CrownChu commented on PR #10771: URL: https://github.com/apache/paimon/pull/10771#issuecomment-6493918294
Fixed the CI compilation regression in d31e015a7e0e591b7f0981fae14f36929fe6e545. The previous commit changed the protected `readIndexed` argument from `GlobalIndexer` to `List<GlobalIndexer>`, but the Flink and Spark local-search paths still passed one indexer. All 12 failing build jobs stopped at these compilation errors; they were not ES test assertion failures. The fix restores the original protected signature while still choosing a reader per index definition internally. It also reuses the supplied indexer for the first definition, so subsequent raw reads retain the metric loaded from the index metadata. Two real-ES-archive regression cases cover this subclass entry point with matching/different definitions and a cosine-index/L2-table/raw-tail combination. Local validation: normal-check `clean package` succeeded for 31 reactor modules on JDK 11, with **503 tests passed, 0 failures/errors/skips**. This includes ESLib (144), core (285), common (56), Lumina (6), Flink vector reads (8), and Spark vector reads (4), using the Flink 1 / Spark 3 / Scala 2.12 profiles. This is targeted local validation, not a claim that the complete CI matrix has passed. The [new CI run](https://github.com/apache/paimon/actions/runs/40855614099) has been triggered. -- 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]
