[
https://issues.apache.org/jira/browse/HIVE-20084?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17627980#comment-17627980
]
Smruti Biswal commented on HIVE-20084:
--------------------------------------
In Hive 3.x you may set {{{}orc.force.positional.evolution=false{}}}, which
will map the columns based on the names instead of position.
> is there any setting to read data based on column names for ORC hive table
> like parquet instead of ordinal position in hive
> ---------------------------------------------------------------------------------------------------------------------------
>
> Key: HIVE-20084
> URL: https://issues.apache.org/jira/browse/HIVE-20084
> Project: Hive
> Issue Type: Task
> Components: Hive
> Affects Versions: 2.1.0
> Reporter: Hazarath
> Priority: Minor
>
> is there any setting to read data based on column names for ORC format like
> parquet instead of ordinal position in hive. I see below settings for Presto
> & Spark. But i dont find any setting in hive for same functionality. If it's
> already there, please route me into it.
>
> Spark – spark.sql.hive.convertMetastoreOrc=true
> Presto – hive.orc.use-column-names=true
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
