Re: [HACKERS] Insert into view - Mailing list pgsql-hackers

From Theo Kramer
Subject Re: [HACKERS] Insert into view
Date
Msg-id 3843EE74.AA2E0AB2@flame.co.za
Whole thread Raw
In response to RE: [HACKERS] Insert into view  ("Ansley, Michael" <Michael.Ansley@intec.co.za>)
List pgsql-hackers
"Ansley, Michael" wrote:
> 
> I think this was covered a little while back, but it runs something like
> this:  a view is a relation, with a select rule (which is the view query).
> When you insert into the view (which, like I said, is just another relation,
> it actually inserts into the view relation.  However, when you select from
> it, of course, the select rule fires, and you don't see any of the
> information.  I suppose you could set up a nice insert rule to insert into
> the base tables of the query if you wanted.  I normally do this through
> stored procs, but this would be essentially the same thing, just nicer
> client-side SQL.

Hmm, interesting.

> I suppose that views could be made so that a tuple insert would fail, but
> you should know your db better ;-)

That I do, just thought a message might assist those that don't :)

--------
Regards
Theo


pgsql-hackers by date:

Previous
From: sk.list@comset.net
Date:
Subject: Re: [ADMIN] When postgres will be faster?
Next
From: Oleg Bartunov
Date:
Subject: Re: [ADMIN] When postgres will be faster?