agno 2.2.13__py3-none-any.whl

agno/vectordb/qdrant/qdrant.py

@@ -0,0 +1,1134 @@
+from hashlib import md5
+from typing import Any, Dict, List, Optional, Union
+
+try:
+    from qdrant_client import AsyncQdrantClient, QdrantClient  # noqa: F401
+    from qdrant_client.http import models
+except ImportError:
+    raise ImportError(
+        "The `qdrant-client` package is not installed. Please install it via `pip install qdrant-client`."
+    )
+
+from agno.filters import FilterExpr
+from agno.knowledge.document import Document
+from agno.knowledge.embedder import Embedder
+from agno.knowledge.reranker.base import Reranker
+from agno.utils.log import log_debug, log_error, log_info, log_warning
+from agno.vectordb.base import VectorDb
+from agno.vectordb.distance import Distance
+from agno.vectordb.search import SearchType
+
+DEFAULT_DENSE_VECTOR_NAME = "dense"
+DEFAULT_SPARSE_VECTOR_NAME = "sparse"
+DEFAULT_SPARSE_MODEL = "Qdrant/bm25"
+
+
+class Qdrant(VectorDb):
+    """Vector DB implementation powered by Qdrant - https://qdrant.tech/"""
+
+    def __init__(
+        self,
+        collection: str,
+        name: Optional[str] = None,
+        description: Optional[str] = None,
+        id: Optional[str] = None,
+        embedder: Optional[Embedder] = None,
+        distance: Distance = Distance.cosine,
+        location: Optional[str] = None,
+        url: Optional[str] = None,
+        port: Optional[int] = 6333,
+        grpc_port: int = 6334,
+        prefer_grpc: bool = False,
+        https: Optional[bool] = None,
+        api_key: Optional[str] = None,
+        prefix: Optional[str] = None,
+        timeout: Optional[float] = None,
+        host: Optional[str] = None,
+        path: Optional[str] = None,
+        reranker: Optional[Reranker] = None,
+        search_type: SearchType = SearchType.vector,
+        dense_vector_name: str = DEFAULT_DENSE_VECTOR_NAME,
+        sparse_vector_name: str = DEFAULT_SPARSE_VECTOR_NAME,
+        hybrid_fusion_strategy: models.Fusion = models.Fusion.RRF,
+        fastembed_kwargs: Optional[dict] = None,
+        **kwargs,
+    ):
+        """
+        Args:
+            collection (str): Name of the Qdrant collection.
+            name (Optional[str]): Name of the vector database.
+            description (Optional[str]): Description of the vector database.
+            embedder (Optional[Embedder]): Optional embedder for automatic vector generation.
+            distance (Distance): Distance metric to use (default: cosine).
+            location (Optional[str]): `":memory:"` for in-memory, or str used as `url`. If `None`, use default host/port.
+            url (Optional[str]): Full URL (scheme, host, port, prefix). Overrides host/port if provided.
+            port (Optional[int]): REST API port (default: 6333).
+            grpc_port (int): gRPC interface port (default: 6334).
+            prefer_grpc (bool): Prefer gRPC over REST if True.
+            https (Optional[bool]): Use HTTPS if True.
+            api_key (Optional[str]): API key for Qdrant Cloud authentication.
+            prefix (Optional[str]): URL path prefix (e.g., "service/v1").
+            timeout (Optional[float]): Request timeout (REST: default 5s, gRPC: unlimited).
+            host (Optional[str]): Qdrant host (default: "localhost" if not specified).
+            path (Optional[str]): Path for local persistence (QdrantLocal).
+            reranker (Optional[Reranker]): Optional reranker for result refinement.
+            search_type (SearchType): Whether to use vector, keyword or hybrid search.
+            dense_vector_name (str): Dense vector name.
+            sparse_vector_name (str): Sparse vector name.
+            hybrid_fusion_strategy (models.Fusion): Strategy for hybrid fusion.
+            fastembed_kwargs (Optional[dict]): Keyword args for `fastembed.SparseTextEmbedding.__init__()`.
+            **kwargs: Keyword args for `qdrant_client.QdrantClient.__init__()`.
+        """
+        # Validate required parameters
+        if not collection:
+            raise ValueError("Collection name must be provided.")
+
+        # Dynamic ID generation based on unique identifiers
+        if id is None:
+            from agno.utils.string import generate_id
+
+            host_identifier = host or location or url or "localhost"
+            seed = f"{host_identifier}#{collection}"
+            id = generate_id(seed)
+
+        # Initialize base class with name, description, and generated ID
+        super().__init__(id=id, name=name, description=description)
+
+        # Collection attributes
+        self.collection: str = collection
+
+        # Embedder for embedding the document contents
+        if embedder is None:
+            from agno.knowledge.embedder.openai import OpenAIEmbedder
+
+            embedder = OpenAIEmbedder()
+            log_info("Embedder not provided, using OpenAIEmbedder as default.")
