mawo-slovnet 1.0.0__tar.gz → 1.0.2__tar.gz

{mawo_slovnet-1.0.0 → mawo_slovnet-1.0.2}/PKG-INFO

@@ -1,10 +1,10 @@
 Metadata-Version: 2.4
 Name: mawo-slovnet
-Version: 1.0.0
+Version: 1.0.2
 Summary: Нейросетевые модели для русского языка: NER, морфология и синтаксис с автоматической загрузкой
 Author-email: MAWO Team <team@mawo.ru>
 Maintainer-email: MAWO Team <team@mawo.ru>
-License-Expression: MIT
+License: MIT
 Project-URL: Homepage, https://github.com/mawo-ru/mawo-slovnet
 Project-URL: Documentation, https://github.com/mawo-ru/mawo-slovnet#readme
 Project-URL: Repository, https://github.com/mawo-ru/mawo-slovnet
@@ -14,6 +14,7 @@ Keywords: nlp,russian,ner,morphology,syntax,slovnet,mawo
 Classifier: Development Status :: 5 - Production/Stable
 Classifier: Intended Audience :: Developers
 Classifier: Intended Audience :: Science/Research
+Classifier: License :: OSI Approved :: MIT License
 Classifier: Natural Language :: Russian
 Classifier: Operating System :: OS Independent
 Classifier: Programming Language :: Python :: 3

{mawo_slovnet-1.0.0 → mawo_slovnet-1.0.2}/mawo_slovnet/__init__.py

@@ -7,12 +7,14 @@ Features:
 - Hybrid mode: DL models + rule-based fallback
 - 100% качество оригинального SlovNet (если модели доступны)
 """
+
 from __future__ import annotations
 
 import logging
 import os
 import sys
-from typing import Any
+from pathlib import Path
+from typing import Any, Optional
 
 logger = logging.getLogger(__name__)
 
@@ -104,7 +106,9 @@ class EnhancedSlovNetLoader:
             self.slovnet_available = True
             logger.info("✅ Original slovnet package available")
         except ImportError:
-            logger.info("ℹ️  Original slovnet package not installed (will try to use numpy-only mode)")
+            logger.info(
+                "ℹ️  Original slovnet package not installed (will try to use numpy-only mode)"
+            )
 
     def ensure_models_downloaded(self) -> bool:
         """Проверяет и загружает модели если нужно.
@@ -123,9 +127,7 @@ class EnhancedSlovNetLoader:
         cache_info = downloader.get_cache_info()
 
         # Check if any models are cached
-        cached_models = [
-            name for name, info in cache_info["models"].items() if info["cached"]
-        ]
+        cached_models = [name for name, info in cache_info["models"].items() if info["cached"]]
 
         if cached_models:
             logger.info(f"✅ Found cached models: {', '.join(cached_models)}")
@@ -184,6 +186,7 @@ class EnhancedSlovNetLoader:
                     sys.path.insert(0, str(model_dir))
 
             # Import slovnet components
+            import slovnet
             from slovnet import NewsEmbedding as _NewsEmbedding
             from slovnet import NewsMorphTagger as _NewsMorphTagger
             from slovnet import NewsNERTagger as _NewsNERTagger
@@ -357,11 +360,19 @@ def get_model_info() -> dict[str, Any]:
 __version__ = "2.0.0-mawo-enhanced"
 __author__ = "MAWO Team (based on SlovNet by Alexander Kukushkin)"
 
+# Алиасы для удобства использования
+NER = NewsNERTagger
+Morph = NewsMorphTagger
+Syntax = NewsSyntaxParser
+
 __all__ = [
     "NewsEmbedding",
     "NewsMorphTagger",
     "NewsNERTagger",
     "NewsSyntaxParser",
+    "NER",
+    "Morph",
+    "Syntax",
     "create_morphology_tagger",
     "download_models",
     "get_model_info",

{mawo_slovnet-1.0.0 → mawo_slovnet-1.0.2}/mawo_slovnet/model_downloader.py

@@ -5,13 +5,16 @@ Based on:
 - SlovNet v0.6.0 (github.com/natasha/slovnet)
 - MAWO offline-first architecture
 """
+
 from __future__ import annotations
 
+import hashlib
 import logging
+import os
 import shutil
 import tarfile
 from pathlib import Path
-from typing import Any
+from typing import Any, Optional
 from urllib.request import urlopen
 
 logger = logging.getLogger(__name__)
@@ -129,7 +132,7 @@ class SlovNetModelDownloader:
             )
 
