\-filters3\fP and 2 MiB input
605
+ .IP \(bu 3
606
+ The filter chain specified by \fB\-\-filters2\fP and 4 MiB input
607
+ .IP \(bu 3
608
+ The filter chain specified by \fB\-\-filters2\fP and 4 MiB input
609
+ .IP \(bu 3
610
+ The default filter chain and 2 MiB input
611
+ .IP \(bu 3
612
+ The default filter chain and 4 MiB input for every block until end of input.
613
+ .RE
614
+ .IP ""
615
+ If one specifies a size that exceeds the encoder's block size (either the
616
+ default value in threaded mode or the value specified with
617
+ \fB\-\-block\-size=\fP\fIsize\fP), the encoder will create additional blocks while
618
+ keeping the boundaries specified in \fIitems\fP. For example, if one specifies
619
+ \fB\-\-block\-size=10MiB\fP \fB\-\-block\-list=5MiB,10MiB,8MiB,12MiB,24MiB\fP and the
620
+ input file is 80 MiB, one will get 11 blocks: 5, 10, 8, 10, 2, 10, 10, 4,
621
+ 10, 10, and 1 MiB.
622
+ .IP ""
623
+ En mode multi\-threadé les tailles de blocs sont stockées dans les en\-têtes
624
+ du bloc. Cela ne se fait pas en mode mono\-threadé, la sortie encodée ne sera
625
+ donc pas identique à celle faite en mode multi\-threadé.
626
+ .TP
627
+ \fB\-\-flush\-timeout=\fP\fItemps_d'attente\fP
628
+ Lors de la compression, si plus que \fItemps_d'attente\fP millisecondes (un
629
+ entier positif) se sont écoulées depuis le précédent vidage et que lire plus
630
+ de données bloquerait, toutes les données d'entrée en attente sont vidées de
631
+ l'encodeur et mises à disposition dans le flux de sortie. Cela peut être
632
+ utile si \fBxz\fP est utilisé pour compresser les données qui sont diffusées
633
+ sur un réseau. Des petites valeurs de \fItemps_d'attente\fP rendent les données
634
+ disponibles à l'extrémité réceptrice avec un léger retard, mais les grandes
635
+ valeurs de \fItemps_d'attente\fP donnent un meilleur taux de compression.
636
+ .IP ""
637
+ Cette option est désactivée par défaut. Si cette option est indiquée plus
638
+ d'une fois, la dernière prend effet. La valeur spéciale de
639
+ \fItemps_d'attente\fP de \fB0\fP peut être utilisée pour explicitement désactiver
640
+ cette option.
641
+ .IP ""
642
+ Cette option n'est pas disponible sur les systèmes qui ne sont pas POSIX.
643
+ .IP ""
644
+ .\" FIXME
645
+ \fBCette option est encore expérimentale.\fP Actuellement, \fBxz\fP ne convient
646
+ pas pour décompresser le flux en temps réel en raison de la façon dont \fBxz\fP
647
+ effectue la mise en mémoire tampon.
648
+ .TP
649
+ \fB\-\-memlimit\-compress=\fP\fIlimite\fP
650
+ Indiquer une limite d'utilisation de la mémoire pour la compression. Si
651
+ cette option est indiquée plusieurs fois, c'est la dernière qui est prise en
652
+ compte.
653
+ .IP ""
654
+ If the compression settings exceed the \fIlimit\fP, \fBxz\fP will attempt to
655
+ adjust the settings downwards so that the limit is no longer exceeded and
656
+ display a notice that automatic adjustment was done. The adjustments are
657
+ done in this order: reducing the number of threads, switching to
658
+ single\-threaded mode if even one thread in multi\-threaded mode exceeds the
659
+ \fIlimit\fP, and finally reducing the LZMA2 dictionary size.
660
+ .IP ""
661
+ When compressing with \fB\-\-format=raw\fP or if \fB\-\-no\-adjust\fP has been
662
+ specified, only the number of threads may be reduced since it can be done
663
+ without affecting the compressed output.
664
+ .IP ""
665
+ If the \fIlimit\fP cannot be met even with the adjustments described above, an
666
+ error is displayed and \fBxz\fP will exit with exit status 1.
667
+ .IP ""
668
+ La \fIlimite\fP peut être indiquée de plusieurs façons :
669
+ .RS
670
+ .IP \(bu 3
671
+ La \fIlimite\fP peut être une valeur absolue en octets. Utiliser un suffixe
672
+ d'entier comme \fBMiB\fP peut être utile. Exemple :
673
+ \fB\-\-memlimit\-compress=80MiB\fP
674
+ .IP \(bu 3
675
+ La \fIlimite\fP peut être indiquée sous forme d'un pourcentage de la mémoire
676
+ physique totale (RAM). Cela peut être particulièrement utile quand la
677
+ variable d'environnement \fBXZ_DEFAULTS\fP est indiquée dans un script
678
+ d'initialisation de l'interpréteur partagé entre différents ordinateurs. De
679
+ cette façon la limite est automatiquement plus grande sur les systèmes avec
680
+ plus de mémoire. Exemple : \fB\-\-memlimit=70%\fP
681
+ .IP \(bu 3
682
+ The \fIlimit\fP can be reset back to its default value by setting it to \fB0\fP.
683
+ This is currently equivalent to setting the \fIlimit\fP to \fBmax\fP (no memory
684
+ usage limit).
685
+ .RE
686
+ .IP ""
687
+ For 32\-bit \fBxz\fP there is a special case: if the \fIlimit\fP would be over
688
+ \fB4020\ MiB\fP, the \fIlimit\fP is set to \fB4020\ MiB\fP. On MIPS32 \fB2000\ MiB\fP
689
+ is used instead. (The values \fB0\fP and \fBmax\fP aren't affected by this. A
690
+ similar feature doesn't exist for decompression.) This can be helpful when
691
+ a 32\-bit executable has access to 4\ GiB address space (2 GiB on MIPS32)
692
+ while hopefully doing no harm in other situations.
693
+ .IP ""
694
+ Voir aussi la section \fButilisation de la mémoire\fP.
695
+ .TP
696
+ \fB\-\-memlimit\-decompress=\fP\fIlimite\fP
697
+ Régler une limite d'utilisation de la mémoire pour la décompression. Cela a
698
+ un effet sur le mode \fB\-\-list\fP. Si l'opération n'est pas possible sans
699
+ dépasser la \fIlimite\fP, \fBxz\fP affichera une erreur et la décompression
700
+ échouera. Voir \fB\-\-memlimit\-compress=\fP\fIlimite\fP pour les manières possibles
701
+ d'indiquer la \fIlimite\fP.
702
+ .TP
703
+ \fB\-\-memlimit\-mt\-decompress=\fP\fIlimit\fP
704
+ Set a memory usage limit for multi\-threaded decompression. This can only
705
+ affect the number of threads; this will never make \fBxz\fP refuse to
706
+ decompress a file. If \fIlimit\fP is too low to allow any multi\-threading, the
707
+ \fIlimit\fP is ignored and \fBxz\fP will continue in single\-threaded mode. Note
708
+ that if also \fB\-\-memlimit\-decompress\fP is used, it will always apply to both
709
+ single\-threaded and multi\-threaded modes, and so the effective \fIlimit\fP for
710
+ multi\-threading will never be higher than the limit set with
711
+ \fB\-\-memlimit\-decompress\fP.
712
+ .IP ""
713
+ In contrast to the other memory usage limit options,
714
+ \fB\-\-memlimit\-mt\-decompress=\fP\fIlimit\fP has a system\-specific default
715
+ \fIlimit\fP. \fBxz \-\-info\-memory\fP can be used to see the current value.
716
+ .IP ""
717
+ This option and its default value exist because without any limit the
718
+ threaded decompressor could end up allocating an insane amount of memory
719
+ with some input files. If the default \fIlimit\fP is too low on your system,
720
+ feel free to increase the \fIlimit\fP but never set it to a value larger than
721
+ the amount of usable RAM as with appropriate input files \fBxz\fP will attempt
722
+ to use that amount of memory even with a low number of threads. Running out
723
+ of memory or swapping will not improve decompression performance.
724
+ .IP ""
725
+ See \fB\-\-memlimit\-compress=\fP\fIlimit\fP for possible ways to specify the
726
+ \fIlimit\fP. Setting \fIlimit\fP to \fB0\fP resets the \fIlimit\fP to the default
727
+ system\-specific value.
728
+ .TP
729
+ \fB\-M\fP \fIlimite\fP, \fB\-\-memlimit=\fP\fIlimite\fP, \fB\-\-memory=\fP\fIlimite\fP
730
+ This is equivalent to specifying \fB\-\-memlimit\-compress=\fP\fIlimit\fP
731
+ \fB\-\-memlimit\-decompress=\fP\fIlimit\fP \fB\-\-memlimit\-mt\-decompress=\fP\fIlimit\fP.
732
+ .TP
733
+ \fB\-\-no\-adjust\fP
734
+ Display an error and exit if the memory usage limit cannot be met without
735
+ adjusting settings that affect the compressed output. That is, this
736
+ prevents \fBxz\fP from switching the encoder from multi\-threaded mode to
737
+ single\-threaded mode and from reducing the LZMA2 dictionary size. Even when
738
+ this option is used the number of threads may be reduced to meet the memory
739
+ usage limit as that won't affect the compressed output.
740
+ .IP ""
741
+ Automatic adjusting is always disabled when creating raw streams
742
+ (\fB\-\-format=raw\fP).
743
+ .TP
744
+ \fB\-T\fP \fIthreads\fP, \fB\-\-threads=\fP\fIthreads\fP
745
+ Specify the number of worker threads to use. Setting \fIthreads\fP to a
746
+ special value \fB0\fP makes \fBxz\fP use up to as many threads as the processor(s)
747
+ on the system support. The actual number of threads can be fewer than
748
+ \fIthreads\fP if the input file is not big enough for threading with the given
749
+ settings or if using more threads would exceed the memory usage limit.
750
+ .IP ""
751
+ The single\-threaded and multi\-threaded compressors produce different
752
+ output. Single\-threaded compressor will give the smallest file size but
753
+ only the output from the multi\-threaded compressor can be decompressed using
754
+ multiple threads. Setting \fIthreads\fP to \fB1\fP will use the single\-threaded
755
+ mode. Setting \fIthreads\fP to any other value, including \fB0\fP, will use the
756
+ multi\-threaded compressor even if the system supports only one hardware
757
+ thread. (\fBxz\fP 5.2.x used single\-threaded mode in this situation.)
758
+ .IP ""
759
+ To use multi\-threaded mode with only one thread, set \fIthreads\fP to \fB+1\fP.
760
+ The \fB+\fP prefix has no effect with values other than \fB1\fP. A memory usage
761
+ limit can still make \fBxz\fP switch to single\-threaded mode unless
762
+ \fB\-\-no\-adjust\fP is used. Support for the \fB+\fP prefix was added in \fBxz\fP
763
+ 5.4.0.
764
+ .IP ""
765
+ If an automatic number of threads has been requested and no memory usage
766
+ limit has been specified, then a system\-specific default soft limit will be
767
+ used to possibly limit the number of threads. It is a soft limit in sense
768
+ that it is ignored if the number of threads becomes one, thus a soft limit
769
+ will never stop \fBxz\fP from compressing or decompressing. This default soft
770
+ limit will not make \fBxz\fP switch from multi\-threaded mode to single\-threaded
771
+ mode. The active limits can be seen with \fBxz \-\-info\-memory\fP.
772
+ .IP ""
773
+ Actuellement, la seule méthode de gestion avec des threads consiste à
774
+ séparer l'entrée en blocs et de les compresser indépendamment les uns des
775
+ autres. La taille par défaut des blocs dépend du niveau de compression et
776
+ peut\-être remplacée avec l'option \fB\-\-block\-size=\fP\fItaille\fP.
777
+ .IP ""
778
+ Threaded decompression only works on files that contain multiple blocks with
779
+ size information in block headers. All large enough files compressed in
780
+ multi\-threaded mode meet this condition, but files compressed in
781
+ single\-threaded mode don't even if \fB\-\-block\-size=\fP\fIsize\fP has been used.
782
+ .IP ""
783
+ The default value for \fIthreads\fP is \fB0\fP. In \fBxz\fP 5.4.x and older the
784
+ default is \fB1\fP.
785
+ .
786
+ .SS "Chaînes de filtres de compresseur personnalisées"
787
+ A custom filter chain allows specifying the compression settings in detail
788
+ instead of relying on the settings associated to the presets. When a custom
789
+ filter chain is specified, preset options (\fB\-0\fP \&...\& \fB\-9\fP and
790
+ \fB\-\-extreme\fP) earlier on the command line are forgotten. If a preset
791
+ option is specified after one or more custom filter chain options, the new
792
+ preset takes effect and the custom filter chain options specified earlier
793
+ are forgotten.
794
+ .PP
795
+ Une chaîne de filtre est comparable à une redirection (pipe) sur la ligne de
796
+ commande. Lors de la compression, les entrées non compressées vont au
797
+ premier filtre, dont la sortie va au prochain filtre (s'il y en a). La
798
+ sortie du dernier filtre est écrite sur le fichier compressé. Le nombre
799
+ maximal de filtres dans la chaîne est quatre, mais habituellement, un chaîne
800
+ de filtre n'a qu'un ou deux filtres.
801
+ .PP
802
+ Beaucoup de filtres ont des limitations sur l'endroit où ils peuvent se
803
+ placer dans la chaîne de filtre : quelques filtres ne peuvent fonctionner
804
+ qu'en tant que dernier filtre dans la chaîne, quelques uns en tant que non
805
+ dernier filtre, et d'autres à n'importe quelle position dans la
806
+ chaîne. Suivant le filtre, cette limitation est soit inhérente au profil du
807
+ filtre, soit existe pour des raisons de sécurité.
808
+ .PP
809
+ A custom filter chain can be specified in two different ways. The options
810
+ \fB\-\-filters=\fP\fIfilters\fP and \fB\-\-filters1=\fP\fIfilters\fP \&...\&
811
+ \fB\-\-filters9=\fP\fIfilters\fP allow specifying an entire filter chain in one
812
+ option using the liblzma filter string syntax. Alternatively, a filter
813
+ chain can be specified by using one or more individual filter options in the
814
+ order they are wanted in the filter chain. That is, the order of the
815
+ individual filter options is significant! When decoding raw streams
816
+ (\fB\-\-format=raw\fP), the filter chain must be specified in the same order as
817
+ it was specified when compressing. Any individual filter or preset options
818
+ specified before the full chain option (\fB\-\-filters=\fP\fIfilters\fP) will be
819
+ forgotten. Individual filters specified after the full chain option will
820
+ reset the filter chain.
821
+ .PP
822
+ Both the full and individual filter options take filter\-specific \fIoptions\fP
823
+ as a comma\-separated list. Extra commas in \fIoptions\fP are ignored. Every
824
+ option has a default value, so specify those you want to change.
825
+ .PP
826
+ Pour voir l'entièreté de la chaîne de filtres et ses \fIoptions\fP, utilisez
827
+ \fBxz \-vv\fP (ce qui est comme utiliser \fB\-\-verbose\fP deux fois). Cela
828
+ fonctionne aussi pour voir les options de chaîne de filtres utilisées par
829
+ les préréglages.
830
+ .TP
831
+ \fB\-\-filters=\fP\fIfilters\fP
832
+ Specify the full filter chain or a preset in a single option. Each filter
833
+ can be separated by spaces or two dashes (\fB\-\-\fP). \fIfilters\fP may need to be
834
+ quoted on the shell command line so it is parsed as a single option. To
835
+ denote \fIoptions\fP, use \fB:\fP or \fB=\fP. A preset can be prefixed with a \fB\-\fP
836
+ and followed with zero or more flags. The only supported flag is \fBe\fP to
837
+ apply the same options as \fB\-\-extreme\fP.
838
+ .TP
839
+ \fB\-\-filters1\fP=\fIfilters\fP ... \fB\-\-filters9\fP=\fIfilters\fP
840
+ Specify up to nine additional filter chains that can be used with
841
+ \fB\-\-block\-list\fP.
842
+ .IP ""
843
+ For example, when compressing an archive with executable files followed by
844
+ text files, the executable part could use a filter chain with a BCJ filter
845
+ and the text part only the LZMA2 filter.
846
+ .TP
847
+ \fB\-\-filters\-help\fP
848
+ Display a help message describing how to specify presets and custom filter
849
+ chains in the \fB\-\-filters\fP and \fB\-\-filters1=\fP\fIfilters\fP \&...\&
850
+ \fB\-\-filters9=\fP\fIfilters\fP options, and exit successfully.
851
+ .TP
852
+ \fB\-\-lzma1\fP[\fB=\fP\fIoptions\fP]
853
+ .PD 0
854
+ .TP
855
+ \fB\-\-lzma2\fP[\fB=\fP\fIoptions\fP]
856
+ .PD
857
+ Ajouter le filtre LZMA1 ou LZMA2 à la chaîne de filtres. Ces filtres ne
858
+ peuvent être utilisés que comme dernier filtre dans la chaîne.
859
+ .IP ""
860
+ LZMA1 est un filtre historique, qui n'est pris en charge presque uniquement
861
+ à cause de l'ancien format de fichier \fB.lzma\fP, qui ne prend en charge que
862
+ LZMA1. LZMA2 est une version mise à jour de LZMA1 pour régler certains
863
+ problèmes pratiques de LZMA1. Le format \fBxz\fP utilise LZMA2 et ne prend pas
864
+ du tout en charge LZMA1. Les taux et vitesses de compression de LZMA1 et
865
+ LZMA2 sont pratiquement identiques.
866
+ .IP ""
867
+ LZMA1 et LZMA2 partagent le même ensemble d'\fIoptions\fP :
868
+ .RS
869
+ .TP
870
+ \fBpreset=\fP\fIpréréglage\fP
871
+ Reset all LZMA1 or LZMA2 \fIoptions\fP to \fIpreset\fP. \fIPreset\fP consist of an
872
+ integer, which may be followed by single\-letter preset modifiers. The
873
+ integer can be from \fB0\fP to \fB9\fP, matching the command line options \fB\-0\fP
874
+ \&...\& \fB\-9\fP. The only supported modifier is currently \fBe\fP, which matches
875
+ \fB\-\-extreme\fP. If no \fBpreset\fP is specified, the default values of LZMA1 or
876
+ LZMA2 \fIoptions\fP are taken from the preset \fB6\fP.
877
+ .TP
878
+ \fBdict=\fP\fItaille\fP
879
+ La \fItaille\fP du dictionnaire (historique du tampon) indique combien d'octets
880
+ des données récement décompressées sont gardés en mémoire. L'algorithme
881
+ essaie de trouver les séquences d'octets répétées (identiques) dans les
882
+ données décompressées et les remplace par les données actuellement dans le
883
+ dictionnaire. Plus gros est le dictionnaire, plus grande est la chance de
884
+ trouver une correspondance. Ainsi, l'augmentation de la \fItaille\fP du
885
+ dictionnaire augmente habituellement le taux de compression, mais un
886
+ dictionnaire plus gros que le fichier non compressé est un gachis de
887
+ mémoire.
888
+ .IP ""
889
+ Généralement la \fItaille\fP du dictionnaire est entre 64\ Kio et 64\ Mio. Le
890
+ minimum étant 4\ Kio. La \fItaille\fP maximale pour la compression est
891
+ habituellement 1,5\ Gio (1536\ Mio). Le décompresseur prend en charge les
892
+ dictionnaires jusqu'à un octet de moins que 4\ Gio, ce qui est le maximum
893
+ pour les formats de flux LZMA1 et LZMA2.
894
+ .IP ""
895
+ La \fItaille\fP du dictionnaire et le chercheur de correspondance (match
896
+ finder) (\fImf\fP) déterminent ensemble l'utilisation de la mémoire de
897
+ l'encodeur LZMA1 ou LZMA2. La même (ou une plus grosse) \fItaille\fP de
898
+ dictionnaire est requise pour décompresser que ce qui a été utilisé pour la
899
+ compression, ainsi l'utilisation de la mémoire du décodeur est déterminée
900
+ par la taille du dictionnaire utilisée lors de la compression. Les en\-têtes
901
+ de \fB.xz\fP stockent la \fItaille\fP de dictionnaire sous la forme 2^\fIn\fP ou
902
+ 2^\fIn\fP + 2^(\fIn\fP\-1), de sorte que ces \fItailles\fP sont quelque peu préférées
903
+ pour la compression. Les autres \fItailles\fP seront arrondies à la hausse
904
+ lorsque stockées dans les en\-têtes de \fB.xz\fP.
905
+ .TP
906
+ \fBlc=\fP\fIlc\fP
907
+ Spécifiez le nombre d'octets de contexte littéraux. Le minimum est \fB0\fP et
908
+ le maximum est \fB4\fP. La valeur par défaut est \fB3\fP. En plus, la somme de
909
+ \fIlc\fP et \fIlp\fP ne doit pas excéder \fB4\fP.
910
+ .IP ""
911
+ Tous les octets qui ne peuvent pas être codés comme des correspondances sont
912
+ codés comme des littéraux. C'est à dire que les littéraux sont simplement
913
+ des octets 8 bits encodés un à la fois.
914
+ .IP ""
915
+ The literal coding makes an assumption that the highest \fIlc\fP bits of the
916
+ previous uncompressed byte correlate with the next byte. For example, in
917
+ typical English text, an upper\-case letter is often followed by a lower\-case
918
+ letter, and a lower\-case letter is usually followed by another lower\-case
919
+ letter. In the US\-ASCII character set, the highest three bits are 010 for
920
+ upper\-case letters and 011 for lower\-case letters. When \fIlc\fP is at least
921
+ 3, the literal coding can take advantage of this property in the
922
+ uncompressed data.
923
+ .IP ""
924
+ The default value (3) is usually good. If you want maximum compression,
925
+ test \fBlc=4\fP. Sometimes it helps a little, and sometimes it makes
926
+ compression worse. If it makes it worse, test \fBlc=2\fP too.
927
+ .TP
928
+ \fBlp=\fP\fIlp\fP
929
+ Indiquer le nombre de bits de position littérale. Le minimum est \fB0\fP et le
930
+ maximum \fB4\fP; par défaut c'est \fB0\fP.
931
+ .IP ""
932
+ \fILp\fP affecte le type d'alignement dans les données décompressées qui est
933
+ présumé lors de l'encodage des littéraux. Voir \fIpb\fP ci dessous pour plus
934
+ d'information sur l'alignement.
935
+ .TP
936
+ \fBpb=\fP\fIpb\fP
937
+ Indiquer le nombre de bits de position. Le minimum est \fB0\fP et le maximum
938
+ \fB4\fP; par défaut \fB2\fP.
939
+ .IP ""
940
+ \fIPb\fP affecte quel genre d'alignement est présumé en général dans les
941
+ données non compressées. Par défaut c'est un alignement de quatre octets
942
+ (2^\fIpb\fP=2^2=4), ce qui est généralement un bon choix lorsqu'il n'y a pas de
943
+ meilleure estimation.
944
+ .IP ""
945
+ When the alignment is known, setting \fIpb\fP accordingly may reduce the file
946
+ size a little. For example, with text files having one\-byte alignment
947
+ (US\-ASCII, ISO\-8859\-*, UTF\-8), setting \fBpb=0\fP can improve compression
948
+ slightly. For UTF\-16 text, \fBpb=1\fP is a good choice. If the alignment is
949
+ an odd number like 3 bytes, \fBpb=0\fP might be the best choice.
950
+ .IP ""
951
+ Même si l'alignement présumé peut être ajusté avec \fIpb\fP et \fIlp\fP, LZMA1 et
952
+ LZMA2 favorisent toujours légèrement l'alignement sur 16 octets. Il peut
953
+ être utile d'en tenir compte lors de la conception de formats de fichiers
954
+ susceptibles d'être souvent compressés avec LZMA1 ou LZMA2.
955
+ .TP
956
+ \fBmf=\fP\fImf\fP
957
+ Match finder has a major effect on encoder speed, memory usage, and
958
+ compression ratio. Usually Hash Chain match finders are faster than Binary
959
+ Tree match finders. The default depends on the \fIpreset\fP: 0 uses \fBhc3\fP,
960
+ 1\(en3 use \fBhc4\fP, and the rest use \fBbt4\fP.
961
+ .IP ""
962
+ Les chercheurs de correspondance suivants sont pris en charge. Les formules
963
+ d'utilisation de la mémoire ci\-dessous sont des approximations grossières
964
+ qui sont les plus proches de la réalité lorsque \fIdict\fP est une puissance de
965
+ deux.
966
+ .RS
967
+ .TP
968
+ \fBhc3\fP
969
+ Chaîne de hachage avec hachage de 2 et 3 octets
970
+ .br
971
+ Valeur minimale pour \fInice\fP : \fB3\fP
972
+ .br
973
+ Utilisation de la mémoire :
974
+ .br
975
+ \fIdict\fP * 7.5 (if \fIdict\fP <= 16 Mio);
976
+ .br
977
+ \fIdict\fP * 5.5 + 64 MiB (si \fIdict\fP > 16 Mio)
978
+ .TP
979
+ \fBhc4\fP
980
+ Chaîne de hachage avec hachage de 2, 3 et 4 octets
981
+ .br
982
+ Valeur minimale pour \fInice\fP : \fB4\fP
983
+ .br
984
+ Utilisation de la mémoire :
985
+ .br
986
+ \fIdict\fP * 7.5 (si \fIdict\fP <= 32 Mio);
987
+ .br
988
+ \fIdict\fP * 6.5 (si \fIdict\fP > 32 Mio)
989
+ .TP
990
+ \fBbt2\fP
991
+ Arbre binaire avec hachage de 2 octets
992
+ .br
993
+ Valeur minimale pour \fInice\fP : \fB2\fP
994
+ .br
995
+ Utilisation de la mémoire : \fIdict\fP * 9.5
996
+ .TP
997
+ \fBbt3\fP
998
+ Arbre binaire avec hachage de 2 et 3 octets
999
+ .br
1000
+ Valeur minimale pour \fInice\fP : \fB3\fP
1001
+ .br
1002
+ Utilisation de la mémoire :
1003
+ .br
1004
+ \fIdict\fP * 11.5 (si \fIdict\fP <= 16 Mio);
1005
+ .br
1006
+ \fIdict\fP * 9.5 + 64 MiB (si \fIdict\fP > 16 Mio)
1007
+ .TP
1008
+ \fBbt4\fP
1009
+ Arbre binaire avec hachage 2, 3 et 4 octets
1010
+ .br
1011
+ Valeur minimale pour \fInice\fP : \fB4\fP
1012
+ .br
1013
+ Utilisation de la mémoire :
1014
+ .br
1015
+ \fIdict\fP * 11.5 (si \fIdict\fP <= 32 Mio);
1016
+ .br
1017
+ \fIdict\fP * 10.5 (si \fIdict\fP > 32 Mio)
1018
+ .RE
1019
+ .TP
1020
+ \fBmode=\fP\fImode\fP
1021
+ Compression \fImode\fP specifies the method to analyze the data produced by the
1022
+ match finder. Supported \fImodes\fP are \fBfast\fP and \fBnormal\fP. The default is
1023
+ \fBfast\fP for \fIpresets\fP 0\(en3 and \fBnormal\fP for \fIpresets\fP 4\(en9.
1024
+ .IP ""
1025
+ Habituellement, \fBfast\fP est utilisé avec les chercheurs de correspondance de
1026
+ chaîne de hachage et \fBnormal\fP avec les chercheurs de correspondance d'arbre
1027
+ binaire. C'est aussi ce que font les \fIpréréglages\fP.
1028
+ .TP
1029
+ \fBnice=\fP\fInice\fP
1030
+ Spécifier ce qui est considéré comme une bonne longueur pour une
1031
+ correspondance. Une fois que la correspondance d'au moins \fInice\fP octets est
1032
+ trouvée, l'algorithme arrête de chercher de meilleures correspondances
1033
+ possibles.
1034
+ .IP ""
1035
+ \fINice\fP can be 2\(en273 bytes. Higher values tend to give better
1036
+ compression ratio at the expense of speed. The default depends on the
1037
+ \fIpreset\fP.
1038
+ .TP
1039
+ \fBdepth=\fP\fIprofondeur\fP
1040
+ Spécifier la profondeur de recherche maximale dans l'outil de recherche de
1041
+ correspondances. La valeur par défaut est \fB0\fP, ce qui fait que le
1042
+ compresseur détermine une \fIprofondeur\fP raisonnable en fonction de \fImf\fP et
1043
+ \fInice\fP.
1044
+ .IP ""
1045
+ Reasonable \fIdepth\fP for Hash Chains is 4\(en100 and 16\(en1000 for Binary
1046
+ Trees. Using very high values for \fIdepth\fP can make the encoder extremely
1047
+ slow with some files. Avoid setting the \fIdepth\fP over 1000 unless you are
1048
+ prepared to interrupt the compression in case it is taking far too long.
1049
+ .RE
1050
+ .IP ""
1051
+ Lors du décodage des flux bruts (\fB\-\-format=raw\fP), LZMA2 nécessite seulement
1052
+ la \fItaille\fP du dictionnaire. LZMA1 nécessite aussi \fIlc\fP, \fIlp\fP et \fIpb\fP.
1053
+ .TP
1054
+ \fB\-\-x86\fP[\fB=\fP\fIoptions\fP]
1055
+ .PD 0
1056
+ .TP
1057
+ \fB\-\-arm\fP[\fB=\fP\fIoptions\fP]
1058
+ .TP
1059
+ \fB\-\-armthumb\fP[\fB=\fP\fIoptions\fP]
1060
+ .TP
1061
+ \fB\-\-arm64\fP[\fB=\fP\fIoptions\fP]
1062
+ .TP
1063
+ \fB\-\-powerpc\fP[\fB=\fP\fIoptions\fP]
1064
+ .TP
1065
+ \fB\-\-ia64\fP[\fB=\fP\fIoptions\fP]
1066
+ .TP
1067
+ \fB\-\-sparc\fP[\fB=\fP\fIoptions\fP]
1068
+ .TP
1069
+ \fB\-\-riscv\fP[\fB=\fP\fIoptions\fP]
1070
+ .PD
1071
+ Ajouter un filtre branch/call/jump (BCJ) à la chaîne de filtres. Ces filtres
1072
+ ne peuvent être utilisés que s'ils ne sont pas le dernier filtre de la
1073
+ chaîne de filtrage.
1074
+ .IP ""
1075
+ A BCJ filter converts relative addresses in the machine code to their
1076
+ absolute counterparts. This doesn't change the size of the data but it
1077
+ increases redundancy, which can help LZMA2 to produce 0\(en15\ % smaller
1078
+ \&\fB.xz\fP file. The BCJ filters are always reversible, so using a BCJ filter
1079
+ for wrong type of data doesn't cause any data loss, although it may make the
1080
+ compression ratio slightly worse. The BCJ filters are very fast and use an
1081
+ insignificant amount of memory.
1082
+ .IP ""
1083
+ Ces filtres BCJ présentent des problèmes connus liés au taux de
1084
+ compression :
1085
+ .RS
1086
+ .IP \(bu 3
1087
+ Some types of files containing executable code (for example, object files,
1088
+ static libraries, and Linux kernel modules) have the addresses in the
1089
+ instructions filled with filler values. These BCJ filters will still do the
1090
+ address conversion, which will make the compression worse with these files.
1091
+ .IP \(bu 3
1092
+ If a BCJ filter is applied on an archive, it is possible that it makes the
1093
+ compression ratio worse than not using a BCJ filter. For example, if there
1094
+ are similar or even identical executables then filtering will likely make
1095
+ the files less similar and thus compression is worse. The contents of
1096
+ non\-executable files in the same archive can matter too. In practice one
1097
+ has to try with and without a BCJ filter to see which is better in each
1098
+ situation.
1099
+ .RE
1100
+ .IP ""
1101
+ Different instruction sets have different alignment: the executable file
1102
+ must be aligned to a multiple of this value in the input data to make the
1103
+ filter work.
1104
+ .RS
1105
+ .RS
1106
+ .PP
1107
+ .TS
1108
+ tab(;);
1109
+ l n l
1110
+ l n l.
1111
+ Filtre;Alignement;Notes
1112
+ x86;1;32 bits ou 64 bits x86
1113
+ ARM;4;
1114
+ ARM\-Thumb;2;
1115
+ ARM64;4;4096\-byte alignment is best
1116
+ PowerPC;4;Grand boutiste seulement
1117
+ IA\-64;16;Itanium
1118
+ SPARC;4;
1119
+ RISC\-V;2;
1120
+ .TE
1121
+ .RE
1122
+ .RE
1123
+ .IP ""
1124
+ Since the BCJ\-filtered data is usually compressed with LZMA2, the
1125
+ compression ratio may be improved slightly if the LZMA2 options are set to
1126
+ match the alignment of the selected BCJ filter. Examples:
1127
+ .RS
1128
+ .IP \(bu 3
1129
+ IA\-64 filter has 16\-byte alignment so \fBpb=4,lp=4,lc=0\fP is good with LZMA2
1130
+ (2^4=16).
1131
+ .IP \(bu 3
1132
+ RISC\-V code has 2\-byte or 4\-byte alignment depending on whether the file
1133
+ contains 16\-bit compressed instructions (the C extension). When 16\-bit
1134
+ instructions are used, \fBpb=2,lp=1,lc=3\fP or \fBpb=1,lp=1,lc=3\fP is good. When
1135
+ 16\-bit instructions aren't present, \fBpb=2,lp=2,lc=2\fP is the best.
1136
+ \fBreadelf \-h\fP can be used to check if "RVC" appears on the "Flags" line.
1137
+ .IP \(bu 3
1138
+ ARM64 is always 4\-byte aligned so \fBpb=2,lp=2,lc=2\fP is the best.
1139
+ .IP \(bu 3
1140
+ The x86 filter is an exception. It's usually good to stick to LZMA2's
1141
+ defaults (\fBpb=2,lp=0,lc=3\fP) when compressing x86 executables.
1142
+ .RE
1143
+ .IP ""
1144
+ Tous les filtres BCJ prennent en charge les mêmes \fIoptions\fP :
1145
+ .RS
1146
+ .TP
1147
+ \fBstart=\fP\fIdécalage\fP
1148
+ Spécifier le \fIdécalage\fP de départ qui est utilisé lors de la conversion
1149
+ entre les adresses relatives et absolues. Le \fIdécalage\fP doit être un
1150
+ multiple de l'alignement du filtre (voir la table ci\-dessus). Sa valeur par
1151
+ défaut est zéro. En pratique, cette dernière convient ; indiquer un
1152
+ \fIdécalage\fP personnalisé est la plupart du temps inutile.
1153
+ .RE
1154
+ .TP
1155
+ \fB\-\-delta\fP[\fB=\fP\fIoptions\fP]
1156
+ Ajouter le filtre Delta à la chaîne de filtres. Le filtre Delta ne peut être
1157
+ utilisé que s'il n'est pas le dernier filtre dans la chaîne.
1158
+ .IP ""
1159
+ Currently only simple byte\-wise delta calculation is supported. It can be
1160
+ useful when compressing, for example, uncompressed bitmap images or
1161
+ uncompressed PCM audio. However, special purpose algorithms may give
1162
+ significantly better results than Delta + LZMA2. This is true especially
1163
+ with audio, which compresses faster and better, for example, with
1164
+ \fBflac\fP(1).
1165
+ .IP ""
1166
+ \fIoptions\fP prises en charge :
1167
+ .RS
1168
+ .TP
1169
+ \fBdist=\fP\fIdistance\fP
1170
+ Specify the \fIdistance\fP of the delta calculation in bytes. \fIdistance\fP must
1171
+ be 1\(en256. The default is 1.
1172
+ .IP ""
1173
+ Par exemple, avec \fBdist=2\fP et une entrée huit octets A1 B1 A2 B3 A3 B5 A4
1174
+ B7, la sortie sera A1 B1 01 02 01 02 01 02.
1175
+ .RE
1176
+ .
1177
+ .SS "Autres options"
1178
+ .TP
1179
+ \fB\-q\fP, \fB\-\-quiet\fP
1180
+ Supprimer les avertissements et les notifications. Indiquer cela deux fois
1181
+ supprimera aussi les erreurs. Cette option n'a aucun effet sur le statut de
1182
+ sortie. Cela dit, même si un avertissement était supprimé, le statut de
1183
+ sortie indiquant un avertissement sera encore utilisé.
1184
+ .TP
1185
+ \fB\-v\fP, \fB\-\-verbose\fP
1186
+ Être bavard. Si l'erreur standard est connectée à un terminal, \fBxz\fP
1187
+ affichera une barre de progression. Indiquer \fB\-\-verbose\fP deux fois donnera
1188
+ une sortie encore plus bavarde.
1189
+ .IP ""
1190
+ La barre de progression montre l'information suivante :
1191
+ .RS
1192
+ .IP \(bu 3
1193
+ Le pourcentage de complétion est montré si la taille du fichier en entrée
1194
+ est connue. Néanmoins, le pourcentage ne peut pas être montré en cas de
1195
+ redirection.
1196
+ .IP \(bu 3
1197
+ Quantité de données compressées produites (compression) ou consommées
1198
+ (décompression).
1199
+ .IP \(bu 3
1200
+ Quantité de données non compressées consommées (compression) ou produites
1201
+ (décompression).
1202
+ .IP \(bu 3
1203
+ Le taux de compression, calculé en divisant la quantité de données
1204
+ compréssées déjà traitées par la quantité de données décompressées déjà
1205
+ traitées.
1206
+ .IP \(bu 3
1207
+ Vitesse de compression ou de décompression. Elle correspond à la quantité de
1208
+ données non compressées consommées (compression) ou produites
1209
+ (décompression) par seconde. Elle apparait quelques secondes après le début
1210
+ du traitement du fichier par \fBxz\fP.
1211
+ .IP \(bu 3
1212
+ Temps écoulé dans le format M:SS ou H:MM:SS.
1213
+ .IP \(bu 3
1214
+ Estimated remaining time is shown only when the size of the input file is
1215
+ known and a couple of seconds have already passed since \fBxz\fP started
1216
+ processing the file. The time is shown in a less precise format which never
1217
+ has any colons, for example, 2 min 30 s.
1218
+ .RE
1219
+ .IP ""
1220
+ When standard error is not a terminal, \fB\-\-verbose\fP will make \fBxz\fP print
1221
+ the filename, compressed size, uncompressed size, compression ratio, and
1222
+ possibly also the speed and elapsed time on a single line to standard error
1223
+ after compressing or decompressing the file. The speed and elapsed time are
1224
+ included only when the operation took at least a few seconds. If the
1225
+ operation didn't finish, for example, due to user interruption, also the
1226
+ completion percentage is printed if the size of the input file is known.
1227
+ .TP
1228
+ \fB\-Q\fP, \fB\-\-no\-warn\fP
1229
+ Ne pas mettre l'état de sortie à \fB2\fP même si une condition méritant un
1230
+ avertissement a été détectée. Cette option n'affecte pas le niveau de
1231
+ verbosité, néanmoins, les deux options \fB\-\-quiet\fP et \fB\-\-no\-warn\fP doivent
1232
+ être utilisées pour ne pas afficher d'avertissements, ni altérer le statut
1233
+ de sortie.
1234
+ .TP
1235
+ \fB\-\-robot\fP
1236
+ Afficher les messages dans un format analysable par une machine. Ceci est
1237
+ destiné à faciliter l'écriture des frontaux qui voudraient utiliser \fBxz\fP
1238
+ plutôt que liblzma, ce qui pourrait être le cas pour différents scripts. La
1239
+ sortie avec cette option activée est destinée à rester stable sur les
1240
+ différentes versions de \fBxz\fP. Consulter le paragraphe \fBROBOT MODE\fP pour
1241
+ les détails.
1242
+ .TP
1243
+ \fB\-\-info\-memory\fP
1244
+ Display, in human\-readable format, how much physical memory (RAM) and how
1245
+ many processor threads \fBxz\fP thinks the system has and the memory usage
1246
+ limits for compression and decompression, and exit successfully.
1247
+ .TP
1248
+ \fB\-h\fP, \fB\-\-help\fP
1249
+ Afficher un message d'aide décrivant les options les plus couramment
1250
+ utilisées et quitter.
1251
+ .TP
1252
+ \fB\-H\fP, \fB\-\-long\-help\fP
1253
+ Afficher un message d'aide décrivant toutes les options de \fBxz\fP et quitter.
1254
+ .TP
1255
+ \fB\-V\fP, \fB\-\-version\fP
1256
+ Afficher le numéro de version de \fBxz\fP et de liblzma dans un format lisible
1257
+ par un humain. Pour obtenir une sortie analysable par la machine, spécifiez
1258
+ \fB\-\-robot\fP avant \fB\-\-version\fP.
1259
+ .
1260
+ .SH "MODE ROBOT"
1261
+ The robot mode is activated with the \fB\-\-robot\fP option. It makes the output
1262
+ of \fBxz\fP easier to parse by other programs. Currently \fB\-\-robot\fP is
1263
+ supported only together with \fB\-\-list\fP, \fB\-\-filters\-help\fP, \fB\-\-info\-memory\fP,
1264
+ and \fB\-\-version\fP. It will be supported for compression and decompression in
1265
+ the future.
1266
+ .
1267
+ .SS "Mode liste"
1268
+ \fBxz \-\-robot \-\-list\fP utilise une sortie séparée par des tabulations. La
1269
+ première colonne de toutes les lignes possède une chaîne qui indique le type
1270
+ d'information trouvée sur cette ligne :
1271
+ .TP
1272
+ \fBname\fP
1273
+ C'est toujours la première ligne au début de la liste d'un fichier. La
1274
+ seconde colonne de la ligne est le nom de fichier.
1275
+ .TP
1276
+ \fBfile\fP
1277
+ Cette ligne contient l'information globale sur le fichier \fB.xz\fP. Cette
1278
+ ligne est toujours écrite après la ligne \fBname\fP.
1279
+ .TP
1280
+ \fBstream\fP
1281
+ Ce type de ligne n'est utilisée que lorsque \fB \-\-verbose\fP a été indiquée. Il
1282
+ y a autant de lignes \fBstream\fP qu'il y a de flux dans le fichier \fB.xz\fP.
1283
+ .TP
1284
+ \fBblock\fP
1285
+ Ce type de ligne n'est utilisé seulement lorsque \fB\-\-verbose\fP a été
1286
+ indiquée. Il y a autant de lignes \fBblock\fP qu'il y a de blocs dans le
1287
+ fichier \fB.xz\fP. Les lignes \fBblock\fP sont affichées après toutes les lignes
1288
+ \fBstream\fP ; les différents types de lignes ne sont pas imbriqués.
1289
+ .TP
1290
+ \fBsummary\fP
1291
+ Ce type de ligne n'est utilisé que lorsque \fB\-\-verbose\fP a été indiqué deux
1292
+ fois. Cette ligne est affichée après toutes les lignes \fBblock\fP. Comme la
1293
+ ligne \fBfile\fP, la ligne \fBsummary\fP contient l'information globale sur le
1294
+ fichier \fB.xz\fP.
1295
+ .TP
1296
+ \fBtotals\fP
1297
+ Cette ligne est toujours la toute dernière ligne de la sortie. Elle affiche
1298
+ les comptes et les tailles totaux.
1299
+ .PP
1300
+ Les colonnes des lignes \fBfile\fP :
1301
+ .PD 0
1302
+ .RS
1303
+ .IP 2. 4
1304
+ Nombre de flux dans le fichier
1305
+ .IP 3. 4
1306
+ Nombre total de blocs dans le ou les flux.
1307
+ .IP 4. 4
1308
+ Taille compressée du fichier
1309
+ .IP 5. 4
1310
+ Taille décompressée du fichier
1311
+ .IP 6. 4
1312
+ Compression ratio, for example, \fB0.123\fP. If ratio is over 9.999, three
1313
+ dashes (\fB\-\-\-\fP) are displayed instead of the ratio.
1314
+ .IP 7. 4
1315
+ Liste de noms de contrôles d'intégrité séparés par des virgules. Les chaînes
1316
+ suivantes sont utilisées pour les types de vérification connus : \fBNone\fP,
1317
+ \fBCRC32\fP, \fBCRC64\fP et \fBSHA256\fP. Pour le types de vérification inconnus,
1318
+ \fBUnknown\-\fP\fIN\fP est utilisé, où \fIN\fP est un identifiant de vérification sous
1319
+ la forme d'un nombre décimal (un ou deux chiffres).
1320
+ .IP 8. 4
1321
+ Taille totale du remplissage du flux dans le fichier
1322
+ .RE
1323
+ .PD
1324
+ .PP
1325
+ Les colonnes des lignes \fBstream\fP :
1326
+ .PD 0
1327
+ .RS
1328
+ .IP 2. 4
1329
+ Numéro de flux (le premier flux a le numéro 1)
1330
+ .IP 3. 4
1331
+ Nombre de blocs dans le flux
1332
+ .IP 4. 4
1333
+ Décalage de départ compressé
1334
+ .IP 5. 4
1335
+ Décalage de départ décompressé
1336
+ .IP 6. 4
1337
+ Taille compressée (ne comprend pas le remplissage du flux)
1338
+ .IP 7. 4
1339
+ Taille décompressée
1340
+ .IP 8. 4
1341
+ Taux de compression
1342
+ .IP 9. 4
1343
+ Nom de la vérification d'intégrité
1344
+ .IP 10. 4
1345
+ Taille du remplissage de flux
1346
+ .RE
1347
+ .PD
1348
+ .PP
1349
+ Les colonnes des lignes \fBblock\fP :
1350
+ .PD 0
1351
+ .RS
1352
+ .IP 2. 4
1353
+ Numéro du flux qui contient ce bloc
1354
+ .IP 3. 4
1355
+ Numéro du bloc relatif au commencement du flux (le premier bloc a pour
1356
+ numéro 1)
1357
+ .IP 4. 4
1358
+ Numéro du bloc relatif au début du fichier
1359
+ .IP 5. 4
1360
+ Décalage de départ compressé relatif au début du fichier
1361
+ .IP 6. 4
1362
+ Décalage de départ décompressé relatif au début du fichier
1363
+ .IP 7. 4
1364
+ Taille compressée totale du bloc (en\-têtes inclus)
1365
+ .IP 8. 4
1366
+ Taille décompressée
1367
+ .IP 9. 4
1368
+ Taux de compression
1369
+ .IP 10. 4
1370
+ Nom de la vérification d'intégrité
1371
+ .RE
1372
+ .PD
1373
+ .PP
1374
+ Si \fB\-\-verbose\fP a été indiqué deux fois, les colonnes additionnelles sont
1375
+ inclues sur les lignes \fBblock\fP. Elles ne sont pas affichées avec un seul
1376
+ \fB\-\-verbose\fP, car l'obtention de ces informations nécessite de nombreuses
1377
+ recherches et peut donc être lente :
1378
+ .PD 0
1379
+ .RS
1380
+ .IP 11. 4
1381
+ Valeur de la vérification d'intégrité en hexadécimal
1382
+ .IP 12. 4
1383
+ Taille d'en\-tête de bloc
1384
+ .IP 13. 4
1385
+ Drapeaux du bloc : \fBc\fP indique que la taille compressée est présente, et
1386
+ \fBu\fP indique que la taille décompréssée est présente. Si le drapeau n'est
1387
+ pas indiqué, un tiret (\fB\-\fP) est affiché à la place pour que la longueur de
1388
+ la chaîne reste fixe. De nouveaux drapeaux pourraient être ajoutés à la fin
1389
+ de la chaîne dans le futur.
1390
+ .IP 14. 4
1391
+ Taille des données effectivement compressées dans le bloc (en excluant
1392
+ l'en\-tête de bloc, le remplissage de bloc et les champs de vérification).
1393
+ .IP 15. 4
1394
+ Quantité de mémoire (en octets) nécessaire pour décompresser ce bloc avec
1395
+ cette version de \fBxz\fP.
1396
+ .IP 16. 4
1397
+ Chaîne de filtrage. Remarquez que la plupart des options utilisées au moment
1398
+ de la compression ne peuvent pas être connues, car seules les options
1399
+ nécessaires pour la décompression sont stockées dans les en\-têtes \fB.xz\fP.
1400
+ .RE
1401
+ .PD
1402
+ .PP
1403
+ Les colonnes des lignes \fBsummary\fP :
1404
+ .PD 0
1405
+ .RS
1406
+ .IP 2. 4
1407
+ Quantité de mémoire (en octets) nécessaire pour décompresser ce fichier avec
1408
+ cette version de \fBxz\fP.
1409
+ .IP 3. 4
1410
+ \fByes\fP ou \fBno\fP indique si tous les en\-têtes de bloc stockent à la fois la
1411
+ taille compressée et la taille décompressée.
1412
+ .PP
1413
+ \fIDepuis\fP \fBxz\fP \fI5.1.2alpha:\fP
1414
+ .IP 4. 4
1415
+ Version minimale de \fBxz\fP nécessaire pour décompresser le fichier.
1416
+ .RE
1417
+ .PD
1418
+ .PP
1419
+ Les colonnes de la ligne \fBtotals\fP :
1420
+ .PD 0
1421
+ .RS
1422
+ .IP 2. 4
1423
+ Nombre de flux
1424
+ .IP 3. 4
1425
+ Nombre de blocs
1426
+ .IP 4. 4
1427
+ Taille compressée
1428
+ .IP 5. 4
1429
+ Taille décompressée
1430
+ .IP 6. 4
1431
+ Taux de compression moyen
1432
+ .IP 7. 4
1433
+ Liste séparée par des virgules des noms de vérification d'intégrité qui
1434
+ étaient présents dans les fichiers
1435
+ .IP 8. 4
1436
+ Taille de remplissage de flux
1437
+ .IP 9. 4
1438
+ Nombre de fichiers. Permet de garder l'ordre des colonnes précédentes comme
1439
+ sur les lignes \fBfile\fP.
1440
+ .PD
1441
+ .RE
1442
+ .PP
1443
+ Si \fB\-\-verbose\fP a été indiqué deux fois, des colonnes supplémentaires sont
1444
+ incluses sur la ligne \fBtotals\fP :
1445
+ .PD 0
1446
+ .RS
1447
+ .IP 10. 4
1448
+ Quantité maximale de mémoire (en octets) nécessaire pour décompresser les
1449
+ fichiers avec cette version de \fBxz\fP.
1450
+ .IP 11. 4
1451
+ \fByes\fP ou \fBno\fP indique si tous les en\-têtes de bloc stockent à la fois la
1452
+ taille compressée et la taille décompressée.
1453
+ .PP
1454
+ \fIDepuis\fP \fBxz\fP \fI5.1.2alpha:\fP
1455
+ .IP 12. 4
1456
+ Version minimale de \fBxz\fP nécessaire pour décompresser le fichier.
1457
+ .RE
1458
+ .PD
1459
+ .PP
1460
+ Les versions futures pourront ajouter de nouveaux types de lignes et de
1461
+ nouvelles colonnes pourront être ajoutées aux types de lignes existants,
1462
+ mais les colonnes existantes ne seront pas modifiées.
1463
+ .
1464
+ .SS "Filters help"
1465
+ \fBxz \-\-robot \-\-filters\-help\fP prints the supported filters in the following
1466
+ format:
1467
+ .PP
1468
+ \fIfilter\fP\fB:\fP\fIoption\fP\fB=<\fP\fIvalue\fP\fB>,\fP\fIoption\fP\fB=<\fP\fIvalue\fP\fB>\fP...
1469
+ .TP
1470
+ \fIfilter\fP
1471
+ Name of the filter
1472
+ .TP
1473
+ \fIoption\fP
1474
+ Name of a filter specific option
1475
+ .TP
1476
+ \fIvalue\fP
1477
+ Numeric \fIvalue\fP ranges appear as \fB<\fP\fImin\fP\fB\-\fP\fImax\fP\fB>\fP. String
1478
+ \fIvalue\fP choices are shown within \fB< >\fP and separated by a \fB|\fP
1479
+ character.
1480
+ .PP
1481
+ Each filter is printed on its own line.
1482
+ .
1483
+ .SS "Information de limite de mémoire"
1484
+ \fBxz \-\-robot \-\-info\-memory\fP prints a single line with multiple tab\-separated
1485
+ columns:
1486
+ .IP 1. 4
1487
+ Total amount of physical memory (RAM) in bytes.
1488
+ .IP 2. 4
1489
+ Memory usage limit for compression in bytes (\fB\-\-memlimit\-compress\fP). A
1490
+ special value of \fB0\fP indicates the default setting which for
1491
+ single\-threaded mode is the same as no limit.
1492
+ .IP 3. 4
1493
+ Memory usage limit for decompression in bytes (\fB\-\-memlimit\-decompress\fP). A
1494
+ special value of \fB0\fP indicates the default setting which for
1495
+ single\-threaded mode is the same as no limit.
1496
+ .IP 4. 4
1497
+ Since \fBxz\fP 5.3.4alpha: Memory usage for multi\-threaded decompression in
1498
+ bytes (\fB\-\-memlimit\-mt\-decompress\fP). This is never zero because a
1499
+ system\-specific default value shown in the column 5 is used if no limit has
1500
+ been specified explicitly. This is also never greater than the value in the
1501
+ column 3 even if a larger value has been specified with
1502
+ \fB\-\-memlimit\-mt\-decompress\fP.
1503
+ .IP 5. 4
1504
+ Since \fBxz\fP 5.3.4alpha: A system\-specific default memory usage limit that is
1505
+ used to limit the number of threads when compressing with an automatic
1506
+ number of threads (\fB\-\-threads=0\fP) and no memory usage limit has been
1507
+ specified (\fB\-\-memlimit\-compress\fP). This is also used as the default value
1508
+ for \fB\-\-memlimit\-mt\-decompress\fP.
1509
+ .IP 6. 4
1510
+ Since \fBxz\fP 5.3.4alpha: Number of available processor threads.
1511
+ .PP
1512
+ Dans le futur, la sortie de \fBxz \-\-robot \-\-info\-memory\fP pourrait avoir plus
1513
+ de colonnes, mais jamais plus qu'une ligne unique.
1514
+ .
1515
+ .SS Version
1516
+ \fBxz \-\-robot \-\-version\fP prints the version number of \fBxz\fP and liblzma in
1517
+ the following format:
1518
+ .PP
1519
+ \fBXZ_VERSION=\fP\fIXYYYZZZS\fP
1520
+ .br
1521
+ \fBLIBLZMA_VERSION=\fP\fIXYYYZZZS\fP
1522
+ .TP
1523
+ \fIX\fP
1524
+ Version majeure.
1525
+ .TP
1526
+ \fIYYY\fP
1527
+ Version mineure. Les numéros pairs sont stables. Les numéros impairs sont
1528
+ des versions alpha ou beta.
1529
+ .TP
1530
+ \fIZZZ\fP
1531
+ Niveau de correctif pour les options stables ou juste un compteur pour les
1532
+ options de développement.
1533
+ .TP
1534
+ \fIS\fP
1535
+ Stabilité. 0 est alpha, 1 est bêta et 2 est stable. \fIS\fP devrait toujours
1536
+ être 2 quand \fIYYY\fP est pair.
1537
+ .PP
1538
+ \fIXYYYZZZS\fP sont identiques sur les deux lignes si \fBxz\fP et liblzma sont
1539
+ issus de la même version d'utilitaires XZ.
1540
+ .PP
1541
+ Exemples : 4.999.9beta est \fB49990091\fP et 5.0.0 est \fB50000002\fP.
1542
+ .
1543
+ .SH "STATUT DE SORTIE"
1544
+ .TP
1545
+ \fB0\fP
1546
+ Tout est bon.
1547
+ .TP
1548
+ \fB1\fP
1549
+ Une erreur est survenue.
1550
+ .TP
1551
+ \fB2\fP
1552
+ Quelquechose méritant un avertissement s'est produit, mais aucune erreur
1553
+ véritable n'est survenue.
1554
+ .PP
1555
+ Les notifications (pas les avertissements ou les erreurs) affichées sur
1556
+ l'erreur standard n'affectent pas le statut de sortie.
1557
+ .
1558
+ .SH ENVIRONNEMENT
1559
+ \fBxz\fP analyse les listes d'options séparées par des espaces à partir des
1560
+ variables d'environnement \fBXZ_DEFAULTS\fP et \fBXZ_OPT\fP, dans cet ordre, avant
1561
+ d'analyser les options de la ligne de commandes. Remarquez que seules les
1562
+ options sont analysées depuis l'environnement des variables ; toutes les
1563
+ non\-options sont ignorées silencieusement. L'analyse est faite avec
1564
+ \fBgetopt_long\fP(3) qui est aussi utilisé pour les arguments de la ligne de
1565
+ commandes.
1566
+ .TP
1567
+ \fBXZ_DEFAULTS\fP
1568
+ Options par défaut propres à l'utilisateur ou pour tout le système. Elles
1569
+ sont le plus souvent définies dans un script d'initialisation de
1570
+ l'interpréteur pour activer le limiteur d'utilisation de la mémoire de \fBxz\fP
1571
+ par défaut. A part pour les scripts d'initialisation de l'interpréteur ou
1572
+ des cas similaires, les sripts ne doivent jamais définir ou désactiver
1573
+ \fBXZ_DEFAULTS\fP.
1574
+ .TP
1575
+ \fBXZ_OPT\fP
1576
+ This is for passing options to \fBxz\fP when it is not possible to set the
1577
+ options directly on the \fBxz\fP command line. This is the case when \fBxz\fP is
1578
+ run by a script or tool, for example, GNU \fBtar\fP(1):
1579
+ .RS
1580
+ .RS
1581
+ .PP
1582
+ .nf
1583
+ \f(CRXZ_OPT=\-2v tar caf foo.tar.xz foo\fR
1584
+ .fi
1585
+ .RE
1586
+ .RE
1587
+ .IP ""
1588
+ Scripts may use \fBXZ_OPT\fP, for example, to set script\-specific default
1589
+ compression options. It is still recommended to allow users to override
1590
+ \fBXZ_OPT\fP if that is reasonable. For example, in \fBsh\fP(1) scripts one may
1591
+ use something like this:
1592
+ .RS
1593
+ .RS
1594
+ .PP
1595
+ .nf
1596
+ \f(CRXZ_OPT=${XZ_OPT\-"\-7e"}
1597
+ export XZ_OPT\fR
1598
+ .fi
1599
+ .RE
1600
+ .RE
1601
+ .
1602
+ .SH "Compatibilité des utilitaires LZMA"
1603
+ La syntaxe de la ligne de commande de \fBxz\fP est quasimment un sur\-ensemble
1604
+ de \fBlzma\fP, \fBunlzma\fP et \fBlzcat\fP comme ils sont trouvés dans les
1605
+ utilitaires LZMA 4.32.x . Dans la pluspart des cas, il est possible de
1606
+ remplacer les outils LZMA par les outils XZ sans casser les scripts
1607
+ existants. Il existe cependant certaines incompatibilités qui peuvent
1608
+ parfois poser des problèmes.
1609
+ .
1610
+ .SS "Niveaux de préréglage de la compression"
1611
+ La numérotation des préréglages de niveau de compression est différente
1612
+ entre les outils \fBxz\fP et LZMA. La différence la plus importante est la
1613
+ manière dont les tailles de dictionnaire sont affectées aux différents
1614
+ préréglages. La taille de dictionnaire est à peu près égale à celle
1615
+ d'utilisation de la mémoire de la décompression.
1616
+ .RS
1617
+ .PP
1618
+ .TS
1619
+ tab(;);
1620
+ c c c
1621
+ c n n.
1622
+ Niveau;xz;Utilitaires LZMA
1623
+ \-0;256 KiB;N/A
1624
+ \-1;1 MiB;64 KiB
1625
+ \-2;2 MiB;1 MiB
1626
+ \-3;4 MiB;512 KiB
1627
+ \-4;4 MiB;1 MiB
1628
+ \-5;8 MiB;2 MiB
1629
+ \-6;8 MiB;4 MiB
1630
+ \-7;16 MiB;8 MiB
1631
+ \-8;32 MiB;16 MiB
1632
+ \-9;64 MiB;32 MiB
1633
+ .TE
1634
+ .RE
1635
+ .PP
1636
+ Les différences de tailles des dictionnaires affectent aussi l'utilisation
1637
+ de la mémoire du compresseur, mais il y a quelques autres différences entre
1638
+ les outils LZMA et les outils XZ, qui rendent la différence encore plus
1639
+ grande :
1640
+ .RS
1641
+ .PP
1642
+ .TS
1643
+ tab(;);
1644
+ c c c
1645
+ c n n.
1646
+ Niveau;xz;Utilitaires LZMA 4.32.x
1647
+ \-0;3 MiB;N/A
1648
+ \-1;9 MiB;2 MiB
1649
+ \-2;17 MiB;12 MiB
1650
+ \-3;32 MiB;12 MiB
1651
+ \-4;48 MiB;16 MiB
1652
+ \-5;94 MiB;26 MiB
1653
+ \-6;94 MiB;45 MiB
1654
+ \-7;186 MiB;83 MiB
1655
+ \-8;370 MiB;159 MiB
1656
+ \-9;674 MiB;311 MiB
1657
+ .TE
1658
+ .RE
1659
+ .PP
1660
+ Le niveau de préréglage par défaut dans les outils LZMA est \fB\-7\fP alors que
1661
+ pour les outils XZ c'est \fB\-6\fP, les deux utilisent ainsi un dictionnaire de
1662
+ 8 Mio par défaut.
1663
+ .
1664
+ .SS "Fichiers .lzma en flux ou non"
1665
+ The uncompressed size of the file can be stored in the \fB.lzma\fP header.
1666
+ LZMA Utils does that when compressing regular files. The alternative is to
1667
+ mark that uncompressed size is unknown and use end\-of\-payload marker to
1668
+ indicate where the decompressor should stop. LZMA Utils uses this method
1669
+ when uncompressed size isn't known, which is the case, for example, in
1670
+ pipes.
1671
+ .PP
1672
+ \fBxz\fP prend en charge la décompression des fichiers \fB.lzma\fP avec ou sans
1673
+ marqueur de fin de charge utile, mais tous les fichiers \fB.lzma\fP créés par
1674
+ \fBxz\fP utiliseront un marqueur de fin de charge utile et ont la taille non
1675
+ compréssée marquée comme inconnue dans l'en\-tête \fB.lzma\fP. Cela peut être un
1676
+ problème dans quelques situations inhabituelles. Par exemple, un
1677
+ décompresseur \fB.lzma\fP dans un périphérique embarqué pourrait ne fonctionner
1678
+ qu'avec des fichiers dont la taille non comprimée est connue. Si vous vous
1679
+ heurtez à ce problème, vous devez utiliser les utilitaires LZMA ou LZMA SDK
1680
+ pour créer des fichiers \fB.lzma\fP avec une taille non compressée connue.
1681
+ .
1682
+ .SS "Fichiers .lzma non pris en charge"
1683
+ Le format \fB.lzma\fP autorise des valeurs \fIlc\fP jusqu'à 8, et des valeurs
1684
+ \fIlp\fP jusqu'à 4. Les outils LZMA peuvent décompresser des fichiers avec tous
1685
+ les \fIlc\fP et \fIlp\fP, mais créez toujours les fichiers avec \fBlc=3\fP et
1686
+ \fBlp=0\fP. Créer des fichiers avec d'autres valeurs \fIlc\fP et \fIlp\fP est
1687
+ possible avec \fBxz\fP et avec LZMA SDK.
1688
+ .PP
1689
+ L'implémentation du filtre LZMA1 dans liblzma nécessite que la somme de
1690
+ \fIlc\fP et \fIlp\fP soit inférieure ou égale à 4. Ainsi, les fichiers \fB.lzma\fP
1691
+ qui excèdent cette limitation ne peuvent pas être décompressés avec \fBxz\fP.
1692
+ .PP
1693
+ Les outils LZMA créent seulement des fichiers \fB.lzma\fP qui ont une taille de
1694
+ dictionnaire de 2^\fIn\fP (une puissance de 2) mais acceptent les fichiers avec
1695
+ toutes les tailles de dictionnaire. Libzlma n'accepte que les fichiers
1696
+ \&\fB.lzma\fP qui ont une taille dictionnaire de 2^\fIn\fP ou
1697
+ 2^\fIn\fP + 2^(\fIn\fP\-1). Cela afin de diminuer les faux positifs lors de la
1698
+ détection des fichiers \fB.lzma\fP.
1699
+ .PP
1700
+ Ces limitations ne devraient pas poser problème en pratique, car
1701
+ pratiquement tous les fichiers \fB.lzma\fP ont été compressés avec des réglages
1702
+ que liblzma accepte.
1703
+ .
1704
+ .SS "Déchets excédentaires"
1705
+ Lors de la décompession, l'utilitaire LZMA ignore silencieusement tout ce
1706
+ qui est après le premier flux \fB.lzma\fP. Dans la majorité des situations,
1707
+ c'est un bogue. Cela veut dire aussi que les outils LZMA ne gèrent pas la
1708
