On Wed, Jul 02, 2003 at 01:23:09PM +0200, Andre Poenitz spake thusly: > > On Wed, Jul 02, 2003 at 02:11:08PM +0300, Martin Vermeer wrote:
> > > Should not be needed if all branches end in 'return'.
> >
> > gcc doesn't seem to think so :-(
>
> I think it does.
>
> Otherwise we'd get much more warnings in insets/*.C
>
> I'd guess you have a 'break' hidden somewhere...
Don't you think I've tried? Currently I have
switch (cmd.action) {
case LFUN_INSET_MODIFY:
{
InsetNoteParams params;
InsetNoteMailer::string2params(cmd.argument, params);
params_.type = params.type;
params_.branch = params.branch;
setButtonLabel();
bv->updateInset(this);
return DISPATCHED;
}
case LFUN_INSET_EDIT:
if (cmd.button() == mouse_button::button3)
break;
return InsetCollapsable::localDispatch(cmd);
case LFUN_INSET_DIALOG_UPDATE:
InsetNoteMailer("note", *this).updateDialog(bv);
return DISPATCHED;
case LFUN_MOUSE_RELEASE:
if (cmd.button() == mouse_button::button3) {
InsetNoteMailer("note", *this).showDialog(bv);
return DISPATCHED;
}
break;
default:
return InsetCollapsable::localDispatch(cmd);
}
//return UNDISPATCHED; // suppresses warning
...and still getting the warning.
> Andre'
Martin
pgp00000.pgp
Description: PGP signature