+
+        self.embedder: Embedder = embedder
+        self.dimensions: Optional[int] = self.embedder.dimensions
+
+        # Distance metric
+        self.distance: Distance = distance
+
+        # Qdrant client instance
+        self._client: Optional[QdrantClient] = None
+
+        # Qdrant async client instance
+        self._async_client: Optional[AsyncQdrantClient] = None
+
+        # Qdrant client arguments
+        self.location: Optional[str] = location
+        self.url: Optional[str] = url
+        self.port: Optional[int] = port
+        self.grpc_port: int = grpc_port
+        self.prefer_grpc: bool = prefer_grpc
+        self.https: Optional[bool] = https
+        self.api_key: Optional[str] = api_key
+        self.prefix: Optional[str] = prefix
+        self.timeout: Optional[float] = timeout
+        self.host: Optional[str] = host
+        self.path: Optional[str] = path
+
+        # Reranker instance
+        self.reranker: Optional[Reranker] = reranker
+
+        # Qdrant client kwargs
+        self.kwargs = kwargs
+
+        self.search_type = search_type
+        self.dense_vector_name = dense_vector_name
+        self.sparse_vector_name = sparse_vector_name
+        self.hybrid_fusion_strategy = hybrid_fusion_strategy
+
+        # TODO(v2.0.0): Remove backward compatibility for unnamed vectors
+        # TODO(v2.0.0): Make named vectors mandatory and simplify the codebase
+        self.use_named_vectors = search_type in [SearchType.hybrid]
+
+        if self.search_type in [SearchType.keyword, SearchType.hybrid]:
+            try:
+                from fastembed import SparseTextEmbedding  # type: ignore
+
+                default_kwargs = {"model_name": DEFAULT_SPARSE_MODEL}
+                if fastembed_kwargs:
+                    default_kwargs.update(fastembed_kwargs)
+
+                # Type ignore for mypy as SparseTextEmbedding constructor accepts flexible kwargs
+                self.sparse_encoder = SparseTextEmbedding(**default_kwargs)  # type: ignore
+
+            except ImportError as e:
+                raise ImportError(
+                    "To use keyword/hybrid search, install the `fastembed` extra with `pip install fastembed`."
+                ) from e
+
+    @property
+    def client(self) -> QdrantClient:
+        if self._client is None:
+            log_debug("Creating Qdrant Client")
+            self._client = QdrantClient(
+                location=self.location,
+                url=self.url,
+                port=self.port,
+                grpc_port=self.grpc_port,
+                prefer_grpc=self.prefer_grpc,
+                https=self.https,
+                api_key=self.api_key,
+                prefix=self.prefix,
+                timeout=int(self.timeout) if self.timeout is not None else None,
+                host=self.host,
+                path=self.path,
+                **self.kwargs,
+            )
+        return self._client
+
+    @property
+    def async_client(self) -> AsyncQdrantClient:
+        """Get or create the async Qdrant client."""
+        if self._async_client is None:
+            log_debug("Creating Async Qdrant Client")
+            self._async_client = AsyncQdrantClient(
+                location=self.location,
+                url=self.url,
+                port=self.port,
+                grpc_port=self.grpc_port,
+                prefer_grpc=self.prefer_grpc,
+                https=self.https,
+                api_key=self.api_key,
+                prefix=self.prefix,
+                timeout=int(self.timeout) if self.timeout is not None else None,
+                host=self.host,
+                path=self.path,
+                **self.kwargs,
+            )
+        return self._async_client
+
+    def create(self) -> None:
+        _distance = models.Distance.COSINE
+        if self.distance == Distance.l2:
+            _distance = models.Distance.EUCLID
+        elif self.distance == Distance.max_inner_product:
+            _distance = models.Distance.DOT
+
+        if not self.exists():
+            log_debug(f"Creating collection: {self.collection}")
+
+            # Configure vectors based on search type
+            if self.search_type == SearchType.vector:
+                # Maintain backward compatibility with unnamed vectors
+                vectors_config = models.VectorParams(size=self.dimensions or 1536, distance=_distance)
+            else:
+                # Use named vectors for hybrid search
+                vectors_config = {
+                    self.dense_vector_name: models.VectorParams(size=self.dimensions or 1536, distance=_distance)
+                }  # type: ignore
+
+            self.client.create_collection(
+                collection_name=self.collection,
+                vectors_config=vectors_config,
+                sparse_vectors_config={self.sparse_vector_name: models.SparseVectorParams()}
+                if self.search_type in [SearchType.keyword, SearchType.hybrid]
+                else None,
+            )
+
+    async def async_create(self) -> None:
+        """Create the collection asynchronously."""
+        # Collection distance
+        _distance = models.Distance.COSINE
+        if self.distance == Distance.l2:
+            _distance = models.Distance.EUCLID
+        elif self.distance == Distance.max_inner_product:
+            _distance = models.Distance.DOT
+
+        if not await self.async_exists():
+            log_debug(f"Creating collection asynchronously: {self.collection}")
+
+            # Configure vectors based on search type
+            if self.search_type == SearchType.vector:
+                # Maintain backward compatibility with unnamed vectors
+                vectors_config = models.VectorParams(size=self.dimensions or 1536, distance=_distance)
+            else:
+                # Use named vectors for hybrid search
+                vectors_config = {
+                    self.dense_vector_name: models.VectorParams(size=self.dimensions or 1536, distance=_distance)
+                }  # type: ignore
+
+            await self.async_client.create_collection(
+                collection_name=self.collection,
+                vectors_config=vectors_config,
+                sparse_vectors_config={self.sparse_vector_name: models.SparseVectorParams()}
+                if self.search_type in [SearchType.keyword, SearchType.hybrid]
+                else None,
+            )
+
+    def doc_exists(self, document: Document) -> bool:
+        """
+        Validating if the document exists or not
+
+        Args:
+            document (Document): Document to validate
+        """
+        if self.client:
+            cleaned_content = document.content.replace("\x00", "\ufffd")
+            doc_id = md5(cleaned_content.encode()).hexdigest()
+            collection_points = self.client.retrieve(
+                collection_name=self.collection,
+                ids=[doc_id],
+            )
+            return len(collection_points) > 0
+        return False
+
+    async def async_doc_exists(self, document: Document) -> bool:
+        """Check if a document exists asynchronously."""
