On 8/1/06, Ian Harding <harding.ian@gmail.com> wrote:
> On 8/1/06, Christopher Browne <cbbrowne@acm.org> wrote:
> > Martha Stewart called it a Good Thing when "Carlo Stonebanks" <cstonebanks@nissenfasteners.com> wrote:
> > > I am interested in finding out a "non-religious" answer to which
> > > procedural language has the richest and most robust implementation
> > > for Postgres. C is at the bottom of my list because of how much
> > > damage runaway code can cause. I also would like a solution which is
> > > platorm-independent; we develop on Windows but may deploy on Linux.
> >
>
>
> >
> > - Doing funky string munging using the SQL functions available in
> > pl/pgsql is likely to be painful;
> >
> > - Doing a lot of DB manipulation in pl/Perl or pl/Tcl or such
> > requires having an extra level of function manipulations that
> > won't be as natural as straight pl/pgsql.
>
> Another important distinguishing characteristic is whether it supports
> set returning functions. I think only plpgsql does right now.
and C, and SQL ;)
in fact, sql functions make the best SRF because they are fast,
basically as fast as a query, but also can be called like this:
select sql_func(); --works!
select plpgsql_func(); --bad
select * from plpgsqlfunc(); works, but the other form is nice in some
situations
merlin