thoth-dbmanager 0.5.1__tar.gz → 0.5.2__tar.gz

{thoth_dbmanager-0.5.1/thoth_dbmanager.egg-info → thoth_dbmanager-0.5.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: thoth_dbmanager
-Version: 0.5.1
+Version: 0.5.2
 Summary: A Python library for managing SQL databases with support for multiple database types, LSH-based similarity search, and a modern plugin architecture.
 Author-email: Marco Pancotti <mp@tylconsulting.it>
 Project-URL: Homepage, https://github.com/mptyl/thoth_dbmanager
@@ -35,10 +35,15 @@ Requires-Dist: mariadb>=1.1.0; extra == "mariadb"
 Provides-Extra: sqlserver
 Requires-Dist: pyodbc>=4.0.0; extra == "sqlserver"
 Provides-Extra: sqlite
+Provides-Extra: embeddings
+Requires-Dist: sentence-transformers>=2.0.0; extra == "embeddings"
+Requires-Dist: numpy>=1.21.0; extra == "embeddings"
 Provides-Extra: all
 Requires-Dist: psycopg2-binary>=2.9.0; extra == "all"
 Requires-Dist: mariadb>=1.1.0; extra == "all"
 Requires-Dist: pyodbc>=4.0.0; extra == "all"
+Requires-Dist: sentence-transformers>=2.0.0; extra == "all"
+Requires-Dist: numpy>=1.21.0; extra == "all"
 Provides-Extra: dev
 Requires-Dist: pytest>=7.0.0; extra == "dev"
 Requires-Dist: pytest-cov>=4.0.0; extra == "dev"

{thoth_dbmanager-0.5.1 → thoth_dbmanager-0.5.2}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "thoth_dbmanager"
-version = "0.5.1"
+version = "0.5.2"
 authors = [
   { name="Marco Pancotti", email="mp@tylconsulting.it" },
 ]
@@ -39,12 +39,15 @@ postgresql = ["psycopg2-binary>=2.9.0"]
 mariadb = ["mariadb>=1.1.0"]
 sqlserver = ["pyodbc>=4.0.0"]
 sqlite = []
+embeddings = ["sentence-transformers>=2.0.0", "numpy>=1.21.0"]
 
 # Convenience groups
 all = [
     "psycopg2-binary>=2.9.0",
     "mariadb>=1.1.0",
     "pyodbc>=4.0.0",
+    "sentence-transformers>=2.0.0",
+    "numpy>=1.21.0",
 ]
 
 # Development dependencies

{thoth_dbmanager-0.5.1 → thoth_dbmanager-0.5.2}/thoth_dbmanager/core/interfaces.py

