ViktorBystrov72 opened a new pull request, #18991: URL: https://github.com/apache/iceberg/pull/18991
Equal decimal literals such as `100.0` and `100.00` can produce different bucket filters, pruning matching partitions. Convert decimal and string literals to the column scale only when that preserves their exact value. Use numeric comparison for decimal set membership so IN/NOT IN agrees with decimal equality. Literals requiring rounding keep their original value, and the bucket hash remains unchanged. Fixes #18988. Test plan: reproduce wrong buckets and IN membership before the fix; cover inclusive/strict projections, trailing zeros, repeated conversion, and fractional bounds that must not round. Full API check: 1346 tests, 1 skipped, no failures/errors; formatting, production/test Checkstyle, and assembly pass. Decimal set lookup uses ordered comparison rather than scale-sensitive hashing. --- **AI Disclosure** - Model: GPT-6.1 (gpt-6.1-sol). - Platform/Tool: OpenAI Codex generated the implementation, regression tests, and this description. - Human Oversight: unreviewed. - Prompt Summary: fix decimal literal scale handling reported in #18988, preserving exact comparisons and bucket compatibility. -- 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]