             # Extract tar archive
-            logger.info("📦 Extracting model archive...")
+            logger.info(f"📦 Extracting model archive...")
             with tarfile.open(tar_path, "r") as tar:
                 tar.extractall(temp_dir)
 
@@ -287,9 +290,7 @@ class SlovNetModelDownloader:
             size_mb = 0
             if cached:
                 # Calculate directory size
-                size_bytes = sum(
-                    f.stat().st_size for f in model_dir.rglob("*") if f.is_file()
-                )
+                size_bytes = sum(f.stat().st_size for f in model_dir.rglob("*") if f.is_file())
                 size_mb = size_bytes / (1024 * 1024)
                 info["total_size_mb"] += size_mb
 

{mawo_slovnet-1.0.0 → mawo_slovnet-1.0.2}/mawo_slovnet.egg-info/PKG-INFO

@@ -1,10 +1,10 @@
 Metadata-Version: 2.4
 Name: mawo-slovnet
-Version: 1.0.0
+Version: 1.0.2
 Summary: Нейросетевые модели для русского языка: NER, морфология и синтаксис с автоматической загрузкой
 Author-email: MAWO Team <team@mawo.ru>
 Maintainer-email: MAWO Team <team@mawo.ru>
-License-Expression: MIT
+License: MIT
 Project-URL: Homepage, https://github.com/mawo-ru/mawo-slovnet
 Project-URL: Documentation, https://github.com/mawo-ru/mawo-slovnet#readme
 Project-URL: Repository, https://github.com/mawo-ru/mawo-slovnet
@@ -14,6 +14,7 @@ Keywords: nlp,russian,ner,morphology,syntax,slovnet,mawo
 Classifier: Development Status :: 5 - Production/Stable
 Classifier: Intended Audience :: Developers
 Classifier: Intended Audience :: Science/Research
+Classifier: License :: OSI Approved :: MIT License
 Classifier: Natural Language :: Russian
 Classifier: Operating System :: OS Independent
 Classifier: Programming Language :: Python :: 3

{mawo_slovnet-1.0.0 → mawo_slovnet-1.0.2}/pyproject.toml

@@ -4,11 +4,11 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "mawo-slovnet"
-version = "1.0.0"
+version = "1.0.2"
 description = "Нейросетевые модели для русского языка: NER, морфология и синтаксис с автоматической загрузкой"
 readme = "README.md"
 requires-python = ">=3.10"
-license = "MIT"
+license = {text = "MIT"}
 authors = [
     {name = "MAWO Team", email = "team@mawo.ru"},
 ]
@@ -28,6 +28,7 @@ classifiers = [
     "Development Status :: 5 - Production/Stable",
     "Intended Audience :: Developers",
     "Intended Audience :: Science/Research",
+    "License :: OSI Approved :: MIT License",
     "Natural Language :: Russian",
     "Operating System :: OS Independent",
     "Programming Language :: Python :: 3",
@@ -77,10 +78,8 @@ target-version = ["py310", "py311", "py312"]
 [tool.ruff]
 line-length = 100
 target-version = "py310"
-
-[tool.ruff.lint]
 select = ["E", "F", "I", "N", "W", "B", "UP"]
-ignore = ["E501", "N802"]  # N802: ignore function name case (NewsNERTagger, etc.)
+ignore = ["E501"]
 
 [tool.mypy]
 python_version = "3.10"

{mawo_slovnet-1.0.0 → mawo_slovnet-1.0.2}/tests/test_integration.py

@@ -3,9 +3,8 @@
 Тестируют библиотеку как самодостаточный проект
 """
 
-from pathlib import Path
-
 import pytest
+from pathlib import Path
 
 
 class TestImports:
@@ -15,8 +14,6 @@ class TestImports:
         """Тест: главный модуль импортируется"""
         try:
             import mawo_slovnet
-
-            assert mawo_slovnet is not None
         except ImportError as e:
             pytest.fail(f"Failed to import mawo_slovnet: {e}")
 
@@ -24,8 +21,6 @@ class TestImports:
         """Тест: класс NER импортируется"""
         try:
             from mawo_slovnet import NER
-
-            assert NER is not None
         except ImportError as e:
             pytest.fail(f"Failed to import NER: {e}")
 
@@ -33,8 +28,6 @@ class TestImports:
         """Тест: класс Morph импортируется"""
         try:
             from mawo_slovnet import Morph
-
-            assert Morph is not None
         except ImportError as e:
             pytest.fail(f"Failed to import Morph: {e}")
 
