Re: Patch for missing implementations of function prototypes.



On Wed, 16 Feb 2000, Sam O'Connor wrote:

> Patches attached.
> 
> Havoc Pennington wrote:
> > 
> > "Sam O'Connor" <oconnor@eiffel.com> writes:
> > > We would like
> > > out customers to be able to type make, have the wrapper
> > > library built along with the application that uses it,
> > > and not have to change it by hand to remove the undefined
> > > functions.
> > >
> > 
> > Your wrapper-generator should be able to trivially special-case these
> > functions, I hope this isn't more than a few lines of code, so
> > hopefully it doesn't actually matter for your customers.
> > 
> > Most likely a patch to clean it up would be accepted of course.
> > 
> > Havoc
> > 
> > --
> >          To unsubscribe: mail gtk-devel-list-request@redhat.com with
> >                        "unsubscribe" as the Subject.

thanks for the patch, it is in CVS now. however, for the glib part:

diff -c -r glib-1.2.7/gmain.c glib-1.2.7.patched/gmain.c
*** glib-1.2.7/gmain.c  Wed Feb 16 03:00:59 2000
--- glib-1.2.7.patched/gmain.c  Wed Feb 16 16:10:06 2000
***************
*** 1386,1389 ****
--- 1386,1397 ----
  g_idle_remove_by_data (gpointer data)
  {
    return g_source_remove_by_funcs_user_data (&idle_funcs, data);
+
+ }
+
+ /* Prevent symbol not defined  errors for binding writers */
+
+ void
+ glib_dummy_decl (void)
+ {
  }


i'm not going to apply this, as the declaration is just what it
says, a *dummy* *declaration* that should not come with an actual
implementation. if people try to link against it, an error should
occour.
for your bindings, you've got two options:

a) add functionality to your wrapper generator to ignore a defined
   set of functions
b) add void glib_dummy_decl (void) {} to your glue code

obviously, a) is the Right Way to do it.

---
ciaoTJ




[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]