+ décompression de fichiers \fB.lzma\fP concaténés.
1709
+ .PP
1710
+ S'il reste des données après le premier flux \fB.lzma\fP, \fBxz\fP considère que
1711
+ le fichier est corrompu sauf si \fB\-\-single\-stream\fP a été utilisé. Cela peut
1712
+ casser des scripts obscurs qui ont supposé que les déchets de fin de ligne
1713
+ sont ignorés.
1714
+ .
1715
+ .SH NOTES
1716
+ .
1717
+ .SS "La sortie compressée peut varier"
1718
+ La sortie compressée exacte produite par les même fichiers non compressés en
1719
+ entrée peut varier en fonction des différentes versions de l'utilitaire XZ,
1720
+ même si les options de compression sont identiques. En effet, il est
1721
+ possible d'améliorer l'encodeur (compression plus rapide ou meilleure) sans
1722
+ affecter le format du fichier. La sortie peut même varier entre différentes
1723
+ compilations de la même version d'utilitaire XZ, si des options de
1724
+ construction différentes sont utilisées.
1725
+ .PP
1726
+ Cela signifie qu'une fois que \fB\-\-rsyncable\fP a été implémenté, les fichiers
1727
+ résultants ne seront pas nécessairement synchronisables avec rsync à moins
1728
+ que les nouveaux et anciens fichiers n'aient été compressés avec la même
1729
+ version de xz. Ce problème peut être résolu si une partie de
1730
+ l'implémentation est gelée pour garantir la stabilité de la sortie rsyncable
1731
+ à travers les versions de xz.
1732
+ .
1733
+ .SS "Décompresseurs .xz embarqués"
1734
+ Les implémentations de décompresseur embarqué comme XZ Embedded ne gèrent
1735
+ pas nécessairement les fichiers créés avec d'autres types de \fIvérification\fP
1736
+ d'intégrité que \fBnone\fP et \fBCRC32\fP. Comme la valeur par défaut est
1737
+ \fB\-\-check=crc64\fP, vous devez utiliser \fB\-\-check=none\fP ou \fB\-\-check=crc32\fP
1738
+ lors de la création de fichiers pour les systèmes embarqués.
1739
+ .PP
1740
+ En dehors des systèmes embarqués, tous les décompresseurs de format \fB.xz\fP
1741
+ gèrent tous les types de \fIvérification\fP ou sont au moins capables de
1742
+ décompresser le fichier sans effectuer la vérification d'intégrité si ce
1743
+ type de \fIvérification\fP particulière n'est pas pris en charge.
1744
+ .PP
1745
+ XZ Embedded prend en charge les filtres BCJ, mais seulement avec le décalage
1746
+ de départ par défaut.
1747
+ .
1748
+ .SH EXEMPLES
1749
+ .
1750
+ .SS Bases
1751
+ Compresser le fichier \fItoto\fP en \fItoto.xz\fP en utilisant le niveau de
1752
+ compression par défaut (\fB\-6\fP) et supprimer \fItoto\fP si la compression
1753
+ réussit :
1754
+ .RS
1755
+ .PP
1756
+ .nf
1757
+ \f(CRxz foo\fR
1758
+ .fi
1759
+ .RE
1760
+ .PP
1761
+ Décompresser \fIbidule.xz\fP en \fIbidule\fP et ne pas supprimer \fIbidule.xz\fP même
1762
+ si la compression réussit :
1763
+ .RS
1764
+ .PP
1765
+ .nf
1766
+ \f(CRxz \-dk bar.xz\fR
1767
+ .fi
1768
+ .RE
1769
+ .PP
1770
+ Create \fIbaz.tar.xz\fP with the preset \fB\-4e\fP (\fB\-4 \-\-extreme\fP), which is
1771
+ slower than the default \fB\-6\fP, but needs less memory for compression and
1772
+ decompression (48\ MiB and 5\ MiB, respectively):
1773
+ .RS
1774
+ .PP
1775
+ .nf
1776
+ \f(CRtar cf \- baz | xz \-4e > baz.tar.xz\fR
1777
+ .fi
1778
+ .RE
1779
+ .PP
1780
+ Un mélange de fichiers compressés et non compressés peuvent être
1781
+ décompressés vers la sortie standard avec une simple commande :
1782
+ .RS
1783
+ .PP
1784
+ .nf
1785
+ \f(CRxz \-dcf a.txt b.txt.xz c.txt d.txt.lzma > abcd.txt\fR
1786
+ .fi
1787
+ .RE
1788
+ .
1789
+ .SS "Compression en parallèle de plusieurs fichiers"
1790
+ Sur GNU et *BSD, \fBfind\fP(1) et \fBxargs\fP(1) peuvent être utilisés pour mettre
1791
+ en parallèle la compression de plusieurs fichiers :
1792
+ .RS
1793
+ .PP
1794
+ .nf
1795
+ \f(CRfind . \-type f \e! \-name '*.xz' \-print0 \e
1796
+ | xargs \-0r \-P4 \-n16 xz \-T1\fR
1797
+ .fi
1798
+ .RE
1799
+ .PP
1800
+ L'option \fBP\fP passée à \fBxargs\fP(1) fixe le nombre de processus \fBxz\fP en
1801
+ parallèles. La meilleure valeur pour l'option \fBn\fP dépend du nombre de
1802
+ fichiers à compresser. S\-il n'y a que quelques fichiers, la valeur sera
1803
+ probablement 1 ; avec des dizaines de milliers de fichiers, 100 ou même plus
1804
+ serait approprié pour réduire le nombre de processus \fBxz\fP que \fBxargs\fP(1)
1805
+ créera éventuellement.
1806
+ .PP
1807
+ L'option \fB\-T1\fP de \fBxz\fP est là pour le forcer en mode mono\-thread, car
1808
+ \fBxargs\fP(1) est utilisé pour contrôler la quantité de mise en parallèle.
1809
+ .
1810
+ .SS "Mode robot"
1811
+ Calculer combien d'octets ont été économisés au total après avoir compressé
1812
+ plusieurs fichiers :
1813
+ .RS
1814
+ .PP
1815
+ .nf
1816
+ \f(CRxz \-\-robot \-\-list *.xz | awk '/^totals/{print $5\-$4}'\fR
1817
+ .fi
1818
+ .RE
1819
+ .PP
1820
+ Un script peut vouloir savoir qu'il utilise une version suffisamment récente
1821
+ de \fBxz\fP. Le script \fBsh\fP(1) suivant vérifie que le numéro de version de
1822
+ l'outil \fBxz\fP soit au minimum 5.0.0. Cette méthode est compatible avec les
1823
+ vieilles versions bêta, qui ne gèrent pas l'option \fB\-\-robot\fP :
1824
+ .RS
1825
+ .PP
1826
+ .nf
1827
+ \f(CRif ! eval "$(xz \-\-robot \-\-version 2> /dev/null)" ||
1828
+ [ "$XZ_VERSION" \-lt 50000002 ]; then
1829
+ echo "Your xz is too old."
1830
+ fi
1831
+ unset XZ_VERSION LIBLZMA_VERSION\fR
1832
+ .fi
1833
+ .RE
1834
+ .PP
1835
+ Régler une limite d'utilisation de la mémoire pour la décompression en
1836
+ utilisant \fBXZ_OPT\fP, mais si une limite a déjà été définie, ne pas
1837
+ l'augmenter :
1838
+ .RS
1839
+ .PP
1840
+ .nf
1841
+ \f(CRNEWLIM=$((123 << 20))\ \ # 123 MiB
1842
+ OLDLIM=$(xz \-\-robot \-\-info\-memory | cut \-f3)
1843
+ if [ $OLDLIM \-eq 0 \-o $OLDLIM \-gt $NEWLIM ]; then
1844
+ XZ_OPT="$XZ_OPT \-\-memlimit\-decompress=$NEWLIM"
1845
+ export XZ_OPT
1846
+ fi\fR
1847
+ .fi
1848
+ .RE
1849
+ .
1850
+ .SS "Chaînes de filtres de compresseur personnalisées"
1851
+ L'utilisation la plus simple des chaînes de filtres personnalisées est la
1852
+ personnalisation d'un préréglage LZMA2. Cela peut être utile, car les
1853
+ préréglages ne couvrent qu'un sous\-ensemble des réglages de compression
1854
+ potentiellement utiles.
1855
+ .PP
1856
+ Les colonnes CompCPU des tableaux des descriptions des options \fB\-0\fP à \fB\-9\fP
1857
+ et \fB\-\-extreme\fP sont utiles lors de la personnalisation des préréglages
1858
+ LZMA2. Voici les parties pertinentes recueillies à partir de ces deux
1859
+ tableaux :
1860
+ .RS
1861
+ .PP
1862
+ .TS
1863
+ tab(;);
1864
+ c c
1865
+ n n.
1866
+ Préréglage;CompCPU
1867
+ \-0;0
1868
+ \-1;1
1869
+ \-2;2
1870
+ \-3;3
1871
+ \-4;4
1872
+ \-5;5
1873
+ \-6;6
1874
+ \-5e;7
1875
+ \-6e;8
1876
+ .TE
1877
+ .RE
1878
+ .PP
1879
+ If you know that a file requires somewhat big dictionary (for example, 32\ MiB) to compress well, but you want to compress it quicker than \fBxz \-8\fP
1880
+ would do, a preset with a low CompCPU value (for example, 1) can be
1881
+ modified to use a bigger dictionary:
1882
+ .RS
1883
+ .PP
1884
+ .nf
1885
+ \f(CRxz \-\-lzma2=preset=1,dict=32MiB foo.tar\fR
1886
+ .fi
1887
+ .RE
1888
+ .PP
1889
+ Avec certains fichiers, la commande ci\-dessus peut être plus rapide que
1890
+ \fBxz\-6\fP tout en compressant bien mieux. Cependant, il faut souligner que
1891
+ seuls certains fichiers bénéficient d'un grand dictionnaire tout en gardant
1892
+ la valeur de CompCPU faible. La siutation la plus évidente où un gros
1893
+ dictionnaire peut baucoup aider, est une archive contenant des fichiers très
1894
+ similaires de quelques megaoctets chacun. La taille de dictionnaire doit
1895
+ être significativement plus grosse que tout fichier individuel pour
1896
+ permettre à LZMA2 de tirer pleinement partie des similarités entre des
1897
+ fichiers consécutifs.
1898
+ .PP
1899
+ Si une utilisation de la mémoire élevée pour la compression et décompression
1900
+ convient, et que le fichier à compresser a une taille de plusieurs centaines
1901
+ de megaoctets, il peut être utile d'utiliser un plus gros dictionnaire que
1902
+ celui fourni par \fBxz\-9\fP (64 Mio) :
1903
+ .RS
1904
+ .PP
1905
+ .nf
1906
+ \f(CRxz \-vv \-\-lzma2=dict=192MiB big_foo.tar\fR
1907
+ .fi
1908
+ .RE
1909
+ .PP
1910
+ Utiliser \fB\-vv\fP (\fB\-\-verbose\-\-verbose\fP) comme dans l'exemple ci\-dessus peut
1911
+ être utile pour voir les besoins en mémoire du compresseur et du
1912
+ décompresseur. Rappelez\-vous qu'utiliser un dictionnaire plus gros que la
1913
+ taille du fichier non compressé est un gachis de mémoire, donc la commande
1914
+ ci\-dessus n'est pas utile pour les petits fichiers.
1915
+ .PP
1916
+ Sometimes the compression time doesn't matter, but the decompressor memory
1917
+ usage has to be kept low, for example, to make it possible to decompress the
1918
+ file on an embedded system. The following command uses \fB\-6e\fP (\fB\-6 \-\-extreme\fP) as a base and sets the dictionary to only 64\ KiB. The
1919
+ resulting file can be decompressed with XZ Embedded (that's why there is
1920
+ \fB\-\-check=crc32\fP) using about 100\ KiB of memory.
1921
+ .RS
1922
+ .PP
1923
+ .nf
1924
+ \f(CRxz \-\-check=crc32 \-\-lzma2=preset=6e,dict=64KiB foo\fR
1925
+ .fi
1926
+ .RE
1927
+ .PP
1928
+ If you want to squeeze out as many bytes as possible, adjusting the number
1929
+ of literal context bits (\fIlc\fP) and number of position bits (\fIpb\fP) can
1930
+ sometimes help. Adjusting the number of literal position bits (\fIlp\fP)
1931
+ might help too, but usually \fIlc\fP and \fIpb\fP are more important. For
1932
+ example, a source code archive contains mostly US\-ASCII text, so something
1933
+ like the following might give slightly (like 0.1\ %) smaller file than \fBxz \-6e\fP (try also without \fBlc=4\fP):
1934
+ .RS
1935
+ .PP
1936
+ .nf
1937
+ \f(CRxz \-\-lzma2=preset=6e,pb=0,lc=4 source_code.tar\fR
1938
+ .fi
1939
+ .RE
1940
+ .PP
1941
+ Using another filter together with LZMA2 can improve compression with
1942
+ certain file types. For example, to compress a x86\-32 or x86\-64 shared
1943
+ library using the x86 BCJ filter:
1944
+ .RS
1945
+ .PP
1946
+ .nf
1947
+ \f(CRxz \-\-x86 \-\-lzma2 libfoo.so\fR
1948
+ .fi
1949
+ .RE
1950
+ .PP
1951
+ Notez que l'ordre des options de filtre est significatif. Si \fB\-\-x86\fP est
1952
+ indiqué après \fB\-\-lzma2\fP, \fBxz\fP donnera une erreur, car il ne peut y avoir
1953
+ aucun filtre après LZMA2, et aussi parce que le filtre BCJ x86 ne peut pas
1954
+ être utilisé comme dernier filtre dans la chaîne.
1955
+ .PP
1956
+ Le filtre Delta associé à LZMA2 peut donner de bons résultats avec les
1957
+ images bitmap. Cela devrait habituellement battre PNG, qui a quelques
1958
+ filtres avancés supplémentaires qu'un simple delta, mais qui utilise Deflate
1959
+ pour la compression effective.
1960
+ .PP
1961
+ The image has to be saved in uncompressed format, for example, as
1962
+ uncompressed TIFF. The distance parameter of the Delta filter is set to
1963
+ match the number of bytes per pixel in the image. For example, 24\-bit RGB
1964
+ bitmap needs \fBdist=3\fP, and it is also good to pass \fBpb=0\fP to LZMA2 to
1965
+ accommodate the three\-byte alignment:
1966
+ .RS
1967
+ .PP
1968
+ .nf
1969
+ \f(CRxz \-\-delta=dist=3 \-\-lzma2=pb=0 foo.tiff\fR
1970
+ .fi
1971
+ .RE
1972
+ .PP
1973
+ If multiple images have been put into a single archive (for example,
1974
+ \&\fB.tar\fP), the Delta filter will work on that too as long as all images have
1975
+ the same number of bytes per pixel.
1976
+ .
1977
+ .SH "VOIR AUSSI"
1978
+ \fBxzdec\fP(1), \fBxzdiff\fP(1), \fBxzgrep\fP(1), \fBxzless\fP(1), \fBxzmore\fP(1),
1979
+ \fBgzip\fP(1), \fBbzip2\fP(1), \fB7z\fP(1)
1980
+ .PP
1981
+ XZ Utils: <https://tukaani.org/xz/>
1982
+ .br
1983
+ XZ Embedded: <https://tukaani.org/xz/embedded.html>
1984
+ .br
1985
+ LZMA SDK: <https://7\-zip.org/sdk.html>
janus/share/man/fr/man1/xzdec.1 ADDED
@@ -0,0 +1,90 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ .\" SPDX-License-Identifier: 0BSD
2
+ .\"
3
+ .\" Author: Lasse Collin
4
+ .\"
5
+ .\" french translation of XZ Utils man
6
+ .\" Copyright (C) 2021 Debian French l10n team <debian-l10n-french@lists.debian.org>
7
+ .\" Translator
8
+ .\" bubu <bubub@no-log.org>, 2021.
9
+ .\"
10
+ .\"*******************************************************************
11
+ .\"
12
+ .\" This file was generated with po4a. Translate the source file.
13
+ .\"
14
+ .\"*******************************************************************
15
+ .TH XZDEC 1 2024\-04\-08 Tukaani "Utilitaires XZ"
16
+ .SH NOM
17
+ xzdec, lzmadec \- Small .xz et .lzma decompresseurs
18
+ .SH SYNOPSIS
19
+ \fBxzdec\fP [\fIoption...\fP] [\fIfichier...\fP]
20
+ .br
21
+ \fBlzmadec\fP [\fIoption...\fP] [\fIfichier...\fP]
22
+ .SH DESCRIPTION
23
+ \fBxzdec\fP est un outil uniquement de décompression, basé sur liblzma pour les
24
+ fichiers \fB.xz\fP (et seulement \fB.xz\fP). \fBxzdec\fP est destiné à remplacer
25
+ \fBxz\fP(1) dans les situations les plus courantes où un script a été écrit
26
+ pour utiliser \fBxz \-\-decompress \-\-stdout\fP (et possiblement quelques autres
27
+ options courantes) pour décompresser des fichiers \fB.xz\fP. \fBlzmadec\fP est
28
+ identique à \fBxzdec\fP, sauf que \fBlzmadec\fP prend en charge les fichiers
29
+ \&\fB.lzma\fP au lieu des fichiers \fB.xz\fP.
30
+ .PP
31
+ Pour réduire la taille de l'exécutable, \fBxzdec\fP ne prend en charge ni le
32
+ multithreading ni la localisation et ne lit pas les options des variables
33
+ d'environnement \fBXZ_DEFAULTS\fP et \fBXZ_OPT\fP. \fBxzdec\fP ne gère pas
34
+ l'affichage d'information sur la progression du traitement : envoyer
35
+ \fBSIGINFO\fP à \fBxzdec\fP ne fait rien, mais envoyer \fBSIGUSR1\fP termine le
36
+ processus au lieu d'afficher de l'information sur la progression.
37
+ .SH OPTIONS
38
+ .TP
39
+ \fB\-d\fP, \fB\-\-decompress\fP, \fB\-\-uncompress\fP
40
+ Ignoré pour la compatibilité avec \fBxz\fP(1), \fBxzdec\fP ne gère que la
41
+ décompression.
42
+ .TP
43
+ \fB\-k\fP, \fB\-\-keep\fP
44
+ Ignoré pour la compatibilité avec \fBxz\fP(1), \fBxzdec\fP ne crée ni ne supprime
45
+ jamais aucun fichier.
46
+ .TP
47
+ \fB\-c\fP, \fB\-\-stdout\fP, \fB\-\-to\-stdout\fP
48
+ Ignoré pour la compatibilité avec \fBxz\fP(1), \fBxzdec\fP écrit toujours les
49
+ données décompressées sur la sortie standard.
50
+ .TP
51
+ \fB\-q\fP, \fB\-\-quiet\fP
52
+ Spécifier cela une fois ne fait rien, car \fBxzdec\fP n'affiche jamais aucun
53
+ avertissement ou notification. Spécifier cela deux fois pour supprimer les
54
+ erreurs.
55
+ .TP
56
+ \fB\-Q\fP, \fB\-\-no\-warn\fP
57
+ Ignoré pour la compatibilité avec \fBxz\fP(1), \fBxzdec\fP n'utilise jamais le
58
+ satut de sortie 2.
59
+ .TP
60
+ \fB\-h\fP, \fB\-\-help\fP
61
+ Afficher un message d'aide et quitter.
62
+ .TP
63
+ \fB\-V\fP, \fB\-\-version\fP
64
+ Afficher le numéro de version de \fBxzdec\fP et liblzma.
65
+ .SH "STATUT DE SORTIE"
66
+ .TP
67
+ \fB0\fP
68
+ Tout s'est bien passé.
69
+ .TP
70
+ \fB1\fP
71
+ Une erreur est survenue.
72
+ .PP
73
+ A la différence de \fBxz\fP(1),\fBxzdec\fP n'a pas de messages d'avertissement, et
74
+ donc le statut de sortie 2 n'est pas utilisé par \fBxzdec\fP.
75
+ .SH NOTES
76
+ Utilisez \fBxz\fP(1) au lieu de \fBxzdec\fP ou \fBlzmadec\fP pour un usage normal de
77
+ tous les jours. \fBxzdec\fP ou \fBlzmadec\fP ne sont utiles que pour les
78
+ situations où il est important d'avoir un plus petit décompresseur que le
79
+ \fBxz\fP(1) complet.
80
+ .PP
81
+ \fBxzdec\fP et \fBlzmadec\fP ne sont en fait pas vraiment si petits. La taille
82
+ peut être encore réduite en abandonnant des fonctionnalités de liblzma au
83
+ moment de la compilation, mais cela ne devrait pas être fait pour des
84
+ exécutables distribués sur des systèmes d'exploitation classique non
85
+ embarqués. Si vous avez besoin d'un décompresseur vraiment petit, pensez à
86
+ utiliser XZ Embedded.
87
+ .SH "VOIR AUSSI"
88
+ \fBxz\fP(1)
89
+ .PP
90
+ XZ Embedded: <https://tukaani.org/xz/embedded.html>
janus/share/man/fr/man1/xzless.1 ADDED
@@ -0,0 +1,49 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ .\" SPDX-License-Identifier: 0BSD
2
+ .\"
3
+ .\" Authors: Andrew Dudman
4
+ .\" Lasse Collin
5
+ .\"
6
+ .\" french translation of XZ Utils man
7
+ .\" Copyright (C) 2021 Debian French l10n team <debian-l10n-french@lists.debian.org>
8
+ .\" Translator
9
+ .\" bubu <bubub@no-log.org>, 2021.
10
+ .\"
11
+ .\" (Note that this file is not based on gzip's zless.1.)
12
+ .\"
13
+ .\"*******************************************************************
14
+ .\"
15
+ .\" This file was generated with po4a. Translate the source file.
16
+ .\"
17
+ .\"*******************************************************************
18
+ .TH XZLESS 1 2024\-02\-12 Tukaani "Utilitaires XZ"
19
+ .SH NOM
20
+ xzless, lzless \- Voir le contenu des fichiers (texte) compressés xz ou lzma
21
+ .SH SYNOPSIS
22
+ \fBxzless\fP [\fIfichier\fP...]
23
+ .br
24
+ \fBlzless\fP [\fIfichier\fP...]
25
+ .SH DESCRIPTION
26
+ \fBxzless\fP is a filter that displays text from compressed files to a
27
+ terminal. Files supported by \fBxz\fP(1) are decompressed; other files are
28
+ assumed to be in uncompressed form already. If no \fIfiles\fP are given,
29
+ \fBxzless\fP reads from standard input.
30
+ .PP
31
+ \fBxzless\fP utilise \fBless\fP(1) pour afficher sa sortie. Contrairement à
32
+ \fBxzmore\fP, son choix d'afficheur ne peut pas être modifié en indiquant une
33
+ variable d'environnement. Les commandes sont basées sur \fBmore\fP(1) et
34
+ \fBvi\fP(1) et permettent des déplacements en avant et en arrière et des
35
+ recherches. Voir le manuel de \fBless\fP(1) pour plus d'information.
36
+ .PP
37
+ La commande nommée \fBlzless\fP est fournie pour la rétrocompatibilité avec les
38
+ utilitaires LZMA.
39
+ .SH ENVIRONNEMENT
40
+ .TP
41
+ \fBLESSMETACHARS\fP
42
+ Une liste de caractères spéciaux pour l'interpréteur. Définis par \fBxzless\fP
43
+ à moins qu'ils ne l'aient déjà été dans l'environnement.
44
+ .TP
45
+ \fBLESSOPEN\fP
46
+ Définir en ligne de commande le décompresseur \fBxz\fP(1) à invoquer pour
47
+ préparer les fichiers en entrée pour \fBless\fP(1).
48
+ .SH "VOIR AUSSI"
49
+ \fBless\fP(1), \fBxz\fP(1), \fBxzmore\fP(1), \fBzless\fP(1)
janus/share/man/man1/lzcat.1 ADDED
@@ -0,0 +1,3198 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ '\" t
2
+ .\" SPDX-License-Identifier: 0BSD
3
+ .\"
4
+ .\" Authors: Lasse Collin
5
+ .\" Jia Tan
6
+ .\"
7
+ .TH XZ 1 "2024-12-30" "Tukaani" "XZ Utils"
8
+ .
9
+ .SH NAME
10
+ xz, unxz, xzcat, lzma, unlzma, lzcat \- Compress or decompress .xz and .lzma files
11
+ .
12
+ .SH SYNOPSIS
13
+ .B xz
14
+ .RI [ option... ]
15
+ .RI [ file... ]
16
+ .
17
+ .SH COMMAND ALIASES
18
+ .B unxz
19
+ is equivalent to
20
+ .BR "xz \-\-decompress" .
21
+ .br
22
+ .B xzcat
23
+ is equivalent to
24
+ .BR "xz \-\-decompress \-\-stdout" .
25
+ .br
26
+ .B lzma
27
+ is equivalent to
28
+ .BR "xz \-\-format=lzma" .
29
+ .br
30
+ .B unlzma
31
+ is equivalent to
32
+ .BR "xz \-\-format=lzma \-\-decompress" .
33
+ .br
34
+ .B lzcat
35
+ is equivalent to
36
+ .BR "xz \-\-format=lzma \-\-decompress \-\-stdout" .
37
+ .PP
38
+ When writing scripts that need to decompress files,
39
+ it is recommended to always use the name
40
+ .B xz
41
+ with appropriate arguments
42
+ .RB ( "xz \-d"
43
+ or
44
+ .BR "xz \-dc" )
45
+ instead of the names
46
+ .B unxz
47
+ and
48
+ .BR xzcat .
49
+ .
50
+ .SH DESCRIPTION
51
+ .B xz
52
+ is a general-purpose data compression tool with
53
+ command line syntax similar to
54
+ .BR gzip (1)
55
+ and
56
+ .BR bzip2 (1).
57
+ The native file format is the
58
+ .B .xz
59
+ format, but the legacy
60
+ .B .lzma
61
+ format used by LZMA Utils and
62
+ raw compressed streams with no container format headers
63
+ are also supported.
64
+ In addition, decompression of the
65
+ .B .lz
66
+ format used by
67
+ .B lzip
68
+ is supported.
69
+ .PP
70
+ .B xz
71
+ compresses or decompresses each
72
+ .I file
73
+ according to the selected operation mode.
74
+ If no
75
+ .I files
76
+ are given or
77
+ .I file
78
+ is
79
+ .BR \- ,
80
+ .B xz
81
+ reads from standard input and writes the processed data
82
+ to standard output.
83
+ .B xz
84
+ will refuse (display an error and skip the
85
+ .IR file )
86
+ to write compressed data to standard output if it is a terminal.
87
+ Similarly,
88
+ .B xz
89
+ will refuse to read compressed data
90
+ from standard input if it is a terminal.
91
+ .PP
92
+ Unless
93
+ .B \-\-stdout
94
+ is specified,
95
+ .I files
96
+ other than
97
+ .B \-
98
+ are written to a new file whose name is derived from the source
99
+ .I file
100
+ name:
101
+ .IP \(bu 3
102
+ When compressing, the suffix of the target file format
103
+ .RB ( .xz
104
+ or
105
+ .BR .lzma )
106
+ is appended to the source filename to get the target filename.
107
+ .IP \(bu 3
108
+ When decompressing, the
109
+ .BR .xz ,
110
+ .BR .lzma ,
111
+ or
112
+ .B .lz
113
+ suffix is removed from the filename to get the target filename.
114
+ .B xz
115
+ also recognizes the suffixes
116
+ .B .txz
117
+ and
118
+ .BR .tlz ,
119
+ and replaces them with the
120
+ .B .tar
121
+ suffix.
122
+ .PP
123
+ If the target file already exists, an error is displayed and the
124
+ .I file
125
+ is skipped.
126
+ .PP
127
+ Unless writing to standard output,
128
+ .B xz
129
+ will display a warning and skip the
130
+ .I file
131
+ if any of the following applies:
132
+ .IP \(bu 3
133
+ .I File
134
+ is not a regular file.
135
+ Symbolic links are not followed,
136
+ and thus they are not considered to be regular files.
137
+ .IP \(bu 3
138
+ .I File
139
+ has more than one hard link.
140
+ .IP \(bu 3
141
+ .I File
142
+ has setuid, setgid, or sticky bit set.
143
+ .IP \(bu 3
144
+ The operation mode is set to compress and the
145
+ .I file
146
+ already has a suffix of the target file format
147
+ .RB ( .xz
148
+ or
149
+ .B .txz
150
+ when compressing to the
151
+ .B .xz
152
+ format, and
153
+ .B .lzma
154
+ or
155
+ .B .tlz
156
+ when compressing to the
157
+ .B .lzma
158
+ format).
159
+ .IP \(bu 3
160
+ The operation mode is set to decompress and the
161
+ .I file
162
+ doesn't have a suffix of any of the supported file formats
163
+ .RB ( .xz ,
164
+ .BR .txz ,
165
+ .BR .lzma ,
166
+ .BR .tlz ,
167
+ or
168
+ .BR .lz ).
169
+ .PP
170
+ After successfully compressing or decompressing the
171
+ .IR file ,
172
+ .B xz
173
+ copies the owner, group, permissions, access time,
174
+ and modification time from the source
175
+ .I file
176
+ to the target file.
177
+ If copying the group fails, the permissions are modified
178
+ so that the target file doesn't become accessible to users
179
+ who didn't have permission to access the source
180
+ .IR file .
181
+ .B xz
182
+ doesn't support copying other metadata like access control lists
183
+ or extended attributes yet.
184
+ .PP
185
+ Once the target file has been successfully closed, the source
186
+ .I file
187
+ is removed unless
188
+ .B \-\-keep
189
+ was specified.
190
+ The source
191
+ .I file
192
+ is never removed if the output is written to standard output
193
+ or if an error occurs.
194
+ .PP
195
+ Sending
196
+ .B SIGINFO
197
+ or
198
+ .B SIGUSR1
199
+ to the
200
+ .B xz
201
+ process makes it print progress information to standard error.
202
+ This has only limited use since when standard error
203
+ is a terminal, using
204
+ .B \-\-verbose
205
+ will display an automatically updating progress indicator.
206
+ .
207
+ .SS "Memory usage"
208
+ The memory usage of
209
+ .B xz
210
+ varies from a few hundred kilobytes to several gigabytes
211
+ depending on the compression settings.
212
+ The settings used when compressing a file determine
213
+ the memory requirements of the decompressor.
214
+ Typically the decompressor needs 5\ % to 20\ % of
215
+ the amount of memory that the compressor needed when
216
+ creating the file.
217
+ For example, decompressing a file created with
218
+ .B xz \-9
219
+ currently requires 65\ MiB of memory.
220
+ Still, it is possible to have
221
+ .B .xz
222
+ files that require several gigabytes of memory to decompress.
223
+ .PP
224
+ Especially users of older systems may find
225
+ the possibility of very large memory usage annoying.
226
+ To prevent uncomfortable surprises,
227
+ .B xz
228
+ has a built-in memory usage limiter, which is disabled by default.
229
+ While some operating systems provide ways to limit
230
+ the memory usage of processes, relying on it
231
+ wasn't deemed to be flexible enough (for example, using
232
+ .BR ulimit (1)
233
+ to limit virtual memory tends to cripple
234
+ .BR mmap (2)).
235
+ .PP
236
+ The memory usage limiter can be enabled with
237
+ the command line option \fB\-\-memlimit=\fIlimit\fR.
238
+ Often it is more convenient to enable the limiter
239
+ by default by setting the environment variable
240
+ .BR XZ_DEFAULTS ,
241
+ for example,
242
+ .BR XZ_DEFAULTS=\-\-memlimit=150MiB .
243
+ It is possible to set the limits separately
244
+ for compression and decompression by using
245
+ .BI \-\-memlimit\-compress= limit
246
+ and \fB\-\-memlimit\-decompress=\fIlimit\fR.
247
+ Using these two options outside
248
+ .B XZ_DEFAULTS
249
+ is rarely useful because a single run of
250
+ .B xz
251
+ cannot do both compression and decompression and
252
+ .BI \-\-memlimit= limit
253
+ (or
254
+ .B \-M
255
+ .IR limit )
256
+ is shorter to type on the command line.
257
+ .PP
258
+ If the specified memory usage limit is exceeded when decompressing,
259
+ .B xz
260
+ will display an error and decompressing the file will fail.
261
+ If the limit is exceeded when compressing,
262
+ .B xz
263
+ will try to scale the settings down so that the limit
264
+ is no longer exceeded (except when using
265
+ .B \-\-format=raw
266
+ or
267
+ .BR \-\-no\-adjust ).
268
+ This way the operation won't fail unless the limit is very small.
269
+ The scaling of the settings is done in steps that don't
270
+ match the compression level presets, for example, if the limit is
271
+ only slightly less than the amount required for
272
+ .BR "xz \-9" ,
273
+ the settings will be scaled down only a little,
274
+ not all the way down to
275
+ .BR "xz \-8" .
276
+ .
277
+ .SS "Concatenation and padding with .xz files"
278
+ It is possible to concatenate
279
+ .B .xz
280
+ files as is.
281
+ .B xz
282
+ will decompress such files as if they were a single
283
+ .B .xz
284
+ file.
285
+ .PP
286
+ It is possible to insert padding between the concatenated parts
287
+ or after the last part.
288
+ The padding must consist of null bytes and the size
289
+ of the padding must be a multiple of four bytes.
290
+ This can be useful, for example, if the
291
+ .B .xz
292
+ file is stored on a medium that measures file sizes
293
+ in 512-byte blocks.
294
+ .PP
295
+ Concatenation and padding are not allowed with
296
+ .B .lzma
297
+ files or raw streams.
298
+ .
299
+ .SH OPTIONS
300
+ .
301
+ .SS "Integer suffixes and special values"
302
+ In most places where an integer argument is expected,
303
+ an optional suffix is supported to easily indicate large integers.
304
+ There must be no space between the integer and the suffix.
305
+ .TP
306
+ .B KiB
307
+ Multiply the integer by 1,024 (2^10).
308
+ .BR Ki ,
309
+ .BR k ,
310
+ .BR kB ,
311
+ .BR K ,
312
+ and
313
+ .B KB
314
+ are accepted as synonyms for
315
+ .BR KiB .
316
+ .TP
317
+ .B MiB
318
+ Multiply the integer by 1,048,576 (2^20).
319
+ .BR Mi ,
320
+ .BR m ,
321
+ .BR M ,
322
+ and
323
+ .B MB
324
+ are accepted as synonyms for
325
+ .BR MiB .
326
+ .TP
327
+ .B GiB
328
+ Multiply the integer by 1,073,741,824 (2^30).
329
+ .BR Gi ,
330
+ .BR g ,
331
+ .BR G ,
332
+ and
333
+ .B GB
334
+ are accepted as synonyms for
335
+ .BR GiB .
336
+ .PP
337
+ The special value
338
+ .B max
339
+ can be used to indicate the maximum integer value
340
+ supported by the option.
341
+ .
342
+ .SS "Operation mode"
343
+ If multiple operation mode options are given,
344
+ the last one takes effect.
345
+ .TP
346
+ .BR \-z ", " \-\-compress
347
+ Compress.
348
+ This is the default operation mode when no operation mode option
349
+ is specified and no other operation mode is implied from
350
+ the command name (for example,
351
+ .B unxz
352
+ implies
353
+ .BR \-\-decompress ).
354
+ .IP ""
355
+ .\" The DESCRIPTION section already says this but it's good to repeat it
356
+ .\" here because the default behavior is a bit dangerous and new users
357
+ .\" in a hurry may skip reading the DESCRIPTION section.
358
+ After successful compression, the source file is removed
359
+ unless writing to standard output or
360
+ .B \-\-keep
361
+ was specified.
362
+ .TP
363
+ .BR \-d ", " \-\-decompress ", " \-\-uncompress
364
+ Decompress.
365
+ .\" The DESCRIPTION section already says this but it's good to repeat it
366
+ .\" here because the default behavior is a bit dangerous and new users
367
+ .\" in a hurry may skip reading the DESCRIPTION section.
368
+ After successful decompression, the source file is removed
369
+ unless writing to standard output or
370
+ .B \-\-keep
371
+ was specified.
372
+ .TP
373
+ .BR \-t ", " \-\-test
374
+ Test the integrity of compressed
375
+ .IR files .
376
+ This option is equivalent to
377
+ .B "\-\-decompress \-\-stdout"
378
+ except that the decompressed data is discarded instead of being
379
+ written to standard output.
380
+ No files are created or removed.
381
+ .TP
382
+ .BR \-l ", " \-\-list
383
+ Print information about compressed
384
+ .IR files .
385
+ No uncompressed output is produced,
386
+ and no files are created or removed.
387
+ In list mode, the program cannot read
388
+ the compressed data from standard
389
+ input or from other unseekable sources.
390
+ .IP ""
391
+ The default listing shows basic information about
392
+ .IR files ,
393
+ one file per line.
394
+ To get more detailed information, use also the
395
+ .B \-\-verbose
396
+ option.
397
+ For even more information, use
398
+ .B \-\-verbose
399
+ twice, but note that this may be slow, because getting all the extra
400
+ information requires many seeks.
401
+ The width of verbose output exceeds
402
+ 80 characters, so piping the output to, for example,
403
+ .B "less\ \-S"
404
+ may be convenient if the terminal isn't wide enough.
405
+ .IP ""
406
+ The exact output may vary between
407
+ .B xz
408
+ versions and different locales.
409
+ For machine-readable output,
410
+ .B \-\-robot \-\-list
411
+ should be used.
412
+ .
413
+ .SS "Operation modifiers"
414
+ .TP
415
+ .BR \-k ", " \-\-keep
416
+ Don't delete the input files.
417
+ .IP ""
418
+ Since
419
+ .B xz
420
+ 5.2.6,
421
+ this option also makes
422
+ .B xz
423
+ compress or decompress even if the input is
424
+ a symbolic link to a regular file,
425
+ has more than one hard link,
426
+ or has the setuid, setgid, or sticky bit set.
427
+ The setuid, setgid, and sticky bits are not copied
428
+ to the target file.
429
+ In earlier versions this was only done with
430
+ .BR \-\-force .
431
+ .TP
432
+ .BR \-f ", " \-\-force
433
+ This option has several effects:
434
+ .RS
435
+ .IP \(bu 3
436
+ If the target file already exists,
437
+ delete it before compressing or decompressing.
438
+ .IP \(bu 3
439
+ Compress or decompress even if the input is
440
+ a symbolic link to a regular file,
441
+ has more than one hard link,
442
+ or has the setuid, setgid, or sticky bit set.
443
+ The setuid, setgid, and sticky bits are not copied
444
+ to the target file.
445
+ .IP \(bu 3
446
+ When used with
447
+ .B \-\-decompress
448
+ .B \-\-stdout
449
+ and
450
+ .B xz
451
+ cannot recognize the type of the source file,
452
+ copy the source file as is to standard output.
453
+ This allows
454
+ .B xzcat
455
+ .B \-\-force
456
+ to be used like
457
+ .BR cat (1)
458
+ for files that have not been compressed with
459
+ .BR xz .
460
+ Note that in future,
461
+ .B xz
462
+ might support new compressed file formats, which may make
463
+ .B xz
464
+ decompress more types of files instead of copying them as is to
465
+ standard output.
466
+ .BI \-\-format= format
467
+ can be used to restrict
468
+ .B xz
469
+ to decompress only a single file format.
470
+ .RE
471
+ .TP
472
+ .BR \-c ", " \-\-stdout ", " \-\-to\-stdout
473
+ Write the compressed or decompressed data to
474
+ standard output instead of a file.
475
+ This implies
476
+ .BR \-\-keep .
477
+ .TP
478
+ .B \-\-single\-stream
479
+ Decompress only the first
480
+ .B .xz
481
+ stream, and
482
+ silently ignore possible remaining input data following the stream.
483
+ Normally such trailing garbage makes
484
+ .B xz
485
+ display an error.
486
+ .IP ""
487
+ .B xz
488
+ never decompresses more than one stream from
489
+ .B .lzma
490
+ files or raw streams, but this option still makes
491
+ .B xz
492
+ ignore the possible trailing data after the
493
+ .B .lzma
494
+ file or raw stream.
495
+ .IP ""
496
+ This option has no effect if the operation mode is not
497
+ .B \-\-decompress
498
+ or
499
+ .BR \-\-test .
500
+ .TP
501
+ .B \-\-no\-sparse
502
+ Disable creation of sparse files.
503
+ By default, if decompressing into a regular file,
504
+ .B xz
505
+ tries to make the file sparse if the decompressed data contains
506
+ long sequences of binary zeros.
507
+ It also works when writing to standard output
508
+ as long as standard output is connected to a regular file
509
+ and certain additional conditions are met to make it safe.
510
+ Creating sparse files may save disk space and speed up
511
+ the decompression by reducing the amount of disk I/O.
512
+ .TP
513
+ \fB\-S\fR \fI.suf\fR, \fB\-\-suffix=\fI.suf
514
+ When compressing, use
515
+ .I .suf
516
+ as the suffix for the target file instead of
517
+ .B .xz
518
+ or
519
+ .BR .lzma .
520
+ If not writing to standard output and
521
+ the source file already has the suffix
522
+ .IR .suf ,
523
+ a warning is displayed and the file is skipped.
524
+ .IP ""
525
+ When decompressing, recognize files with the suffix
526
+ .I .suf
527
+ in addition to files with the
528
+ .BR .xz ,
529
+ .BR .txz ,
530
+ .BR .lzma ,
531
+ .BR .tlz ,
532
+ or
533
+ .B .lz
534
+ suffix.
535
+ If the source file has the suffix
536
+ .IR .suf ,
537
+ the suffix is removed to get the target filename.
538
+ .IP ""
539
+ When compressing or decompressing raw streams
540
+ .RB ( \-\-format=raw ),
541
+ the suffix must always be specified unless
542
+ writing to standard output,
543
+ because there is no default suffix for raw streams.
544
+ .TP
545
+ \fB\-\-files\fR[\fB=\fIfile\fR]
546
+ Read the filenames to process from
547
+ .IR file ;
548
+ if
549
+ .I file
550
+ is omitted, filenames are read from standard input.
551
+ Filenames must be terminated with the newline character.
552
+ A dash
553
+ .RB ( \- )
554
+ is taken as a regular filename; it doesn't mean standard input.
555
+ If filenames are given also as command line arguments, they are
556
+ processed before the filenames read from
557
+ .IR file .
558
+ .TP
559
+ \fB\-\-files0\fR[\fB=\fIfile\fR]
560
+ This is identical to \fB\-\-files\fR[\fB=\fIfile\fR] except
561
+ that each filename must be terminated with the null character.
562
+ .
563
+ .SS "Basic file format and compression options"
564
+ .TP
565
+ \fB\-F\fR \fIformat\fR, \fB\-\-format=\fIformat
566
+ Specify the file
567
+ .I format
568
+ to compress or decompress:
569
+ .RS
570
+ .TP
571
+ .B auto
572
+ This is the default.
573
+ When compressing,
574
+ .B auto
575
+ is equivalent to
576
+ .BR xz .
577
+ When decompressing,
578
+ the format of the input file is automatically detected.
579
+ Note that raw streams (created with
580
+ .BR \-\-format=raw )
581
+ cannot be auto-detected.
582
+ .TP
583
+ .B xz
584
+ Compress to the
585
+ .B .xz
586
+ file format, or accept only
587
+ .B .xz
588
+ files when decompressing.
589
+ .TP
590
+ .BR lzma ", " alone
591
+ Compress to the legacy
592
+ .B .lzma
593
+ file format, or accept only
594
+ .B .lzma
595
+ files when decompressing.
596
+ The alternative name
597
+ .B alone
598
+ is provided for backwards compatibility with LZMA Utils.
599
+ .TP
600
+ .B lzip
601
+ Accept only
602
+ .B .lz
603
+ files when decompressing.
604
+ Compression is not supported.
605
+ .IP ""
606
+ The
607
+ .B .lz
608
+ format version 0 and the unextended version 1 are supported.
609
+ Version 0 files were produced by
610
+ .B lzip
611
+ 1.3 and older.
612
+ Such files aren't common but may be found from file archives
613
+ as a few source packages were released in this format.
614
+ People might have old personal files in this format too.
615
+ Decompression support for the format version 0 was removed in
616
+ .B lzip
617
+ 1.18.
618
+ .IP ""
619
+ .B lzip
620
+ 1.4 and later create files in the format version 1.
621
+ The sync flush marker extension to the format version 1 was added in
622
+ .B lzip
623
+ 1.6.
624
+ This extension is rarely used and isn't supported by
625
+ .B xz
626
+ (diagnosed as corrupt input).
627
+ .TP
628
+ .B raw
629
+ Compress or uncompress a raw stream (no headers).
630
+ This is meant for advanced users only.
631
+ To decode raw streams, you need use
632
+ .B \-\-format=raw
633
+ and explicitly specify the filter chain,
634
+ which normally would have been stored in the container headers.
635
+ .RE
636
+ .TP
637
+ \fB\-C\fR \fIcheck\fR, \fB\-\-check=\fIcheck
638
+ Specify the type of the integrity check.
639
+ The check is calculated from the uncompressed data and
640
+ stored in the
641
+ .B .xz
642
+ file.
643
+ This option has an effect only when compressing into the
644
+ .B .xz
645
+ format; the
646
+ .B .lzma
647
+ format doesn't support integrity checks.
648
+ The integrity check (if any) is verified when the
649
+ .B .xz
650
+ file is decompressed.
651
+ .IP ""
652
+ Supported
653
+ .I check
654
+ types:
655
+ .RS
656
+ .TP
657
+ .B none
658
+ Don't calculate an integrity check at all.
659
+ This is usually a bad idea.
660
+ This can be useful when integrity of the data is verified
661
+ by other means anyway.
662
+ .TP
663
+ .B crc32
664
+ Calculate CRC32 using the polynomial from IEEE-802.3 (Ethernet).
665
+ .TP
666
+ .B crc64
667
+ Calculate CRC64 using the polynomial from ECMA-182.
668
+ This is the default, since it is slightly better than CRC32
669
+ at detecting damaged files and the speed difference is negligible.
670
+ .TP
671
+ .B sha256
672
+ Calculate SHA-256.
673
+ This is somewhat slower than CRC32 and CRC64.
674
+ .RE
675
+ .IP ""
676
+ Integrity of the
677
+ .B .xz
678
+ headers is always verified with CRC32.
679
+ It is not possible to change or disable it.
680
+ .TP
681
+ .B \-\-ignore\-check
682
+ Don't verify the integrity check of the compressed data when decompressing.
683
+ The CRC32 values in the
684
+ .B .xz
685
+ headers will still be verified normally.
686
+ .IP ""
687
+ .B "Do not use this option unless you know what you are doing."
688
+ Possible reasons to use this option:
689
+ .RS
690
+ .IP \(bu 3
691
+ Trying to recover data from a corrupt .xz file.
692
+ .IP \(bu 3
693
+ Speeding up decompression.
694
+ This matters mostly with SHA-256 or
695
+ with files that have compressed extremely well.
696
+ It's recommended to not use this option for this purpose
697
+ unless the file integrity is verified externally in some other way.
698
+ .RE
699
+ .TP
700
+ .BR \-0 " ... " \-9
701
+ Select a compression preset level.
702
+ The default is
703
+ .BR \-6 .
704
+ If multiple preset levels are specified,
705
+ the last one takes effect.
706
+ If a custom filter chain was already specified, setting
707
+ a compression preset level clears the custom filter chain.
708
+ .IP ""
709
+ The differences between the presets are more significant than with
710
+ .BR gzip (1)
711
+ and
712
+ .BR bzip2 (1).
713
+ The selected compression settings determine
714
+ the memory requirements of the decompressor,
715
+ thus using a too high preset level might make it painful
716
+ to decompress the file on an old system with little RAM.
717
+ Specifically,
718
+ .B "it's not a good idea to blindly use \-9 for everything"
719
+ like it often is with
720
+ .BR gzip (1)
721
+ and
722
+ .BR bzip2 (1).
723
+ .RS
724
+ .TP
725
+ .BR "\-0" " ... " "\-3"
726
+ These are somewhat fast presets.
727
+ .B \-0
728
+ is sometimes faster than
729
+ .B "gzip \-9"
730
+ while compressing much better.
731
+ The higher ones often have speed comparable to
732
+ .BR bzip2 (1)
733
+ with comparable or better compression ratio,
734
+ although the results
735
+ depend a lot on the type of data being compressed.
736
+ .TP
737
+ .BR "\-4" " ... " "\-6"
738
+ Good to very good compression while keeping
739
+ decompressor memory usage reasonable even for old systems.
740
+ .B \-6
741
+ is the default, which is usually a good choice
742
+ for distributing files that need to be decompressible
743
+ even on systems with only 16\ MiB RAM.
744
+ .RB ( \-5e
745
+ or
746
+ .B \-6e
747
+ may be worth considering too.
748
+ See
749
+ .BR \-\-extreme .)
750
+ .TP
751
+ .B "\-7 ... \-9"
752
+ These are like
753
+ .B \-6
754
+ but with higher compressor and decompressor memory requirements.
755
+ These are useful only when compressing files bigger than
756
+ 8\ MiB, 16\ MiB, and 32\ MiB, respectively.
757
+ .RE
758
+ .IP ""
759
+ On the same hardware, the decompression speed is approximately
760
+ a constant number of bytes of compressed data per second.
761
+ In other words, the better the compression,
762
+ the faster the decompression will usually be.
763
+ This also means that the amount of uncompressed output
764
+ produced per second can vary a lot.
765
+ .IP ""
766
+ The following table summarises the features of the presets:
767
+ .RS
768
+ .RS
769
+ .PP
770
+ .TS
771
+ tab(;);
772
+ c c c c c
773
+ n n n n n.
774
+ Preset;DictSize;CompCPU;CompMem;DecMem
775
+ \-0;256 KiB;0;3 MiB;1 MiB
776
+ \-1;1 MiB;1;9 MiB;2 MiB
777
+ \-2;2 MiB;2;17 MiB;3 MiB
778
+ \-3;4 MiB;3;32 MiB;5 MiB
779
+ \-4;4 MiB;4;48 MiB;5 MiB
780
+ \-5;8 MiB;5;94 MiB;9 MiB
781
+ \-6;8 MiB;6;94 MiB;9 MiB
782
+ \-7;16 MiB;6;186 MiB;17 MiB
783
+ \-8;32 MiB;6;370 MiB;33 MiB
784
+ \-9;64 MiB;6;674 MiB;65 MiB
785
+ .TE
786
+ .RE
787
+ .RE
788
+ .IP ""
789
+ Column descriptions:
790
+ .RS
791
+ .IP \(bu 3
792
+ DictSize is the LZMA2 dictionary size.
793
+ It is waste of memory to use a dictionary bigger than
794
+ the size of the uncompressed file.
795
+ This is why it is good to avoid using the presets
796
+ .BR \-7 " ... " \-9
797
+ when there's no real need for them.
798
+ At
799
+ .B \-6
800
+ and lower, the amount of memory wasted is
801
+ usually low enough to not matter.
802
+ .IP \(bu 3
803
+ CompCPU is a simplified representation of the LZMA2 settings
804
+ that affect compression speed.
805
+ The dictionary size affects speed too,
806
+ so while CompCPU is the same for levels
807
+ .BR \-6 " ... " \-9 ,
808
+ higher levels still tend to be a little slower.
809
+ To get even slower and thus possibly better compression, see
810
+ .BR \-\-extreme .
811
+ .IP \(bu 3
812
+ CompMem contains the compressor memory requirements
813
+ in the single-threaded mode.
814
+ It may vary slightly between
815
+ .B xz
816
+ versions.
817
+ .IP \(bu 3
818
+ DecMem contains the decompressor memory requirements.
819
+ That is, the compression settings determine
820
+ the memory requirements of the decompressor.
821
+ The exact decompressor memory usage is slightly more than
822
+ the LZMA2 dictionary size, but the values in the table
823
+ have been rounded up to the next full MiB.
824
+ .RE
825
+ .IP ""
826
+ Memory requirements of the multi-threaded mode are
827
+ significantly higher than that of the single-threaded mode.
828
+ With the default value of
829
+ .BR \-\-block\-size ,
830
+ each thread needs 3*3*DictSize plus CompMem or DecMem.
831
+ For example, four threads with preset
832
+ .B \-6
833
+ needs 660\(en670\ MiB of memory.
834
+ .TP
835
+ .BR \-e ", " \-\-extreme
836
+ Use a slower variant of the selected compression preset level
837
+ .RB ( \-0 " ... " \-9 )
838
+ to hopefully get a little bit better compression ratio,
839
+ but with bad luck this can also make it worse.
840
+ Decompressor memory usage is not affected,
841
+ but compressor memory usage increases a little at preset levels
842
+ .BR \-0 " ... " \-3 .
843
+ .IP ""
844
+ Since there are two presets with dictionary sizes
845
+ 4\ MiB and 8\ MiB, the presets
846
+ .B \-3e
847
+ and
848
+ .B \-5e
849
+ use slightly faster settings (lower CompCPU) than
850
+ .B \-4e
851
+ and
852
+ .BR \-6e ,
853
+ respectively.
854
+ That way no two presets are identical.
855
+ .RS
856
+ .RS
857
+ .PP
858
+ .TS
859
+ tab(;);
860
+ c c c c c
861
+ n n n n n.
862
+ Preset;DictSize;CompCPU;CompMem;DecMem
863
+ \-0e;256 KiB;8;4 MiB;1 MiB
864
+ \-1e;1 MiB;8;13 MiB;2 MiB
865
+ \-2e;2 MiB;8;25 MiB;3 MiB
866
+ \-3e;4 MiB;7;48 MiB;5 MiB
867
+ \-4e;4 MiB;8;48 MiB;5 MiB
868
+ \-5e;8 MiB;7;94 MiB;9 MiB
869
+ \-6e;8 MiB;8;94 MiB;9 MiB
870
+ \-7e;16 MiB;8;186 MiB;17 MiB
871
+ \-8e;32 MiB;8;370 MiB;33 MiB
872
+ \-9e;64 MiB;8;674 MiB;65 MiB
873
+ .TE
874
+ .RE
875
+ .RE
876
+ .IP ""
877
+ For example, there are a total of four presets that use
878
+ 8\ MiB dictionary, whose order from the fastest to the slowest is
879
+ .BR \-5 ,
880
+ .BR \-6 ,
881
+ .BR \-5e ,
882
+ and
883
+ .BR \-6e .
884
+ .TP
885
+ .B \-\-fast
886
+ .PD 0
887
+ .TP
888
+ .B \-\-best
889
+ .PD
890
+ These are somewhat misleading aliases for
891
+ .B \-0
892
+ and
893
+ .BR \-9 ,
894
+ respectively.
895
+ These are provided only for backwards compatibility
896
+ with LZMA Utils.
897
+ Avoid using these options.
898
+ .TP
899
+ .BI \-\-block\-size= size
900
+ When compressing to the
901
+ .B .xz
902
+ format, split the input data into blocks of
903
+ .I size
904
+ bytes.
905
+ The blocks are compressed independently from each other,
906
+ which helps with multi-threading and
907
+ makes limited random-access decompression possible.
908
+ This option is typically used to override the default
909
+ block size in multi-threaded mode,
910
+ but this option can be used in single-threaded mode too.
911
+ .IP ""
912
+ In multi-threaded mode about three times
913
+ .I size
914
+ bytes will be allocated in each thread for buffering input and output.
915
+ The default
916
+ .I size
917
+ is three times the LZMA2 dictionary size or 1 MiB,
918
+ whichever is more.
919
+ Typically a good value is 2\(en4 times
920
+ the size of the LZMA2 dictionary or at least 1 MiB.
921
+ Using
922
+ .I size
923
+ less than the LZMA2 dictionary size is waste of RAM
924
+ because then the LZMA2 dictionary buffer will never get fully used.
925
+ In multi-threaded mode,
926
+ the sizes of the blocks are stored in the block headers.
927
+ This size information is required for multi-threaded decompression.
928
+ .IP ""
929
+ In single-threaded mode no block splitting is done by default.
930
+ Setting this option doesn't affect memory usage.
931
+ No size information is stored in block headers,
932
+ thus files created in single-threaded mode
933
+ won't be identical to files created in multi-threaded mode.
934
+ The lack of size information also means that
935
+ .B xz
936
+ won't be able decompress the files in multi-threaded mode.
937
+ .TP
938
+ .BI \-\-block\-list= items
939
+ When compressing to the
940
+ .B .xz
941
+ format, start a new block with an optional custom filter chain after
942
+ the given intervals of uncompressed data.
943
+ .IP ""
944
+ The
945
+ .I items
946
+ are a comma-separated list.
947
+ Each item consists of an optional filter chain number
948
+ between 0 and 9 followed by a colon
949
+ .RB ( : )
950
+ and a required size of uncompressed data.
951
+ Omitting an item (two or more consecutive commas) is a
952
+ shorthand to use the size and filters of the previous item.
953
+ .IP ""
954
+ If the input file is bigger than the sum of
955
+ the sizes in
956
+ .IR items ,
957
+ the last item is repeated until the end of the file.
958
+ A special value of
959
+ .B 0
960
+ may be used as the last size to indicate that
961
+ the rest of the file should be encoded as a single block.
962
+ .IP ""
963
+ An alternative filter chain for each block can be
964
+ specified in combination with the
965
+ .BI \-\-filters1= filters
966
+ \&...\&
967
+ .BI \-\-filters9= filters
968
+ options.
969
+ These options define filter chains with an identifier
970
+ between 1\(en9.
971
+ Filter chain 0 can be used to refer to the default filter chain,
972
+ which is the same as not specifying a filter chain.
973
+ The filter chain identifier can be used before the uncompressed
974
+ size, followed by a colon
975
+ .RB ( : ).
976
+ For example, if one specifies
977
+ .B \-\-block\-list=1:2MiB,3:2MiB,2:4MiB,,2MiB,0:4MiB
978
+ then blocks will be created using:
979
+ .RS
980
+ .IP \(bu 3
981
+ The filter chain specified by
982
+ .B \-\-filters1
983
+ and 2 MiB input
984
+ .IP \(bu 3
985
+ The filter chain specified by
986
+ .B \-\-filters3
987
+ and 2 MiB input
988
+ .IP \(bu 3
989
+ The filter chain specified by
990
+ .B \-\-filters2
991
+ and 4 MiB input
992
+ .IP \(bu 3
993
+ The filter chain specified by
994
+ .B \-\-filters2
995
+ and 4 MiB input
996
+ .IP \(bu 3
997
+ The default filter chain and 2 MiB input
998
+ .IP \(bu 3
999
+ The default filter chain and 4 MiB input for every block until
1000
+ end of input.
1001
+ .RE
1002
+ .IP ""
1003
+ If one specifies a size that exceeds the encoder's block size
1004
+ (either the default value in threaded mode or
1005
+ the value specified with \fB\-\-block\-size=\fIsize\fR),
1006
+ the encoder will create additional blocks while
1007
+ keeping the boundaries specified in
1008
+ .IR items .
1009
+ For example, if one specifies
1010
+ .B \-\-block\-size=10MiB
1011
+ .B \-\-block\-list=5MiB,10MiB,8MiB,12MiB,24MiB
1012
+ and the input file is 80 MiB,
1013
+ one will get 11 blocks:
1014
+ 5, 10, 8, 10, 2, 10, 10, 4, 10, 10, and 1 MiB.
1015
+ .IP ""
1016
+ In multi-threaded mode the sizes of the blocks
1017
+ are stored in the block headers.
1018
+ This isn't done in single-threaded mode,
1019
+ so the encoded output won't be
1020
+ identical to that of the multi-threaded mode.
1021
+ .TP
1022
+ .BI \-\-flush\-timeout= timeout
1023
+ When compressing, if more than
1024
+ .I timeout
1025
+ milliseconds (a positive integer) has passed since the previous flush and
1026
+ reading more input would block,
1027
+ all the pending input data is flushed from the encoder and
1028
+ made available in the output stream.
1029
+ This can be useful if
1030
+ .B xz
1031
+ is used to compress data that is streamed over a network.
1032
+ Small
1033
+ .I timeout
1034
+ values make the data available at the receiving end
1035
+ with a small delay, but large
1036
+ .I timeout
1037
+ values give better compression ratio.
1038
+ .IP ""
1039
+ This feature is disabled by default.
1040
+ If this option is specified more than once, the last one takes effect.
1041
+ The special
1042
+ .I timeout
1043
+ value of
1044
+ .B 0
1045
+ can be used to explicitly disable this feature.
1046
+ .IP ""
1047
+ This feature is not available on non-POSIX systems.
1048
+ .IP ""
1049
+ .\" FIXME
1050
+ .B "This feature is still experimental."
1051
+ Currently
1052
+ .B xz
1053
+ is unsuitable for decompressing the stream in real time due to how
1054
+ .B xz
1055
+ does buffering.
1056
+ .TP
1057
+ .BI \-\-memlimit\-compress= limit
1058
+ Set a memory usage limit for compression.
1059
+ If this option is specified multiple times,
1060
+ the last one takes effect.
1061
+ .IP ""
1062
+ If the compression settings exceed the
1063
+ .IR limit ,
1064
+ .B xz
1065
+ will attempt to adjust the settings downwards so that
1066
+ the limit is no longer exceeded and display a notice that
1067
+ automatic adjustment was done.
1068
+ The adjustments are done in this order:
1069
+ reducing the number of threads,
1070
+ switching to single-threaded mode
1071
+ if even one thread in multi-threaded mode exceeds the
1072
+ .IR limit ,
1073
+ and finally reducing the LZMA2 dictionary size.
1074
+ .IP ""
1075
+ When compressing with
1076
+ .B \-\-format=raw
1077
+ or if
1078
+ .B \-\-no\-adjust
1079
+ has been specified,
1080
+ only the number of threads may be reduced
1081
+ since it can be done without affecting the compressed output.
1082
+ .IP ""
1083
+ If the
1084
+ .I limit
1085
+ cannot be met even with the adjustments described above,
1086
+ an error is displayed and
1087
+ .B xz
1088
+ will exit with exit status 1.
1089
+ .IP ""
1090
+ The
1091
+ .I limit
1092
+ can be specified in multiple ways:
1093
+ .RS
1094
+ .IP \(bu 3
1095
+ The
1096
+ .I limit
1097
+ can be an absolute value in bytes.
1098
+ Using an integer suffix like
1099
+ .B MiB
1100
+ can be useful.
1101
+ Example:
1102
+ .B "\-\-memlimit\-compress=80MiB"
1103
+ .IP \(bu 3
1104
+ The
1105
+ .I limit
1106
+ can be specified as a percentage of total physical memory (RAM).
1107
+ This can be useful especially when setting the
1108
+ .B XZ_DEFAULTS
1109
+ environment variable in a shell initialization script
1110
+ that is shared between different computers.
1111
+ That way the limit is automatically bigger
1112
+ on systems with more memory.
1113
+ Example:
1114
+ .B "\-\-memlimit\-compress=70%"
1115
+ .IP \(bu 3
1116
+ The
1117
+ .I limit
1118
+ can be reset back to its default value by setting it to
1119
+ .BR 0 .
1120
+ This is currently equivalent to setting the
1121
+ .I limit
1122
+ to
1123
