0lai0 commented on PR #4653: URL: https://github.com/apache/datafusion-comet/pull/4653#issuecomment-6669568630
Thanks @andygrove for review . All three addressed: - **Title and test names:** retitled to `fix: find a SparkError anywhere in the error source chain at the JNI boundary` and dropped `#4517` from the two `CometCodegenSuite` test names. - **`CheckedBinaryExpr`:** added unit tests for both cases: `Context(External(SparkError))` comes back as `External(SparkErrorWithContext)` with the expression's context, and `External(SparkErrorWithContext)` keeps its inner context. - **A query that hits it today:** yes, the hash join build side. When the build side runs in the same native plan as the join (an aggregate already partitioned on the join key), DataFusion wraps its error in `Shared`. Added a `CometJoinSuite` test with `SHUFFLE_HASH` over `1 / sum(v)`; it fails on main with `CometNativeException` and passes here. I couldn't find a `Context` or `Diagnostic` producer in DataFusion 55.2 or Comet, so the description says those cases are defensive. -- 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]
