[webkit-reviews] review requested: [Bug 17066] [GTK] Improve frameloader signals : [Attachment 18937] Improve load signals Take 3

[bugzilla-daemon at webkit.org]

Christian Dywan <christian at imendio.com> has asked  for review:
Bug 17066: [GTK] Improve frameloader signals
http://bugs.webkit.org/show_bug.cgi?id=17066

Attachment 18937: Improve load signals Take 3
http://bugs.webkit.org/attachment.cgi?id=18937&action=edit

------- Additional Comments from Christian Dywan <christian at imendio.com>
One part, terminology wise, was the distinction between "progress" and "load".
The former is global in relation to the view and not part of the "load" as
such. The original 'load-finished' was not the end of the load but in fact the
end of the "progress". So I replaced it to match web frame's "load-done" in the
proper meaning. In general it's now possible to use a web view or a web frame
on its own since both receive all meaningful signals.

As for enhancements, I'm introducing "redirect-scheduled", "redirect-done" and
"redirect-cancelled" and "document-load-done". Further more errors are reported
in the form of gerrors.

I tried to explain all the signals in code, ie. gtkdoc. If that is not
sufficient/ clear please say so.

A few slightly higher level changes:
. Renaming 'status-bar-text-changed' to "statusbar-text-changed"
. Removing 'selection-changed' and 'icon-loaded' since they are virtually dead
code; better introduce them again properly or they just cause confusion.
. Adding all signals to the web view struct so they can be cleanly overridden
in derived classes.
. A bit of related code/ docs cleanup.

Note that this patch is not 100% clean, I'm sure I overlooked one or two
things. I am going to fix these soon while testing a bit more and depending on
your feedback.