Merge branch 'main' into dev/1.35

Author: dirkkul

docs/changelog.rst

@@ -1,6 +1,12 @@
 Changelog
 =========
 
+Version 4.18.3
+--------------
+This patch version includes:
+    - Fixes wrong deprecation warning in PyCharm when creating a collection without vectorizers
+    - Fixes warnings with boolean filters
+
 Version 4.18.2
 --------------
 This patch version includes:

integration/test_collection_filter.py

@@ -278,13 +278,14 @@ def test_filters_comparison(
     ],
 )
 def test_filters_contains(
+    recwarn: pytest.WarningsRecorder,
     collection_factory: CollectionFactory,
     weaviate_filter: _FilterValue,
     results: List[int],
     require_version: Optional[tuple[int, int, int]],
 ) -> None:
     collection = collection_factory(
-        vectorizer_config=Configure.Vectorizer.none(),
+        vector_config=Configure.Vectors.self_provided(),
         properties=[
             Property(name="text", data_type=DataType.TEXT),
             Property(name="texts", data_type=DataType.TEXT_ARRAY),
@@ -380,6 +381,12 @@ def test_filters_contains(
     uuids = [uuids[result] for result in results]
     assert all(obj.uuid in uuids for obj in objects)
 
+    # Check for warnings to make sure booleans are handled as their correct type and are not sent as ints
+    if len(recwarn) != 0:
+        for rwarning in recwarn.list:
+            print(rwarning.message)
+    assert len(recwarn) == 0
+
 
 @pytest.mark.parametrize(
     "weaviate_filter,results",

weaviate/collections/collections/async_.pyi

@@ -28,9 +28,6 @@ from weaviate.connect.v4 import ConnectionAsync
 
 class _CollectionsAsync(_CollectionsBase[ConnectionAsync]):
     @overload
-    @deprecated(
-        'Using the "vector_index_config" argument is deprecated. Instead, define the vector index for each specific vectorizer supplied to the "vectorizers_config" argument.'
-    )
     async def create(
         self,
         name: str,
@@ -44,7 +41,7 @@ class _CollectionsAsync(_CollectionsBase[ConnectionAsync]):
         replication_config: Optional[_ReplicationConfigCreate] = None,
         reranker_config: Optional[_RerankerProvider] = None,
         sharding_config: Optional[_ShardingConfigCreate] = None,
-        vector_index_config: _VectorIndexConfigCreate,
+        vector_index_config: Optional[_VectorIndexConfigCreate] = None,
         vectorizer_config: Optional[
             Union[_VectorizerConfigCreate, List[_NamedVectorConfigCreate]]
         ] = None,
@@ -55,7 +52,7 @@ class _CollectionsAsync(_CollectionsBase[ConnectionAsync]):
     ) -> CollectionAsync[Properties, References]: ...
     @overload
     @deprecated(
-        'Using the "vectorizer_config" argument is deprecated. Instead, use the "vectorizers_config" argument.'
+        'Using the "vectorizer_config" argument is deprecated. Instead, use the "vector_config" argument.'
     )
     async def create(
         self,
@@ -70,14 +67,19 @@ class _CollectionsAsync(_CollectionsBase[ConnectionAsync]):
         replication_config: Optional[_ReplicationConfigCreate] = None,
         reranker_config: Optional[_RerankerProvider] = None,
         sharding_config: Optional[_ShardingConfigCreate] = None,
-        vector_index_config: Optional[_VectorIndexConfigCreate] = None,
-        vectorizer_config: Union[_VectorizerConfigCreate, List[_NamedVectorConfigCreate]],
+        vector_index_config: _VectorIndexConfigCreate,
+        vectorizer_config: Optional[
+            Union[_VectorizerConfigCreate, List[_NamedVectorConfigCreate]]
+        ] = None,
         vector_config: Optional[Union[_VectorConfigCreate, List[_VectorConfigCreate]]] = None,
         data_model_properties: Optional[Type[Properties]] = None,
         data_model_references: Optional[Type[References]] = None,
         skip_argument_validation: bool = False,
     ) -> CollectionAsync[Properties, References]: ...
     @overload
+    @deprecated(
+        'Using the "vectorizer_config" argument is deprecated. Instead, use the "vector_config" argument.'
+    )
     async def create(
         self,
         name: str,
@@ -92,9 +94,7 @@ class _CollectionsAsync(_CollectionsBase[ConnectionAsync]):
         reranker_config: Optional[_RerankerProvider] = None,
         sharding_config: Optional[_ShardingConfigCreate] = None,
         vector_index_config: Optional[_VectorIndexConfigCreate] = None,
-        vectorizer_config: Optional[
-            Union[_VectorizerConfigCreate, List[_NamedVectorConfigCreate]]
-        ] = None,
+        vectorizer_config: Union[_VectorizerConfigCreate, List[_NamedVectorConfigCreate]],
         vector_config: Optional[Union[_VectorConfigCreate, List[_VectorConfigCreate]]] = None,
         data_model_properties: Optional[Type[Properties]] = None,
         data_model_references: Optional[Type[References]] = None,

