taowenjun opened a new issue, #13128:
URL: https://github.com/apache/seatunnel/issues/13128

   ### Search before asking
   
   I searched existing open and closed issues and pull requests. #13085 
implements opt-in inheritance for existing ORC tables; I did not find a 
separate implementation for Parquet table compression inheritance.
   
   ### Description
   
   An existing Hive Parquet table can declare `parquet.compression=SNAPPY`, but 
the Hive sink currently uses its own compression configuration unless the job 
repeats the codec. This can make the generated files inconsistent with the 
table's declared compression.
   
   I would like to contribute a separate Parquet follow-up to #13087, as 
suggested in 
https://github.com/apache/seatunnel/issues/13087#issuecomment-6593771299. This 
issue does not expand the scope of #13087 or #13085, which remain focused on 
ORC.
   
   ### Proposed behavior
   
   - Reuse the default-off `use_table_compression` option introduced by #13085, 
once that prerequisite is available.
   - For a Parquet table that exists when the sink is initialized, use explicit 
`compress_codec` first, then `parquet.compression` when inheritance is enabled, 
then the existing default.
   - Explicit `NONE` must override a compressed table property. Missing, null 
or blank properties retain existing behavior.
   - Normalize whitespace and case; accept Parquet's `UNCOMPRESSED` as the 
uncompressed alias. Reject unknown or incompatible inherited codecs instead of 
silently writing another codec.
   - Keep automatic table creation and unrelated formats unchanged, and do not 
modify Hive table metadata.
   
   ### Validation
   
   Add configuration tests for the option default, precedence, normalization, 
null/blank properties, codec validation and format isolation. Verify actual 
Parquet footer codecs and exact row round trips, including inherited SNAPPY and 
explicit NONE. Add matching Hive integration coverage and English/Chinese 
documentation.
   
   This work depends on the opt-in option in #13085. I will keep the Parquet 
changes separate for review and rebase onto dev after that prerequisite merges. 
May this follow-up be assigned to me (`taowenjun`)?
   


-- 
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