[
https://issues.apache.org/jira/browse/SPARK-29667?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16987712#comment-16987712
]
Marco Gaido commented on SPARK-29667:
-------------------------------------
I can agree more with you [~hyukjin.kwon]. I think that having different
coercion rules for the two types of IN is very confusing. It'd be great for
such things to be consistent among all the framework in order to avoid
"surprises" for users IMHO.
> implicitly convert mismatched datatypes on right side of "IN" operator
> ----------------------------------------------------------------------
>
> Key: SPARK-29667
> URL: https://issues.apache.org/jira/browse/SPARK-29667
> Project: Spark
> Issue Type: Improvement
> Components: SQL
> Affects Versions: 2.4.3
> Reporter: Jessie Lin
> Priority: Minor
>
> Ran into error on this sql
> Mismatched columns:
> {code}
> [(a.`id`:decimal(28,0), db1.table1.`id`:decimal(18,0))]
> {code}
> the sql and clause
> {code}
> AND a.id in (select id from db1.table1 where col1 = 1 group by id)
> {code}
> Once I cast {{decimal(18,0)}} to {{decimal(28,0)}} explicitly above, the sql
> ran just fine. Can the sql engine cast implicitly in this case?
>
>
>
>
--
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
