[Search] Include Issue Cleanup starting 2026-08-01-preview release (#48767)

* Fix issue in python Collection

* Fix serialization perf regression

* Include issue in changelog

* Address copilot comments

* Revert generated skill documentation edits

Co-authored-by: efrainretana <141282336+efrainretana@users.noreply.github.com>

* Update API surface

* Fix accidental change

---------

Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com>

Author: efrainretana

sdk/search/azure-search-documents/.github/skills/azure-search-documents/scripts/apply_generator_workarounds.py

@@ -92,7 +92,7 @@ class Replacement:
         (
             "class SearchIndexerKnowledgeStoreTableProjectionSelector(TypedDict, total=False):  "
             "# pylint: disable=name-too-long\n"
-                """    \"\"\"Description for what data to store in Azure Tables.
+            """    \"\"\"Description for what data to store in Azure Tables.
 
     :ivar referenceKeyName: Name of reference key to different projection.
     :vartype referenceKeyName: str
@@ -209,4 +209,4 @@ def main() -> int:
 
 
 if __name__ == "__main__":
-    raise SystemExit(main())
+    raise SystemExit(main())

sdk/search/azure-search-documents/CHANGELOG.md

@@ -70,6 +70,8 @@
 
 ### Bugs Fixed
 
+- Improved large document upload and merge performance by avoiding redundant serialization. #46860
+- Made `SearchFieldDataType.Collection` visible to static type checkers. #47929
 - Normalized `SearchIndexClient.update_knowledge_source_file` and its asynchronous equivalent to
   use the name-first signature `(name, file_id, body)`, consistent with the other File knowledge
   source operations.

sdk/search/azure-search-documents/api.md

@@ -23,25 +23,25 @@ namespace azure.search.documents
 
         def add_delete_actions(
                 self, 
-                *documents: Union[List[Dict], List[List[Dict]]], 
+                *documents: Union[Dict[str, Any], List[Dict[str, Any]]], 
                 **kwargs: Any
             ) -> List[IndexAction]: ...
 
         def add_merge_actions(
                 self, 
-                *documents: Union[List[Dict], List[List[Dict]]], 
+                *documents: Union[Dict[str, Any], List[Dict[str, Any]]], 
                 **kwargs: Any
             ) -> List[IndexAction]: ...
 
         def add_merge_or_upload_actions(
                 self, 
-                *documents: Union[List[Dict], List[List[Dict]]], 
+                *documents: Union[Dict[str, Any], List[Dict[str, Any]]], 
                 **kwargs: Any
             ) -> List[IndexAction]: ...
 
         def add_upload_actions(
                 self, 
-                *documents: Union[List[Dict], List[List[Dict]]], 
+                *documents: Union[Dict[str, Any], List[Dict[str, Any]]], 
                 **kwargs: Any
             ) -> List[IndexAction]: ...
 
@@ -13750,25 +13750,25 @@ namespace azure.search.documents.models
 
         def add_delete_actions(
                 self, 
-                *documents: Union[List[Dict], List[List[Dict]]], 
+                *documents: Union[Dict[str, Any], List[Dict[str, Any]]], 
                 **kwargs: Any
             ) -> List[IndexAction]: ...
 
         def add_merge_actions(
                 self, 
-                *documents: Union[List[Dict], List[List[Dict]]], 
+                *documents: Union[Dict[str, Any], List[Dict[str, Any]]], 
                 **kwargs: Any
             ) -> List[IndexAction]: ...
 
         def add_merge_or_upload_actions(
                 self, 
-                *documents: Union[List[Dict], List[List[Dict]]], 
+                *documents: Union[Dict[str, Any], List[Dict[str, Any]]], 
                 **kwargs: Any
             ) -> List[IndexAction]: ...
 
         def add_upload_actions(
                 self, 
-                *documents: Union[List[Dict], List[List[Dict]]], 
+                *documents: Union[Dict[str, Any], List[Dict[str, Any]]], 
                 **kwargs: Any
             ) -> List[IndexAction]: ...
 

sdk/search/azure-search-documents/api.metadata.yml

@@ -1,3 +1,3 @@
-apiMdSha256: 084e314c75fadd638286314c23b600310d948b8748f479b39b840db568f7b0f4
+apiMdSha256: 036650e18f0e3316265c4dd74ee48e52fc99ea2167f7302d443f7974c4099501
 parserVersion: 0.3.31
 pythonVersion: 3.12.1

sdk/search/azure-search-documents/azure/search/documents/indexes/models/_patch.py

