Github user icexelloss commented on the issue:
https://github.com/apache/spark/pull/22104
I think another way to fix this is to move the logic to `ExtractPythonUDF`
to ignore `FileScanExec` `DataSourceScanExec` and `DataSourceV2ScanExec`
instead of changing all three rules. The downside is that if a XScanExec node
with a Python UDF pushed filter throws exception somewhere else, we need to fix
that too. Not sure which way is better. But either way, it would be good to
create test case with data source and data source V2... Would appreciate some
advise on how to create such relation in test
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org
