Feature: Qdrant filters supports (#5446)

# Support Qdrant filters

Qdrant has an [extensive filtering
system](https://qdrant.tech/documentation/concepts/filtering/) with rich
type support. This PR makes it possible to use the filters in Langchain
by passing an additional param to both the
`similarity_search_with_score` and `similarity_search` methods.

## Who can review?

@dev2049 @hwchase17

---------

Co-authored-by: Dev 2049 <dev.dev2049@gmail.com>

docs/modules/indexes/vectorstores/examples/qdrant.ipynb

@@ -399,6 +399,31 @@
     "print(f\"\\nScore: {score}\")"
    ]
   },
+  {
+   "cell_type": "markdown",
+   "source": [
+    "### Metadata filtering\n",
+    "\n",
+    "Qdrant has an [extensive filtering system](https://qdrant.tech/documentation/concepts/filtering/) with rich type support. It is also possible to use the filters in Langchain, by passing an additional param to both the `similarity_search_with_score` and `similarity_search` methods."
+   ],
+   "metadata": {
+    "collapsed": false
+   }
+  },
+  {
+   "cell_type": "markdown",
+   "source": [
+    "```python\n",
+    "from qdrant_client.http import models as rest\n",
+    "\n",
+    "query = \"What did the president say about Ketanji Brown Jackson\"\n",
+    "found_docs = qdrant.similarity_search_with_score(query, filter=rest.Filter(...))\n",
+    "```"
+   ],
+   "metadata": {
+    "collapsed": false
+   }
+  },
   {
    "cell_type": "markdown",
    "id": "c58c30bf",

langchain/vectorstores/qdrant.py

@@ -27,10 +27,11 @@ from langchain.vectorstores import VectorStore
 from langchain.vectorstores.utils import maximal_marginal_relevance
 
 if TYPE_CHECKING:
+    from qdrant_client.conversions import common_types
     from qdrant_client.http import models as rest
 
-
+    DictFilter = Dict[str, Union[str, int, bool, dict, list]]
-MetadataFilter = Dict[str, Union[str, int, bool, dict, list]]
+    MetadataFilter = Union[DictFilter, common_types.Filter]
 
 
 class Qdrant(VectorStore):
@@ -234,10 +235,21 @@ class Qdrant(VectorStore):
             List of Documents most similar to the query and score for each.
         """
 
+        if filter is not None and isinstance(filter, dict):
+            warnings.warn(
+                "Using dict as a `filter` is deprecated. Please use qdrant-client "
+                "filters directly: "
+                "https://qdrant.tech/documentation/concepts/filtering/",
+                DeprecationWarning,
+            )
+            qdrant_filter = self._qdrant_filter_from_dict(filter)
+        else:
+            qdrant_filter = filter
+
         results = self.client.search(
             collection_name=self.collection_name,
             query_vector=self._embed_query(query),
-            query_filter=self._qdrant_filter_from_dict(filter),
+            query_filter=qdrant_filter,
             with_payload=True,
             limit=k,
         )
@@ -519,7 +531,7 @@ class Qdrant(VectorStore):
         return out
 
     def _qdrant_filter_from_dict(
-        self, filter: Optional[MetadataFilter]
+        self, filter: Optional[DictFilter]
     ) -> Optional[rest.Filter]:
         from qdrant_client.http import models as rest
 

pyproject.toml

@@ -306,7 +306,7 @@ extended_testing = [
  "html2text",
  "py-trello",
  "scikit-learn",
- "pyspark",
+ "pyspark"
 ]
 
 [tool.ruff]

tests/integration_tests/vectorstores/test_qdrant.py

@@ -2,6 +2,7 @@
 from typing import Callable, Optional
 
 import pytest
+from qdrant_client.http import models as rest
 
 from langchain.docstore.document import Document
 from langchain.embeddings.base import Embeddings
@@ -129,6 +130,45 @@ def test_qdrant_similarity_search_filters(batch_size: int) -> None:
     ]
 
 
+def test_qdrant_similarity_search_filters_with_qdrant_filters() -> None:
+    """Test end to end construction and search."""
+    texts = ["foo", "bar", "baz"]
+    metadatas = [
+        {"page": i, "details": {"page": i + 1, "pages": [i + 2, -1]}}
+        for i in range(len(texts))
+    ]
+    docsearch = Qdrant.from_texts(
+        texts,
+        ConsistentFakeEmbeddings(),
+        metadatas=metadatas,
+        location=":memory:",
+    )
+
+    qdrant_filter = rest.Filter(
+        must=[
+            rest.FieldCondition(
+                key="metadata.page",
+                match=rest.MatchValue(value=1),
+            ),
+            rest.FieldCondition(
+                key="metadata.details.page",
+                match=rest.MatchValue(value=2),
+            ),
+            rest.FieldCondition(
+                key="metadata.details.pages",
+                match=rest.MatchAny(any=[3]),
+            ),
+        ]
+    )
+    output = docsearch.similarity_search("foo", k=1, filter=qdrant_filter)
+    assert output == [
+        Document(
+            page_content="bar",
+            metadata={"page": 1, "details": {"page": 2, "pages": [3, -1]}},
+        )
+    ]
+
+
 @pytest.mark.parametrize("batch_size", [1, 64])
 @pytest.mark.parametrize(
     ["content_payload_key", "metadata_payload_key"],