[patch] Fix PR fortran/27698 emit error when rejecting invalid name

[Steve Kargl]

On Wed, Jul 05, 2006 at 01:18:42PM -0700, Steve Kargl wrote:
> On Wed, Jul 05, 2006 at 09:13:11PM +0200, Paul Thomas wrote:
> > Bernhard,
> > 
> > >PS: My machine is currently too busy, so this patchlet is not tested so
> > >far. I'd be glad if someone could take care of testing it. Feel free to
> > >apply if it's ok. Apologies for my lazyness and TIA
> > >
> > Please test it; I think that it will break quite a lot of things but I 
> > could be wrong. 
> > 
> > gfc_match_name is often called as the first, or one of a series, of 
> > attempted matches.  A non-match does not necessarily mean an error, 
> > depending on the context.  I thought of doing the same but did not even try 
> > it for this reason.
> > 
> > Prove me wrong!
> 
> The only failures in the testsuite are
> 
> FAIL: gfortran.dg/do_iterator.f90  -O  (test for excess errors)
> FAIL: gfortran.dg/nullify_2.f90  -O   (test for errors, line 5)
> FAIL: gfortran.dg/nullify_2.f90  -O  (test for excess errors)
> FAIL: gfortran.dg/pr19936_1.f90  -O   (test for errors, line 4)
> FAIL: gfortran.dg/pr19936_1.f90  -O  (test for excess errors)
> 
> Inspection of the code shows that Bernhard's error is
> triggered by illegal code in these files.  I do, however,
> need to study the situation with a suite of legal code.

I just ran this patch against my large colllection of legal
Fortran (with extensions :-), and the only failure was in the
Moware package.  The failure occurs with and without the patch.  It
involves a *.F90 that needs to be preprocessed an Moware program,
and may have screwed up the Makefile sufficiently to get the
preprocessing mangled.

-- 
Steve