Implement improvements on the QdrantVectorSearchTool

- Allow search filters to be set at the constructor level
- Fix issue that prevented multiple records from being returned

Author: danielfsbarreto

crewai_tools/tools/qdrant_vector_search_tool/qdrant_search_tool.py

@@ -1,11 +1,10 @@
 import json
 import os
-from typing import Any, Callable, Optional, Type, List
-
+from typing import Any, Callable, List, Optional, Type
 
 try:
     from qdrant_client import QdrantClient
-    from qdrant_client.http.models import Filter, FieldCondition, MatchValue
+    from qdrant_client.http.models import FieldCondition, Filter, MatchValue
 
     QDRANT_AVAILABLE = True
 except ImportError:
@@ -57,8 +56,8 @@ class QdrantVectorSearchTool(BaseTool):
     description: str = "A tool to search the Qdrant database for relevant information on internal documents."
     args_schema: Type[BaseModel] = QdrantToolSchema
     query: Optional[str] = None
-    filter_by: Optional[str] = None
-    filter_value: Optional[str] = None
+    filter_conditions: Optional[list[tuple[str, Any]]] = []
+    search_filter: Optional[Filter] = None
     collection_name: Optional[str] = None
     limit: Optional[int] = Field(default=3)
     score_threshold: float = Field(default=0.35)
@@ -101,6 +100,18 @@ def __init__(self, **kwargs):
                     "The 'qdrant-client' package is required to use the QdrantVectorSearchTool. "
                     "Please install it with: uv add qdrant-client"
                 )
+        if kwargs.get("filter_conditions"):
+            must_conditions = []
+            for filter_condition in kwargs.get("filter_conditions"):
+                must_conditions.append(
+                    FieldCondition(
+                        key=filter_condition[0],
+                        match=MatchValue(value=filter_condition[1]),
+                    )
+                )
+            self.search_filter = Filter(must=must_conditions)
+        else:
+            self.search_filter = None
 
     def _run(
         self,
@@ -126,14 +137,26 @@ def _run(
         if not self.qdrant_url:
             raise ValueError("QDRANT_URL is not set")
 
-        # Create filter if filter parameters are provided
-        search_filter = None
+        # If filter_by and filter_value are provided, add them to the search filter
+        # without modifying the original search filter
+        search_filter = self.search_filter.copy() if self.search_filter else None
         if filter_by and filter_value:
-            search_filter = Filter(
-                must=[
+            if (
+                search_filter
+                and hasattr(search_filter, "must")
+                and isinstance(search_filter.must, list)
+            ):
+                search_filter.must.append(
                     FieldCondition(key=filter_by, match=MatchValue(value=filter_value))
-                ]
-            )
+                )
+            else:
+                search_filter = Filter(
+                    must=[
+                        FieldCondition(
+                            key=filter_by, match=MatchValue(value=filter_value)
+                        )
+                    ]
+                )
 
         # Search in Qdrant using the built-in query method
         query_vector = (
@@ -151,12 +174,11 @@ def _run(
 
         # Format results similar to storage implementation
         results = []
-        # Extract the list of ScoredPoint objects from the tuple
-        for point in search_results:
+        for point in search_results.points:
             result = {
-                "metadata": point[1][0].payload.get("metadata", {}),
-                "context": point[1][0].payload.get("text", ""),
-                "distance": point[1][0].score,
+                "distance": point.score,
+                "metadata": point.payload.get("metadata", {}),
+                "context": point.payload.get("text", ""),
             }
             results.append(result)