+ .B max
1124
+ (no memory usage limit).
1125
+ .RE
1126
+ .IP ""
1127
+ For 32-bit
1128
+ .B xz
1129
+ there is a special case: if the
1130
+ .I limit
1131
+ would be over
1132
+ .BR "4020\ MiB" ,
1133
+ the
1134
+ .I limit
1135
+ is set to
1136
+ .BR "4020\ MiB" .
1137
+ On MIPS32
1138
+ .B "2000\ MiB"
1139
+ is used instead.
1140
+ (The values
1141
+ .B 0
1142
+ and
1143
+ .B max
1144
+ aren't affected by this.
1145
+ A similar feature doesn't exist for decompression.)
1146
+ This can be helpful when a 32-bit executable has access
1147
+ to 4\ GiB address space (2 GiB on MIPS32)
1148
+ while hopefully doing no harm in other situations.
1149
+ .IP ""
1150
+ See also the section
1151
+ .BR "Memory usage" .
1152
+ .TP
1153
+ .BI \-\-memlimit\-decompress= limit
1154
+ Set a memory usage limit for decompression.
1155
+ This also affects the
1156
+ .B \-\-list
1157
+ mode.
1158
+ If the operation is not possible without exceeding the
1159
+ .IR limit ,
1160
+ .B xz
1161
+ will display an error and decompressing the file will fail.
1162
+ See
1163
+ .BI \-\-memlimit\-compress= limit
1164
+ for possible ways to specify the
1165
+ .IR limit .
1166
+ .TP
1167
+ .BI \-\-memlimit\-mt\-decompress= limit
1168
+ Set a memory usage limit for multi-threaded decompression.
1169
+ This can only affect the number of threads;
1170
+ this will never make
1171
+ .B xz
1172
+ refuse to decompress a file.
1173
+ If
1174
+ .I limit
1175
+ is too low to allow any multi-threading, the
1176
+ .I limit
1177
+ is ignored and
1178
+ .B xz
1179
+ will continue in single-threaded mode.
1180
+ Note that if also
1181
+ .B \-\-memlimit\-decompress
1182
+ is used,
1183
+ it will always apply to both single-threaded and multi-threaded modes,
1184
+ and so the effective
1185
+ .I limit
1186
+ for multi-threading will never be higher than the limit set with
1187
+ .BR \-\-memlimit\-decompress .
1188
+ .IP ""
1189
+ In contrast to the other memory usage limit options,
1190
+ .BI \-\-memlimit\-mt\-decompress= limit
1191
+ has a system-specific default
1192
+ .IR limit .
1193
+ .B "xz \-\-info\-memory"
1194
+ can be used to see the current value.
1195
+ .IP ""
1196
+ This option and its default value exist
1197
+ because without any limit the threaded decompressor
1198
+ could end up allocating an insane amount of memory with some input files.
1199
+ If the default
1200
+ .I limit
1201
+ is too low on your system,
1202
+ feel free to increase the
1203
+ .I limit
1204
+ but never set it to a value larger than the amount of usable RAM
1205
+ as with appropriate input files
1206
+ .B xz
1207
+ will attempt to use that amount of memory
1208
+ even with a low number of threads.
1209
+ Running out of memory or swapping
1210
+ will not improve decompression performance.
1211
+ .IP ""
1212
+ See
1213
+ .BI \-\-memlimit\-compress= limit
1214
+ for possible ways to specify the
1215
+ .IR limit .
1216
+ Setting
1217
+ .I limit
1218
+ to
1219
+ .B 0
1220
+ resets the
1221
+ .I limit
1222
+ to the default system-specific value.
1223
+ .TP
1224
+ \fB\-M\fR \fIlimit\fR, \fB\-\-memlimit=\fIlimit\fR, \fB\-\-memory=\fIlimit
1225
+ This is equivalent to specifying
1226
+ .BI \-\-memlimit\-compress= limit
1227
+ .BI \-\-memlimit-decompress= limit
1228
+ \fB\-\-memlimit\-mt\-decompress=\fIlimit\fR.
1229
+ .TP
1230
+ .B \-\-no\-adjust
1231
+ Display an error and exit if the memory usage limit cannot be
1232
+ met without adjusting settings that affect the compressed output.
1233
+ That is, this prevents
1234
+ .B xz
1235
+ from switching the encoder from multi-threaded mode to single-threaded mode
1236
+ and from reducing the LZMA2 dictionary size.
1237
+ Even when this option is used the number of threads may be reduced
1238
+ to meet the memory usage limit as that won't affect the compressed output.
1239
+ .IP ""
1240
+ Automatic adjusting is always disabled when creating raw streams
1241
+ .RB ( \-\-format=raw ).
1242
+ .TP
1243
+ \fB\-T\fR \fIthreads\fR, \fB\-\-threads=\fIthreads
1244
+ Specify the number of worker threads to use.
1245
+ Setting
1246
+ .I threads
1247
+ to a special value
1248
+ .B 0
1249
+ makes
1250
+ .B xz
1251
+ use up to as many threads as the processor(s) on the system support.
1252
+ The actual number of threads can be fewer than
1253
+ .I threads
1254
+ if the input file is not big enough
1255
+ for threading with the given settings or
1256
+ if using more threads would exceed the memory usage limit.
1257
+ .IP ""
1258
+ The single-threaded and multi-threaded compressors produce different output.
1259
+ Single-threaded compressor will give the smallest file size but
1260
+ only the output from the multi-threaded compressor can be decompressed
1261
+ using multiple threads.
1262
+ Setting
1263
+ .I threads
1264
+ to
1265
+ .B 1
1266
+ will use the single-threaded mode.
1267
+ Setting
1268
+ .I threads
1269
+ to any other value, including
1270
+ .BR 0 ,
1271
+ will use the multi-threaded compressor
1272
+ even if the system supports only one hardware thread.
1273
+ .RB ( xz
1274
+ 5.2.x
1275
+ used single-threaded mode in this situation.)
1276
+ .IP ""
1277
+ To use multi-threaded mode with only one thread, set
1278
+ .I threads
1279
+ to
1280
+ .BR +1 .
1281
+ The
1282
+ .B +
1283
+ prefix has no effect with values other than
1284
+ .BR 1 .
1285
+ A memory usage limit can still make
1286
+ .B xz
1287
+ switch to single-threaded mode unless
1288
+ .B \-\-no\-adjust
1289
+ is used.
1290
+ Support for the
1291
+ .B +
1292
+ prefix was added in
1293
+ .B xz
1294
+ 5.4.0.
1295
+ .IP ""
1296
+ If an automatic number of threads has been requested and
1297
+ no memory usage limit has been specified,
1298
+ then a system-specific default soft limit will be used to possibly
1299
+ limit the number of threads.
1300
+ It is a soft limit in sense that it is ignored
1301
+ if the number of threads becomes one,
1302
+ thus a soft limit will never stop
1303
+ .B xz
1304
+ from compressing or decompressing.
1305
+ This default soft limit will not make
1306
+ .B xz
1307
+ switch from multi-threaded mode to single-threaded mode.
1308
+ The active limits can be seen with
1309
+ .BR "xz \-\-info\-memory" .
1310
+ .IP ""
1311
+ Currently the only threading method is to split the input into
1312
+ blocks and compress them independently from each other.
1313
+ The default block size depends on the compression level and
1314
+ can be overridden with the
1315
+ .BI \-\-block\-size= size
1316
+ option.
1317
+ .IP ""
1318
+ Threaded decompression only works on files that contain
1319
+ multiple blocks with size information in block headers.
1320
+ All large enough files compressed in multi-threaded mode
1321
+ meet this condition,
1322
+ but files compressed in single-threaded mode don't even if
1323
+ .BI \-\-block\-size= size
1324
+ has been used.
1325
+ .IP ""
1326
+ The default value for
1327
+ .I threads
1328
+ is
1329
+ .BR 0 .
1330
+ In
1331
+ .B xz
1332
+ 5.4.x and older the default is
1333
+ .BR 1 .
1334
+ .
1335
+ .SS "Custom compressor filter chains"
1336
+ A custom filter chain allows specifying
1337
+ the compression settings in detail instead of relying on
1338
+ the settings associated to the presets.
1339
+ When a custom filter chain is specified,
1340
+ preset options
1341
+ .RB ( \-0
1342
+ \&...\&
1343
+ .B \-9
1344
+ and
1345
+ .BR \-\-extreme )
1346
+ earlier on the command line are forgotten.
1347
+ If a preset option is specified
1348
+ after one or more custom filter chain options,
1349
+ the new preset takes effect and
1350
+ the custom filter chain options specified earlier are forgotten.
1351
+ .PP
1352
+ A filter chain is comparable to piping on the command line.
1353
+ When compressing, the uncompressed input goes to the first filter,
1354
+ whose output goes to the next filter (if any).
1355
+ The output of the last filter gets written to the compressed file.
1356
+ The maximum number of filters in the chain is four,
1357
+ but typically a filter chain has only one or two filters.
1358
+ .PP
1359
+ Many filters have limitations on where they can be
1360
+ in the filter chain:
1361
+ some filters can work only as the last filter in the chain,
1362
+ some only as a non-last filter, and some work in any position
1363
+ in the chain.
1364
+ Depending on the filter, this limitation is either inherent to
1365
+ the filter design or exists to prevent security issues.
1366
+ .PP
1367
+ A custom filter chain can be specified in two different ways.
1368
+ The options
1369
+ .BI \-\-filters= filters
1370
+ and
1371
+ .BI \-\-filters1= filters
1372
+ \&...\&
1373
+ .BI \-\-filters9= filters
1374
+ allow specifying an entire filter chain in one option using the
1375
+ liblzma filter string syntax.
1376
+ Alternatively, a filter chain can be specified by using one or more
1377
+ individual filter options in the order they are wanted in the filter chain.
1378
+ That is, the order of the individual filter options is significant!
1379
+ When decoding raw streams
1380
+ .RB ( \-\-format=raw ),
1381
+ the filter chain must be specified in the same order as
1382
+ it was specified when compressing.
1383
+ Any individual filter or preset options specified before the full
1384
+ chain option
1385
+ (\fB\-\-filters=\fIfilters\fR)
1386
+ will be forgotten.
1387
+ Individual filters specified after the full chain option will reset the
1388
+ filter chain.
1389
+ .PP
1390
+ Both the full and individual filter options take filter-specific
1391
+ .I options
1392
+ as a comma-separated list.
1393
+ Extra commas in
1394
+ .I options
1395
+ are ignored.
1396
+ Every option has a default value, so
1397
+ specify those you want to change.
1398
+ .PP
1399
+ To see the whole filter chain and
1400
+ .IR options ,
1401
+ use
1402
+ .B "xz \-vv"
1403
+ (that is, use
1404
+ .B \-\-verbose
1405
+ twice).
1406
+ This works also for viewing the filter chain options used by presets.
1407
+ .TP
1408
+ .BI \-\-filters= filters
1409
+ Specify the full filter chain or a preset in a single option.
1410
+ Each filter can be separated by spaces or two dashes
1411
+ .RB ( \-\- ).
1412
+ .I filters
1413
+ may need to be quoted on the shell command line so it is
1414
+ parsed as a single option.
1415
+ To denote
1416
+ .IR options ,
1417
+ use
1418
+ .B :
1419
+ or
1420
+ .BR = .
1421
+ A preset can be prefixed with a
1422
+ .B \-
1423
+ and followed with zero or more flags.
1424
+ The only supported flag is
1425
+ .B e
1426
+ to apply the same options as
1427
+ .BR \-\-extreme .
1428
+ .TP
1429
+ \fB\-\-filters1\fR=\fIfilters\fR ... \fB\-\-filters9\fR=\fIfilters
1430
+ Specify up to nine additional filter chains that can be used with
1431
+ .BR \-\-block\-list .
1432
+ .IP ""
1433
+ For example, when compressing an archive with executable files
1434
+ followed by text files, the executable part could use a filter
1435
+ chain with a BCJ filter and the text part only the LZMA2 filter.
1436
+ .TP
1437
+ .B \-\-filters-help
1438
+ Display a help message describing how to specify presets and
1439
+ custom filter chains in the
1440
+ .B \-\-filters
1441
+ and
1442
+ .BI \-\-filters1= filters
1443
+ \&...\&
1444
+ .BI \-\-filters9= filters
1445
+ options, and exit successfully.
1446
+ .TP
1447
+ \fB\-\-lzma1\fR[\fB=\fIoptions\fR]
1448
+ .PD 0
1449
+ .TP
1450
+ \fB\-\-lzma2\fR[\fB=\fIoptions\fR]
1451
+ .PD
1452
+ Add LZMA1 or LZMA2 filter to the filter chain.
1453
+ These filters can be used only as the last filter in the chain.
1454
+ .IP ""
1455
+ LZMA1 is a legacy filter,
1456
+ which is supported almost solely due to the legacy
1457
+ .B .lzma
1458
+ file format, which supports only LZMA1.
1459
+ LZMA2 is an updated
1460
+ version of LZMA1 to fix some practical issues of LZMA1.
1461
+ The
1462
+ .B .xz
1463
+ format uses LZMA2 and doesn't support LZMA1 at all.
1464
+ Compression speed and ratios of LZMA1 and LZMA2
1465
+ are practically the same.
1466
+ .IP ""
1467
+ LZMA1 and LZMA2 share the same set of
1468
+ .IR options :
1469
+ .RS
1470
+ .TP
1471
+ .BI preset= preset
1472
+ Reset all LZMA1 or LZMA2
1473
+ .I options
1474
+ to
1475
+ .IR preset .
1476
+ .I Preset
1477
+ consist of an integer, which may be followed by single-letter
1478
+ preset modifiers.
1479
+ The integer can be from
1480
+ .B 0
1481
+ to
1482
+ .BR 9 ,
1483
+ matching the command line options
1484
+ .B \-0
1485
+ \&...\&
1486
+ .BR \-9 .
1487
+ The only supported modifier is currently
1488
+ .BR e ,
1489
+ which matches
1490
+ .BR \-\-extreme .
1491
+ If no
1492
+ .B preset
1493
+ is specified, the default values of LZMA1 or LZMA2
1494
+ .I options
1495
+ are taken from the preset
1496
+ .BR 6 .
1497
+ .TP
1498
+ .BI dict= size
1499
+ Dictionary (history buffer)
1500
+ .I size
1501
+ indicates how many bytes of the recently processed
1502
+ uncompressed data is kept in memory.
1503
+ The algorithm tries to find repeating byte sequences (matches) in
1504
+ the uncompressed data, and replace them with references
1505
+ to the data currently in the dictionary.
1506
+ The bigger the dictionary, the higher is the chance
1507
+ to find a match.
1508
+ Thus, increasing dictionary
1509
+ .I size
1510
+ usually improves compression ratio, but
1511
+ a dictionary bigger than the uncompressed file is waste of memory.
1512
+ .IP ""
1513
+ Typical dictionary
1514
+ .I size
1515
+ is from 64\ KiB to 64\ MiB.
1516
+ The minimum is 4\ KiB.
1517
+ The maximum for compression is currently 1.5\ GiB (1536\ MiB).
1518
+ The decompressor already supports dictionaries up to
1519
+ one byte less than 4\ GiB, which is the maximum for
1520
+ the LZMA1 and LZMA2 stream formats.
1521
+ .IP ""
1522
+ Dictionary
1523
+ .I size
1524
+ and match finder
1525
+ .RI ( mf )
1526
+ together determine the memory usage of the LZMA1 or LZMA2 encoder.
1527
+ The same (or bigger) dictionary
1528
+ .I size
1529
+ is required for decompressing that was used when compressing,
1530
+ thus the memory usage of the decoder is determined
1531
+ by the dictionary size used when compressing.
1532
+ The
1533
+ .B .xz
1534
+ headers store the dictionary
1535
+ .I size
1536
+ either as
1537
+ .RI "2^" n
1538
+ or
1539
+ .RI "2^" n " + 2^(" n "\-1),"
1540
+ so these
1541
+ .I sizes
1542
+ are somewhat preferred for compression.
1543
+ Other
1544
+ .I sizes
1545
+ will get rounded up when stored in the
1546
+ .B .xz
1547
+ headers.
1548
+ .TP
1549
+ .BI lc= lc
1550
+ Specify the number of literal context bits.
1551
+ The minimum is 0 and the maximum is 4; the default is 3.
1552
+ In addition, the sum of
1553
+ .I lc
1554
+ and
1555
+ .I lp
1556
+ must not exceed 4.
1557
+ .IP ""
1558
+ All bytes that cannot be encoded as matches
1559
+ are encoded as literals.
1560
+ That is, literals are simply 8-bit bytes
1561
+ that are encoded one at a time.
1562
+ .IP ""
1563
+ The literal coding makes an assumption that the highest
1564
+ .I lc
1565
+ bits of the previous uncompressed byte correlate
1566
+ with the next byte.
1567
+ For example, in typical English text, an upper-case letter is
1568
+ often followed by a lower-case letter, and a lower-case
1569
+ letter is usually followed by another lower-case letter.
1570
+ In the US-ASCII character set, the highest three bits are 010
1571
+ for upper-case letters and 011 for lower-case letters.
1572
+ When
1573
+ .I lc
1574
+ is at least 3, the literal coding can take advantage of
1575
+ this property in the uncompressed data.
1576
+ .IP ""
1577
+ The default value (3) is usually good.
1578
+ If you want maximum compression, test
1579
+ .BR lc=4 .
1580
+ Sometimes it helps a little, and
1581
+ sometimes it makes compression worse.
1582
+ If it makes it worse, test
1583
+ .B lc=2
1584
+ too.
1585
+ .TP
1586
+ .BI lp= lp
1587
+ Specify the number of literal position bits.
1588
+ The minimum is 0 and the maximum is 4; the default is 0.
1589
+ .IP ""
1590
+ .I Lp
1591
+ affects what kind of alignment in the uncompressed data is
1592
+ assumed when encoding literals.
1593
+ See
1594
+ .I pb
1595
+ below for more information about alignment.
1596
+ .TP
1597
+ .BI pb= pb
1598
+ Specify the number of position bits.
1599
+ The minimum is 0 and the maximum is 4; the default is 2.
1600
+ .IP ""
1601
+ .I Pb
1602
+ affects what kind of alignment in the uncompressed data is
1603
+ assumed in general.
1604
+ The default means four-byte alignment
1605
+ .RI (2^ pb =2^2=4),
1606
+ which is often a good choice when there's no better guess.
1607
+ .IP ""
1608
+ When the alignment is known, setting
1609
+ .I pb
1610
+ accordingly may reduce the file size a little.
1611
+ For example, with text files having one-byte
1612
+ alignment (US-ASCII, ISO-8859-*, UTF-8), setting
1613
+ .B pb=0
1614
+ can improve compression slightly.
1615
+ For UTF-16 text,
1616
+ .B pb=1
1617
+ is a good choice.
1618
+ If the alignment is an odd number like 3 bytes,
1619
+ .B pb=0
1620
+ might be the best choice.
1621
+ .IP ""
1622
+ Even though the assumed alignment can be adjusted with
1623
+ .I pb
1624
+ and
1625
+ .IR lp ,
1626
+ LZMA1 and LZMA2 still slightly favor 16-byte alignment.
1627
+ It might be worth taking into account when designing file formats
1628
+ that are likely to be often compressed with LZMA1 or LZMA2.
1629
+ .TP
1630
+ .BI mf= mf
1631
+ Match finder has a major effect on encoder speed,
1632
+ memory usage, and compression ratio.
1633
+ Usually Hash Chain match finders are faster than Binary Tree
1634
+ match finders.
1635
+ The default depends on the
1636
+ .IR preset :
1637
+ 0 uses
1638
+ .BR hc3 ,
1639
+ 1\(en3
1640
+ use
1641
+ .BR hc4 ,
1642
+ and the rest use
1643
+ .BR bt4 .
1644
+ .IP ""
1645
+ The following match finders are supported.
1646
+ The memory usage formulas below are rough approximations,
1647
+ which are closest to the reality when
1648
+ .I dict
1649
+ is a power of two.
1650
+ .RS
1651
+ .TP
1652
+ .B hc3
1653
+ Hash Chain with 2- and 3-byte hashing
1654
+ .br
1655
+ Minimum value for
1656
+ .IR nice :
1657
+ 3
1658
+ .br
1659
+ Memory usage:
1660
+ .br
1661
+ .I dict
1662
+ * 7.5 (if
1663
+ .I dict
1664
+ <= 16 MiB);
1665
+ .br
1666
+ .I dict
1667
+ * 5.5 + 64 MiB (if
1668
+ .I dict
1669
+ > 16 MiB)
1670
+ .TP
1671
+ .B hc4
1672
+ Hash Chain with 2-, 3-, and 4-byte hashing
1673
+ .br
1674
+ Minimum value for
1675
+ .IR nice :
1676
+ 4
1677
+ .br
1678
+ Memory usage:
1679
+ .br
1680
+ .I dict
1681
+ * 7.5 (if
1682
+ .I dict
1683
+ <= 32 MiB);
1684
+ .br
1685
+ .I dict
1686
+ * 6.5 (if
1687
+ .I dict
1688
+ > 32 MiB)
1689
+ .TP
1690
+ .B bt2
1691
+ Binary Tree with 2-byte hashing
1692
+ .br
1693
+ Minimum value for
1694
+ .IR nice :
1695
+ 2
1696
+ .br
1697
+ Memory usage:
1698
+ .I dict
1699
+ * 9.5
1700
+ .TP
1701
+ .B bt3
1702
+ Binary Tree with 2- and 3-byte hashing
1703
+ .br
1704
+ Minimum value for
1705
+ .IR nice :
1706
+ 3
1707
+ .br
1708
+ Memory usage:
1709
+ .br
1710
+ .I dict
1711
+ * 11.5 (if
1712
+ .I dict
1713
+ <= 16 MiB);
1714
+ .br
1715
+ .I dict
1716
+ * 9.5 + 64 MiB (if
1717
+ .I dict
1718
+ > 16 MiB)
1719
+ .TP
1720
+ .B bt4
1721
+ Binary Tree with 2-, 3-, and 4-byte hashing
1722
+ .br
1723
+ Minimum value for
1724
+ .IR nice :
1725
+ 4
1726
+ .br
1727
+ Memory usage:
1728
+ .br
1729
+ .I dict
1730
+ * 11.5 (if
1731
+ .I dict
1732
+ <= 32 MiB);
1733
+ .br
1734
+ .I dict
1735
+ * 10.5 (if
1736
+ .I dict
1737
+ > 32 MiB)
1738
+ .RE
1739
+ .TP
1740
+ .BI mode= mode
1741
+ Compression
1742
+ .I mode
1743
+ specifies the method to analyze
1744
+ the data produced by the match finder.
1745
+ Supported
1746
+ .I modes
1747
+ are
1748
+ .B fast
1749
+ and
1750
+ .BR normal .
1751
+ The default is
1752
+ .B fast
1753
+ for
1754
+ .I presets
1755
+ 0\(en3 and
1756
+ .B normal
1757
+ for
1758
+ .I presets
1759
+ 4\(en9.
1760
+ .IP ""
1761
+ Usually
1762
+ .B fast
1763
+ is used with Hash Chain match finders and
1764
+ .B normal
1765
+ with Binary Tree match finders.
1766
+ This is also what the
1767
+ .I presets
1768
+ do.
1769
+ .TP
1770
+ .BI nice= nice
1771
+ Specify what is considered to be a nice length for a match.
1772
+ Once a match of at least
1773
+ .I nice
1774
+ bytes is found, the algorithm stops
1775
+ looking for possibly better matches.
1776
+ .IP ""
1777
+ .I Nice
1778
+ can be 2\(en273 bytes.
1779
+ Higher values tend to give better compression ratio
1780
+ at the expense of speed.
1781
+ The default depends on the
1782
+ .IR preset .
1783
+ .TP
1784
+ .BI depth= depth
1785
+ Specify the maximum search depth in the match finder.
1786
+ The default is the special value of 0,
1787
+ which makes the compressor determine a reasonable
1788
+ .I depth
1789
+ from
1790
+ .I mf
1791
+ and
1792
+ .IR nice .
1793
+ .IP ""
1794
+ Reasonable
1795
+ .I depth
1796
+ for Hash Chains is 4\(en100 and 16\(en1000 for Binary Trees.
1797
+ Using very high values for
1798
+ .I depth
1799
+ can make the encoder extremely slow with some files.
1800
+ Avoid setting the
1801
+ .I depth
1802
+ over 1000 unless you are prepared to interrupt
1803
+ the compression in case it is taking far too long.
1804
+ .RE
1805
+ .IP ""
1806
+ When decoding raw streams
1807
+ .RB ( \-\-format=raw ),
1808
+ LZMA2 needs only the dictionary
1809
+ .IR size .
1810
+ LZMA1 needs also
1811
+ .IR lc ,
1812
+ .IR lp ,
1813
+ and
1814
+ .IR pb .
1815
+ .TP
1816
+ \fB\-\-x86\fR[\fB=\fIoptions\fR]
1817
+ .PD 0
1818
+ .TP
1819
+ \fB\-\-arm\fR[\fB=\fIoptions\fR]
1820
+ .TP
1821
+ \fB\-\-armthumb\fR[\fB=\fIoptions\fR]
1822
+ .TP
1823
+ \fB\-\-arm64\fR[\fB=\fIoptions\fR]
1824
+ .TP
1825
+ \fB\-\-powerpc\fR[\fB=\fIoptions\fR]
1826
+ .TP
1827
+ \fB\-\-ia64\fR[\fB=\fIoptions\fR]
1828
+ .TP
1829
+ \fB\-\-sparc\fR[\fB=\fIoptions\fR]
1830
+ .TP
1831
+ \fB\-\-riscv\fR[\fB=\fIoptions\fR]
1832
+ .PD
1833
+ Add a branch/call/jump (BCJ) filter to the filter chain.
1834
+ These filters can be used only as a non-last filter
1835
+ in the filter chain.
1836
+ .IP ""
1837
+ A BCJ filter converts relative addresses in
1838
+ the machine code to their absolute counterparts.
1839
+ This doesn't change the size of the data
1840
+ but it increases redundancy,
1841
+ which can help LZMA2 to produce 0\(en15\ % smaller
1842
+ .B .xz
1843
+ file.
1844
+ The BCJ filters are always reversible,
1845
+ so using a BCJ filter for wrong type of data
1846
+ doesn't cause any data loss, although it may make
1847
+ the compression ratio slightly worse.
1848
+ The BCJ filters are very fast and
1849
+ use an insignificant amount of memory.
1850
+ .IP ""
1851
+ These BCJ filters have known problems related to
1852
+ the compression ratio:
1853
+ .RS
1854
+ .IP \(bu 3
1855
+ Some types of files containing executable code
1856
+ (for example, object files, static libraries, and Linux kernel modules)
1857
+ have the addresses in the instructions filled with filler values.
1858
+ These BCJ filters will still do the address conversion,
1859
+ which will make the compression worse with these files.
1860
+ .IP \(bu 3
1861
+ If a BCJ filter is applied on an archive,
1862
+ it is possible that it makes the compression ratio
1863
+ worse than not using a BCJ filter.
1864
+ For example, if there are similar or even identical executables
1865
+ then filtering will likely make the files less similar
1866
+ and thus compression is worse.
1867
+ The contents of non-executable files in the same archive can matter too.
1868
+ In practice one has to try with and without a BCJ filter to see
1869
+ which is better in each situation.
1870
+ .RE
1871
+ .IP ""
1872
+ Different instruction sets have different alignment:
1873
+ the executable file must be aligned to a multiple of
1874
+ this value in the input data to make the filter work.
1875
+ .RS
1876
+ .RS
1877
+ .PP
1878
+ .TS
1879
+ tab(;);
1880
+ l n l
1881
+ l n l.
1882
+ Filter;Alignment;Notes
1883
+ x86;1;32-bit or 64-bit x86
1884
+ ARM;4;
1885
+ ARM-Thumb;2;
1886
+ ARM64;4;4096-byte alignment is best
1887
+ PowerPC;4;Big endian only
1888
+ IA-64;16;Itanium
1889
+ SPARC;4;
1890
+ RISC-V;2;
1891
+ .TE
1892
+ .RE
1893
+ .RE
1894
+ .IP ""
1895
+ Since the BCJ-filtered data is usually compressed with LZMA2,
1896
+ the compression ratio may be improved slightly if
1897
+ the LZMA2 options are set to match the
1898
+ alignment of the selected BCJ filter.
1899
+ Examples:
1900
+ .RS
1901
+ .IP \(bu 3
1902
+ IA-64 filter has 16-byte alignment so
1903
+ .B pb=4,lp=4,lc=0
1904
+ is good
1905
+ with LZMA2 (2^4=16).
1906
+ .IP \(bu 3
1907
+ RISC-V code has 2-byte or 4-byte alignment
1908
+ depending on whether the file contains
1909
+ 16-bit compressed instructions (the C extension).
1910
+ When 16-bit instructions are used,
1911
+ .B pb=2,lp=1,lc=3
1912
+ or
1913
+ .B pb=1,lp=1,lc=3
1914
+ is good.
1915
+ When 16-bit instructions aren't present,
1916
+ .B pb=2,lp=2,lc=2
1917
+ is the best.
1918
+ .B readelf \-h
1919
+ can be used to check if "RVC"
1920
+ appears on the "Flags" line.
1921
+ .IP \(bu 3
1922
+ ARM64 is always 4-byte aligned so
1923
+ .B pb=2,lp=2,lc=2
1924
+ is the best.
1925
+ .IP \(bu 3
1926
+ The x86 filter is an exception.
1927
+ It's usually good to stick to LZMA2's defaults
1928
+ .RB ( pb=2,lp=0,lc=3 )
1929
+ when compressing x86 executables.
1930
+ .RE
1931
+ .IP ""
1932
+ All BCJ filters support the same
1933
+ .IR options :
1934
+ .RS
1935
+ .TP
1936
+ .BI start= offset
1937
+ Specify the start
1938
+ .I offset
1939
+ that is used when converting between relative
1940
+ and absolute addresses.
1941
+ The
1942
+ .I offset
1943
+ must be a multiple of the alignment of the filter
1944
+ (see the table above).
1945
+ The default is zero.
1946
+ In practice, the default is good; specifying a custom
1947
+ .I offset
1948
+ is almost never useful.
1949
+ .RE
1950
+ .TP
1951
+ \fB\-\-delta\fR[\fB=\fIoptions\fR]
1952
+ Add the Delta filter to the filter chain.
1953
+ The Delta filter can be only used as a non-last filter
1954
+ in the filter chain.
1955
+ .IP ""
1956
+ Currently only simple byte-wise delta calculation is supported.
1957
+ It can be useful when compressing, for example, uncompressed bitmap images
1958
+ or uncompressed PCM audio.
1959
+ However, special purpose algorithms may give significantly better
1960
+ results than Delta + LZMA2.
1961
+ This is true especially with audio,
1962
+ which compresses faster and better, for example, with
1963
+ .BR flac (1).
1964
+ .IP ""
1965
+ Supported
1966
+ .IR options :
1967
+ .RS
1968
+ .TP
1969
+ .BI dist= distance
1970
+ Specify the
1971
+ .I distance
1972
+ of the delta calculation in bytes.
1973
+ .I distance
1974
+ must be 1\(en256.
1975
+ The default is 1.
1976
+ .IP ""
1977
+ For example, with
1978
+ .B dist=2
1979
+ and eight-byte input A1 B1 A2 B3 A3 B5 A4 B7, the output will be
1980
+ A1 B1 01 02 01 02 01 02.
1981
+ .RE
1982
+ .
1983
+ .SS "Other options"
1984
+ .TP
1985
+ .BR \-q ", " \-\-quiet
1986
+ Suppress warnings and notices.
1987
+ Specify this twice to suppress errors too.
1988
+ This option has no effect on the exit status.
1989
+ That is, even if a warning was suppressed,
1990
+ the exit status to indicate a warning is still used.
1991
+ .TP
1992
+ .BR \-v ", " \-\-verbose
1993
+ Be verbose.
1994
+ If standard error is connected to a terminal,
1995
+ .B xz
1996
+ will display a progress indicator.
1997
+ Specifying
1998
+ .B \-\-verbose
1999
+ twice will give even more verbose output.
2000
+ .IP ""
2001
+ The progress indicator shows the following information:
2002
+ .RS
2003
+ .IP \(bu 3
2004
+ Completion percentage is shown
2005
+ if the size of the input file is known.
2006
+ That is, the percentage cannot be shown in pipes.
2007
+ .IP \(bu 3
2008
+ Amount of compressed data produced (compressing)
2009
+ or consumed (decompressing).
2010
+ .IP \(bu 3
2011
+ Amount of uncompressed data consumed (compressing)
2012
+ or produced (decompressing).
2013
+ .IP \(bu 3
2014
+ Compression ratio, which is calculated by dividing
2015
+ the amount of compressed data processed so far by
2016
+ the amount of uncompressed data processed so far.
2017
+ .IP \(bu 3
2018
+ Compression or decompression speed.
2019
+ This is measured as the amount of uncompressed data consumed
2020
+ (compression) or produced (decompression) per second.
2021
+ It is shown after a few seconds have passed since
2022
+ .B xz
2023
+ started processing the file.
2024
+ .IP \(bu 3
2025
+ Elapsed time in the format M:SS or H:MM:SS.
2026
+ .IP \(bu 3
2027
+ Estimated remaining time is shown
2028
+ only when the size of the input file is
2029
+ known and a couple of seconds have already passed since
2030
+ .B xz
2031
+ started processing the file.
2032
+ The time is shown in a less precise format which
2033
+ never has any colons, for example, 2 min 30 s.
2034
+ .RE
2035
+ .IP ""
2036
+ When standard error is not a terminal,
2037
+ .B \-\-verbose
2038
+ will make
2039
+ .B xz
2040
+ print the filename, compressed size, uncompressed size,
2041
+ compression ratio, and possibly also the speed and elapsed time
2042
+ on a single line to standard error after compressing or
2043
+ decompressing the file.
2044
+ The speed and elapsed time are included only when
2045
+ the operation took at least a few seconds.
2046
+ If the operation didn't finish, for example, due to user interruption,
2047
+ also the completion percentage is printed
2048
+ if the size of the input file is known.
2049
+ .TP
2050
+ .BR \-Q ", " \-\-no\-warn
2051
+ Don't set the exit status to 2
2052
+ even if a condition worth a warning was detected.
2053
+ This option doesn't affect the verbosity level, thus both
2054
+ .B \-\-quiet
2055
+ and
2056
+ .B \-\-no\-warn
2057
+ have to be used to not display warnings and
2058
+ to not alter the exit status.
2059
+ .TP
2060
+ .B \-\-robot
2061
+ Print messages in a machine-parsable format.
2062
+ This is intended to ease writing frontends that want to use
2063
+ .B xz
2064
+ instead of liblzma, which may be the case with various scripts.
2065
+ The output with this option enabled is meant to be stable across
2066
+ .B xz
2067
+ releases.
2068
+ See the section
2069
+ .B "ROBOT MODE"
2070
+ for details.
2071
+ .TP
2072
+ .B \-\-info\-memory
2073
+ Display, in human-readable format, how much physical memory (RAM)
2074
+ and how many processor threads
2075
+ .B xz
2076
+ thinks the system has and the memory usage limits for compression
2077
+ and decompression, and exit successfully.
2078
+ .TP
2079
+ .BR \-h ", " \-\-help
2080
+ Display a help message describing the most commonly used options,
2081
+ and exit successfully.
2082
+ .TP
2083
+ .BR \-H ", " \-\-long\-help
2084
+ Display a help message describing all features of
2085
+ .BR xz ,
2086
+ and exit successfully
2087
+ .TP
2088
+ .BR \-V ", " \-\-version
2089
+ Display the version number of
2090
+ .B xz
2091
+ and liblzma in human readable format.
2092
+ To get machine-parsable output, specify
2093
+ .B \-\-robot
2094
+ before
2095
+ .BR \-\-version .
2096
+ .
2097
+ .SH "ROBOT MODE"
2098
+ The robot mode is activated with the
2099
+ .B \-\-robot
2100
+ option.
2101
+ It makes the output of
2102
+ .B xz
2103
+ easier to parse by other programs.
2104
+ Currently
2105
+ .B \-\-robot
2106
+ is supported only together with
2107
+ .BR \-\-list ,
2108
+ .BR \-\-filters\-help ,
2109
+ .BR \-\-info\-memory ,
2110
+ and
2111
+ .BR \-\-version .
2112
+ It will be supported for compression and
2113
+ decompression in the future.
2114
+ .
2115
+ .SS "List mode"
2116
+ .B "xz \-\-robot \-\-list"
2117
+ uses tab-separated output.
2118
+ The first column of every line has a string
2119
+ that indicates the type of the information found on that line:
2120
+ .TP
2121
+ .B name
2122
+ This is always the first line when starting to list a file.
2123
+ The second column on the line is the filename.
2124
+ .TP
2125
+ .B file
2126
+ This line contains overall information about the
2127
+ .B .xz
2128
+ file.
2129
+ This line is always printed after the
2130
+ .B name
2131
+ line.
2132
+ .TP
2133
+ .B stream
2134
+ This line type is used only when
2135
+ .B \-\-verbose
2136
+ was specified.
2137
+ There are as many
2138
+ .B stream
2139
+ lines as there are streams in the
2140
+ .B .xz
2141
+ file.
2142
+ .TP
2143
+ .B block
2144
+ This line type is used only when
2145
+ .B \-\-verbose
2146
+ was specified.
2147
+ There are as many
2148
+ .B block
2149
+ lines as there are blocks in the
2150
+ .B .xz
2151
+ file.
2152
+ The
2153
+ .B block
2154
+ lines are shown after all the
2155
+ .B stream
2156
+ lines; different line types are not interleaved.
2157
+ .TP
2158
+ .B summary
2159
+ This line type is used only when
2160
+ .B \-\-verbose
2161
+ was specified twice.
2162
+ This line is printed after all
2163
+ .B block
2164
+ lines.
2165
+ Like the
2166
+ .B file
2167
+ line, the
2168
+ .B summary
2169
+ line contains overall information about the
2170
+ .B .xz
2171
+ file.
2172
+ .TP
2173
+ .B totals
2174
+ This line is always the very last line of the list output.
2175
+ It shows the total counts and sizes.
2176
+ .PP
2177
+ The columns of the
2178
+ .B file
2179
+ lines:
2180
+ .PD 0
2181
+ .RS
2182
+ .IP 2. 4
2183
+ Number of streams in the file
2184
+ .IP 3. 4
2185
+ Total number of blocks in the stream(s)
2186
+ .IP 4. 4
2187
+ Compressed size of the file
2188
+ .IP 5. 4
2189
+ Uncompressed size of the file
2190
+ .IP 6. 4
2191
+ Compression ratio, for example,
2192
+ .BR 0.123 .
2193
+ If ratio is over 9.999, three dashes
2194
+ .RB ( \-\-\- )
2195
+ are displayed instead of the ratio.
2196
+ .IP 7. 4
2197
+ Comma-separated list of integrity check names.
2198
+ The following strings are used for the known check types:
2199
+ .BR None ,
2200
+ .BR CRC32 ,
2201
+ .BR CRC64 ,
2202
+ and
2203
+ .BR SHA\-256 .
2204
+ For unknown check types,
2205
+ .BI Unknown\- N
2206
+ is used, where
2207
+ .I N
2208
+ is the Check ID as a decimal number (one or two digits).
2209
+ .IP 8. 4
2210
+ Total size of stream padding in the file
2211
+ .RE
2212
+ .PD
2213
+ .PP
2214
+ The columns of the
2215
+ .B stream
2216
+ lines:
2217
+ .PD 0
2218
+ .RS
2219
+ .IP 2. 4
2220
+ Stream number (the first stream is 1)
2221
+ .IP 3. 4
2222
+ Number of blocks in the stream
2223
+ .IP 4. 4
2224
+ Compressed start offset
2225
+ .IP 5. 4
2226
+ Uncompressed start offset
2227
+ .IP 6. 4
2228
+ Compressed size (does not include stream padding)
2229
+ .IP 7. 4
2230
+ Uncompressed size
2231
+ .IP 8. 4
2232
+ Compression ratio
2233
+ .IP 9. 4
2234
+ Name of the integrity check
2235
+ .IP 10. 4
2236
+ Size of stream padding
2237
+ .RE
2238
+ .PD
2239
+ .PP
2240
+ The columns of the
2241
+ .B block
2242
+ lines:
2243
+ .PD 0
2244
+ .RS
2245
+ .IP 2. 4
2246
+ Number of the stream containing this block
2247
+ .IP 3. 4
2248
+ Block number relative to the beginning of the stream
2249
+ (the first block is 1)
2250
+ .IP 4. 4
2251
+ Block number relative to the beginning of the file
2252
+ .IP 5. 4
2253
+ Compressed start offset relative to the beginning of the file
2254
+ .IP 6. 4
2255
+ Uncompressed start offset relative to the beginning of the file
2256
+ .IP 7. 4
2257
+ Total compressed size of the block (includes headers)
2258
+ .IP 8. 4
2259
+ Uncompressed size
2260
+ .IP 9. 4
2261
+ Compression ratio
2262
+ .IP 10. 4
2263
+ Name of the integrity check
2264
+ .RE
2265
+ .PD
2266
+ .PP
2267
+ If
2268
+ .B \-\-verbose
2269
+ was specified twice, additional columns are included on the
2270
+ .B block
2271
+ lines.
2272
+ These are not displayed with a single
2273
+ .BR \-\-verbose ,
2274
+ because getting this information requires many seeks
2275
+ and can thus be slow:
2276
+ .PD 0
2277
+ .RS
2278
+ .IP 11. 4
2279
+ Value of the integrity check in hexadecimal
2280
+ .IP 12. 4
2281
+ Block header size
2282
+ .IP 13. 4
2283
+ Block flags:
2284
+ .B c
2285
+ indicates that compressed size is present, and
2286
+ .B u
2287
+ indicates that uncompressed size is present.
2288
+ If the flag is not set, a dash
2289
+ .RB ( \- )
2290
+ is shown instead to keep the string length fixed.
2291
+ New flags may be added to the end of the string in the future.
2292
+ .IP 14. 4
2293
+ Size of the actual compressed data in the block (this excludes
2294
+ the block header, block padding, and check fields)
2295
+ .IP 15. 4
2296
+ Amount of memory (in bytes) required to decompress
2297
+ this block with this
2298
+ .B xz
2299
+ version
2300
+ .IP 16. 4
2301
+ Filter chain.
2302
+ Note that most of the options used at compression time
2303
+ cannot be known, because only the options
2304
+ that are needed for decompression are stored in the
2305
+ .B .xz
2306
+ headers.
2307
+ .RE
2308
+ .PD
2309
+ .PP
2310
+ The columns of the
2311
+ .B summary
2312
+ lines:
2313
+ .PD 0
2314
+ .RS
2315
+ .IP 2. 4
2316
+ Amount of memory (in bytes) required to decompress
2317
+ this file with this
2318
+ .B xz
2319
+ version
2320
+ .IP 3. 4
2321
+ .B yes
2322
+ or
2323
+ .B no
2324
+ indicating if all block headers have both compressed size and
2325
+ uncompressed size stored in them
2326
+ .PP
2327
+ .I Since
2328
+ .B xz
2329
+ .I 5.1.2alpha:
2330
+ .IP 4. 4
2331
+ Minimum
2332
+ .B xz
2333
+ version required to decompress the file
2334
+ .RE
2335
+ .PD
2336
+ .PP
2337
+ The columns of the
2338
+ .B totals
2339
+ line:
2340
+ .PD 0
2341
+ .RS
2342
+ .IP 2. 4
2343
+ Number of streams
2344
+ .IP 3. 4
2345
+ Number of blocks
2346
+ .IP 4. 4
2347
+ Compressed size
2348
+ .IP 5. 4
2349
+ Uncompressed size
2350
+ .IP 6. 4
2351
+ Average compression ratio
2352
+ .IP 7. 4
2353
+ Comma-separated list of integrity check names
2354
+ that were present in the files
2355
+ .IP 8. 4
2356
+ Stream padding size
2357
+ .IP 9. 4
2358
+ Number of files.
2359
+ This is here to
2360
+ keep the order of the earlier columns the same as on
2361
+ .B file
2362
+ lines.
2363
+ .PD
2364
+ .RE
2365
+ .PP
2366
+ If
2367
+ .B \-\-verbose
2368
+ was specified twice, additional columns are included on the
2369
+ .B totals
2370
+ line:
2371
+ .PD 0
2372
+ .RS
2373
+ .IP 10. 4
2374
+ Maximum amount of memory (in bytes) required to decompress
2375
+ the files with this
2376
+ .B xz
2377
+ version
2378
+ .IP 11. 4
2379
+ .B yes
2380
+ or
2381
+ .B no
2382
+ indicating if all block headers have both compressed size and
2383
+ uncompressed size stored in them
2384
+ .PP
2385
+ .I Since
2386
+ .B xz
2387
+ .I 5.1.2alpha:
2388
+ .IP 12. 4
2389
+ Minimum
2390
+ .B xz
2391
+ version required to decompress the file
2392
+ .RE
2393
+ .PD
2394
+ .PP
2395
+ Future versions may add new line types and
2396
+ new columns can be added to the existing line types,
2397
+ but the existing columns won't be changed.
2398
+ .
2399
+ .SS "Filters help"
2400
+ .B "xz \-\-robot \-\-filters-help"
2401
+ prints the supported filters in the following format:
2402
+ .PP
2403
+ \fIfilter\fB:\fIoption\fB=<\fIvalue\fB>,\fIoption\fB=<\fIvalue\fB>\fR...
2404
+ .TP
2405
+ .I filter
2406
+ Name of the filter
2407
+ .TP
2408
+ .I option
2409
+ Name of a filter specific option
2410
+ .TP
2411
+ .I value
2412
+ Numeric
2413
+ .I value
2414
+ ranges appear as
2415
+ \fB<\fImin\fB\-\fImax\fB>\fR.
2416
+ String
2417
+ .I value
2418
+ choices are shown within
2419
+ .B "< >"
2420
+ and separated by a
2421
+ .B |
2422
+ character.
2423
+ .PP
2424
+ Each filter is printed on its own line.
2425
+ .
2426
+ .SS "Memory limit information"
2427
+ .B "xz \-\-robot \-\-info\-memory"
2428
+ prints a single line with multiple tab-separated columns:
2429
+ .IP 1. 4
2430
+ Total amount of physical memory (RAM) in bytes.
2431
+ .IP 2. 4
2432
+ Memory usage limit for compression in bytes
2433
+ .RB ( \-\-memlimit\-compress ).
2434
+ A special value of
2435
+ .B 0
2436
+ indicates the default setting
2437
+ which for single-threaded mode is the same as no limit.
2438
+ .IP 3. 4
2439
+ Memory usage limit for decompression in bytes
2440
+ .RB ( \-\-memlimit\-decompress ).
2441
+ A special value of
2442
+ .B 0
2443
+ indicates the default setting
2444
+ which for single-threaded mode is the same as no limit.
2445
+ .IP 4. 4
2446
+ Since
2447
+ .B xz
2448
+ 5.3.4alpha:
2449
+ Memory usage for multi-threaded decompression in bytes
2450
+ .RB ( \-\-memlimit\-mt\-decompress ).
2451
+ This is never zero because a system-specific default value
2452
+ shown in the column 5
2453
+ is used if no limit has been specified explicitly.
2454
+ This is also never greater than the value in the column 3
2455
+ even if a larger value has been specified with
2456
+ .BR \-\-memlimit\-mt\-decompress .
2457
+ .IP 5. 4
2458
+ Since
2459
+ .B xz
2460
+ 5.3.4alpha:
2461
+ A system-specific default memory usage limit
2462
+ that is used to limit the number of threads
2463
+ when compressing with an automatic number of threads
2464
+ .RB ( \-\-threads=0 )
2465
+ and no memory usage limit has been specified
2466
+ .RB ( \-\-memlimit\-compress ).
2467
+ This is also used as the default value for
2468
+ .BR \-\-memlimit\-mt\-decompress .
2469
+ .IP 6. 4
2470
+ Since
2471
+ .B xz
2472
+ 5.3.4alpha:
2473
+ Number of available processor threads.
2474
+ .PP
2475
+ In the future, the output of
2476
+ .B "xz \-\-robot \-\-info\-memory"
2477
+ may have more columns, but never more than a single line.
2478
+ .
2479
+ .SS Version
2480
+ .B "xz \-\-robot \-\-version"
2481
+ prints the version number of
2482
+ .B xz
2483
+ and liblzma in the following format:
2484
+ .PP
2485
+ .BI XZ_VERSION= XYYYZZZS
2486
+ .br
2487
+ .BI LIBLZMA_VERSION= XYYYZZZS
2488
+ .TP
2489
+ .I X
2490
+ Major version.
2491
+ .TP
2492
+ .I YYY
2493
+ Minor version.
2494
+ Even numbers are stable.
2495
+ Odd numbers are alpha or beta versions.
2496
+ .TP
2497
+ .I ZZZ
2498
+ Patch level for stable releases or
2499
+ just a counter for development releases.
2500
+ .TP
2501
+ .I S
2502
+ Stability.
2503
+ 0 is alpha, 1 is beta, and 2 is stable.
2504
+ .I S
2505
+ should be always 2 when
2506
+ .I YYY
2507
+ is even.
2508
+ .PP
2509
+ .I XYYYZZZS
2510
+ are the same on both lines if
2511
+ .B xz
2512
+ and liblzma are from the same XZ Utils release.
2513
+ .PP
2514
+ Examples: 4.999.9beta is
2515
+ .B 49990091
2516
+ and
2517
+ 5.0.0 is
2518
+ .BR 50000002 .
2519
+ .
2520
+ .SH "EXIT STATUS"
2521
+ .TP
2522
+ .B 0
2523
+ All is good.
2524
+ .TP
2525
+ .B 1
2526
+ An error occurred.
2527
+ .TP
2528
+ .B 2
2529
+ Something worth a warning occurred,
2530
+ but no actual errors occurred.
2531
+ .PP
2532
+ Notices (not warnings or errors) printed on standard error
2533
+ don't affect the exit status.
2534
+ .
2535
+ .SH ENVIRONMENT
2536
+ .B xz
2537
+ parses space-separated lists of options
2538
+ from the environment variables
2539
+ .B XZ_DEFAULTS
2540
+ and
2541
+ .BR XZ_OPT ,
2542
+ in this order, before parsing the options from the command line.
2543
+ Note that only options are parsed from the environment variables;
2544
+ all non-options are silently ignored.
2545
+ Parsing is done with
2546
+ .BR getopt_long (3)
2547
+ which is used also for the command line arguments.
2548
+ .TP
2549
+ .B XZ_DEFAULTS
2550
+ User-specific or system-wide default options.
2551
+ Typically this is set in a shell initialization script to enable
2552
+ .BR xz 's
2553
+ memory usage limiter by default.
2554
+ Excluding shell initialization scripts
2555
+ and similar special cases, scripts must never set or unset
2556
+ .BR XZ_DEFAULTS .
2557
+ .TP
2558
+ .B XZ_OPT
2559
+ This is for passing options to
2560
+ .B xz
2561
+ when it is not possible to set the options directly on the
2562
+ .B xz
2563
+ command line.
2564
+ This is the case when
2565
+ .B xz
2566
+ is run by a script or tool, for example, GNU
2567
+ .BR tar (1):
2568
+ .RS
2569
+ .RS
2570
+ .PP
2571
+ .nf
2572
+ .ft CR
2573
+ XZ_OPT=\-2v tar caf foo.tar.xz foo
2574
+ .ft R
2575
+ .fi
2576
+ .RE
2577
+ .RE
2578
+ .IP ""
2579
+ Scripts may use
2580
+ .BR XZ_OPT ,
2581
+ for example, to set script-specific default compression options.
2582
+ It is still recommended to allow users to override
2583
+ .B XZ_OPT
2584
+ if that is reasonable.
2585
+ For example, in
2586
+ .BR sh (1)
2587
+ scripts one may use something like this:
2588
+ .RS
2589
+ .RS
2590
+ .PP
2591
+ .nf
2592
+ .ft CR
2593
+ XZ_OPT=${XZ_OPT\-"\-7e"}
2594
+ export XZ_OPT
2595
+ .ft R
2596
+ .fi
2597
+ .RE
2598
+ .RE
2599
+ .
2600
+ .SH "LZMA UTILS COMPATIBILITY"
2601
+ The command line syntax of
2602
+ .B xz
2603
+ is practically a superset of
2604
+ .BR lzma ,
2605
+ .BR unlzma ,
2606
+ and
2607
+ .B lzcat
2608
+ as found from LZMA Utils 4.32.x.
2609
+ In most cases, it is possible to replace
2610
+ LZMA Utils with XZ Utils without breaking existing scripts.
2611
+ There are some incompatibilities though,
2612
+ which may sometimes cause problems.
2613
+ .
2614
+ .SS "Compression preset levels"
2615
+ The numbering of the compression level presets is not identical in
2616
+ .B xz
2617
+ and LZMA Utils.
2618
+ The most important difference is how dictionary sizes
2619
+ are mapped to different presets.
2620
+ Dictionary size is roughly equal to the decompressor memory usage.
2621
+ .RS
2622
+ .PP
2623
+ .TS
2624
+ tab(;);
2625
+ c c c
2626
+ c n n.
2627
+ Level;xz;LZMA Utils
2628
+ \-0;256 KiB;N/A
2629
+ \-1;1 MiB;64 KiB
2630
+ \-2;2 MiB;1 MiB
2631
+ \-3;4 MiB;512 KiB
2632
+ \-4;4 MiB;1 MiB
2633
+ \-5;8 MiB;2 MiB
2634
+ \-6;8 MiB;4 MiB
2635
+ \-7;16 MiB;8 MiB
2636
+ \-8;32 MiB;16 MiB
2637
+ \-9;64 MiB;32 MiB
2638
+ .TE
2639
+ .RE
2640
+ .PP
2641
+ The dictionary size differences affect
2642
+ the compressor memory usage too,
2643
+ but there are some other differences between
2644
+ LZMA Utils and XZ Utils, which
2645
+ make the difference even bigger:
2646
+ .RS
2647
+ .PP
2648
+ .TS
2649
+ tab(;);
2650
+ c c c
2651
+ c n n.
2652
+ Level;xz;LZMA Utils 4.32.x
2653
+ \-0;3 MiB;N/A
2654
+ \-1;9 MiB;2 MiB
2655
+ \-2;17 MiB;12 MiB
2656
+ \-3;32 MiB;12 MiB
2657
+ \-4;48 MiB;16 MiB
2658
+ \-5;94 MiB;26 MiB
2659
+ \-6;94 MiB;45 MiB
2660
+ \-7;186 MiB;83 MiB
2661
+ \-8;370 MiB;159 MiB
2662
+ \-9;674 MiB;311 MiB
2663
+ .TE
2664
+ .RE
2665
+ .PP
2666
+ The default preset level in LZMA Utils is
2667
+ .B \-7
2668
+ while in XZ Utils it is
2669
+ .BR \-6 ,
2670
+ so both use an 8 MiB dictionary by default.
2671
+ .
2672
+ .SS "Streamed vs. non-streamed .lzma files"
2673
+ The uncompressed size of the file can be stored in the
2674
+ .B .lzma
2675
+ header.
2676
+ LZMA Utils does that when compressing regular files.
2677
+ The alternative is to mark that uncompressed size is unknown
2678
+ and use end-of-payload marker to indicate
2679
+ where the decompressor should stop.
2680
+ LZMA Utils uses this method when uncompressed size isn't known,
2681
+ which is the case, for example, in pipes.
2682
+ .PP
2683
+ .B xz
2684
+ supports decompressing
2685
+ .B .lzma
2686
+ files with or without end-of-payload marker, but all
2687
+ .B .lzma
2688
+ files created by
2689
+ .B xz
2690
+ will use end-of-payload marker and have uncompressed size
2691
+ marked as unknown in the
2692
+ .B .lzma
2693
+ header.
2694
+ This may be a problem in some uncommon situations.
2695
+ For example, a
2696
+ .B .lzma
2697
+ decompressor in an embedded device might work
2698
+ only with files that have known uncompressed size.
2699
+ If you hit this problem, you need to use LZMA Utils
2700
+ or LZMA SDK to create
2701
+ .B .lzma
2702
+ files with known uncompressed size.
2703
+ .
2704
+ .SS "Unsupported .lzma files"
2705
+ The
2706
+ .B .lzma
2707
+ format allows
2708
+ .I lc
2709
+ values up to 8, and
2710
+ .I lp
2711
+ values up to 4.
2712
+ LZMA Utils can decompress files with any
2713
+ .I lc
2714
+ and
2715
+ .IR lp ,
2716
+ but always creates files with
2717
+ .B lc=3
2718
+ and
2719
+ .BR lp=0 .
2720
+ Creating files with other
2721
+ .I lc
2722
+ and
2723
+ .I lp
2724
+ is possible with
2725
+ .B xz
2726
+ and with LZMA SDK.
2727
+ .PP
2728
+ The implementation of the LZMA1 filter in liblzma
2729
+ requires that the sum of
2730
+ .I lc
2731
+ and
2732
+ .I lp
2733
+ must not exceed 4.
2734
+ Thus,
2735
+ .B .lzma
2736
+ files, which exceed this limitation, cannot be decompressed with
2737
+ .BR xz .
2738
+ .PP
2739
+ LZMA Utils creates only
2740
+ .B .lzma
2741
+ files which have a dictionary size of
2742
+ .RI "2^" n
2743
+ (a power of 2) but accepts files with any dictionary size.
2744
+ liblzma accepts only
2745
+ .B .lzma
2746
+ files which have a dictionary size of
2747
+ .RI "2^" n
2748
+ or
2749
+ .RI "2^" n " + 2^(" n "\-1)."
2750
+ This is to decrease false positives when detecting
2751
+ .B .lzma
2752
+ files.
2753
+ .PP
2754
+ These limitations shouldn't be a problem in practice,
2755
+ since practically all
2756
+ .B .lzma
2757
+ files have been compressed with settings that liblzma will accept.
2758
+ .
2759
+ .SS "Trailing garbage"
2760
+ When decompressing,
2761
+ LZMA Utils silently ignore everything after the first
2762
+ .B .lzma
2763
+ stream.
2764
+ In most situations, this is a bug.
2765
+ This also means that LZMA Utils
2766
+ don't support decompressing concatenated
2767
+ .B .lzma
2768
+ files.
2769
+ .PP
2770
+ If there is data left after the first
2771
+ .B .lzma
2772
+ stream,
2773
+ .B xz
2774
+ considers the file to be corrupt unless
2775
+ .B \-\-single\-stream
2776
+ was used.
2777
+ This may break obscure scripts which have
2778
+ assumed that trailing garbage is ignored.
2779
+ .
2780
+ .SH NOTES
2781
+ .
2782
+ .SS "Compressed output may vary"
2783
+ The exact compressed output produced from
2784
+ the same uncompressed input file
2785
+ may vary between XZ Utils versions even if
2786
+ compression options are identical.
2787
+ This is because the encoder can be improved
2788
+ (faster or better compression)
2789
+ without affecting the file format.
2790
+ The output can vary even between different
2791
+ builds of the same XZ Utils version,
2792
+ if different build options are used.
2793
+ .PP
2794
+ The above means that once
2795
+ .B \-\-rsyncable
2796
+ has been implemented,
2797
+ the resulting files won't necessarily be rsyncable
2798
+ unless both old and new files have been compressed
2799
+ with the same xz version.
2800
+ This problem can be fixed if a part of the encoder
2801
+ implementation is frozen to keep rsyncable output
2802
+ stable across xz versions.
2803
+ .
2804
+ .SS "Embedded .xz decompressors"
2805
+ Embedded
2806
+ .B .xz
2807
+ decompressor implementations like XZ Embedded don't necessarily
2808
+ support files created with integrity
2809
+ .I check
2810
+ types other than
2811
+ .B none
2812
+ and
2813
+ .BR crc32 .
2814
+ Since the default is
2815
+ .BR \-\-check=crc64 ,
2816
+ you must use
2817
+ .B \-\-check=none
2818
+ or
2819
+ .B \-\-check=crc32
2820
+ when creating files for embedded systems.
2821
+ .PP
2822
+ Outside embedded systems, all
2823
+ .B .xz
2824
+ format decompressors support all the
2825
+ .I check
2826
+ types, or at least are able to decompress
2827
+ the file without verifying the
2828
+ integrity check if the particular
2829
+ .I check
2830
+ is not supported.
2831
+ .PP
2832
+ XZ Embedded supports BCJ filters,
2833
+ but only with the default start offset.
2834
+ .
2835
+ .SH EXAMPLES
2836
+ .
2837
+ .SS Basics
2838
+ Compress the file
2839
+ .I foo
2840
+ into
2841
+ .I foo.xz
2842
+ using the default compression level
2843
+ .RB ( \-6 ),
2844
+ and remove
2845
+ .I foo
2846
+ if compression is successful:
2847
+ .RS
2848
+ .PP
2849
+ .nf
2850
+ .ft CR
2851
+ xz foo
2852
+ .ft R
2853
+ .fi
2854
+ .RE
2855
+ .PP
2856
+ Decompress
2857
+ .I bar.xz
2858
+ into
2859
+ .I bar
2860
+ and don't remove
2861
+ .I bar.xz
2862
+ even if decompression is successful:
2863
+ .RS
2864
+ .PP
2865
+ .nf
2866
+ .ft CR
2867
+ xz \-dk bar.xz
2868
+ .ft R
2869
+ .fi
2870
+ .RE
2871
+ .PP
2872
+ Create
2873
+ .I baz.tar.xz
2874
+ with the preset
2875
+ .B \-4e
2876
+ .RB ( "\-4 \-\-extreme" ),
2877
+ which is slower than the default
2878
+ .BR \-6 ,
2879
+ but needs less memory for compression and decompression (48\ MiB
2880
+ and 5\ MiB, respectively):
2881
+ .RS
2882
+ .PP
2883
+ .nf
2884
+ .ft CR
2885
+ tar cf \- baz | xz \-4e > baz.tar.xz
2886
+ .ft R
2887
+ .fi
2888
+ .RE
2889
+ .PP
2890
+ A mix of compressed and uncompressed files can be decompressed
2891
+ to standard output with a single command:
2892
+ .RS
2893
+ .PP
2894
+ .nf
2895
+ .ft CR
2896
+ xz \-dcf a.txt b.txt.xz c.txt d.txt.lzma > abcd.txt
2897
+ .ft R
2898
+ .fi
2899
+ .RE
2900
+ .
2901
+ .SS "Parallel compression of many files"
2902
+ On GNU and *BSD,
2903
+ .BR find (1)
2904
+ and
2905
+ .BR xargs (1)
2906
+ can be used to parallelize compression of many files:
2907
+ .RS
2908
+ .PP
2909
+ .nf
2910
+ .ft CR
2911
+ find . \-type f \e! \-name '*.xz' \-print0 \e
2912
+ | xargs \-0r \-P4 \-n16 xz \-T1
2913
+ .ft R
2914
+ .fi
2915
+ .RE
2916
+ .PP
2917
+ The
2918
+ .B \-P
2919
+ option to
2920
+ .BR xargs (1)
2921
+ sets the number of parallel
2922
+ .B xz
2923
+ processes.
2924
+ The best value for the
2925
+ .B \-n
2926
+ option depends on how many files there are to be compressed.
2927
+ If there are only a couple of files,
2928
+ the value should probably be 1;
2929
+ with tens of thousands of files,
2930
+ 100 or even more may be appropriate to reduce the number of
2931
+ .B xz
2932
+ processes that
2933
+ .BR xargs (1)
2934
+ will eventually create.
2935
+ .PP
2936
+ The option
2937
+ .B \-T1
2938
+ for
2939
+ .B xz
2940
+ is there to force it to single-threaded mode, because
2941
+ .BR xargs (1)
2942
+ is used to control the amount of parallelization.
2943
+ .
2944
+ .SS "Robot mode"
2945
+ Calculate how many bytes have been saved in total
2946
+ after compressing multiple files:
2947
+ .RS
2948
+ .PP
2949
+ .nf
2950
+ .ft CR
2951
