a-narsudinov opened a new pull request, #20644: URL: https://github.com/apache/hudi/pull/20644
### Describe the issue this Pull Request addresses With column-name case resolution enabled, the Trino connector translates index predicates by looking up each column in the Avro table schema. Hidden columns such as `$path`, `$file_size` and `$file_modified_time` have no physical schema field, so predicates on these columns can fail during query planning. Queries combining physical-column and hidden-column predicates are affected as well. ### Summary and Changelog Exclude synthesized columns from Hudi index predicate translation while preserving SQL filtering. - Remove hidden columns from the tuple domain passed to index strategies before resolving physical column names. - Preserve physical-column predicates for index-based pruning. - Keep the original scan predicates and Trino residual filtering intact so hidden-column conditions are still evaluated. - Add SQL coverage with Hudi metadata and column-name case resolution enabled. - Verify hidden-only and mixed predicates, unmatched paths, positive file sizes and non-null modification times. Validation: both tests in `TestHudiHiddenColumnPredicates` passed. ### Impact Allows queries with hidden-column predicates to execute when column-name case resolution is enabled. Hidden-column conditions remain enforced without requiring those columns to exist in the Hudi schema or indexes. No public API, configuration or storage-format changes. ### Risk Level low — The change only removes synthesized columns from index predicate translation. SQL tests verify that hidden-column filtering remains effective and that mixed predicates return the expected results. ### Documentation Update none ### Contributor's checklist - [x] Read through [contributor's guide](https://hudi.apache.org/contribute/how-to-contribute) - [x] Enough context is provided in the sections above - [x] Adequate tests were added if applicable -- 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]
