[Scons-users] F90PPCOM vs F90COM not invoked correctly

[Luke Robison]

On Thu, Apr 16, 2020 at 10:34 AM Mats Wichmann <mats at wichmann.us> wrote:

> On 4/16/20 9:12 AM, Bill Deegan wrote:
> > When you run under msys, which python are you using? windows or msys?
> > The code in question is in SCons/Tools/f*.py FortranCommon.py gfortran.py
> >
> > As far as I can see there's no F90PPSuffixes defined by default?
>

Bill,

 I'm using the msys python.  which python and scons -v both point to msys
paths.

The lack of F90PPsuffixes confused me as well, but it works for native
Linux, I'm just not sure how.


>
> the Tool/__init__.py module has some logic that adds a C scanner on a
> capital F suffix (and also on .fpp and .FPP).  I assume that's what
> causes cpp to run on it.  Look for CSuffixes.
>
> Mats,

Interesting to note that the list is missing .F90 .F95 .F03 .F08.  I think
the scanner would only affect the dependency generation, right?  The
Fortran Tool already defines it's own scanners, I'm not sure why any
fortran file extensions would be in that CScanner list.


All,
While looking at Tools/FortranCommon.py, I think I found the answer in
ComputeFortranSuffixes:

s = suffixes[0]
sup = s.upper()
if Scons.Util.case_sensitive_suffixes(s, sup):

sure enough in Util.py, there is essentially a hard-coded flag to make msys
(and cygwin) platforms return False, indicating case-insensitive platform.
I'm going to try to override that and see if I observe any ill effects.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://pairlist4.pair.net/pipermail/scons-users/attachments/20200416/96880758/attachment.html>