@@ -19,6 +19,8 @@
 )
 
 if TYPE_CHECKING:
+    from enum import Enum
+
     from ._models import (
         AzureActiveDirectoryApplicationCredentials,
         DataChangeDetectionPolicy,
@@ -170,19 +172,52 @@ def _collection_helper(typ: Any) -> str:
 
 # Re-export SearchFieldDataType with Collection method
 # The Collection method is added at runtime via monkey-patching
-SearchFieldDataType = _SearchFieldDataType
-SearchFieldDataType.Collection = staticmethod(_collection_helper)  # type: ignore[attr-defined]
+if TYPE_CHECKING:
 
-# Backward-compatible aliases (old camelCase names -> new UPPER_CASE names)
-SearchFieldDataType.String = SearchFieldDataType.STRING  # type: ignore[attr-defined]
-SearchFieldDataType.Int32 = SearchFieldDataType.INT32  # type: ignore[attr-defined]
-SearchFieldDataType.Int64 = SearchFieldDataType.INT64  # type: ignore[attr-defined]
-SearchFieldDataType.Single = SearchFieldDataType.SINGLE  # type: ignore[attr-defined]
-SearchFieldDataType.Double = SearchFieldDataType.DOUBLE  # type: ignore[attr-defined]
-SearchFieldDataType.Boolean = SearchFieldDataType.BOOLEAN  # type: ignore[attr-defined]
-SearchFieldDataType.DateTimeOffset = SearchFieldDataType.DATE_TIME_OFFSET  # type: ignore[attr-defined]
-SearchFieldDataType.GeographyPoint = SearchFieldDataType.GEOGRAPHY_POINT  # type: ignore[attr-defined]
-SearchFieldDataType.ComplexType = SearchFieldDataType.COMPLEX  # type: ignore[attr-defined]
+    # pylint: disable=enum-must-inherit-case-insensitive-enum-meta,enum-must-be-uppercase
+    class SearchFieldDataType(str, Enum):
+        STRING = "Edm.String"
+        INT32 = "Edm.Int32"
+        INT64 = "Edm.Int64"
+        DOUBLE = "Edm.Double"
+        BOOLEAN = "Edm.Boolean"
+        DATE_TIME_OFFSET = "Edm.DateTimeOffset"
+        GEOGRAPHY_POINT = "Edm.GeographyPoint"
+        COMPLEX = "Edm.ComplexType"
+        SINGLE = "Edm.Single"
+        HALF = "Edm.Half"
+        INT16 = "Edm.Int16"
+        S_BYTE = "Edm.SByte"
+        BYTE = "Edm.Byte"
+        String = "Edm.String"
+        Int32 = "Edm.Int32"
+        Int64 = "Edm.Int64"
+        Single = "Edm.Single"
+        Double = "Edm.Double"
+        Boolean = "Edm.Boolean"
+        DateTimeOffset = "Edm.DateTimeOffset"
+        GeographyPoint = "Edm.GeographyPoint"
+        ComplexType = "Edm.ComplexType"
+
+        @staticmethod
+        def Collection(typ: Union[str, "SearchFieldDataType"]) -> str:
+            return _collection_helper(typ)
+
+    # pylint: enable=enum-must-inherit-case-insensitive-enum-meta,enum-must-be-uppercase
+
+else:
+    SearchFieldDataType = _SearchFieldDataType
+    SearchFieldDataType.Collection = staticmethod(_collection_helper)  # type: ignore[attr-defined]
+    # Backward-compatible aliases (old camelCase names -> new UPPER_CASE names)
+    SearchFieldDataType.String = SearchFieldDataType.STRING  # type: ignore[attr-defined]
+    SearchFieldDataType.Int32 = SearchFieldDataType.INT32  # type: ignore[attr-defined]
+    SearchFieldDataType.Int64 = SearchFieldDataType.INT64  # type: ignore[attr-defined]
+    SearchFieldDataType.Single = SearchFieldDataType.SINGLE  # type: ignore[attr-defined]
+    SearchFieldDataType.Double = SearchFieldDataType.DOUBLE  # type: ignore[attr-defined]
+    SearchFieldDataType.Boolean = SearchFieldDataType.BOOLEAN  # type: ignore[attr-defined]
+    SearchFieldDataType.DateTimeOffset = SearchFieldDataType.DATE_TIME_OFFSET  # type: ignore[attr-defined]
+    SearchFieldDataType.GeographyPoint = SearchFieldDataType.GEOGRAPHY_POINT  # type: ignore[attr-defined]
+    SearchFieldDataType.ComplexType = SearchFieldDataType.COMPLEX  # type: ignore[attr-defined]
 
 
 def Collection(typ: Any) -> str:
@@ -257,7 +292,7 @@ def SimpleField(
     :rtype:  SearchField
     """
     # If type is an enum, get its value; otherwise use it as-is
-    field_type = type.value if hasattr(type, "value") else type
+    field_type = type.value if isinstance(type, SearchFieldDataType) else type
     result: Dict[str, Any] = {
         "name": name,
         "type": field_type,

sdk/search/azure-search-documents/azure/search/documents/models/_patch.py

@@ -9,29 +9,47 @@
 Follow our quickstart for examples: https://aka.ms/azsdk/python/dpcodegen/python/customize
 """
 
