nsivabalan commented on issue #17043:
URL: https://github.com/apache/hudi/issues/17043#issuecomment-4800737348
This appears to have been fixed on master by [HUDI-9468] (commit
e040dfdd6b84, PR #13365, merged 2025-06-28), which superseded the originally
linked PR #13397.
In current master, `HoodieDatasetBulkInsertHelper.bulkInsert` removed the
`WriteOperationType` switch and now unconditionally aligns nullability against
the table/Hoodie schema for every bulk-insert-as-Row path (see
`hudi-client/hudi-spark-client/src/main/scala/org/apache/hudi/HoodieDatasetBulkInsertHelper.scala:159`):
```scala
val schema = HoodieSchemaConversionUtils.alignFieldsNullability(
dataset.schema, HoodieSchema.parse(writeConfig.getSchema))
```
The helper (now `HoodieSchemaConversionUtils.alignFieldsNullability`) also
recursively aligns nullability for nested struct/array/map fields, which is
broader than the original PR #13397 proposed.
If you're still hitting a nullable-schema inconsistency on the latest master
under a specific scenario (writer config / partitioner / nested types), could
you share a minimal repro? Otherwise this can likely be closed as resolved by
HUDI-9468.
--
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]
