ErikBPF commented on PR #5732: URL: https://github.com/apache/datafusion-comet/pull/5732#issuecomment-5647663069
Thanks for the feedback. Addressed the remaining requests in 7cc641a2c. - Skip unbindable list/map/struct unary null predicates before native serialization, avoiding per-task bind warnings while retaining Spark post-filtering. - Removed the obsolete struct-null scan fallback. Native results now cover NULL/NOT NULL, whole-struct equality, and non-null structs whose children are all null. - Replaced the panicking list constructor with `GenericListArray::try_new(...)?`; the regression reproduced the panic before the fix and now verifies error propagation. - Updated the Iceberg guide and opened #5883 for safe positive-polarity partial-conjunction pruning. Conservative conjunction behavior remains in this PR. Validation: 723 Rust spark-expr tests passed; the full Spark 4.1 Iceberg suite passed 101 tests with one existing SPARK-55626 cancellation. All 12 focused cases passed on Spark 3.5 / Scala 2.12. Native build, whole-reactor packaging, Clippy with warnings denied, semantic/syntactic Scalafix, Spotless, Scalastyle, Prettier, and whitespace checks passed. The focused complex-null run emitted no bind warnings. -- 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]
