File size: 19,624 Bytes
8fd238c
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
from functools import lru_cache
from logging import getLogger
from typing import List, Optional

from .constant import (
    COMMON_SAFE_ASCII_CHARACTERS,
    TRACE,
    UNICODE_SECONDARY_RANGE_KEYWORD,
)
from .utils import (
    is_accentuated,
    is_arabic,
    is_arabic_isolated_form,
    is_case_variable,
    is_cjk,
    is_emoticon,
    is_hangul,
    is_hiragana,
    is_katakana,
    is_latin,
    is_punctuation,
    is_separator,
    is_symbol,
    is_thai,
    is_unprintable,
    remove_accent,
    unicode_range,
)


class MessDetectorPlugin:
    """
    Base abstract class used for mess detection plugins.
    All detectors MUST extend and implement given methods.
    """

    def eligible(self, character: str) -> bool:
        """
        Determine if given character should be fed in.
        """
        raise NotImplementedError  # pragma: nocover

    def feed(self, character: str) -> None:
        """
        The main routine to be executed upon character.
        Insert the logic in witch the text would be considered chaotic.
        """
        raise NotImplementedError  # pragma: nocover

    def reset(self) -> None:  # pragma: no cover
        """
        Permit to reset the plugin to the initial state.
        """
        raise NotImplementedError

    @property
    def ratio(self) -> float:
        """
        Compute the chaos ratio based on what your feed() has seen.
        Must NOT be lower than 0.; No restriction gt 0.
        """
        raise NotImplementedError  # pragma: nocover


class TooManySymbolOrPunctuationPlugin(MessDetectorPlugin):
    def __init__(self) -> None:
        self._punctuation_count: int = 0
        self._symbol_count: int = 0
        self._character_count: int = 0

        self._last_printable_char: Optional[str] = None
        self._frenzy_symbol_in_word: bool = False

    def eligible(self, character: str) -> bool:
        return character.isprintable()

    def feed(self, character: str) -> None:
        self._character_count += 1

        if (
            character != self._last_printable_char
            and character not in COMMON_SAFE_ASCII_CHARACTERS
        ):
            if is_punctuation(character):
                self._punctuation_count += 1
            elif (
                character.isdigit() is False
                and is_symbol(character)
                and is_emoticon(character) is False
            ):
                self._symbol_count += 2

        self._last_printable_char = character

    def reset(self) -> None:  # pragma: no cover
        self._punctuation_count = 0
        self._character_count = 0
        self._symbol_count = 0

    @property
    def ratio(self) -> float:
        if self._character_count == 0:
            return 0.0

        ratio_of_punctuation: float = (
            self._punctuation_count + self._symbol_count
        ) / self._character_count

        return ratio_of_punctuation if ratio_of_punctuation >= 0.3 else 0.0


class TooManyAccentuatedPlugin(MessDetectorPlugin):
    def __init__(self) -> None:
        self._character_count: int = 0
        self._accentuated_count: int = 0

    def eligible(self, character: str) -> bool:
        return character.isalpha()

    def feed(self, character: str) -> None:
        self._character_count += 1

        if is_accentuated(character):
            self._accentuated_count += 1

    def reset(self) -> None:  # pragma: no cover
        self._character_count = 0
        self._accentuated_count = 0

    @property
    def ratio(self) -> float:
        if self._character_count < 8:
            return 0.0

        ratio_of_accentuation: float = self._accentuated_count / self._character_count
        return ratio_of_accentuation if ratio_of_accentuation >= 0.35 else 0.0


class UnprintablePlugin(MessDetectorPlugin):
    def __init__(self) -> None:
        self._unprintable_count: int = 0
        self._character_count: int = 0

    def eligible(self, character: str) -> bool:
        return True

    def feed(self, character: str) -> None:
        if is_unprintable(character):
            self._unprintable_count += 1
        self._character_count += 1

    def reset(self) -> None:  # pragma: no cover
        self._unprintable_count = 0

    @property
    def ratio(self) -> float:
        if self._character_count == 0:
            return 0.0

        return (self._unprintable_count * 8) / self._character_count


class SuspiciousDuplicateAccentPlugin(MessDetectorPlugin):
    def __init__(self) -> None:
        self._successive_count: int = 0
        self._character_count: int = 0

        self._last_latin_character: Optional[str] = None

    def eligible(self, character: str) -> bool:
        return character.isalpha() and is_latin(character)

    def feed(self, character: str) -> None:
        self._character_count += 1
        if (
            self._last_latin_character is not None
            and is_accentuated(character)
            and is_accentuated(self._last_latin_character)
        ):
            if character.isupper() and self._last_latin_character.isupper():
                self._successive_count += 1
            # Worse if its the same char duplicated with different accent.
            if remove_accent(character) == remove_accent(self._last_latin_character):
                self._successive_count += 1
        self._last_latin_character = character

    def reset(self) -> None:  # pragma: no cover
        self._successive_count = 0
        self._character_count = 0
        self._last_latin_character = None

