Re: giomm API review needed

[José Alburquerque <jaalburquerque cox net>]

Jonathon Jongsma wrote:
> On Tue, 2008-01-15 at 15:55 +0100, Murray Cumming wrote:
>   
> > On Wed, 2008-01-09 at 17:58 +0100, Murray Cumming wrote:
> >     
> > > giomm is now in glibmm. But we need more people to take a look and
> > > review the API. We'll have to freeze the API quite soon for GNOME 2.22,
> > > though I'm not sure exactly when that will be for glib and glibmm.
> > >
> > > I am gradually adding documentation, but I'm only up to
> > > Gio::FileEnumerator so far, alphabetically:
> > > http://www.gtkmm.org/docs/glibmm-2.4/docs/reference/html/namespaceGio.html
> > >
> > > Gio::File is probably a good place to start, though I don't yet have a
> > > good idea of the API as a whole:
> > > http://www.gtkmm.org/docs/glibmm-2.4/docs/reference/html/classGio_1_1File.html#_details
> > >       
> > There are several TODO comments in the .hg files if someone would like
> > to help. It's simple stuff.
> >
> >     
>
> I finally got some time to sit down and play around with the giomm stuff
> tonight, but unfortunately I haven't gotten very far yet.  I can't even
> get a simple example to run.  Has anybody else written any test/sample
> programs with giomm? Am I doing something wrong here?
>
> $ cat giomm-test.cc 
> #include <giomm.h>
>
> // no overloads exist without a cancellable argument yet, so create a
> dummy NULL
> // cancellable
> Glib::RefPtr<Gio::Cancellable> not_cancellable;
>
> int main (int argc, char** argv)
> {
>     Gio::init();
>     Glib::RefPtr<Gio::File> f = Gio::File::create_for_path
> ("/home/jonathon/tmp/giomm.cc");
>     Glib::RefPtr<Gio::FileInputStream> stream = f->read
> (not_cancellable);
>     return 0;
> }
>
> $ ./giomm-test 
> Segmentation fault (core dumped)
>
> gdb says:
> Program received signal SIGSEGV, Segmentation fault.
> [Switching to Thread -1214220608 (LWP 12799)]
> 0x08053311 in ?? ()
> (gdb) bt
> #0  0x08053311 in ?? ()
> #1  0xbf8d8758 in ?? ()
> #2  0xb7dcd1ea in IA__g_type_create_instance (type=0) at gtype.c:1569
> Backtrace stopped: previous frame inner to this frame (corrupt stack?)
>
>   
I think that though the docs say the Cancellable can be NULL this is 
probably only true for for the C API.  You may need to have a "real" 
Cancellable.

-Jose