[postgis-devel] run_test.pl

[Sandro Santilli]

On Thu, May 31, 2012 at 08:20:09AM -0700, Paul Ramsey wrote:
> On Thu, May 31, 2012 at 8:15 AM, Sandro Santilli <strk at keybit.net> wrote:
> > On Wed, May 30, 2012 at 03:18:14PM -0400, Paul Ramsey wrote:
> >> OK, run_test.pl now works for me on all tests, under OSX. I would
> >> expect it does the same for Linux, but a test would be nice. I'm going
> >> to see if it works for my under Windows/MinGW. Trying it under
> >> Windows/CMD will be a different adventure.
> >
> > Can't locate Text/Diff.pm in @INC (@INC contains: /etc/perl /usr/local/lib/perl/5.10.1 /usr/local/share/perl/5.10.1 /usr/lib/perl5 /usr/share/perl5 /usr/lib/perl/5.10 /usr/share/perl/5.10 /usr/local/lib/site_perl .) at ./run_test.pl line 3.
> >
> > Can we avoid raising the dependencies ?
> 
> It would change the format of the output diffs. Right now everything
> is the SAME as before. Unified diffs on output. If we avoid Text::Diff
> we have to write our own little diff routine, which won't be as clever
> or as nice.

I don't think anyone relies on the format of those outputs.
Actually, I would even like it more to have a custom format,
so to finally distinguish the expected line from the obtained one.
You can always run your preferred diff command on the files
themselves afterwards.

> Running cpan is too hard? The other dependency which is
> less standard is File::Which, which can probably be avoided or not
> used if we accept a less pretty script.

I never learnt how to run cpan effectively. And by effectively I include
doing it w/out having the root on a machine. Even then, it'd be the first
dependency on a non-core perl feature. I can take a look at rolling a
diff if you solve the File::Which.

--strk; 

  ,------o-. 
  |   __/  |    Delivering high quality PostGIS 2.0 !
  |  / 2.0 |    http://strk.keybit.net - http://vizzuality.com
  `-o------'