+ xz \-\-robot \-\-list *.xz | awk '/^totals/{print $5\-$4}'
2952
+ .ft R
2953
+ .fi
2954
+ .RE
2955
+ .PP
2956
+ A script may want to know that it is using new enough
2957
+ .BR xz .
2958
+ The following
2959
+ .BR sh (1)
2960
+ script checks that the version number of the
2961
+ .B xz
2962
+ tool is at least 5.0.0.
2963
+ This method is compatible with old beta versions,
2964
+ which didn't support the
2965
+ .B \-\-robot
2966
+ option:
2967
+ .RS
2968
+ .PP
2969
+ .nf
2970
+ .ft CR
2971
+ if ! eval "$(xz \-\-robot \-\-version 2> /dev/null)" ||
2972
+ [ "$XZ_VERSION" \-lt 50000002 ]; then
2973
+ echo "Your xz is too old."
2974
+ fi
2975
+ unset XZ_VERSION LIBLZMA_VERSION
2976
+ .ft R
2977
+ .fi
2978
+ .RE
2979
+ .PP
2980
+ Set a memory usage limit for decompression using
2981
+ .BR XZ_OPT ,
2982
+ but if a limit has already been set, don't increase it:
2983
+ .RS
2984
+ .PP
2985
+ .nf
2986
+ .ft CR
2987
+ NEWLIM=$((123 << 20))\ \ # 123 MiB
2988
+ OLDLIM=$(xz \-\-robot \-\-info\-memory | cut \-f3)
2989
+ if [ $OLDLIM \-eq 0 \-o $OLDLIM \-gt $NEWLIM ]; then
2990
+ XZ_OPT="$XZ_OPT \-\-memlimit\-decompress=$NEWLIM"
2991
+ export XZ_OPT
2992
+ fi
2993
+ .ft R
2994
+ .fi
2995
+ .RE
2996
+ .
2997
+ .SS "Custom compressor filter chains"
2998
+ The simplest use for custom filter chains is
2999
+ customizing a LZMA2 preset.
3000
+ This can be useful,
3001
+ because the presets cover only a subset of the
3002
+ potentially useful combinations of compression settings.
3003
+ .PP
3004
+ The CompCPU columns of the tables
3005
+ from the descriptions of the options
3006
+ .BR "\-0" " ... " "\-9"
3007
+ and
3008
+ .B \-\-extreme
3009
+ are useful when customizing LZMA2 presets.
3010
+ Here are the relevant parts collected from those two tables:
3011
+ .RS
3012
+ .PP
3013
+ .TS
3014
+ tab(;);
3015
+ c c
3016
+ n n.
3017
+ Preset;CompCPU
3018
+ \-0;0
3019
+ \-1;1
3020
+ \-2;2
3021
+ \-3;3
3022
+ \-4;4
3023
+ \-5;5
3024
+ \-6;6
3025
+ \-5e;7
3026
+ \-6e;8
3027
+ .TE
3028
+ .RE
3029
+ .PP
3030
+ If you know that a file requires
3031
+ somewhat big dictionary (for example, 32\ MiB) to compress well,
3032
+ but you want to compress it quicker than
3033
+ .B "xz \-8"
3034
+ would do, a preset with a low CompCPU value (for example, 1)
3035
+ can be modified to use a bigger dictionary:
3036
+ .RS
3037
+ .PP
3038
+ .nf
3039
+ .ft CR
3040
+ xz \-\-lzma2=preset=1,dict=32MiB foo.tar
3041
+ .ft R
3042
+ .fi
3043
+ .RE
3044
+ .PP
3045
+ With certain files, the above command may be faster than
3046
+ .B "xz \-6"
3047
+ while compressing significantly better.
3048
+ However, it must be emphasized that only some files benefit from
3049
+ a big dictionary while keeping the CompCPU value low.
3050
+ The most obvious situation,
3051
+ where a big dictionary can help a lot,
3052
+ is an archive containing very similar files
3053
+ of at least a few megabytes each.
3054
+ The dictionary size has to be significantly bigger
3055
+ than any individual file to allow LZMA2 to take
3056
+ full advantage of the similarities between consecutive files.
3057
+ .PP
3058
+ If very high compressor and decompressor memory usage is fine,
3059
+ and the file being compressed is
3060
+ at least several hundred megabytes, it may be useful
3061
+ to use an even bigger dictionary than the 64 MiB that
3062
+ .B "xz \-9"
3063
+ would use:
3064
+ .RS
3065
+ .PP
3066
+ .nf
3067
+ .ft CR
3068
+ xz \-vv \-\-lzma2=dict=192MiB big_foo.tar
3069
+ .ft R
3070
+ .fi
3071
+ .RE
3072
+ .PP
3073
+ Using
3074
+ .B \-vv
3075
+ .RB ( "\-\-verbose \-\-verbose" )
3076
+ like in the above example can be useful
3077
+ to see the memory requirements
3078
+ of the compressor and decompressor.
3079
+ Remember that using a dictionary bigger than
3080
+ the size of the uncompressed file is waste of memory,
3081
+ so the above command isn't useful for small files.
3082
+ .PP
3083
+ Sometimes the compression time doesn't matter,
3084
+ but the decompressor memory usage has to be kept low, for example,
3085
+ to make it possible to decompress the file on an embedded system.
3086
+ The following command uses
3087
+ .B \-6e
3088
+ .RB ( "\-6 \-\-extreme" )
3089
+ as a base and sets the dictionary to only 64\ KiB.
3090
+ The resulting file can be decompressed with XZ Embedded
3091
+ (that's why there is
3092
+ .BR \-\-check=crc32 )
3093
+ using about 100\ KiB of memory.
3094
+ .RS
3095
+ .PP
3096
+ .nf
3097
+ .ft CR
3098
+ xz \-\-check=crc32 \-\-lzma2=preset=6e,dict=64KiB foo
3099
+ .ft R
3100
+ .fi
3101
+ .RE
3102
+ .PP
3103
+ If you want to squeeze out as many bytes as possible,
3104
+ adjusting the number of literal context bits
3105
+ .RI ( lc )
3106
+ and number of position bits
3107
+ .RI ( pb )
3108
+ can sometimes help.
3109
+ Adjusting the number of literal position bits
3110
+ .RI ( lp )
3111
+ might help too, but usually
3112
+ .I lc
3113
+ and
3114
+ .I pb
3115
+ are more important.
3116
+ For example, a source code archive contains mostly US-ASCII text,
3117
+ so something like the following might give
3118
+ slightly (like 0.1\ %) smaller file than
3119
+ .B "xz \-6e"
3120
+ (try also without
3121
+ .BR lc=4 ):
3122
+ .RS
3123
+ .PP
3124
+ .nf
3125
+ .ft CR
3126
+ xz \-\-lzma2=preset=6e,pb=0,lc=4 source_code.tar
3127
+ .ft R
3128
+ .fi
3129
+ .RE
3130
+ .PP
3131
+ Using another filter together with LZMA2 can improve
3132
+ compression with certain file types.
3133
+ For example, to compress a x86-32 or x86-64 shared library
3134
+ using the x86 BCJ filter:
3135
+ .RS
3136
+ .PP
3137
+ .nf
3138
+ .ft CR
3139
+ xz \-\-x86 \-\-lzma2 libfoo.so
3140
+ .ft R
3141
+ .fi
3142
+ .RE
3143
+ .PP
3144
+ Note that the order of the filter options is significant.
3145
+ If
3146
+ .B \-\-x86
3147
+ is specified after
3148
+ .BR \-\-lzma2 ,
3149
+ .B xz
3150
+ will give an error,
3151
+ because there cannot be any filter after LZMA2,
3152
+ and also because the x86 BCJ filter cannot be used
3153
+ as the last filter in the chain.
3154
+ .PP
3155
+ The Delta filter together with LZMA2
3156
+ can give good results with bitmap images.
3157
+ It should usually beat PNG,
3158
+ which has a few more advanced filters than simple
3159
+ delta but uses Deflate for the actual compression.
3160
+ .PP
3161
+ The image has to be saved in uncompressed format,
3162
+ for example, as uncompressed TIFF.
3163
+ The distance parameter of the Delta filter is set
3164
+ to match the number of bytes per pixel in the image.
3165
+ For example, 24-bit RGB bitmap needs
3166
+ .BR dist=3 ,
3167
+ and it is also good to pass
3168
+ .B pb=0
3169
+ to LZMA2 to accommodate the three-byte alignment:
3170
+ .RS
3171
+ .PP
3172
+ .nf
3173
+ .ft CR
3174
+ xz \-\-delta=dist=3 \-\-lzma2=pb=0 foo.tiff
3175
+ .ft R
3176
+ .fi
3177
+ .RE
3178
+ .PP
3179
+ If multiple images have been put into a single archive (for example,
3180
+ .BR .tar ),
3181
+ the Delta filter will work on that too as long as all images
3182
+ have the same number of bytes per pixel.
3183
+ .
3184
+ .SH "SEE ALSO"
3185
+ .BR xzdec (1),
3186
+ .BR xzdiff (1),
3187
+ .BR xzgrep (1),
3188
+ .BR xzless (1),
3189
+ .BR xzmore (1),
3190
+ .BR gzip (1),
3191
+ .BR bzip2 (1),
3192
+ .BR 7z (1)
3193
+ .PP
3194
+ XZ Utils: <https://tukaani.org/xz/>
3195
+ .br
3196
+ XZ Embedded: <https://tukaani.org/xz/embedded.html>
3197
+ .br
3198
+ LZMA SDK: <https://7-zip.org/sdk.html>
janus/share/man/man1/lzmore.1 ADDED
@@ -0,0 +1,65 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ .\" SPDX-License-Identifier: 0BSD
2
+ .\"
3
+ .\" Authors: Andrew Dudman
4
+ .\" Lasse Collin
5
+ .\"
6
+ .\" (Note that this file is based on xzless.1 instead of gzip's zmore.1.)
7
+ .\"
8
+ .TH XZMORE 1 "2024-02-12" "Tukaani" "XZ Utils"
9
+ .SH NAME
10
+ xzmore, lzmore \- view xz or lzma compressed (text) files
11
+ .
12
+ .SH SYNOPSIS
13
+ .B xzmore
14
+ .RI [ file ...]
15
+ .br
16
+ .B lzmore
17
+ .RI [ file ...]
18
+ .
19
+ .SH DESCRIPTION
20
+ .B xzmore
21
+ displays text from compressed files to a terminal using
22
+ .BR more (1).
23
+ Files supported by
24
+ .BR xz (1)
25
+ are decompressed;
26
+ other files are assumed to be in uncompressed form already.
27
+ If no
28
+ .I files
29
+ are given,
30
+ .B xzmore
31
+ reads from standard input.
32
+ See the
33
+ .BR more (1)
34
+ manual for the keyboard commands.
35
+ .PP
36
+ Note that scrolling backwards might not be possible
37
+ depending on the implementation of
38
+ .BR more (1).
39
+ This is because
40
+ .B xzmore
41
+ uses a pipe to pass the decompressed data to
42
+ .BR more (1).
43
+ .BR xzless (1)
44
+ uses
45
+ .BR less (1)
46
+ which provides more advanced features.
47
+ .PP
48
+ The command
49
+ .B lzmore
50
+ is provided for backward compatibility with LZMA Utils.
51
+ .
52
+ .SH ENVIRONMENT
53
+ .TP
54
+ .B PAGER
55
+ If
56
+ .B PAGER
57
+ is set,
58
+ its value is used as the pager instead of
59
+ .BR more (1).
60
+ .
61
+ .SH "SEE ALSO"
62
+ .BR more (1),
63
+ .BR xz (1),
64
+ .BR xzless (1),
65
+ .BR zmore (1)
janus/share/man/man1/sqlite3.1 ADDED
@@ -0,0 +1,161 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ .\" Hey, EMACS: -*- nroff -*-
2
+ .\" First parameter, NAME, should be all caps
3
+ .\" Second parameter, SECTION, should be 1-8, maybe w/ subsection
4
+ .\" other parameters are allowed: see man(7), man(1)
5
+ .TH SQLITE3 1 "Fri Aug 11 23:50:12 CET 2023"
6
+ .\" Please adjust this date whenever revising the manpage.
7
+ .\"
8
+ .\" Some roff macros, for reference:
9
+ .\" .nh disable hyphenation
10
+ .\" .hy enable hyphenation
11
+ .\" .ad l left justify
12
+ .\" .ad b justify to both left and right margins
13
+ .\" .nf disable filling
14
+ .\" .fi enable filling
15
+ .\" .br insert line break
16
+ .\" .sp <n> insert n+1 empty lines
17
+ .\" for manpage-specific macros, see man(7)
18
+ .SH NAME
19
+ .B sqlite3
20
+ \- A command line interface for SQLite version 3
21
+
22
+ .SH SYNOPSIS
23
+ .B sqlite3
24
+ .RI [ options ]
25
+ .RI [ databasefile ]
26
+ .RI [ SQL ]
27
+
28
+ .SH SUMMARY
29
+ .PP
30
+ .B sqlite3
31
+ is a terminal-based front-end to the SQLite library that can evaluate
32
+ queries interactively and display the results in multiple formats.
33
+ .B sqlite3
34
+ can also be used within shell scripts and other applications to provide
35
+ batch processing features.
36
+
37
+ .SH DESCRIPTION
38
+ To start a
39
+ .B sqlite3
40
+ interactive session, invoke the
41
+ .B sqlite3
42
+ command and optionally provide the name of a database file. If the
43
+ database file does not exist, it will be created. If the database file
44
+ does exist, it will be opened.
45
+
46
+ For example, to create a new database file named "mydata.db", create
47
+ a table named "memos" and insert a couple of records into that table:
48
+ .sp
49
+ $
50
+ .B sqlite3 mydata.db
51
+ .br
52
+ SQLite version 3.43.0 2023-08-11 17:45:23
53
+ .br
54
+ Enter ".help" for usage hints.
55
+ .br
56
+ sqlite>
57
+ .B create table memos(text, priority INTEGER);
58
+ .br
59
+ sqlite>
60
+ .B insert into memos values('deliver project description', 10);
61
+ .br
62
+ sqlite>
63
+ .B insert into memos values('lunch with Christine', 100);
64
+ .br
65
+ sqlite>
66
+ .B select * from memos;
67
+ .br
68
+ deliver project description|10
69
+ .br
70
+ lunch with Christine|100
71
+ .br
72
+ sqlite>
73
+ .sp
74
+
75
+ If no database name is supplied, the ATTACH sql command can be used
76
+ to attach to existing or create new database files. ATTACH can also
77
+ be used to attach to multiple databases within the same interactive
78
+ session. This is useful for migrating data between databases,
79
+ possibly changing the schema along the way.
80
+
81
+ Optionally, a SQL statement or set of SQL statements can be supplied as
82
+ a single argument. Multiple statements should be separated by
83
+ semi-colons.
84
+
85
+ For example:
86
+ .sp
87
+ $
88
+ .B sqlite3 -line mydata.db 'select * from memos where priority > 20;'
89
+ .br
90
+ text = lunch with Christine
91
+ .br
92
+ priority = 100
93
+ .br
94
+ .sp
95
+
96
+ .SS SQLITE META-COMMANDS
97
+ .PP
98
+ The interactive interpreter offers a set of meta-commands that can be
99
+ used to control the output format, examine the currently attached
100
+ database files, or perform administrative operations upon the
101
+ attached databases (such as rebuilding indices). Meta-commands are
102
+ always prefixed with a dot (.).
103
+
104
+ A list of available meta-commands can be viewed at any time by issuing
105
+ the '.help' command. For example:
106
+ .sp
107
+ sqlite>
108
+ .B .help
109
+ .nf
110
+ .tr %.
111
+ ...
112
+ .sp
113
+ .fi
114
+
115
+ The available commands differ by version and build options, so they
116
+ are not listed here. Please refer to your local copy for all available
117
+ options.
118
+
119
+
120
+ .SH INIT FILE
121
+ .B sqlite3
122
+ reads an initialization file to set the configuration of the
123
+ interactive environment. Throughout initialization, any previously
124
+ specified setting can be overridden. The sequence of initialization is
125
+ as follows:
126
+
127
+ o The default configuration is established as follows:
128
+
129
+ .sp
130
+ .nf
131
+ .cc |
132
+ mode = LIST
133
+ separator = "|"
134
+ main prompt = "sqlite> "
135
+ continue prompt = " ...> "
136
+ |cc .
137
+ .sp
138
+ .fi
139
+
140
+ o If the file
141
+ .B ${XDG_CONFIG_HOME}/sqlite3/sqliterc
142
+ or
143
+ .B ~/.sqliterc
144
+ exists, the first of those to be found is processed during startup.
145
+ It should generally only contain meta-commands.
146
+
147
+ o If the -init option is present, the specified file is processed.
148
+
149
+ o All other command line options are processed.
150
+
151
+ .SH SEE ALSO
152
+ https://sqlite.org/cli.html
153
+ .br
154
+ https://sqlite.org/fiddle (a WebAssembly build of the CLI app)
155
+ .br
156
+ The sqlite3-doc package.
157
+ .SH AUTHOR
158
+ This manual page was originally written by Andreas Rottmann
159
+ <rotty@debian.org>, for the Debian GNU/Linux system (but may be used
160
+ by others). It was subsequently revised by Bill Bumgarner <bbum@mac.com>,
161
+ Laszlo Boszormenyi <gcs@debian.hu>, and the sqlite3 developers.
janus/share/man/man1/unxz.1 ADDED
@@ -0,0 +1,3198 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ '\" t
2
+ .\" SPDX-License-Identifier: 0BSD
3
+ .\"
4
+ .\" Authors: Lasse Collin
5
+ .\" Jia Tan
6
+ .\"
7
+ .TH XZ 1 "2024-12-30" "Tukaani" "XZ Utils"
8
+ .
9
+ .SH NAME
10
+ xz, unxz, xzcat, lzma, unlzma, lzcat \- Compress or decompress .xz and .lzma files
11
+ .
12
+ .SH SYNOPSIS
13
+ .B xz
14
+ .RI [ option... ]
15
+ .RI [ file... ]
16
+ .
17
+ .SH COMMAND ALIASES
18
+ .B unxz
19
+ is equivalent to
20
+ .BR "xz \-\-decompress" .
21
+ .br
22
+ .B xzcat
23
+ is equivalent to
24
+ .BR "xz \-\-decompress \-\-stdout" .
25
+ .br
26
+ .B lzma
27
+ is equivalent to
28
+ .BR "xz \-\-format=lzma" .
29
+ .br
30
+ .B unlzma
31
+ is equivalent to
32
+ .BR "xz \-\-format=lzma \-\-decompress" .
33
+ .br
34
+ .B lzcat
35
+ is equivalent to
36
+ .BR "xz \-\-format=lzma \-\-decompress \-\-stdout" .
37
+ .PP
38
+ When writing scripts that need to decompress files,
39
+ it is recommended to always use the name
40
+ .B xz
41
+ with appropriate arguments
42
+ .RB ( "xz \-d"
43
+ or
44
+ .BR "xz \-dc" )
45
+ instead of the names
46
+ .B unxz
47
+ and
48
+ .BR xzcat .
49
+ .
50
+ .SH DESCRIPTION
51
+ .B xz
52
+ is a general-purpose data compression tool with
53
+ command line syntax similar to
54
+ .BR gzip (1)
55
+ and
56
+ .BR bzip2 (1).
57
+ The native file format is the
58
+ .B .xz
59
+ format, but the legacy
60
+ .B .lzma
61
+ format used by LZMA Utils and
62
+ raw compressed streams with no container format headers
63
+ are also supported.
64
+ In addition, decompression of the
65
+ .B .lz
66
+ format used by
67
+ .B lzip
68
+ is supported.
69
+ .PP
70
+ .B xz
71
+ compresses or decompresses each
72
+ .I file
73
+ according to the selected operation mode.
74
+ If no
75
+ .I files
76
+ are given or
77
+ .I file
78
+ is
79
+ .BR \- ,
80
+ .B xz
81
+ reads from standard input and writes the processed data
82
+ to standard output.
83
+ .B xz
84
+ will refuse (display an error and skip the
85
+ .IR file )
86
+ to write compressed data to standard output if it is a terminal.
87
+ Similarly,
88
+ .B xz
89
+ will refuse to read compressed data
90
+ from standard input if it is a terminal.
91
+ .PP
92
+ Unless
93
+ .B \-\-stdout
94
+ is specified,
95
+ .I files
96
+ other than
97
+ .B \-
98
+ are written to a new file whose name is derived from the source
99
+ .I file
100
+ name:
101
+ .IP \(bu 3
102
+ When compressing, the suffix of the target file format
103
+ .RB ( .xz
104
+ or
105
+ .BR .lzma )
106
+ is appended to the source filename to get the target filename.
107
+ .IP \(bu 3
108
+ When decompressing, the
109
+ .BR .xz ,
110
+ .BR .lzma ,
111
+ or
112
+ .B .lz
113
+ suffix is removed from the filename to get the target filename.
114
+ .B xz
115
+ also recognizes the suffixes
116
+ .B .txz
117
+ and
118
+ .BR .tlz ,
119
+ and replaces them with the
120
+ .B .tar
121
+ suffix.
122
+ .PP
123
+ If the target file already exists, an error is displayed and the
124
+ .I file
125
+ is skipped.
126
+ .PP
127
+ Unless writing to standard output,
128
+ .B xz
129
+ will display a warning and skip the
130
+ .I file
131
+ if any of the following applies:
132
+ .IP \(bu 3
133
+ .I File
134
+ is not a regular file.
135
+ Symbolic links are not followed,
136
+ and thus they are not considered to be regular files.
137
+ .IP \(bu 3
138
+ .I File
139
+ has more than one hard link.
140
+ .IP \(bu 3
141
+ .I File
142
+ has setuid, setgid, or sticky bit set.
143
+ .IP \(bu 3
144
+ The operation mode is set to compress and the
145
+ .I file
146
+ already has a suffix of the target file format
147
+ .RB ( .xz
148
+ or
149
+ .B .txz
150
+ when compressing to the
151
+ .B .xz
152
+ format, and
153
+ .B .lzma
154
+ or
155
+ .B .tlz
156
+ when compressing to the
157
+ .B .lzma
158
+ format).
159
+ .IP \(bu 3
160
+ The operation mode is set to decompress and the
161
+ .I file
162
+ doesn't have a suffix of any of the supported file formats
163
+ .RB ( .xz ,
164
+ .BR .txz ,
165
+ .BR .lzma ,
166
+ .BR .tlz ,
167
+ or
168
+ .BR .lz ).
169
+ .PP
170
+ After successfully compressing or decompressing the
171
+ .IR file ,
172
+ .B xz
173
+ copies the owner, group, permissions, access time,
174
+ and modification time from the source
175
+ .I file
176
+ to the target file.
177
+ If copying the group fails, the permissions are modified
178
+ so that the target file doesn't become accessible to users
179
+ who didn't have permission to access the source
180
+ .IR file .
181
+ .B xz
182
+ doesn't support copying other metadata like access control lists
183
+ or extended attributes yet.
184
+ .PP
185
+ Once the target file has been successfully closed, the source
186
+ .I file
187
+ is removed unless
188
+ .B \-\-keep
189
+ was specified.
190
+ The source
191
+ .I file
192
+ is never removed if the output is written to standard output
193
+ or if an error occurs.
194
+ .PP
195
+ Sending
196
+ .B SIGINFO
197
+ or
198
+ .B SIGUSR1
199
+ to the
200
+ .B xz
201
+ process makes it print progress information to standard error.
202
+ This has only limited use since when standard error
203
+ is a terminal, using
204
+ .B \-\-verbose
205
+ will display an automatically updating progress indicator.
206
+ .
207
+ .SS "Memory usage"
208
+ The memory usage of
209
+ .B xz
210
+ varies from a few hundred kilobytes to several gigabytes
211
+ depending on the compression settings.
212
+ The settings used when compressing a file determine
213
+ the memory requirements of the decompressor.
214
+ Typically the decompressor needs 5\ % to 20\ % of
215
+ the amount of memory that the compressor needed when
216
+ creating the file.
217
+ For example, decompressing a file created with
218
+ .B xz \-9
219
+ currently requires 65\ MiB of memory.
220
+ Still, it is possible to have
221
+ .B .xz
222
+ files that require several gigabytes of memory to decompress.
223
+ .PP
224
+ Especially users of older systems may find
225
+ the possibility of very large memory usage annoying.
226
+ To prevent uncomfortable surprises,
227
+ .B xz
228
+ has a built-in memory usage limiter, which is disabled by default.
229
+ While some operating systems provide ways to limit
230
+ the memory usage of processes, relying on it
231
+ wasn't deemed to be flexible enough (for example, using
232
+ .BR ulimit (1)
233
+ to limit virtual memory tends to cripple
234
+ .BR mmap (2)).
235
+ .PP
236
+ The memory usage limiter can be enabled with
237
+ the command line option \fB\-\-memlimit=\fIlimit\fR.
238
+ Often it is more convenient to enable the limiter
239
+ by default by setting the environment variable
240
+ .BR XZ_DEFAULTS ,
241
+ for example,
242
+ .BR XZ_DEFAULTS=\-\-memlimit=150MiB .
243
+ It is possible to set the limits separately
244
+ for compression and decompression by using
245
+ .BI \-\-memlimit\-compress= limit
246
+ and \fB\-\-memlimit\-decompress=\fIlimit\fR.
247
+ Using these two options outside
248
+ .B XZ_DEFAULTS
249
+ is rarely useful because a single run of
250
+ .B xz
251
+ cannot do both compression and decompression and
252
+ .BI \-\-memlimit= limit
253
+ (or
254
+ .B \-M
255
+ .IR limit )
256
+ is shorter to type on the command line.
257
+ .PP
258
+ If the specified memory usage limit is exceeded when decompressing,
259
+ .B xz
260
+ will display an error and decompressing the file will fail.
261
+ If the limit is exceeded when compressing,
262
+ .B xz
263
+ will try to scale the settings down so that the limit
264
+ is no longer exceeded (except when using
265
+ .B \-\-format=raw
266
+ or
267
+ .BR \-\-no\-adjust ).
268
+ This way the operation won't fail unless the limit is very small.
269
+ The scaling of the settings is done in steps that don't
270
+ match the compression level presets, for example, if the limit is
271
+ only slightly less than the amount required for
272
+ .BR "xz \-9" ,
273
+ the settings will be scaled down only a little,
274
+ not all the way down to
275
+ .BR "xz \-8" .
276
+ .
277
+ .SS "Concatenation and padding with .xz files"
278
+ It is possible to concatenate
279
+ .B .xz
280
+ files as is.
281
+ .B xz
282
+ will decompress such files as if they were a single
283
+ .B .xz
284
+ file.
285
+ .PP
286
+ It is possible to insert padding between the concatenated parts
287
+ or after the last part.
288
+ The padding must consist of null bytes and the size
289
+ of the padding must be a multiple of four bytes.
290
+ This can be useful, for example, if the
291
+ .B .xz
292
+ file is stored on a medium that measures file sizes
293
+ in 512-byte blocks.
294
+ .PP
295
+ Concatenation and padding are not allowed with
296
+ .B .lzma
297
+ files or raw streams.
298
+ .
299
+ .SH OPTIONS
300
+ .
301
+ .SS "Integer suffixes and special values"
302
+ In most places where an integer argument is expected,
303
+ an optional suffix is supported to easily indicate large integers.
304
+ There must be no space between the integer and the suffix.
305
+ .TP
306
+ .B KiB
307
+ Multiply the integer by 1,024 (2^10).
308
+ .BR Ki ,
309
+ .BR k ,
310
+ .BR kB ,
311
+ .BR K ,
312
+ and
313
+ .B KB
314
+ are accepted as synonyms for
315
+ .BR KiB .
316
+ .TP
317
+ .B MiB
318
+ Multiply the integer by 1,048,576 (2^20).
319
+ .BR Mi ,
320
+ .BR m ,
321
+ .BR M ,
322
+ and
323
+ .B MB
324
+ are accepted as synonyms for
325
+ .BR MiB .
326
+ .TP
327
+ .B GiB
328
+ Multiply the integer by 1,073,741,824 (2^30).
329
+ .BR Gi ,
330
+ .BR g ,
331
+ .BR G ,
332
+ and
333
+ .B GB
334
+ are accepted as synonyms for
335
+ .BR GiB .
336
+ .PP
337
+ The special value
338
+ .B max
339
+ can be used to indicate the maximum integer value
340
+ supported by the option.
341
+ .
342
+ .SS "Operation mode"
343
+ If multiple operation mode options are given,
344
+ the last one takes effect.
345
+ .TP
346
+ .BR \-z ", " \-\-compress
347
+ Compress.
348
+ This is the default operation mode when no operation mode option
349
+ is specified and no other operation mode is implied from
350
+ the command name (for example,
351
+ .B unxz
352
+ implies
353
+ .BR \-\-decompress ).
354
+ .IP ""
355
+ .\" The DESCRIPTION section already says this but it's good to repeat it
356
+ .\" here because the default behavior is a bit dangerous and new users
357
+ .\" in a hurry may skip reading the DESCRIPTION section.
358
+ After successful compression, the source file is removed
359
+ unless writing to standard output or
360
+ .B \-\-keep
361
+ was specified.
362
+ .TP
363
+ .BR \-d ", " \-\-decompress ", " \-\-uncompress
364
+ Decompress.
365
+ .\" The DESCRIPTION section already says this but it's good to repeat it
366
+ .\" here because the default behavior is a bit dangerous and new users
367
+ .\" in a hurry may skip reading the DESCRIPTION section.
368
+ After successful decompression, the source file is removed
369
+ unless writing to standard output or
370
+ .B \-\-keep
371
+ was specified.
372
+ .TP
373
+ .BR \-t ", " \-\-test
374
+ Test the integrity of compressed
375
+ .IR files .
376
+ This option is equivalent to
377
+ .B "\-\-decompress \-\-stdout"
378
+ except that the decompressed data is discarded instead of being
379
+ written to standard output.
380
+ No files are created or removed.
381
+ .TP
382
+ .BR \-l ", " \-\-list
383
+ Print information about compressed
384
+ .IR files .
385
+ No uncompressed output is produced,
386
+ and no files are created or removed.
387
+ In list mode, the program cannot read
388
+ the compressed data from standard
389
+ input or from other unseekable sources.
390
+ .IP ""
391
+ The default listing shows basic information about
392
+ .IR files ,
393
+ one file per line.
394
+ To get more detailed information, use also the
395
+ .B \-\-verbose
396
+ option.
397
+ For even more information, use
398
+ .B \-\-verbose
399
+ twice, but note that this may be slow, because getting all the extra
400
+ information requires many seeks.
401
+ The width of verbose output exceeds
402
+ 80 characters, so piping the output to, for example,
403
+ .B "less\ \-S"
404
+ may be convenient if the terminal isn't wide enough.
405
+ .IP ""
406
+ The exact output may vary between
407
+ .B xz
408
+ versions and different locales.
409
+ For machine-readable output,
410
+ .B \-\-robot \-\-list
411
+ should be used.
412
+ .
413
+ .SS "Operation modifiers"
414
+ .TP
415
+ .BR \-k ", " \-\-keep
416
+ Don't delete the input files.
417
+ .IP ""
418
+ Since
419
+ .B xz
420
+ 5.2.6,
421
+ this option also makes
422
+ .B xz
423
+ compress or decompress even if the input is
424
+ a symbolic link to a regular file,
425
+ has more than one hard link,
426
+ or has the setuid, setgid, or sticky bit set.
427
+ The setuid, setgid, and sticky bits are not copied
428
+ to the target file.
429
+ In earlier versions this was only done with
430
+ .BR \-\-force .
431
+ .TP
432
+ .BR \-f ", " \-\-force
433
+ This option has several effects:
434
+ .RS
435
+ .IP \(bu 3
436
+ If the target file already exists,
437
+ delete it before compressing or decompressing.
438
+ .IP \(bu 3
439
+ Compress or decompress even if the input is
440
+ a symbolic link to a regular file,
441
+ has more than one hard link,
442
+ or has the setuid, setgid, or sticky bit set.
443
+ The setuid, setgid, and sticky bits are not copied
444
+ to the target file.
445
+ .IP \(bu 3
446
+ When used with
447
+ .B \-\-decompress
448
+ .B \-\-stdout
449
+ and
450
+ .B xz
451
+ cannot recognize the type of the source file,
452
+ copy the source file as is to standard output.
453
+ This allows
454
+ .B xzcat
455
+ .B \-\-force
456
+ to be used like
457
+ .BR cat (1)
458
+ for files that have not been compressed with
459
+ .BR xz .
460
+ Note that in future,
461
+ .B xz
462
+ might support new compressed file formats, which may make
463
+ .B xz
464
+ decompress more types of files instead of copying them as is to
465
+ standard output.
466
+ .BI \-\-format= format
467
+ can be used to restrict
468
+ .B xz
469
+ to decompress only a single file format.
470
+ .RE
471
+ .TP
472
+ .BR \-c ", " \-\-stdout ", " \-\-to\-stdout
473
+ Write the compressed or decompressed data to
474
+ standard output instead of a file.
475
+ This implies
476
+ .BR \-\-keep .
477
+ .TP
478
+ .B \-\-single\-stream
479
+ Decompress only the first
480
+ .B .xz
481
+ stream, and
482
+ silently ignore possible remaining input data following the stream.
483
+ Normally such trailing garbage makes
484
+ .B xz
485
+ display an error.
486
+ .IP ""
487
+ .B xz
488
+ never decompresses more than one stream from
489
+ .B .lzma
490
+ files or raw streams, but this option still makes
491
+ .B xz
492
+ ignore the possible trailing data after the
493
+ .B .lzma
494
+ file or raw stream.
495
+ .IP ""
496
+ This option has no effect if the operation mode is not
497
+ .B \-\-decompress
498
+ or
499
+ .BR \-\-test .
500
+ .TP
501
+ .B \-\-no\-sparse
502
+ Disable creation of sparse files.
503
+ By default, if decompressing into a regular file,
504
+ .B xz
505
+ tries to make the file sparse if the decompressed data contains
506
+ long sequences of binary zeros.
507
+ It also works when writing to standard output
508
+ as long as standard output is connected to a regular file
509
+ and certain additional conditions are met to make it safe.
510
+ Creating sparse files may save disk space and speed up
511
+ the decompression by reducing the amount of disk I/O.
512
+ .TP
513
+ \fB\-S\fR \fI.suf\fR, \fB\-\-suffix=\fI.suf
514
+ When compressing, use
515
+ .I .suf
516
+ as the suffix for the target file instead of
517
+ .B .xz
518
+ or
519
+ .BR .lzma .
520
+ If not writing to standard output and
521
+ the source file already has the suffix
522
+ .IR .suf ,
523
+ a warning is displayed and the file is skipped.
524
+ .IP ""
525
+ When decompressing, recognize files with the suffix
526
+ .I .suf
527
+ in addition to files with the
528
+ .BR .xz ,
529
+ .BR .txz ,
530
+ .BR .lzma ,
531
+ .BR .tlz ,
532
+ or
533
+ .B .lz
534
+ suffix.
535
+ If the source file has the suffix
536
+ .IR .suf ,
537
+ the suffix is removed to get the target filename.
538
+ .IP ""
539
+ When compressing or decompressing raw streams
540
+ .RB ( \-\-format=raw ),
541
+ the suffix must always be specified unless
542
+ writing to standard output,
543
+ because there is no default suffix for raw streams.
544
+ .TP
545
+ \fB\-\-files\fR[\fB=\fIfile\fR]
546
+ Read the filenames to process from
547
+ .IR file ;
548
+ if
549
+ .I file
550
+ is omitted, filenames are read from standard input.
551
+ Filenames must be terminated with the newline character.
552
+ A dash
553
+ .RB ( \- )
554
+ is taken as a regular filename; it doesn't mean standard input.
555
+ If filenames are given also as command line arguments, they are
556
+ processed before the filenames read from
557
+ .IR file .
558
+ .TP
559
+ \fB\-\-files0\fR[\fB=\fIfile\fR]
560
+ This is identical to \fB\-\-files\fR[\fB=\fIfile\fR] except
561
+ that each filename must be terminated with the null character.
562
+ .
563
+ .SS "Basic file format and compression options"
564
+ .TP
565
+ \fB\-F\fR \fIformat\fR, \fB\-\-format=\fIformat
566
+ Specify the file
567
+ .I format
568
+ to compress or decompress:
569
+ .RS
570
+ .TP
571
+ .B auto
572
+ This is the default.
573
+ When compressing,
574
+ .B auto
575
+ is equivalent to
576
+ .BR xz .
577
+ When decompressing,
578
+ the format of the input file is automatically detected.
579
+ Note that raw streams (created with
580
+ .BR \-\-format=raw )
581
+ cannot be auto-detected.
582
+ .TP
583
+ .B xz
584
+ Compress to the
585
+ .B .xz
586
+ file format, or accept only
587
+ .B .xz
588
+ files when decompressing.
589
+ .TP
590
+ .BR lzma ", " alone
591
+ Compress to the legacy
592
+ .B .lzma
593
+ file format, or accept only
594
+ .B .lzma
595
+ files when decompressing.
596
+ The alternative name
597
+ .B alone
598
+ is provided for backwards compatibility with LZMA Utils.
599
+ .TP
600
+ .B lzip
601
+ Accept only
602
+ .B .lz
603
+ files when decompressing.
604
+ Compression is not supported.
605
+ .IP ""
606
+ The
607
+ .B .lz
608
+ format version 0 and the unextended version 1 are supported.
609
+ Version 0 files were produced by
610
+ .B lzip
611
+ 1.3 and older.
612
+ Such files aren't common but may be found from file archives
613
+ as a few source packages were released in this format.
614
+ People might have old personal files in this format too.
615
+ Decompression support for the format version 0 was removed in
616
+ .B lzip
617
+ 1.18.
618
+ .IP ""
619
+ .B lzip
620
+ 1.4 and later create files in the format version 1.
621
+ The sync flush marker extension to the format version 1 was added in
622
+ .B lzip
623
+ 1.6.
624
+ This extension is rarely used and isn't supported by
625
+ .B xz
626
+ (diagnosed as corrupt input).
627
+ .TP
628
+ .B raw
629
+ Compress or uncompress a raw stream (no headers).
630
+ This is meant for advanced users only.
631
+ To decode raw streams, you need use
632
+ .B \-\-format=raw
633
+ and explicitly specify the filter chain,
634
+ which normally would have been stored in the container headers.
635
+ .RE
636
+ .TP
637
+ \fB\-C\fR \fIcheck\fR, \fB\-\-check=\fIcheck
638
+ Specify the type of the integrity check.
639
+ The check is calculated from the uncompressed data and
640
+ stored in the
641
+ .B .xz
642
+ file.
643
+ This option has an effect only when compressing into the
644
+ .B .xz
645
+ format; the
646
+ .B .lzma
647
+ format doesn't support integrity checks.
648
+ The integrity check (if any) is verified when the
649
+ .B .xz
650
+ file is decompressed.
651
+ .IP ""
652
+ Supported
653
+ .I check
654
+ types:
655
+ .RS
656
+ .TP
657
+ .B none
658
+ Don't calculate an integrity check at all.
659
+ This is usually a bad idea.
660
+ This can be useful when integrity of the data is verified
661
+ by other means anyway.
662
+ .TP
663
+ .B crc32
664
+ Calculate CRC32 using the polynomial from IEEE-802.3 (Ethernet).
665
+ .TP
666
+ .B crc64
667
+ Calculate CRC64 using the polynomial from ECMA-182.
668
+ This is the default, since it is slightly better than CRC32
669
+ at detecting damaged files and the speed difference is negligible.
670
+ .TP
671
+ .B sha256
672
+ Calculate SHA-256.
673
+ This is somewhat slower than CRC32 and CRC64.
674
+ .RE
675
+ .IP ""
676
+ Integrity of the
677
+ .B .xz
678
+ headers is always verified with CRC32.
679
+ It is not possible to change or disable it.
680
+ .TP
681
+ .B \-\-ignore\-check
682
+ Don't verify the integrity check of the compressed data when decompressing.
683
+ The CRC32 values in the
684
+ .B .xz
685
+ headers will still be verified normally.
686
+ .IP ""
687
+ .B "Do not use this option unless you know what you are doing."
688
+ Possible reasons to use this option:
689
+ .RS
690
+ .IP \(bu 3
691
+ Trying to recover data from a corrupt .xz file.
692
+ .IP \(bu 3
693
+ Speeding up decompression.
694
+ This matters mostly with SHA-256 or
695
+ with files that have compressed extremely well.
696
+ It's recommended to not use this option for this purpose
697
+ unless the file integrity is verified externally in some other way.
698
+ .RE
699
+ .TP
700
+ .BR \-0 " ... " \-9
701
+ Select a compression preset level.
702
+ The default is
703
+ .BR \-6 .
704
+ If multiple preset levels are specified,
705
+ the last one takes effect.
706
+ If a custom filter chain was already specified, setting
707
+ a compression preset level clears the custom filter chain.
708
+ .IP ""
709
+ The differences between the presets are more significant than with
710
+ .BR gzip (1)
711
+ and
712
+ .BR bzip2 (1).
713
+ The selected compression settings determine
714
+ the memory requirements of the decompressor,
715
+ thus using a too high preset level might make it painful
716
+ to decompress the file on an old system with little RAM.
717
+ Specifically,
718
+ .B "it's not a good idea to blindly use \-9 for everything"
719
+ like it often is with
720
+ .BR gzip (1)
721
+ and
722
+ .BR bzip2 (1).
723
+ .RS
724
+ .TP
725
+ .BR "\-0" " ... " "\-3"
726
+ These are somewhat fast presets.
727
+ .B \-0
728
+ is sometimes faster than
729
+ .B "gzip \-9"
730
+ while compressing much better.
731
+ The higher ones often have speed comparable to
732
+ .BR bzip2 (1)
733
+ with comparable or better compression ratio,
734
+ although the results
735
+ depend a lot on the type of data being compressed.
736
+ .TP
737
+ .BR "\-4" " ... " "\-6"
738
+ Good to very good compression while keeping
739
+ decompressor memory usage reasonable even for old systems.
740
+ .B \-6
741
+ is the default, which is usually a good choice
742
+ for distributing files that need to be decompressible
743
+ even on systems with only 16\ MiB RAM.
744
+ .RB ( \-5e
745
+ or
746
+ .B \-6e
747
+ may be worth considering too.
748
+ See
749
+ .BR \-\-extreme .)
750
+ .TP
751
+ .B "\-7 ... \-9"
752
+ These are like
753
+ .B \-6
754
+ but with higher compressor and decompressor memory requirements.
755
+ These are useful only when compressing files bigger than
756
+ 8\ MiB, 16\ MiB, and 32\ MiB, respectively.
757
+ .RE
758
+ .IP ""
759
+ On the same hardware, the decompression speed is approximately
760
+ a constant number of bytes of compressed data per second.
761
+ In other words, the better the compression,
762
+ the faster the decompression will usually be.
763
+ This also means that the amount of uncompressed output
764
+ produced per second can vary a lot.
765
+ .IP ""
766
+ The following table summarises the features of the presets:
767
+ .RS
768
+ .RS
769
+ .PP
770
+ .TS
771
+ tab(;);
772
+ c c c c c
773
+ n n n n n.
774
+ Preset;DictSize;CompCPU;CompMem;DecMem
775
+ \-0;256 KiB;0;3 MiB;1 MiB
776
+ \-1;1 MiB;1;9 MiB;2 MiB
777
+ \-2;2 MiB;2;17 MiB;3 MiB
778
+ \-3;4 MiB;3;32 MiB;5 MiB
779
+ \-4;4 MiB;4;48 MiB;5 MiB
780
+ \-5;8 MiB;5;94 MiB;9 MiB
781
+ \-6;8 MiB;6;94 MiB;9 MiB
782
+ \-7;16 MiB;6;186 MiB;17 MiB
783
+ \-8;32 MiB;6;370 MiB;33 MiB
784
+ \-9;64 MiB;6;674 MiB;65 MiB
785
+ .TE
786
+ .RE
787
+ .RE
788
+ .IP ""
789
+ Column descriptions:
790
+ .RS
791
+ .IP \(bu 3
792
+ DictSize is the LZMA2 dictionary size.
793
+ It is waste of memory to use a dictionary bigger than
794
+ the size of the uncompressed file.
795
+ This is why it is good to avoid using the presets
796
+ .BR \-7 " ... " \-9
797
+ when there's no real need for them.
798
+ At
799
+ .B \-6
800
+ and lower, the amount of memory wasted is
801
+ usually low enough to not matter.
802
+ .IP \(bu 3
803
+ CompCPU is a simplified representation of the LZMA2 settings
804
+ that affect compression speed.
805
+ The dictionary size affects speed too,
806
+ so while CompCPU is the same for levels
807
+ .BR \-6 " ... " \-9 ,
808
+ higher levels still tend to be a little slower.
809
+ To get even slower and thus possibly better compression, see
810
+ .BR \-\-extreme .
811
+ .IP \(bu 3
812
+ CompMem contains the compressor memory requirements
813
+ in the single-threaded mode.
814
+ It may vary slightly between
815
+ .B xz
816
+ versions.
817
+ .IP \(bu 3
818
+ DecMem contains the decompressor memory requirements.
819
+ That is, the compression settings determine
820
+ the memory requirements of the decompressor.
821
+ The exact decompressor memory usage is slightly more than
822
+ the LZMA2 dictionary size, but the values in the table
823
+ have been rounded up to the next full MiB.
824
+ .RE
825
+ .IP ""
826
+ Memory requirements of the multi-threaded mode are
827
+ significantly higher than that of the single-threaded mode.
828
+ With the default value of
829
+ .BR \-\-block\-size ,
830
+ each thread needs 3*3*DictSize plus CompMem or DecMem.
831
+ For example, four threads with preset
832
+ .B \-6
833
+ needs 660\(en670\ MiB of memory.
834
+ .TP
835
+ .BR \-e ", " \-\-extreme
836
+ Use a slower variant of the selected compression preset level
837
+ .RB ( \-0 " ... " \-9 )
838
+ to hopefully get a little bit better compression ratio,
839
+ but with bad luck this can also make it worse.
840
+ Decompressor memory usage is not affected,
841
+ but compressor memory usage increases a little at preset levels
842
+ .BR \-0 " ... " \-3 .
843
+ .IP ""
844
+ Since there are two presets with dictionary sizes
845
+ 4\ MiB and 8\ MiB, the presets
846
+ .B \-3e
847
+ and
848
+ .B \-5e
849
+ use slightly faster settings (lower CompCPU) than
850
+ .B \-4e
851
+ and
852
+ .BR \-6e ,
853
+ respectively.
854
+ That way no two presets are identical.
855
+ .RS
856
+ .RS
857
+ .PP
858
+ .TS
859
+ tab(;);
860
+ c c c c c
861
+ n n n n n.
862
+ Preset;DictSize;CompCPU;CompMem;DecMem
863
+ \-0e;256 KiB;8;4 MiB;1 MiB
864
+ \-1e;1 MiB;8;13 MiB;2 MiB
865
+ \-2e;2 MiB;8;25 MiB;3 MiB
866
+ \-3e;4 MiB;7;48 MiB;5 MiB
867
+ \-4e;4 MiB;8;48 MiB;5 MiB
868
+ \-5e;8 MiB;7;94 MiB;9 MiB
869
+ \-6e;8 MiB;8;94 MiB;9 MiB
870
+ \-7e;16 MiB;8;186 MiB;17 MiB
871
+ \-8e;32 MiB;8;370 MiB;33 MiB
872
+ \-9e;64 MiB;8;674 MiB;65 MiB
873
+ .TE
874
+ .RE
875
+ .RE
876
+ .IP ""
877
+ For example, there are a total of four presets that use
878
+ 8\ MiB dictionary, whose order from the fastest to the slowest is
879
+ .BR \-5 ,
880
+ .BR \-6 ,
881
+ .BR \-5e ,
882
+ and
883
+ .BR \-6e .
884
+ .TP
885
+ .B \-\-fast
886
+ .PD 0
887
+ .TP
888
+ .B \-\-best
889
+ .PD
890
+ These are somewhat misleading aliases for
891
+ .B \-0
892
+ and
893
+ .BR \-9 ,
894
+ respectively.
895
+ These are provided only for backwards compatibility
896
+ with LZMA Utils.
897
+ Avoid using these options.
898
+ .TP
899
+ .BI \-\-block\-size= size
900
+ When compressing to the
901
+ .B .xz
902
+ format, split the input data into blocks of
903
+ .I size
904
+ bytes.
905
+ The blocks are compressed independently from each other,
906
+ which helps with multi-threading and
907
+ makes limited random-access decompression possible.
908
+ This option is typically used to override the default
909
+ block size in multi-threaded mode,
910
+ but this option can be used in single-threaded mode too.
911
+ .IP ""
912
+ In multi-threaded mode about three times
913
+ .I size
914
+ bytes will be allocated in each thread for buffering input and output.
915
+ The default
916
+ .I size
917
+ is three times the LZMA2 dictionary size or 1 MiB,
918
+ whichever is more.
919
+ Typically a good value is 2\(en4 times
920
+ the size of the LZMA2 dictionary or at least 1 MiB.
921
+ Using
922
+ .I size
923
+ less than the LZMA2 dictionary size is waste of RAM
924
+ because then the LZMA2 dictionary buffer will never get fully used.
925
+ In multi-threaded mode,
926
+ the sizes of the blocks are stored in the block headers.
927
+ This size information is required for multi-threaded decompression.
928
+ .IP ""
929
+ In single-threaded mode no block splitting is done by default.
930
+ Setting this option doesn't affect memory usage.
931
+ No size information is stored in block headers,
932
+ thus files created in single-threaded mode
933
+ won't be identical to files created in multi-threaded mode.
934
+ The lack of size information also means that
935
+ .B xz
936
+ won't be able decompress the files in multi-threaded mode.
937
+ .TP
938
+ .BI \-\-block\-list= items
939
+ When compressing to the
940
+ .B .xz
941
+ format, start a new block with an optional custom filter chain after
942
+ the given intervals of uncompressed data.
943
+ .IP ""
944
+ The
945
+ .I items
946
+ are a comma-separated list.
947
+ Each item consists of an optional filter chain number
948
+ between 0 and 9 followed by a colon
949
+ .RB ( : )
950
+ and a required size of uncompressed data.
951
+ Omitting an item (two or more consecutive commas) is a
952
+ shorthand to use the size and filters of the previous item.
953
+ .IP ""
954
+ If the input file is bigger than the sum of
955
+ the sizes in
956
+ .IR items ,
957
+ the last item is repeated until the end of the file.
958
+ A special value of
959
+ .B 0
960
+ may be used as the last size to indicate that
961
+ the rest of the file should be encoded as a single block.
962
+ .IP ""
963
+ An alternative filter chain for each block can be
964
+ specified in combination with the
965
+ .BI \-\-filters1= filters
966
+ \&...\&
967
+ .BI \-\-filters9= filters
968
+ options.
969
+ These options define filter chains with an identifier
970
+ between 1\(en9.
971
+ Filter chain 0 can be used to refer to the default filter chain,
972
+ which is the same as not specifying a filter chain.
973
+ The filter chain identifier can be used before the uncompressed
974
+ size, followed by a colon
975
+ .RB ( : ).
976
+ For example, if one specifies
977
+ .B \-\-block\-list=1:2MiB,3:2MiB,2:4MiB,,2MiB,0:4MiB
978
+ then blocks will be created using:
979
+ .RS
980
+ .IP \(bu 3
981
+ The filter chain specified by
982
+ .B \-\-filters1
983
+ and 2 MiB input
984
+ .IP \(bu 3
985
+ The filter chain specified by
986
+ .B \-\-filters3
987
+ and 2 MiB input
988
+ .IP \(bu 3
989
+ The filter chain specified by
990
+ .B \-\-filters2
991
+ and 4 MiB input
992
+ .IP \(bu 3
993
+ The filter chain specified by
994
+ .B \-\-filters2
995
+ and 4 MiB input
996
+ .IP \(bu 3
997
+ The default filter chain and 2 MiB input
998
+ .IP \(bu 3
999
+ The default filter chain and 4 MiB input for every block until
1000
+ end of input.
1001
+ .RE
1002
+ .IP ""
1003
+ If one specifies a size that exceeds the encoder's block size
1004
+ (either the default value in threaded mode or
1005
+ the value specified with \fB\-\-block\-size=\fIsize\fR),
1006
+ the encoder will create additional blocks while
1007
+ keeping the boundaries specified in
1008
+ .IR items .
1009
+ For example, if one specifies
1010
+ .B \-\-block\-size=10MiB
1011
+ .B \-\-block\-list=5MiB,10MiB,8MiB,12MiB,24MiB
1012
+ and the input file is 80 MiB,
1013
+ one will get 11 blocks:
1014
+ 5, 10, 8, 10, 2, 10, 10, 4, 10, 10, and 1 MiB.
1015
+ .IP ""
1016
+ In multi-threaded mode the sizes of the blocks
1017
+ are stored in the block headers.
1018
+ This isn't done in single-threaded mode,
1019
+ so the encoded output won't be
1020
+ identical to that of the multi-threaded mode.
1021
+ .TP
1022
+ .BI \-\-flush\-timeout= timeout
1023
+ When compressing, if more than
1024
+ .I timeout
1025
+ milliseconds (a positive integer) has passed since the previous flush and
1026
+ reading more input would block,
1027
+ all the pending input data is flushed from the encoder and
1028
+ made available in the output stream.
1029
+ This can be useful if
1030
+ .B xz
1031
+ is used to compress data that is streamed over a network.
1032
+ Small
1033
+ .I timeout
1034
+ values make the data available at the receiving end
1035
+ with a small delay, but large
1036
+ .I timeout
1037
+ values give better compression ratio.
1038
+ .IP ""
1039
+ This feature is disabled by default.
1040
+ If this option is specified more than once, the last one takes effect.
1041
+ The special
1042
+ .I timeout
1043
+ value of
1044
+ .B 0
1045
+ can be used to explicitly disable this feature.
1046
+ .IP ""
1047
+ This feature is not available on non-POSIX systems.
1048
+ .IP ""
1049
+ .\" FIXME
1050
+ .B "This feature is still experimental."
1051
+ Currently
1052
+ .B xz
1053
+ is unsuitable for decompressing the stream in real time due to how
1054
+ .B xz
1055
+ does buffering.
1056
+ .TP
1057
+ .BI \-\-memlimit\-compress= limit
1058
+ Set a memory usage limit for compression.
1059
+ If this option is specified multiple times,
1060
+ the last one takes effect.
1061
+ .IP ""
1062
+ If the compression settings exceed the
1063
+ .IR limit ,
1064
+ .B xz
1065
+ will attempt to adjust the settings downwards so that
1066
+ the limit is no longer exceeded and display a notice that
1067
+ automatic adjustment was done.
1068
+ The adjustments are done in this order:
1069
+ reducing the number of threads,
1070
+ switching to single-threaded mode
1071
+ if even one thread in multi-threaded mode exceeds the
1072
+ .IR limit ,
1073
+ and finally reducing the LZMA2 dictionary size.
1074
+ .IP ""
1075
+ When compressing with
1076
+ .B \-\-format=raw
1077
+ or if
1078
+ .B \-\-no\-adjust
1079
+ has been specified,
1080
+ only the number of threads may be reduced
1081
+ since it can be done without affecting the compressed output.
1082
+ .IP ""
1083
+ If the
1084
+ .I limit
1085
+ cannot be met even with the adjustments described above,
1086
+ an error is displayed and
1087
+ .B xz
1088
+ will exit with exit status 1.
1089
+ .IP ""
1090
+ The
1091
+ .I limit
1092
+ can be specified in multiple ways:
1093
+ .RS
1094
+ .IP \(bu 3
1095
+ The
1096
+ .I limit
1097
+ can be an absolute value in bytes.
1098
+ Using an integer suffix like
1099
+ .B MiB
1100
+ can be useful.
1101
+ Example:
1102
+ .B "\-\-memlimit\-compress=80MiB"
1103
+ .IP \(bu 3
1104
+ The
1105
+ .I limit
1106
+ can be specified as a percentage of total physical memory (RAM).
1107
+ This can be useful especially when setting the
1108
+ .B XZ_DEFAULTS
1109
+ environment variable in a shell initialization script
1110
+ that is shared between different computers.
1111
+ That way the limit is automatically bigger
1112
+ on systems with more memory.
1113
+ Example:
1114
+ .B "\-\-memlimit\-compress=70%"
1115
+ .IP \(bu 3
1116
+ The
1117
+ .I limit
1118
+ can be reset back to its default value by setting it to
1119
+ .BR 0 .
1120
+ This is currently equivalent to setting the
1121
+ .I limit
1122
+ to
1123
+ .B max
1124
+ (no memory usage limit).
1125
+ .RE
1126
+ .IP ""
1127
+ For 32-bit
1128
+ .B xz
1129
+ there is a special case: if the
1130
+ .I limit
1131
+ would be over
1132
+ .BR "4020\ MiB" ,
1133
+ the
1134
+ .I limit
1135
+ is set to
1136
+ .BR "4020\ MiB" .
1137
+ On MIPS32
1138
+ .B "2000\ MiB"
1139
+ is used instead.
1140
+ (The values
1141
+ .B 0
1142
+ and
1143
+ .B max
1144
+ aren't affected by this.
1145
+ A similar feature doesn't exist for decompression.)
1146
+ This can be helpful when a 32-bit executable has access
1147
+ to 4\ GiB address space (2 GiB on MIPS32)
1148
+ while hopefully doing no harm in other situations.
1149
+ .IP ""
1150
+ See also the section
1151
+ .BR "Memory usage" .
1152
+ .TP
1153
+ .BI \-\-memlimit\-decompress= limit
1154
+ Set a memory usage limit for decompression.
1155
+ This also affects the
1156
+ .B \-\-list
1157
+ mode.
1158
+ If the operation is not possible without exceeding the
1159
+ .IR limit ,
1160
+ .B xz
1161
+ will display an error and decompressing the file will fail.
1162
+ See
1163
+ .BI \-\-memlimit\-compress= limit
1164
+ for possible ways to specify the
1165
+ .IR limit .
1166
+ .TP
1167
+ .BI \-\-memlimit\-mt\-decompress= limit
1168
+ Set a memory usage limit for multi-threaded decompression.
1169
+ This can only affect the number of threads;
1170
+ this will never make
1171
+ .B xz
1172
+ refuse to decompress a file.
1173
+ If
1174
+ .I limit
1175
+ is too low to allow any multi-threading, the
1176
+ .I limit
1177
+ is ignored and
1178
+ .B xz
1179
+ will continue in single-threaded mode.
1180
+ Note that if also
1181
+ .B \-\-memlimit\-decompress
1182
+ is used,
1183
+ it will always apply to both single-threaded and multi-threaded modes,
1184
+ and so the effective
1185
+ .I limit
1186
+ for multi-threading will never be higher than the limit set with
1187
+ .BR \-\-memlimit\-decompress .
1188
+ .IP ""
1189
+ In contrast to the other memory usage limit options,
1190
+ .BI \-\-memlimit\-mt\-decompress= limit
1191
+ has a system-specific default
1192
+ .IR limit .
1193
+ .B "xz \-\-info\-memory"
1194
+ can be used to see the current value.
1195
+ .IP ""
1196
+ This option and its default value exist
1197
+ because without any limit the threaded decompressor
1198
+ could end up allocating an insane amount of memory with some input files.
1199
+ If the default
1200
+ .I limit
1201
+ is too low on your system,
1202
+ feel free to increase the
1203
+ .I limit
1204
+ but never set it to a value larger than the amount of usable RAM
1205
+ as with appropriate input files
1206
+ .B xz
1207
+ will attempt to use that amount of memory
1208
+ even with a low number of threads.
1209
+ Running out of memory or swapping
1210
+ will not improve decompression performance.
1211
+ .IP ""
1212
+ See
1213
+ .BI \-\-memlimit\-compress= limit
1214
+ for possible ways to specify the
1215
+ .IR limit .
1216
+ Setting
1217
+ .I limit
1218
+ to
1219
+ .B 0
1220
+ resets the
1221
+ .I limit
1222
+ to the default system-specific value.
1223
+ .TP
1224
+ \fB\-M\fR \fIlimit\fR, \fB\-\-memlimit=\fIlimit\fR, \fB\-\-memory=\fIlimit
1225
+ This is equivalent to specifying
1226
+ .BI \-\-memlimit\-compress= limit
1227
+ .BI \-\-memlimit-decompress= limit
1228
+ \fB\-\-memlimit\-mt\-decompress=\fIlimit\fR.
1229
+ .TP
1230
+ .B \-\-no\-adjust
1231
+ Display an error and exit if the memory usage limit cannot be
1232
+ met without adjusting settings that affect the compressed output.
1233
+ That is, this prevents
1234
+ .B xz
1235
+ from switching the encoder from multi-threaded mode to single-threaded mode
1236
+ and from reducing the LZMA2 dictionary size.
1237
+ Even when this option is used the number of threads may be reduced
1238
+ to meet the memory usage limit as that won't affect the compressed output.
1239
+ .IP ""
1240
+ Automatic adjusting is always disabled when creating raw streams
1241
+ .RB ( \-\-format=raw ).
1242
+ .TP
1243
+ \fB\-T\fR \fIthreads\fR, \fB\-\-threads=\fIthreads
1244
+ Specify the number of worker threads to use.
1245
+ Setting
1246
+ .I threads
1247
+ to a special value
1248
+ .B 0
1249
+ makes
1250
+ .B xz
1251
+ use up to as many threads as the processor(s) on the system support.
