srielau commented on PR #58549: URL: https://github.com/apache/spark/pull/58549#issuecomment-5719121938
Scope-freeze follow-up is pushed in 5b7fd7adbc9. - Removed UDT-storage CHAR/VARCHAR assignment support and its support-only tests; UDT recursion now remains only for deterministic rejection. - Reverted Connect server local-relation schema-policy handling and the unused state conversion overload. - Spark Connect local `createDataFrame` now rejects recursive CHAR/VARCHAR schemas before the shared Arrow mapper can admit them. - Fixed default policy so classic `createDataFrame` performs CHAR/VARCHAR assignment checks while exposing STRING unless first-class types are enabled. - Added focused default-policy coverage for row, Arrow-optimized, native Arrow scalar/iterator, pandas scalar-iterator UDFs, and classic local/Arrow DataFrame creation. Validation: - `EvaluatePythonSuite`: 12 tests passed. - Focused `ArrowConvertersSuite` policy test passed. - Python `compileall` passed for all changed Python files. - PySpark runtime modules could not start locally because this checkout lacks the assembled SparkSubmit/Logging classpath; CI is independent and will run them. - SQL domain re-review found no remaining P1/P2/P3 issues. -- 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]
