chasingegg commented on pull request #35168:
URL: https://github.com/apache/spark/pull/35168#issuecomment-1014070459
@cloud-fan Hello, Besides the PushProjectionThroughUnion as I explained
above, I guess the code below causes this bug,
```
private def collectFromPlan(plan: SparkPlan): Array[T] = {
val fromRow = resolvedEnc.createDeserializer()
plan.executeCollect().map(fromRow)
}
```
This `fromRow` method would bindReference for deserialization, in the second
child of Union, it will take the first column twice as the result because the
first child of Union has duplicate columns, and the first child determines the
final union result's attributes.
If you have time, could you please take a look, thx~
--
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]