    @property
    def ratio(self) -> float:
        if self._character_count == 0:
            return 0.0

        return (self._successive_count * 2) / self._character_count


class SuspiciousRange(MessDetectorPlugin):
    def __init__(self) -> None:
        self._suspicious_successive_range_count: int = 0
        self._character_count: int = 0
        self._last_printable_seen: Optional[str] = None

    def eligible(self, character: str) -> bool:
        return character.isprintable()

    def feed(self, character: str) -> None:
        self._character_count += 1

        if (
            character.isspace()
            or is_punctuation(character)
            or character in COMMON_SAFE_ASCII_CHARACTERS
        ):
            self._last_printable_seen = None
            return

        if self._last_printable_seen is None:
            self._last_printable_seen = character
            return

        unicode_range_a: Optional[str] = unicode_range(self._last_printable_seen)
        unicode_range_b: Optional[str] = unicode_range(character)

        if is_suspiciously_successive_range(unicode_range_a, unicode_range_b):
            self._suspicious_successive_range_count += 1

        self._last_printable_seen = character

    def reset(self) -> None:  # pragma: no cover
        self._character_count = 0
        self._suspicious_successive_range_count = 0
        self._last_printable_seen = None

    @property
    def ratio(self) -> float:
        if self._character_count <= 24:
            return 0.0

        ratio_of_suspicious_range_usage: float = (
            self._suspicious_successive_range_count * 2
        ) / self._character_count

        return ratio_of_suspicious_range_usage


class SuperWeirdWordPlugin(MessDetectorPlugin):
    def __init__(self) -> None:
        self._word_count: int = 0
        self._bad_word_count: int = 0
        self._foreign_long_count: int = 0

        self._is_current_word_bad: bool = False
        self._foreign_long_watch: bool = False

        self._character_count: int = 0
        self._bad_character_count: int = 0

        self._buffer: str = ""
        self._buffer_accent_count: int = 0

    def eligible(self, character: str) -> bool:
        return True

    def feed(self, character: str) -> None:
        if character.isalpha():
            self._buffer += character
            if is_accentuated(character):
                self._buffer_accent_count += 1
            if (
                self._foreign_long_watch is False
                and (is_latin(character) is False or is_accentuated(character))
                and is_cjk(character) is False
                and is_hangul(character) is False
                and is_katakana(character) is False
                and is_hiragana(character) is False
                and is_thai(character) is False
            ):
                self._foreign_long_watch = True
            return
        if not self._buffer:
            return
        if (
            character.isspace() or is_punctuation(character) or is_separator(character)
        ) and self._buffer:
            self._word_count += 1
            buffer_length: int = len(self._buffer)

            self._character_count += buffer_length

            if buffer_length >= 4:
                if self._buffer_accent_count / buffer_length > 0.34:
                    self._is_current_word_bad = True
                # Word/Buffer ending with an upper case accentuated letter are so rare,
                # that we will consider them all as suspicious. Same weight as foreign_long suspicious.
                if (
                    is_accentuated(self._buffer[-1])
                    and self._buffer[-1].isupper()
                    and all(_.isupper() for _ in self._buffer) is False
                ):
                    self._foreign_long_count += 1
                    self._is_current_word_bad = True
            if buffer_length >= 24 and self._foreign_long_watch:
                camel_case_dst = [
                    i
                    for c, i in zip(self._buffer, range(0, buffer_length))
                    if c.isupper()
                ]
                probable_camel_cased: bool = False

                if camel_case_dst and (len(camel_case_dst) / buffer_length <= 0.3):
                    probable_camel_cased = True

                if not probable_camel_cased:
                    self._foreign_long_count += 1
                    self._is_current_word_bad = True

            if self._is_current_word_bad:
                self._bad_word_count += 1
                self._bad_character_count += len(self._buffer)
                self._is_current_word_bad = False

            self._foreign_long_watch = False
            self._buffer = ""
            self._buffer_accent_count = 0
        elif (
            character not in {"<", ">", "-", "=", "~", "|", "_"}
            and character.isdigit() is False
            and is_symbol(character)
        ):
            self._is_current_word_bad = True
            self._buffer += character

    def reset(self) -> None:  # pragma: no cover
        self._buffer = ""
        self._is_current_word_bad = False
        self._foreign_long_watch = False
        self._bad_word_count = 0
        self._word_count = 0
        self._character_count = 0
        self._bad_character_count = 0
        self._foreign_long_count = 0