1252
+ The actual number of threads can be fewer than
1253
+ .I threads
1254
+ if the input file is not big enough
1255
+ for threading with the given settings or
1256
+ if using more threads would exceed the memory usage limit.
1257
+ .IP ""
1258
+ The single-threaded and multi-threaded compressors produce different output.
1259
+ Single-threaded compressor will give the smallest file size but
1260
+ only the output from the multi-threaded compressor can be decompressed
1261
+ using multiple threads.
1262
+ Setting
1263
+ .I threads
1264
+ to
1265
+ .B 1
1266
+ will use the single-threaded mode.
1267
+ Setting
1268
+ .I threads
1269
+ to any other value, including
1270
+ .BR 0 ,
1271
+ will use the multi-threaded compressor
1272
+ even if the system supports only one hardware thread.
1273
+ .RB ( xz
1274
+ 5.2.x
1275
+ used single-threaded mode in this situation.)
1276
+ .IP ""
1277
+ To use multi-threaded mode with only one thread, set
1278
+ .I threads
1279
+ to
1280
+ .BR +1 .
1281
+ The
1282
+ .B +
1283
+ prefix has no effect with values other than
1284
+ .BR 1 .
1285
+ A memory usage limit can still make
1286
+ .B xz
1287
+ switch to single-threaded mode unless
1288
+ .B \-\-no\-adjust
1289
+ is used.
1290
+ Support for the
1291
+ .B +
1292
+ prefix was added in
1293
+ .B xz
1294
+ 5.4.0.
1295
+ .IP ""
1296
+ If an automatic number of threads has been requested and
1297
+ no memory usage limit has been specified,
1298
+ then a system-specific default soft limit will be used to possibly
1299
+ limit the number of threads.
1300
+ It is a soft limit in sense that it is ignored
1301
+ if the number of threads becomes one,
1302
+ thus a soft limit will never stop
1303
+ .B xz
1304
+ from compressing or decompressing.
1305
+ This default soft limit will not make
1306
+ .B xz
1307
+ switch from multi-threaded mode to single-threaded mode.
1308
+ The active limits can be seen with
1309
+ .BR "xz \-\-info\-memory" .
1310
+ .IP ""
1311
+ Currently the only threading method is to split the input into
1312
+ blocks and compress them independently from each other.
1313
+ The default block size depends on the compression level and
1314
+ can be overridden with the
1315
+ .BI \-\-block\-size= size
1316
+ option.
1317
+ .IP ""
1318
+ Threaded decompression only works on files that contain
1319
+ multiple blocks with size information in block headers.
1320
+ All large enough files compressed in multi-threaded mode
1321
+ meet this condition,
1322
+ but files compressed in single-threaded mode don't even if
1323
+ .BI \-\-block\-size= size
1324
+ has been used.
1325
+ .IP ""
1326
+ The default value for
1327
+ .I threads
1328
+ is
1329
+ .BR 0 .
1330
+ In
1331
+ .B xz
1332
+ 5.4.x and older the default is
1333
+ .BR 1 .
1334
+ .
1335
+ .SS "Custom compressor filter chains"
1336
+ A custom filter chain allows specifying
1337
+ the compression settings in detail instead of relying on
1338
+ the settings associated to the presets.
1339
+ When a custom filter chain is specified,
1340
+ preset options
1341
+ .RB ( \-0
1342
+ \&...\&
1343
+ .B \-9
1344
+ and
1345
+ .BR \-\-extreme )
1346
+ earlier on the command line are forgotten.
1347
+ If a preset option is specified
1348
+ after one or more custom filter chain options,
1349
+ the new preset takes effect and
1350
+ the custom filter chain options specified earlier are forgotten.
1351
+ .PP
1352
+ A filter chain is comparable to piping on the command line.
1353
+ When compressing, the uncompressed input goes to the first filter,
1354
+ whose output goes to the next filter (if any).
1355
+ The output of the last filter gets written to the compressed file.
1356
+ The maximum number of filters in the chain is four,
1357
+ but typically a filter chain has only one or two filters.
1358
+ .PP
1359
+ Many filters have limitations on where they can be
1360
+ in the filter chain:
1361
+ some filters can work only as the last filter in the chain,
1362
+ some only as a non-last filter, and some work in any position
1363
+ in the chain.
1364
+ Depending on the filter, this limitation is either inherent to
1365
+ the filter design or exists to prevent security issues.
1366
+ .PP
1367
+ A custom filter chain can be specified in two different ways.
1368
+ The options
1369
+ .BI \-\-filters= filters
1370
+ and
1371
+ .BI \-\-filters1= filters
1372
+ \&...\&
1373
+ .BI \-\-filters9= filters
1374
+ allow specifying an entire filter chain in one option using the
1375
+ liblzma filter string syntax.
1376
+ Alternatively, a filter chain can be specified by using one or more
1377
+ individual filter options in the order they are wanted in the filter chain.
1378
+ That is, the order of the individual filter options is significant!
1379
+ When decoding raw streams
1380
+ .RB ( \-\-format=raw ),
1381
+ the filter chain must be specified in the same order as
1382
+ it was specified when compressing.
1383
+ Any individual filter or preset options specified before the full
1384
+ chain option
1385
+ (\fB\-\-filters=\fIfilters\fR)
1386
+ will be forgotten.
1387
+ Individual filters specified after the full chain option will reset the
1388
+ filter chain.
1389
+ .PP
1390
+ Both the full and individual filter options take filter-specific
1391
+ .I options
1392
+ as a comma-separated list.
1393
+ Extra commas in
1394
+ .I options
1395
+ are ignored.
1396
+ Every option has a default value, so
1397
+ specify those you want to change.
1398
+ .PP
1399
+ To see the whole filter chain and
1400
+ .IR options ,
1401
+ use
1402
+ .B "xz \-vv"
1403
+ (that is, use
1404
+ .B \-\-verbose
1405
+ twice).
1406
+ This works also for viewing the filter chain options used by presets.
1407
+ .TP
1408
+ .BI \-\-filters= filters
1409
+ Specify the full filter chain or a preset in a single option.
1410
+ Each filter can be separated by spaces or two dashes
1411
+ .RB ( \-\- ).
1412
+ .I filters
1413
+ may need to be quoted on the shell command line so it is
1414
+ parsed as a single option.
1415
+ To denote
1416
+ .IR options ,
1417
+ use
1418
+ .B :
1419
+ or
1420
+ .BR = .
1421
+ A preset can be prefixed with a
1422
+ .B \-
1423
+ and followed with zero or more flags.
1424
+ The only supported flag is
1425
+ .B e
1426
+ to apply the same options as
1427
+ .BR \-\-extreme .
1428
+ .TP
1429
+ \fB\-\-filters1\fR=\fIfilters\fR ... \fB\-\-filters9\fR=\fIfilters
1430
+ Specify up to nine additional filter chains that can be used with
1431
+ .BR \-\-block\-list .
1432
+ .IP ""
1433
+ For example, when compressing an archive with executable files
1434
+ followed by text files, the executable part could use a filter
1435
+ chain with a BCJ filter and the text part only the LZMA2 filter.
1436
+ .TP
1437
+ .B \-\-filters-help
1438
+ Display a help message describing how to specify presets and
1439
+ custom filter chains in the
1440
+ .B \-\-filters
1441
+ and
1442
+ .BI \-\-filters1= filters
1443
+ \&...\&
1444
+ .BI \-\-filters9= filters
1445
+ options, and exit successfully.
1446
+ .TP
1447
+ \fB\-\-lzma1\fR[\fB=\fIoptions\fR]
1448
+ .PD 0
1449
+ .TP
1450
+ \fB\-\-lzma2\fR[\fB=\fIoptions\fR]
1451
+ .PD
1452
+ Add LZMA1 or LZMA2 filter to the filter chain.
1453
+ These filters can be used only as the last filter in the chain.
1454
+ .IP ""
1455
+ LZMA1 is a legacy filter,
1456
+ which is supported almost solely due to the legacy
1457
+ .B .lzma
1458
+ file format, which supports only LZMA1.
1459
+ LZMA2 is an updated
1460
+ version of LZMA1 to fix some practical issues of LZMA1.
1461
+ The
1462
+ .B .xz
1463
+ format uses LZMA2 and doesn't support LZMA1 at all.
1464
+ Compression speed and ratios of LZMA1 and LZMA2
1465
+ are practically the same.
1466
+ .IP ""
1467
+ LZMA1 and LZMA2 share the same set of
1468
+ .IR options :
1469
+ .RS
1470
+ .TP
1471
+ .BI preset= preset
1472
+ Reset all LZMA1 or LZMA2
1473
+ .I options
1474
+ to
1475
+ .IR preset .
1476
+ .I Preset
1477
+ consist of an integer, which may be followed by single-letter
1478
+ preset modifiers.
1479
+ The integer can be from
1480
+ .B 0
1481
+ to
1482
+ .BR 9 ,
1483
+ matching the command line options
1484
+ .B \-0
1485
+ \&...\&
1486
+ .BR \-9 .
1487
+ The only supported modifier is currently
1488
+ .BR e ,
1489
+ which matches
1490
+ .BR \-\-extreme .
1491
+ If no
1492
+ .B preset
1493
+ is specified, the default values of LZMA1 or LZMA2
1494
+ .I options
1495
+ are taken from the preset
1496
+ .BR 6 .
1497
+ .TP
1498
+ .BI dict= size
1499
+ Dictionary (history buffer)
1500
+ .I size
1501
+ indicates how many bytes of the recently processed
1502
+ uncompressed data is kept in memory.
1503
+ The algorithm tries to find repeating byte sequences (matches) in
1504
+ the uncompressed data, and replace them with references
1505
+ to the data currently in the dictionary.
1506
+ The bigger the dictionary, the higher is the chance
1507
+ to find a match.
1508
+ Thus, increasing dictionary
1509
+ .I size
1510
+ usually improves compression ratio, but
1511
+ a dictionary bigger than the uncompressed file is waste of memory.
1512
+ .IP ""
1513
+ Typical dictionary
1514
+ .I size
1515
+ is from 64\ KiB to 64\ MiB.
1516
+ The minimum is 4\ KiB.
1517
+ The maximum for compression is currently 1.5\ GiB (1536\ MiB).
1518
+ The decompressor already supports dictionaries up to
1519
+ one byte less than 4\ GiB, which is the maximum for
1520
+ the LZMA1 and LZMA2 stream formats.
1521
+ .IP ""
1522
+ Dictionary
1523
+ .I size
1524
+ and match finder
1525
+ .RI ( mf )
1526
+ together determine the memory usage of the LZMA1 or LZMA2 encoder.
1527
+ The same (or bigger) dictionary
1528
+ .I size
1529
+ is required for decompressing that was used when compressing,
1530
+ thus the memory usage of the decoder is determined
1531
+ by the dictionary size used when compressing.
1532
+ The
1533
+ .B .xz
1534
+ headers store the dictionary
1535
+ .I size
1536
+ either as
1537
+ .RI "2^" n
1538
+ or
1539
+ .RI "2^" n " + 2^(" n "\-1),"
1540
+ so these
1541
+ .I sizes
1542
+ are somewhat preferred for compression.
1543
+ Other
1544
+ .I sizes
1545
+ will get rounded up when stored in the
1546
+ .B .xz
1547
+ headers.
1548
+ .TP
1549
+ .BI lc= lc
1550
+ Specify the number of literal context bits.
1551
+ The minimum is 0 and the maximum is 4; the default is 3.
1552
+ In addition, the sum of
1553
+ .I lc
1554
+ and
1555
+ .I lp
1556
+ must not exceed 4.
1557
+ .IP ""
1558
+ All bytes that cannot be encoded as matches
1559
+ are encoded as literals.
1560
+ That is, literals are simply 8-bit bytes
1561
+ that are encoded one at a time.
1562
+ .IP ""
1563
+ The literal coding makes an assumption that the highest
1564
+ .I lc
1565
+ bits of the previous uncompressed byte correlate
1566
+ with the next byte.
1567
+ For example, in typical English text, an upper-case letter is
1568
+ often followed by a lower-case letter, and a lower-case
1569
+ letter is usually followed by another lower-case letter.
1570
+ In the US-ASCII character set, the highest three bits are 010
1571
+ for upper-case letters and 011 for lower-case letters.
1572
+ When
1573
+ .I lc
1574
+ is at least 3, the literal coding can take advantage of
1575
+ this property in the uncompressed data.
1576
+ .IP ""
1577
+ The default value (3) is usually good.
1578
+ If you want maximum compression, test
1579
+ .BR lc=4 .
1580
+ Sometimes it helps a little, and
1581
+ sometimes it makes compression worse.
1582
+ If it makes it worse, test
1583
+ .B lc=2
1584
+ too.
1585
+ .TP
1586
+ .BI lp= lp
1587
+ Specify the number of literal position bits.
1588
+ The minimum is 0 and the maximum is 4; the default is 0.
1589
+ .IP ""
1590
+ .I Lp
1591
+ affects what kind of alignment in the uncompressed data is
1592
+ assumed when encoding literals.
1593
+ See
1594
+ .I pb
1595
+ below for more information about alignment.
1596
+ .TP
1597
+ .BI pb= pb
1598
+ Specify the number of position bits.
1599
+ The minimum is 0 and the maximum is 4; the default is 2.
1600
+ .IP ""
1601
+ .I Pb
1602
+ affects what kind of alignment in the uncompressed data is
1603
+ assumed in general.
1604
+ The default means four-byte alignment
1605
+ .RI (2^ pb =2^2=4),
1606
+ which is often a good choice when there's no better guess.
1607
+ .IP ""
1608
+ When the alignment is known, setting
1609
+ .I pb
1610
+ accordingly may reduce the file size a little.
1611
+ For example, with text files having one-byte
1612
+ alignment (US-ASCII, ISO-8859-*, UTF-8), setting
1613
+ .B pb=0
1614
+ can improve compression slightly.
1615
+ For UTF-16 text,
1616
+ .B pb=1
1617
+ is a good choice.
1618
+ If the alignment is an odd number like 3 bytes,
1619
+ .B pb=0
1620
+ might be the best choice.
1621
+ .IP ""
1622
+ Even though the assumed alignment can be adjusted with
1623
+ .I pb
1624
+ and
1625
+ .IR lp ,
1626
+ LZMA1 and LZMA2 still slightly favor 16-byte alignment.
1627
+ It might be worth taking into account when designing file formats
1628
+ that are likely to be often compressed with LZMA1 or LZMA2.
1629
+ .TP
1630
+ .BI mf= mf
1631
+ Match finder has a major effect on encoder speed,
1632
+ memory usage, and compression ratio.
1633
+ Usually Hash Chain match finders are faster than Binary Tree
1634
+ match finders.
1635
+ The default depends on the
1636
+ .IR preset :
1637
+ 0 uses
1638
+ .BR hc3 ,
1639
+ 1\(en3
1640
+ use
1641
+ .BR hc4 ,
1642
+ and the rest use
1643
+ .BR bt4 .
1644
+ .IP ""
1645
+ The following match finders are supported.
1646
+ The memory usage formulas below are rough approximations,
1647
+ which are closest to the reality when
1648
+ .I dict
1649
+ is a power of two.
1650
+ .RS
1651
+ .TP
1652
+ .B hc3
1653
+ Hash Chain with 2- and 3-byte hashing
1654
+ .br
1655
+ Minimum value for
1656
+ .IR nice :
1657
+ 3
1658
+ .br
1659
+ Memory usage:
1660
+ .br
1661
+ .I dict
1662
+ * 7.5 (if
1663
+ .I dict
1664
+ <= 16 MiB);
1665
+ .br
1666
+ .I dict
1667
+ * 5.5 + 64 MiB (if
1668
+ .I dict
1669
+ > 16 MiB)
1670
+ .TP
1671
+ .B hc4
1672
+ Hash Chain with 2-, 3-, and 4-byte hashing
1673
+ .br
1674
+ Minimum value for
1675
+ .IR nice :
1676
+ 4
1677
+ .br
1678
+ Memory usage:
1679
+ .br
1680
+ .I dict
1681
+ * 7.5 (if
1682
+ .I dict
1683
+ <= 32 MiB);
1684
+ .br
1685
+ .I dict
1686
+ * 6.5 (if
1687
+ .I dict
1688
+ > 32 MiB)
1689
+ .TP
1690
+ .B bt2
1691
+ Binary Tree with 2-byte hashing
1692
+ .br
1693
+ Minimum value for
1694
+ .IR nice :
1695
+ 2
1696
+ .br
1697
+ Memory usage:
1698
+ .I dict
1699
+ * 9.5
1700
+ .TP
1701
+ .B bt3
1702
+ Binary Tree with 2- and 3-byte hashing
1703
+ .br
1704
+ Minimum value for
1705
+ .IR nice :
1706
+ 3
1707
+ .br
1708
+ Memory usage:
1709
+ .br
1710
+ .I dict
1711
+ * 11.5 (if
1712
+ .I dict
1713
+ <= 16 MiB);
1714
+ .br
1715
+ .I dict
1716
+ * 9.5 + 64 MiB (if
1717
+ .I dict
1718
+ > 16 MiB)
1719
+ .TP
1720
+ .B bt4
1721
+ Binary Tree with 2-, 3-, and 4-byte hashing
1722
+ .br
1723
+ Minimum value for
1724
+ .IR nice :
1725
+ 4
1726
+ .br
1727
+ Memory usage:
1728
+ .br
1729
+ .I dict
1730
+ * 11.5 (if
1731
+ .I dict
1732
+ <= 32 MiB);
1733
+ .br
1734
+ .I dict
1735
+ * 10.5 (if
1736
+ .I dict
1737
+ > 32 MiB)
1738
+ .RE
1739
+ .TP
1740
+ .BI mode= mode
1741
+ Compression
1742
+ .I mode
1743
+ specifies the method to analyze
1744
+ the data produced by the match finder.
1745
+ Supported
1746
+ .I modes
1747
+ are
1748
+ .B fast
1749
+ and
1750
+ .BR normal .
1751
+ The default is
1752
+ .B fast
1753
+ for
1754
+ .I presets
1755
+ 0\(en3 and
1756
+ .B normal
1757
+ for
1758
+ .I presets
1759
+ 4\(en9.
1760
+ .IP ""
1761
+ Usually
1762
+ .B fast
1763
+ is used with Hash Chain match finders and
1764
+ .B normal
1765
+ with Binary Tree match finders.
1766
+ This is also what the
1767
+ .I presets
1768
+ do.
1769
+ .TP
1770
+ .BI nice= nice
1771
+ Specify what is considered to be a nice length for a match.
1772
+ Once a match of at least
1773
+ .I nice
1774
+ bytes is found, the algorithm stops
1775
+ looking for possibly better matches.
1776
+ .IP ""
1777
+ .I Nice
1778
+ can be 2\(en273 bytes.
1779
+ Higher values tend to give better compression ratio
1780
+ at the expense of speed.
1781
+ The default depends on the
1782
+ .IR preset .
1783
+ .TP
1784
+ .BI depth= depth
1785
+ Specify the maximum search depth in the match finder.
1786
+ The default is the special value of 0,
1787
+ which makes the compressor determine a reasonable
1788
+ .I depth
1789
+ from
1790
+ .I mf
1791
+ and
1792
+ .IR nice .
1793
+ .IP ""
1794
+ Reasonable
1795
+ .I depth
1796
+ for Hash Chains is 4\(en100 and 16\(en1000 for Binary Trees.
1797
+ Using very high values for
1798
+ .I depth
1799
+ can make the encoder extremely slow with some files.
1800
+ Avoid setting the
1801
+ .I depth
1802
+ over 1000 unless you are prepared to interrupt
1803
+ the compression in case it is taking far too long.
1804
+ .RE
1805
+ .IP ""
1806
+ When decoding raw streams
1807
+ .RB ( \-\-format=raw ),
1808
+ LZMA2 needs only the dictionary
1809
+ .IR size .
1810
+ LZMA1 needs also
1811
+ .IR lc ,
1812
+ .IR lp ,
1813
+ and
1814
+ .IR pb .
1815
+ .TP
1816
+ \fB\-\-x86\fR[\fB=\fIoptions\fR]
1817
+ .PD 0
1818
+ .TP
1819
+ \fB\-\-arm\fR[\fB=\fIoptions\fR]
1820
+ .TP
1821
+ \fB\-\-armthumb\fR[\fB=\fIoptions\fR]
1822
+ .TP
1823
+ \fB\-\-arm64\fR[\fB=\fIoptions\fR]
1824
+ .TP
1825
+ \fB\-\-powerpc\fR[\fB=\fIoptions\fR]
1826
+ .TP
1827
+ \fB\-\-ia64\fR[\fB=\fIoptions\fR]
1828
+ .TP
1829
+ \fB\-\-sparc\fR[\fB=\fIoptions\fR]
1830
+ .TP
1831
+ \fB\-\-riscv\fR[\fB=\fIoptions\fR]
1832
+ .PD
1833
+ Add a branch/call/jump (BCJ) filter to the filter chain.
1834
+ These filters can be used only as a non-last filter
1835
+ in the filter chain.
1836
+ .IP ""
1837
+ A BCJ filter converts relative addresses in
1838
+ the machine code to their absolute counterparts.
1839
+ This doesn't change the size of the data
1840
+ but it increases redundancy,
1841
+ which can help LZMA2 to produce 0\(en15\ % smaller
1842
+ .B .xz
1843
+ file.
1844
+ The BCJ filters are always reversible,
1845
+ so using a BCJ filter for wrong type of data
1846
+ doesn't cause any data loss, although it may make
1847
+ the compression ratio slightly worse.
1848
+ The BCJ filters are very fast and
1849
+ use an insignificant amount of memory.
1850
+ .IP ""
1851
+ These BCJ filters have known problems related to
1852
+ the compression ratio:
1853
+ .RS
1854
+ .IP \(bu 3
1855
+ Some types of files containing executable code
1856
+ (for example, object files, static libraries, and Linux kernel modules)
1857
+ have the addresses in the instructions filled with filler values.
1858
+ These BCJ filters will still do the address conversion,
1859
+ which will make the compression worse with these files.
1860
+ .IP \(bu 3
1861
+ If a BCJ filter is applied on an archive,
1862
+ it is possible that it makes the compression ratio
1863
+ worse than not using a BCJ filter.
1864
+ For example, if there are similar or even identical executables
1865
+ then filtering will likely make the files less similar
1866
+ and thus compression is worse.
1867
+ The contents of non-executable files in the same archive can matter too.
1868
+ In practice one has to try with and without a BCJ filter to see
1869
+ which is better in each situation.
1870
+ .RE
1871
+ .IP ""
1872
+ Different instruction sets have different alignment:
1873
+ the executable file must be aligned to a multiple of
1874
+ this value in the input data to make the filter work.
1875
+ .RS
1876
+ .RS
1877
+ .PP
1878
+ .TS
1879
+ tab(;);
1880
+ l n l
1881
+ l n l.
1882
+ Filter;Alignment;Notes
1883
+ x86;1;32-bit or 64-bit x86
1884
+ ARM;4;
1885
+ ARM-Thumb;2;
1886
+ ARM64;4;4096-byte alignment is best
1887
+ PowerPC;4;Big endian only
1888
+ IA-64;16;Itanium
1889
+ SPARC;4;
1890
+ RISC-V;2;
1891
+ .TE
1892
+ .RE
1893
+ .RE
1894
+ .IP ""
1895
+ Since the BCJ-filtered data is usually compressed with LZMA2,
1896
+ the compression ratio may be improved slightly if
1897
+ the LZMA2 options are set to match the
1898
+ alignment of the selected BCJ filter.
1899
+ Examples:
1900
+ .RS
1901
+ .IP \(bu 3
1902
+ IA-64 filter has 16-byte alignment so
1903
+ .B pb=4,lp=4,lc=0
1904
+ is good
1905
+ with LZMA2 (2^4=16).
1906
+ .IP \(bu 3
1907
+ RISC-V code has 2-byte or 4-byte alignment
1908
+ depending on whether the file contains
1909
+ 16-bit compressed instructions (the C extension).
1910
+ When 16-bit instructions are used,
1911
+ .B pb=2,lp=1,lc=3
1912
+ or
1913
+ .B pb=1,lp=1,lc=3
1914
+ is good.
1915
+ When 16-bit instructions aren't present,
1916
+ .B pb=2,lp=2,lc=2
1917
+ is the best.
1918
+ .B readelf \-h
1919
+ can be used to check if "RVC"
1920
+ appears on the "Flags" line.
1921
+ .IP \(bu 3
1922
+ ARM64 is always 4-byte aligned so
1923
+ .B pb=2,lp=2,lc=2
1924
+ is the best.
1925
+ .IP \(bu 3
1926
+ The x86 filter is an exception.
1927
+ It's usually good to stick to LZMA2's defaults
1928
+ .RB ( pb=2,lp=0,lc=3 )
1929
+ when compressing x86 executables.
1930
+ .RE
1931
+ .IP ""
1932
+ All BCJ filters support the same
1933
+ .IR options :
1934
+ .RS
1935
+ .TP
1936
+ .BI start= offset
1937
+ Specify the start
1938
+ .I offset
1939
+ that is used when converting between relative
1940
+ and absolute addresses.
1941
+ The
1942
+ .I offset
1943
+ must be a multiple of the alignment of the filter
1944
+ (see the table above).
1945
+ The default is zero.
1946
+ In practice, the default is good; specifying a custom
1947
+ .I offset
1948
+ is almost never useful.
1949
+ .RE
1950
+ .TP
1951
+ \fB\-\-delta\fR[\fB=\fIoptions\fR]
1952
+ Add the Delta filter to the filter chain.
1953
+ The Delta filter can be only used as a non-last filter
1954
+ in the filter chain.
1955
+ .IP ""
1956
+ Currently only simple byte-wise delta calculation is supported.
1957
+ It can be useful when compressing, for example, uncompressed bitmap images
1958
+ or uncompressed PCM audio.
1959
+ However, special purpose algorithms may give significantly better
1960
+ results than Delta + LZMA2.
1961
+ This is true especially with audio,
1962
+ which compresses faster and better, for example, with
1963
+ .BR flac (1).
1964
+ .IP ""
1965
+ Supported
1966
+ .IR options :
1967
+ .RS
1968
+ .TP
1969
+ .BI dist= distance
1970
+ Specify the
1971
+ .I distance
1972
+ of the delta calculation in bytes.
1973
+ .I distance
1974
+ must be 1\(en256.
1975
+ The default is 1.
1976
+ .IP ""
1977
+ For example, with
1978
+ .B dist=2
1979
+ and eight-byte input A1 B1 A2 B3 A3 B5 A4 B7, the output will be
1980
+ A1 B1 01 02 01 02 01 02.
1981
+ .RE
1982
+ .
1983
+ .SS "Other options"
1984
+ .TP
1985
+ .BR \-q ", " \-\-quiet
1986
+ Suppress warnings and notices.
1987
+ Specify this twice to suppress errors too.
1988
+ This option has no effect on the exit status.
1989
+ That is, even if a warning was suppressed,
1990
+ the exit status to indicate a warning is still used.
1991
+ .TP
1992
+ .BR \-v ", " \-\-verbose
1993
+ Be verbose.
1994
+ If standard error is connected to a terminal,
1995
+ .B xz
1996
+ will display a progress indicator.
1997
+ Specifying
1998
+ .B \-\-verbose
1999
+ twice will give even more verbose output.
2000
+ .IP ""
2001
+ The progress indicator shows the following information:
2002
+ .RS
2003
+ .IP \(bu 3
2004
+ Completion percentage is shown
2005
+ if the size of the input file is known.
2006
+ That is, the percentage cannot be shown in pipes.
2007
+ .IP \(bu 3
2008
+ Amount of compressed data produced (compressing)
2009
+ or consumed (decompressing).
2010
+ .IP \(bu 3
2011
+ Amount of uncompressed data consumed (compressing)
2012
+ or produced (decompressing).
2013
+ .IP \(bu 3
2014
+ Compression ratio, which is calculated by dividing
2015
+ the amount of compressed data processed so far by
2016
+ the amount of uncompressed data processed so far.
2017
+ .IP \(bu 3
2018
+ Compression or decompression speed.
2019
+ This is measured as the amount of uncompressed data consumed
2020
+ (compression) or produced (decompression) per second.
2021
+ It is shown after a few seconds have passed since
2022
+ .B xz
2023
+ started processing the file.
2024
+ .IP \(bu 3
2025
+ Elapsed time in the format M:SS or H:MM:SS.
2026
+ .IP \(bu 3
2027
+ Estimated remaining time is shown
2028
+ only when the size of the input file is
2029
+ known and a couple of seconds have already passed since
2030
+ .B xz
2031
+ started processing the file.
2032
+ The time is shown in a less precise format which
2033
+ never has any colons, for example, 2 min 30 s.
2034
+ .RE
2035
+ .IP ""
2036
+ When standard error is not a terminal,
2037
+ .B \-\-verbose
2038
+ will make
2039
+ .B xz
2040
+ print the filename, compressed size, uncompressed size,
2041
+ compression ratio, and possibly also the speed and elapsed time
2042
+ on a single line to standard error after compressing or
2043
+ decompressing the file.
2044
+ The speed and elapsed time are included only when
2045
+ the operation took at least a few seconds.
2046
+ If the operation didn't finish, for example, due to user interruption,
2047
+ also the completion percentage is printed
2048
+ if the size of the input file is known.
2049
+ .TP
2050
+ .BR \-Q ", " \-\-no\-warn
2051
+ Don't set the exit status to 2
2052
+ even if a condition worth a warning was detected.
2053
+ This option doesn't affect the verbosity level, thus both
2054
+ .B \-\-quiet
2055
+ and
2056
+ .B \-\-no\-warn
2057
+ have to be used to not display warnings and
2058
+ to not alter the exit status.
2059
+ .TP
2060
+ .B \-\-robot
2061
+ Print messages in a machine-parsable format.
2062
+ This is intended to ease writing frontends that want to use
2063
+ .B xz
2064
+ instead of liblzma, which may be the case with various scripts.
2065
+ The output with this option enabled is meant to be stable across
2066
+ .B xz
2067
+ releases.
2068
+ See the section
2069
+ .B "ROBOT MODE"
2070
+ for details.
2071
+ .TP
2072
+ .B \-\-info\-memory
2073
+ Display, in human-readable format, how much physical memory (RAM)
2074
+ and how many processor threads
2075
+ .B xz
2076
+ thinks the system has and the memory usage limits for compression
2077
+ and decompression, and exit successfully.
2078
+ .TP
2079
+ .BR \-h ", " \-\-help
2080
+ Display a help message describing the most commonly used options,
2081
+ and exit successfully.
2082
+ .TP
2083
+ .BR \-H ", " \-\-long\-help
2084
+ Display a help message describing all features of
2085
+ .BR xz ,
2086
+ and exit successfully
2087
+ .TP
2088
+ .BR \-V ", " \-\-version
2089
+ Display the version number of
2090
+ .B xz
2091
+ and liblzma in human readable format.
2092
+ To get machine-parsable output, specify
2093
+ .B \-\-robot
2094
+ before
2095
+ .BR \-\-version .
2096
+ .
2097
+ .SH "ROBOT MODE"
2098
+ The robot mode is activated with the
2099
+ .B \-\-robot
2100
+ option.
2101
+ It makes the output of
2102
+ .B xz
2103
+ easier to parse by other programs.
2104
+ Currently
2105
+ .B \-\-robot
2106
+ is supported only together with
2107
+ .BR \-\-list ,
2108
+ .BR \-\-filters\-help ,
2109
+ .BR \-\-info\-memory ,
2110
+ and
2111
+ .BR \-\-version .
2112
+ It will be supported for compression and
2113
+ decompression in the future.
2114
+ .
2115
+ .SS "List mode"
2116
+ .B "xz \-\-robot \-\-list"
2117
+ uses tab-separated output.
2118
+ The first column of every line has a string
2119
+ that indicates the type of the information found on that line:
2120
+ .TP
2121
+ .B name
2122
+ This is always the first line when starting to list a file.
2123
+ The second column on the line is the filename.
2124
+ .TP
2125
+ .B file
2126
+ This line contains overall information about the
2127
+ .B .xz
2128
+ file.
2129
+ This line is always printed after the
2130
+ .B name
2131
+ line.
2132
+ .TP
2133
+ .B stream
2134
+ This line type is used only when
2135
+ .B \-\-verbose
2136
+ was specified.
2137
+ There are as many
2138
+ .B stream
2139
+ lines as there are streams in the
2140
+ .B .xz
2141
+ file.
2142
+ .TP
2143
+ .B block
2144
+ This line type is used only when
2145
+ .B \-\-verbose
2146
+ was specified.
2147
+ There are as many
2148
+ .B block
2149
+ lines as there are blocks in the
2150
+ .B .xz
2151
+ file.
2152
+ The
2153
+ .B block
2154
+ lines are shown after all the
2155
+ .B stream
2156
+ lines; different line types are not interleaved.
2157
+ .TP
2158
+ .B summary
2159
+ This line type is used only when
2160
+ .B \-\-verbose
2161
+ was specified twice.
2162
+ This line is printed after all
2163
+ .B block
2164
+ lines.
2165
+ Like the
2166
+ .B file
2167
+ line, the
2168
+ .B summary
2169
+ line contains overall information about the
2170
+ .B .xz
2171
+ file.
2172
+ .TP
2173
+ .B totals
2174
+ This line is always the very last line of the list output.
2175
+ It shows the total counts and sizes.
2176
+ .PP
2177
+ The columns of the
2178
+ .B file
2179
+ lines:
2180
+ .PD 0
2181
+ .RS
2182
+ .IP 2. 4
2183
+ Number of streams in the file
2184
+ .IP 3. 4
2185
+ Total number of blocks in the stream(s)
2186
+ .IP 4. 4
2187
+ Compressed size of the file
2188
+ .IP 5. 4
2189
+ Uncompressed size of the file
2190
+ .IP 6. 4
2191
+ Compression ratio, for example,
2192
+ .BR 0.123 .
2193
+ If ratio is over 9.999, three dashes
2194
+ .RB ( \-\-\- )
2195
+ are displayed instead of the ratio.
2196
+ .IP 7. 4
2197
+ Comma-separated list of integrity check names.
2198
+ The following strings are used for the known check types:
2199
+ .BR None ,
2200
+ .BR CRC32 ,
2201
+ .BR CRC64 ,
2202
+ and
2203
+ .BR SHA\-256 .
2204
+ For unknown check types,
2205
+ .BI Unknown\- N
2206
+ is used, where
2207
+ .I N
2208
+ is the Check ID as a decimal number (one or two digits).
2209
+ .IP 8. 4
2210
+ Total size of stream padding in the file
2211
+ .RE
2212
+ .PD
2213
+ .PP
2214
+ The columns of the
2215
+ .B stream
2216
+ lines:
2217
+ .PD 0
2218
+ .RS
2219
+ .IP 2. 4
2220
+ Stream number (the first stream is 1)
2221
+ .IP 3. 4
2222
+ Number of blocks in the stream
2223
+ .IP 4. 4
2224
+ Compressed start offset
2225
+ .IP 5. 4
2226
+ Uncompressed start offset
2227
+ .IP 6. 4
2228
+ Compressed size (does not include stream padding)
2229
+ .IP 7. 4
2230
+ Uncompressed size
2231
+ .IP 8. 4
2232
+ Compression ratio
2233
+ .IP 9. 4
2234
+ Name of the integrity check
2235
+ .IP 10. 4
2236
+ Size of stream padding
2237
+ .RE
2238
+ .PD
2239
+ .PP
2240
+ The columns of the
2241
+ .B block
2242
+ lines:
2243
+ .PD 0
2244
+ .RS
2245
+ .IP 2. 4
2246
+ Number of the stream containing this block
2247
+ .IP 3. 4
2248
+ Block number relative to the beginning of the stream
2249
+ (the first block is 1)
2250
+ .IP 4. 4
2251
+ Block number relative to the beginning of the file
2252
+ .IP 5. 4
2253
+ Compressed start offset relative to the beginning of the file
2254
+ .IP 6. 4
2255
+ Uncompressed start offset relative to the beginning of the file
2256
+ .IP 7. 4
2257
+ Total compressed size of the block (includes headers)
2258
+ .IP 8. 4
2259
+ Uncompressed size
2260
+ .IP 9. 4
2261
+ Compression ratio
2262
+ .IP 10. 4
2263
+ Name of the integrity check
2264
+ .RE
2265
+ .PD
2266
+ .PP
2267
+ If
2268
+ .B \-\-verbose
2269
+ was specified twice, additional columns are included on the
2270
+ .B block
2271
+ lines.
2272
+ These are not displayed with a single
2273
+ .BR \-\-verbose ,
2274
+ because getting this information requires many seeks
2275
+ and can thus be slow:
2276
+ .PD 0
2277
+ .RS
2278
+ .IP 11. 4
2279
+ Value of the integrity check in hexadecimal
2280
+ .IP 12. 4
2281
+ Block header size
2282
+ .IP 13. 4
2283
+ Block flags:
2284
+ .B c
2285
+ indicates that compressed size is present, and
2286
+ .B u
2287
+ indicates that uncompressed size is present.
2288
+ If the flag is not set, a dash
2289
+ .RB ( \- )
2290
+ is shown instead to keep the string length fixed.
2291
+ New flags may be added to the end of the string in the future.
2292
+ .IP 14. 4
2293
+ Size of the actual compressed data in the block (this excludes
2294
+ the block header, block padding, and check fields)
2295
+ .IP 15. 4
2296
+ Amount of memory (in bytes) required to decompress
2297
+ this block with this
2298
+ .B xz
2299
+ version
2300
+ .IP 16. 4
2301
+ Filter chain.
2302
+ Note that most of the options used at compression time
2303
+ cannot be known, because only the options
2304
+ that are needed for decompression are stored in the
2305
+ .B .xz
2306
+ headers.
2307
+ .RE
2308
+ .PD
2309
+ .PP
2310
+ The columns of the
2311
+ .B summary
2312
+ lines:
2313
+ .PD 0
2314
+ .RS
2315
+ .IP 2. 4
2316
+ Amount of memory (in bytes) required to decompress
2317
+ this file with this
2318
+ .B xz
2319
+ version
2320
+ .IP 3. 4
2321
+ .B yes
2322
+ or
2323
+ .B no
2324
+ indicating if all block headers have both compressed size and
2325
+ uncompressed size stored in them
2326
+ .PP
2327
+ .I Since
2328
+ .B xz
2329
+ .I 5.1.2alpha:
2330
+ .IP 4. 4
2331
+ Minimum
2332
+ .B xz
2333
+ version required to decompress the file
2334
+ .RE
2335
+ .PD
2336
+ .PP
2337
+ The columns of the
2338
+ .B totals
2339
+ line:
2340
+ .PD 0
2341
+ .RS
2342
+ .IP 2. 4
2343
+ Number of streams
2344
+ .IP 3. 4
2345
+ Number of blocks
2346
+ .IP 4. 4
2347
+ Compressed size
2348
+ .IP 5. 4
2349
+ Uncompressed size
2350
+ .IP 6. 4
2351
+ Average compression ratio
2352
+ .IP 7. 4
2353
+ Comma-separated list of integrity check names
2354
+ that were present in the files
2355
+ .IP 8. 4
2356
+ Stream padding size
2357
+ .IP 9. 4
2358
+ Number of files.
2359
+ This is here to
2360
+ keep the order of the earlier columns the same as on
2361
+ .B file
2362
+ lines.
2363
+ .PD
2364
+ .RE
2365
+ .PP
2366
+ If
2367
+ .B \-\-verbose
2368
+ was specified twice, additional columns are included on the
2369
+ .B totals
2370
+ line:
2371
+ .PD 0
2372
+ .RS
2373
+ .IP 10. 4
2374
+ Maximum amount of memory (in bytes) required to decompress
2375
+ the files with this
2376
+ .B xz
2377
+ version
2378
+ .IP 11. 4
2379
+ .B yes
2380
+ or
2381
+ .B no
2382
+ indicating if all block headers have both compressed size and
2383
+ uncompressed size stored in them
2384
+ .PP
2385
+ .I Since
2386
+ .B xz
2387
+ .I 5.1.2alpha:
2388
+ .IP 12. 4
2389
+ Minimum
2390
+ .B xz
2391
+ version required to decompress the file
2392
+ .RE
2393
+ .PD
2394
+ .PP
2395
+ Future versions may add new line types and
2396
+ new columns can be added to the existing line types,
2397
+ but the existing columns won't be changed.
2398
+ .
2399
+ .SS "Filters help"
2400
+ .B "xz \-\-robot \-\-filters-help"
2401
+ prints the supported filters in the following format:
2402
+ .PP
2403
+ \fIfilter\fB:\fIoption\fB=<\fIvalue\fB>,\fIoption\fB=<\fIvalue\fB>\fR...
2404
+ .TP
2405
+ .I filter
2406
+ Name of the filter
2407
+ .TP
2408
+ .I option
2409
+ Name of a filter specific option
2410
+ .TP
2411
+ .I value
2412
+ Numeric
2413
+ .I value
2414
+ ranges appear as
2415
+ \fB<\fImin\fB\-\fImax\fB>\fR.
2416
+ String
2417
+ .I value
2418
+ choices are shown within
2419
+ .B "< >"
2420
+ and separated by a
2421
+ .B |
2422
+ character.
2423
+ .PP
2424
+ Each filter is printed on its own line.
2425
+ .
2426
+ .SS "Memory limit information"
2427
+ .B "xz \-\-robot \-\-info\-memory"
2428
+ prints a single line with multiple tab-separated columns:
2429
+ .IP 1. 4
2430
+ Total amount of physical memory (RAM) in bytes.
2431
+ .IP 2. 4
2432
+ Memory usage limit for compression in bytes
2433
+ .RB ( \-\-memlimit\-compress ).
2434
+ A special value of
2435
+ .B 0
2436
+ indicates the default setting
2437
+ which for single-threaded mode is the same as no limit.
2438
+ .IP 3. 4
2439
+ Memory usage limit for decompression in bytes
2440
+ .RB ( \-\-memlimit\-decompress ).
2441
+ A special value of
2442
+ .B 0
2443
+ indicates the default setting
2444
+ which for single-threaded mode is the same as no limit.
2445
+ .IP 4. 4
2446
+ Since
2447
+ .B xz
2448
+ 5.3.4alpha:
2449
+ Memory usage for multi-threaded decompression in bytes
2450
+ .RB ( \-\-memlimit\-mt\-decompress ).
2451
+ This is never zero because a system-specific default value
2452
+ shown in the column 5
2453
+ is used if no limit has been specified explicitly.
2454
+ This is also never greater than the value in the column 3
2455
+ even if a larger value has been specified with
2456
+ .BR \-\-memlimit\-mt\-decompress .
2457
+ .IP 5. 4
2458
+ Since
2459
+ .B xz
2460
+ 5.3.4alpha:
2461
+ A system-specific default memory usage limit
2462
+ that is used to limit the number of threads
2463
+ when compressing with an automatic number of threads
2464
+ .RB ( \-\-threads=0 )
2465
+ and no memory usage limit has been specified
2466
+ .RB ( \-\-memlimit\-compress ).
2467
+ This is also used as the default value for
2468
+ .BR \-\-memlimit\-mt\-decompress .
2469
+ .IP 6. 4
2470
+ Since
2471
+ .B xz
2472
+ 5.3.4alpha:
2473
+ Number of available processor threads.
2474
+ .PP
2475
+ In the future, the output of
2476
+ .B "xz \-\-robot \-\-info\-memory"
2477
+ may have more columns, but never more than a single line.
2478
+ .
2479
+ .SS Version
2480
+ .B "xz \-\-robot \-\-version"
2481
+ prints the version number of
2482
+ .B xz
2483
+ and liblzma in the following format:
2484
+ .PP
2485
+ .BI XZ_VERSION= XYYYZZZS
2486
+ .br
2487
+ .BI LIBLZMA_VERSION= XYYYZZZS
2488
+ .TP
2489
+ .I X
2490
+ Major version.
2491
+ .TP
2492
+ .I YYY
2493
+ Minor version.
2494
+ Even numbers are stable.
2495
+ Odd numbers are alpha or beta versions.
2496
+ .TP
2497
+ .I ZZZ
2498
+ Patch level for stable releases or
2499
+ just a counter for development releases.
2500
+ .TP
2501
+ .I S
2502
+ Stability.
2503
+ 0 is alpha, 1 is beta, and 2 is stable.
2504
+ .I S
2505
+ should be always 2 when
2506
+ .I YYY
2507
+ is even.
2508
+ .PP
2509
+ .I XYYYZZZS
2510
+ are the same on both lines if
2511
+ .B xz
2512
+ and liblzma are from the same XZ Utils release.
2513
+ .PP
2514
+ Examples: 4.999.9beta is
2515
+ .B 49990091
2516
+ and
2517
+ 5.0.0 is
2518
+ .BR 50000002 .
2519
+ .
2520
+ .SH "EXIT STATUS"
2521
+ .TP
2522
+ .B 0
2523
+ All is good.
2524
+ .TP
2525
+ .B 1
2526
+ An error occurred.
2527
+ .TP
2528
+ .B 2
2529
+ Something worth a warning occurred,
2530
+ but no actual errors occurred.
2531
+ .PP
2532
+ Notices (not warnings or errors) printed on standard error
2533
+ don't affect the exit status.
2534
+ .
2535
+ .SH ENVIRONMENT
2536
+ .B xz
2537
+ parses space-separated lists of options
2538
+ from the environment variables
2539
+ .B XZ_DEFAULTS
2540
+ and
2541
+ .BR XZ_OPT ,
2542
+ in this order, before parsing the options from the command line.
2543
+ Note that only options are parsed from the environment variables;
2544
+ all non-options are silently ignored.
2545
+ Parsing is done with
2546
+ .BR getopt_long (3)
2547
+ which is used also for the command line arguments.
2548
+ .TP
2549
+ .B XZ_DEFAULTS
2550
+ User-specific or system-wide default options.
2551
+ Typically this is set in a shell initialization script to enable
2552
+ .BR xz 's
2553
+ memory usage limiter by default.
2554
+ Excluding shell initialization scripts
2555
+ and similar special cases, scripts must never set or unset
2556
+ .BR XZ_DEFAULTS .
2557
+ .TP
2558
+ .B XZ_OPT
2559
+ This is for passing options to
2560
+ .B xz
2561
+ when it is not possible to set the options directly on the
2562
+ .B xz
2563
+ command line.
2564
+ This is the case when
2565
+ .B xz
2566
+ is run by a script or tool, for example, GNU
2567
+ .BR tar (1):
2568
+ .RS
2569
+ .RS
2570
+ .PP
2571
+ .nf
2572
+ .ft CR
2573
+ XZ_OPT=\-2v tar caf foo.tar.xz foo
2574
+ .ft R
2575
+ .fi
2576
+ .RE
2577
+ .RE
2578
+ .IP ""
2579
+ Scripts may use
2580
+ .BR XZ_OPT ,
2581
+ for example, to set script-specific default compression options.
2582
+ It is still recommended to allow users to override
2583
+ .B XZ_OPT
2584
+ if that is reasonable.
2585
+ For example, in
2586
+ .BR sh (1)
2587
+ scripts one may use something like this:
2588
+ .RS
2589
+ .RS
2590
+ .PP
2591
+ .nf
2592
+ .ft CR
2593
+ XZ_OPT=${XZ_OPT\-"\-7e"}
2594
+ export XZ_OPT
2595
+ .ft R
2596
+ .fi
2597
+ .RE
2598
+ .RE
2599
+ .
2600
+ .SH "LZMA UTILS COMPATIBILITY"
2601
+ The command line syntax of
2602
+ .B xz
2603
+ is practically a superset of
2604
+ .BR lzma ,
2605
+ .BR unlzma ,
2606
+ and
2607
+ .B lzcat
2608
+ as found from LZMA Utils 4.32.x.
2609
+ In most cases, it is possible to replace
2610
+ LZMA Utils with XZ Utils without breaking existing scripts.
2611
+ There are some incompatibilities though,
2612
+ which may sometimes cause problems.
2613
+ .
2614
+ .SS "Compression preset levels"
2615
+ The numbering of the compression level presets is not identical in
2616
+ .B xz
2617
+ and LZMA Utils.
2618
+ The most important difference is how dictionary sizes
2619
+ are mapped to different presets.
2620
+ Dictionary size is roughly equal to the decompressor memory usage.
2621
+ .RS
2622
+ .PP
2623
+ .TS
2624
+ tab(;);
2625
+ c c c
2626
+ c n n.
2627
+ Level;xz;LZMA Utils
2628
+ \-0;256 KiB;N/A
2629
+ \-1;1 MiB;64 KiB
2630
+ \-2;2 MiB;1 MiB
2631
+ \-3;4 MiB;512 KiB
2632
+ \-4;4 MiB;1 MiB
2633
+ \-5;8 MiB;2 MiB
2634
+ \-6;8 MiB;4 MiB
2635
+ \-7;16 MiB;8 MiB
2636
+ \-8;32 MiB;16 MiB
2637
+ \-9;64 MiB;32 MiB
2638
+ .TE
2639
+ .RE
2640
+ .PP
2641
+ The dictionary size differences affect
2642
+ the compressor memory usage too,
2643
+ but there are some other differences between
2644
+ LZMA Utils and XZ Utils, which
2645
+ make the difference even bigger:
2646
+ .RS
2647
+ .PP
2648
+ .TS
2649
+ tab(;);
2650
+ c c c
2651
+ c n n.
2652
+ Level;xz;LZMA Utils 4.32.x
2653
+ \-0;3 MiB;N/A
2654
+ \-1;9 MiB;2 MiB
2655
+ \-2;17 MiB;12 MiB
2656
+ \-3;32 MiB;12 MiB
2657
+ \-4;48 MiB;16 MiB
2658
+ \-5;94 MiB;26 MiB
2659
+ \-6;94 MiB;45 MiB
2660
+ \-7;186 MiB;83 MiB
2661
+ \-8;370 MiB;159 MiB
2662
+ \-9;674 MiB;311 MiB
2663
+ .TE
2664
+ .RE
2665
+ .PP
2666
+ The default preset level in LZMA Utils is
2667
+ .B \-7
2668
+ while in XZ Utils it is
2669
+ .BR \-6 ,
2670
+ so both use an 8 MiB dictionary by default.
2671
+ .
2672
+ .SS "Streamed vs. non-streamed .lzma files"
2673
+ The uncompressed size of the file can be stored in the
2674
+ .B .lzma
2675
+ header.
2676
+ LZMA Utils does that when compressing regular files.
2677
+ The alternative is to mark that uncompressed size is unknown
2678
+ and use end-of-payload marker to indicate
2679
+ where the decompressor should stop.
2680
+ LZMA Utils uses this method when uncompressed size isn't known,
2681
+ which is the case, for example, in pipes.
2682
+ .PP
2683
+ .B xz
2684
+ supports decompressing
2685
+ .B .lzma
2686
+ files with or without end-of-payload marker, but all
2687
+ .B .lzma
2688
+ files created by
2689
+ .B xz
2690
+ will use end-of-payload marker and have uncompressed size
2691
+ marked as unknown in the
2692
+ .B .lzma
2693
+ header.
2694
+ This may be a problem in some uncommon situations.
2695
+ For example, a
2696
+ .B .lzma
2697
+ decompressor in an embedded device might work
2698
+ only with files that have known uncompressed size.
2699
+ If you hit this problem, you need to use LZMA Utils
2700
+ or LZMA SDK to create
2701
+ .B .lzma
2702
+ files with known uncompressed size.
2703
+ .
2704
+ .SS "Unsupported .lzma files"
2705
+ The
2706
+ .B .lzma
2707
+ format allows
2708
+ .I lc
2709
+ values up to 8, and
2710
+ .I lp
2711
+ values up to 4.
2712
+ LZMA Utils can decompress files with any
2713
+ .I lc
2714
+ and
2715
+ .IR lp ,
2716
+ but always creates files with
2717
+ .B lc=3
2718
+ and
2719
+ .BR lp=0 .
2720
+ Creating files with other
2721
+ .I lc
2722
+ and
2723
+ .I lp
2724
+ is possible with
2725
+ .B xz
2726
+ and with LZMA SDK.
2727
+ .PP
2728
+ The implementation of the LZMA1 filter in liblzma
2729
+ requires that the sum of
2730
+ .I lc
2731
+ and
2732
+ .I lp
2733
+ must not exceed 4.
2734
+ Thus,
2735
+ .B .lzma
2736
+ files, which exceed this limitation, cannot be decompressed with
2737
+ .BR xz .
2738
+ .PP
2739
+ LZMA Utils creates only
2740
+ .B .lzma
2741
+ files which have a dictionary size of
2742
+ .RI "2^" n
2743
+ (a power of 2) but accepts files with any dictionary size.
2744
+ liblzma accepts only
2745
+ .B .lzma
2746
+ files which have a dictionary size of
2747
+ .RI "2^" n
2748
+ or
2749
+ .RI "2^" n " + 2^(" n "\-1)."
2750
+ This is to decrease false positives when detecting
2751
+ .B .lzma
2752
+ files.
2753
+ .PP
2754
+ These limitations shouldn't be a problem in practice,
2755
+ since practically all
2756
+ .B .lzma
2757
+ files have been compressed with settings that liblzma will accept.
2758
+ .
2759
+ .SS "Trailing garbage"
2760
+ When decompressing,
2761
+ LZMA Utils silently ignore everything after the first
2762
+ .B .lzma
2763
+ stream.
2764
+ In most situations, this is a bug.
2765
+ This also means that LZMA Utils
2766
+ don't support decompressing concatenated
2767
+ .B .lzma
2768
+ files.
2769
+ .PP
2770
+ If there is data left after the first
2771
+ .B .lzma
2772
+ stream,
2773
+ .B xz
2774
+ considers the file to be corrupt unless
2775
+ .B \-\-single\-stream
2776
+ was used.
2777
+ This may break obscure scripts which have
2778
+ assumed that trailing garbage is ignored.
2779
+ .
2780
+ .SH NOTES
2781
+ .
2782
+ .SS "Compressed output may vary"
2783
+ The exact compressed output produced from
2784
+ the same uncompressed input file
2785
+ may vary between XZ Utils versions even if
2786
+ compression options are identical.
2787
+ This is because the encoder can be improved
2788
+ (faster or better compression)
2789
+ without affecting the file format.
2790
+ The output can vary even between different
2791
+ builds of the same XZ Utils version,
2792
+ if different build options are used.
2793
+ .PP
2794
+ The above means that once
2795
+ .B \-\-rsyncable
2796
+ has been implemented,
2797
+ the resulting files won't necessarily be rsyncable
2798
+ unless both old and new files have been compressed
2799
+ with the same xz version.
2800
+ This problem can be fixed if a part of the encoder
2801
+ implementation is frozen to keep rsyncable output
2802
+ stable across xz versions.
2803
+ .
2804
+ .SS "Embedded .xz decompressors"
2805
+ Embedded
2806
+ .B .xz
2807
+ decompressor implementations like XZ Embedded don't necessarily
2808
+ support files created with integrity
2809
+ .I check
2810
+ types other than
2811
+ .B none
2812
+ and
2813
+ .BR crc32 .
2814
+ Since the default is
2815
+ .BR \-\-check=crc64 ,
2816
+ you must use
2817
+ .B \-\-check=none
2818
+ or
2819
+ .B \-\-check=crc32
2820
+ when creating files for embedded systems.
2821
+ .PP
2822
+ Outside embedded systems, all
2823
+ .B .xz
2824
+ format decompressors support all the
2825
+ .I check
2826
+ types, or at least are able to decompress
2827
+ the file without verifying the
2828
+ integrity check if the particular
2829
+ .I check
2830
+ is not supported.
2831
+ .PP
2832
+ XZ Embedded supports BCJ filters,
2833
+ but only with the default start offset.
2834
+ .
2835
+ .SH EXAMPLES
2836
+ .
2837
+ .SS Basics
2838
+ Compress the file
2839
+ .I foo
2840
+ into
2841
+ .I foo.xz
2842
+ using the default compression level
2843
+ .RB ( \-6 ),
2844
+ and remove
2845
+ .I foo
2846
+ if compression is successful:
2847
+ .RS
2848
+ .PP
2849
+ .nf
2850
+ .ft CR
2851
+ xz foo
2852
+ .ft R
2853
+ .fi
2854
+ .RE
2855
+ .PP
2856
+ Decompress
2857
+ .I bar.xz
2858
+ into
2859
+ .I bar
2860
+ and don't remove
2861
+ .I bar.xz
2862
+ even if decompression is successful:
2863
+ .RS
2864
+ .PP
2865
+ .nf
2866
+ .ft CR
2867
+ xz \-dk bar.xz
2868
+ .ft R
2869
+ .fi
2870
+ .RE
2871
+ .PP
2872
+ Create
2873
+ .I baz.tar.xz
2874
+ with the preset
2875
+ .B \-4e
2876
+ .RB ( "\-4 \-\-extreme" ),
2877
+ which is slower than the default
2878
+ .BR \-6 ,
2879
+ but needs less memory for compression and decompression (48\ MiB
2880
+ and 5\ MiB, respectively):
2881
+ .RS
2882
+ .PP
2883
+ .nf
2884
+ .ft CR
2885
+ tar cf \- baz | xz \-4e > baz.tar.xz
2886
+ .ft R
2887
+ .fi
2888
+ .RE
2889
+ .PP
2890
+ A mix of compressed and uncompressed files can be decompressed
2891
+ to standard output with a single command:
2892
+ .RS
2893
+ .PP
2894
+ .nf
2895
+ .ft CR
2896
+ xz \-dcf a.txt b.txt.xz c.txt d.txt.lzma > abcd.txt
2897
+ .ft R
2898
+ .fi
2899
+ .RE
2900
+ .
2901
+ .SS "Parallel compression of many files"
2902
+ On GNU and *BSD,
2903
+ .BR find (1)
2904
+ and
2905
+ .BR xargs (1)
2906
+ can be used to parallelize compression of many files:
2907
+ .RS
2908
+ .PP
2909
+ .nf
2910
+ .ft CR
2911
+ find . \-type f \e! \-name '*.xz' \-print0 \e
2912
+ | xargs \-0r \-P4 \-n16 xz \-T1
2913
+ .ft R
2914
+ .fi
2915
+ .RE
2916
+ .PP
2917
+ The
2918
+ .B \-P
2919
+ option to
2920
+ .BR xargs (1)
2921
+ sets the number of parallel
2922
+ .B xz
2923
+ processes.
2924
+ The best value for the
2925
+ .B \-n
2926
+ option depends on how many files there are to be compressed.
2927
+ If there are only a couple of files,
2928
+ the value should probably be 1;
2929
+ with tens of thousands of files,
2930
+ 100 or even more may be appropriate to reduce the number of
2931
+ .B xz
2932
+ processes that
2933
+ .BR xargs (1)
2934
+ will eventually create.
2935
+ .PP
2936
+ The option
2937
+ .B \-T1
2938
+ for
2939
+ .B xz
2940
+ is there to force it to single-threaded mode, because
2941
+ .BR xargs (1)
2942
+ is used to control the amount of parallelization.
2943
+ .
2944
+ .SS "Robot mode"
2945
+ Calculate how many bytes have been saved in total
2946
+ after compressing multiple files:
2947
+ .RS
2948
+ .PP
2949
+ .nf
2950
+ .ft CR
2951
+ xz \-\-robot \-\-list *.xz | awk '/^totals/{print $5\-$4}'
2952
+ .ft R
2953
+ .fi
2954
+ .RE
2955
+ .PP
2956
+ A script may want to know that it is using new enough
2957
+ .BR xz .
2958
+ The following
2959
+ .BR sh (1)
2960
+ script checks that the version number of the
2961
+ .B xz
2962
+ tool is at least 5.0.0.
2963
+ This method is compatible with old beta versions,
2964
+ which didn't support the
2965
+ .B \-\-robot
2966
+ option:
2967
+ .RS
2968
+ .PP
2969
+ .nf
2970
+ .ft CR
2971
+ if ! eval "$(xz \-\-robot \-\-version 2> /dev/null)" ||
2972
+ [ "$XZ_VERSION" \-lt 50000002 ]; then
2973
+ echo "Your xz is too old."
2974
+ fi
2975
+ unset XZ_VERSION LIBLZMA_VERSION
2976
+ .ft R
2977
+ .fi
2978
+ .RE
2979
+ .PP
2980
+ Set a memory usage limit for decompression using
2981
+ .BR XZ_OPT ,
2982
+ but if a limit has already been set, don't increase it:
2983
+ .RS
2984
+ .PP
2985
+ .nf
2986
+ .ft CR
2987
+ NEWLIM=$((123 << 20))\ \ # 123 MiB
2988
+ OLDLIM=$(xz \-\-robot \-\-info\-memory | cut \-f3)
2989
+ if [ $OLDLIM \-eq 0 \-o $OLDLIM \-gt $NEWLIM ]; then
2990
+ XZ_OPT="$XZ_OPT \-\-memlimit\-decompress=$NEWLIM"
2991
+ export XZ_OPT
2992
+ fi
2993
+ .ft R
2994
+ .fi
2995
+ .RE
2996
+ .
2997
+ .SS "Custom compressor filter chains"
2998
+ The simplest use for custom filter chains is
2999
+ customizing a LZMA2 preset.
3000
+ This can be useful,
3001
+ because the presets cover only a subset of the
3002
+ potentially useful combinations of compression settings.
3003
+ .PP
3004
+ The CompCPU columns of the tables
3005
+ from the descriptions of the options
3006
+ .BR "\-0" " ... " "\-9"
3007
+ and
3008
+ .B \-\-extreme
3009
+ are useful when customizing LZMA2 presets.
3010
+ Here are the relevant parts collected from those two tables:
3011
+ .RS
3012
+ .PP
3013
+ .TS
3014
+ tab(;);
3015
+ c c
3016
+ n n.
3017
+ Preset;CompCPU
3018
+ \-0;0
3019
+ \-1;1
3020
+ \-2;2
3021
+ \-3;3
3022
+ \-4;4
3023
+ \-5;5
3024
+ \-6;6
3025
+ \-5e;7
3026
+ \-6e;8
3027
+ .TE
3028
+ .RE
3029
+ .PP
3030
+ If you know that a file requires
3031
+ somewhat big dictionary (for example, 32\ MiB) to compress well,
3032
+ but you want to compress it quicker than
3033
+ .B "xz \-8"
3034
+ would do, a preset with a low CompCPU value (for example, 1)
3035
+ can be modified to use a bigger dictionary:
3036
+ .RS
3037
+ .PP
3038
+ .nf
3039
+ .ft CR
3040
+ xz \-\-lzma2=preset=1,dict=32MiB foo.tar
3041
+ .ft R
3042
+ .fi
3043
+ .RE
3044
+ .PP
3045
+ With certain files, the above command may be faster than
3046
+ .B "xz \-6"
3047
+ while compressing significantly better.
3048
+ However, it must be emphasized that only some files benefit from
3049
+ a big dictionary while keeping the CompCPU value low.
3050
+ The most obvious situation,
3051
+ where a big dictionary can help a lot,
3052
+ is an archive containing very similar files
3053
+ of at least a few megabytes each.
3054
+ The dictionary size has to be significantly bigger
3055
+ than any individual file to allow LZMA2 to take
3056
+ full advantage of the similarities between consecutive files.
3057
+ .PP
3058
+ If very high compressor and decompressor memory usage is fine,
3059
+ and the file being compressed is
3060
+ at least several hundred megabytes, it may be useful
3061
+ to use an even bigger dictionary than the 64 MiB that
3062
+ .B "xz \-9"
3063
+ would use:
3064
+ .RS
3065
+ .PP
3066
+ .nf
3067
+ .ft CR
3068
+ xz \-vv \-\-lzma2=dict=192MiB big_foo.tar
3069
+ .ft R
3070
+ .fi
3071
+ .RE
3072
+ .PP
3073
+ Using
3074
+ .B \-vv
3075
+ .RB ( "\-\-verbose \-\-verbose" )
3076
+ like in the above example can be useful
3077
+ to see the memory requirements
3078
+ of the compressor and decompressor.
3079
+ Remember that using a dictionary bigger than
3080
+ the size of the uncompressed file is waste of memory,
3081
+ so the above command isn't useful for small files.
3082
+ .PP
3083
+ Sometimes the compression time doesn't matter,
3084
+ but the decompressor memory usage has to be kept low, for example,
3085
+ to make it possible to decompress the file on an embedded system.
3086
+ The following command uses
3087
+ .B \-6e
3088
+ .RB ( "\-6 \-\-extreme" )
3089
+ as a base and sets the dictionary to only 64\ KiB.
3090
