Github user jinxing64 commented on the issue:
https://github.com/apache/spark/pull/21252
> Instead of touching inside of TakeOrderedAndProjectExec, how about we
don't replace Sort + Limit with TakeOrderedAndProjectExec when reaching the
threshold?
Yes, the code will be much cleaner. I updated the change.
Note that all data will still be sorted if above the threshold and all data
will be within one partition after the limit operator
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