@@ -279,3 +279,84 @@ class DbPlugin(ABC):
         if not self.adapter:
             raise RuntimeError("Plugin not initialized")
         return self.adapter.get_unique_values()
+    
+    def get_embedding_function(self):
+        """
+        Get the embedding function for similarity computations.
+        
+        Returns:
+            SafeSentenceTransformer: An embedding function with encode method
+        """
+        try:
+            # Import SafeSentenceTransformer
+            try:
+                from sentence_transformers import SentenceTransformer
+                import logging
+                
+                logger = logging.getLogger(__name__)
+                
+                class SafeSentenceTransformer:
+                    """
+                    Wrapper for SentenceTransformer that handles PyTorch meta tensor issues.
+                    """
+                    def __init__(self, model_name_or_path: str):
+                        self.model_name_or_path = model_name_or_path
+                        self._model = None
+
+                    def _get_model(self):
+                        """Lazy initialization of the SentenceTransformer model."""
+                        if self._model is None:
+                            try:
+                                logger.info(f"Initializing SentenceTransformer with model: {self.model_name_or_path}")
+                                self._model = SentenceTransformer(
+                                    model_name_or_path=self.model_name_or_path,
+                                    device='cpu'  # Explicitly set device to CPU to avoid meta tensor issues
+                                )
+                                logger.info("SentenceTransformer initialized successfully")
+                            except Exception as e:
+                                logger.error(f"Failed to initialize SentenceTransformer: {e}")
+                                # Try alternative initialization approach
+                                try:
+                                    logger.info("Trying alternative initialization approach...")
+                                    self._model = SentenceTransformer(self.model_name_or_path)
+                                    # Move to CPU explicitly after initialization
+                                    self._model = self._model.to('cpu')
+                                    logger.info("Alternative initialization successful")
+                                except Exception as e2:
+                                    logger.error(f"Alternative initialization also failed: {e2}")
+                                    raise e2
+                        return self._model
+
+                    def encode(self, sentences, **kwargs):
+                        """Encode sentences using the underlying SentenceTransformer model."""
+                        model = self._get_model()
+                        return model.encode(sentences, **kwargs)
+                
+                return SafeSentenceTransformer(
+                    model_name_or_path="paraphrase-multilingual-MiniLM-L12-v2"
+                )
+                
+            except ImportError:
+                import logging
+                logger = logging.getLogger(__name__)
+                logger.warning("sentence_transformers not available, creating dummy embedding function")
+                # Create a dummy embedding function for testing
+                class DummyEmbeddingFunction:
+                    def encode(self, sentences, **kwargs):
+                        import numpy as np
+                        # Return dummy embeddings - same shape for all sentences
+                        return np.random.rand(len(sentences), 384)  # 384 is typical embedding size
+                
+                return DummyEmbeddingFunction()
+                
+        except Exception as e:
+            import logging
+            logger = logging.getLogger(__name__)
+            logger.error(f"Failed to create embedding function: {e}")
+            # Return a basic dummy function as fallback
+            class BasicDummyEmbeddingFunction:
+                def encode(self, sentences, **kwargs):
+                    import numpy as np
+                    return np.random.rand(len(sentences), 384)
+            
+            return BasicDummyEmbeddingFunction()

{thoth_dbmanager-0.5.1 → thoth_dbmanager-0.5.2}/thoth_dbmanager/plugins/sqlite.py

@@ -225,3 +225,4 @@ class SQLitePlugin(DbPlugin):
             return self.adapter.health_check()
         else:
             return False
+    

{thoth_dbmanager-0.5.1 → thoth_dbmanager-0.5.2/thoth_dbmanager.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: thoth_dbmanager
-Version: 0.5.1
+Version: 0.5.2
 Summary: A Python library for managing SQL databases with support for multiple database types, LSH-based similarity search, and a modern plugin architecture.
 Author-email: Marco Pancotti <mp@tylconsulting.it>
 Project-URL: Homepage, https://github.com/mptyl/thoth_dbmanager
@@ -35,10 +35,15 @@ Requires-Dist: mariadb>=1.1.0; extra == "mariadb"
 Provides-Extra: sqlserver
 Requires-Dist: pyodbc>=4.0.0; extra == "sqlserver"
 Provides-Extra: sqlite
+Provides-Extra: embeddings
+Requires-Dist: sentence-transformers>=2.0.0; extra == "embeddings"
+Requires-Dist: numpy>=1.21.0; extra == "embeddings"
 Provides-Extra: all
 Requires-Dist: psycopg2-binary>=2.9.0; extra == "all"
 Requires-Dist: mariadb>=1.1.0; extra == "all"
 Requires-Dist: pyodbc>=4.0.0; extra == "all"
+Requires-Dist: sentence-transformers>=2.0.0; extra == "all"
+Requires-Dist: numpy>=1.21.0; extra == "all"
 Provides-Extra: dev
 Requires-Dist: pytest>=7.0.0; extra == "dev"
 Requires-Dist: pytest-cov>=4.0.0; extra == "dev"

{thoth_dbmanager-0.5.1 → thoth_dbmanager-0.5.2}/thoth_dbmanager.egg-info/requires.txt

@@ -9,6 +9,8 @@ requests>=2.25.0
 psycopg2-binary>=2.9.0
 mariadb>=1.1.0
 pyodbc>=4.0.0
+sentence-transformers>=2.0.0
+numpy>=1.21.0
 
 [dev]
 pytest>=7.0.0
@@ -22,6 +24,10 @@ twine>=4.0.0
 psutil>=5.8.0
 docker>=6.0.0
 
+[embeddings]
+sentence-transformers>=2.0.0
+numpy>=1.21.0
+
 [mariadb]
 mariadb>=1.1.0