Dag H. Wanvik created DERBY-5957:
------------------------------------
Summary: 0 returned instead of NULL with correlated scalar
subquery in select list
Key: DERBY-5957
URL: https://issues.apache.org/jira/browse/DERBY-5957
Project: Derby
Issue Type: Bug
Components: SQL
Affects Versions: 10.9.1.0
Reporter: Dag H. Wanvik
In DERBY-5954 we saw an example query that returned 0 instead of NULL if a
correlated scalar subquery in a select list didn't find a row, and another
column in the select list has a values, i.e.:
s.executeUpdate("create table derby5954( a int )");
s.executeUpdate("insert into derby5954 values 1,2,3,4,5,6,7");
ResultSet rs = s.executeQuery(
"SELECT a_t1, (" +
" SELECT a_t2 FROM (" +
" SELECT a as a_t2 FROM derby5954) " +
" as T2" +
" where T2.a_t2 = T1.a_t1 + 1 ) " +
" as a_outer" +
" FROM (SELECT a as a_t1 from derby5954) as T1");
gives
1 2
2 3
3 4
4 5
5 6
6 7
7 0
where the last row's second column should be NULL, not "0"., since in the
predicate T2.a_t2 = T1.a_t1 + 1
there will be no T2.a_t2 which matches 7 (T1.a_t1's last row) + 1.
--
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
