Re: Bug in pg_dump

Michael,

* Stephen Frost (sfrost(at)snowman(dot)net) wrote:
> * Michael Paquier (michael(dot)paquier(at)gmail(dot)com) wrote:
> > >> Subject: [PATCH 2/3] Make prove_check install contents of current directory as well
> > >
> > > This is really an independent thing, no? I don't see any particular
> > > problem with it, for my part.
> >
> > Yes, that's an independent thing, but my guess is that it would be
> > good to have a real test case this time to be sure that this does not
> > break again, at least on master where test/modules is an ideal place.
>
> I've been looking at this, but noticed the following in
> src/test/Makefile:
>
> # We want to recurse to all subdirs for all standard targets, except that
> # installcheck and install should not recurse into the subdirectory "modules".

Hrmpf, that hadn't fixed it as I thought, I had another issue which was
making it appear to work.

The other modules work because they use pg_regress and pass it an
'--extra-install' option, so perhaps adding this makes sense after all,
though I'm a bit nervous that we're doing double-duty with this
approach as some things clearly do get installed by the first 'install'.

Peter, if you have a minute, could you take a look at this thread and
discussion of having TAP tests under src/test/modules which need to
install an extension? I think it's something we certainly want to
support but I'm not sure it's a good idea to just run install in every
directory that has a prove_check.

I'm going to move forward with the actual bug fix. We can certainly add
the test in later, once we've got this all sorted.

Thanks!

Stephen