+ The resulting file can be decompressed with XZ Embedded
3091
+ (that's why there is
3092
+ .BR \-\-check=crc32 )
3093
+ using about 100\ KiB of memory.
3094
+ .RS
3095
+ .PP
3096
+ .nf
3097
+ .ft CR
3098
+ xz \-\-check=crc32 \-\-lzma2=preset=6e,dict=64KiB foo
3099
+ .ft R
3100
+ .fi
3101
+ .RE
3102
+ .PP
3103
+ If you want to squeeze out as many bytes as possible,
3104
+ adjusting the number of literal context bits
3105
+ .RI ( lc )
3106
+ and number of position bits
3107
+ .RI ( pb )
3108
+ can sometimes help.
3109
+ Adjusting the number of literal position bits
3110
+ .RI ( lp )
3111
+ might help too, but usually
3112
+ .I lc
3113
+ and
3114
+ .I pb
3115
+ are more important.
3116
+ For example, a source code archive contains mostly US-ASCII text,
3117
+ so something like the following might give
3118
+ slightly (like 0.1\ %) smaller file than
3119
+ .B "xz \-6e"
3120
+ (try also without
3121
+ .BR lc=4 ):
3122
+ .RS
3123
+ .PP
3124
+ .nf
3125
+ .ft CR
3126
+ xz \-\-lzma2=preset=6e,pb=0,lc=4 source_code.tar
3127
+ .ft R
3128
+ .fi
3129
+ .RE
3130
+ .PP
3131
+ Using another filter together with LZMA2 can improve
3132
+ compression with certain file types.
3133
+ For example, to compress a x86-32 or x86-64 shared library
3134
+ using the x86 BCJ filter:
3135
+ .RS
3136
+ .PP
3137
+ .nf
3138
+ .ft CR
3139
+ xz \-\-x86 \-\-lzma2 libfoo.so
3140
+ .ft R
3141
+ .fi
3142
+ .RE
3143
+ .PP
3144
+ Note that the order of the filter options is significant.
3145
+ If
3146
+ .B \-\-x86
3147
+ is specified after
3148
+ .BR \-\-lzma2 ,
3149
+ .B xz
3150
+ will give an error,
3151
+ because there cannot be any filter after LZMA2,
3152
+ and also because the x86 BCJ filter cannot be used
3153
+ as the last filter in the chain.
3154
+ .PP
3155
+ The Delta filter together with LZMA2
3156
+ can give good results with bitmap images.
3157
+ It should usually beat PNG,
3158
+ which has a few more advanced filters than simple
3159
+ delta but uses Deflate for the actual compression.
3160
+ .PP
3161
+ The image has to be saved in uncompressed format,
3162
+ for example, as uncompressed TIFF.
3163
+ The distance parameter of the Delta filter is set
3164
+ to match the number of bytes per pixel in the image.
3165
+ For example, 24-bit RGB bitmap needs
3166
+ .BR dist=3 ,
3167
+ and it is also good to pass
3168
+ .B pb=0
3169
+ to LZMA2 to accommodate the three-byte alignment:
3170
+ .RS
3171
+ .PP
3172
+ .nf
3173
+ .ft CR
3174
+ xz \-\-delta=dist=3 \-\-lzma2=pb=0 foo.tiff
3175
+ .ft R
3176
+ .fi
3177
+ .RE
3178
+ .PP
3179
+ If multiple images have been put into a single archive (for example,
3180
+ .BR .tar ),
3181
+ the Delta filter will work on that too as long as all images
3182
+ have the same number of bytes per pixel.
3183
+ .
3184
+ .SH "SEE ALSO"
3185
+ .BR xzdec (1),
3186
+ .BR xzdiff (1),
3187
+ .BR xzgrep (1),
3188
+ .BR xzless (1),
3189
+ .BR xzmore (1),
3190
+ .BR gzip (1),
3191
+ .BR bzip2 (1),
3192
+ .BR 7z (1)
3193
+ .PP
3194
+ XZ Utils: <https://tukaani.org/xz/>
3195
+ .br
3196
+ XZ Embedded: <https://tukaani.org/xz/embedded.html>
3197
+ .br
3198
+ LZMA SDK: <https://7-zip.org/sdk.html>
janus/share/man/man1/xzgrep.1 ADDED
@@ -0,0 +1,144 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ .\" SPDX-License-Identifier: 0BSD
2
+ .\"
3
+ .\" Authors: Lasse Collin
4
+ .\" Jia Tan
5
+ .\"
6
+ .\" (Note that this file is not based on gzip's zgrep.1.)
7
+ .\"
8
+ .TH XZGREP 1 "2024-02-13" "Tukaani" "XZ Utils"
9
+ .SH NAME
10
+ xzgrep \- search possibly-compressed files for patterns
11
+ .
12
+ .SH SYNOPSIS
13
+ .B xzgrep
14
+ .RI [ option... ]
15
+ .RI [ pattern_list ]
16
+ .RI [ file... ]
17
+ .br
18
+ .B xzegrep
19
+ \&...
20
+ .br
21
+ .B xzfgrep
22
+ \&...
23
+ .br
24
+ .B lzgrep
25
+ \&...
26
+ .br
27
+ .B lzegrep
28
+ \&...
29
+ .br
30
+ .B lzfgrep
31
+ \&...
32
+ .
33
+ .SH DESCRIPTION
34
+ .B xzgrep
35
+ invokes
36
+ .BR grep (1)
37
+ on uncompressed contents of files.
38
+ The formats of the
39
+ .I files
40
+ are determined from the filename suffixes.
41
+ Any
42
+ .I file
43
+ with a suffix supported by
44
+ .BR xz (1),
45
+ .BR gzip (1),
46
+ .BR bzip2 (1),
47
+ .BR lzop (1),
48
+ .BR zstd (1),
49
+ or
50
+ .BR lz4 (1)
51
+ will be decompressed;
52
+ all other files are assumed to be uncompressed.
53
+ .PP
54
+ If no
55
+ .I files
56
+ are specified or
57
+ .I file
58
+ is
59
+ .B \-
60
+ then standard input is read.
61
+ When reading from standard input, only files supported by
62
+ .BR xz (1)
63
+ are decompressed.
64
+ Other files are assumed to be in uncompressed form already.
65
+ .PP
66
+ Most
67
+ .I options
68
+ of
69
+ .BR grep (1)
70
+ are supported.
71
+ However, the following options are not supported:
72
+ .IP "" 4
73
+ .BR \-r ,
74
+ .B \-\-recursive
75
+ .IP "" 4
76
+ .BR \-R ,
77
+ .B \-\-dereference\-recursive
78
+ .IP "" 4
79
+ .BR \-d ,
80
+ .BI \-\-directories= action
81
+ .IP "" 4
82
+ .BR \-Z ,
83
+ .B \-\-null
84
+ .IP "" 4
85
+ .BR \-z ,
86
+ .B \-\-null\-data
87
+ .IP "" 4
88
+ .BI \-\-include= glob
89
+ .IP "" 4
90
+ .BI \-\-exclude= glob
91
+ .IP "" 4
92
+ .BI \-\-exclude\-from= file
93
+ .IP "" 4
94
+ .BI \-\-exclude\-dir= glob
95
+ .PP
96
+ .B xzegrep
97
+ is an alias for
98
+ .BR "xzgrep \-E" .
99
+ .B xzfgrep
100
+ is an alias for
101
+ .BR "xzgrep \-F" .
102
+ .PP
103
+ The commands
104
+ .BR lzgrep ,
105
+ .BR lzegrep ,
106
+ and
107
+ .B lzfgrep
108
+ are provided for backward compatibility with LZMA Utils.
109
+ .
110
+ .SH EXIT STATUS
111
+ .TP
112
+ 0
113
+ At least one match was found from at least one of the input files.
114
+ No errors occurred.
115
+ .TP
116
+ 1
117
+ No matches were found from any of the input files.
118
+ No errors occurred.
119
+ .TP
120
+ >1
121
+ One or more errors occurred.
122
+ It is unknown if matches were found.
123
+ .
124
+ .SH ENVIRONMENT
125
+ .TP
126
+ .B GREP
127
+ If
128
+ .B GREP
129
+ is set to a non-empty value,
130
+ it is used instead of
131
+ .BR "grep" ,
132
+ .BR "grep \-E" ,
133
+ or
134
+ .BR "grep \-F" .
135
+ .
136
+ .SH "SEE ALSO"
137
+ .BR grep (1),
138
+ .BR xz (1),
139
+ .BR gzip (1),
140
+ .BR bzip2 (1),
141
+ .BR lzop (1),
142
+ .BR zstd (1),
143
+ .BR lz4 (1),
144
+ .BR zgrep (1)
janus/share/man/man1/xzless.1 ADDED
@@ -0,0 +1,67 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ .\" SPDX-License-Identifier: 0BSD
2
+ .\"
3
+ .\" Authors: Andrew Dudman
4
+ .\" Lasse Collin
5
+ .\"
6
+ .\" (Note that this file is not based on gzip's zless.1.)
7
+ .\"
8
+ .TH XZLESS 1 "2024-02-12" "Tukaani" "XZ Utils"
9
+ .SH NAME
10
+ xzless, lzless \- view xz or lzma compressed (text) files
11
+ .SH SYNOPSIS
12
+ .B xzless
13
+ .RI [ file ...]
14
+ .br
15
+ .B lzless
16
+ .RI [ file ...]
17
+ .SH DESCRIPTION
18
+ .B xzless
19
+ is a filter that displays text from compressed files to a terminal.
20
+ Files supported by
21
+ .BR xz (1)
22
+ are decompressed;
23
+ other files are assumed to be in uncompressed form already.
24
+ If no
25
+ .I files
26
+ are given,
27
+ .B xzless
28
+ reads from standard input.
29
+ .PP
30
+ .B xzless
31
+ uses
32
+ .BR less (1)
33
+ to present its output.
34
+ Unlike
35
+ .BR xzmore ,
36
+ its choice of pager cannot be altered by
37
+ setting an environment variable.
38
+ Commands are based on both
39
+ .BR more (1)
40
+ and
41
+ .BR vi (1)
42
+ and allow back and forth movement and searching.
43
+ See the
44
+ .BR less (1)
45
+ manual for more information.
46
+ .PP
47
+ The command named
48
+ .B lzless
49
+ is provided for backward compatibility with LZMA Utils.
50
+ .SH ENVIRONMENT
51
+ .TP
52
+ .B LESSMETACHARS
53
+ A list of characters special to the shell.
54
+ Set by
55
+ .B xzless
56
+ unless it is already set in the environment.
57
+ .TP
58
+ .B LESSOPEN
59
+ Set to a command line to invoke the
60
+ .BR xz (1)
61
+ decompressor for preprocessing the input files to
62
+ .BR less (1).
63
+ .SH "SEE ALSO"
64
+ .BR less (1),
65
+ .BR xz (1),
66
+ .BR xzmore (1),
67
+ .BR zless (1)
janus/share/man/man1/xzmore.1 ADDED
@@ -0,0 +1,65 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ .\" SPDX-License-Identifier: 0BSD
2
+ .\"
3
+ .\" Authors: Andrew Dudman
4
+ .\" Lasse Collin
5
+ .\"
6
+ .\" (Note that this file is based on xzless.1 instead of gzip's zmore.1.)
7
+ .\"
8
+ .TH XZMORE 1 "2024-02-12" "Tukaani" "XZ Utils"
9
+ .SH NAME
10
+ xzmore, lzmore \- view xz or lzma compressed (text) files
11
+ .
12
+ .SH SYNOPSIS
13
+ .B xzmore
14
+ .RI [ file ...]
15
+ .br
16
+ .B lzmore
17
+ .RI [ file ...]
18
+ .
19
+ .SH DESCRIPTION
20
+ .B xzmore
21
+ displays text from compressed files to a terminal using
22
+ .BR more (1).
23
+ Files supported by
24
+ .BR xz (1)
25
+ are decompressed;
26
+ other files are assumed to be in uncompressed form already.
27
+ If no
28
+ .I files
29
+ are given,
30
+ .B xzmore
31
+ reads from standard input.
32
+ See the
33
+ .BR more (1)
34
+ manual for the keyboard commands.
35
+ .PP
36
+ Note that scrolling backwards might not be possible
37
+ depending on the implementation of
38
+ .BR more (1).
39
+ This is because
40
+ .B xzmore
41
+ uses a pipe to pass the decompressed data to
42
+ .BR more (1).
43
+ .BR xzless (1)
44
+ uses
45
+ .BR less (1)
46
+ which provides more advanced features.
47
+ .PP
48
+ The command
49
+ .B lzmore
50
+ is provided for backward compatibility with LZMA Utils.
51
+ .
52
+ .SH ENVIRONMENT
53
+ .TP
54
+ .B PAGER
55
+ If
56
+ .B PAGER
57
+ is set,
58
+ its value is used as the pager instead of
59
+ .BR more (1).
60
+ .
61
+ .SH "SEE ALSO"
62
+ .BR more (1),
63
+ .BR xz (1),
64
+ .BR xzless (1),
65
+ .BR zmore (1)
janus/share/man/pt_BR/man1/lzcat.1 ADDED
@@ -0,0 +1,1951 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ '\" t
2
+ .\" SPDX-License-Identifier: 0BSD
3
+ .\"
4
+ .\" Authors: Lasse Collin
5
+ .\" Jia Tan
6
+ .\"
7
+ .\" Brazilian Portuguese translations for xz package
8
+ .\" Traduções em português brasileiro para o pacote xz.
9
+ .\" Rafael Fontenelle <rafaelff@gnome.org>, 2022-2023.
10
+ .\"
11
+ .\"*******************************************************************
12
+ .\"
13
+ .\" This file was generated with po4a. Translate the source file.
14
+ .\"
15
+ .\"*******************************************************************
16
+ .TH XZ 1 2024\-12\-30 Tukaani "XZ Utils"
17
+ .
18
+ .SH NOME
19
+ xz, unxz, xzcat, lzma, unlzma, lzcat \- Compacta ou descompacta arquivos .xz
20
+ e .lzma
21
+ .
22
+ .SH SINOPSE
23
+ \fBxz\fP [\fIopção...\fP] [\fIarquivo...\fP]
24
+ .
25
+ .SH "COMANDOS APELIDOS"
26
+ \fBunxz\fP é equivalente a \fBxz \-\-decompress\fP.
27
+ .br
28
+ \fBxzcat\fP é equivalente a \fBxz \-\-decompress \-\-stdout\fP.
29
+ .br
30
+ \fBlzma\fP é equivalente a \fBxz \-\-format=lzma\fP.
31
+ .br
32
+ \fBunlzma\fP é equivalente a \fBxz \-\-format=lzma \-\-decompress\fP.
33
+ .br
34
+ \fBlzcat\fP é equivalente a \fBxz \-\-format=lzma \-\-decompress \-\-stdout\fP.
35
+ .PP
36
+ Ao escrever scripts que precisam descompactar arquivos, é recomendável
37
+ sempre usar o nome \fBxz\fP com os argumentos apropriados (\fBxz \-d\fP ou \fBxz \-dc\fP) em vez dos nomes \fBunxz\fP e \fBxzcat\fP.
38
+ .
39
+ .SH DESCRIÇÃO
40
+ \fBxz\fP é uma ferramenta de compactação de dados de uso geral com sintaxe de
41
+ linha de comando semelhante ao \fBgzip\fP(1) e ao \fBbzip2\fP(1). O formato de
42
+ arquivo nativo é o formato \fB.xz\fP, mas o formato legado \fB.lzma\fP usado por
43
+ LZMA Utils e fluxos compactados brutos sem cabeçalhos de formato de
44
+ contêiner também são suportados. Além disso, a descompactação do formato
45
+ \&\fB.lz\fP usado por \fBlzip\fP é suportada.
46
+ .PP
47
+ \fBxz\fP compacta ou descompacta cada \fIarquivo\fP de acordo com o modo de
48
+ operação selecionado. Se nenhum \fIarquivo\fP for fornecido ou \fIarquivo\fP for
49
+ \fB\-\fP, \fBxz\fP lê da entrada padrão e grava os dados processados na saída
50
+ padrão. \fBxz\fP recusará (exibirá um erro e ignorará o \fIarquivo\fP) para gravar
51
+ dados compactados na saída padrão se for um terminal. Da mesma forma, \fBxz\fP
52
+ se recusará a ler dados compactados da entrada padrão se for um terminal.
53
+ .PP
54
+ A menos que \fB\-\-stdout\fP seja especificado, \fIarquivos\fP diferentes de \fB\-\fP
55
+ são gravados em um novo arquivo cujo nome é derivado do nome \fIarquivo\fP de
56
+ origem:
57
+ .IP \(bu 3
58
+ Ao compactar, o sufixo do formato de arquivo de destino (\fB.xz\fP ou \fB.lzma\fP)
59
+ é anexado ao nome do arquivo de origem para obter o nome do arquivo de
60
+ destino.
61
+ .IP \(bu 3
62
+ Ao descompactar, o sufixo \fB.xz\fP, \fB.lzma\fP ou \fB.lz\fP é removido do nome do
63
+ arquivo para obter o nome do arquivo de destino. \fBxz\fP também reconhece os
64
+ sufixos \fB.txz\fP e \fB.tlz\fP e os substitui pelo sufixo \fB.tar\fP.
65
+ .PP
66
+ Se o arquivo de destino já existir, um erro será exibido e \fIarquivo\fP será
67
+ ignorado.
68
+ .PP
69
+ A menos que grave na saída padrão, \fBxz\fP exibirá um aviso e pulará o
70
+ \fIarquivo\fP se qualquer um dos seguintes se aplicar:
71
+ .IP \(bu 3
72
+ \fIArquivo\fP não é um arquivo normal. Links simbólicos não são seguidos e,
73
+ portanto, não são considerados arquivos comuns.
74
+ .IP \(bu 3
75
+ \fIArquivo\fP tem mais de um link físico.
76
+ .IP \(bu 3
77
+ \fIFile\fP tem setuid, setgid ou sticky bit definido.
78
+ .IP \(bu 3
79
+ O modo de operação está definido para compactar e o \fIarquivo\fP já possui um
80
+ sufixo do formato de arquivo de destino (\fB.xz\fP ou \fB.txz\fP ao compactar para
81
+ o formato \fB.xz\fP e \fB.lzma \fP ou \fB.tlz\fP ao compactar para o formato
82
+ \&\fB.lzma\fP).
83
+ .IP \(bu 3
84
+ O modo de operação está definido para descompactar e o \fIarquivo\fP não possui
85
+ um sufixo de nenhum dos formatos de arquivo suportados (\fB.xz\fP, \fB.txz\fP,
86
+ \&\fB.lzma\fP, \fB.tlz\fP , ou \fB.lz\fP).
87
+ .PP
88
+ Depois de compactar ou descompactar com êxito o \fIarquivo\fP, o \fBxz\fP copia o
89
+ dono, grupo, permissões, horário de acesso e horário de modificação do
90
+ \fIarquivo\fP de origem para o arquivo de destino. Se a cópia do grupo falhar,
91
+ as permissões serão modificadas para que o arquivo de destino não se torne
92
+ acessível a usuários que não têm permissão para acessar o \fIarquivo\fP de
93
+ origem. \fBxz\fP ainda não oferece suporte à cópia de outros metadados, como
94
+ listas de controle de acesso ou atributos estendidos.
95
+ .PP
96
+ Depois que o arquivo de destino for fechado com êxito, o \fIarquivo\fP de
97
+ origem será removido, a menos que \fB\-\-keep\fP tenha sido especificado. O
98
+ \fIarquivo\fP de origem nunca é removido se a saída for gravada na saída padrão
99
+ ou se ocorrer um erro.
100
+ .PP
101
+ O envio de \fBSIGINFO\fP ou \fBSIGUSR1\fP para o processo do \fBxz\fP faz com que ele
102
+ imprima informações de andamento para erro padrão. Isso tem uso limitado,
103
+ pois quando o erro padrão é um terminal, usar \fB\-\-verbose\fP exibirá um
104
+ indicador de progresso de atualização automática.
105
+ .
106
+ .SS "Uso de memória"
107
+ O uso de memória de \fBxz\fP varia de algumas centenas de kilobytes a vários
108
+ gigabytes, dependendo das configurações de compactação. As configurações
109
+ usadas ao compactar um arquivo determinam os requisitos de memória do
110
+ descompactador. Normalmente, o descompactador precisa de 5\ % a 20\ % da
111
+ quantidade de memória que o compactador precisou ao criar o arquivo. Por
112
+ exemplo, descompactar um arquivo criado com \fBxz \-9\fP atualmente requer 65\ MiB de memória. Ainda assim, é possível ter arquivos \fB.xz\fP que requerem
113
+ vários gigabytes de memória para descompactar.
114
+ .PP
115
+ Especialmente os usuários de sistemas mais antigos podem achar irritante a
116
+ possibilidade de uso de memória muito grande. Para evitar surpresas
117
+ desconfortáveis, o \fBxz\fP possui um limitador de uso de memória embutido, que
118
+ está desabilitado por padrão. Embora alguns sistemas operacionais forneçam
119
+ maneiras de limitar o uso de memória dos processos, confiar nele não foi
120
+ considerado flexível o suficiente (por exemplo, usar \fBulimit\fP(1) para
121
+ limitar a memória virtual tende a prejudicar \fBmmap\fP(2)).
122
+ .PP
123
+ O limitador de uso de memória pode ser ativado com a opção de linha de
124
+ comando \fB\-\-memlimit=\fP\fIlimite\fP. Geralmente é mais conveniente habilitar o
125
+ limitador por padrão definindo a variável de ambiente \fBXZ_DEFAULTS\fP, por
126
+ exemplo, \fBXZ_DEFAULTS=\-\-memlimit=150MiB\fP. É possível definir os limites
127
+ separadamente para compactação e descompactação usando
128
+ \fB\-\-memlimit\-compress=\fP\fIlimite\fP e \fB\-\-memlimit\-decompress=\fP\fIlimite\fP. Usar
129
+ essas duas opções fora de \fBXZ_DEFAULTS\fP raramente é útil porque uma única
130
+ execução de \fBxz\fP não pode fazer compactação e descompactação e
131
+ \fB\-\-memlimit=\fP\fIlimite\fP (ou \fB\-M\fP \fIlimite\fP ) é mais curto para digitar na
132
+ linha de comando.
133
+ .PP
134
+ Se o limite de uso de memória especificado for excedido durante a
135
+ descompactação, \fBxz\fP exibirá um erro e a descompactação do arquivo
136
+ falhará. Se o limite for excedido durante a compactação, \fBxz\fP tentará
137
+ reduzir as configurações para que o limite não seja mais excedido (exceto ao
138
+ usar \fB\-\-format=raw\fP ou \fB\-\-no\-adjust\fP). Dessa forma, a operação não
139
+ falhará, a menos que o limite seja muito pequeno. A escala das configurações
140
+ é feita em etapas que não correspondem às predefinições do nível de
141
+ compactação, por exemplo, se o limite for apenas um pouco menor que o valor
142
+ necessário para \fBxz \-9\fP, as configurações serão reduzidas apenas um pouco ,
143
+ não até \fBxz \-8\fP.
144
+ .
145
+ .SS "Concatenação e preenchimento com arquivos .xz"
146
+ É possível concatenar arquivos \fB.xz\fP como estão. \fBxz\fP irá descompactar
147
+ tais arquivos como se fossem um único arquivo \fB.xz\fP.
148
+ .PP
149
+ É possível inserir preenchimento entre as partes concatenadas ou após a
150
+ última parte. O preenchimento deve consistir em bytes nulos e o tamanho do
151
+ preenchimento deve ser um múltiplo de quatro bytes. Isso pode ser útil, por
152
+ exemplo, se o arquivo \fB.xz\fP for armazenado em uma mídia que mede tamanhos
153
+ de arquivo em blocos de 512 bytes.
154
+ .PP
155
+ Concatenação e preenchimento não são permitidos com arquivos \fB.lzma\fP ou
156
+ fluxos brutos.
157
+ .
158
+ .SH OPÇÕES
159
+ .
160
+ .SS "Sufixos inteiros e valores especiais"
161
+ Na maioria dos lugares onde um argumento inteiro é esperado, um sufixo
162
+ opcional é suportado para indicar facilmente números inteiros grandes. Não
163
+ deve haver espaço entre o número inteiro e o sufixo.
164
+ .TP
165
+ \fBKiB\fP
166
+ Multiplica o inteiro por 1.024 (2^10). \fBKi\fP, \fBk\fP, \fBkB\fP, \fBK\fP e \fBKB\fP são
167
+ aceitos como sinônimos de \fBKiB\fP.
168
+ .TP
169
+ \fBMiB\fP
170
+ Multiplica o número inteiro por 1.048.576 (2^20). \fBMi\fP, \fBm\fP, \fBM\fP e \fBMB\fP
171
+ são aceitos como sinônimos de \fBMiB\fP.
172
+ .TP
173
+ \fBGiB\fP
174
+ Multiplica o número inteiro por 1.073.741.824 (2^30). \fBGi\fP, \fBg\fP, \fBG\fP e
175
+ \fBGB\fP são aceitos como sinônimos de \fBGiB\fP.
176
+ .PP
177
+ O valor especial \fBmax\fP pode ser usado para indicar o valor inteiro máximo
178
+ suportado pela opção.
179
+ .
180
+ .SS "Modo de operação"
181
+ Se várias opções de modo de operação forem dadas, a última entrará em vigor.
182
+ .TP
183
+ \fB\-z\fP, \fB\-\-compress\fP
184
+ Compacta. Este é o modo de operação padrão quando nenhuma opção de modo de
185
+ operação é especificada e nenhum outro modo de operação está implícito no
186
+ nome do comando (por exemplo, \fBunxz\fP implica em \fB\-\-decompress\fP).
187
+ .IP ""
188
+ .\" The DESCRIPTION section already says this but it's good to repeat it
189
+ .\" here because the default behavior is a bit dangerous and new users
190
+ .\" in a hurry may skip reading the DESCRIPTION section.
191
+ After successful compression, the source file is removed unless writing to
192
+ standard output or \fB\-\-keep\fP was specified.
193
+ .TP
194
+ \fB\-d\fP, \fB\-\-decompress\fP, \fB\-\-uncompress\fP
195
+ .\" The DESCRIPTION section already says this but it's good to repeat it
196
+ .\" here because the default behavior is a bit dangerous and new users
197
+ .\" in a hurry may skip reading the DESCRIPTION section.
198
+ Decompress. After successful decompression, the source file is removed
199
+ unless writing to standard output or \fB\-\-keep\fP was specified.
200
+ .TP
201
+ \fB\-t\fP, \fB\-\-test\fP
202
+ Testa a integridade de \fIarquivos\fP compactados. Esta opção é equivalente a
203
+ \fB\-\-decompress \-\-stdout\fP exceto que os dados descompactados são descartados
204
+ em vez de serem gravados na saída padrão. Nenhum arquivo é criado ou
205
+ removido.
206
+ .TP
207
+ \fB\-l\fP, \fB\-\-list\fP
208
+ Imprime informações sobre \fIarquivos\fP compactados. Nenhuma saída
209
+ descompactada é produzida e nenhum arquivo é criado ou removido. No modo de
210
+ lista, o programa não pode ler os dados compactados da entrada padrão ou de
211
+ outras fontes não pesquisáveis.
212
+ .IP ""
213
+ A listagem padrão mostra informações básicas sobre \fIarquivos\fP, um arquivo
214
+ por linha. Para obter informações mais detalhadas, use também a opção
215
+ \fB\-\-verbose\fP. Para obter ainda mais informações, use \fB\-\-verbose\fP duas
216
+ vezes, mas observe que isso pode ser lento, porque obter todas as
217
+ informações extras requer muitas buscas. A largura da saída detalhada excede
218
+ 80 caracteres, portanto, canalizar a saída para, por exemplo, \fBless\ \-S\fP
219
+ pode ser conveniente se o terminal não tiver largura o suficiente.
220
+ .IP ""
221
+ A saída exata pode variar entre versões \fBxz\fP e localidades diferentes. Para
222
+ saída legível por máquina, \fB\-\-robot \-\-list\fP deve ser usado.
223
+ .
224
+ .SS "Modificadores de operação"
225
+ .TP
226
+ \fB\-k\fP, \fB\-\-keep\fP
227
+ Não exclui os arquivos de entrada.
228
+ .IP ""
229
+ Desde \fBxz\fP 5.2.6, esta opção também faz \fBxz\fP compactar ou descompactar
230
+ mesmo se a entrada for um link simbólico para um arquivo comum, tiver mais
231
+ de um link físico ou tiver o setuid, setgid ou sticky bit definir. Os bits
232
+ setuid, setgid e sticky não são copiados para o arquivo de destino. Nas
233
+ versões anteriores, isso era feito apenas com \fB\-\-force\fP.
234
+ .TP
235
+ \fB\-f\fP, \fB\-\-force\fP
236
+ Esta opção tem vários efeitos:
237
+ .RS
238
+ .IP \(bu 3
239
+ Se o arquivo de destino já existir, o exclui antes de compactar ou
240
+ descompactar.
241
+ .IP \(bu 3
242
+ Compacta ou descompacta, mesmo que a entrada seja um link simbólico para um
243
+ arquivo normal, tenha mais de um link físico ou tenha setuid, setgid ou
244
+ sticky bit definido. Os bits setuid, setgid e sticky não são copiados para o
245
+ arquivo de destino.
246
+ .IP \(bu 3
247
+ Quando usado com \fB\-\-decompress\fP \fB\-\-stdout\fP e \fBxz\fP não consegue reconhecer
248
+ o tipo do arquivo de origem, copia o arquivo de origem como está na saída
249
+ padrão. Isso permite que \fBxzcat\fP \fB\-\-force\fP seja usado como \fBcat\fP(1) para
250
+ arquivos que não foram compactados com \fBxz\fP. Observe que, no futuro, o
251
+ \fBxz\fP pode oferecer suporte a novos formatos de arquivo compactado, o que
252
+ pode fazer com que o \fBxz\fP descompacte mais tipos de arquivos em vez de
253
+ copiá\-los como na saída padrão. \fB\-\-format=\fP\fIformato\fP pode ser usado para
254
+ restringir \fBxz\fP para descompactar apenas um único formato de arquivo.
255
+ .RE
256
+ .TP
257
+ \fB\-c\fP, \fB\-\-stdout\fP, \fB\-\-to\-stdout\fP
258
+ Grava os dados compactados ou descompactados na saída padrão em vez de em um
259
+ arquivo. Isso implica em \fB\-\-keep\fP.
260
+ .TP
261
+ \fB\-\-single\-stream\fP
262
+ Descompacta apenas o primeiro fluxo de \fB.xz\fP e ignora silenciosamente
263
+ possíveis dados de entrada restantes após o fluxo. Normalmente, esse
264
+ restante posterior sem uso faz com que \fBxz\fP exiba um erro.
265
+ .IP ""
266
+ \fBxz\fP nunca descompacta mais de um fluxo de arquivos \fB.lzma\fP ou fluxos
267
+ brutos, mas esta opção ainda faz \fBxz\fP ignorar os possíveis dados
268
+ posteriores após o arquivo \fB.lzma\fP ou fluxo bruto.
269
+ .IP ""
270
+ Esta opção não tem efeito se o modo de operação não for \fB\-\-decompress\fP ou
271
+ \fB\-\-test\fP.
272
+ .TP
273
+ \fB\-\-no\-sparse\fP
274
+ Desativa a criação de arquivos esparsos. Por padrão, ao descompactar em um
275
+ arquivo normal, \fBxz\fP tenta tornar o arquivo esparso se os dados
276
+ descompactados contiverem longas sequências de zeros binários. Ele também
277
+ funciona ao gravar na saída padrão, desde que a saída padrão esteja
278
+ conectada a um arquivo normal e certas condições adicionais sejam atendidas
279
+ para torná\-la segura. A criação de arquivos esparsos pode economizar espaço
280
+ em disco e acelerar a descompactação, reduzindo a quantidade de E/S do
281
+ disco.
282
+ .TP
283
+ \fB\-S\fP \fI.suf\fP, \fB\-\-suffix=\fP\fI.suf\fP
284
+ Ao compactar, usa \fI.suf\fP como sufixo para o arquivo de destino em vez de
285
+ \&\fB.xz\fP ou \fB.lzma\fP. Se não estiver gravando na saída padrão e o arquivo de
286
+ origem já tiver o sufixo \fI.suf\fP, um aviso será exibido e o arquivo será
287
+ ignorado.
288
+ .IP ""
289
+ Ao descompactar, reconhece arquivos com o sufixo \fI.suf\fP além de arquivos
290
+ com o sufixo \fB.xz\fP, \fB.txz\fP, \fB.lzma\fP, \fB.tlz\fP ou \fB.lz\fP . Se o arquivo de
291
+ origem tiver o sufixo \fI.suf\fP, o sufixo será removido para obter o nome do
292
+ arquivo de destino.
293
+ .IP ""
294
+ Ao compactar ou descompactar fluxos brutos (\fB\-\-format=raw\fP), o sufixo
295
+ sempre deve ser especificado, a menos que seja gravado na saída padrão,
296
+ porque não há sufixo padrão para fluxos brutos.
297
+ .TP
298
+ \fB\-\-files\fP[\fB=\fP\fIarquivo\fP]
299
+ Lê os nomes dos arquivos a serem processados em \fIarquivo\fP; se \fIarquivo\fP
300
+ for omitido, os nomes dos arquivos serão lidos da entrada padrão. Os nomes
301
+ de arquivo devem terminar com o caractere de nova linha. Um traço (\fB\-\fP) é
302
+ considerado um nome de arquivo regular; não significa entrada padrão. Se os
303
+ nomes de arquivo forem fornecidos também como argumentos de linha de
304
+ comando, eles serão processados antes da leitura dos nomes de arquivo de
305
+ \fIarquivo\fP.
306
+ .TP
307
+ \fB\-\-files0\fP[\fB=\fP\fIarquivo\fP]
308
+ Isso é idêntico a \fB\-\-files\fP[\fB=\fP\fIarquivo\fP], exceto que cada nome de
309
+ arquivo deve ser finalizado com o caractere nulo.
310
+ .
311
+ .SS "Opções básicas de formato de arquivo e de compactação"
312
+ .TP
313
+ \fB\-F\fP \fIformato\fP, \fB\-\-format=\fP\fIformato\fP
314
+ Especifica o \fIformato\fP de arquivo para compactar ou descompactar:
315
+ .RS
316
+ .TP
317
+ \fBauto\fP
318
+ Este é o padrão. Ao compactar, \fBauto\fP é equivalente a \fBxz\fP. Ao
319
+ descompactar, o formato do arquivo de entrada é detectado
320
+ automaticamente. Observe que os fluxos brutos (criados com \fB\-\-format=raw\fP)
321
+ não podem ser detectados automaticamente.
322
+ .TP
323
+ \fBxz\fP
324
+ Compacta no formato de arquivo \fB.xz\fP ou aceite apenas arquivos \fB.xz\fP ao
325
+ descompactar.
326
+ .TP
327
+ \fBlzma\fP, \fBalone\fP
328
+ Compacta no formato de arquivo legado \fB.lzma\fP ou aceite apenas arquivos
329
+ \&\fB.lzma\fP ao descompactar. O nome alternativo \fBalone\fP é fornecido para
330
+ compatibilidade com versões anteriores do LZMA Utils.
331
+ .TP
332
+ \fBlzip\fP
333
+ Aceita apenas arquivos \fB.lz\fP ao descompactar. Sem suporte a compactação.
334
+ .IP ""
335
+ O formato \fB.lz\fP versão 0 e a versão não estendida 1 são suportados. Os
336
+ arquivos da versão 0 foram produzidos por \fBlzip\fP 1.3 e anteriores. Esses
337
+ arquivos não são comuns, mas podem ser encontrados em arquivos compactados,
338
+ pois alguns pacotes de origem foram lançados nesse formato. As pessoas
339
+ também podem ter arquivos pessoais antigos neste formato. O suporte de
340
+ descompactação para o formato versão 0 foi removido em \fBlzip\fP 1.18.
341
+ .IP ""
342
+ \fBlzip\fP 1.4 e posteriores criam arquivos no formato versão 1. A extensão do
343
+ marcador de descarga de sincronização para o formato versão 1 foi adicionada
344
+ em \fBlzip\fP 1.6. Esta extensão raramente é usada e não é suportada por \fBxz\fP
345
+ (diagnosticada como entrada corrompida).
346
+ .TP
347
+ \fBraw\fP
348
+ Compacta ou descompacta um fluxo bruto (sem cabeçalhos). Isso é destinado
349
+ apenas a usuários avançados. Para decodificar fluxos brutos, você precisa
350
+ usar \fB\-\-format=raw\fP e especificar explicitamente a cadeia de filtros, que
351
+ normalmente seria armazenada nos cabeçalhos do contêiner.
352
+ .RE
353
+ .TP
354
+ \fB\-C\fP \fIverificação\fP, \fB\-\-check=\fP\fIverificação\fP
355
+ Especifica o tipo de verificação de integridade. A verificação é calculada a
356
+ partir dos dados descompactados e armazenados no arquivo \fB.xz\fP. Esta opção
357
+ tem efeito somente ao compactar no formato \fB.xz\fP; o formato \fB.lzma\fP não
358
+ oferece suporte a verificações de integridade. A verificação de integridade
359
+ (se for o caso) é verificada quando o arquivo \fB.xz\fP é descompactado.
360
+ .IP ""
361
+ Tipos de \fIverificação\fP suportados:
362
+ .RS
363
+ .TP
364
+ \fBnone\fP
365
+ Não calcula uma verificação de integridade. Isso geralmente é uma má
366
+ ideia. Pode ser útil quando a integridade dos dados é verificada por outros
367
+ meios.
368
+ .TP
369
+ \fBcrc32\fP
370
+ Calcula CRC32 usando o polinômio do IEEE\-802.3 (Ethernet).
371
+ .TP
372
+ \fBcrc64\fP
373
+ Calcula CRC64 usando o polinômio de ECMA\-182. Este é o padrão, pois é um
374
+ pouco melhor que o CRC32 na detecção de arquivos danificados e a diferença
375
+ de velocidade é insignificante.
376
+ .TP
377
+ \fBsha256\fP
378
+ Calcula SHA\-256. Isso é um pouco mais lento do que CRC32 e CRC64.
379
+ .RE
380
+ .IP ""
381
+ A integridade dos cabeçalhos de \fB.xz\fP é sempre verificada com CRC32. Não é
382
+ possível alterá\-la ou desativá\-la.
383
+ .TP
384
+ \fB\-\-ignore\-check\fP
385
+ Não confere a verificação de integridade dos dados compactados ao
386
+ descompactar. Os valores CRC32 nos cabeçalhos \fB.xz\fP ainda serão conferidos
387
+ normalmente.
388
+ .IP ""
389
+ \fBNão use esta opção a menos que saiba o que está fazendo.\fP Possíveis razões
390
+ para usar esta opção:
391
+ .RS
392
+ .IP \(bu 3
393
+ Tentativa de recuperar dados de um arquivo .xz corrompido.
394
+ .IP \(bu 3
395
+ Acelerar a descompactação. Isso é importante principalmente com SHA\-256 ou
396
+ com arquivos extremamente bem compactados. É recomendável não usar essa
397
+ opção para essa finalidade, a menos que a integridade do arquivo seja
398
+ verificada externamente de alguma outra forma.
399
+ .RE
400
+ .TP
401
+ \fB\-0\fP ... \fB\-9\fP
402
+ Seleciona um nível de predefinição de compactação. O padrão é \fB\-6\fP. Se
403
+ vários níveis de predefinição forem especificados, o último terá efeito. Se
404
+ uma cadeia de filtro personalizada já foi especificada, especificar um nível
405
+ de predefinição de compactação limpa a cadeia de filtro personalizada.
406
+ .IP ""
407
+ As diferenças entre as predefinições são mais significativas do que com
408
+ \fBgzip\fP(1) e \fBbzip2\fP(1). As configurações de compactação selecionadas
409
+ determinam os requisitos de memória do descompactador, portanto, usar um
410
+ nível de predefinição muito alto pode dificultar a descompactação do arquivo
411
+ em um sistema antigo com pouca RAM. Especificamente, \fBnão é uma boa ideia usar cegamente \-9 para tudo\fP como costuma acontecer com \fBgzip\fP(1) e
412
+ \fBbzip2\fP(1).
413
+ .RS
414
+ .TP
415
+ \fB\-0\fP ... \fB\-3\fP
416
+ Estas são predefinições um tanto rápidas. \fB\-0\fP às vezes é mais rápida que
417
+ \fBgzip \-9\fP ao mesmo tempo que compacta muito melhor. As mais altas
418
+ geralmente têm velocidade comparável ao \fBbzip2\fP(1) com taxa de compactação
419
+ comparável ou melhor, embora os resultados dependam muito do tipo de dados
420
+ que estão sendo compactados.
421
+ .TP
422
+ \fB\-4\fP ... \fB\-6\fP
423
+ Compactação boa a muito boa, mantendo o uso de memória do descompactador
424
+ razoável mesmo para sistemas antigos. \fB\-6\fP é o padrão, que geralmente é uma
425
+ boa escolha para distribuir arquivos que precisam ser descompactáveis, mesmo
426
+ em sistemas com apenas 16\ MiB de RAM. (\fB\-5e\fP ou \fB\-6e\fP também vale a pena
427
+ considerar. Veja \fB\-\-extreme\fP.)
428
+ .TP
429
+ \fB\-7 ... \-9\fP
430
+ Eles são como \fB\-6\fP, mas com requisitos de memória de compressor e
431
+ descompressor mais altos. Eles são úteis apenas ao compactar arquivos
432
+ maiores que 8\ MiB, 16\ MiB e 32\ MiB, respectivamente.
433
+ .RE
434
+ .IP ""
435
+ No mesmo hardware, a velocidade de descompactação é aproximadamente um
436
+ número constante de bytes de dados compactados por segundo. Em outras
437
+ palavras, quanto melhor a compactação, mais rápida será a
438
+ descompactação. Isso também significa que a quantidade de saída não
439
+ compactada produzida por segundo pode variar muito.
440
+ .IP ""
441
+ A tabela a seguir resume os recursos das predefinições:
442
+ .RS
443
+ .RS
444
+ .PP
445
+ .TS
446
+ tab(;);
447
+ c c c c c
448
+ n n n n n.
449
+ Predefinição;DicTam;CompCPU;CompMem;DecMem
450
+ \-0;256 KiB;0;3 MiB;1 MiB
451
+ \-1;1 MiB;1;9 MiB;2 MiB
452
+ \-2;2 MiB;2;17 MiB;3 MiB
453
+ \-3;4 MiB;3;32 MiB;5 MiB
454
+ \-4;4 MiB;4;48 MiB;5 MiB
455
+ \-5;8 MiB;5;94 MiB;9 MiB
456
+ \-6;8 MiB;6;94 MiB;9 MiB
457
+ \-7;16 MiB;6;186 MiB;17 MiB
458
+ \-8;32 MiB;6;370 MiB;33 MiB
459
+ \-9;64 MiB;6;674 MiB;65 MiB
460
+ .TE
461
+ .RE
462
+ .RE
463
+ .IP ""
464
+ Descrições das colunas:
465
+ .RS
466
+ .IP \(bu 3
467
+ DicTam é o tamanho do dicionário LZMA2. É desperdício de memória usar um
468
+ dicionário maior que o tamanho do arquivo descompactado. É por isso que é
469
+ bom evitar usar as predefinições \fB\-7\fP ... \fB\-9\fP quando não há real
470
+ necessidade deles. Em \fB\-6\fP e inferior, a quantidade de memória desperdiçada
471
+ geralmente é baixa o suficiente para não importar.
472
+ .IP \(bu 3
473
+ CompCPU é uma representação simplificada das configurações LZMA2 que afetam
474
+ a velocidade de compactação. O tamanho do dicionário também afeta a
475
+ velocidade, portanto, embora o CompCPU seja o mesmo para os níveis \fB\-6\fP
476
+ \&... \fB\-9\fP, níveis mais altos ainda tendem a ser um pouco mais lentos. Para
477
+ obter uma compactação ainda mais lenta e possivelmente melhor, consulte
478
+ \fB\-\-extreme\fP.
479
+ .IP \(bu 3
480
+ CompMem contains the compressor memory requirements in the single\-threaded
481
+ mode. It may vary slightly between \fBxz\fP versions.
482
+ .IP \(bu 3
483
+ DecMem contém os requisitos de memória do descompactador. Ou seja, as
484
+ configurações de compactação determinam os requisitos de memória do
485
+ descompactador. O uso exato da memória do descompactador é um pouco maior do
486
+ que o tamanho do dicionário LZMA2, mas os valores na tabela foram
487
+ arredondados para o próximo MiB completo.
488
+ .RE
489
+ .IP ""
490
+ Memory requirements of the multi\-threaded mode are significantly higher than
491
+ that of the single\-threaded mode. With the default value of
492
+ \fB\-\-block\-size\fP, each thread needs 3*3*DictSize plus CompMem or DecMem. For
493
+ example, four threads with preset \fB\-6\fP needs 660\(en670\ MiB of memory.
494
+ .TP
495
+ \fB\-e\fP, \fB\-\-extreme\fP
496
+ Usa uma variante mais lenta do nível de predefinição de compactação
497
+ selecionado (\fB\-0\fP ... \fB\-9\fP) para obter uma taxa de compactação um pouco
498
+ melhor, mas, com azar, isso também pode piorar. O uso da memória do
499
+ descompressor não é afetado, mas o uso da memória do compactador aumenta um
500
+ pouco nos níveis de predefinição \fB\-0\fP ... \fB\-3\fP.
501
+ .IP ""
502
+ Como existem duas predefinições com tamanhos de dicionário 4\ MiB e 8\ MiB,
503
+ as predefinições \fB\-3e\fP e \fB\-5e\fP usam configurações um pouco mais rápidas
504
+ (CompCPU inferior) do que \fB\-4e\fP e \fB\-6e\fP, respectivamente. Dessa forma, não
505
+ há duas predefinições idênticas.
506
+ .RS
507
+ .RS
508
+ .PP
509
+ .TS
510
+ tab(;);
511
+ c c c c c
512
+ n n n n n.
513
+ Predefinição;DicTam;CompCPU;CompMem;DecMem
514
+ \-0e;256 KiB;8;4 MiB;1 MiB
515
+ \-1e;1 MiB;8;13 MiB;2 MiB
516
+ \-2e;2 MiB;8;25 MiB;3 MiB
517
+ \-3e;4 MiB;7;48 MiB;5 MiB
518
+ \-4e;4 MiB;8;48 MiB;5 MiB
519
+ \-5e;8 MiB;7;94 MiB;9 MiB
520
+ \-6e;8 MiB;8;94 MiB;9 MiB
521
+ \-7e;16 MiB;8;186 MiB;17 MiB
522
+ \-8e;32 MiB;8;370 MiB;33 MiB
523
+ \-9e;64 MiB;8;674 MiB;65 MiB
524
+ .TE
525
+ .RE
526
+ .RE
527
+ .IP ""
528
+ Por exemplo, há um total de quatro predefinições que usam o dicionário 8\ MiB, cuja ordem do mais rápido ao mais lento é \fB\-5\fP, \fB\-6\fP, \fB\-5e\fP e
529
+ \fB\-6e\fP.
530
+ .TP
531
+ \fB\-\-fast\fP
532
+ .PD 0
533
+ .TP
534
+ \fB\-\-best\fP
535
+ .PD
536
+ Esses são apelidos um tanto enganosos para \fB\-0\fP e \fB\-9\fP,
537
+ respectivamente. Eles são fornecidos apenas para compatibilidade com versões
538
+ anteriores do LZMA Utils. Evite usar essas opções.
539
+ .TP
540
+ \fB\-\-block\-size=\fP\fItamanho\fP
541
+ Ao compactar para o formato \fB.xz\fP, divida os dados de entrada em blocos de
542
+ \fItamanho\fP bytes. Os blocos são compactados independentemente uns dos
543
+ outros, o que ajuda no multi\-threading e torna possível a descompactação
544
+ limitada de acesso aleatório. Essa opção normalmente é usada para substituir
545
+ o tamanho de bloco padrão no modo multi\-thread, mas também pode ser usada em
546
+ thread única.
547
+ .IP ""
548
+ In multi\-threaded mode about three times \fIsize\fP bytes will be allocated in
549
+ each thread for buffering input and output. The default \fIsize\fP is three
550
+ times the LZMA2 dictionary size or 1 MiB, whichever is more. Typically a
551
+ good value is 2\(en4 times the size of the LZMA2 dictionary or at least 1
552
+ MiB. Using \fIsize\fP less than the LZMA2 dictionary size is waste of RAM
553
+ because then the LZMA2 dictionary buffer will never get fully used. In
554
+ multi\-threaded mode, the sizes of the blocks are stored in the block
555
+ headers. This size information is required for multi\-threaded
556
+ decompression.
557
+ .IP ""
558
+ In single\-threaded mode no block splitting is done by default. Setting this
559
+ option doesn't affect memory usage. No size information is stored in block
560
+ headers, thus files created in single\-threaded mode won't be identical to
561
+ files created in multi\-threaded mode. The lack of size information also
562
+ means that \fBxz\fP won't be able decompress the files in multi\-threaded mode.
563
+ .TP
564
+ \fB\-\-block\-list=\fP\fIitems\fP
565
+ When compressing to the \fB.xz\fP format, start a new block with an optional
566
+ custom filter chain after the given intervals of uncompressed data.
567
+ .IP ""
568
+ The \fIitems\fP are a comma\-separated list. Each item consists of an optional
569
+ filter chain number between 0 and 9 followed by a colon (\fB:\fP) and a
570
+ required size of uncompressed data. Omitting an item (two or more
571
+ consecutive commas) is a shorthand to use the size and filters of the
572
+ previous item.
573
+ .IP ""
574
+ If the input file is bigger than the sum of the sizes in \fIitems\fP, the last
575
+ item is repeated until the end of the file. A special value of \fB0\fP may be
576
+ used as the last size to indicate that the rest of the file should be
577
+ encoded as a single block.
578
+ .IP ""
579
+ An alternative filter chain for each block can be specified in combination
580
+ with the \fB\-\-filters1=\fP\fIfilters\fP \&...\& \fB\-\-filters9=\fP\fIfilters\fP options.
581
+ These options define filter chains with an identifier between 1\(en9.
582
+ Filter chain 0 can be used to refer to the default filter chain, which is
583
+ the same as not specifying a filter chain. The filter chain identifier can
584
+ be used before the uncompressed size, followed by a colon (\fB:\fP). For
585
+ example, if one specifies \fB\-\-block\-list=1:2MiB,3:2MiB,2:4MiB,,2MiB,0:4MiB\fP
586
+ then blocks will be created using:
587
+ .RS
588
+ .IP \(bu 3
589
+ The filter chain specified by \fB\-\-filters1\fP and 2 MiB input
590
+ .IP \(bu 3
591
+ The filter chain specified by \fB\-\-filters3\fP and 2 MiB input
592
+ .IP \(bu 3
593
+ The filter chain specified by \fB\-\-filters2\fP and 4 MiB input
594
+ .IP \(bu 3
595
+ The filter chain specified by \fB\-\-filters2\fP and 4 MiB input
596
+ .IP \(bu 3
597
+ The default filter chain and 2 MiB input
598
+ .IP \(bu 3
599
+ The default filter chain and 4 MiB input for every block until end of input.
600
+ .RE
601
+ .IP ""
602
+ If one specifies a size that exceeds the encoder's block size (either the
603
+ default value in threaded mode or the value specified with
604
+ \fB\-\-block\-size=\fP\fIsize\fP), the encoder will create additional blocks while
605
+ keeping the boundaries specified in \fIitems\fP. For example, if one specifies
606
+ \fB\-\-block\-size=10MiB\fP \fB\-\-block\-list=5MiB,10MiB,8MiB,12MiB,24MiB\fP and the
607
+ input file is 80 MiB, one will get 11 blocks: 5, 10, 8, 10, 2, 10, 10, 4,
608
+ 10, 10, and 1 MiB.
609
+ .IP ""
610
+ No modo multi\-thread, os tamanhos dos blocos são armazenados nos cabeçalhos
611
+ dos blocos. Isso não é feito no modo de thread única, portanto, a saída
612
+ codificada não será idêntica à do modo multi\-thread.
613
+ .TP
614
+ \fB\-\-flush\-timeout=\fP\fItempo_limite\fP
615
+ Ao compactar, se mais de \fItempo_limite\fP milissegundos (um número inteiro
616
+ positivo) se passaram desde a liberação anterior e a leitura de mais entrada
617
+ seria bloqueada, todos os dados de entrada pendentes serão liberados do
618
+ codificador e disponibilizados no fluxo de saída. Isso pode ser útil se
619
+ \fBxz\fP for usado para compactar dados transmitidos por uma rede. Valores
620
+ \fItempo_limite\fP pequenos tornam os dados disponíveis na extremidade
621
+ receptora com um pequeno atraso, mas valores \fItempo_limite\fP grandes
622
+ oferecem melhor taxa de compactação.
623
+ .IP ""
624
+ Esse recurso está desabilitado por padrão. Se esta opção for especificada
625
+ mais de uma vez, a última terá efeito. O valor especial \fItempo_limite\fP de
626
+ \fB0\fP pode ser usado para desabilitar explicitamente esse recurso.
627
+ .IP ""
628
+ Este recurso não está disponível em sistemas não\-POSIX.
629
+ .IP ""
630
+ .\" FIXME
631
+ \fBEste recurso ainda é experimental.\fP Atualmente, \fBxz\fP não é adequado para
632
+ descompactar o fluxo em tempo real devido à forma como \fBxz\fP faz o buffer.
633
+ .TP
634
+ \fB\-\-memlimit\-compress=\fP\fIlimite\fP
635
+ Define um limite de uso de memória para compactação. Se esta opção for
636
+ especificada várias vezes, a última entrará em vigor.
637
+ .IP ""
638
+ Se as configurações de compactação excederem o \fIlimite\fP, \fBxz\fP tentará
639
+ ajustar as configurações para baixo para que o limite não seja mais excedido
640
+ e exibirá um aviso de que o ajuste automático foi feito. Os ajustes são
641
+ feitos nesta ordem: reduzindo o número de encadeamentos, alternando para o
642
+ modo sigle\-thread se até mesmo uma thread no modo multi\-thread exceder o
643
+ \fIlimite\fP e, finalmente, reduzindo o tamanho do dicionário LZMA2.
644
+ .IP ""
645
+ Ao compactar com \fB\-\-format=raw\fP ou se \fB\-\-no\-adjust\fP tiver sido
646
+ especificado, apenas o número de threads pode ser reduzido, pois isso pode
647
+ ser feito sem afetar a saída compactada.
648
+ .IP ""
649
+ Se o \fIlimite\fP não puder ser alcançado mesmo com os ajustes descritos acima,
650
+ um erro será exibido e \fBxz\fP sairá com status de saída 1.
651
+ .IP ""
652
+ O \fIlimite\fP pode ser especificado de várias maneiras:
653
+ .RS
654
+ .IP \(bu 3
655
+ O \fIlimite\fP pode ser um valor absoluto em bytes. Usar um sufixo inteiro como
656
+ \fBMiB\fP pode ser útil. Exemplo: \fB\-\-memlimit\-compress=80MiB\fP
657
+ .IP \(bu 3
658
+ O \fIlimite\fP pode ser especificado como uma porcentagem da memória física
659
+ total (RAM). Isso pode ser útil especialmente ao definir a variável de
660
+ ambiente \fBXZ_DEFAULTS\fP em um script de inicialização de shell que é
661
+ compartilhado entre diferentes computadores. Dessa forma o limite é
662
+ automaticamente maior em sistemas com mais memória. Exemplo:
663
+ \fB\-\-memlimit\-compress=70%\fP
664
+ .IP \(bu 3
665
+ O \fIlimite\fP pode ser redefinido para seu valor padrão, definindo\-o como
666
+ \fB0\fP. Atualmente, isso equivale a definir \fIlimite\fP como \fBmax\fP (sem limite
667
+ de uso de memória).
668
+ .RE
669
+ .IP ""
670
+ Para \fBxz\fP de 32 bits, há um caso especial: se o \fIlimite\fP estiver acima de
671
+ \fB4020\ MiB\fP, o \fIlimite\fP é definido como \fB4020\ MiB\fP. No MIPS32 \fB2000\ MiB\fP é usado em seu lugar. (Os valores \fB0\fP e \fBmax\fP não são afetados por
672
+ isso. Um recurso semelhante não existe para descompactação.) Isso pode ser
673
+ útil quando um executável de 32 bits tem acesso a espaço de endereço de 4\ GiB (2 GiB no MIPS32) enquanto espero não causar danos em outras situações.
674
+ .IP ""
675
+ Consulte também a seção \fBUso de memória\fP.
676
+ .TP
677
+ \fB\-\-memlimit\-decompress=\fP\fIlimite\fP
678
+ Define um limite de uso de memória para descompactação. Isso também afeta o
679
+ modo \fB\-\-list\fP. Se a operação não for possível sem exceder o \fIlimite\fP,
680
+ \fBxz\fP exibirá um erro e a descompactação do arquivo falhará. Consulte
681
+ \fB\-\-memlimit\-compress=\fP\fIlimite\fP para possíveis maneiras de especificar o
682
+ \fIlimite\fP.
683
+ .TP
684
+ \fB\-\-memlimit\-mt\-decompress=\fP\fIlimite\fP
685
+ Define um limite de uso de memória para descompactação multi\-thread. Isso
686
+ pode afetar apenas o número de threads; isso nunca fará com que \fBxz\fP se
687
+ recuse a descompactar um arquivo. Se \fIlimite\fP for muito baixo para permitir
688
+ qualquer multi\-thread, o \fIlimite\fP será ignorado e \fBxz\fP continuará no modo
689
+ de thread única. Observe que se \fB\-\-memlimit\-decompress\fP também for usado,
690
+ ele sempre se aplicará aos modos de thread única e multi\-thread e, portanto,
691
+ o \fIlimite\fP efetivo para multi\-threading nunca será maior que o limite
692
+ definido com \fB\-\-memlimit\-decompress\fP.
693
+ .IP ""
694
+ Em contraste com as outras opções de limite de uso de memória,
695
+ \fB\-\-memlimit\-mt\-decompress=\fP\fIlimite\fP tem um padrão \fIlimite\fP específico do
696
+ sistema. \fBxz \-\-info\-memory\fP pode ser usado para ver o valor atual.
697
+ .IP ""
698
+ Esta opção e seu valor padrão existem porque, sem qualquer limite, o
699
+ descompactador usando threads pode acabar alocando uma quantidade insana de
700
+ memória com alguns arquivos de entrada. Se o \fIlimite\fP padrão for muito
701
+ baixo em seu sistema, sinta\-se à vontade para aumentar o \fIlimite\fP, mas
702
+ nunca defina\-o para um valor maior que a quantidade de RAM utilizável, pois
703
+ com os arquivos de entrada apropriados \fBxz\fP tentará usar essa quantidade de
704
+ memória mesmo com um baixo número de threads. Ficar sem memória ou trocar
705
+ não melhorará o desempenho da descompactação.
706
+ .IP ""
707
+ Consulte \fB\-\-memlimit\-compress=\fP\fIlimite\fP para possíveis maneiras de
708
+ especificar o \fIlimite\fP. Definir \fIlimite\fP como \fB0\fP redefine \fIlimite\fP para
709
+ o valor padrão específico do sistema.
710
+ .TP
711
+ \fB\-M\fP \fIlimite\fP, \fB\-\-memlimit=\fP\fIlimite\fP, \fB\-\-memory=\fP\fIlimite\fP
712
+ Isso é equivalente a especificar \fB\-\-memlimit\-compress=\fP\fIlimite\fP
713
+ \fB\-\-memlimit\-decompress=\fP\fIlimite\fP \fB\-\-memlimit\-mt\-decompress=\fP\fIlimite\fP.
714
+ .TP
715
+ \fB\-\-no\-adjust\fP
716
+ Exibe um erro e saia se o limite de uso de memória não puder ser atendido
717
+ sem ajustar as configurações que afetam a saída compactada. Ou seja, isso
718
+ evita que \fBxz\fP alterne o codificador do modo multi\-thread para o modo
719
+ encadeado único e reduza o tamanho do dicionário LZMA2. Mesmo quando esta
720
+ opção é usada, o número de threads pode ser reduzido para atender ao limite
721
+ de uso de memória, pois isso não afetará a saída compactada.
722
+ .IP ""
723
+ O ajuste automático é sempre desativado ao criar fluxos brutos
724
+ (\fB\-\-format=raw\fP).
725
+ .TP
726
+ \fB\-T\fP \fIthreads\fP, \fB\-\-threads=\fP\fIthreads\fP
727
+ Especifica o número de threads de trabalho a serem usados. Definir
728
+ \fIthreads\fP para um valor especial \fB0\fP faz com que \fBxz\fP use tantos threads
729
+ quanto o(s) processador(es) no suporte do sistema. O número real de
730
+ encadeamentos pode ser menor que \fIthreads\fP se o arquivo de entrada não for
731
+ grande o suficiente para subdividir em threads com as configurações
732
+ fornecidas ou se o uso de mais threads exceder o limite de uso de memória.
733
+ .IP ""
734
+ Os compactadores usando thread única e várias threads produzem saídas
735
+ diferentes. O compactador de thread única fornecerá o menor tamanho de
736
+ arquivo, mas apenas a saída do compactador de várias threads pode ser
737
+ descompactada usando várias threads. Definir \fIthreads\fP como \fB1\fP usará o
738
+ modo de thread única. Definir \fIthreads\fP para qualquer outro valor,
739
+ incluindo \fB0\fP, usará o compressor de várias threads, mesmo que o sistema
740
+ tenha suporte a apenas uma thread de hardware. (\fBxz\fP 5.2.x usou o modo de
741
+ thread única nesta situação.)
742
+ .IP ""
743
+ Para usar o modo de várias threads com apenas uma thread, defina \fIthreads\fP
744
+ como \fB+1\fP. O prefixo \fB+\fP não tem efeito com valores diferentes de \fB1\fP. Um
745
+ limite de uso de memória ainda pode fazer \fBxz\fP alternar para o modo de
746
+ thread única, a menos que \fB\-\-no\-adjust\fP seja usado. O suporte para o
747
+ prefixo \fB+\fP foi adicionado no \fBxz\fP 5.4.0.
748
+ .IP ""
749
+ Se um número automático de threads foi solicitado e nenhum limite de uso de
750
+ memória foi especificado, um limite flexível padrão específico do sistema
751
+ será usado para possivelmente limitar o número de threads. É um limite
752
+ flexível no sentido de que é ignorado se o número de threads se tornar um,
753
+ portanto, um limite flexível nunca impedirá \fBxz\fP de compactar ou
754
+ descompactar. Este limite flexível padrão não fará com que \fBxz\fP alterne do
755
+ modo de várias threads para o modo de thread única. Os limites ativos podem
756
+ ser vistos com \fBxz \-\-info\-memory\fP.
757
+ .IP ""
758
+ Atualmente, o único método de threading é dividir a entrada em blocos e
759
+ comprimi\-los independentemente um do outro. O tamanho padrão do bloco
760
+ depende do nível de compactação e pode ser substituído com a opção
761
+ \fB\-\-block\-size=\fP\fItamanho\fP.
762
+ .IP ""
763
+ A descompactação em threads funciona apenas em arquivos que contêm vários
764
+ blocos com informações de tamanho nos cabeçalhos dos blocos. Todos os
765
+ arquivos grandes o suficiente compactados no modo de várias thread atendem a
766
+ essa condição, mas os arquivos compactados no modo de thread única não
767
+ atendem, mesmo se \fB\-\-block\-size=\fP\fItamanho\fP tiver sido usado.
768
+ .IP ""
769
+ The default value for \fIthreads\fP is \fB0\fP. In \fBxz\fP 5.4.x and older the
770
+ default is \fB1\fP.
771
+ .
772
+ .SS "Cadeias de filtro de compressor personalizadas"
773
+ Uma cadeia de filtro personalizada permite especificar as configurações de
774
+ compactação em detalhes, em vez de confiar nas configurações associadas às
775
+ predefinições. Quando uma cadeia de filtro personalizada é especificada, as
776
+ opções predefinidas (\fB\-0\fP \&...\& \fB\-9\fP e \fB\-\-extreme\fP) anteriores na linha
777
+ de comando são esquecidas. Se uma opção predefinida for especificada após
778
+ uma ou mais opções de cadeia de filtros personalizados, a nova predefinição
779
+ entrará em vigor e as opções de cadeia de filtros personalizados
780
+ especificadas anteriormente serão esquecidas.
781
+ .PP
782
+ Uma cadeia de filtro é comparável à tubulação na linha de comando. Ao
783
+ compactar, a entrada descompactada vai para o primeiro filtro, cuja saída
784
+ vai para o próximo filtro (se houver). A saída do último filtro é gravada no
785
+ arquivo compactado. O número máximo de filtros na cadeia é quatro, mas
786
+ normalmente uma cadeia de filtros tem apenas um ou dois filtros.
787
+ .PP
788
+ Muitos filtros têm limitações sobre onde podem estar na cadeia de filtros:
789
+ alguns filtros podem funcionar apenas como o último filtro na cadeia, alguns
790
+ apenas como filtro não\-último e alguns funcionam em qualquer posição na
791
+ cadeia. Dependendo do filtro, essa limitação é inerente ao projeto do filtro
792
+ ou existe para evitar problemas de segurança.
793
+ .PP
794
+ A custom filter chain can be specified in two different ways. The options
795
+ \fB\-\-filters=\fP\fIfilters\fP and \fB\-\-filters1=\fP\fIfilters\fP \&...\&
796
+ \fB\-\-filters9=\fP\fIfilters\fP allow specifying an entire filter chain in one
797
+ option using the liblzma filter string syntax. Alternatively, a filter
798
+ chain can be specified by using one or more individual filter options in the