+from enum import Enum
 from typing import Any, Dict, List, Tuple, Union, cast, Optional
 from azure.core.exceptions import HttpResponseError
 
+from .._utils.model_base import SdkJSONEncoder
 from ._enums import IndexActionType
 from ._models import IndexAction
 from ._models import IndexDocumentsBatch as IndexDocumentsBatchGenerated
 
+_ENUM_ENCODER_MARKER = "_azure_search_documents_handles_enum"
 
-def _flatten_args(args: Tuple[Union[List[Dict[Any, Any]], List[List[Dict[Any, Any]]]], ...]) -> List[Dict]:
+
+def _patch_sdk_json_encoder() -> None:
+    default = SdkJSONEncoder.default
+    if getattr(default, _ENUM_ENCODER_MARKER, False):
+        return
+
+    def default_with_enum(self: SdkJSONEncoder, value: Any) -> Any:
+        if isinstance(value, Enum):
+            return value.value
+        return default(self, value)
+
+    setattr(default_with_enum, _ENUM_ENCODER_MARKER, True)
+    setattr(SdkJSONEncoder, "default", default_with_enum)
+
+
+def _flatten_args(args: Tuple[Union[Dict[str, Any], List[Dict[str, Any]]], ...]) -> List[Dict[str, Any]]:
     """Flatten variadic arguments into a single list of documents.
 
     Supports both:
     - add_upload_actions([doc1, doc2])  # single list
     - add_upload_actions(doc1, doc2)    # multiple args
 
-    :param args: Variadic arguments containing documents or lists of documents.
-    :type args: Tuple[Union[List[Dict[Any, Any]], List[List[Dict[Any, Any]]]], ...]
+    :param args: Variadic arguments containing documents or a list of documents.
+    :type args: tuple[dict[str, Any] or list[dict[str, Any]], ...]
     :return: A flattened list of document dictionaries.
     :rtype: List[Dict]
     """
     if len(args) == 1 and isinstance(args[0], (list, tuple)):
-        return cast(List[Dict], args[0])
-    return cast(List[Dict], args)
+        return list(cast(List[Dict[str, Any]], args[0]))
+    return list(cast(Tuple[Dict[str, Any], ...], args))
 
 
 class RequestEntityTooLargeError(HttpResponseError):
@@ -53,7 +71,9 @@ def __init__(self, *, actions: Optional[List[IndexAction]] = None) -> None:
     def __repr__(self) -> str:
         return "<IndexDocumentsBatch [{} actions]>".format(len(self.actions) if self.actions else 0)[:1024]
 
-    def add_upload_actions(self, *documents: Union[List[Dict], List[List[Dict]]], **kwargs: Any) -> List[IndexAction]:
+    def add_upload_actions(
+        self, *documents: Union[Dict[str, Any], List[Dict[str, Any]]], **kwargs: Any
+    ) -> List[IndexAction]:
         # pylint: disable=unused-argument
         """Add documents to upload to the Azure search index.
 
@@ -69,7 +89,9 @@ def add_upload_actions(self, *documents: Union[List[Dict], List[List[Dict]]], **
         """
         return self._extend_batch(_flatten_args(documents), IndexActionType.UPLOAD)
 
-    def add_delete_actions(self, *documents: Union[List[Dict], List[List[Dict]]], **kwargs: Any) -> List[IndexAction]:
+    def add_delete_actions(
+        self, *documents: Union[Dict[str, Any], List[Dict[str, Any]]], **kwargs: Any
+    ) -> List[IndexAction]:
         # pylint: disable=unused-argument
         """Add documents to delete from the Azure search index.
 