@@ -42,8 +35,6 @@ class TestImports:
         """Тест: класс Syntax импортируется"""
         try:
             from mawo_slovnet import Syntax
-
-            assert Syntax is not None
         except ImportError as e:
             pytest.fail(f"Failed to import Syntax: {e}")
 
@@ -67,7 +58,7 @@ class TestNERInitialization:
 
         ner = NER()
         # Проверяем, что модель загружена
-        assert hasattr(ner, 'model') or hasattr(ner, '_model')
+        assert hasattr(ner, "model") or hasattr(ner, "_model")
 
 
 class TestMorphInitialization:
@@ -127,7 +118,7 @@ class TestNERFunctionality:
         assert len(result) > 0, "Expected to find at least one entity"
 
         # Проверяем, что есть сущность типа PER
-        has_person = any(span.type == 'PER' for span in result if hasattr(span, 'type'))
+        has_person = any(span.type == "PER" for span in result if hasattr(span, "type"))
         assert has_person, "Expected to find PER entity for 'Александр Пушкин'"
 
     def test_ner_finds_location(self):
@@ -139,7 +130,7 @@ class TestNERFunctionality:
         result = ner(text)
 
         assert len(result) > 0
-        has_location = any(span.type == 'LOC' for span in result if hasattr(span, 'type'))
+        has_location = any(span.type == "LOC" for span in result if hasattr(span, "type"))
         assert has_location, "Expected to find LOC entity for 'Санкт-Петербурге'"
 
     def test_ner_empty_text(self):
@@ -190,7 +181,7 @@ class TestMorphFunctionality:
         assert len(result) > 0
         # Проверяем, что есть разметка
         first_token = result[0]
-        assert hasattr(first_token, 'pos') or hasattr(first_token, 'tag')
+        assert hasattr(first_token, "pos") or hasattr(first_token, "tag")
 
     def test_morph_empty_text(self):
         """Тест: Morph обрабатывает пустой текст"""
@@ -230,7 +221,7 @@ class TestSyntaxFunctionality:
         assert len(result) > 0
         # Проверяем, что есть синтаксические связи
         first_token = result[0]
-        assert hasattr(first_token, 'head') or hasattr(first_token, 'rel')
+        assert hasattr(first_token, "head") or hasattr(first_token, "rel")
 
 
 class TestDataFiles:
@@ -238,32 +229,35 @@ class TestDataFiles:
 
     def test_ner_model_file_exists(self):
         """Тест: файл модели NER существует"""
+        from pathlib import Path
         import mawo_slovnet
 
         # Находим директорию модуля
         module_path = Path(mawo_slovnet.__file__).parent
 
         # Проверяем наличие модели NER
-        ner_model = module_path / 'slovnet_ner_news_v1.tar.neural.gz'
+        ner_model = module_path / "slovnet_ner_news_v1.tar.neural.gz"
         assert ner_model.exists(), f"NER model not found at {ner_model}"
         assert ner_model.stat().st_size > 1_000_000, "NER model file is too small"
 
     def test_morph_model_file_exists(self):
         """Тест: файл модели Morph существует"""
+        from pathlib import Path
         import mawo_slovnet
 
         module_path = Path(mawo_slovnet.__file__).parent
-        morph_model = module_path / 'slovnet_morph_news_v1.tar.neural.gz'
+        morph_model = module_path / "slovnet_morph_news_v1.tar.neural.gz"
 
         assert morph_model.exists(), f"Morph model not found at {morph_model}"
         assert morph_model.stat().st_size > 1_000_000, "Morph model file is too small"
 
     def test_syntax_model_file_exists(self):
         """Тест: файл модели Syntax существует"""
+        from pathlib import Path
         import mawo_slovnet
 
         module_path = Path(mawo_slovnet.__file__).parent
-        syntax_model = module_path / 'slovnet_syntax_news_v1.tar.neural.gz'
+        syntax_model = module_path / "slovnet_syntax_news_v1.tar.neural.gz"
 
         assert syntax_model.exists(), f"Syntax model not found at {syntax_model}"
         assert syntax_model.stat().st_size > 1_000_000, "Syntax model file is too small"
@@ -349,5 +343,5 @@ class TestMultipleInstances:
         assert syntax_result is not None
 
 
-if __name__ == '__main__':
-    pytest.main([__file__, '-v', '--tb=short'])
+if __name__ == "__main__":
+    pytest.main([__file__, "-v", "--tb=short"])