+        cleaned_content = document.content.replace("\x00", "\ufffd")
+        doc_id = md5(cleaned_content.encode()).hexdigest()
+        collection_points = await self.async_client.retrieve(
+            collection_name=self.collection,
+            ids=[doc_id],
+        )
+        return len(collection_points) > 0
+
+    def name_exists(self, name: str) -> bool:
+        """
+        Validates if a document with the given name exists in the collection.
+
+        Args:
+            name (str): The name of the document to check.
+
+        Returns:
+            bool: True if a document with the given name exists, False otherwise.
+        """
+        if self.client:
+            scroll_result = self.client.scroll(
+                collection_name=self.collection,
+                scroll_filter=models.Filter(
+                    must=[models.FieldCondition(key="name", match=models.MatchValue(value=name))]
+                ),
+                limit=1,
+            )
+            return len(scroll_result[0]) > 0
+        return False
+
+    async def async_name_exists(self, name: str) -> bool:  # type: ignore[override]
+        """
+        Asynchronously validates if a document with the given name exists in the collection.
+
+        Args:
+            name (str): The name of the document to check.
+
+        Returns:
+            bool: True if a document with the given name exists, False otherwise.
+        """
+        if self.async_client:
+            scroll_result = await self.async_client.scroll(
+                collection_name=self.collection,
+                scroll_filter=models.Filter(
+                    must=[models.FieldCondition(key="name", match=models.MatchValue(value=name))]
+                ),
+                limit=1,
+            )
+            return len(scroll_result[0]) > 0
+        return False
+
+    def insert(
+        self,
+        content_hash: str,
+        documents: List[Document],
+        filters: Optional[Dict[str, Any]] = None,
+        batch_size: int = 10,
+    ) -> None:
+        """
+        Insert documents into the database.
+
+        Args:
+            documents (List[Document]): List of documents to insert
+            filters (Optional[Dict[str, Any]]): Filters to apply while inserting documents
+            batch_size (int): Batch size for inserting documents
+        """
+        log_debug(f"Inserting {len(documents)} documents")
+        points = []
+        for document in documents:
+            cleaned_content = document.content.replace("\x00", "\ufffd")
+            doc_id = md5(cleaned_content.encode()).hexdigest()
+
+            # TODO(v2.0.0): Remove conditional vector naming logic
+            if self.use_named_vectors:
+                vector = {self.dense_vector_name: document.embedding}
+            else:
+                vector = document.embedding  # type: ignore
+
+            if self.search_type == SearchType.vector:
+                # For vector search, maintain backward compatibility with unnamed vectors
+                document.embed(embedder=self.embedder)
+                vector = document.embedding  # type: ignore
+            else:
+                # For other search types, use named vectors
+                vector = {}
+                if self.search_type in [SearchType.hybrid]:
+                    document.embed(embedder=self.embedder)
+                    vector[self.dense_vector_name] = document.embedding
+
+                if self.search_type in [SearchType.keyword, SearchType.hybrid]:
+                    vector[self.sparse_vector_name] = next(
+                        iter(self.sparse_encoder.embed([document.content]))
+                    ).as_object()  # type: ignore
+
+            # Create payload with document properties
+            payload = {
+                "name": document.name,
+                "meta_data": document.meta_data,
+                "content": cleaned_content,
+                "usage": document.usage,
+                "content_id": document.content_id,
+                "content_hash": content_hash,
+            }
+
+            # Add filters as metadata if provided
+            if filters:
+                # Merge filters with existing metadata
+                if "meta_data" not in payload:
+                    payload["meta_data"] = {}
+                payload["meta_data"].update(filters)  # type: ignore
+
+            points.append(
+                models.PointStruct(
+                    id=doc_id,
+                    vector=vector,  # type: ignore
+                    payload=payload,
+                )
+            )
+            log_debug(f"Inserted document: {document.name} ({document.meta_data})")
+        if len(points) > 0:
+            self.client.upsert(collection_name=self.collection, wait=False, points=points)
+        log_debug(f"Upsert {len(points)} documents")
+
+    async def async_insert(
+        self, content_hash: str, documents: List[Document], filters: Optional[Dict[str, Any]] = None
+    ) -> None:
+        """
+        Insert documents asynchronously.