@@ -90,7 +112,9 @@ def add_delete_actions(self, *documents: Union[List[Dict], List[List[Dict]]], **
         """
         return self._extend_batch(_flatten_args(documents), IndexActionType.DELETE)
 
-    def add_merge_actions(self, *documents: Union[List[Dict], List[List[Dict]]], **kwargs: Any) -> List[IndexAction]:
+    def add_merge_actions(
+        self, *documents: Union[Dict[str, Any], List[Dict[str, Any]]], **kwargs: Any
+    ) -> List[IndexAction]:
         # pylint: disable=unused-argument
         """Add documents to merge in to existing documents in the Azure search
         index.
@@ -109,7 +133,7 @@ def add_merge_actions(self, *documents: Union[List[Dict], List[List[Dict]]], **k
         return self._extend_batch(_flatten_args(documents), IndexActionType.MERGE)
 
     def add_merge_or_upload_actions(
-        self, *documents: Union[List[Dict], List[List[Dict]]], **kwargs: Any
+        self, *documents: Union[Dict[str, Any], List[Dict[str, Any]]], **kwargs: Any
     ) -> List[IndexAction]:
         # pylint: disable=unused-argument
         """Add documents to merge in to existing documents in the Azure search
@@ -175,7 +199,7 @@ def enqueue_actions(self, new_actions: Union[IndexAction, List[IndexAction]], **
         else:
             self._actions.extend(new_actions)
 
-    def _extend_batch(self, documents: List[Dict], action_type: str) -> List[IndexAction]:
+    def _extend_batch(self, documents: List[Dict[str, Any]], action_type: str) -> List[IndexAction]:
         """Internal helper to extend the batch with new actions.
 
         :param documents: The documents to add
@@ -190,9 +214,8 @@ def _extend_batch(self, documents: List[Dict], action_type: str) -> List[IndexAc
 
         new_actions = []
         for doc in documents:
-            action_dict = {"@search.action": action_type}
-            action_dict.update(doc)
-            action = IndexAction(action_dict)
+            action = IndexAction(action_type=action_type)
+            action.update(doc)
             new_actions.append(action)
 
         self._actions.extend(new_actions)
@@ -215,3 +238,5 @@ def patch_sdk():
     you can't accomplish using the techniques described in
     https://aka.ms/azsdk/python/dpcodegen/python/customize
     """
+
+    _patch_sdk_json_encoder()

sdk/search/azure-search-documents/samples/sample_knowledge_base_crud.py

@@ -73,6 +73,7 @@ def update_knowledge_base():
     # [START update_knowledge_base]
     from azure.core.credentials import AzureKeyCredential
     from azure.search.documents.indexes import SearchIndexClient
+
     index_client = SearchIndexClient(service_endpoint, AzureKeyCredential(key))
     knowledge_base = index_client.get_knowledge_base(knowledge_base_name)
     knowledge_base.tags = {"environment": "sample", "owner": "retrieval-team"}

sdk/search/azure-search-documents/samples/sample_knowledge_base_crud_async.py

@@ -76,6 +76,7 @@ async def update_knowledge_base_async():
     # [START update_knowledge_base_async]
     from azure.core.credentials import AzureKeyCredential
     from azure.search.documents.indexes.aio import SearchIndexClient
+
     index_client = SearchIndexClient(service_endpoint, AzureKeyCredential(key))
     async with index_client:
         knowledge_base = await index_client.get_knowledge_base(knowledge_base_name)

sdk/search/azure-search-documents/samples/sample_knowledge_source_crud.py

@@ -215,11 +215,7 @@ def create_private_blob_knowledge_source():  # pylint: disable=too-many-locals
         assert indexer_client.get_data_source_connection(data_source_name).name == data_source_name
         assert indexer_client.get_indexer_status(indexer_name).status is not None
         generated_index = index_client.get_index(generated_index_name)
-        analyzers = {
-            str(field.analyzer_name)
-            for field in generated_index.fields
-            if field.analyzer_name is not None
-        }
+        analyzers = {str(field.analyzer_name) for field in generated_index.fields if field.analyzer_name is not None}
         assert os.environ["AZURE_SEARCH_EXPECTED_ANALYZER"] in analyzers
         assert os.environ["AZURE_SEARCH_EXPECTED_FALLBACK_ANALYZER"] in analyzers
         print(f"Verified generated resources and analyzers: {sorted(analyzers)}")

sdk/search/azure-search-documents/samples/sample_knowledge_source_crud_async.py

@@ -223,9 +223,7 @@ async def create_private_blob_knowledge_source_async():  # pylint: disable=too-m
             assert (await indexer_client.get_indexer_status(indexer_name)).status is not None
             generated_index = await index_client.get_index(generated_index_name)
             analyzers = {
-                str(field.analyzer_name)
-                for field in generated_index.fields
-                if field.analyzer_name is not None
+                str(field.analyzer_name) for field in generated_index.fields if field.analyzer_name is not None
             }
             assert os.environ["AZURE_SEARCH_EXPECTED_ANALYZER"] in analyzers
             assert os.environ["AZURE_SEARCH_EXPECTED_FALLBACK_ANALYZER"] in analyzers