Re: Documentation Request

[Bob Caryl <bob fis-cal com>]

Murray Cumming wrote:
> On Mon, 2005-03-21 at 11:00 -0600, Bob Caryl wrote:
>  
> > Murray Cumming wrote:
> >    
> > > On Mon, 2005-03-21 at 10:35 -0600, Bob Caryl wrote:
> > >
> > >      
> > > > Hello everyone,
> > > >
> > > > Can someone tell me to what the formal parameters of the following 
> > > > template refer?  I am guessing that the path is for the row that has a 
> > > > child row toggled, and the iterator is for the toggled child row.  Am I 
> > > > correct?
> > > >
> > > > Glib::SignalProxy2< void, const TreeModel::Path&, const 
> > > > TreeModel::iterator& >Gtk::TreeModel::signal_row_has_child_toggled()
> > > >   
> > > >        
> > > The similar function of the same name gives some clue:
> > > signal_row_has_child_toggled
> > > http://gtkmm.org/docs/gtkmm-2.4/docs/reference/html/classGtk_1_1TreeModel.html#a16
> > >
> > > signals are not well documented in gtkmm. I'm still looking for a way to
> > > get their documentation automatically from the GTK+ documentation.
> > > Eventaully we will give up and do it manually. In this case, even GTK+
> > > does not have any docs for the signal:
> > > http://developer.gnome.org/doc/API/2.0/gtk/GtkTreeModel.html#GtkTreeModel-row-has-child-toggled
> > >
> > > That deserves a bugzilla bug for GTK+.
> > >
> > > Luckily, gtkmm has less signals than GTK+, so it is a bit clearer.
> > >
> > >
> > >      
> > Ok... from what I see from the links you provided it appears that I will 
> > know only the parent row's path and iterator.  Does this mean that I 
> > will have to scan all the child rows for this parent row to discover the 
> > data state of the child that was toggled?  I see a test program to find 
> > out exactly what I'm getting passed to me in this case.  I will report 
> > to the list what these parms actually are.
> >    
>
> I'm not sure exactly what you are trying to do. I guess, but I don't
> know, that this signal is emitted when the last child is removed, or the
> first child is added.
>
> But if you are just interested in whether a Gtk::CellRendererToggle has
> been toggled, then handle its "toggled" signal. This should be well
> covered in the book and the examples.
>
>  
Hey Murray,

As it turns out, what I really wanted was to connect to the 
"signal_changed" funtion of a Gtk::TreeView::Selection object.   As for 
what I'm doing:  <deep breath>

My launcher application configures a menu based on the contents of an 
xml file.  I am writing a configuration utility that parses that same 
xml file into a Gtk::TreeStore, storeing all the elements necessary to 
create a new xml file with an abbreviated version of the full menu.  
Along with the fields I need for a new xml file, I also insert the a 
"bool" data type to get an editable check box.

What I need to do is:  When a row is selected (and probably when the 
state of the check box is toggled) I need to determine the ancestor of 
the current row if it has one.  I then need to determine if any of its 
children's checkboxes are "activated" and then activate the ancestor's 
check box, proceeding recursively up the treestore until I hit the top 
level.

So, what I am struggling with now is deriving my own version of 
Gtk::TreeModel so that I can call "iter_parent_vfunc" to get the 
iterator of a particular row's parent.  I have derived a class, but when 
I try to reference it using Glib::RefPtr<> so that 
Gtk::TreeView::Selection::get_model can populate it, I get all kinds of 
errors.  My derivation is very simple:

class MyTreeModel : public Gtk::TreeModel
{
   public:
      virtual bool iter_parent_vfunc(const iterator& child, iterator& 
parent) (Gtk::TreeModel::iter_parent_vfunc(child,parent););
}

My problem is that when I declare an instance of this new class:

Glib::RefPtr<MyTreeModel> mymodel;

The compiler has a hissy and I still  cannot use my new class.  If I 
declare it this way:

MyTreeModel *mymodel;

and try to use (MyTreeModel *)Gtk::TreeView::TreeSelection::get_model() 
to get my model object, the compiler again complains that it cannot 
convert Glib::RefPtr<Gtk::TreeModel> to MyTreeModel * .

Any ideas on what I'm doing wrong?

Thanks,

Bob Caryl