jackylee-ch opened a new pull request, #984: URL: https://github.com/apache/paimon-rust/pull/984
`WHEN NOT MATCHED THEN INSERT VALUES (s.a, s.b, ...)` without an explicit column list built an empty column->expression map, so `insert_select_clause` emitted `NULL` for every target column and dropped the source values. The merge then inserted all-NULL rows (or failed on a non-null column) instead of the source data — silent corruption for a standard SQL / Spark write pattern. The existing MERGE INSERT tests all pass an explicit column list, so this went uncovered. Map the VALUES to the table's columns by position when no column list is given, and reject a value/column count mismatch. Explicit-column and `INSERT *` clauses are unchanged. Added an end-to-end MERGE test using a column-less INSERT; it inserts all-NULL rows before the fix. -- 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]
