QuakeWang opened a new pull request, #8284:
URL: https://github.com/apache/paimon/pull/8284

   ### Purpose
   
   Vector search splits must keep the vector index selection tied to the vector 
column as the primary index field, because vector/full-text readers rebuild the 
search indexer from `indexFieldId`.
   
   The scalar pre-filter path is different: a multi-column global index can 
satisfy a filter field stored in `extraFieldIds`. Previously `VectorScanImpl` 
only matched filter fields against `indexFieldId`, so vector search missed 
usable scalar indexes when the filtered column was an extra field.
   
   This PR matches vector filter fields against all indexed fields while 
keeping vector/full-text search columns restricted to primary index fields, and 
adds regression coverage for that contract.
   
   ### Tests
   
   - `mvn -pl paimon-core -am -Pfast-build -DfailIfNoTests=false 
-Dtest=FullTextSearchBuilderTest test`
   - `mvn -pl paimon-core -am -Pfast-build -DfailIfNoTests=false 
-Dtest=VectorSearchBuilderTest#testVectorSearchRequiresVectorColumnAsPrimaryField
 test`
   - `mvn -pl paimon-core -am -Pfast-build -DfailIfNoTests=false 
-Dtest=VectorSearchBuilderTest#testVectorSearchWithBTreePreFilter test`
   - `mvn -pl paimon-core -am -Pfast-build -DfailIfNoTests=false 
-Dtest=VectorSearchBuilderTest#testScanPartialRangeIntersection test`


-- 
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