JingsongLi opened a new pull request, #955: URL: https://github.com/apache/paimon-rust/pull/955
## Summary - Align primary-key merge engines with Java for partial update sequence groups, field aggregation, DELETE / UPDATE_BEFORE handling, whole-row and sequence-group deletes, and cross-partition updates. - Implement every Java field aggregator registered by Paimon, including collect, merge-map, nested update, Roaring 32/64, HLL/Theta sketches, and primary-key; add Java-compatible retract and option behavior to existing numeric, value, and listagg functions. - Use the same partial-update merge function for both buffered writes and reads so the two paths cannot drift. ## Java parity checks - Compared merge functions and aggregators under `paimon-core/.../mergetree/compact` and their option validation. - Added cross-language byte fixtures for Java Roaring/HLL/Theta formats, including Theta UpdateSketch input, and regression tests for tombstones, sparse inserts after deletes, sequence groups, retracts, and cross-partition routing. ## Verification - `cargo test -p paimon --lib --quiet` (3,282 passed; 6 ignored) - `cargo test --locked -p paimon --all-targets --features fulltext,vortex --quiet` (passed) - `cargo clippy --locked --all-targets --workspace --features fulltext,vortex -- -D warnings` (passed) - `cargo fmt --all` and `git diff --check` (passed) ## Existing limitation Writing partial-update or aggregation tables with `deletion-vectors.enabled=true` still requires a separate deletion-vector write path; this PR preserves the existing explicit rejection rather than writing incorrect data. -- 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]
