SqueakyCleanText 0.2.1__tar.gz → 0.2.2__tar.gz

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (28) hide show
  1. {SqueakyCleanText-0.2.1/SqueakyCleanText.egg-info → SqueakyCleanText-0.2.2}/PKG-INFO +1 -1
  2. {SqueakyCleanText-0.2.1 → SqueakyCleanText-0.2.2/SqueakyCleanText.egg-info}/PKG-INFO +1 -1
  3. {SqueakyCleanText-0.2.1 → SqueakyCleanText-0.2.2}/SqueakyCleanText.egg-info/requires.txt +1 -1
  4. {SqueakyCleanText-0.2.1 → SqueakyCleanText-0.2.2}/sct/config.py +4 -2
  5. {SqueakyCleanText-0.2.1 → SqueakyCleanText-0.2.2}/setup.py +2 -2
  6. {SqueakyCleanText-0.2.1 → SqueakyCleanText-0.2.2}/LICENSE +0 -0
  7. {SqueakyCleanText-0.2.1 → SqueakyCleanText-0.2.2}/MANIFEST.in +0 -0
  8. {SqueakyCleanText-0.2.1 → SqueakyCleanText-0.2.2}/README.md +0 -0
  9. {SqueakyCleanText-0.2.1 → SqueakyCleanText-0.2.2}/SqueakyCleanText.egg-info/SOURCES.txt +0 -0
  10. {SqueakyCleanText-0.2.1 → SqueakyCleanText-0.2.2}/SqueakyCleanText.egg-info/dependency_links.txt +0 -0
  11. {SqueakyCleanText-0.2.1 → SqueakyCleanText-0.2.2}/SqueakyCleanText.egg-info/entry_points.txt +0 -0
  12. {SqueakyCleanText-0.2.1 → SqueakyCleanText-0.2.2}/SqueakyCleanText.egg-info/top_level.txt +0 -0
  13. {SqueakyCleanText-0.2.1 → SqueakyCleanText-0.2.2}/sct/__init__.py +0 -0
  14. {SqueakyCleanText-0.2.1 → SqueakyCleanText-0.2.2}/sct/scripts/__init__.py +0 -0
  15. {SqueakyCleanText-0.2.1 → SqueakyCleanText-0.2.2}/sct/scripts/download_nltk_stopwords.py +0 -0
  16. {SqueakyCleanText-0.2.1 → SqueakyCleanText-0.2.2}/sct/sct.py +0 -0
  17. {SqueakyCleanText-0.2.1 → SqueakyCleanText-0.2.2}/sct/utils/__init__.py +0 -0
  18. {SqueakyCleanText-0.2.1 → SqueakyCleanText-0.2.2}/sct/utils/constants.py +0 -0
  19. {SqueakyCleanText-0.2.1 → SqueakyCleanText-0.2.2}/sct/utils/contact.py +0 -0
  20. {SqueakyCleanText-0.2.1 → SqueakyCleanText-0.2.2}/sct/utils/datetime.py +0 -0
  21. {SqueakyCleanText-0.2.1 → SqueakyCleanText-0.2.2}/sct/utils/ner.py +0 -0
  22. {SqueakyCleanText-0.2.1 → SqueakyCleanText-0.2.2}/sct/utils/normtext.py +0 -0
  23. {SqueakyCleanText-0.2.1 → SqueakyCleanText-0.2.2}/sct/utils/resources.py +0 -0
  24. {SqueakyCleanText-0.2.1 → SqueakyCleanText-0.2.2}/sct/utils/special.py +0 -0
  25. {SqueakyCleanText-0.2.1 → SqueakyCleanText-0.2.2}/sct/utils/stopwords.py +0 -0
  26. {SqueakyCleanText-0.2.1 → SqueakyCleanText-0.2.2}/setup.cfg +0 -0
  27. {SqueakyCleanText-0.2.1 → SqueakyCleanText-0.2.2}/tests/__init__.py +0 -0
  28. {SqueakyCleanText-0.2.1 → SqueakyCleanText-0.2.2}/tests/test_sct.py +0 -0
{SqueakyCleanText-0.2.1/SqueakyCleanText.egg-info → SqueakyCleanText-0.2.2}/PKG-INFO RENAMED
@@ -1,6 +1,6 @@
1
1
  Metadata-Version: 2.1
2
2
  Name: SqueakyCleanText
3
- Version: 0.2.1
3
+ Version: 0.2.2
4
4
  Summary: A comprehensive text cleaning and preprocessing pipeline.
5
5
  Home-page: https://github.com/rhnfzl/SqueakyCleanText
6
6
  Author: Rehan Fazal
{SqueakyCleanText-0.2.1 → SqueakyCleanText-0.2.2/SqueakyCleanText.egg-info}/PKG-INFO RENAMED
@@ -1,6 +1,6 @@
1
1
  Metadata-Version: 2.1
2
2
  Name: SqueakyCleanText
3
- Version: 0.2.1
3
+ Version: 0.2.2
4
4
  Summary: A comprehensive text cleaning and preprocessing pipeline.
5
5
  Home-page: https://github.com/rhnfzl/SqueakyCleanText
6
6
  Author: Rehan Fazal
{SqueakyCleanText-0.2.1 → SqueakyCleanText-0.2.2}/SqueakyCleanText.egg-info/requires.txt RENAMED
@@ -1,4 +1,4 @@
1
- lingua-language-detector>=2.0.0
1
+ lingua-language-detector>=2.0.2
2
2
  nltk>=3.8
3
3
  emoji>=2.8
4
4
  ftfy>=6.1
{SqueakyCleanText-0.2.1 → SqueakyCleanText-0.2.2}/sct/config.py RENAMED
@@ -49,7 +49,9 @@ POSITIONAL_TAGS = ['PER', 'LOC', 'ORG']
49
49
  NER_CONFIDENCE_THRESHOLD = 0.85
50
50
  LANGUAGE = None
51
51
 
52
- NER_MODELS_LIST = ["FacebookAI/xlm-roberta-large-finetuned-conll03-english",
52
+ # Order of the model is Important : English Model, Dutch Model, German Model, Spanish Model, MULTILINGUAL Model
53
+ NER_MODELS_LIST = ["FacebookAI/xlm-roberta-large-finetuned-conll03-english",
53
54
  "FacebookAI/xlm-roberta-large-finetuned-conll02-dutch",
54
55
  "FacebookAI/xlm-roberta-large-finetuned-conll03-german",
55
- "FacebookAI/xlm-roberta-large-finetuned-conll02-spanish",
56
+ "FacebookAI/xlm-roberta-large-finetuned-conll02-spanish",
57
+ "Babelscape/wikineural-multilingual-ner"]
{SqueakyCleanText-0.2.1 → SqueakyCleanText-0.2.2}/setup.py RENAMED
@@ -2,7 +2,7 @@ from setuptools import setup, find_packages
2
2
 
3
3
  setup(
4
4
  name='SqueakyCleanText',
5
- version='0.2.1',
5
+ version='0.2.2',
6
6
  author='Rehan Fazal',
7
7
  description='A comprehensive text cleaning and preprocessing pipeline.',
8
8
  long_description=open('README.md', encoding='utf-8').read(),
@@ -11,7 +11,7 @@ setup(
11
11
  license='MIT',
12
12
  packages=find_packages(),
13
13
  install_requires=[
14
- 'lingua-language-detector>=2.0.0',
14
+ 'lingua-language-detector>=2.0.2',
15
15
  'nltk>=3.8',
16
16
  'emoji>=2.8',
17
17
  'ftfy>=6.1',