[ 
https://issues.apache.org/jira/browse/METAMODEL-128?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14387268#comment-14387268
 ] 

ASF GitHub Bot commented on METAMODEL-128:
------------------------------------------

Github user kaspersorensen commented on a diff in the pull request:

    https://github.com/apache/metamodel/pull/13#discussion_r27423309
  
    --- Diff: 
core/src/main/java/org/apache/metamodel/query/parser/SelectItemParser.java ---
    @@ -160,6 +164,14 @@ public SelectItem findSelectItem(String expression) 
throws MultipleSelectItemsPa
                     return new SelectItem(subQuerySelectItem, fromItem);
                 }
             }
    +        
    +        //if the expression is alias of some select item defined return 
clone of that select item
    +        List<SelectItem> allSelectItems = 
MetaModelHelper.getAllSelectItems(_query);
    --- End diff --
    
    Also reflecting a bit on this ... Do we need to have this 
getAllSelectItems(...) method?
    
    What I mean is, could there be aliased select items anywhere else than in 
the SELECT clause? I cannot imagine it really ... For instance in WHERE clause 
I have never seen anything a la:
    
    SELECT ... FROM ... WHERE name AS n = "Kasper"
    
    If we assume that this never happens then you can get rid of the 
getAllSelectItems(...) method which is already a bit greedy (makes new lists 
etc.). If I am wrong about the assumption (I seriously am in doubt) then I 
think the solution is good!


> Ordering on aggregation value not working
> -----------------------------------------
>
>                 Key: METAMODEL-128
>                 URL: https://issues.apache.org/jira/browse/METAMODEL-128
>             Project: Apache MetaModel
>          Issue Type: Bug
>    Affects Versions: 4.3.2
>            Reporter: Ashish Mukherjee
>
> I am using a CSV Data Context example with multiple fields - occurrences, 
> first_name, last_name. 
> The query -
> SELECT SUM(persons.csv.occurrences) AS X, persons.csv.first_name FROM 
> ashish.persons.csv GROUP BY persons.csv.first_name ORDER BY X
> This query always orders the results by first_name and not the aggregation 
> value.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Reply via email to