LuciferYang opened a new pull request, #9771:
URL: https://github.com/apache/paimon/pull/9771
### Purpose
close #9770
`StringUtils.toLowerCaseIfNeed`/`toLowerCase`/`toUpperCase` converted with
the JVM default locale; under a Turkish or Azeri default locale the
case-sensitive=false identifier matching (CDC table mapping, computed columns,
Arrow readers) silently broke for columns containing `I`/`i`.
Convert with `Locale.ROOT`, and align the record side of the same CDC flow:
`CdcRecord.fieldNameLowerCase` also used the default locale, so fixing only the
schema side would newly diverge the two halves of the record-schema join under
tr/az (previously both sides mangled identically and data still flowed).
### Tests
- `StringUtilsTest`: tr-TR default locale, `toLowerCaseIfNeed("INDEX")` →
`index` (RED on master: `ındex`), plus `toLowerCase`/`toUpperCase` assertions;
locale restored in finally.
- `CdcRecordTest`: `fieldNameLowerCase` maps `INDEX` → `index` and not
`ındex` under tr-TR.
### API and Format
For every locale except tr/az/lt, `Locale.ROOT` conversion is byte-identical
to the default-locale result, so existing behavior is unchanged. The
`upper()`/`lower()` computed columns now produce ROOT output under tr —
consistent with other SQL engines.
### Documentation
None.
--
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]
