Re: Gtk2::MozEmbed + DOM

[Torsten Schoenfeld <kaffeetisch gmx de>]

On Sun, 2005-03-27 at 18:24 +0200, Scott Lanning wrote:

> Did you copy the signal documentation from somewhere?

Yes, I copied most of the docs from

  http://www.mozilla.org/unix/gtk-embedding.html

It doesn't have docs for the dom_* signals, though.

>    I noticed some of the POD in GtkMozEmbed.xs is incorrect.
> For example, =for only takes one "paragraph" and it must follow
> the =for tag immediately (see `perldoc perlpod`). These should
> be replaced with =begin and =end, if I'm not mistaken. Also,
> I think blank lines are required after =cut, etc. Hm, the POD
> generated seems okay, though. I shouldn't worry about it.
> The POD generated for the dom event functions is messed up
> though because they use ALIAS. I didn't look at that yet.

Well, the POD in the XS files is not really standard, but rather an own
dialect that is parsed by Glib::GenPod.  See `perldoc Glib::GenPod` for
more.

> Finally, I didn't write any tests yet. But it's a four-day weekend,
> so I have some time.

I think it's not going to be easy to test this stuff in an automated
fashion.  You'd have to synthesize events and emit them via
Glib::Object::signal_emit, but that would mean having constructors for
the two event types.

> Something that worries me a little bit...
> Gtk2::MozEmbed relies on gtkmozembed, in Mozilla. I heard
> that Mozilla development will stop and Firefox will take its
> place. I don't know if Firefox has the same gtkmozembed.
> I looked in firefox-1.01 source, and it has an identical
> gtkmozembed.h, so maybe there's no cause for worry. But I wonder
> if we tried adding nsIWebBrowser/contentDOMWindow stuff for accessing
> the DOM, whether that would all be different in Firefox.

Firefox seems to ship its own version of the gtkmozembed stuff.  The
pkg-config file is called firefox-gtkmozembed, so perhaps it's enough to
test for the existence of either of them (via `pkg-config --exists´) and
use whatever is there.

-- 
Bye,
-Torsten