Re: select/insert/delete

From: Robert Freeman <robertgfreeman_at_yahoo.com>
Date: Wed, 8 Jun 2011 14:59:12 -0700 (PDT)
Message-ID: <205448.97247.qm_at_web113202.mail.gq1.yahoo.com>

See my earlier response. You can build a function that you call in the DELETE that will remove the rows in question.

 Robert G. Freeman
Master Principal Consultant, Oracle Corporation, Oracle ACE Author of various books on RMAN, New Features and this shorter signature line. Blog: http://robertgfreeman.blogspot.com

Note: THIS EMAIL IS NOT AN OFFICIAL ORACLE SUPPORT COMMUNICATION. It is just the opinion of one Oracle employee. I can be wrong, have been wrong in the past and will be wrong in the future. If your problem is a critical production problem, you should always contact Oracle support for assistance. Statements in this email in no way represent Oracle Corporation or any subsidiaries and reflect only the opinion of the author of this email.

---

From: Adam Musch <ahmusch_at_gmail.com>
To: knecht.stefan_at_gmail.com
Cc: michaeljmoore_at_gmail.com; ORACLE-L <oracle-l_at_freelists.org> Sent: Wed, June 8, 2011 12:33:39 PM
Subject: Re: select/insert/delete

The WHEN MATCHED THEN UPDATE ... DELETE only permits deleting from the target table, and I don't think it can you can build a key-preserved view as the target of the join, especially when the target table doesn't have the matching rows.

After reading the OP's most recent post post, I suspect that the only practical choices are to:

• Build a PL/SQL collections of records for manipulation, validation and migration:
• Use a global temporary table as a staging where one does as much manipulation and validation at the SQL level. This should mitigate some of the concurrency issues, as you're working with a copy of the data.
• Exploit the LOG ERRORS DML clause to capture the records which fail referential integrity validation. If you're not using RI, well, that solution will be less effective.

On Wed, Jun 8, 2011 at 3:13 AM, Stefan Knecht <knecht.stefan_at_gmail.com> wrote:
> I saw merge was suggested, but without mentioning that merge actually can do
> deletes as well in recent versions...
> I haven't tested it, but what if you merge into a view that joins you
> table_a and table_b together. Then it might fulfill your 3 statements in 1
> go.
> Stefan
>

-- 
Adam Musch
ahmusch_at_gmail.com
--
http://www.freelists.org/webpage/oracle-l
--
http://www.freelists.org/webpage/oracle-l

Received on Wed Jun 08 2011 - 16:59:12 CDT