cshuo opened a new issue, #19183: URL: https://github.com/apache/hudi/issues/19183
### Background With the semantics of `RecordContext#convertValueToEngineType` clarified (it now returns the raw engine-native type that can be stored directly as a field value inside an engine record, e.g. a raw `UTF8String` for Spark; callers that need a comparable value go through `convertOrderingValueToEngineType` / `ensureComparability`, which wrap types that no longer implement `compareTo`, see SPARK-46832), `RecordContext#convertPartitionValueToEngineType` becomes redundant: - The base-class default implementation simply forwards to `convertValueToEngineType`. - The only override, `BaseSparkInternalRecordContext#convertPartitionValueToEngineType` (`String` -> `UTF8String.fromString`), now behaves identically to the Spark implementation of `convertValueToEngineType`. ### Proposed cleanup - Remove `convertPartitionValueToEngineType` from `RecordContext` and its override in `BaseSparkInternalRecordContext`. - Update its only call site, `HoodieFileGroupReader` (building partition filter fields/values), to call `convertValueToEngineType` directly. This is a follow-up cleanup to the native log format work (see #19118). -- 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]
