Python: fix(python): Fix InMemoryCollection upsert with key storage_name alias

python/semantic_kernel/connectors/in_memory.py

@@ -655,7 +655,31 @@ async def _inner_upsert(self, records: Sequence[Any], **kwargs: Any) -> Sequence
         return updated_keys
 
     def _deserialize_store_models_to_dicts(self, records: Sequence[Any], **kwargs: Any) -> Sequence[dict[str, Any]]:
-        return records
+        if not (isinstance(self.record_type, type) and issubclass(self.record_type, BaseModel)):
+            return records
+        # Pydantic records are stored under their Python field names (they are
+        # serialized with model_dump()), but the shared deserializer for
+        # Pydantic models expects store dicts keyed by storage name and maps
+        # them back to field names. Rename name -> storage name in one pass
+        # so a data field whose name overlaps the key's storage name (or vice
+        # versa) keeps its own value instead of overwriting the key's.
+        converted_records: list[dict[str, Any]] = []
+        for record in records:
+            renames = {
+                field.storage_name: record[field.name]
+                for field in self.definition.fields
+                if field.storage_name and field.storage_name != field.name and field.name in record
+            }
+            if not renames:
+                converted_records.append(record)
+                continue
+            # Drop every renamed field's original name, then apply all
+            # renames at once so overlapping names cannot overwrite each other.
+            renamed_names = {field.name for field in self.definition.fields if field.storage_name in renames}
+            converted = {key: value for key, value in record.items() if key not in renamed_names}
+            converted.update(renames)
+            converted_records.append(converted)
+        return converted_records
 
     def _serialize_data_model_to_dict(self, record: TModel, **kwargs: Any) -> OneOrList[dict[str, Any]]:
         """Keep track of key names before records lose their model type."""

python/tests/unit/connectors/memory/test_in_memory.py

@@ -2,12 +2,12 @@
 
 import ast
 
-from pydantic import ConfigDict, Field, model_serializer
+from pydantic import BaseModel, ConfigDict, Field, model_serializer
 from pytest import fixture, mark, raises
 
 from semantic_kernel.connectors.in_memory import InMemoryCollection, InMemoryStore
 from semantic_kernel.data._shared import default_dynamic_filter_function
-from semantic_kernel.data.vector import DistanceFunction
+from semantic_kernel.data.vector import DistanceFunction, VectorStoreCollectionDefinition, VectorStoreField
 from semantic_kernel.exceptions.vector_store_exceptions import VectorStoreOperationException
 
 
@@ -518,3 +518,28 @@ async def test_large_sequence_repeat_filter_is_blocked(collection):
 
     with raises(VectorStoreOperationException, match="Sequence repetition in filter expressions exceeds the maximum"):
         collection._get_filtered_records(type("opt", (), {"filter": "lambda x: [0] * 2000000000"})())
+
+
+async def test_upsert_pydantic_key_storage_name_overlapping_data_field():
+    # The key field's storage name ("stored_id") is also the name of a data
+    # field. Pydantic records serialize with logical names, so the key must
+    # come from "id", not from the data field named "stored_id".
+    definition = VectorStoreCollectionDefinition(
+        fields=[
+            VectorStoreField("key", name="id", type="str", storage_name="stored_id"),
+            VectorStoreField("data", name="stored_id", type="str", storage_name="payload"),
+            VectorStoreField("vector", name="vector", type="float", dimensions=2),
+        ]
+    )
+
+    class OverlappingRecord(BaseModel):
+        id: str
+        stored_id: str
+        vector: list[float]
+
+    collection = InMemoryCollection(collection_name="test", record_type=OverlappingRecord, definition=definition)
+    key = await collection.upsert(OverlappingRecord(id="actual-key", stored_id="payload-value", vector=[1.0, 0.0]))
+    assert key == "actual-key"
+    result = await collection.get("actual-key", include_vectors=True)
+    assert result.id == "actual-key"
+    assert result.stored_id == "payload-value"