wayne-x-wang commented on PR #7792: URL: https://github.com/apache/kyuubi/pull/7792#issuecomment-6460012275
@pan3793 Thanks for the detailed review. I agree that input-table collection should cover subqueries consistently across plan nodes, rather than handling only `Filter`. I’ll update the patch to: - Collect subquery input tables generically, gated by `legacy=false`, while preserving the existing scalar-subquery path for column lineage. I’ll also document why the table-collection traversal intentionally discards the returned column mappings. - Add coverage for subqueries in the SELECT list, JOIN conditions, and REPLACE WHERE predicates where supported, plus explicit column-lineage assertions for `SELECT (SELECT t1.col1)`. - Remove the duplicate EXISTS test, reuse table fixtures for the analysis-only cases, and replace the diagnostic dumps with useful assertion context. - Remove the redundant configuration-restoration assertions and query execution/result-row checks, while retaining cleanup and configuration restoration in `finally`. The earlier execution checks confirmed that the reported SQL succeeds independently of lineage extraction; I agree that the committed parser tests only need the analyzed plans and lineage assertions. I’ll verify the new regression cases before applying the changes, then rerun the lineage suite on Spark 3.5 and 4.2. -- 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]