+
+        Args:
+            documents (List[Document]): List of documents to insert
+            filters (Optional[Dict[str, Any]]): Filters to apply while inserting documents
+        """
+        log_debug(f"Inserting {len(documents)} documents asynchronously")
+
+        # Apply batch embedding when needed for vector or hybrid search
+        if self.search_type in [SearchType.vector, SearchType.hybrid]:
+            if self.embedder.enable_batch and hasattr(self.embedder, "async_get_embeddings_batch_and_usage"):
+                # Use batch embedding when enabled and supported
+                try:
+                    # Extract content from all documents
+                    doc_contents = [doc.content for doc in documents]
+
+                    # Get batch embeddings and usage
+                    embeddings, usages = await self.embedder.async_get_embeddings_batch_and_usage(doc_contents)
+
+                    # Process documents with pre-computed embeddings
+                    for j, doc in enumerate(documents):
+                        try:
+                            if j < len(embeddings):
+                                doc.embedding = embeddings[j]
+                                doc.usage = usages[j] if j < len(usages) else None
+                        except Exception as e:
+                            log_error(f"Error assigning batch embedding to document '{doc.name}': {e}")
+
+                except Exception as e:
+                    # Check if this is a rate limit error - don't fall back as it would make things worse
+                    error_str = str(e).lower()
+                    is_rate_limit = any(
+                        phrase in error_str
+                        for phrase in ["rate limit", "too many requests", "429", "trial key", "api calls / minute"]
+                    )
+
+                    if is_rate_limit:
+                        log_error(f"Rate limit detected during batch embedding. {e}")
+                        raise e
+                    else:
+                        log_warning(f"Async batch embedding failed, falling back to individual embeddings: {e}")
+                        # Fall back to individual embedding
+                        for doc in documents:
+                            if self.search_type in [SearchType.vector, SearchType.hybrid]:
+                                doc.embed(embedder=self.embedder)
+            else:
+                # Use individual embedding
+                for doc in documents:
+                    if self.search_type in [SearchType.vector, SearchType.hybrid]:
+                        doc.embed(embedder=self.embedder)
+
+        async def process_document(document):
+            cleaned_content = document.content.replace("\x00", "\ufffd")
+            doc_id = md5(cleaned_content.encode()).hexdigest()
+
+            if self.search_type == SearchType.vector:
+                # For vector search, maintain backward compatibility with unnamed vectors
+                vector = document.embedding  # Already embedded above
+            else:
+                # For other search types, use named vectors
+                vector = {}
+                if self.search_type in [SearchType.hybrid]:
+                    vector[self.dense_vector_name] = document.embedding  # Already embedded above
+
+                if self.search_type in [SearchType.keyword, SearchType.hybrid]:
+                    vector[self.sparse_vector_name] = next(
+                        iter(self.sparse_encoder.embed([document.content]))
+                    ).as_object()  # type: ignore
+
+            if self.search_type in [SearchType.keyword, SearchType.hybrid]:
+                vector[self.sparse_vector_name] = next(iter(self.sparse_encoder.embed([document.content]))).as_object()
+
+            # Create payload with document properties
+            payload = {
+                "name": document.name,
+                "meta_data": document.meta_data,
+                "content": cleaned_content,
+                "usage": document.usage,
+                "content_id": document.content_id,
+                "content_hash": content_hash,
+            }
+
+            # Add filters as metadata if provided
+            if filters:
+                # Merge filters with existing metadata
+                if "meta_data" not in payload:
+                    payload["meta_data"] = {}
+                payload["meta_data"].update(filters)
+
+            log_debug(f"Inserted document asynchronously: {document.name} ({document.meta_data})")
+            return models.PointStruct(  # type: ignore
+                id=doc_id,
+                vector=vector,  # type: ignore
+                payload=payload,
+            )
+
+        import asyncio
+
+        # Process all documents in parallel
+        points = await asyncio.gather(*[process_document(doc) for doc in documents])
+
+        if len(points) > 0:
+            await self.async_client.upsert(collection_name=self.collection, wait=False, points=points)
+        log_debug(f"Upserted {len(points)} documents asynchronously")
+
+    def upsert(self, content_hash: str, documents: List[Document], filters: Optional[Dict[str, Any]] = None) -> None:
+        """
+        Upsert documents into the database.
+
+        Args:
+            documents (List[Document]): List of documents to upsert
+            filters (Optional[Dict[str, Any]]): Filters to apply while upserting
+        """
+        log_debug("Redirecting the request to insert")
+        if self.content_hash_exists(content_hash):
+            self._delete_by_content_hash(content_hash)
+        self.insert(content_hash=content_hash, documents=documents, filters=filters)
+
+    async def async_upsert(
+        self, content_hash: str, documents: List[Document], filters: Optional[Dict[str, Any]] = None
+    ) -> None:
+        """Upsert documents asynchronously."""
+        log_debug("Redirecting the async request to async_insert")
+        await self.async_insert(content_hash=content_hash, documents=documents, filters=filters)
+
+    def search(
+        self, query: str, limit: int = 5, filters: Optional[Union[Dict[str, Any], List[FilterExpr]]] = None
+    ) -> List[Document]:
+        """
+        Search for documents in the collection.
+
+        Args:
+            query (str): Query to search for
+            limit (int): Number of search results to return
+            filters (Optional[Dict[str, Any]]): Filters to apply while searching
+        """
+
+        if isinstance(filters, List):
+            log_warning("Filters Expressions are not supported in Qdrant. No filters will be applied.")
