Merge 05d3fffbdf into 21bae296f2

llama_stack/providers/remote/vector_io/chroma/chroma.py

@@ -112,6 +112,9 @@ class ChromaIndex(EmbeddingIndex):
     ) -> QueryChunksResponse:
         raise NotImplementedError("Keyword search is not supported in Chroma")
 
+    async def delete_chunk(self, chunk_id: str) -> None:
+        raise NotImplementedError("delete_chunk is not supported in Chroma")
+
     async def query_hybrid(
         self,
         embedding: NDArray,
@@ -208,3 +211,6 @@ class ChromaVectorIOAdapter(OpenAIVectorStoreMixin, VectorIO, VectorDBsProtocolP
         index = VectorDBWithIndex(vector_db, ChromaIndex(self.client, collection), self.inference_api)
         self.cache[vector_db_id] = index
         return index
+
+    async def _delete_openai_chunk_from_vector_store(self, store_id: str, chunk_id: str) -> None:
+        raise NotImplementedError("OpenAI Vector Stores API is not supported in Chroma")

llama_stack/providers/remote/vector_io/milvus/milvus.py

@@ -247,6 +247,16 @@ class MilvusIndex(EmbeddingIndex):
     ) -> QueryChunksResponse:
         raise NotImplementedError("Hybrid search is not supported in Milvus")
 
+    async def delete_chunk(self, chunk_id: str) -> None:
+        """Remove a chunk from the Milvus collection."""
+        try:
+            await asyncio.to_thread(
+                self.client.delete, collection_name=self.collection_name, filter=f'chunk_id == "{chunk_id}"'
+            )
+        except Exception as e:
+            logger.error(f"Error deleting chunk {chunk_id} from Milvus collection {self.collection_name}: {e}")
+            raise
+
 
 class MilvusVectorIOAdapter(OpenAIVectorStoreMixin, VectorIO, VectorDBsProtocolPrivate):
     def __init__(
@@ -369,3 +379,12 @@ class MilvusVectorIOAdapter(OpenAIVectorStoreMixin, VectorIO, VectorDBsProtocolP
                 )
 
         return await index.query_chunks(query, params)
+
+    async def _delete_openai_chunk_from_vector_store(self, store_id: str, chunk_id: str) -> None:
+        """Delete a chunk from a milvus vector store."""
+        index = await self._get_and_cache_vector_db_index(store_id)
+        if not index:
+            raise ValueError(f"Vector DB {store_id} not found")
+
+        # Use the index's delete_chunk method
+        await index.index.delete_chunk(chunk_id)

llama_stack/providers/remote/vector_io/pgvector/__init__.py

@@ -12,6 +12,6 @@ from .config import PGVectorVectorIOConfig
 async def get_adapter_impl(config: PGVectorVectorIOConfig, deps: dict[Api, ProviderSpec]):
     from .pgvector import PGVectorVectorIOAdapter
 
-    impl = PGVectorVectorIOAdapter(config, deps[Api.inference])
+    impl = PGVectorVectorIOAdapter(config, deps[Api.inference], deps.get(Api.files, None))
     await impl.initialize()
     return impl

llama_stack/providers/remote/vector_io/pgvector/pgvector.py

@@ -99,7 +99,7 @@ class PGVectorIndex(EmbeddingIndex):
         for i, chunk in enumerate(chunks):
             values.append(
                 (
-                    f"{chunk.metadata['document_id']}:chunk-{i}",
+                    f"{chunk.chunk_id}",
                     Json(chunk.model_dump()),
                     embeddings[i].tolist(),
                 )
@@ -159,6 +159,11 @@ class PGVectorIndex(EmbeddingIndex):
         with self.conn.cursor(cursor_factory=psycopg2.extras.DictCursor) as cur:
             cur.execute(f"DROP TABLE IF EXISTS {self.table_name}")
 
+    async def delete_chunk(self, chunk_id: str) -> None:
+        """Remove a chunk from the PostgreSQL table."""
+        with self.conn.cursor(cursor_factory=psycopg2.extras.DictCursor) as cur:
+            cur.execute(f"DELETE FROM {self.table_name} WHERE id = %s", (chunk_id,))
+
 
 class PGVectorVectorIOAdapter(OpenAIVectorStoreMixin, VectorIO, VectorDBsProtocolPrivate):
     def __init__(
@@ -265,3 +270,12 @@ class PGVectorVectorIOAdapter(OpenAIVectorStoreMixin, VectorIO, VectorDBsProtoco
         index = PGVectorIndex(vector_db, vector_db.embedding_dimension, self.conn)
         self.cache[vector_db_id] = VectorDBWithIndex(vector_db, index, self.inference_api)
         return self.cache[vector_db_id]
+
+    async def _delete_openai_chunk_from_vector_store(self, store_id: str, chunk_id: str) -> None:
+        """Delete a chunk from a PostgreSQL vector store."""
+        index = await self._get_and_cache_vector_db_index(store_id)
+        if not index:
+            raise ValueError(f"Vector DB {store_id} not found")
+
+        # Use the index's delete_chunk method
+        await index.index.delete_chunk(chunk_id)

llama_stack/providers/remote/vector_io/qdrant/qdrant.py

@@ -82,6 +82,9 @@ class QdrantIndex(EmbeddingIndex):
 
         await self.client.upsert(collection_name=self.collection_name, points=points)
 
+    async def delete_chunk(self, chunk_id: str) -> None:
+        raise NotImplementedError("delete_chunk is not supported in qdrant")
+
     async def query_vector(self, embedding: NDArray, k: int, score_threshold: float) -> QueryChunksResponse:
         results = (
             await self.client.query_points(
@@ -307,3 +310,6 @@ class QdrantVectorIOAdapter(VectorIO, VectorDBsProtocolPrivate):
         file_id: str,
     ) -> VectorStoreFileObject:
         raise NotImplementedError("OpenAI Vector Stores API is not supported in Qdrant")
+
+    async def _delete_openai_chunk_from_vector_store(self, store_id: str, chunk_id: str) -> None:
+        raise NotImplementedError("OpenAI Vector Stores API is not supported in Qdrant")

llama_stack/providers/remote/vector_io/weaviate/weaviate.py

@@ -66,6 +66,9 @@ class WeaviateIndex(EmbeddingIndex):
         # TODO: make this async friendly
         collection.data.insert_many(data_objects)
 
+    async def delete_chunk(self, chunk_id: str) -> None:
+        raise NotImplementedError("delete_chunk is not supported in Chroma")
+
     async def query_vector(self, embedding: NDArray, k: int, score_threshold: float) -> QueryChunksResponse:
         collection = self.client.collections.get(self.collection_name)
 
@@ -264,3 +267,6 @@ class WeaviateVectorIOAdapter(
 
     async def _delete_openai_vector_store_file_from_storage(self, store_id: str, file_id: str) -> None:
         raise NotImplementedError("OpenAI Vector Stores API is not supported in Weaviate")
+
+    async def _delete_openai_chunk_from_vector_store(self, store_id: str, chunk_id: str) -> None:
+        raise NotImplementedError("OpenAI Vector Stores API is not supported in Weaviate")