List:General Discussion« Previous MessageNext Message »
From:Matt Neimeyer Date:July 14 2009 8:00pm
Subject:Re: Removing Duplicate Records
View as plain text  
Ah... Yes. Good point. I like this because I was planning on keeping
the output somewhere for a while. (In case we need an "accounting" at
some point) So it will be easy enough to dump what's being deleted to
the screen while we loop over our candidates.


On Tue, Jul 14, 2009 at 10:16 AM, Nathan Sullivan<nsullivan@stripped> wrote:
> If you went with option #3, you could avoid the looping by using (tCount - 1) as the
> LIMIT in
> the delete statement instead of always using 1.

>> 3. Do... "SELECT count(*) AS tCount,OrgID,ContID FROM OrgContLink
>> GROUP BY OrgID,ContID HAVING tCount > 1". Then for every record in the
>> result "DELETE FROM OrgContLink WHERE OrgID=X AND ContID=Y LIMIT 1".
>> Then repeat until no results are found.
Removing Duplicate RecordsMatt Neimeyer14 Jul
  • RE: Removing Duplicate RecordsNathan Sullivan14 Jul
    • Re: Removing Duplicate RecordsMatt Neimeyer14 Jul
  • Re: Removing Duplicate RecordsMarcus Bointon14 Jul
    • Re: Removing Duplicate RecordsBrent Baisley14 Jul