+            filters = None
+
+        filters = self._format_filters(filters or {})  # type: ignore
+        if self.search_type == SearchType.vector:
+            results = self._run_vector_search_sync(query, limit, filters)  # type: ignore
+        elif self.search_type == SearchType.keyword:
+            results = self._run_keyword_search_sync(query, limit, filters)  # type: ignore
+        elif self.search_type == SearchType.hybrid:
+            results = self._run_hybrid_search_sync(query, limit, filters)  # type: ignore
+        else:
+            raise ValueError(f"Unsupported search type: {self.search_type}")
+
+        return self._build_search_results(results, query)
+
+    async def async_search(
+        self, query: str, limit: int = 5, filters: Optional[Union[Dict[str, Any], List[FilterExpr]]] = None
+    ) -> List[Document]:
+        if isinstance(filters, List):
+            log_warning("Filters Expressions are not supported in Qdrant. No filters will be applied.")
+            filters = None
+
+        filters = self._format_filters(filters or {})  # type: ignore
+        if self.search_type == SearchType.vector:
+            results = await self._run_vector_search_async(query, limit, filters)  # type: ignore
+        elif self.search_type == SearchType.keyword:
+            results = await self._run_keyword_search_async(query, limit, filters)  # type: ignore
+        elif self.search_type == SearchType.hybrid:
+            results = await self._run_hybrid_search_async(query, limit, filters)  # type: ignore
+        else:
+            raise ValueError(f"Unsupported search type: {self.search_type}")
+
+        return self._build_search_results(results, query)
+
+    def _run_hybrid_search_sync(
+        self,
+        query: str,
+        limit: int,
+        filters: Optional[Union[Dict[str, Any], List[FilterExpr]]],
+    ) -> List[models.ScoredPoint]:
+        dense_embedding = self.embedder.get_embedding(query)
+        sparse_embedding = next(iter(self.sparse_encoder.embed([query]))).as_object()
+        call = self.client.query_points(
+            collection_name=self.collection,
+            prefetch=[
+                models.Prefetch(
+                    query=models.SparseVector(**sparse_embedding),  # type: ignore  # type: ignore
+                    limit=limit,
+                    using=self.sparse_vector_name,
+                ),
+                models.Prefetch(query=dense_embedding, limit=limit, using=self.dense_vector_name),
+            ],
+            query=models.FusionQuery(fusion=self.hybrid_fusion_strategy),
+            with_vectors=True,
+            with_payload=True,
+            limit=limit,
+            query_filter=filters,
+        )
+        return call.points
+
+    def _run_vector_search_sync(
+        self,
+        query: str,
+        limit: int,
+        filters: Optional[Union[Dict[str, Any], List[FilterExpr]]],
+    ) -> List[models.ScoredPoint]:
+        dense_embedding = self.embedder.get_embedding(query)
+
+        # TODO(v2.0.0): Remove this conditional and always use named vectors
+        if self.use_named_vectors:
+            call = self.client.query_points(
+                collection_name=self.collection,
+                query=dense_embedding,
+                with_vectors=True,
+                with_payload=True,
+                limit=limit,
+                query_filter=filters,
+                using=self.dense_vector_name,
+            )
+        else:
+            # Backward compatibility mode - use unnamed vector
+            call = self.client.query_points(
+                collection_name=self.collection,
+                query=dense_embedding,
+                with_vectors=True,
+                with_payload=True,
+                limit=limit,
+                query_filter=filters,
+            )
+        return call.points
+
+    def _run_keyword_search_sync(
+        self,
+        query: str,
+        limit: int,
+        filters: Optional[Union[Dict[str, Any], List[FilterExpr]]],
+    ) -> List[models.ScoredPoint]:
+        sparse_embedding = next(iter(self.sparse_encoder.embed([query]))).as_object()
+        call = self.client.query_points(
+            collection_name=self.collection,
+            query=models.SparseVector(**sparse_embedding),  # type: ignore
+            with_vectors=True,
+            with_payload=True,
+            limit=limit,
+            using=self.sparse_vector_name,
+            query_filter=filters,
+        )
+        return call.points
+
+    async def _run_vector_search_async(
+        self,
+        query: str,
+        limit: int,
+        filters: Optional[Dict[str, Any]],
+    ) -> List[models.ScoredPoint]:
+        dense_embedding = self.embedder.get_embedding(query)
+
+        # TODO(v2.0.0): Remove this conditional and always use named vectors
+        if self.use_named_vectors:
+            call = await self.async_client.query_points(
+                collection_name=self.collection,
+                query=dense_embedding,
+                with_vectors=True,
+                with_payload=True,
+                limit=limit,
+                query_filter=filters,
+                using=self.dense_vector_name,
+            )
+        else:
+            # Backward compatibility mode - use unnamed vector
+            call = await self.async_client.query_points(
+                collection_name=self.collection,
+                query=dense_embedding,
+                with_vectors=True,
+                with_payload=True,
+                limit=limit,
+                query_filter=filters,
+            )
+        return call.points
+
+    async def _run_keyword_search_async(
+        self,
+        query: str,
+        limit: int,
+        filters: Optional[Dict[str, Any]],
+    ) -> List[models.ScoredPoint]:
+        sparse_embedding = next(iter(self.sparse_encoder.embed([query]))).as_object()
+        call = await self.async_client.query_points(
+            collection_name=self.collection,
+            query=models.SparseVector(**sparse_embedding),  # type: ignore
+            with_vectors=True,
