[
https://issues.apache.org/jira/browse/HIVE-11080?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14601896#comment-14601896
]
Matt McCline commented on HIVE-11080:
-------------------------------------
The only place I remember it being used was in the Tez's ReduceRecordSource for
error reporting.
{code}
/* batch.toString depends on this */
batch.setValueWriters(valueStringWriters
.toArray(new VectorExpressionWriter[0]));
rowString = batch.toString();
{code}
So, it is fine if we have toString just use the vector type instead of Hive
types. Or, it is also fine if we eliminate this usage all together, too.
> Modify VectorizedRowBatch.toString() to not depend on VectorExpressionWriter
> ----------------------------------------------------------------------------
>
> Key: HIVE-11080
> URL: https://issues.apache.org/jira/browse/HIVE-11080
> Project: Hive
> Issue Type: Sub-task
> Reporter: Owen O'Malley
> Assignee: Owen O'Malley
> Attachments: HIVE-11080.patch
>
>
> Currently the VectorizedRowBatch.toString method uses the
> VectorExpressionWriter to convert the row batch to a string.
> Since the string is only used for printing error messages, I'd propose making
> the toString use the types of the vector batch instead of the object
> inspector.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
