TheR1sing3un opened a new pull request, #10427: URL: https://github.com/apache/paimon/pull/10427
### Purpose Enable Ray `read_by_row_id` and `update_by_row_id` for data-evolution tables with deletion vectors. Routed reads now preserve the owning split's deletion files and snapshot metadata, so tombstones are filtered at the pinned snapshot. Update workers validate requested IDs against that snapshot's deletion vectors before rewriting each group while preserving physical row positions. Deleted read IDs are omitted; deleted update IDs raise an error. A delete committed after update preparation remains effective. Document these semantics and remove the former blanket rejection. ### Tests All 46 row-ID read/update tests passed (45 in the full run, followed by the corrected and expanded DV update test). Coverage includes multiple files, duplicate IDs, snapshot pinning across deletes, all-deleted results with typed schema, live-row updates with physical gaps, rejection of deleted updates, and a concurrent delete between update preparation and commit. Flake8 and `git diff --check` passed. No benchmark code included. -- 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]
