JingsongLi opened a new pull request, #902: URL: https://github.com/apache/paimon-rust/pull/902
## Summary - Resolve BLOB views lazily for `LIMIT` reads, after ordinary-column filtering and quota slicing of the current batch. This avoids loading an invalid or expensive upstream reference outside the returned rows. - Cache descriptors for references seen in prior batches; keep the existing prescan path for unlimited reads to avoid changing its request pattern. - Preserve existing behavior for predicates on view values, which still need candidate view resolution before evaluation. ## Verification - Reproduced the prior failure: `LIMIT 2` tried to resolve an invalid reference in row 3 and returned an error. - The integration regression now checks repeated view references across batches, an invalid reference after the quota, and an ordinary-column predicate. - `cargo test -p paimon table::data_evolution_reader:: --lib --no-default-features --features storage-memory,storage-fs` (80 passed) - `cargo test -p paimon --test rest_catalog_test --no-default-features --features storage-memory,storage-fs,storage-oss` (79 passed) - `cargo fmt --check` and `git diff --check` passed. -- 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]
