pg.ddx.io  pgsql-sql@postgresql.org mailing list archive  
help / color / mirror / Atom feed
From: dipesh <dipesh.mistry@mobilefundas.com>
To: felix@crowfix.com
Cc: pgsql-sql@postgresql.org
Subject: Re: Delete with foreign keys
Date: Sat, 23 Feb 2008 12:42:13 +0530
Message-ID: <47BFC74D.9090200@mobilefundas.com> (raw)
In-Reply-To: <20080222180021.GA7700@crowfix.com>
References: <20080222180021.GA7700@crowfix.com>

felix@crowfix.com wrote:
> I have three tables --
>
>     CREATE TABLE name (id INT PRIMARY KEY, str VARCHAR(20));
>     CREATE TABLE place (id INT PRIMARY KEY, name_id INT REFERENCES name(id));
>     CREATE TABLE data (id INT PRIMARY KEY, place_id INT REFERENCES place(id));
>
> I want to delete all place and data rows which reference specific
> names, but not the names themselves.  I can do it like this:
>
>     DELETE FROM data d WHERE exists (SELECT id from place p WHERE d.place_id = p.id AND p.name_id IN (SELECT id FROM name WHERE str LIKE 'Fredonia%'));
>
>     DELETE FROM place WHERE name_id IN (SELECT id FROM name WHERE str LIKE 'Fredonia%');
>
> but it seems rather roundabout, and I wonder whether the EXISTS and IN
> business is slow.  Is there some way to do it using JOINs?  I think of
> something like this:
>
>     DELETE FROM place p, name n WHERE p.name_id = n.id AND n.str LIKE 'Fredonia%';
>
> but I don't want to delete the name rows.  Then I think of this:
>
>     DELETE FROM place p WHERE p.name_id = name.id AND name.str LIKE 'Fredonia%';
>
> but I feel uneasy about the two separate name references when the
> table is not named in the FROM clause.  Maybe that's just my novicity.
>
> I also wonder about getting fancy and ending up with SQL specific to a
> database; I don't have any plans to migrate, but I try to avoid
> branding my SQL.
>
>   
Hello, I am Dipesh Mistry from Ahmedabad-India.
I have the solution for you.
First drop the constraint on table data.
Example.
    alter table data drop constraint data_place_id_fkey;
And in second step add new Constraint.
Example.
    alter table data add constraint data_place_id_fkey FOREIGN 
KEY(place_id) REFERENCES place(id) ON DELETE CASCADE ON UPDATE CASCADE;

Now when you test query like,

delete from place where name_id in(select id from name where str='dip%');

So, you can try this above step.

-- 
With Warm Regards,
Dipesh Mistry
Information Technology Dept.
GaneshaSpeaks.com




view thread (3+ messages)  latest in thread

Message-ID: <47BFC74D.9090200@mobilefundas.com>
Permalink:  ../47BFC74D.9090200@mobilefundas.com/
Also on:    postgresql.org/message-id/47BFC74D.9090200@mobilefundas.com

 · 

reply

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Reply to all the recipients using the --to and --cc options:
  reply via email

  To: pgsql-sql@postgresql.org
  Cc: dipesh.mistry@mobilefundas.com, felix@crowfix.com
  Subject: Re: Delete with foreign keys
  In-Reply-To: <47BFC74D.9090200@mobilefundas.com>

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

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