linliu-code commented on PR #19610:
URL: https://github.com/apache/hudi/pull/19610#issuecomment-5401664515

   Re-synced with the internal copy after the latest review round — both are 
now `+944/-29` over the same 4 files.
   
   Changes in this push, per review on onehouseinc/hudi-internal#2197 (keeping 
the detailed discussion on that thread, as agreed):
   
   - `globalIndexTypes` and `isGlobalIndexEnabled` folded into a single map 
keyed by index type, so an index cannot be recorded as global without also 
recording the `update.partition.path` config that decides the re-keying 
exemption. Semantics unchanged — unmapped still yields `false`.
   - Comments cut by about half (-94 lines net across both spark copies), 
keeping only what is not visible in the code.
   - `w/in` -> `within` in the resolution error.
   - New CoW test covering the accepting half of `update.partition.path=true`: 
a source that carries a *new* partition value must relocate the record. It 
does; mutation-checked. Plus a value assertion in the dedupe test and an ENG 
ref where MOR is excluded.
   
   31 tests, 31 passed locally.


-- 
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