DomainKey-Signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma;
	h=message-id:date:from:to:subject:cc:in-reply-to:mime-version
	:content-type:content-transfer-encoding:content-disposition
	:references;
	b=uvBRggsqjwl0DRoQpSNV6mLsRvWaSOpkulN9KG/zxiWa8T3WT5vhdfoU49+JIUeJ8z
	cxSo2tA9kW1J3pTWEVVmRlwwGeG9whmtSjmng14VQh1wepqgeChMaagP8bZ+ADLVkp4z
	EWjf3DunAwOaVSj2hNeK5wTA7KHE8l2WxAmno=
Message-ID: <4dc28b160810281312m7f9c7635sf60502bef468a77b@mail.gmail.com>
Date: Tue, 28 Oct 2008 15:12:59 -0500
From: "Joe Halpin" <jhalpin100@gmail.com>
To: "Tom Lane" <tgl@sss.pgh.pa.us>
Subject: Re: Problem with trigger function in C
Cc: pgsql-interfaces@postgresql.org
In-Reply-To: <15986.1225224113@sss.pgh.pa.us>
MIME-Version: 1.0
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
Content-Disposition: inline
References: <4dc28b160810281234g1292831awf4b8716d64f7f405@mail.gmail.com>
	<15986.1225224113@sss.pgh.pa.us>

Sorry I forgot to add that part. Here's how I did that:

sn=# create function xqueryTrigger() returns trigger as
'/usr/local/lib/postgresql/xqueryTrigger.so' language c;
CREATE FUNCTION
sn=# create trigger tbefore before insert or update on at_xparam
for each row execute procedure xquerytrigger();
CREATE TRIGGER

Is the function name case sensitive? I had originally named it
xqueryTrigger, but after some error messages which referenced
"xquerytrigger" I renamed it to be all lower case.

Thanks

Joe


On Tue, Oct 28, 2008 at 3:01 PM, Tom Lane <tgl@sss.pgh.pa.us> wrote:
> "Joe Halpin" <jhalpin100@gmail.com> writes:
>> I'm new at this and I apologize if I got the wrong list, but I'm
>> trying to write a trigger function in C which is linked in with
>> libxqilla and is supposed to run an xquery against some xml we have
>> stored in a column. I find that when the function is entered fcinfo is
>> null, which causes a segfault.
>
> AFAIK the only way fcinfo would be null is if PG didn't think that the
> function was V1 call convention.  I see that you do have the
> PG_FUNCTION_INFO_V1 macro in there, which eliminates the most obvious
> explanation.  Are you sure that the SQL-level CREATE FUNCTION command
> references this function name and not some other one?
>
>                        regards, tom lane
>