rakam-systems-vectorstore 0.1.1rc7__py3-none-any.whl

rakam_systems_vectorstore/__init__.py

@@ -0,0 +1,93 @@
+"""
+AI Vector Store Package
+
+Provides modular, configurable vector storage and retrieval components
+leveraging PGVector for efficient semantic search.
+
+Key Components:
+- ConfigurablePgVectorStore: Enhanced vector store with full configuration support
+- AdaptiveLoader: Automatically detects and processes various file types
+- ConfigurableEmbeddings: Multi-backend embedding model support
+- VectorStoreConfig: Comprehensive configuration management
+
+Quick Start:
+    from ai_vectorstore import ConfigurablePgVectorStore, VectorStoreConfig
+    
+    config = VectorStoreConfig()
+    store = ConfigurablePgVectorStore(config=config)
+    store.setup()
+"""
+
+from rakam_systems_vectorstore.core import Node, NodeMetadata, VSFile
+from rakam_systems_vectorstore.config import (
+    VectorStoreConfig,
+    EmbeddingConfig,
+    DatabaseConfig,
+    SearchConfig,
+    IndexConfig,
+    load_config,
+)
+
+# Lazy imports for Django model-dependent components
+# These will be imported when accessed, after Django setup
+
+
+def __getattr__(name):
+    """Lazy import Django-dependent components."""
+    if name == "ConfigurablePgVectorStore":
+        from rakam_systems_vectorstore.components.vectorstore.configurable_pg_vector_store import (
+            ConfigurablePgVectorStore,
+        )
+        return ConfigurablePgVectorStore
+    elif name == "PgVectorStore":
+        from rakam_systems_vectorstore.components.vectorstore.pg_vector_store import PgVectorStore
+        return PgVectorStore
+    elif name == "AdaptiveLoader":
+        from rakam_systems_vectorstore.components.loader.adaptive_loader import AdaptiveLoader
+        return AdaptiveLoader
+    elif name == "create_adaptive_loader":
+        from rakam_systems_vectorstore.components.loader.adaptive_loader import create_adaptive_loader
+        return create_adaptive_loader
+    elif name == "ConfigurableEmbeddings":
+        from rakam_systems_vectorstore.components.embedding_model.configurable_embeddings import (
+            ConfigurableEmbeddings,
+        )
+        return ConfigurableEmbeddings
+    elif name == "create_embedding_model":
+        from rakam_systems_vectorstore.components.embedding_model.configurable_embeddings import (
+            create_embedding_model,
+        )
+        return create_embedding_model
+    elif name == "FaissVectorStore":
+        from rakam_systems_vectorstore.components.vectorstore.faiss_vector_store import FaissStore as FaissVectorStore
+        return FaissVectorStore
+    raise AttributeError(f"module {__name__!r} has no attribute {name!r}")
+
+
+__version__ = "1.0.0"
+
+__all__ = [
+    # Core data structures
+    "Node",
+    "NodeMetadata",
+    "VSFile",
+
+    # Configuration
+    "VectorStoreConfig",
+    "EmbeddingConfig",
+    "DatabaseConfig",
+    "SearchConfig",
+    "IndexConfig",
+    "load_config",
+
+    # New configurable components
+    "ConfigurablePgVectorStore",
+    "AdaptiveLoader",
+    "create_adaptive_loader",
+    "ConfigurableEmbeddings",
+    "create_embedding_model",
+
+    # Original components (backward compatibility)
+    "PgVectorStore",
+    "FaissVectorStore",
+]

rakam_systems_vectorstore/components/chunker/__init__.py

@@ -0,0 +1,19 @@
+from .text_chunker import TextChunker, create_text_chunker
+from .advanced_chunker import (
+    AdvancedChunker,
+    create_chunker,
+    MDTableSerializerProvider,
+    ImgPlaceholderSerializerProvider,
+    ImgAnnotationSerializerProvider,
+)
+
+__all__ = [
+    "TextChunker",
+    "create_text_chunker",
+    "AdvancedChunker",
+    "create_chunker",
+    "MDTableSerializerProvider",
+    "ImgPlaceholderSerializerProvider",
+    "ImgAnnotationSerializerProvider",
+]
+