UniTok 3.0.8__tar.gz → 3.0.9__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 (29) hide show
  1. {UniTok-3.0.8 → UniTok-3.0.9}/PKG-INFO +1 -1
  2. {UniTok-3.0.8 → UniTok-3.0.9}/UniTok/meta.py +8 -1
  3. {UniTok-3.0.8 → UniTok-3.0.9}/UniTok.egg-info/PKG-INFO +1 -1
  4. {UniTok-3.0.8 → UniTok-3.0.9}/setup.py +1 -1
  5. {UniTok-3.0.8 → UniTok-3.0.9}/README.md +0 -0
  6. {UniTok-3.0.8 → UniTok-3.0.9}/UniTok/__init__.py +0 -0
  7. {UniTok-3.0.8 → UniTok-3.0.9}/UniTok/analysis/__init__.py +0 -0
  8. {UniTok-3.0.8 → UniTok-3.0.9}/UniTok/analysis/lengths.py +0 -0
  9. {UniTok-3.0.8 → UniTok-3.0.9}/UniTok/analysis/plot.py +0 -0
  10. {UniTok-3.0.8 → UniTok-3.0.9}/UniTok/cols.py +0 -0
  11. {UniTok-3.0.8 → UniTok-3.0.9}/UniTok/column.py +0 -0
  12. {UniTok-3.0.8 → UniTok-3.0.9}/UniTok/global_setting.py +0 -0
  13. {UniTok-3.0.8 → UniTok-3.0.9}/UniTok/tok/__init__.py +0 -0
  14. {UniTok-3.0.8 → UniTok-3.0.9}/UniTok/tok/bert_tok.py +0 -0
  15. {UniTok-3.0.8 → UniTok-3.0.9}/UniTok/tok/entity_tok.py +0 -0
  16. {UniTok-3.0.8 → UniTok-3.0.9}/UniTok/tok/id_tok.py +0 -0
  17. {UniTok-3.0.8 → UniTok-3.0.9}/UniTok/tok/number_tok.py +0 -0
  18. {UniTok-3.0.8 → UniTok-3.0.9}/UniTok/tok/seq_tok.py +0 -0
  19. {UniTok-3.0.8 → UniTok-3.0.9}/UniTok/tok/split_tok.py +0 -0
  20. {UniTok-3.0.8 → UniTok-3.0.9}/UniTok/tok/tok.py +0 -0
  21. {UniTok-3.0.8 → UniTok-3.0.9}/UniTok/unidep.py +0 -0
  22. {UniTok-3.0.8 → UniTok-3.0.9}/UniTok/unitok.py +0 -0
  23. {UniTok-3.0.8 → UniTok-3.0.9}/UniTok/vocab.py +0 -0
  24. {UniTok-3.0.8 → UniTok-3.0.9}/UniTok/vocabs.py +0 -0
  25. {UniTok-3.0.8 → UniTok-3.0.9}/UniTok.egg-info/SOURCES.txt +0 -0
  26. {UniTok-3.0.8 → UniTok-3.0.9}/UniTok.egg-info/dependency_links.txt +0 -0
  27. {UniTok-3.0.8 → UniTok-3.0.9}/UniTok.egg-info/requires.txt +0 -0
  28. {UniTok-3.0.8 → UniTok-3.0.9}/UniTok.egg-info/top_level.txt +0 -0
  29. {UniTok-3.0.8 → UniTok-3.0.9}/setup.cfg +0 -0
{UniTok-3.0.8 → UniTok-3.0.9}/PKG-INFO RENAMED
@@ -1,6 +1,6 @@
1
1
  Metadata-Version: 2.1
2
2
  Name: UniTok
3
- Version: 3.0.8
3
+ Version: 3.0.9
4
4
  Summary: Unified Tokenizer
5
5
  Home-page: https://github.com/Jyonn/UnifiedTokenizer
6
6
  Author: Jyonn Liu
{UniTok-3.0.8 → UniTok-3.0.9}/UniTok/meta.py RENAMED
@@ -41,10 +41,17 @@ class Voc:
41
41
  }
42
42
 
