[
https://issues.apache.org/jira/browse/SPARK-53873?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Andy Sautins updated SPARK-53873:
---------------------------------
External issue URL: https://github.com/apache/spark/pull/52528
> ExplodeBase.eval Iterate directly on input
> ------------------------------------------
>
> Key: SPARK-53873
> URL: https://issues.apache.org/jira/browse/SPARK-53873
> Project: Spark
> Issue Type: Improvement
> Components: SQL
> Affects Versions: 4.0.1
> Reporter: Andy Sautins
> Priority: Trivial
>
> It was noticed that `ExplodeBase.eval` returns an IterableOnce[InternalRow].
> The current implementation creates a pre-allocated array, populates the array
> appropriately, and returns the Array. This works as the is an implicit
> conversion from Array to IterableOnce.
> However Allocating and populating an array does not seem to provide benefits
> over exposing an iterator over the input data type.
> A proposed PR removes the creation and population of the array and instead
> returns a IterableOnce object that iterates over the underlying input.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
