On 16 Jan 2003 11:47:20 -0200
Johan Dahlin <[EMAIL PROTECTED]> wrote:
> > > Strange.
> > > I seems that the code generator was interrupted when generating atk.c
> > > Try to remove atk.c (it's generated) and the build/ directory (it's used
> > > by distutils)
> > >
> > > If it still doesn't work, could you send me a more detailed bug report
> > > with gcc version, python version and the broken atk.c.
> >
> > I removed the atk.c file, the build dir, did cvs update, but rebuilding gave the
>same results.
> > gcc 2.95.4, on debian sarge. python 2.2. I have attached atk.c.
>
> Looking on the output of atk.c it looks like the error is in the
> following part of the code, codegen/codegen.py line 665
>
> # This is working well
> bymod = {}
> for module, pyname, cname in imports:
> bymod.setdefault(module, []).append((pyname, cname))
> fp.write(' PyObject *module;\n\n')
>
> # Here it stops
> for module in bymod:
> fp.write(' if ((module = PyImport_ImportModule(....)
> fp.write(' PyObject *moddict = PyModul....)
>
>
> It might be due to a bogus import statement of some kind.
>
> Could you try to add a debug output and print out bmod just before that.
> And modify setup.py to print out the argument send to the codegenerator?
> (add a print s on line 225 in setup.py)
Strange things are happening. When I add some print statements to codegen.py, they are
not printed to the screen but into atk.c!
The value of bymod printed there is: {'gobject': [('GObject', 'PyGObject_Type')]}
I have included both atk.c and codegen.py in a attachment.
Bye and Thanks,
Martijn
_______________________________________________
pygtk mailing list [EMAIL PROTECTED]
http://www.daa.com.au/mailman/listinfo/pygtk
Read the PyGTK FAQ: http://www.async.com.br/faq/pygtk/
