koodin9 commented on PR #18003: URL: https://github.com/apache/iceberg/pull/18003#issuecomment-5577982091
@bryanck I'd appreciate your review, since you raised the read-performance concern that stopped #14797. This PR is part 1 of 4. It adds `EqualityDeleteConverter`, which resolves the keys of a commit's equality delete files to row positions in the base snapshot and writes deletion vectors, following the approach of Flink's `ConvertEqualityDeletes` but run inside the sink's coordinator before the commit. The later parts wire it into the coordinator, add the delta writer, and add integration tests. Ordering the series this way means no merged state ever produces equality deletes without the conversion in place. Nothing calls the converter yet, so there is no behavior change here. If the direction looks wrong to you, I'd rather hear it on this PR than after the writer lands. -- 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]
