urosstan-db commented on PR #58958: URL: https://github.com/apache/spark/pull/58958#issuecomment-5779538689
**Reviewed by Claude (Claude Code)** — no issues found. Both production changes are correct and minimally scoped: - The `partitionColumn` guard in `JDBCScanBuilder.pushDownJoin` avoids the invalid `query` + `partitionColumn` combination that `JDBCOptions` rejects. Checking only this side is sufficient because `isOtherSideCompatibleForJoin` already requires both sides' options (including `partitionColumn`) to be identical before a join is attempted. - The AVG-rewrite normalization change makes the second normalization pass consistent with the first-pass and grouping-expression normalization, so the rewritten `SUM`/`COUNT` resolves against the pushed-join aliases and aggregates the correct column. The two added regression tests exercise both fixed behaviors and fail on the pre-fix code (the wrong AVG result, and the `JDBCOptions` rejection of `query` + `partitionColumn`). The three earlier inline review requests — the join-pushdown=true test path with not-pushed and full-result assertions, the `idCol` inner val, and the `rowsWithJoinPushdown` naming — are all implemented. The note that non-H2 database integration suites were not run is an accurate, accepted limitation. No blocking issues; the change looks ready. -- 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] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
