Rahul Challapalli created DRILL-2228:
----------------------------------------

             Summary: Projecting '*' returns all nulls when we have flatten in 
a filter and order by
                 Key: DRILL-2228
                 URL: https://issues.apache.org/jira/browse/DRILL-2228
             Project: Apache Drill
          Issue Type: Bug
          Components: Functions - Drill
            Reporter: Rahul Challapalli
            Assignee: Jason Altekruse


git.commit.id.abbrev=3d863b5

The below query returns all nulls currently :
{code}
0: jdbc:drill:schema=dfs_eea> select * from `data.json` where 2 in (select 
flatten(lst_lst[1]) from `data.json`) order by flatten(lst_lst[1]);
+------------+
|     *      |
+------------+
| null       |
| null       |
| null       |
| null       |
| null       |
| null       |
| null       |
| null       |
| null       |
| null       |
+------------+
{code}

There seems to be another issue here since the no of records returned also does 
not look right. I will raise a separate JIRA for that.


The issue goes away, if we do an order by without the flatten. Below query works
{code}
select * from `data.json` where 2 in (select flatten(lst_lst[1]) from 
`data.json`) order by uid;
{code}

Attached the data files



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

Reply via email to