Brunda10 commented on code in PR #54597:
URL: https://github.com/apache/airflow/pull/54597#discussion_r2299921707
##########
airflow-core/src/airflow/api_fastapi/core_api/services/public/common.py:
##########
@@ -72,3 +73,71 @@ def handle_bulk_update(self, action: BulkUpdateAction[T],
results: BulkActionRes
def handle_bulk_delete(self, action: BulkDeleteAction[T], results:
BulkActionResponse) -> None:
"""Bulk delete entities."""
raise NotImplementedError
+
+
+class PatchUtil:
+ """
+ Utility class for applying patch operations with support for update masks.
+
+ This helper is used to update only selected fields of a Pydantic model
instance,
+ while ensuring:
+ - Only fields present in the update_mask are modified.
+ - Non-updatable fields can be excluded explicitly.
+ - Both raw field names and aliases are handled correctly.
+ """
+
+ @staticmethod
+ def apply_patch_with_update_mask(
+ model,
+ patch_body,
+ update_mask: list[str] | None,
+ non_update_fields: set[str] | None = None,
+ ):
+ """
+ Apply a patch to the given model using the provided update mask.
+
+ Args:
+ model: The model instance to update.
+ patch_body: Pydantic model containing patch data.
+ update_mask (list[str] | None): Optional list of fields to update.
+ non_update_fields (set[str] | None): Fields that should not be
updated.
+
+ Returns:
+ The updated model instance.
+
+ Raises:
+ HTTPException: If invalid fields are provided in update_mask.
+ """
+ print("Inside patch")
+
+ # Always dump without aliases for internal validation
+ raw_data = patch_body.model_dump(by_alias=False)
+
+ fields_set = set(patch_body.model_fields_set)
+ allowed_fields = set(patch_body.model_fields.keys())
+
+ print(fields_set, update_mask)
+ if update_mask:
+ invalid = set(update_mask) - allowed_fields
+ if invalid:
+ raise HTTPException(
+ status_code=status.HTTP_400_BAD_REQUEST,
+ detail=f"Invalid fields in update_mask: {invalid}",
+ )
+ fields_set = fields_set.intersection(update_mask)
+
+ if non_update_fields:
+ fields_set = fields_set - non_update_fields
+
+ # Validate only the subset of fields we want to update
+ print("Validating data with fields_set:", fields_set, raw_data)
+ validated_data = {key: raw_data[key] for key in fields_set if key in
raw_data}
Review Comment:
I initially added the error to make it clear when they passed an
unsupported field in update_mask. But I agree that ignoring invalid fields via
intersection is more consistent with the other APIs. I’ll update the code
accordingly.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
