Is record handle available to a check constraint stored procedure call?

Started by Gauthier, Daveabout 14 years ago2 messagesgeneral
Jump to latest
#1Gauthier, Dave
dave.gauthier@intel.com

v8.3.4 on Linux

I have a check constraint on a column. The constraint decides pass/fail based on the returned status of a stored procedure call that returns either "OK" or "NO". So when the stored procedure is called, there's a living attempt to insert or update a record.

Question: Is there a handle on the record being inserted or updated? Something like what's available in a trigger function (new.col1, old.col2, etc...)? If so, what does it look like?

Thanks in Advance

#2Jasen Betts
jasen@xnet.co.nz
In reply to: Gauthier, Dave (#1)
Re: Is record handle available to a check constraint stored procedure call?

On 2012-03-23, Gauthier, Dave <dave.gauthier@intel.com> wrote:

--_000_0AD01C53605506449BA127FB8B99E5E10C358B82FMSMSX105amrcor_
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: quoted-printable

v8.3.4 on Linux

I have a check constraint on a column. The constraint decides pass/fail ba=
sed on the returned status of a stored procedure call that returns either "=
OK" or "NO". So when the stored procedure is called, there's a living atte=
mpt to insert or update a record.

Question: Is there a handle on the record being inserted or updated? Some=
thing like what's available in a trigger function (new.col1, old.col2, etc.=
..)? If so, what does it look like?

it looks like the name of the table

create function somefunc( a foo ) returns boolean as
$$ begin return a.id <> 7 ; end $$ language plpgsql;

alter table foo add check ( somefunc(foo));

--
⚂⚃ 100% natural