43
43
  def merge(self, other):
44
+ cols = self.cols.copy()
45
+ for col in other.cols:
46
+ for _col in cols:
47
+ if col.name == _col.name:
48
+ break
49
+ else:
50
+ cols.append(col)
44
51
  return Voc(
45
52
  name=self.name,
46
53
  size=self.size,
47
- cols=list(set(self.cols + other.cols)),
54
+ cols=cols,
48
55
  )
49
56
 
50
57
 
{UniTok-3.0.8 → UniTok-3.0.9}/UniTok.egg-info/PKG-INFO RENAMED
@@ -1,6 +1,6 @@
1
1
  Metadata-Version: 2.1
2
2
  Name: UniTok
3
- Version: 3.0.8
3
+ Version: 3.0.9
4
4
  Summary: Unified Tokenizer
5
5
  Home-page: https://github.com/Jyonn/UnifiedTokenizer
6
6
  Author: Jyonn Liu
{UniTok-3.0.8 → UniTok-3.0.9}/setup.py RENAMED
@@ -6,7 +6,7 @@ long_description = (this_directory / "README.md").read_text()
6
6
 
7
7
  setup(
8
8
  name='UniTok',
9
- version='3.0.8',
9
+ version='3.0.9',
10
10
  keywords=['token', 'tokenizer', 'bert'],
11
11
  description='Unified Tokenizer',
12
12
  long_description=long_description,
{UniTok-3.0.8 → UniTok-3.0.9}/README.md RENAMED
File without changes
{UniTok-3.0.8 → UniTok-3.0.9}/UniTok/__init__.py RENAMED
File without changes
{UniTok-3.0.8 → UniTok-3.0.9}/UniTok/analysis/plot.py RENAMED
File without changes
{UniTok-3.0.8 → UniTok-3.0.9}/UniTok/cols.py RENAMED
File without changes
{UniTok-3.0.8 → UniTok-3.0.9}/UniTok/column.py RENAMED
File without changes
{UniTok-3.0.8 → UniTok-3.0.9}/UniTok/global_setting.py RENAMED
File without changes
{UniTok-3.0.8 → UniTok-3.0.9}/UniTok/tok/__init__.py RENAMED
File without changes
{UniTok-3.0.8 → UniTok-3.0.9}/UniTok/tok/bert_tok.py RENAMED
File without changes
{UniTok-3.0.8 → UniTok-3.0.9}/UniTok/tok/entity_tok.py RENAMED
File without changes
{UniTok-3.0.8 → UniTok-3.0.9}/UniTok/tok/id_tok.py RENAMED
File without changes
{UniTok-3.0.8 → UniTok-3.0.9}/UniTok/tok/number_tok.py RENAMED
File without changes
{UniTok-3.0.8 → UniTok-3.0.9}/UniTok/tok/seq_tok.py RENAMED
File without changes
{UniTok-3.0.8 → UniTok-3.0.9}/UniTok/tok/split_tok.py RENAMED
File without changes
{UniTok-3.0.8 → UniTok-3.0.9}/UniTok/tok/tok.py RENAMED
File without changes
{UniTok-3.0.8 → UniTok-3.0.9}/UniTok/unidep.py RENAMED
File without changes
{UniTok-3.0.8 → UniTok-3.0.9}/UniTok/unitok.py RENAMED
File without changes
{UniTok-3.0.8 → UniTok-3.0.9}/UniTok/vocab.py RENAMED
File without changes
{UniTok-3.0.8 → UniTok-3.0.9}/UniTok/vocabs.py RENAMED
File without changes
{UniTok-3.0.8 → UniTok-3.0.9}/setup.cfg RENAMED
File without changes