davidzollo opened a new pull request, #11928: URL: https://github.com/apache/seatunnel/pull/11928
### Purpose of this pull request Replacement for #10943 on an Apache-owned maintainer branch so current-head CI can run from the main repository. This keeps the original fix intent from Asish Kumar's commit `824dac5619d818e88e193d2f4bc860ec4a3b2795` and reconciles it with the current `dev` branch: - normalize configured INT96 field names before they are stored in `writePathsAsInt96` - normalize mixed-case row field names before record writes - normalize nested row field names the same way as the Parquet schema - preserve the current `dev` branch's safe row-field access path With these changes, mixed-case fields such as `createTime` and configured fixed-byte fields such as `F3_Bytes` resolve against the normalized Parquet schema consistently when INT96 encoding is enabled. ### Why this replacement PR exists The original contributor PR #10943 is still open, but this release batch needs a maintainer-owned branch in `apache/seatunnel` instead of pushing follow-up commits into the contributor fork. ### Does this PR introduce _any_ user-facing change? Yes. Parquet sinks can now write INT96 timestamp and fixed-byte fields whose SeaTunnel column names contain uppercase letters without requiring lower-case aliases. ### How was this patch tested? Per the current maintainer validation rule for SeaTunnel in this workflow, local validation was limited to formatting and diff hygiene: - `./mvnw spotless:apply -pl seatunnel-connectors-v2/connector-file/connector-file-base -am -nsu` - `git diff --check` GitHub CI on this maintainer-owned branch is the source of truth for compile, test, and runtime validation. -- 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]
