pg.ddx.io  pgsql-sql@postgresql.org mailing list archive  
help / color / mirror / Atom feed
Primary Key Bugs
2+ messages / 2 participants
[nested] [flat]

* Primary Key Bugs
@ 1999-04-25 17:59  Tim Perdue <perdue@raccoon.com>
  0 siblings, 1 reply; 2+ messages in thread

From: Tim Perdue @ 1999-04-25 17:59 UTC (permalink / raw)
  To: pgsql-sql@hub.org

Somehow I have records in my database with duplicate primary keys. Because
of this, I'm not able to update a lot of records, because Postgres then
complains that I'm trying to insert a duplicate primary key.

Can anyone suggest a way to select the duplicate ids out of the table so I
can change them?

I can't really do a SELECT DISTINCT on a 3GB database table or my machine
would likely go down.

Basically, I just want a query like this:

select * into tbl_tmp from tbl_mail where mailid is duplicated

8-)

Any suggestions?

Tim Perdue
PHPBuilder.com / GotoCity.com / Geocrawler.com






^ permalink  raw  reply  [nested|flat] 2+ messages in thread

* Re: [SQL] Primary Key Bugs
@ 1999-04-26 12:44  José Soares <jose@sferacarta.com>
  parent: Tim Perdue <perdue@raccoon.com>
  0 siblings, 0 replies; 2+ messages in thread

From: José Soares @ 1999-04-26 12:44 UTC (permalink / raw)
  To: Tim Perdue <perdue@raccoon.com>; +Cc: pgsql-sql@hub.org



Tim Perdue ha scritto:

> Somehow I have records in my database with duplicate primary keys. Because
> of this, I'm not able to update a lot of records, because Postgres then
> complains that I'm trying to insert a duplicate primary key.
>
> Can anyone suggest a way to select the duplicate ids out of the table so I
> can change them?
>
> I can't really do a SELECT DISTINCT on a 3GB database table or my machine
> would likely go down.
>
> Basically, I just want a query like this:
>
> select * into tbl_tmp from tbl_mail where mailid is duplicated
>
> 8-)
>
> Any suggestions?
>
> Tim Perdue
> PHPBuilder.com / GotoCity.com / Geocrawler.com

The SQL expression to retrieve duplicate rows is:

select *  from tbl_mail where mailid in (
select mailid from tbl_mail group by mailid having count(mailid) > 1
);

But currently PostgreSQL have problems with having (see TODO):
        * subqueries containing HAVING return incorrect results

Therefore you have to query the tbl_mail in two times.

1) select mailid from tbl_mail group by mailid having count(mailid) > 1
2) select *  from tbl_mail where mailid = (replace with values returned at
point 1)

José






^ permalink  raw  reply  [nested|flat] 2+ messages in thread


end of thread, other threads:[~1999-04-26 12:44 UTC | newest]

Thread overview: 2+ messages (download: mbox mbox.gz follow: Atom feed)
-- links below jump to the message on this page --
1999-04-25 17:59 Primary Key Bugs Tim Perdue <perdue@raccoon.com>
1999-04-26 12:44 ` José Soares <jose@sferacarta.com>

This inbox is served by DDX for PostgreSQL; see mirroring instructions
for how to clone and mirror all data and code used for this inbox