Make SupabaseVectorStore a subclass of BasePydanticVectorStore

llama-index-integrations/vector_stores/llama-index-vector-stores-supabase/llama_index/vector_stores/supabase/base.py

@@ -1,14 +1,16 @@
 import logging
 import math
 from collections import defaultdict
-from typing import Any, List
+from typing import Any, List, Optional
 
 import vecs
+from vecs.collection import Collection
 from llama_index.core.constants import DEFAULT_EMBEDDING_DIM
 from llama_index.core.schema import BaseNode, TextNode
+from llama_index.core.bridge.pydantic import PrivateAttr
 from llama_index.core.vector_stores.types import (
     MetadataFilters,
-    VectorStore,
+    BasePydanticVectorStore,
     VectorStoreQuery,
     VectorStoreQueryResult,
 )
@@ -22,7 +24,7 @@
 logger = logging.getLogger(__name__)
 
 
-class SupabaseVectorStore(VectorStore):
+class SupabaseVectorStore(BasePydanticVectorStore):
     """Supbabase Vector.
 
     In this vector store, embeddings are stored in Postgres table using pgvector.
@@ -41,6 +43,8 @@ class SupabaseVectorStore(VectorStore):
 
     stores_text = True
     flat_metadata = False
+    _client: Optional[Any] = PrivateAttr()
+    _collection: Optional[Collection] = PrivateAttr()
 
     def __init__(
         self,
@@ -49,17 +53,17 @@ def __init__(
         dimension: int = DEFAULT_EMBEDDING_DIM,
         **kwargs: Any,
     ) -> None:
-        """Init params."""
-        client = vecs.create_client(postgres_connection_string)
+        super().__init__()
+        self._client = vecs.create_client(postgres_connection_string)
 
         try:
-            self._collection = client.get_collection(name=collection_name)
+            self._collection = self._client.get_collection(name=collection_name)
         except CollectionNotFound:
             logger.info(
                 f"Collection {collection_name} does not exist, "
                 f"try creating one with dimension={dimension}"
             )
-            self._collection = client.create_collection(
+            self._collection = self._client.create_collection(
                 name=collection_name, dimension=dimension
             )
 
@@ -71,7 +75,7 @@ def client(self) -> None:
     def _to_vecs_filters(self, filters: MetadataFilters) -> Any:
         """Convert llama filters to vecs filters. $eq is the only supported operator."""
         vecs_filter = defaultdict(list)
-        filter_cond = f"${filters.condition}"
+        filter_cond = f"${filters.condition.value}"
 
         for f in filters.legacy_filters():
             sub_filter = {}

llama-index-integrations/vector_stores/llama-index-vector-stores-supabase/tests/test_vector_stores_supabase.py

@@ -1,7 +1,7 @@
-from llama_index.core.vector_stores.types import VectorStore
+from llama_index.core.vector_stores.types import BasePydanticVectorStore
 from llama_index.vector_stores.supabase import SupabaseVectorStore
 
 
 def test_class():
     names_of_base_classes = [b.__name__ for b in SupabaseVectorStore.__mro__]
-    assert VectorStore.__name__ in names_of_base_classes
+    assert BasePydanticVectorStore.__name__ in names_of_base_classes