    @property
    def ratio(self) -> float:
        if self._word_count <= 10 and self._foreign_long_count == 0:
            return 0.0

        return self._bad_character_count / self._character_count


class CjkInvalidStopPlugin(MessDetectorPlugin):
    """
    GB(Chinese) based encoding often render the stop incorrectly when the content does not fit and
    can be easily detected. Searching for the overuse of '丅' and '丄'.
    """

    def __init__(self) -> None:
        self._wrong_stop_count: int = 0
        self._cjk_character_count: int = 0

    def eligible(self, character: str) -> bool:
        return True

    def feed(self, character: str) -> None:
        if character in {"丅", "丄"}:
            self._wrong_stop_count += 1
            return
        if is_cjk(character):
            self._cjk_character_count += 1

    def reset(self) -> None:  # pragma: no cover
        self._wrong_stop_count = 0
        self._cjk_character_count = 0

    @property
    def ratio(self) -> float:
        if self._cjk_character_count < 16:
            return 0.0
        return self._wrong_stop_count / self._cjk_character_count


class ArchaicUpperLowerPlugin(MessDetectorPlugin):
    def __init__(self) -> None:
        self._buf: bool = False

        self._character_count_since_last_sep: int = 0

        self._successive_upper_lower_count: int = 0
        self._successive_upper_lower_count_final: int = 0

        self._character_count: int = 0

        self._last_alpha_seen: Optional[str] = None
        self._current_ascii_only: bool = True

    def eligible(self, character: str) -> bool:
        return True

    def feed(self, character: str) -> None:
        is_concerned = character.isalpha() and is_case_variable(character)
        chunk_sep = is_concerned is False

        if chunk_sep and self._character_count_since_last_sep > 0:
            if (
                self._character_count_since_last_sep <= 64
                and character.isdigit() is False
                and self._current_ascii_only is False
            ):
                self._successive_upper_lower_count_final += (
                    self._successive_upper_lower_count
                )

            self._successive_upper_lower_count = 0
            self._character_count_since_last_sep = 0
            self._last_alpha_seen = None
            self._buf = False
            self._character_count += 1
            self._current_ascii_only = True

            return

        if self._current_ascii_only is True and character.isascii() is False:
            self._current_ascii_only = False

        if self._last_alpha_seen is not None:
            if (character.isupper() and self._last_alpha_seen.islower()) or (
                character.islower() and self._last_alpha_seen.isupper()
            ):
                if self._buf is True:
                    self._successive_upper_lower_count += 2
                    self._buf = False
                else:
                    self._buf = True
            else:
                self._buf = False

        self._character_count += 1
        self._character_count_since_last_sep += 1
        self._last_alpha_seen = character

    def reset(self) -> None:  # pragma: no cover
        self._character_count = 0
        self._character_count_since_last_sep = 0
        self._successive_upper_lower_count = 0
        self._successive_upper_lower_count_final = 0
        self._last_alpha_seen = None
        self._buf = False
        self._current_ascii_only = True

    @property
    def ratio(self) -> float:
        if self._character_count == 0:
            return 0.0

        return self._successive_upper_lower_count_final / self._character_count


class ArabicIsolatedFormPlugin(MessDetectorPlugin):
    def __init__(self) -> None:
        self._character_count: int = 0
        self._isolated_form_count: int = 0

    def reset(self) -> None:  # pragma: no cover
        self._character_count = 0
        self._isolated_form_count = 0

    def eligible(self, character: str) -> bool:
        return is_arabic(character)

    def feed(self, character: str) -> None:
        self._character_count += 1

        if is_arabic_isolated_form(character):
            self._isolated_form_count += 1

