andygrove opened a new pull request, #6511: URL: https://github.com/apache/datafusion-comet/pull/6511
## Which issue does this PR close? Closes #6505 on `branch-1.1`, for 1.1.0-rc2. ## Rationale for this change This is the `branch-1.1` backport of #6510. #5237 made the native Parquet scan fill in `_metadata.file_block_start` and `_metadata.file_block_length`, and when Spark splits a file, DataFusion and Spark's parquet-mr reader assign row groups to splits differently, so rows can report the wrong split's values. 1.0.0 fell back to Spark for these columns. #6510 has the details. ## What changes are included in this PR? A clean cherry-pick (`-x`) of #6510's commit. A scan that reads either block column falls back to Spark, and the other `_metadata` columns stay native. ## How are these changes tested? On this branch, the `_metadata` tests in `ParquetReadV1Suite`, including the new one, and `CometScanRuleSuite` pass on the Spark 4.1 profile, 15 tests. -- 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]
