Re: Completing PL support for Event Triggers

On Fri, 2013-09-13 at 22:40 +0200, Dimitri Fontaine wrote:
> Please find attached to this email three patches covering the missing PL
> support for Event Triggers: pltcl, plperl and plpython.

You introduced some new compiler warnings, please fix those.

http://pgci.eisentraut.org/jenkins/view/PostgreSQL/job/postgresql_commitfest_world/80/warnings19Result/new/?

In the source code, I'm dubious about the use of is_dml_trigger. I can
see where you are coming from, but in most of the code, a trigger is a
trigger and an event trigger is an event trigger. Let's not introduce
more terminology.

Other opinions on that?

> Due to “platform” problems here tonight and the CF deadline, the
> plpython patch is known not to pass regression tests on my machine. The
> code is fully rebased and compiles without warning, though, so I'm still
> entering it into this CF: hopefully I will figure out what's wrong with
> my local plpython platform support here early next week.

For me, the plpython patch causes an (well, many) assertion failures in
the regression tests, because this change is wrong:

if (!found)
{
! /* Haven't found it, create a new cache entry */
! entry->proc = PLy_procedure_create(procTup, fn_oid,
! is_dml_trigger, is_evt_trigger);
if (use_cache)
entry->proc = proc;
}

When that is fixed, I get more failures and segfaults later.

Please give this another look. I'll review the Perl and Tcl things more
closely in the meantime.