limadog9 opened a new pull request, #25857: URL: https://github.com/apache/datafusion/pull/25857
## Which issue does this PR close? Closes #25829. ## Rationale for this change `DataFrame::fill_null` and `fill_nan` report a schema error when a column is named `Name` or `a.b`, even if that column is not being filled. Rebuilding the projection parses these existing names as SQL identifiers, changing their case or treating a dot as a table qualifier. ## What changes are included in this PR? Build column expressions directly from each schema entry, preserving the exact name and table qualifier in the replacement, untouched-column, and failed-cast paths. Two existing Windows lint blockers are also corrected so the required all-target/all-feature Clippy check can run: a redundant semicolon and the intentionally async interface of the unavailable FIFO example. Neither changes runtime behavior. ## What is the testing strategy for this PR? Ten DataFrame API regression cases cover both methods with uppercase and dotted names, untouched columns, uncastable replacement values, and duplicate qualified column names passed through a join. All ten new cases fail with the expected schema errors without the fix; the eight existing fill tests pass on that baseline. - `cargo fmt --all` passes. - All 18 focused DataFrame fill tests pass (10 new regressions and 8 existing tests). - `cargo clippy --all-targets --all-features -- -D warnings` passes with zero warnings. No existing benchmark exercises these DataFrame methods; execution kernels are unchanged. ## Are there any user-facing changes? Both methods accept existing column names literally and preserve qualifiers when passing columns through. Public API signatures and replacement/casting behavior are unchanged. -- 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]
