sam-1112 commented on PR #5957: URL: https://github.com/apache/datafusion-comet/pull/5957#issuecomment-5959076097
I added the symmetric `ColumnarToRowExec` branch for row-output stages whose reverted root remains columnar. The `DISTRIBUTE BY` test now runs with AQE on and off, and the AQE-on case explicitly checks for a coalesced read over a Comet shuffle. There is also a scan-only regression test for a reverted vectorized Spark scan. I also updated map-stage fallback to use `CometSparkToColumnarExec` before native shuffle, since Spark columnar batches do not satisfy its Arrow input contract. New cases cover both row-based and columnar fallback roots. -- 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]
