nsivabalan commented on issue #17044: URL: https://github.com/apache/hudi/issues/17044#issuecomment-4800739714
This appears to be fixed on master by [bc7a766ff3e](https://github.com/apache/hudi/commit/bc7a766ff3e810caaf18d288cc3a386d0c94d008) (PR #17936, *feat: use ScanOperation for Spark 3.3 and 3.4 partition pruning*). The new `Spark33HoodiePruneFileSourcePartitions` rule (`hudi-spark-datasource/hudi-spark3.3.x/src/main/scala/org/apache/spark/sql/hudi/analysis/Spark33HoodiePruneFileSourcePartitions.scala`) switches Spark 3.3 from `PhysicalOperation` to `ScanOperation` and explicitly retains deterministic filters: ```scala case op @ ScanOperation(projects, filters, lr @ LogicalRelation(HoodieRelationMatcher(fileIndex), _, _, _)) if !fileIndex.hasPredicatesPushedDown => val deterministicFilters = filters.filter(f => f.deterministic && !SubqueryExpression.hasSubquery(f)) ``` `HoodieAnalysis.scala` routes Spark 3.3 to this implementation while Spark 3.4+ keeps the unified `Spark3HoodiePruneFileSourcePartitions` (post-SPARK-39764). Could you verify on a recent master build and close if resolved? -- 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]
