This is the mail archive of the fortran@gcc.gnu.org mailing list for the GNU Fortran project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: RFC: Renaming 'enum gfc_generic_isym_id'


On Monday 28 May 2007 17:17:41 Bud Davis wrote:
> --- Daniel Franke <franke.daniel@gmail.com> wrote:
> >  (1) rename 'enum gfc_generic_isym_id' to 'enum> gfc_isym_id'
> >  (2) add missing intrinsics to 'enum gfc_isym_id'
> >  (3) add an additional 'id' argument to 'add_sym_[012345][s]'
> >  (4) change all callers, add the corresponding 'gfc_isym_id'
> >  (5) rename member 'generic_id' in 'struct gfc_intrinsic_sym' to 'id'
> >  (6) change all callers
>
> suggestion:  make this change seperate from fixing any
> PR or changing any functionality.  then it is much
> easier to see what is important and what is not when
> looking at diff's.

This is what I had in mind :)


> overall, i love your idea.   while you are doing this,
> you may want to make sure the difference between
> generic and others is not lost....the previous names
> mentioned generic and the significance might be
> important...something to think about.

Over all source files (gcc/fortran/*.[ch]), 'generic_id' is used 25 times, 
most notably as an argument to switch (11 times). As far as I see, all of 
them could easily work on 'id' as they do with 'generic_id'. The generic 
aspect stems from intrinsic.c (make_generic) where functions as ABS are 
overloaded with specific names as CABS, DABS, ... . There is no such 
genericity with subroutines. Thus, I think we would not loose any information 
if the 'generic' part of the name is removed. But the point you raise is the 
reason for the request: the hive-mind may know better.

I'll go ahead and prepare a patch following the outline given above. A patch 
tells more than vague descriptions.

Regards
	Daniel


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]