    @property
    def ratio(self) -> float:
        if self._character_count < 8:
            return 0.0

        isolated_form_usage: float = self._isolated_form_count / self._character_count

        return isolated_form_usage


@lru_cache(maxsize=1024)
def is_suspiciously_successive_range(
    unicode_range_a: Optional[str], unicode_range_b: Optional[str]
) -> bool:
    """
    Determine if two Unicode range seen next to each other can be considered as suspicious.
    """
    if unicode_range_a is None or unicode_range_b is None:
        return True

    if unicode_range_a == unicode_range_b:
        return False

    if "Latin" in unicode_range_a and "Latin" in unicode_range_b:
        return False

    if "Emoticons" in unicode_range_a or "Emoticons" in unicode_range_b:
        return False

    # Latin characters can be accompanied with a combining diacritical mark
    # eg. Vietnamese.
    if ("Latin" in unicode_range_a or "Latin" in unicode_range_b) and (
        "Combining" in unicode_range_a or "Combining" in unicode_range_b
    ):
        return False

    keywords_range_a, keywords_range_b = unicode_range_a.split(
        " "
    ), unicode_range_b.split(" ")

    for el in keywords_range_a:
        if el in UNICODE_SECONDARY_RANGE_KEYWORD:
            continue
        if el in keywords_range_b:
            return False

    # Japanese Exception
    range_a_jp_chars, range_b_jp_chars = (
        unicode_range_a
        in (
            "Hiragana",
            "Katakana",
        ),
        unicode_range_b in ("Hiragana", "Katakana"),
    )
    if (range_a_jp_chars or range_b_jp_chars) and (
        "CJK" in unicode_range_a or "CJK" in unicode_range_b
    ):
        return False
    if range_a_jp_chars and range_b_jp_chars:
        return False

    if "Hangul" in unicode_range_a or "Hangul" in unicode_range_b:
        if "CJK" in unicode_range_a or "CJK" in unicode_range_b:
            return False
        if unicode_range_a == "Basic Latin" or unicode_range_b == "Basic Latin":
            return False

    # Chinese/Japanese use dedicated range for punctuation and/or separators.
    if ("CJK" in unicode_range_a or "CJK" in unicode_range_b) or (
        unicode_range_a in ["Katakana", "Hiragana"]
        and unicode_range_b in ["Katakana", "Hiragana"]
    ):
        if "Punctuation" in unicode_range_a or "Punctuation" in unicode_range_b:
            return False
        if "Forms" in unicode_range_a or "Forms" in unicode_range_b:
            return False
        if unicode_range_a == "Basic Latin" or unicode_range_b == "Basic Latin":
            return False

    return True


@lru_cache(maxsize=2048)
def mess_ratio(
    decoded_sequence: str, maximum_threshold: float = 0.2, debug: bool = False
) -> float:
    """
    Compute a mess ratio given a decoded bytes sequence. The maximum threshold does stop the computation earlier.
    """

    detectors: List[MessDetectorPlugin] = [
        md_class() for md_class in MessDetectorPlugin.__subclasses__()
    ]

    length: int = len(decoded_sequence) + 1

    mean_mess_ratio: float = 0.0

    if length < 512:
        intermediary_mean_mess_ratio_calc: int = 32
    elif length <= 1024:
        intermediary_mean_mess_ratio_calc = 64
    else:
        intermediary_mean_mess_ratio_calc = 128

    for character, index in zip(decoded_sequence + "\n", range(length)):
        for detector in detectors:
            if detector.eligible(character):
                detector.feed(character)

        if (
            index > 0 and index % intermediary_mean_mess_ratio_calc == 0
        ) or index == length - 1:
            mean_mess_ratio = sum(dt.ratio for dt in detectors)

            if mean_mess_ratio >= maximum_threshold:
                break

    if debug:
        logger = getLogger("charset_normalizer")

        logger.log(
            TRACE,
            "Mess-detector extended-analysis start. "
            f"intermediary_mean_mess_ratio_calc={intermediary_mean_mess_ratio_calc} mean_mess_ratio={mean_mess_ratio} "
            f"maximum_threshold={maximum_threshold}",
        )

        if len(decoded_sequence) > 16:
            logger.log(TRACE, f"Starting with: {decoded_sequence[:16]}")
            logger.log(TRACE, f"Ending with: {decoded_sequence[-16::]}")

        for dt in detectors:  # pragma: nocover
            logger.log(TRACE, f"{dt.__class__}: {dt.ratio}")

    return round(mean_mess_ratio, 3)