Re: set constraints behavior - Mailing list pgsql-hackers

From Stephan Szabo
Subject Re: set constraints behavior
Date
Msg-id 20020503103842.P71279-100000@megazone23.bigpanda.com
Whole thread Raw
In response to set constraints behavior  (Neil Conway <nconway@klamath.dyndns.org>)
Responses Re: set constraints behavior  (Neil Conway <nconway@klamath.dyndns.org>)
List pgsql-hackers
On Fri, 3 May 2002, Neil Conway wrote:

> Hi all,
>
> The SQL92 spec has this to say about SET CONSTRAINTS DEFERRED:
>
>     a) If ALL is specified, then the constraint mode in TXN of all
>        constraints that are DEFERRABLE is set to deferred.
>
>     b) Otherwise, the constraint mode in TXN for the constraints
>        identified by the <constraint name>s in the <constraint name
>        list> is set to deferred.
>
> (section 14.2, page 401)
>
> My reading of this: if you specify ALL, only the constraints marked
> as DEFERRABLE are affected. If you specify a specific constraint,
> it is deferred, whether the constraint is marked as DEFERRABLE or
> not.
>
> Current Postgres behavior is incompatible with this interpretation:

I think you missed Syntax Rule 2:
"The constraint specified by <constraint name> shall be DEFERRABLE"



pgsql-hackers by date:

Previous
From: Neil Conway
Date:
Subject: set constraints behavior
Next
From: mlw
Date:
Subject: Re: HEADS UP: Win32/OS2/BeOS native ports