JingsongLi commented on PR #9822: URL: https://github.com/apache/paimon/pull/9822#issuecomment-5832651229
Re-reviewed head `b0fd1ea` for production. The Cassandra `WRITETIME` use case supplies clear end-to-end value. The latest fix allocates synthetic changelog metadata IDs above the full recursive physical field-ID set, addressing my prior nested-field collision; it also adds a Flink SQL read/write regression with a nested `ROW` and metadata alias. Local JDK 8 core suites passed: `ChangelogEventMetadataTest` 2/2, `SchemaValidationTest` 72/72, and `LookupChangelogMergeFunctionWrapperTest` 13/13. I attempted the four Flink integration cases locally, but each failed at its initial manifest write before the changed logic because this isolated checkout loads an incompatible Avro jar (`DataFileWriter.setEncoder` `NoSuchMethodError`), the same local classpath issue observed in the earlier review. Exact-head Flink 1/2, Spark, JDK 8/11, E2E, docs and licensing CI are all green. `git diff --check` passed. I found no remaining code blocker; keep this PR open for merge. -- 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]
