[
https://issues.apache.org/jira/browse/HIVE-568?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Bill Graham updated HIVE-568:
-----------------------------
Attachment: HIVE-568.patch
Please code review this patch for validity. It includes a fix, as well as a
unit test that fails without the fix.
In the case where a tableDesc isn't set on the FetchTask object, I look for a
partDesc. If one is found, I use the tableDesc from the first partDesc to get
the deserializer. This seems like a reasonable thing to do, but please review
if this logic is valid and won't cause other issues.
> Driver.getSchema() throws an Exception with partitioned tables
> --------------------------------------------------------------
>
> Key: HIVE-568
> URL: https://issues.apache.org/jira/browse/HIVE-568
> Project: Hadoop Hive
> Issue Type: Bug
> Components: Clients
> Reporter: Bill Graham
> Attachments: HIVE-568.patch
>
> Original Estimate: 2h
> Remaining Estimate: 2h
>
> When doing a select all on a partitioned table, the Driver.getSchema() method
> throws an NPE because tableDesc isn't set on the FetchTask object.
> Partitioned tables have partDesc set instead of tableDesc.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
