Tom Lane <tgl@sss.pgh.pa.us> wrote:
> Bill Moran <wmoran@potentialtech.com> writes:
> > Let's take the following fictional scenerio:
>
> > BEGIN;
> > INSERT INTO table1 VALUES ('somestring');
> > INSERT INTO table1 VALUES ('anotherstring');
> > SELECT user_defined_function();
> > COMMIT;
>
> > In this case, user_defined_function() does a lot more table manipulation.
> > I don't want that to be done if any statement prior fails, but it seems as
> > if it's always done, regardless. It seems as if the second INSERT is not
> > executed if the first fails, but the function is always called.
>
> Sorry, I don't believe a word of that. If the first insert fails,
> everything will be rejected until COMMIT.
>
> Possibly you need to show a less fictionalized version of your problem.
You're right, Tom. I can't get a simplified reproduction of the problem.
That means that the problem is occurring somewhere else in my program.
--
Bill Moran
Potential Technologies
http://www.potentialtech.com