+            with_payload=True,
+            limit=limit,
+            using=self.sparse_vector_name,
+            query_filter=filters,
+        )
+        return call.points
+
+    async def _run_hybrid_search_async(
+        self,
+        query: str,
+        limit: int,
+        filters: Optional[Union[Dict[str, Any], List[FilterExpr]]],
+    ) -> List[models.ScoredPoint]:
+        dense_embedding = self.embedder.get_embedding(query)
+        sparse_embedding = next(iter(self.sparse_encoder.embed([query]))).as_object()
+        call = await self.async_client.query_points(
+            collection_name=self.collection,
+            prefetch=[
+                models.Prefetch(
+                    query=models.SparseVector(**sparse_embedding),  # type: ignore  # type: ignore
+                    limit=limit,
+                    using=self.sparse_vector_name,
+                ),
+                models.Prefetch(query=dense_embedding, limit=limit, using=self.dense_vector_name),
+            ],
+            query=models.FusionQuery(fusion=self.hybrid_fusion_strategy),
+            with_vectors=True,
+            with_payload=True,
+            limit=limit,
+            query_filter=filters,
+        )
+        return call.points
+
+    def _build_search_results(self, results, query: str) -> List[Document]:
+        search_results: List[Document] = []
+
+        for result in results:
+            if result.payload is None:
+                continue
+            search_results.append(
+                Document(
+                    name=result.payload["name"],
+                    meta_data=result.payload["meta_data"],
+                    content=result.payload["content"],
+                    embedder=self.embedder,
+                    embedding=result.vector,  # type: ignore
+                    usage=result.payload.get("usage"),
+                    content_id=result.payload.get("content_id"),
+                )
+            )
+
+        if self.reranker:
+            search_results = self.reranker.rerank(query=query, documents=search_results)
+
+        log_info(f"Found {len(search_results)} documents")
+        return search_results
+
+    def _format_filters(self, filters: Optional[Dict[str, Any]]) -> Optional[models.Filter]:
+        if filters:
+            filter_conditions = []
+            for key, value in filters.items():
+                # If key contains a dot already, assume it's in the correct format
+                # Otherwise, assume it's a metadata field and add the prefix
+                if "." not in key and not key.startswith("meta_data."):
+                    # This is a simple field name, assume it's metadata
+                    key = f"meta_data.{key}"
+
+                if isinstance(value, dict):
+                    # Handle nested dictionaries
+                    for sub_key, sub_value in value.items():
+                        filter_conditions.append(
+                            models.FieldCondition(key=f"{key}.{sub_key}", match=models.MatchValue(value=sub_value))
+                        )
+                else:
+                    # Handle direct key-value pairs
+                    filter_conditions.append(models.FieldCondition(key=key, match=models.MatchValue(value=value)))
+
+            if filter_conditions:
+                return models.Filter(must=filter_conditions)  # type: ignore
+
+        return None
+
+    def optimize(self) -> None:
+        pass
+
+    def drop(self) -> None:
+        if self.exists():
+            log_debug(f"Deleting collection: {self.collection}")
+            self.client.delete_collection(self.collection)
+
+    async def async_drop(self) -> None:
+        """Drop the collection asynchronously."""
+        if await self.async_exists():
+            log_debug(f"Deleting collection asynchronously: {self.collection}")
+            await self.async_client.delete_collection(self.collection)
+
+    def exists(self) -> bool:
+        """Check if the collection exists."""
+        return self.client.collection_exists(collection_name=self.collection)
+
+    async def async_exists(self) -> bool:
+        """Check if the collection exists asynchronously."""
+        return await self.async_client.collection_exists(collection_name=self.collection)
+
+    def get_count(self) -> int:
+        count_result: models.CountResult = self.client.count(collection_name=self.collection, exact=True)
+        return count_result.count
+
+    def point_exists(self, id: str) -> bool:
+        """Check if a point with the given ID exists in the collection."""
+        try:
+            log_info(f"Checking if point with ID '{id}' (type: {type(id)}) exists in collection '{self.collection}'")
+            points = self.client.retrieve(
+                collection_name=self.collection, ids=[id], with_payload=False, with_vectors=False
+            )
+            log_info(f"Retrieved {len(points)} points for ID '{id}'")
+            if len(points) > 0:
+                log_info(f"Found point with ID: {points[0].id} (type: {type(points[0].id)})")
+            return len(points) > 0
+        except Exception as e:
+            log_info(f"Error checking if point {id} exists: {e}")
+            return False
+
+    def delete(self) -> bool:
+        return self.client.delete_collection(collection_name=self.collection)
+
+    def delete_by_id(self, id: str) -> bool:
+        try:
+            # Check if point exists before deletion
+            if not self.point_exists(id):
+                log_warning(f"Point with ID {id} does not exist")
+                return True
+
+            self.client.delete(
+                collection_name=self.collection,
+                points_selector=models.PointIdsList(points=[id]),
+                wait=True,  # Wait for the operation to complete
+            )
+            return True
+
+        except Exception as e:
+            log_info(f"Error deleting point with ID {id}: {e}")
+            return False
+
+    def delete_by_name(self, name: str) -> bool:
+        """Delete all points that have the specified name in their payload (precise match)."""
