From: Achilleas Mantzios <achill@matrix.gatewaynet.com>
Organization: Dynacom Tankers Mgmt
To: pgsql-sql@postgresql.org
Subject: Re: delete on cascade
Date: Mon, 23 Oct 2006 12:00:20 +0300
User-Agent: KMail/1.9.3
References: <200610231049.16507.fluca1978@infinito.it>
In-Reply-To: <200610231049.16507.fluca1978@infinito.it>
MIME-Version: 1.0
Content-Type: text/plain;
  charset="iso-8859-7"
Content-Transfer-Encoding: quoted-printable
Content-Disposition: inline
Message-Id: <200610231200.20932.achill@matrix.gatewaynet.com>

=D3=F4=E9=F2 =C4=E5=F5=F4=DD=F1=E1 23 =CF=EA=F4=FE=E2=F1=E9=EF=F2 2006 11:4=
9, =EF/=E7 Luca Ferrari =DD=E3=F1=E1=F8=E5:
> Hi all,
> I guess this is an already asked question, but I didn't found an answer, =
so
> apologize me. Imagine I've got two tables:
> skill(id,description)		// primary key =3D> id
> family(id,description)	// primary key =3D> id
> and I want to associate skills to families:
> ass_sf(id_skill,id_family)	// appropriate foreign keys
>
> Tables are already created and the database is running. Now I'd like to
> implement a delete cascade, thus when I delete a skill also its associati=
on
> with the family must be deleted. I gues I should have declared the skill
> table as follows:
> CREATE TABLE skill
> (
> 	id varchar(20) on delete cascade,
> 	description varchar(50),
> 	primary key(id)
> );
>
> right? The problem is: how can I alter the table to add the column

not right!!!
ON DELETE CASCADE is specified in FOREIGN KEY contsraints.
So that would be in ass_sf table.
If you find ALTER TABLE ass_sf command hard to run, then drob your=20
ass_sf table and define it like

CREATE TABLE ass_sf (
=2E....
skill_id INT REFERENCES skill(id) ON DELETE CASCADE,
familly_id INT REFERENCES family(id) ON DELETE CASCADE
);

The above causes cascade deletes not only on skills but on families also.

> constraint now, without redeclaring the table?
> Thanks very much for helping me.
> Luca
>
> ---------------------------(end of broadcast)---------------------------
> TIP 9: In versions below 8.0, the planner will ignore your desire to
>        choose an index scan if your joining column's datatypes do not
>        match

=2D-=20
Achilleas Mantzios