guilhermelmello commited on
Commit
4baed46
1 Parent(s): 9b698a8

Update loading script v1.2

Browse files
Files changed (1) hide show
  1. corpus-carolina.py +30 -24
corpus-carolina.py CHANGED
@@ -14,6 +14,7 @@
14
  # limitations under the License.
15
  """Carolina Corpus"""
16
 
 
17
  from lxml import etree
18
  import os
19
  import datasets
@@ -30,7 +31,7 @@ robust volume of texts of varied typology in contemporary Brazilian Portuguese
30
 
31
 
32
  _CITATION = r"""
33
- @misc{corpusCarolinaV1.1,
34
  title={
35
  Carolina:
36
  The Open Corpus for Linguistics and Artificial Intelligence},
@@ -52,7 +53,7 @@ _CITATION = r"""
52
  Palma, Mayara Feliciano},
53
  howpublished={\url{https://sites.usp.br/corpuscarolina/corpus}},
54
  year={2022},
55
- note={Version 1.1 (Ada)},
56
  }
57
  """
58
 
@@ -81,13 +82,13 @@ def _taxonomies():
81
  leg="legislative branch",
82
  pub="public domain works",
83
  soc="social media",
84
- uni="university_domains",
85
  wik="wikis",
86
  )
87
 
88
 
89
- _VERSION = "1.1.0"
90
- _CORPUS_URL = "corpus/{taxonomy}/"
91
  _CHECKSUM_FNAME = _CORPUS_URL + "checksum.sha256"
92
 
93
 
@@ -140,6 +141,7 @@ class Carolina(datasets.GeneratorBasedBuilder):
140
 
141
  return datasets.DatasetInfo(
142
  description=_DESCRIPTION,
 
143
  homepage=_HOMEPAGE,
144
  citation=_CITATION,
145
  features=features,
@@ -153,23 +155,21 @@ class Carolina(datasets.GeneratorBasedBuilder):
153
  else:
154
  taxonomies = [_taxonomies()[self.config.taxonomy]]
155
 
156
- zip_urls = dict()
157
- for taxonomy in taxonomies:
158
- # download checksum file
159
- checksum_path = _CHECKSUM_FNAME.format(taxonomy=taxonomy)
160
- checksum_path = dl_manager.download(checksum_path)
161
-
162
- tax_url = _CORPUS_URL.format(taxonomy=taxonomy)
163
-
164
- # extract and build zip urls
165
- with open(checksum_path, encoding="utf-8") as cfile:
166
  for line in cfile:
167
- fname = line.split()[1]
168
- if fname.endswith(".xml.zip"):
169
- zip_url = tax_url + fname # download url
170
- fname = os.path.split(fname)[1] # removes subdirs
171
- fname = fname[:-4] # removes .zip
172
- zip_urls[fname] = zip_url # xml -> zip url
173
 
174
  # extractions are made in cache folders and
175
  # the path returned is the folder path, not the
@@ -200,11 +200,17 @@ class Carolina(datasets.GeneratorBasedBuilder):
200
  # parse xml file
201
  for _, tei in etree.iterparse(path, **parser_params):
202
  header = tei.find(f"{TEI_NS}teiHeader")
 
 
 
 
 
 
 
203
 
204
  example = {
205
- "meta": etree.tostring(
206
- header, encoding="utf-8").decode("utf-8"),
207
- "text": tei.find(f".//{TEI_NS}body/{TEI_NS}p").text
208
  }
209
  yield _key, example
210
  _key += 1
 
14
  # limitations under the License.
15
  """Carolina Corpus"""
16
 
17
+ from collections import defaultdict
18
  from lxml import etree
19
  import os
20
  import datasets
 
31
 
32
 
33
  _CITATION = r"""
34
+ @misc{corpusCarolinaV1.2,
35
  title={
36
  Carolina:
37
  The Open Corpus for Linguistics and Artificial Intelligence},
 
53
  Palma, Mayara Feliciano},
54
  howpublished={\url{https://sites.usp.br/corpuscarolina/corpus}},
55
  year={2022},
56
+ note={Version 1.2 (Ada)},
57
  }
58
  """
59
 
 
82
  leg="legislative branch",
83
  pub="public domain works",
84
  soc="social media",
85
+ uni="university domains",
86
  wik="wikis",
87
  )
88
 
89
 
90
+ _VERSION = "1.2.0"
91
+ _CORPUS_URL = "corpus/{tax}/"
92
  _CHECKSUM_FNAME = _CORPUS_URL + "checksum.sha256"
93
 
94
 
 
141
 
142
  return datasets.DatasetInfo(
143
  description=_DESCRIPTION,
144
+ supervised_keys=None,
145
  homepage=_HOMEPAGE,
146
  citation=_CITATION,
147
  features=features,
 
155
  else:
156
  taxonomies = [_taxonomies()[self.config.taxonomy]]
157
 
158
+ # download checksum files
159
+ checksum_urls = {t: _CHECKSUM_FNAME.format(tax=t) for t in taxonomies}
160
+ checksum_paths = dl_manager.download(checksum_urls)
161
+
162
+ # prepare xml file name and zip urls
163
+ zip_urls = dict() # xml
164
+ for tax, cpath in checksum_paths.items():
165
+ tax_path = _CORPUS_URL.format(tax=tax)
166
+ with open(cpath, encoding="utf-8") as cfile:
 
167
  for line in cfile:
168
+ xml_tax_path = line.split()[1] # xml file inside taxonomy
169
+ zip_fname = xml_tax_path + ".zip" # zip file inside taxonomy
170
+ xml_fname = xml_tax_path.split('/')[-1] # xml file name only
171
+ zip_fpath = os.path.join(tax_path, zip_fname) # path inside corpus
172
+ zip_urls[xml_fname] = zip_fpath # xml file -> zip path
 
173
 
174
  # extractions are made in cache folders and
175
  # the path returned is the folder path, not the
 
200
  # parse xml file
201
  for _, tei in etree.iterparse(path, **parser_params):
202
  header = tei.find(f"{TEI_NS}teiHeader")
203
+
204
+ meta = etree.tostring(
205
+ header, encoding="utf-8").decode("utf-8")
206
+ text = ' '.join([e.text
207
+ for e in tei.findall(f".//{TEI_NS}body/{TEI_NS}p")
208
+ if e.text is not None
209
+ ])
210
 
211
  example = {
212
+ "meta": meta,
213
+ "text": text
 
214
  }
215
  yield _key, example
216
  _key += 1