Re: Improved \df(+) in psql + backward-compatibility

On Mon, Aug 29, 2005 at 11:13:29AM -0400, Tom Lane wrote:
> David Fetter <david@fetter.org> writes:
> > I've noticed that \df doesn't do quite what it might when a
> > function is created with named input parameters.  Please find
> > enclosed a patch against CVS TIP that does this better.
>
> Meanwhile, getting back to the actual merits of the patch ... this
> is not right yet, because it will do the wrong thing when there are
> OUT parameters.

Right.  I'd tried doing something with INOUT and OUT parameters, but I
wasn't able to figure out how to do with oid[] what I'd do with
oidvector.  On the bright side, what I did does do the right thing if
there are named IN parameters, which was part of what I was trying to
fix.

> (The proargnames array includes both IN and OUT params, and you
> can't assume that proargnames and proargtypes have corresponding
> subscripts.)  It would probably be a good idea to discuss what
> display we want for a function with OUT parameters, anyway.  The
> strict columnar representation that \df currently uses doesn't scale
> very well :-(

Speaking of said psql's columnar representations, what about the
alignment thing proposed earlier where an embedded newline doesn't
mess up the alignment of everything else?  Is there some generic way
to handle this?

Cheers,
D
--
David Fetter david@fetter.org http://fetter.org/
phone: +1 510 893 6100   mobile: +1 415 235 3778

Remember to vote!