refactor: moving all language related files in languages folder

AndyTheFactory · Jan 23, 2024 · 9a48739 · 9a48739
1 parent 7a08fc2
commit 9a48739
Show file tree

Hide file tree

Showing 10 changed files with 3 additions and 2 deletions.
diff --git a/.pre-commit-config.yaml b/.pre-commit-config.yaml
@@ -42,6 +42,7 @@ repos:
                 .*/stopwords.*\.txt|
                 tests/data/.*|
                 newspaper/languages.py|
+                newspaper/languages/.*|
                 newspaper/resources/.*
             )$
     additional_dependencies:

diff --git a/newspaper/language/__init__.py b/newspaper/language/__init__.py
diff --git a/newspaper/languages.py → newspaper/languages/__init__.py b/newspaper/languages.py → newspaper/languages/__init__.py
diff --git a/newspaper/language/ar.py → newspaper/languages/ar.py b/newspaper/language/ar.py → newspaper/languages/ar.py
diff --git a/newspaper/language/bn.py → newspaper/languages/bn.py b/newspaper/language/bn.py → newspaper/languages/bn.py
diff --git a/newspaper/language/ja.py → newspaper/languages/ja.py b/newspaper/language/ja.py → newspaper/languages/ja.py
diff --git a/newspaper/language/ko.py → newspaper/languages/ko.py b/newspaper/language/ko.py → newspaper/languages/ko.py
diff --git a/newspaper/language/th.py → newspaper/languages/th.py b/newspaper/language/th.py → newspaper/languages/th.py
diff --git a/newspaper/language/zh.py → newspaper/languages/zh.py b/newspaper/language/zh.py → newspaper/languages/zh.py
diff --git a/newspaper/text.py b/newspaper/text.py
@@ -82,11 +82,11 @@ def __init__(self, language="en"):
             with open(stopwordsFile, "r", encoding="utf-8") as f:
                 self._cached_stop_words[language] = set(f.read().splitlines())
 
-        lang_module = Path(__file__).parent / "language" / f"{language}.py"
+        lang_module = Path(__file__).parent / "languages" / f"{language}.py"
         if lang_module.exists():
             import importlib
 
-            module = importlib.import_module(f"newspaper.language.{language}")
+            module = importlib.import_module(f"newspaper.languages.{language}")
             if not hasattr(module, "tokenizer"):
                 raise ValueError(
                     f"Language module {lang_module} has no tokenizer function!"