weaviate/collections/collections/sync.pyi

@@ -28,9 +28,6 @@ from weaviate.connect.v4 import ConnectionSync
 
 class _Collections(_CollectionsBase[ConnectionSync]):
     @overload
-    @deprecated(
-        'Using the "vector_index_config" argument is deprecated. Instead, define the vector index for each specific vectorizer supplied to the "vector_config" argument.'
-    )
     def create(
         self,
         name: str,
@@ -44,7 +41,7 @@ class _Collections(_CollectionsBase[ConnectionSync]):
         replication_config: Optional[_ReplicationConfigCreate] = None,
         reranker_config: Optional[_RerankerProvider] = None,
         sharding_config: Optional[_ShardingConfigCreate] = None,
-        vector_index_config: _VectorIndexConfigCreate,
+        vector_index_config: Optional[_VectorIndexConfigCreate] = None,
         vectorizer_config: Optional[
             Union[_VectorizerConfigCreate, List[_NamedVectorConfigCreate]]
         ] = None,
@@ -55,7 +52,7 @@ class _Collections(_CollectionsBase[ConnectionSync]):
     ) -> Collection[Properties, References]: ...
     @overload
     @deprecated(
-        'Using the "vectorizer_config" argument is deprecated. Instead, use the "vector_config" argument.'
+        'Using the "vector_index_config" argument is deprecated. Instead, define the vector index for each specific vectorizer supplied to the "vector_config" argument.'
     )
     def create(
         self,
@@ -70,14 +67,19 @@ class _Collections(_CollectionsBase[ConnectionSync]):
         replication_config: Optional[_ReplicationConfigCreate] = None,
         reranker_config: Optional[_RerankerProvider] = None,
         sharding_config: Optional[_ShardingConfigCreate] = None,
-        vector_index_config: Optional[_VectorIndexConfigCreate] = None,
-        vectorizer_config: Union[_VectorizerConfigCreate, List[_NamedVectorConfigCreate]],
+        vector_index_config: _VectorIndexConfigCreate,
+        vectorizer_config: Optional[
+            Union[_VectorizerConfigCreate, List[_NamedVectorConfigCreate]]
+        ] = None,
         vector_config: Optional[Union[_VectorConfigCreate, List[_VectorConfigCreate]]] = None,
         data_model_properties: Optional[Type[Properties]] = None,
         data_model_references: Optional[Type[References]] = None,
         skip_argument_validation: bool = False,
     ) -> Collection[Properties, References]: ...
     @overload
+    @deprecated(
+        'Using the "vectorizer_config" argument is deprecated. Instead, use the "vector_config" argument.'
+    )
     def create(
         self,
         name: str,
@@ -92,9 +94,7 @@ class _Collections(_CollectionsBase[ConnectionSync]):
         reranker_config: Optional[_RerankerProvider] = None,
         sharding_config: Optional[_ShardingConfigCreate] = None,
         vector_index_config: Optional[_VectorIndexConfigCreate] = None,
-        vectorizer_config: Optional[
-            Union[_VectorizerConfigCreate, List[_NamedVectorConfigCreate]]
-        ] = None,
+        vectorizer_config: Union[_VectorizerConfigCreate, List[_NamedVectorConfigCreate]],
         vector_config: Optional[Union[_VectorConfigCreate, List[_VectorConfigCreate]]] = None,
         data_model_properties: Optional[Type[Properties]] = None,
         data_model_references: Optional[Type[References]] = None,

weaviate/collections/filters.py

@@ -40,18 +40,30 @@ def convert(weav_filter: Optional[_Filters]) -> Optional[base_pb2.Filters]:
 
     @staticmethod
     def __value_filter(weav_filter: _FilterValue) -> base_pb2.Filters:
+        operator = weav_filter.operator._to_grpc()
+        target = _FilterToGRPC.__to_target(weav_filter.target)
+        if isinstance(weav_filter.value, bool):
+            # bool is a subclass of int in Python, so we need to handle it before the int check. Also for whatever reason
+            # the generated code from the proto files does not accept None for value_boolean, while it does for all other types.
+            return base_pb2.Filters(
+                operator=operator,
+                value_boolean=weav_filter.value,
+                target=target,
+            )
+
         return base_pb2.Filters(
-            operator=weav_filter.operator._to_grpc(),
+            operator=operator,
             value_text=_FilterToGRPC.__filter_to_text(weav_filter.value),
-            value_int=weav_filter.value if isinstance(weav_filter.value, int) else None,
-            value_boolean=weav_filter.value if isinstance(weav_filter.value, bool) else None,  # type: ignore
+            value_int=weav_filter.value
+            if isinstance(weav_filter.value, int) and not isinstance(weav_filter.value, bool)
+            else None,
             value_number=(weav_filter.value if isinstance(weav_filter.value, float) else None),
+            value_boolean_array=_FilterToGRPC.__filter_to_bool_list(weav_filter.value),
             value_int_array=_FilterToGRPC.__filter_to_int_list(weav_filter.value),
             value_number_array=_FilterToGRPC.__filter_to_float_list(weav_filter.value),
             value_text_array=_FilterToGRPC.__filter_to_text_list(weav_filter.value),
-            value_boolean_array=_FilterToGRPC.__filter_to_bool_list(weav_filter.value),
             value_geo=_FilterToGRPC.__filter_to_geo(weav_filter.value),
-            target=_FilterToGRPC.__to_target(weav_filter.target),
+            target=target,
         )
 
     @staticmethod
@@ -137,7 +149,12 @@ def __filter_to_float_list(value: FilterValues) -> Optional[base_pb2.NumberArray
 
     @staticmethod
     def __filter_to_int_list(value: FilterValues) -> Optional[base_pb2.IntArray]:
-        if not isinstance(value, list) or not isinstance(value[0], int):
+        # bool is a subclass of int in Python, so the check must ensure it's not a bool
+        if (
+            not isinstance(value, list)
+            or not isinstance(value[0], int)
+            or isinstance(value[0], bool)
+        ):
             return None
 
         return base_pb2.IntArray(values=cast(List[int], value))