+        try:
+            log_info(f"Attempting to delete all points with name: {name}")
+
+            # Create a filter to find all points with the specified name (precise match)
+            filter_condition = models.Filter(
+                must=[models.FieldCondition(key="name", match=models.MatchValue(value=name))]
+            )
+
+            # First, count how many points will be deleted
+            count_result = self.client.count(collection_name=self.collection, count_filter=filter_condition, exact=True)
+
+            if count_result.count == 0:
+                log_warning(f"No points found with name: {name}")
+                return True
+
+            log_info(f"Found {count_result.count} points to delete with name: {name}")
+
+            # Delete all points matching the filter
+            result = self.client.delete(
+                collection_name=self.collection,
+                points_selector=filter_condition,
+                wait=True,  # Wait for the operation to complete
+            )
+
+            # Check if the deletion was successful
+            if result.status == models.UpdateStatus.COMPLETED:
+                log_info(f"Successfully deleted {count_result.count} points with name: {name}")
+                return True
+            else:
+                log_warning(f"Deletion failed for name {name}. Status: {result.status}")
+                return False
+
+        except Exception as e:
+            log_warning(f"Error deleting points with name {name}: {e}")
+            return False
+
+    def delete_by_metadata(self, metadata: Dict[str, Any]) -> bool:
+        """Delete all points where the given metadata is contained in the meta_data payload field."""
+        try:
+            log_info(f"Attempting to delete all points with metadata: {metadata}")
+
+            # Create filter conditions for each metadata key-value pair
+            filter_conditions = []
+            for key, value in metadata.items():
+                # Use the meta_data prefix since that's how metadata is stored in the payload
+                filter_conditions.append(
+                    models.FieldCondition(key=f"meta_data.{key}", match=models.MatchValue(value=value))
+                )
+
+            # Create a filter that requires ALL metadata conditions to match
+            filter_condition = models.Filter(must=filter_conditions)  # type: ignore
+
+            # First, count how many points will be deleted
+            count_result = self.client.count(collection_name=self.collection, count_filter=filter_condition, exact=True)
+
+            if count_result.count == 0:
+                log_warning(f"No points found with metadata: {metadata}")
+                return True
+
+            log_info(f"Found {count_result.count} points to delete with metadata: {metadata}")
+
+            # Delete all points matching the filter
+            result = self.client.delete(
+                collection_name=self.collection,
+                points_selector=filter_condition,
+                wait=True,  # Wait for the operation to complete
+            )
+
+            # Check if the deletion was successful
+            if result.status == models.UpdateStatus.COMPLETED:
+                log_info(f"Successfully deleted {count_result.count} points with metadata: {metadata}")
+                return True
+            else:
+                log_warning(f"Deletion failed for metadata {metadata}. Status: {result.status}")
+                return False
+
+        except Exception as e:
+            log_warning(f"Error deleting points with metadata {metadata}: {e}")
+            return False
+
+    def delete_by_content_id(self, content_id: str) -> bool:
+        """Delete all points that have the specified content_id in their payload."""
+        try:
+            log_info(f"Attempting to delete all points with content_id: {content_id}")
+
+            # Create a filter to find all points with the specified content_id
+            filter_condition = models.Filter(
+                must=[models.FieldCondition(key="content_id", match=models.MatchValue(value=content_id))]
+            )
+
+            # First, count how many points will be deleted
+            count_result = self.client.count(collection_name=self.collection, count_filter=filter_condition, exact=True)
+
+            if count_result.count == 0:
+                log_warning(f"No points found with content_id: {content_id}")
+                return True
+
+            log_info(f"Found {count_result.count} points to delete with content_id: {content_id}")
+
+            # Delete all points matching the filter
+            result = self.client.delete(
+                collection_name=self.collection,
+                points_selector=filter_condition,
+                wait=True,  # Wait for the operation to complete
+            )
+
+            # Check if the deletion was successful
+            if result.status == models.UpdateStatus.COMPLETED:
+                log_info(f"Successfully deleted {count_result.count} points with content_id: {content_id}")
+                return True
+            else:
+                log_warning(f"Deletion failed for content_id {content_id}. Status: {result.status}")
+                return False
+
+        except Exception as e:
+            log_warning(f"Error deleting points with content_id {content_id}: {e}")
+            return False
+
+    def id_exists(self, id: str) -> bool:
+        """Check if a point with the given ID exists in the collection.
+
+        Args:
+            id (str): The ID to check.
+
+        Returns:
+            bool: True if the point exists, False otherwise.
+        """
+        try:
+            points = self.client.retrieve(
+                collection_name=self.collection, ids=[id], with_payload=False, with_vectors=False
+            )
+            return len(points) > 0
+        except Exception as e:
+            log_info(f"Error checking if point {id} exists: {e}")
+            return False
+
+    def content_hash_exists(self, content_hash: str) -> bool:
+        """Check if any points with the given content hash exist in the collection.
