[
https://issues.apache.org/jira/browse/SPARK-32051?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17146722#comment-17146722
]
Jungtaek Lim edited comment on SPARK-32051 at 6/27/20, 1:42 AM:
----------------------------------------------------------------
OK so it needs to explicitly return Unit. Thanks for the information.
[~srowen] Do you have some idea around this? My guess is that Scala 2.12
enables leveraging functional interface into lambda, and that leads confusions.
I haven't played with Java side, so I'm not sure removing one of two methods
would work, and even it works it may lead backward incompatibility. I have no
idea with Scala 2.13 so a bit afraid to try out some changes.
was (Author: kabhwan):
OK so it needs to explicitly return Unit. Thanks for the information.
[~srowen] Do you have some idea around this? My guess is that Scala 2.12
enables leveraging functional interface into lambda, and that leads confusions.
I haven't played with Java side, so I'm not sure removing one of two methods
would work, and even it works it may lead backward incompatibility.
> Dataset.foreachPartition returns object
> ---------------------------------------
>
> Key: SPARK-32051
> URL: https://issues.apache.org/jira/browse/SPARK-32051
> Project: Spark
> Issue Type: Bug
> Components: SQL
> Affects Versions: 3.0.0
> Reporter: Frank Oosterhuis
> Priority: Major
>
> I'm trying to map values from the Dataset[Row], but since 3.0.0 this fails.
> In 3.0.0 I'm dealing with an error: "Error:(28, 38) value map is not a member
> of Object"
>
> This is the simplest code that works in 2.4.x, but fails in 3.0.0:
> {code:scala}
> spark.range(100)
> .repartition(10)
> .foreachPartition(part => println(part.toList))
> {code}
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