799
+ order they are wanted in the filter chain. That is, the order of the
800
+ individual filter options is significant! When decoding raw streams
801
+ (\fB\-\-format=raw\fP), the filter chain must be specified in the same order as
802
+ it was specified when compressing. Any individual filter or preset options
803
+ specified before the full chain option (\fB\-\-filters=\fP\fIfilters\fP) will be
804
+ forgotten. Individual filters specified after the full chain option will
805
+ reset the filter chain.
806
+ .PP
807
+ Both the full and individual filter options take filter\-specific \fIoptions\fP
808
+ as a comma\-separated list. Extra commas in \fIoptions\fP are ignored. Every
809
+ option has a default value, so specify those you want to change.
810
+ .PP
811
+ Para ver toda a cadeia de filtros e \fIopções\fP, use \fBxz \-vv\fP (isto é, use
812
+ \fB\-\-verbose\fP duas vezes). Isso também funciona para visualizar as opções da
813
+ cadeia de filtros usadas pelas predefinições.
814
+ .TP
815
+ \fB\-\-filters=\fP\fIfilters\fP
816
+ Specify the full filter chain or a preset in a single option. Each filter
817
+ can be separated by spaces or two dashes (\fB\-\-\fP). \fIfilters\fP may need to be
818
+ quoted on the shell command line so it is parsed as a single option. To
819
+ denote \fIoptions\fP, use \fB:\fP or \fB=\fP. A preset can be prefixed with a \fB\-\fP
820
+ and followed with zero or more flags. The only supported flag is \fBe\fP to
821
+ apply the same options as \fB\-\-extreme\fP.
822
+ .TP
823
+ \fB\-\-filters1\fP=\fIfilters\fP ... \fB\-\-filters9\fP=\fIfilters\fP
824
+ Specify up to nine additional filter chains that can be used with
825
+ \fB\-\-block\-list\fP.
826
+ .IP ""
827
+ For example, when compressing an archive with executable files followed by
828
+ text files, the executable part could use a filter chain with a BCJ filter
829
+ and the text part only the LZMA2 filter.
830
+ .TP
831
+ \fB\-\-filters\-help\fP
832
+ Display a help message describing how to specify presets and custom filter
833
+ chains in the \fB\-\-filters\fP and \fB\-\-filters1=\fP\fIfilters\fP \&...\&
834
+ \fB\-\-filters9=\fP\fIfilters\fP options, and exit successfully.
835
+ .TP
836
+ \fB\-\-lzma1\fP[\fB=\fP\fIopções\fP]
837
+ .PD 0
838
+ .TP
839
+ \fB\-\-lzma2\fP[\fB=\fP\fIopções\fP]
840
+ .PD
841
+ Adiciona o filtro LZMA1 ou LZMA2 à cadeia de filtros. Esses filtros podem
842
+ ser usados apenas como o último filtro na cadeia.
843
+ .IP ""
844
+ LZMA1 é um filtro legado, que é suportado quase exclusivamente devido ao
845
+ formato de arquivo legado \fB.lzma\fP, que suporta apenas LZMA1. LZMA2 é uma
846
+ versão atualizada do LZMA1 para corrigir alguns problemas práticos do
847
+ LZMA1. O formato \fB.xz\fP usa LZMA2 e não suporta LZMA1. A velocidade de
848
+ compactação e as proporções de LZMA1 e LZMA2 são praticamente as mesmas.
849
+ .IP ""
850
+ LZMA1 e LZMA2 compartilham o mesmo conjunto de \fIopções\fP:
851
+ .RS
852
+ .TP
853
+ \fBpreset=\fP\fIpredefinição\fP
854
+ Redefine todas as \fIopções\fP de LZMA1 ou LZMA2 para
855
+ \fIpredefinição\fP. \fIPredefinição\fP consiste em um número inteiro, que pode ser
856
+ seguido por modificadores de predefinição de uma única letra. O inteiro pode
857
+ ser de \fB0\fP a \fB9\fP, correspondendo às opções de linha de comando \fB\-0\fP
858
+ \&...\& \fB\-9\fP. O único modificador suportado atualmente é \fBe\fP, que
859
+ corresponde a \fB\-\-extreme\fP. Se nenhum \fBpreset\fP for especificado, os valores
860
+ padrão das \fIopções\fP LZMA1 ou LZMA2 serão obtidos da predefinição \fB6\fP.
861
+ .TP
862
+ \fBdict=\fP\fItamanho\fP
863
+ O \fItamanho\fP do dicionário (buffer de histórico) indica quantos bytes dos
864
+ dados não compactados processados recentemente são mantidos na memória. O
865
+ algoritmo tenta encontrar sequências de bytes repetidos (correspondências)
866
+ nos dados não compactados e substituí\-los por referências aos dados
867
+ atualmente no dicionário. Quanto maior o dicionário, maior a chance de
868
+ encontrar uma correspondência. Portanto, aumentar o dicionário \fItamanho\fP
869
+ geralmente melhora a taxa de compactação, mas um dicionário maior que o
870
+ arquivo não compactado é um desperdício de memória.
871
+ .IP ""
872
+ Um \fItamanho\fP de dicionário típico é de 64\ KiB a 64\ MiB. O mínimo é 4\ KiB. O máximo para compactação é atualmente 1,5\ GiB (1536\ MiB). O
873
+ descompactador já oferece suporte a dicionários de até um byte a menos de 4\ GiB, que é o máximo para os formatos de fluxo LZMA1 e LZMA2.
874
+ .IP ""
875
+ O \fItamanho\fP de dicionário e o localizador de correspondência (\fImf\fP) juntos
876
+ determinam o uso de memória do codificador LZMA1 ou LZMA2. O mesmo (ou
877
+ maior) \fItamanho\fP de dicionário é necessário para descompactar que foi usado
878
+ durante a compactação, portanto, o uso de memória do decodificador é
879
+ determinado pelo tamanho do dicionário usado durante a compactação. Os
880
+ cabeçalhos \fB.xz\fP armazenam o \fItamanho\fP de dicionário como 2^\fIn\fP ou 2^\fIn\fP
881
+ + 2^(\fIn\fP\-1), então esses \fItamanhos\fP são um tanto preferidos para
882
+ compactação. Outros \fItamanhos\fP serão arredondados quando armazenados nos
883
+ cabeçalhos \fB.xz\fP.
884
+ .TP
885
+ \fBlc=\fP\fIlc\fP
886
+ Especifica o número de bits de contexto literais. O mínimo é 0 e o máximo é
887
+ 4; o padrão é 3. Al��m disso, a soma de \fIlc\fP e \fIlp\fP não deve exceder 4.
888
+ .IP ""
889
+ Todos os bytes que não podem ser codificados como correspondências são
890
+ codificados como literais. Ou seja, literais são simplesmente bytes de 8
891
+ bits que são codificados um de cada vez.
892
+ .IP ""
893
+ A codificação literal assume que os bits \fIlc\fP mais altos do byte não
894
+ compactado anterior se correlacionam com o próximo byte. Por exemplo, em um
895
+ texto típico em inglês, uma letra maiúscula geralmente é seguida por uma
896
+ letra minúscula, e uma letra minúscula geralmente é seguida por outra letra
897
+ minúscula. No conjunto de caracteres US\-ASCII, os três bits mais altos são
898
+ 010 para letras maiúsculas e 011 para letras minúsculas. Quando \fIlc\fP é pelo
899
+ menos 3, a codificação literal pode aproveitar essa propriedade nos dados
900
+ não compactados.
901
+ .IP ""
902
+ O valor padrão (3) geralmente é bom. Se você deseja compactação máxima,
903
+ experimente \fBlc=4\fP. Às vezes ajuda um pouco, às vezes piora a
904
+ compactação. Se piorar, experimente \fBlc=2\fP também.
905
+ .TP
906
+ \fBlp=\fP\fIlp\fP
907
+ Especifica o número de bits de posição literal. O mínimo é 0 e o máximo é 4;
908
+ o padrão é 0.
909
+ .IP ""
910
+ \fILp\fP afeta que tipo de alinhamento nos dados não compactados é assumido ao
911
+ codificar literais. Consulte \fIpb\fP abaixo para obter mais informações sobre
912
+ alinhamento.
913
+ .TP
914
+ \fBpb=\fP\fIpb\fP
915
+ Especifica o número de bits de posição. O mínimo é 0 e o máximo é 4; o
916
+ padrão é 2.
917
+ .IP ""
918
+ \fIPb\fP afeta que tipo de alinhamento nos dados não compactados é assumido em
919
+ geral. O padrão significa alinhamento de quatro bytes (2^\fIpb\fP=2^2=4), que
920
+ geralmente é uma boa escolha quando não há melhor estimativa.
921
+ .IP ""
922
+ Quando o alinhamento é conhecido, definir \fIpb\fP adequadamente pode reduzir
923
+ um pouco o tamanho do arquivo. Por exemplo, com arquivos de texto com
924
+ alinhamento de um byte (US\-ASCII, ISO\-8859\-*, UTF\-8), a configuração \fBpb=0\fP
925
+ pode melhorar um pouco a compactação. Para texto UTF\-16, \fBpb=1\fP é uma boa
926
+ escolha. Se o alinhamento for um número ímpar como 3 bytes, \fBpb=0\fP pode ser
927
+ a melhor escolha.
928
+ .IP ""
929
+ Embora o alinhamento assumido possa ser ajustado com \fIpb\fP e \fIlp\fP, LZMA1 e
930
+ LZMA2 ainda favorecem ligeiramente o alinhamento de 16 bytes. Pode valer a
931
+ pena levar em consideração ao projetar formatos de arquivo que provavelmente
932
+ serão compactados com LZMA1 ou LZMA2.
933
+ .TP
934
+ \fBmf=\fP\fImf\fP
935
+ O localizador de correspondência tem um efeito importante na velocidade do
936
+ codificador, uso de memória e taxa de compactação. Normalmente, os
937
+ localizadores de correspondência de Hash Chain são mais rápidos do que os
938
+ localizadores de correspondência de árvore binária. O padrão depende do
939
+ \fIpredefinição\fP: 0 usa \fBhc3\fP, 1\(en3 usa \fBhc4\fP e o resto usa \fBbt4\fP.
940
+ .IP ""
941
+ Os seguintes localizadores de correspondência são suportados. As fórmulas de
942
+ uso de memória abaixo são aproximações aproximadas, que estão mais próximas
943
+ da realidade quando \fIdict\fP é uma potência de dois.
944
+ .RS
945
+ .TP
946
+ \fBhc3\fP
947
+ Cadeia de hashs com hashing de 2 e 3 bytes
948
+ .br
949
+ Valor mínimo para \fInice\fP: 3
950
+ .br
951
+ Uso de memória:
952
+ .br
953
+ \fIdict\fP * 7.5 (if \fIdict\fP <= 16 MiB);
954
+ .br
955
+ \fIdict\fP * 5.5 + 64 MiB (if \fIdict\fP > 16 MiB)
956
+ .TP
957
+ \fBhc4\fP
958
+ Cadeia de hashs com hashing de 2, 3 e 4 bytes
959
+ .br
960
+ Valor mínimo para \fInice\fP: 4
961
+ .br
962
+ Uso de memória:
963
+ .br
964
+ \fIdict\fP * 7.5 (if \fIdict\fP <= 32 MiB);
965
+ .br
966
+ \fIdict\fP * 6.5 (if \fIdict\fP > 32 MiB)
967
+ .TP
968
+ \fBbt2\fP
969
+ Árvore binária com hashing de 2 bytes
970
+ .br
971
+ Valor mínimo para \fInice\fP: 2
972
+ .br
973
+ Uso de memória: \fIdict\fP * 9.5
974
+ .TP
975
+ \fBbt3\fP
976
+ Árvore binária com hashing de 2 e 3 bytes
977
+ .br
978
+ Valor mínimo para \fInice\fP: 3
979
+ .br
980
+ Uso de memória:
981
+ .br
982
+ \fIdict\fP * 11.5 (if \fIdict\fP <= 16 MiB);
983
+ .br
984
+ \fIdict\fP * 9.5 + 64 MiB (if \fIdict\fP > 16 MiB)
985
+ .TP
986
+ \fBbt4\fP
987
+ Árvore binária com hashing de 2, 3 e 4 bytes
988
+ .br
989
+ Valor mínimo para \fInice\fP: 4
990
+ .br
991
+ Uso de memória:
992
+ .br
993
+ \fIdict\fP * 11.5 (if \fIdict\fP <= 32 MiB);
994
+ .br
995
+ \fIdict\fP * 10.5 (if \fIdict\fP > 32 MiB)
996
+ .RE
997
+ .TP
998
+ \fBmode=\fP\fImodo\fP
999
+ O \fImodo\fP de compactação especifica o método para analisar os dados
1000
+ produzidos pelo localizador de correspondência. Os \fImodos\fP suportados são
1001
+ \fBfast\fP e \fBnormal\fP. O padrão é \fBfast\fP para \fIpredefinições\fP 0\(en3 e
1002
+ \fBnormal\fP para \fIpredefinições\fP 4\(en9.
1003
+ .IP ""
1004
+ Normalmente, \fBfast\fP é usado com localizadores de correspondência cadeia de
1005
+ hashs e \fBnormal\fP com localizadores de correspondência de árvore
1006
+ binária. Isso também é o que os \fIpredefinições\fP fazem.
1007
+ .TP
1008
+ \fBnice=\fP\fInice\fP
1009
+ Especifica o que é considerado um bom comprimento para uma
1010
+ correspondência. Uma vez que uma correspondência de pelo menos \fInice\fP bytes
1011
+ é encontrada, o algoritmo para de procurar correspondências possivelmente
1012
+ melhores.
1013
+ .IP ""
1014
+ \fINice\fP pode ser 2\(en273 bytes. Valores mais altos tendem a fornecer melhor
1015
+ taxa de compactação em detrimento da velocidade. O padrão depende do
1016
+ \fIpredefinição\fP.
1017
+ .TP
1018
+ \fBdepth=\fP\fIprofundidade\fP
1019
+ Especifica a profundidade máxima de pesquisa no localizador de
1020
+ correspondências. O padrão é o valor especial de 0, que faz com que o
1021
+ compressor determine um \fIprofundidade\fP razoável de \fImf\fP e \fInice\fP.
1022
+ .IP ""
1023
+ Uma \fIprofundidade\fP razoável para cadeias de hash é 4\(en100 e 16\(en1000
1024
+ para árvores binárias. Usar valores muito altos para \fIprofundidade\fP pode
1025
+ tornar o codificador extremamente lento com alguns arquivos. Evite definir
1026
+ \fIprofundidade\fP acima de 1000 a menos que você esteja preparado para
1027
+ interromper a compactação caso ela esteja demorando muito.
1028
+ .RE
1029
+ .IP ""
1030
+ Ao decodificar fluxos brutos (\fB\-\-format=raw\fP), o LZMA2 precisa apenas do
1031
+ dicionário \fItamanho\fP. LZMA1 também precisa de \fIlc\fP, \fIlp\fP e \fIpb\fP.
1032
+ .TP
1033
+ \fB\-\-x86\fP[\fB=\fP\fIopções\fP]
1034
+ .PD 0
1035
+ .TP
1036
+ \fB\-\-arm\fP[\fB=\fP\fIopções\fP]
1037
+ .TP
1038
+ \fB\-\-armthumb\fP[\fB=\fP\fIopções\fP]
1039
+ .TP
1040
+ \fB\-\-arm64\fP[\fB=\fP\fIopções\fP]
1041
+ .TP
1042
+ \fB\-\-powerpc\fP[\fB=\fP\fIopções\fP]
1043
+ .TP
1044
+ \fB\-\-ia64\fP[\fB=\fP\fIopções\fP]
1045
+ .TP
1046
+ \fB\-\-sparc\fP[\fB=\fP\fIopções\fP]
1047
+ .TP
1048
+ \fB\-\-riscv\fP[\fB=\fP\fIopções\fP]
1049
+ .PD
1050
+ Adiciona um filtro de ramificação/chamada/salto (BCJ) à cadeia de
1051
+ filtros. Esses filtros podem ser usados apenas como um filtro não último na
1052
+ cadeia de filtros.
1053
+ .IP ""
1054
+ Um filtro BCJ converte endereços relativos no código de máquina em suas
1055
+ contrapartes absolutas. Isso não altera o tamanho dos dados, mas aumenta a
1056
+ redundância, o que pode ajudar o LZMA2 a produzir um arquivo \fB.xz\fP 0\(en15\ % menor. Os filtros BCJ são sempre reversíveis, portanto, usar um filtro BCJ
1057
+ para o tipo errado de dados não causa nenhuma perda de dados, embora possa
1058
+ piorar um pouco a taxa de compactação.Os filtros BCJ são muito rápidos e
1059
+ usam uma quantidade insignificante de memória.
1060
+ .IP ""
1061
+ Esses filtros BCJ têm problemas conhecidos relacionados à taxa de
1062
+ compactação:
1063
+ .RS
1064
+ .IP \(bu 3
1065
+ Alguns tipos de arquivos contendo código executável (por exemplo, arquivos
1066
+ de objeto, bibliotecas estáticas e módulos do kernel do Linux) têm os
1067
+ endereços nas instruções preenchidos com valores de preenchimento. Esses
1068
+ filtros BCJ ainda vão fazer a conversão de endereço, o que vai piorar a
1069
+ compactação desses arquivos.
1070
+ .IP \(bu 3
1071
+ Se um filtro BCJ for aplicado em um arquivo, é possível que isso torne a
1072
+ taxa de compactação pior do que não usar um filtro BCJ. Por exemplo, se
1073
+ houver executáveis semelhantes ou mesmo idênticos, a filtragem provavelmente
1074
+ tornará os arquivos menos semelhantes e, portanto, a compactação será
1075
+ pior. O conteúdo de arquivos não executáveis no mesmo arquivo também pode
1076
+ ser importante. Na prática tem que tentar com e sem filtro BCJ para ver qual
1077
+ é melhor em cada situação.
1078
+ .RE
1079
+ .IP ""
1080
+ Conjuntos de instruções diferentes têm alinhamento diferente: o arquivo
1081
+ executável deve ser alinhado a um múltiplo desse valor nos dados de entrada
1082
+ para fazer o filtro funcionar.
1083
+ .RS
1084
+ .RS
1085
+ .PP
1086
+ .TS
1087
+ tab(;);
1088
+ l n l
1089
+ l n l.
1090
+ Filtro;Alinhamento;Observações
1091
+ x86;1;x86 32 bits ou 64 bits
1092
+ ARM;4;
1093
+ ARM\-Thumb;2;
1094
+ ARM64;4;Alinhamento de 4096 bytes
1095
+ ;;é melhor
1096
+ PowerPC;4;Somente big endian
1097
+ IA\-64;16;Itanium
1098
+ SPARC;4;
1099
+ RISC\-V;2;
1100
+ .TE
1101
+ .RE
1102
+ .RE
1103
+ .IP ""
1104
+ Since the BCJ\-filtered data is usually compressed with LZMA2, the
1105
+ compression ratio may be improved slightly if the LZMA2 options are set to
1106
+ match the alignment of the selected BCJ filter. Examples:
1107
+ .RS
1108
+ .IP \(bu 3
1109
+ IA\-64 filter has 16\-byte alignment so \fBpb=4,lp=4,lc=0\fP is good with LZMA2
1110
+ (2^4=16).
1111
+ .IP \(bu 3
1112
+ RISC\-V code has 2\-byte or 4\-byte alignment depending on whether the file
1113
+ contains 16\-bit compressed instructions (the C extension). When 16\-bit
1114
+ instructions are used, \fBpb=2,lp=1,lc=3\fP or \fBpb=1,lp=1,lc=3\fP is good. When
1115
+ 16\-bit instructions aren't present, \fBpb=2,lp=2,lc=2\fP is the best.
1116
+ \fBreadelf \-h\fP can be used to check if "RVC" appears on the "Flags" line.
1117
+ .IP \(bu 3
1118
+ ARM64 is always 4\-byte aligned so \fBpb=2,lp=2,lc=2\fP is the best.
1119
+ .IP \(bu 3
1120
+ The x86 filter is an exception. It's usually good to stick to LZMA2's
1121
+ defaults (\fBpb=2,lp=0,lc=3\fP) when compressing x86 executables.
1122
+ .RE
1123
+ .IP ""
1124
+ Todos os filtros BCJ suportam as mesmas \fIopções\fP:
1125
+ .RS
1126
+ .TP
1127
+ \fBstart=\fP\fIdeslocamento\fP
1128
+ Especifica o \fIdeslocamento\fP inicial que é usado na conversão entre
1129
+ endereços relativos e absolutos. O \fIdeslocamento\fP deve ser um múltiplo do
1130
+ alinhamento do filtro (ver tabela acima). O padrão é zero. Na prática, o
1131
+ padrão é bom; especificar um \fIdeslocamento\fP personalizado quase nunca é
1132
+ útil.
1133
+ .RE
1134
+ .TP
1135
+ \fB\-\-delta\fP[\fB=\fP\fIopções\fP]
1136
+ Adiciona o filtro Delta à cadeia de filtros. O filtro Delta só pode ser
1137
+ usado como filtro não\-último na cadeia de filtros.
1138
+ .IP ""
1139
+ Atualmente, apenas o cálculo simples de delta byte a byte é suportado. Pode
1140
+ ser útil ao compactar, por exemplo, imagens bitmap não compactadas ou áudio
1141
+ PCM não compactado. No entanto, algoritmos de propósito especial podem
1142
+ fornecer resultados significativamente melhores do que Delta + LZMA2. Isso é
1143
+ verdade especialmente com áudio, que compacta mais rápido e melhor, por
1144
+ exemplo, com \fBflac\fP(1).
1145
+ .IP ""
1146
+ \fIOpções\fP suportadas:
1147
+ .RS
1148
+ .TP
1149
+ \fBdist=\fP\fIdistância\fP
1150
+ Especifica a \fIdistância\fP do cálculo delta em bytes. \fIdistância\fP deve ser
1151
+ 1\(en256. O padrão é 1.
1152
+ .IP ""
1153
+ Por exemplo, com \fBdist=2\fP e entrada de oito bytes A1 B1 A2 B3 A3 B5 A4 B7,
1154
+ a saída será A1 B1 01 02 01 02 01 02.
1155
+ .RE
1156
+ .
1157
+ .SS "Outras opções"
1158
+ .TP
1159
+ \fB\-q\fP, \fB\-\-quiet\fP
1160
+ Suprime avisos e avisos. Especifique isso duas vezes para suprimir erros
1161
+ também. Esta opção não tem efeito no status de saída. Ou seja, mesmo que um
1162
+ aviso tenha sido suprimido, o status de saída para indicar um aviso ainda é
1163
+ usado.
1164
+ .TP
1165
+ \fB\-v\fP, \fB\-\-verbose\fP
1166
+ Ser detalhado. Se o erro padrão estiver conectado a um terminal, \fBxz\fP
1167
+ exibirá um indicador de progresso. Especifique \fB\-\-verbose\fP duas vezes dará
1168
+ uma saída ainda mais detalhada.
1169
+ .IP ""
1170
+ O indicador de progresso mostra as seguintes informações:
1171
+ .RS
1172
+ .IP \(bu 3
1173
+ A porcentagem de conclusão é mostrada se o tamanho do arquivo de entrada for
1174
+ conhecido. Ou seja, a porcentagem não pode ser mostrada em encadeamentos
1175
+ (pipe).
1176
+ .IP \(bu 3
1177
+ Quantidade de dados compactados produzidos (compactando) ou consumidos
1178
+ (descompactando).
1179
+ .IP \(bu 3
1180
+ Quantidade de dados não compactados consumidos (compactação) ou produzidos
1181
+ (descompactação).
1182
+ .IP \(bu 3
1183
+ Taxa de compactação, que é calculada dividindo a quantidade de dados
1184
+ compactados processados até o momento pela quantidade de dados não
1185
+ compactados processados até o momento.
1186
+ .IP \(bu 3
1187
+ Velocidade de compactação ou descompactação. Isso é medido como a quantidade
1188
+ de dados não compactados consumidos (compactação) ou produzidos
1189
+ (descompactação) por segundo. É mostrado após alguns segundos desde que
1190
+ \fBxz\fP começou a processar o arquivo.
1191
+ .IP \(bu 3
1192
+ Tempo decorrido no formato M:SS ou H:MM:SS.
1193
+ .IP \(bu 3
1194
+ O tempo restante estimado é mostrado apenas quando o tamanho do arquivo de
1195
+ entrada é conhecido e alguns segundos já se passaram desde que \fBxz\fP começou
1196
+ a processar o arquivo. A hora é mostrada em um formato menos preciso que
1197
+ nunca tem dois pontos, por exemplo, 2 min 30 s.
1198
+ .RE
1199
+ .IP ""
1200
+ Quando o erro padrão não é um terminal, \fB\-\-verbose\fP fará com que \fBxz\fP
1201
+ imprima o nome do arquivo, tamanho compactado, tamanho não compactado, taxa
1202
+ de compactação e possivelmente também a velocidade e o tempo decorrido em
1203
+ uma única linha para o erro padrão após a compactação ou descompactando o
1204
+ arquivo. A velocidade e o tempo decorrido são incluídos apenas quando a
1205
+ operação leva pelo menos alguns segundos. Se a operação não foi concluída,
1206
+ por exemplo, devido à interrupção do usuário, também é impressa a
1207
+ porcentagem de conclusão se o tamanho do arquivo de entrada for conhecido.
1208
+ .TP
1209
+ \fB\-Q\fP, \fB\-\-no\-warn\fP
1210
+ Não define o status de saída como 2, mesmo que uma condição digna de um
1211
+ aviso tenha sido detectada. Esta opção não afeta o nível de detalhamento,
1212
+ portanto, tanto \fB\-\-quiet\fP quanto \fB\-\-no\-warn\fP devem ser usados para não
1213
+ exibir avisos e não alterar o status de saída.
1214
+ .TP
1215
+ \fB\-\-robot\fP
1216
+ Imprime mensagens em um formato analisável por máquina. Isso visa facilitar
1217
+ a criação de frontends que desejam usar \fBxz\fP em vez de liblzma, o que pode
1218
+ ser o caso de vários scripts. A saída com esta opção habilitada deve ser
1219
+ estável em versões \fBxz\fP. Consulte a seção \fBMODO ROBÔ\fP para obter detalhes.
1220
+ .TP
1221
+ \fB\-\-info\-memory\fP
1222
+ Exibe, em formato legível por humanos, quanta memória física (RAM) e quantos
1223
+ threads de processador \fBxz\fP acredita que o sistema possui e os limites de
1224
+ uso de memória para compactação e descompactação e saia com êxito.
1225
+ .TP
1226
+ \fB\-h\fP, \fB\-\-help\fP
1227
+ Exibe uma mensagem de ajuda descrevendo as opções mais usadas e sai com
1228
+ sucesso.
1229
+ .TP
1230
+ \fB\-H\fP, \fB\-\-long\-help\fP
1231
+ Exibe uma mensagem de ajuda descrevendo todos os recursos de \fBxz\fP e sai com
1232
+ sucesso
1233
+ .TP
1234
+ \fB\-V\fP, \fB\-\-version\fP
1235
+ Exibe o número da versão de \fBxz\fP e liblzma em formato legível por
1236
+ humanos. Para obter uma saída analisável por máquina, especifique \fB\-\-robot\fP
1237
+ antes de \fB\-\-version\fP.
1238
+ .
1239
+ .SH "MODO ROBÔ"
1240
+ The robot mode is activated with the \fB\-\-robot\fP option. It makes the output
1241
+ of \fBxz\fP easier to parse by other programs. Currently \fB\-\-robot\fP is
1242
+ supported only together with \fB\-\-list\fP, \fB\-\-filters\-help\fP, \fB\-\-info\-memory\fP,
1243
+ and \fB\-\-version\fP. It will be supported for compression and decompression in
1244
+ the future.
1245
+ .
1246
+ .SS "Modo lista"
1247
+ \fBxz \-\-robot \-\-list\fP usa saída separada por tabulações. A primeira coluna de
1248
+ cada linha possui uma string que indica o tipo de informação encontrada
1249
+ naquela linha:
1250
+ .TP
1251
+ \fBname\fP
1252
+ Esta é sempre a primeira linha ao começar a listar um arquivo. A segunda
1253
+ coluna na linha é o nome do arquivo.
1254
+ .TP
1255
+ \fBfile\fP
1256
+ Esta linha contém informações gerais sobre o arquivo \fB.xz\fP. Esta linha é
1257
+ sempre impressa após a linha \fBname\fP.
1258
+ .TP
1259
+ \fBstream\fP
1260
+ Este tipo de linha é usado somente quando \fB\-\-verbose\fP foi
1261
+ especificado. Existem tantas linhas de \fBstream\fP quanto fluxos no arquivo
1262
+ \&\fB.xz\fP.
1263
+ .TP
1264
+ \fBblock\fP
1265
+ Este tipo de linha é usado somente quando \fB\-\-verbose\fP foi
1266
+ especificado. Existem tantas linhas \fBblock\fP quanto blocos no arquivo
1267
+ \&\fB.xz\fP. As linhas \fBblock\fP são mostradas após todas as linhas \fBstream\fP;
1268
+ diferentes tipos de linha não são intercalados.
1269
+ .TP
1270
+ \fBsummary\fP
1271
+ Este tipo de linha é usado apenas quando \fB\-\-verbose\fP foi especificado duas
1272
+ vezes. Esta linha é impressa após todas as linhas de \fBblock\fP. Assim como a
1273
+ linha \fBarquivo\fP, a linha \fBsummary\fP contém informações gerais sobre o
1274
+ arquivo \fB.xz\fP.
1275
+ .TP
1276
+ \fBtotals\fP
1277
+ Esta linha é sempre a última linha da saída da lista. Ele mostra as
1278
+ contagens totais e tamanhos.
1279
+ .PP
1280
+ As colunas das linhas \fBfile\fP:
1281
+ .PD 0
1282
+ .RS
1283
+ .IP 2. 4
1284
+ Número de fluxos no arquivo
1285
+ .IP 3. 4
1286
+ Número total de blocos no(s) fluxo(s)
1287
+ .IP 4. 4
1288
+ Tamanho compactado do arquivo
1289
+ .IP 5. 4
1290
+ Uncompressed size of the file
1291
+ .IP 6. 4
1292
+ Taxa de compactação, por exemplo, \fB0.123\fP. Se a proporção for superior a
1293
+ 9.999, serão exibidos três traços (\fB\-\-\-\fP) em vez da proporção.
1294
+ .IP 7. 4
1295
+ Lista separada por vírgulas de nomes de verificação de integridade. As
1296
+ seguintes strings são usadas para os tipos de verificação conhecidos:
1297
+ \fBNone\fP, \fBCRC32\fP, \fBCRC64\fP e \fBSHA\-256\fP. Para tipos de verificações
1298
+ desconhecidos, \fBUnknown\-\fP\fIN\fP é usado, onde \fIN\fP é o ID do cheque como um
1299
+ número decimal (um ou dois dígitos).
1300
+ .IP 8. 4
1301
+ Tamanho total do preenchimento de fluxo no arquivo
1302
+ .RE
1303
+ .PD
1304
+ .PP
1305
+ As colunas das linhas \fBstream\fP:
1306
+ .PD 0
1307
+ .RS
1308
+ .IP 2. 4
1309
+ Número do fluxo (o primeiro fluxo é 1)
1310
+ .IP 3. 4
1311
+ Número de blocos no fluxo
1312
+ .IP 4. 4
1313
+ Deslocamento inicial compactado
1314
+ .IP 5. 4
1315
+ Deslocamento inicial descompactado
1316
+ .IP 6. 4
1317
+ Tamanho compactado (não inclui preenchimento de fluxo)
1318
+ .IP 7. 4
1319
+ Tamanho descompactado
1320
+ .IP 8. 4
1321
+ Taxa de compactação
1322
+ .IP 9. 4
1323
+ Nome da verificação de integridade
1324
+ .IP 10. 4
1325
+ Tamanho do preenchimento do fluxo
1326
+ .RE
1327
+ .PD
1328
+ .PP
1329
+ As colunas das linhas \fBblock\fP:
1330
+ .PD 0
1331
+ .RS
1332
+ .IP 2. 4
1333
+ Número do fluxo que contém este bloco
1334
+ .IP 3. 4
1335
+ Número do bloco relativo ao início do fluxo (o primeiro bloco é 1)
1336
+ .IP 4. 4
1337
+ Número do bloco relativo ao início do arquivo
1338
+ .IP 5. 4
1339
+ Deslocamento inicial compactado em relação ao início do arquivo
1340
+ .IP 6. 4
1341
+ Deslocamento inicial descompactado em relação ao início do arquivo
1342
+ .IP 7. 4
1343
+ Tamanho total compactado do bloco (inclui cabeçalhos)
1344
+ .IP 8. 4
1345
+ Tamanho descompactado
1346
+ .IP 9. 4
1347
+ Taxa de compactação
1348
+ .IP 10. 4
1349
+ Nome da verificação de integridade
1350
+ .RE
1351
+ .PD
1352
+ .PP
1353
+ Se \fB\-\-verbose\fP for especificado duas vezes, colunas adicionais serão
1354
+ incluídas nas linhas \fBblock\fP. Eles não são exibidos com um único
1355
+ \fB\-\-verbose\fP, porque obter essas informações requer muitas buscas e,
1356
+ portanto, pode ser lento:
1357
+ .PD 0
1358
+ .RS
1359
+ .IP 11. 4
1360
+ Valor da verificação de integridade em hexadecimal
1361
+ .IP 12. 4
1362
+ Tamanho do cabeçalho do bloco
1363
+ .IP 13. 4
1364
+ Sinalizadores de bloco: \fBc\fP indica que o tamanho compactado está presente e
1365
+ \fBu\fP indica que o tamanho não compactado está presente. Se o sinalizador não
1366
+ estiver definido, um traço (\fB\-\fP) será exibido para manter o comprimento da
1367
+ string fixo. Novos sinalizadores podem ser adicionados ao final da string no
1368
+ futuro.
1369
+ .IP 14. 4
1370
+ Tamanho dos dados reais compactados no bloco (isso exclui o cabeçalho do
1371
+ bloco, o preenchimento do bloco e os campos de verificação)
1372
+ .IP 15. 4
1373
+ Quantidade de memória (em bytes) necessária para descompactar este bloco com
1374
+ esta versão \fBxz\fP
1375
+ .IP 16. 4
1376
+ Cadeia de filtro. Observe que a maioria das opções usadas no momento da
1377
+ compactação não pode ser conhecida, pois apenas as opções necessárias para a
1378
+ descompactação são armazenadas nos cabeçalhos \fB.xz\fP.
1379
+ .RE
1380
+ .PD
1381
+ .PP
1382
+ As colunas das linhas \fBsummary\fP:
1383
+ .PD 0
1384
+ .RS
1385
+ .IP 2. 4
1386
+ Quantidade de memória (em bytes) necessária para descompactar este arquivo
1387
+ com esta versão do \fBxz\fP
1388
+ .IP 3. 4
1389
+ \fByes\fP ou \fBno\fP indicando se todos os cabeçalhos de bloco têm tamanho
1390
+ compactado e tamanho não compactado armazenados neles
1391
+ .PP
1392
+ \fIDesde\fP \fBxz\fP \fI5.1.2alpha:\fP
1393
+ .IP 4. 4
1394
+ Versão mínima do \fBxz\fP necessária para descompactar o arquivo
1395
+ .RE
1396
+ .PD
1397
+ .PP
1398
+ As colunas da linha \fBtotals\fP:
1399
+ .PD 0
1400
+ .RS
1401
+ .IP 2. 4
1402
+ Número de fluxos
1403
+ .IP 3. 4
1404
+ Número de blocos
1405
+ .IP 4. 4
1406
+ Tamanho compactado
1407
+ .IP 5. 4
1408
+ Tamanho descompactado
1409
+ .IP 6. 4
1410
+ Taxa de compactação média
1411
+ .IP 7. 4
1412
+ Lista separada por vírgulas de nomes de verificação de integridade que
1413
+ estavam presentes nos arquivos
1414
+ .IP 8. 4
1415
+ Tamanho do preenchimento do fluxo
1416
+ .IP 9. 4
1417
+ Número de arquivos. Isso está aqui para manter a ordem das colunas
1418
+ anteriores a mesma das linhas \fBfile\fP.
1419
+ .PD
1420
+ .RE
1421
+ .PP
1422
+ Se \fB\-\-verbose\fP for especificado duas vezes, colunas adicionais serão
1423
+ incluídas na linha \fBtotals\fP:
1424
+ .PD 0
1425
+ .RS
1426
+ .IP 10. 4
1427
+ Quantidade máxima de memória (em bytes) necessária para descompactar os
1428
+ arquivos com esta versão do \fBxz\fP
1429
+ .IP 11. 4
1430
+ \fByes\fP ou \fBno\fP indicando se todos os cabeçalhos de bloco têm tamanho
1431
+ compactado e tamanho não compactado armazenados neles
1432
+ .PP
1433
+ \fIDesde\fP \fBxz\fP \fI5.1.2alpha:\fP
1434
+ .IP 12. 4
1435
+ Versão mínima do \fBxz\fP necessária para descompactar o arquivo
1436
+ .RE
1437
+ .PD
1438
+ .PP
1439
+ Versões futuras podem adicionar novos tipos de linha e novas colunas podem
1440
+ ser adicionadas aos tipos de linha existentes, mas as colunas existentes não
1441
+ serão alteradas.
1442
+ .
1443
+ .SS "Filters help"
1444
+ \fBxz \-\-robot \-\-filters\-help\fP prints the supported filters in the following
1445
+ format:
1446
+ .PP
1447
+ \fIfilter\fP\fB:\fP\fIoption\fP\fB=<\fP\fIvalue\fP\fB>,\fP\fIoption\fP\fB=<\fP\fIvalue\fP\fB>\fP...
1448
+ .TP
1449
+ \fIfilter\fP
1450
+ Name of the filter
1451
+ .TP
1452
+ \fIoption\fP
1453
+ Name of a filter specific option
1454
+ .TP
1455
+ \fIvalue\fP
1456
+ Numeric \fIvalue\fP ranges appear as \fB<\fP\fImin\fP\fB\-\fP\fImax\fP\fB>\fP. String
1457
+ \fIvalue\fP choices are shown within \fB< >\fP and separated by a \fB|\fP
1458
+ character.
1459
+ .PP
1460
+ Each filter is printed on its own line.
1461
+ .
1462
+ .SS "Informações de limite de memória"
1463
+ \fBxz \-\-robot \-\-info\-memory\fP prints a single line with multiple tab\-separated
1464
+ columns:
1465
+ .IP 1. 4
1466
+ Quantidade total de memória física (RAM) em bytes.
1467
+ .IP 2. 4
1468
+ Limite de uso de memória para compactação em bytes
1469
+ (\fB\-\-memlimit\-compress\fP). Um valor especial de \fB0\fP indica a configuração
1470
+ padrão que para o modo de thread única é o mesmo que sem limite.
1471
+ .IP 3. 4
1472
+ Limite de uso de memória para descompactação em bytes
1473
+ (\fB\-\-memlimit\-decompress\fP). Um valor especial de \fB0\fP indica a configuração
1474
+ padrão que para o modo de thread única é o mesmo que sem limite.
1475
+ .IP 4. 4
1476
+ Desde \fBxz\fP 5.3.4alpha: Uso de memória para descompactação com várias thread
1477
+ em bytes (\fB\-\-memlimit\-mt\-decompress\fP). Isso nunca é zero porque um valor
1478
+ padrão específico do sistema mostrado na coluna 5 é usado se nenhum limite
1479
+ for especificado explicitamente. Isso também nunca é maior que o valor na
1480
+ coluna 3, mesmo que um valor maior tenha sido especificado com
1481
+ \fB\-\-memlimit\-mt\-decompress\fP.
1482
+ .IP 5. 4
1483
+ Desde \fBxz\fP 5.3.4alpha: Um limite de uso de memória padrão específico do
1484
+ sistema que é usado para limitar o número de threads ao compactar com um
1485
+ número automático de threads (\fB\-\-threads=0\fP) e nenhum limite de uso de
1486
+ memória foi especificado (\fB\-\-memlimit\-compress\fP). Isso também é usado como
1487
+ o valor padrão para \fB\-\-memlimit\-mt\-decompress\fP.
1488
+ .IP 6. 4
1489
+ Desde \fBxz\fP 5.3.4alpha: Número de threads de processador disponíveis.
1490
+ .PP
1491
+ No futuro, a saída de \fBxz \-\-robot \-\-info\-memory\fP pode ter mais colunas, mas
1492
+ nunca mais do que uma única linha.
1493
+ .
1494
+ .SS Versão
1495
+ \fBxz \-\-robot \-\-version\fP prints the version number of \fBxz\fP and liblzma in
1496
+ the following format:
1497
+ .PP
1498
+ \fBXZ_VERSION=\fP\fIXYYYZZZS\fP
1499
+ .br
1500
+ \fBLIBLZMA_VERSION=\fP\fIXYYYZZZS\fP
1501
+ .TP
1502
+ \fIX\fP
1503
+ Versão principal.
1504
+ .TP
1505
+ \fIYYY\fP
1506
+ Versão menor. Números pares são estáveis. Os números ímpares são versões
1507
+ alfa ou beta.
1508
+ .TP
1509
+ \fIZZZ\fP
1510
+ Nível de patch para versões estáveis ou apenas um contador para versões de
1511
+ desenvolvimento.
1512
+ .TP
1513
+ \fIS\fP
1514
+ Estabilidade. 0 é alfa, 1 é beta e 2 é estável. \fIS\fP deve ser sempre 2
1515
+ quando \fIYYY\fP for par.
1516
+ .PP
1517
+ \fIXYYYZZZS\fP são iguais em ambas as linhas se \fBxz\fP e liblzma forem da mesma
1518
+ versão do XZ Utils.
1519
+ .PP
1520
+ Exemplos: 4.999.9beta é \fB49990091\fP e 5.0.0 é \fB50000002\fP.
1521
+ .
1522
+ .SH "STATUS DE SAÍDA"
1523
+ .TP
1524
+ \fB0\fP
1525
+ Está tudo bem.
1526
+ .TP
1527
+ \fB1\fP
1528
+ Ocorreu um erro.
1529
+ .TP
1530
+ \fB2\fP
1531
+ Algo digno de um aviso ocorreu, mas ocorreu nenhum erro real.
1532
+ .PP
1533
+ Observações (não avisos ou erros) impressas no erro padrão não afetam o
1534
+ status de saída.
1535
+ .
1536
+ .SH AMBIENTE
1537
+ \fBxz\fP analisa listas de opções separadas por espaços das variáveis de
1538
+ ambiente \fBXZ_DEFAULTS\fP e \fBXZ_OPT\fP, nesta ordem, antes de analisar as
1539
+ opções da linha de comando. Observe que apenas as opções são analisadas a
1540
+ partir das variáveis de ambiente; todas as não opções são silenciosamente
1541
+ ignoradas. A análise é feita com \fBgetopt_long\fP(3) que também é usado para
1542
+ os argumentos da linha de comando.
1543
+ .TP
1544
+ \fBXZ_DEFAULTS\fP
1545
+ Opções padrão específicas do usuário ou de todo o sistema. Normalmente, isso
1546
+ é definido em um script de inicialização do shell para habilitar o limitador
1547
+ de uso de memória do \fBxz\fP por padrão. Excluindo scripts de inicialização de
1548
+ shell e casos especiais semelhantes, os scripts nunca devem definir ou
1549
+ remover a definição de \fBXZ_DEFAULTS\fP.
1550
+ .TP
1551
+ \fBXZ_OPT\fP
1552
+ Isso é para passar opções para \fBxz\fP quando não é possível definir as opções
1553
+ diretamente na linha de comando \fBxz\fP. Este é o caso quando \fBxz\fP é
1554
+ executado por um script ou ferramenta, por exemplo, GNU \fBtar\fP(1):
1555
+ .RS
1556
+ .RS
1557
+ .PP
1558
+ .nf
1559
+ \f(CRXZ_OPT=\-2v tar caf foo.tar.xz foo\fR
1560
+ .fi
1561
+ .RE
1562
+ .RE
1563
+ .IP ""
1564
+ Os scripts podem usar \fBXZ_OPT\fP, por exemplo, para definir opções de
1565
+ compactação padrão específicas do script. Ainda é recomendável permitir que
1566
+ os usuários substituam \fBXZ_OPT\fP se isso for razoável. Por exemplo, em
1567
+ scripts \fBsh\fP(1) pode\-se usar algo assim:
1568
+ .RS
1569
+ .RS
1570
+ .PP
1571
+ .nf
1572
+ \f(CRXZ_OPT=${XZ_OPT\-"\-7e"}
1573
+ export XZ_OPT\fR
1574
+ .fi
1575
+ .RE
1576
+ .RE
1577
+ .
1578
+ .SH "COMPATIBILIDADE COM LZMA UTILS"
1579
+ A sintaxe da linha de comando do \fBxz\fP é praticamente um superconjunto de
1580
+ \fBlzma\fP, \fBunlzma\fP e \fBlzcat\fP conforme encontrado no LZMA Utils 4.32.x. Na
1581
+ maioria dos casos, é possível substituir LZMA Utils por XZ Utils sem
1582
+ interromper os scripts existentes. Existem algumas incompatibilidades,
1583
+ porém, que às vezes podem causar problemas.
1584
+ .
1585
+ .SS "Níveis de predefinição de compactação"
1586
+ A numeração das predefinições de nível de compactação não é idêntica em
1587
+ \fBxz\fP e LZMA Utils. A diferença mais importante é como os tamanhos dos
1588
+ dicionários são mapeados para diferentes predefinições. O tamanho do
1589
+ dicionário é aproximadamente igual ao uso de memória do descompactador.
1590
+ .RS
1591
+ .PP
1592
+ .TS
1593
+ tab(;);
1594
+ c c c
1595
+ c n n.
1596
+ Nível;xz;LZMA Utils
1597
+ \-0;256 KiB;N/D
1598
+ \-1;1 MiB;64 KiB
1599
+ \-2;2 MiB;1 MiB
1600
+ \-3;4 MiB;512 KiB
1601
+ \-4;4 MiB;1 MiB
1602
+ \-5;8 MiB;2 MiB
1603
+ \-6;8 MiB;4 MiB
1604
+ \-7;16 MiB;8 MiB
1605
+ \-8;32 MiB;16 MiB
1606
+ \-9;64 MiB;32 MiB
1607
+ .TE
1608
+ .RE
1609
+ .PP
1610
+ As diferenças de tamanho do dicionário também afetam o uso da memória do
1611
+ compressor, mas existem algumas outras diferenças entre LZMA Utils e XZ
1612
+ Utils, que tornam a diferença ainda maior:
1613
+ .RS
1614
+ .PP
1615
+ .TS
1616
+ tab(;);
1617
+ c c c
1618
+ c n n.
1619
+ Nível;xz;LZMA Utils 4.32.x
1620
+ \-0;3 MiB;N/D
1621
+ \-1;9 MiB;2 MiB
1622
+ \-2;17 MiB;12 MiB
1623
+ \-3;32 MiB;12 MiB
1624
+ \-4;48 MiB;16 MiB
1625
+ \-5;94 MiB;26 MiB
1626
+ \-6;94 MiB;45 MiB
1627
+ \-7;186 MiB;83 MiB
1628
+ \-8;370 MiB;159 MiB
1629
+ \-9;674 MiB;311 MiB
1630
+ .TE
1631
+ .RE
1632
+ .PP
1633
+ O nível de predefinição padrão no LZMA Utils é \fB\-7\fP enquanto no XZ Utils é
1634
+ \fB\-6\fP, então ambos usam um dicionário de 8 MiB por padrão.
1635
+ .
1636
+ .SS "Arquivos .lzma em um fluxo versus sem ser em um fluxo"
1637
+ O tamanho descompactado do arquivo pode ser armazenado no cabeçalho de
1638
+ \&\fB.lzma\fP. O LZMA Utils faz isso ao compactar arquivos comuns. A alternativa
1639
+ é marcar que o tamanho não compactado é desconhecido e usar o marcador de
1640
+ fim de carga útil para indicar onde o descompactador deve parar. O LZMA
1641
+ Utils usa este método quando o tamanho não compactado não é conhecido, como
1642
+ é o caso, por exemplo, de encadeamentos (pipes).
1643
+ .PP
1644
+ \fBxz\fP oferece suporte à descompactação de arquivos \fB.lzma\fP com ou sem
1645
+ marcador de fim de carga útil, mas todos os arquivos \fB.lzma\fP criados por
1646
+ \fBxz\fP usarão marcador de fim de carga útil e terão o tamanho descompactado
1647
+ marcado como desconhecido no cabeçalho de \fB.lzma\fP. Isso pode ser um
1648
+ problema em algumas situações incomuns. Por exemplo, um descompactador de
1649
+ \&\fB.lzma\fP em um dispositivo embarcado pode funcionar apenas com arquivos que
1650
+ tenham tamanho descompactado conhecido. Se você encontrar esse problema,
1651
+ precisará usar o LZMA Utils ou o LZMA SDK para criar arquivos \fB.lzma\fP com
1652
+ tamanho descompactado conhecido.
1653
+ .
1654
+ .SS "Arquivos .lzma não suportados"
1655
+ O formato \fB.lzma\fP permite valores \fIlc\fP até 8 e valores \fIlp\fP até 4. LZMA
1656
+ Utils pode descompactar arquivos com qualquer \fIlc\fP e \fIlp\fP, mas sempre cria
1657
+ arquivos com \fBlc=3\fP e \fBlp=0\fP. Criar arquivos com outros \fIlc\fP e \fIlp\fP é
1658
+ possível com \fBxz\fP e com LZMA SDK.
1659
+ .PP
1660
+ A implementação do filtro LZMA1 em liblzma requer que a soma de \fIlc\fP e
1661
+ \fIlp\fP não exceda 4. Assim, arquivos \fB.lzma\fP, que excedam esta limitação,
1662
+ não podem ser descompactados com \fBxz\fP.
1663
+ .PP
1664
+ LZMA Utils cria apenas arquivos \fB.lzma\fP que possuem um tamanho de
1665
+ dicionário de 2^\fIn\fP (uma potência de 2), mas aceita arquivos com qualquer
1666
+ tamanho de dicionário. liblzma aceita apenas arquivos \fB.lzma\fP que tenham um
1667
+ tamanho de dicionário de 2^\fIn\fP ou 2^\fIn\fP + 2^(\fIn\fP\-1). Isso é para diminuir
1668
+ os falsos positivos ao detectar arquivos \fB.lzma\fP.
1669
+ .PP
1670
+ Essas limitações não devem ser um problema na prática, já que praticamente
1671
+ todos os arquivos \fB.lzma\fP foram compactados com configurações que o liblzma
1672
+ aceitará.
1673
+ .
1674
+ .SS "Lixo à direita"
1675
+ Ao descompactar, o LZMA Utils silenciosamente ignora tudo após o primeiro
1676
+ fluxo \fB.lzma\fP. Na maioria das situações, isso é um bug. Isso também
1677
+ significa que o LZMA Utils não oferece suporte a descompactação de arquivos
1678
+ \&\fB.lzma\fP concatenados.
1679
+ .PP
1680
+ Se houver dados restantes após o primeiro fluxo de \fB.lzma\fP, \fBxz\fP considera
1681
+ o arquivo corrompido, a menos que \fB\-\-single\-stream\fP tenha sido usado. Isso
1682
+ pode quebrar scripts obscuros que presumiram que o lixo à direita é
1683
+ ignorado.
1684
+ .
1685
+ .SH NOTAS
1686
+ .
1687
+ .SS "A saída compactada pode variar"
1688
+ A saída compactada exata produzida a partir do mesmo arquivo de entrada não
1689
+ compactado pode variar entre as versões do XZ Utils, mesmo se as opções de
1690
+ compactação forem idênticas. Isso ocorre porque o codificador pode ser
1691
+ aprimorado (compactação mais rápida ou melhor) sem afetar o formato do
1692
+ arquivo. A saída pode variar mesmo entre diferentes compilações da mesma
1693
+ versão do XZ Utils, se diferentes opções de compilação forem usadas.
1694
+ .PP
1695
+ A informação acima significa que, uma vez que \fB\-\-rsyncable\fP tenha sido
1696
+ implementado, os arquivos resultantes não serão necessariamente
1697
+ "rsyncáveis", a menos que os arquivos antigos e novos tenham sido
1698
+ compactados com a mesma versão xz. Esse problema pode ser corrigido se uma
1699
+ parte da implementação do codificador for congelada para manter a saída de
1700
+ rsyncable estável nas versões do xz.
1701
+ .
1702
+ .SS "Descompactadores .xz embarcados"
1703
+ As implementações do descompactador \fB.xz\fP embarcados, como o XZ Embedded,
1704
+ não oferecem necessariamente suporte a arquivos criados com tipos de
1705
+ \fIverificações\fP de integridade diferentes de \fBnone\fP e \fBcrc32\fP. Como o
1706
+ padrão é \fB\-\-check=crc64\fP, você deve usar \fB\-\-check=none\fP ou
1707
+ \fB\-\-check=crc32\fP ao criar arquivos para sistemas embarcados.
1708
+ .PP
1709
+ Fora dos sistemas embarcados, todos os descompactadores de formato \fB.xz\fP
1710
+ oferecem suporte a todos os tipos de \fIverificação\fP ou, pelo menos, são
1711
+ capazes de descompactar o arquivo sem verificar a verificação de integridade
1712
+ se a \fIverificação\fP específica não for suportada.
1713
+ .PP
1714
+ XZ Embedded oferece suporte a filtros BCJ, mas apenas com o deslocamento
1715
+ inicial padrão.
1716
+ .
1717
+ .SH EXEMPLOS
1718
+ .
1719
+ .SS Básico
1720
+ Compactar o arquivo \fIfoo\fP em \fIfoo.xz\fP usando o nível de compactação padrão
1721
+ (\fB\-6\fP) e remover \fIfoo\fP se a compactação for bem\-sucedida:
1722
+ .RS
1723
+ .PP
1724
+ .nf
1725
+ \f(CRxz foo\fR
1726
+ .fi
1727
+ .RE
1728
+ .PP
1729
+ Descompactar \fIbar.xz\fP em \fIbar\fP e não remover \fIbar.xz\fP mesmo se a
1730
+ descompactação for bem\-sucedida:
1731
+ .RS
1732
+ .PP
1733
+ .nf
1734
+ \f(CRxz \-dk bar.xz\fR
1735
+ .fi
1736
+ .RE
1737
+ .PP
1738
+ Criar \fIbaz.tar.xz\fP com a predefinição \fB\-4e\fP (\fB\-4 \-\-extreme\fP), que é mais
1739
+ lenta que o padrão \fB\-6\fP, mas precisa de menos memória para compactação e
1740
+ descompactação (48 \ MiB e 5\ MiB, respectivamente):
1741
+ .RS
1742
+ .PP
1743
+ .nf
1744
+ \f(CRtar cf \- baz | xz \-4e > baz.tar.xz\fR
1745
+ .fi
1746
+ .RE
1747
+ .PP
1748
+ Uma mistura de arquivos compactados e descompactados pode ser descompactada
1749
+ para a saída padrão com um único comando:
1750
+ .RS
1751
+ .PP
1752
+ .nf
1753
+ \f(CRxz \-dcf a.txt b.txt.xz c.txt d.txt.lzma > abcd.txt\fR
1754
+ .fi
1755
+ .RE
1756
+ .
1757
+ .SS "Compactação paralela de muitos arquivos"
1758
+ No GNU e *BSD, \fBfind\fP(1) e \fBxargs\fP(1) podem ser usados para paralelizar a
1759
+ compactação de muitos arquivos:
1760
+ .RS
1761
+ .PP
1762
+ .nf
1763
+ \f(CRfind . \-type f \e! \-name '*.xz' \-print0 \e
1764
+ | xargs \-0r \-P4 \-n16 xz \-T1\fR
1765
+ .fi
1766
+ .RE
1767
+ .PP
1768
+ A opção \fB\-P\fP para \fBxargs\fP(1) define o número de processos paralelos do
1769
+ \fBxz\fP. O melhor valor para a opção \fB\-n\fP depende de quantos arquivos devem
1770
+ ser compactados. Se houver apenas alguns arquivos, o valor provavelmente
1771
+ deve ser 1; com dezenas de milhares de arquivos, 100 ou até mais podem ser
1772
+ apropriados para reduzir o número de processos de \fBxz\fP que \fBxargs\fP(1)
1773
+ eventualmente criará.
1774
+ .PP
1775
+ A opção \fB\-T1\fP para \fBxz\fP existe para forçá\-lo ao modo de thread única,
1776
+ porque \fBxargs\fP(1) é usado para controlar a quantidade de paralelização.
1777
+ .
1778
+ .SS "Modo robô"
1779
+ Calcular quantos bytes foram salvos no total depois de compactar vários
1780
+ arquivos:
1781
+ .RS
1782
+ .PP
1783
+ .nf
1784
+ \f(CRxz \-\-robot \-\-list *.xz | awk '/^totals/{print $5\-$4}'\fR
1785
+ .fi
1786
+ .RE
1787
+ .PP
1788
+ Um script pode querer saber que está usando \fBxz\fP novo o suficiente. O
1789
+ seguinte script \fBsh\fP(1) verifica se o número da versão da ferramenta \fBxz\fP
1790
+ é pelo menos 5.0.0. Este método é compatível com versões beta antigas, que
1791
+ não suportavam a opção \fB\-\-robot\fP:
1792
+ .RS
1793
+ .PP
1794
+ .nf
1795
+ \f(CRif ! eval "$(xz \-\-robot \-\-version 2> /dev/null)" ||
1796
+ [ "$XZ_VERSION" \-lt 50000002 ]; then
1797
+ echo "Your xz is too old."
1798
+ fi
1799
+ unset XZ_VERSION LIBLZMA_VERSION\fR
1800
+ .fi
1801
+ .RE
1802
+ .PP
1803
+ Definir um limite de uso de memória para descompactação usando \fBXZ_OPT\fP,
1804
+ mas se um limite já tiver sido definido, não o aumentar:
1805
+ .RS
1806
+ .PP
1807
+ .nf
1808
+ \f(CRNEWLIM=$((123 << 20))\ \ # 123 MiB
1809
+ OLDLIM=$(xz \-\-robot \-\-info\-memory | cut \-f3)
1810
+ if [ $OLDLIM \-eq 0 \-o $OLDLIM \-gt $NEWLIM ]; then
1811
+ XZ_OPT="$XZ_OPT \-\-memlimit\-decompress=$NEWLIM"
1812
+ export XZ_OPT
1813
+ fi\fR
1814
+ .fi
1815
+ .RE
1816
+ .
1817
+ .SS "Cadeias de filtro de compressor personalizadas"
1818
+ O uso mais simples para cadeias de filtro personalizadas é personalizar uma
1819
+ predefinição LZMA2. Isso pode ser útil, porque as predefinições abrangem
1820
+ apenas um subconjunto das combinações potencialmente úteis de configurações
1821
+ de compactação.
1822
+ .PP
1823
+ As colunas CompCPU das tabelas das descrições das opções \fB\-0\fP ... \fB\-9\fP e
1824
+ \fB\-\-extreme\fP são úteis ao personalizar as predefinições LZMA2. Aqui estão as
1825
+ partes relevantes coletadas dessas duas tabelas:
1826
+ .RS
1827
+ .PP
1828
+ .TS
1829
+ tab(;);
1830
+ c c
1831
+ n n.
1832
+ Predefinição;CompCPU
1833
+ \-0;0
1834
+ \-1;1
1835
+ \-2;2
1836
+ \-3;3
1837
+ \-4;4
1838
+ \-5;5
1839
+ \-6;6
1840
+ \-5e;7
1841
+ \-6e;8
1842
+ .TE
1843
+ .RE
1844
+ .PP
1845
+ Se você sabe que um arquivo requer um dicionário um tanto grande (por
1846
+ exemplo, 32\ MiB) para compactar bem, mas deseja comprimi\-lo mais
1847
+ rapidamente do que \fBxz \-8\fP faria, uma predefinição com um valor CompCPU
1848
+ baixo (por exemplo, 1) pode ser modificado para usar um dicionário maior:
1849
+ .RS
1850
+ .PP
1851
+ .nf
1852
+ \f(CRxz \-\-lzma2=preset=1,dict=32MiB foo.tar\fR
1853
+ .fi
1854
+ .RE
1855
+ .PP
1856
+ Com certos arquivos, o comando acima pode ser mais rápido que \fBxz \-6\fP
1857
+ enquanto compacta significativamente melhor. No entanto, deve\-se enfatizar
1858
+ que apenas alguns arquivos se beneficiam de um grande dicionário, mantendo o
1859
+ valor CompCPU baixo. A situação mais óbvia, onde um grande dicionário pode
1860
+ ajudar muito, é um arquivo contendo arquivos muito semelhantes de pelo menos
1861
+ alguns megabytes cada. O tamanho do dicionário deve ser significativamente
1862
+ maior do que qualquer arquivo individual para permitir que o LZMA2 aproveite
1863
+ ao máximo as semelhanças entre arquivos consecutivos.
1864
+ .PP
1865
+ Se o uso muito alto de memória do compactador e do descompactador for bom e
1866
+ o arquivo que está sendo compactado tiver pelo menos várias centenas de
1867
+ megabytes, pode ser útil usar um dicionário ainda maior do que os 64 MiB que
1868
+ o \fBxz \-9\fP usaria:
1869
+ .RS
1870
+ .PP
1871
+ .nf
1872
+ \f(CRxz \-vv \-\-lzma2=dict=192MiB big_foo.tar\fR
1873
+ .fi
1874
+ .RE
1875
+ .PP
1876
+ Usar \fB\-vv\fP (\fB\-\-verbose \-\-verbose\fP) como no exemplo acima pode ser útil
1877
+ para ver os requisitos de memória do compactador e do
1878
+ descompactador. Lembre\-se que usar um dicionário maior que o tamanho do
1879
+ arquivo descompactado é desperdício de memória, então o comando acima não é
1880
+ útil para arquivos pequenos.
1881
+ .PP
1882
+ Às vezes, o tempo de compactação não importa, mas o uso de memória do
1883
+ descompactador deve ser mantido baixo, por exemplo, para possibilitar a
1884
+ descompactação do arquivo em um sistema embarcado. O comando a seguir usa
1885
+ \fB\-6e\fP (\fB\-6 \-\-extreme\fP) como base e define o dicionário como apenas 64\ KiB. O arquivo resultante pode ser descompactado com XZ Embedded (é por isso
1886
+ que existe \fB\-\-check=crc32\fP) usando cerca de 100\ KiB de memória.
1887
+ .RS
1888
+ .PP
1889
+ .nf
1890
+ \f(CRxz \-\-check=crc32 \-\-lzma2=preset=6e,dict=64KiB foo\fR
1891
+ .fi
1892
+ .RE
1893
+ .PP
1894
+ Se você deseja espremer o máximo de bytes possível, ajustar o número de bits
1895
+ de contexto literal (\fIlc\fP) e o número de bits de posição (\fIpb\fP) às vezes
1896
+ pode ajudar. Ajustar o número de bits de posição literal (\fIlp\fP) também pode
1897
+ ajudar, mas geralmente \fIlc\fP e \fIpb\fP são mais importantes. Por exemplo, um
1898
+ arquivo de código\-fonte contém principalmente texto US\-ASCII, então algo
1899
+ como o seguinte pode fornecer um arquivo ligeiramente (como 0,1\ %) menor
1900
+ que \fBxz \-6e\fP (tente também sem \fBlc=4\fP):
1901
+ .RS
1902
+ .PP
1903
+ .nf
1904
+ \f(CRxz \-\-lzma2=preset=6e,pb=0,lc=4 source_code.tar\fR
1905
+ .fi
1906
+ .RE
1907
+ .PP
1908
+ O uso de outro filtro junto com o LZMA2 pode melhorar a compactação com
1909
+ determinados tipos de arquivo. Por exemplo, para compactar uma biblioteca
1910
+ compartilhada x86\-32 ou x86\-64 usando o filtro x86 BCJ:
1911
+ .RS
1912
+ .PP
1913
+ .nf
1914
+ \f(CRxz \-\-x86 \-\-lzma2 libfoo.so\fR
1915
+ .fi
1916
+ .RE
1917
+ .PP
1918
+ Observe que a ordem das opções de filtro é significativa. Se \fB\-\-x86\fP for
1919
+ especificado após \fB\-\-lzma2\fP, \fBxz\fP dará um erro, porque não pode haver
1920
+ nenhum filtro após LZMA2 e também porque o filtro x86 BCJ não pode ser usado
1921
+ como o último filtro em a corrente.
1922
+ .PP
1923
+ O filtro Delta junto com LZMA2 pode dar bons resultados com imagens
1924
+ bitmap. Ele geralmente deve superar o PNG, que possui alguns filtros mais
1925
+ avançados do que o delta simples, mas usa Deflate para a compactação real.
1926
+ .PP
1927
+ A imagem deve ser salva em formato não compactado, por exemplo, como TIFF
1928
+ não compactado. O parâmetro de distância do filtro Delta é definido para
1929
+ corresponder ao número de bytes por pixel na imagem. Por exemplo, bitmap RGB
1930
+ de 24 bits precisa de \fBdist=3\fP, e também é bom passar \fBpb=0\fP para LZMA2
1931
+ para acomodar o alinhamento de três bytes:
1932
+ .RS
1933
+ .PP
1934
+ .nf
1935
+ \f(CRxz \-\-delta=dist=3 \-\-lzma2=pb=0 foo.tiff\fR
1936
+ .fi
1937
+ .RE
1938
+ .PP
1939
+ Se várias imagens foram colocadas em um único arquivo (por exemplo,
1940
+ \&\fB.tar\fP), o filtro Delta também funcionará, desde que todas as imagens
1941
+ tenham o mesmo número de bytes por pixel.
1942
+ .
1943
+ .SH "VEJA TAMBÉM"
1944
+ \fBxzdec\fP(1), \fBxzdiff\fP(1), \fBxzgrep\fP(1), \fBxzless\fP(1), \fBxzmore\fP(1),
1945
+ \fBgzip\fP(1), \fBbzip2\fP(1), \fB7z\fP(1)
1946
+ .PP
1947
+ XZ Utils: <https://tukaani.org/xz/>
1948
+ .br
1949
+ XZ Embedded: <https://tukaani.org/xz/embedded.html>
1950
+ .br
1951
+ LZMA SDK: <https://7\-zip.org/sdk.html>
janus/share/man/uk/man1/lzdiff.1 ADDED
@@ -0,0 +1,55 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ .\" SPDX-License-Identifier: 0BSD
2
+ .\"
3
+ .\" Authors: Lasse Collin
4
+ .\" Jia Tan
5
+ .\"
6
+ .\" Ukrainian translation for xz-man.
7
+ .\" Yuri Chornoivan <yurchor@ukr.net>, 2019, 2022, 2023, 2024.
8
+ .\"
9
+ .\" (Note that this file is not based on gzip's zdiff.1.)
10
+ .\"
11
+ .\"*******************************************************************
12
+ .\"
13
+ .\" This file was generated with po4a. Translate the source file.
14
+ .\"
15
+ .\"*******************************************************************
16
+ .TH XZDIFF 1 "13 лютого 2024 року" Tukaani "XZ Utils"
17
+ .SH НАЗВА
18
+ xzcmp, xzdiff, lzcmp, lzdiff — порівняння стиснених файлів
19
+ .
20
+ .SH "КОРОТКИЙ ОПИС"
21
+ \fBxzcmp\fP [\fIпараметр...\fP] \fIфайл1\fP [\fIфайл2\fP]
22
+ .br
23
+ \fBxzdiff\fP \&...
24
+ .br
25
+ \fBlzcmp\fP \&...
26
+ .br
27
+ \fBlzdiff\fP \&...
28
+ .
29
+ .SH ОПИС
30
+ \fBxzcmp\fP і \fBxzdiff\fP порівнюють нестиснені дані двох файлів. Нестиснені дані
31
+ та параметри буде передано \fBcmp\fP(1) або \fBdiff\fP(1), якщо не вказано
32
+ \fB\-\-help\fP або \fB\-\-version\fP.
33
+ .PP
34
+ Якщо вказано одразу \fIфайл1\fP і \fIфайл2\fP, це можуть бути нестиснені файли або
35
+ файли у форматах, які може бути розпаковано за допомогою \fBxz\fP(1),
36
+ \fBgzip\fP(1), \fBbzip2\fP(1), \fBlzop\fP(1), \fBzstd\fP(1) або \fBlz4\fP(1). Потрібні
37
+ команди розпаковування визначаються з суфіксів назв файлів \fIфайл1\fP і
38
+ \fIфайл2\fP. Файли із невідомими програмі суфіксами вважатимуться або
39
+ нестисненими або такими, що мають формат, дані у якому може розпаковувати
40
+ \fBxz\fP(1).
41
+ .PP
42
+ Якщо вказано лише одну назву файла, \fIфайл1\fP повинен мати суфікс
43
+ підтримуваного формату стискання, а назвою \fIфайл2\fP має бути назва \fIфайл1\fP
44
+ з вилученим суфіксом формату стискання.
45
+ .PP
46
+ Працездатність команд \fBlzcmp\fP і \fBlzdiff\fP забезпечено для зворотної
47
+ сумісності із LZMA Utils.
48
+ .
49
+ .SH "СТАН ВИХОДУ"
50
+ Якщо станеться помилка розпаковування, станом виходу буде \fB2\fP. Інакше,
51
+ станом виходу буде \fBcmp\fP(1) або буде використано \fBdiff\fP(1).
52
+ .
53
+ .SH "ДИВ. ТАКОЖ"
54
+ \fBcmp\fP(1), \fBdiff\fP(1), \fBxz\fP(1), \fBgzip\fP(1), \fBbzip2\fP(1), \fBlzop\fP(1),
55
+ \fBzstd\fP(1), \fBlz4\fP(1)
janus/share/zoneinfo/Asia/Almaty ADDED
Binary file (618 Bytes). View file
 
