Re: removing duplicated constraints

On 2 Mar 2002 at 11:09, Stephan Szabo wrote:

> On Sat, 2 Mar 2002, Dan Langille wrote:
>
> > On 2 Mar 2002 at 10:19, Dan Langille wrote:
> >
> > > CREATE CONSTRAINT TRIGGER "<unnamed>" AFTER INSERT OR UPDATE ON "ports"
> > > FROM "categories" NOT DEFERRABLE INITIALLY IMMEDIATE FOR EACH ROW
> > > EXECUTE PROCEDURE "RI_FKey_check_ins" ('<unnamed>', 'ports',
> > > 'categories', 'UNSPECIFIED', 'category_id', 'id');
> > >
> > > CREATE CONSTRAINT TRIGGER "<unnamed>" AFTER INSERT OR UPDATE ON "ports"
> > > FROM "categories" NOT DEFERRABLE INITIALLY IMMEDIATE FOR EACH ROW
> > > EXECUTE PROCEDURE "RI_FKey_check_ins" ('<unnamed>', 'ports',
> > > 'categories', 'UNSPECIFIED', 'category_id', 'id');
> >
> > How is an unnamed trigger deleted?  During a dump, I can find information
> > such as this:
>
> The "<unnamed>" is the constraint name given.  It also gets a trigger name
> (see system table pg_trigger) which you can use with DROP TRIGGER (although
> in the case of the fk constraint triggers you need to double quote the
> name).

I found the list of duplicate constraints by eyeballing the files I
created using the methods mentioned in previous post (mainly grep and
sort).  Then I would search in the original dump file for the the CREATE
CONSTRAINT statement.  Using the name found there
(RI_ConstraintTrigger_998199), I did this:

  drop TRIGGER "RI_ConstraintTrigger_998189" on categories;

Making sense?
--
Dan Langille
The FreeBSD Diary - http://freebsddiary.org/ - practical examples