comphead opened a new pull request, #3473: URL: https://github.com/apache/iceberg-rust/pull/3473
## Which issue does this PR close? - Closes #3472. ## What changes are included in this PR? An empty projection (`project_field_ids == []`, e.g. `COUNT(*)` or Spark's `df.isEmpty`) was excluded from the metadata-only path in `FileScanTaskReader`, so every column of every data file was read. Removing the `!field_ids.is_empty()` guard makes it take the existing metadata-only path: the `RowNumber` virtual column supplies the row count and `ProjectionMask::none` reads no data columns. The stale comments that described the old behavior are updated. ## Are these changes tested? Yes. `test_empty_projection_reads_no_data_columns` asserts that an empty projection reads fewer bytes than projecting the wide column. It fails without the fix (2838 vs 2787 bytes) and passes with it. The existing `test_empty_projection_preserves_row_count` still covers the row count. The full `iceberg` lib suite passes (1976 tests), and clippy and fmt are clean. ## AI Disclosure Drafted with Claude Code (Anthropic). Opened as a draft pending the author's review. -- 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]
