[
https://issues.apache.org/jira/browse/TAJO-1900?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14948320#comment-14948320
]
ASF GitHub Bot commented on TAJO-1900:
--------------------------------------
Github user jihoonson commented on the pull request:
https://github.com/apache/tajo/pull/792#issuecomment-146462466
I've verified with 'mvn clean install' with jdk7.
> When a record column and its child column are retrieved together, the record
> column might not be inferred as record type properly
> ---------------------------------------------------------------------------------------------------------------------------------
>
> Key: TAJO-1900
> URL: https://issues.apache.org/jira/browse/TAJO-1900
> Project: Tajo
> Issue Type: Bug
> Components: Planner/Optimizer
> Reporter: Jihoon Son
> Assignee: Jihoon Son
> Attachments: TAJO-1900.patch
>
>
> For example, given a following query, the column 'glossary' can be inferred
> as the TEXT type rather than the RECORD type.
> {noformat}
> default> select glossary from self_desc_table2 where
> char_length(glossary."GlossDiv".title) > 0
> {noformat}
> This is because columns of the query are inferred in an arbitrary order, and
> there isn't any routine to resolve the conflict of inferred data type.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
