spark-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Swapnil Chougule (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (SPARK-25334) Default SessionCatalog should support UDFs
Date Thu, 27 Sep 2018 18:51:00 GMT

    [ https://issues.apache.org/jira/browse/SPARK-25334?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16630890#comment-16630890
] 

Swapnil Chougule commented on SPARK-25334:
------------------------------------------

+1

My work has also been impacted by same.
SessionCatalog calls registerFunction to add a function to function registry. However, makeFunctionExpression
supports only UserDefinedAggregateFunction.

 

Can we have handler for scala udf, java udf & java udaf as well in SessionCatalog?

> Default SessionCatalog should support UDFs
> ------------------------------------------
>
>                 Key: SPARK-25334
>                 URL: https://issues.apache.org/jira/browse/SPARK-25334
>             Project: Spark
>          Issue Type: Task
>          Components: SQL
>    Affects Versions: 2.3.1
>            Reporter: Tomasz Gawęda
>            Priority: Major
>
> SessionCatalog calls registerFunction to add a function to function registry. However,
makeFunctionExpression supports only UserDefinedAggregateFunction.
> We should make makeFunctionExpression support UserDefinedFunction, as it's one of functions
type.
> Currently we can use persistent functions only with Hive metastore, but "create function"
command works also with default SessionCatalog. It sometimes cause user confusion, like in
https://stackoverflow.com/questions/52164488/spark-hive-udf-no-handler-for-udaf-analysis-exception/52170519



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@spark.apache.org
For additional commands, e-mail: issues-help@spark.apache.org


Mime
View raw message