You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@drill.apache.org by "Victoria Markman (JIRA)" <ji...@apache.org> on 2015/01/22 00:31:35 UTC

[jira] [Created] (DRILL-2053) Column names are case sensitive if column is coming from WITH clause

Victoria Markman created DRILL-2053:
---------------------------------------

             Summary: Column names are case sensitive if column is coming from WITH clause
                 Key: DRILL-2053
                 URL: https://issues.apache.org/jira/browse/DRILL-2053
             Project: Apache Drill
          Issue Type: Bug
          Components: Query Planning & Optimization
    Affects Versions: 0.8.0
            Reporter: Victoria Markman
            Assignee: Jinfeng Ni


test.json
{code}
{"customerid":100,"customername":"AAA"}
{"customerid":101,"customername":"BBB"}
{"customerid":102,"customername":"CCC"}
{code}

Wrong result:
{code}
0: jdbc:drill:schema=dfs> with a as ( select * from `test.json` ) select * from a, `test.json` b where a.CUSTOMERID = b.CUSTOMERID;
+------------+--------------+-------------+---------------+
| customerid | customername | customerid0 | customername0 |
+------------+--------------+-------------+---------------+
+------------+--------------+-------------+---------------+
No rows selected (0.202 seconds)
{code}

Correct result, when column name matches the case of the column name in the json file:
{code}
0: jdbc:drill:schema=dfs> with a as ( select * from `test.json` ) select * from a, `test.json` b where a.customerid = b.customerid;
+------------+--------------+-------------+---------------+
| customerid | customername | customerid0 | customername0 |
+------------+--------------+-------------+---------------+
| 100        | AAA          | 100         | AAA           |
| 101        | BBB          | 101         | BBB           |
| 102        | CCC          | 102         | CCC           |
+------------+--------------+-------------+---------------+
3 rows selected (0.204 seconds)
{code}

Correct result when column does not match case, but is coming directly from the table:
{code}
0: jdbc:drill:schema=dfs> with a as ( select * from `test.json` ) select * from a, `test.json` b where a.customerid = b.customerID;
+------------+--------------+-------------+---------------+
| customerid | customername | customerid0 | customername0 |
+------------+--------------+-------------+---------------+
| 100        | AAA          | 100         | AAA           |
| 101        | BBB          | 101         | BBB           |
| 102        | CCC          | 102         | CCC           |
+------------+--------------+-------------+---------------+
3 rows selected (0.197 seconds)
{code}

If you change case of a column name that comes from subquery (WITH clause), this is where it goes all wrong:
{code}
0: jdbc:drill:schema=dfs> with a as ( select * from `test.json` ) select * from a, `test.json` b where a.Customerid = b.customerid;
+------------+--------------+-------------+---------------+
| customerid | customername | customerid0 | customername0 |
+------------+--------------+-------------+---------------+
+------------+--------------+-------------+---------------+
No rows selected (0.186 seconds)
{code}




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