amoghrajesh commented on code in PR #76575:
URL: https://github.com/apache/airflow/pull/76575#discussion_r4546561451


##########
dev/registry/registry_contract_models.py:
##########
@@ -146,17 +146,33 @@ class ModuleContract(BaseModel):
     provider_name: str | None = None
     supports_durable_execution: bool = False
     supports_deferrable: bool = False
+    durable_gating_provider: str | None = None

Review Comment:
   Good qn. The merge step now runs from the workflow's checkout, not the tag, 
like `make_mnt_writeable.sh` already does. So main's contract always reads 
main's data.
   
   Checked with `providers/2026-11-03`'s files: its contract rejects an entry 
main wrote, this branch's accepts both. Re-publishing an old wave drops the new 
fields for its providers until the next publish from main.
   
   Thats my thought, do you disagree?



##########
dev/registry/merge_registry_data.py:
##########
@@ -88,7 +91,7 @@ def merge(
     merged_modules.sort(key=lambda m: date_by_provider.get(m["provider_id"], 
""), reverse=True)
 
     providers_payload = validate_providers_catalog({"providers": 
merged_providers})
-    modules_payload = validate_modules_catalog({"modules": merged_modules})
+    modules_payload = validate_extracted_modules_catalog({"modules": 
merged_modules})

Review Comment:
   Back on `validate_modules_catalog`. Added a test that merge rejects 
`durable_mechanism` and `durable_gating_class`.



-- 
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]

Reply via email to