JingsongLi opened a new pull request, #11319:
URL: https://github.com/apache/paimon/pull/11319

   ### Purpose
   
   After an index build, a normal Data Evolution scalar query could silently 
omit newly appended matches. Default batch scans to `full` so they scan 
uncovered rows from the selected snapshot. Explicit `scalar-index.search-mode` 
still takes precedence over `global-index.search-mode`. Scalar filters inside 
vector and full-text search keep their `fast` default.
   
   Persist the effective mode in new reader-side index splits so recovery 
replays the selection made during planning.
   
   ### Tests
   
   - Reproduced the missing appended match with both planning-time and 
reader-side Java index evaluation and with the Python scanner. The tests now 
cover default behavior and explicit overrides.
   - 295 Core tests and 4 Flink sorted-index integration tests passed.
   - PyPaimon focused suites covering scalar indexes, vector filters, composite 
indexes and DML passed with 332 tests. The full-text suite skipped 33 tests in 
the local environment.
   


-- 
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