janus/share/zoneinfo/Asia/Barnaul ADDED
Binary file (753 Bytes). View file
 
janus/share/zoneinfo/Asia/Bishkek ADDED
Binary file (618 Bytes). View file
 
janus/share/zoneinfo/Asia/Calcutta ADDED
Binary file (220 Bytes). View file
 
janus/share/zoneinfo/Asia/Chungking ADDED
Binary file (393 Bytes). View file
 
janus/share/zoneinfo/Asia/Dhaka ADDED
Binary file (231 Bytes). View file
 
janus/share/zoneinfo/Asia/Gaza ADDED
Binary file (2.95 kB). View file
 
janus/share/zoneinfo/Asia/Hebron ADDED
Binary file (2.97 kB). View file
 
janus/share/zoneinfo/Asia/Katmandu ADDED
Binary file (161 Bytes). View file
 
janus/share/zoneinfo/Asia/Khandyga ADDED
Binary file (775 Bytes). View file
 
janus/share/zoneinfo/Asia/Makassar ADDED
Binary file (190 Bytes). View file
 
janus/share/zoneinfo/Asia/Novokuznetsk ADDED
Binary file (726 Bytes). View file
 
janus/share/zoneinfo/Asia/Pontianak ADDED
Binary file (247 Bytes). View file
 
janus/share/zoneinfo/Asia/Pyongyang ADDED
Binary file (183 Bytes). View file
 
janus/share/zoneinfo/Asia/Rangoon ADDED
Binary file (187 Bytes). View file
 
janus/share/zoneinfo/Asia/Samarkand ADDED
Binary file (366 Bytes). View file
 
janus/share/zoneinfo/Asia/Seoul ADDED
Binary file (415 Bytes). View file
 
janus/share/zoneinfo/Asia/Shanghai ADDED
Binary file (393 Bytes). View file
 
janus/share/zoneinfo/Asia/Thimphu ADDED
Binary file (154 Bytes). View file
 
janus/share/zoneinfo/Asia/Ulan_Bator ADDED
Binary file (594 Bytes). View file
 
janus/share/zoneinfo/Asia/Ust-Nera ADDED
Binary file (771 Bytes). View file
 
janus/share/zoneinfo/Asia/Yerevan ADDED
Binary file (708 Bytes). View file
 
janus/share/zoneinfo/Cuba ADDED
Binary file (1.12 kB). View file
 
janus/share/zoneinfo/EST ADDED
Binary file (149 Bytes). View file
 
janus/share/zoneinfo/EST5EDT ADDED
Binary file (1.74 kB). View file
 
janus/share/zoneinfo/Egypt ADDED
Binary file (1.31 kB). View file
 
janus/share/zoneinfo/Factory ADDED
Binary file (113 Bytes). View file
 
janus/share/zoneinfo/GB ADDED
Binary file (1.6 kB). View file
 
janus/share/zoneinfo/GMT ADDED
Binary file (111 Bytes). View file