[
https://issues.apache.org/jira/browse/DERBY-4442?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12780601#action_12780601
]
Knut Anders Hatlen commented on DERBY-4442:
-------------------------------------------
Thanks, Dag. That's good news. I ran derbyall and suites.All with
always_prn.diff, and only one test failed:
DistinctTest.testDistinctInsertWithGeneratedColumn(). That test case is
intended to fail if DERBY-3 is fixed, according to its comments, so that should
be fine:
* See DERBY-3. If that bug is fixed, the first query after the comment
* below will fail.
I'll try to clean up the patch and get it into a committable shape.
> Evaluation of default value and identity in an INSERT result set evaluated
> too early.
> -------------------------------------------------------------------------------------
>
> Key: DERBY-4442
> URL: https://issues.apache.org/jira/browse/DERBY-4442
> Project: Derby
> Issue Type: Bug
> Components: SQL
> Reporter: Dag H. Wanvik
> Attachments: always_prn.diff, insert.diff
>
>
> In contrast to generated column, which are evaluated when the next row from
> the result set to be inserted, currently default values and identity columns
> are generated "early", that is as part of avaluating the subquery (SELECT or
> VALUES as the case may be).
> This does not currently cause a user visible bug in Derby, but it lies behind
> DERBY-3 and the effect Bryan observed in DERBY-4.
> Additionally, "early" computation has given rise to much special handling and
> ensuing bugs, cf. DERBY-1644, DERBY-4413, DERBY-4419, DERBY-4425 and others.
> DERBY-4397 requires this fix for correct behaviour with INSERT.
> See also
> https://issues.apache.org/jira/browse/DERBY-4413?focusedCommentId=12769532&page=com.atlassian.jira.plugin.system.issuetabpanels%3Acomment-tabpanel#action_12769532
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
