[jira] [Updated] (TAJO-186) Improve column resolving method

["Hyunsik Choi (JIRA)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/TAJO-186?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Hyunsik Choi updated TAJO-186:
------------------------------

    Summary: Improve column resolving method  (was: Improve column resolve method)
    
> Improve column resolving method
> -------------------------------
>
>                 Key: TAJO-186
>                 URL: https://issues.apache.org/jira/browse/TAJO-186
>             Project: Tajo
>          Issue Type: Bug
>          Components: planner/optimizer
>            Reporter: Hyunsik Choi
>            Assignee: Hyunsik Choi
>             Fix For: 0.2-incubating
>
>
> The current column resolving method (LogicalPlanner::createFieldEval) usually finds columns from relations in the current query block. If not found, it tries to find columns from other query blocks.
> However, this way cannot resolve some cases. Please take a look at the below example:
> {noformat}
> select col1 as rank from table1 order by rank
> {noformat}
> In the above case, the column 'rank' cannot be found in any relation. So, in this case, we currently use LogicalPlan::findColumnFromChildNode method to resolve columns from a child node.
> But, this way cannot resolve a column 'total' in the below cases:
> {noformat}
> select ..., sum(col1) as total ... having total > 1
> {noformat}
> Resolving columns should also consider the current node's output schema.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira