fix: handle non-dict intermediate value in nested metadata filters (#11649)

Author: i-anubhav-anand

haystack/utils/filters.py

@@ -195,8 +195,9 @@ def _comparison_condition(condition: dict[str, Any], document: Document | ByteSt
         parts = field.split(".")
         document_value = getattr(document, parts[0])
         for part in parts[1:]:
-            if part not in document_value:
-                # If a field is not found we treat it as None
+            if not isinstance(document_value, dict) or part not in document_value:
+                # If a field is not found (or an intermediate value is not a dict,
+                # e.g. None) we treat it as None
                 document_value = None
                 break
             document_value = document_value[part]

releasenotes/notes/fix-nested-meta-filter-non-dict-intermediate-a5e768b6da3bcb37.yaml

@@ -0,0 +1,8 @@
+---
+fixes:
+  - |
+    Fixed a ``TypeError`` when filtering documents on a nested metadata field
+    (e.g. ``meta.user.age``) where an intermediate value is not a dictionary
+    (for example ``None`` or a scalar). Such intermediate values are now treated
+    as a missing field (``None``) instead of raising, which previously crashed
+    ``filter_documents`` for stores like ``InMemoryDocumentStore``.

test/utils/test_filters.py

@@ -46,6 +46,18 @@
         False,
         id="== operator with None filter value",
     ),
+    pytest.param(
+        {"field": "meta.user.age", "operator": "==", "value": 30},
+        Document(meta={"user": {"age": 30}}),
+        True,
+        id="== operator with nested meta field",
+    ),
+    pytest.param(
+        {"field": "meta.user.age", "operator": "==", "value": 30},
+        Document(meta={"user": None}),
+        False,
+        id="== operator with nested field on non-dict intermediate value",
+    ),
     # != operator params
     pytest.param(
         {"field": "meta.name", "operator": "!=", "value": "test"},