+
+        Args:
+            content_hash (str): The content hash to check.
+
+        Returns:
+            bool: True if points with the content hash exist, False otherwise.
+        """
+        try:
+            # Create a filter to find points with the specified content_hash
+            filter_condition = models.Filter(
+                must=[models.FieldCondition(key="content_hash", match=models.MatchValue(value=content_hash))]
+            )
+
+            # Count how many points match the filter
+            count_result = self.client.count(collection_name=self.collection, count_filter=filter_condition, exact=True)
+            return count_result.count > 0
+        except Exception as e:
+            log_info(f"Error checking if content_hash {content_hash} exists: {e}")
+            return False
+
+    def _delete_by_content_hash(self, content_hash: str) -> bool:
+        """Delete all points that have the specified content_hash in their payload.
+
+        Args:
+            content_hash (str): The content hash to delete.
+
+        Returns:
+            bool: True if points were deleted successfully, False otherwise.
+        """
+        try:
+            log_info(f"Attempting to delete all points with content_hash: {content_hash}")
+
+            # Create a filter to find all points with the specified content_hash
+            filter_condition = models.Filter(
+                must=[models.FieldCondition(key="content_hash", match=models.MatchValue(value=content_hash))]
+            )
+
+            # First, count how many points will be deleted
+            count_result = self.client.count(collection_name=self.collection, count_filter=filter_condition, exact=True)
+
+            if count_result.count == 0:
+                log_warning(f"No points found with content_hash: {content_hash}")
+                return True
+
+            log_info(f"Found {count_result.count} points to delete with content_hash: {content_hash}")
+
+            # Delete all points matching the filter
+            result = self.client.delete(
+                collection_name=self.collection,
+                points_selector=filter_condition,
+                wait=True,  # Wait for the operation to complete
+            )
+
+            # Check if the deletion was successful
+            if result.status == models.UpdateStatus.COMPLETED:
+                log_info(f"Successfully deleted {count_result.count} points with content_hash: {content_hash}")
+                return True
+            else:
+                log_warning(f"Deletion failed for content_hash {content_hash}. Status: {result.status}")
+                return False
+
+        except Exception as e:
+            log_warning(f"Error deleting points with content_hash {content_hash}: {e}")
+            return False
+
+    def update_metadata(self, content_id: str, metadata: Dict[str, Any]) -> None:
+        """
+        Update the metadata for documents with the given content_id.
+
+        Args:
+            content_id (str): The content ID to update
+            metadata (Dict[str, Any]): The metadata to update
+        """
+        try:
+            if not self.client:
+                log_error("Client not initialized")
+                return
+
+            # Create filter for content_id
+            filter_condition = models.Filter(
+                must=[models.FieldCondition(key="content_id", match=models.MatchValue(value=content_id))]
+            )
+
+            # Search for points with the given content_id
+            search_result = self.client.scroll(
+                collection_name=self.collection,
+                scroll_filter=filter_condition,
+                limit=10000,  # Get all matching points
+                with_payload=True,
+                with_vectors=False,
+            )
+
+            if not search_result[0]:  # search_result is a tuple (points, next_page_offset)
+                log_error(f"No documents found with content_id: {content_id}")
+                return
+
+            points = search_result[0]
+            update_operations = []
+
+            # Prepare update operations for each point
+            for point in points:
+                point_id = point.id
+                current_payload = point.payload or {}
+
+                # Merge existing metadata with new metadata
+                updated_payload = current_payload.copy()
+                updated_payload.update(metadata)
+
+                if "filters" not in updated_payload:
+                    updated_payload["filters"] = {}
+                if isinstance(updated_payload["filters"], dict):
+                    updated_payload["filters"].update(metadata)
+                else:
+                    updated_payload["filters"] = metadata
+
+                # Create set payload operation
+                update_operations.append(models.SetPayload(payload=updated_payload, points=[point_id]))
+
+            # Execute all updates
+            for operation in update_operations:
+                self.client.set_payload(
+                    collection_name=self.collection, payload=operation.payload, points=operation.points
+                )
+
+            log_debug(f"Updated metadata for {len(update_operations)} documents with content_id: {content_id}")
+
+        except Exception as e:
+            log_error(f"Error updating metadata for content_id '{content_id}': {e}")
+            raise
+
+    def close(self) -> None:
+        """Close the Qdrant client connections."""
+        if self._client is not None:
+            try:
+                self._client.close()
+                log_debug("Qdrant client closed successfully")
+            except Exception as e:
+                log_debug(f"Error closing Qdrant client: {e}")
+            finally:
+                self._client = None
+
+    async def async_close(self) -> None:
+        """Close the Qdrant client connections asynchronously."""
+        if self._async_client is not None:
+            try:
+                await self._async_client.close()
+                log_debug("Async Qdrant client closed successfully")
+            except Exception as e:
+                log_debug(f"Error closing async Qdrant client: {e}")
+            finally:
+                self._async_client = None
+
+    def get_supported_search_types(self) -> List[str]:
+        """Get the supported search types for this vector database."""
+        return [SearchType.vector, SearchType.keyword, SearchType.hybrid]