thswlsqls opened a new issue, #8626: URL: https://github.com/apache/paimon/issues/8626
**Search before asking** - [x] I searched in the [issues](https://github.com/apache/paimon/issues) and found nothing similar. **Paimon version** master @ 4fe3cfdd1 / 2.0-SNAPSHOT **Compute Engine** Spark **Minimal reproduce step** Query an ORC table with an `isNaN` filter, e.g. `SELECT * FROM t WHERE c = float('nan')`. `SparkFilterConverter` builds `PredicateBuilder.isNaN(index)`, which is pushed to the ORC reader. `OrcPredicateFunctionVisitor` (paimon-format/.../orc/filter/OrcPredicateFunctionVisitor.java) does not override `visitIsNaN`, so the `FunctionVisitor.visitIsNaN` default runs and throws `UnsupportedOperationException`. `OrcFileFormat.createReaderFactory()` calls `pred.visit(VISITOR)` with no try/catch, so reader creation fails and the query errors out. **What doesn't meet your expectations?** Expected: the `isNaN` filter is simply not pushed down (engine evaluates it), as the Parquet path already does — `ParquetFilters.convert` catches `UnsupportedOperationException` and skips push-down. Actual: the ORC path throws and the query fails. **Anything else?** Fix: override `visitIsNaN` to `return Optional.empty()`, matching the sibling unsupported ops (`visitStartsWith`/`visitEndsWith`/`visitContains`/`visitLike`) in the same class. ORC SearchArgument has no isNaN leaf. **Are you willing to submit a PR?** - [x] I'm willing to submit a PR! -- 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]
