Re: Using scalar function as set-returning: bug or feature?

"Tels" <nospam-pg-abuse(at)bloodgate(dot)com> writes:
> On Mon, February 12, 2018 5:03 pm, Tom Lane wrote:
>> ... However, my pending patch at
>> https://commitfest.postgresql.org/17/1439/
>> gets rid of the use of DTYPE_ROW for composite types, and once that
>> is in it might well be reasonable to just throw a flat-out error for
>> wrong number of source values for a DTYPE_ROW target. I can't
>> immediately think of any good reason why you'd want to allow for
>> the number of INTO items not matching what the query produces.

> Perl lets you set a fixed number of multiple variables from an array and
> discard the rest like so:
> my ($a, $b) = (1,2,3);
> I'm not sure if you mean exactly the scenario as in the attached test
> case, but this works in plpgsql, too, and would be a shame to lose.

Well, that's exactly the issue. Whether that's a handy feature or
a foot-gun that hides bugs depends entirely on your point of view.
Personally, this is not the kind of behavior I really want from a
programming language ;-). And I'm sure that if plpgsql were still
enforcing the old rules, and someone came along with a proposal to
relax that to be more like Perl, it'd be laughed down.

Still, backwards compatibility is worth something. I don't really
have a strong opinion about whether to change it or not.

regards, tom lane