Flesh out more of ChromiumBridge.cpp eliminating all but two methods on...

Flesh out more of ChromiumBridge.cpp eliminating all but two methods on
TemporaryGlue.

Adds WebKitClient::userAgent() function to avoid a webkit_glue dependency.

R=dglazkov
BUG=none
TEST=none


Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=30609

[dglazkov, 2009-10-30 17:41:49]

LGTM after nits:

http://codereview.chromium.org/350003/diff/1/2
File webkit/api/src/ChromiumBridge.cpp (right):

http://codereview.chromium.org/350003/diff/1/2#newcode86
Line 86: static ChromeClientImpl* getChromeClientImpl(Widget* widget)
should be chromeClientImpl(..) according to WebKit style.

http://codereview.chromium.org/350003/diff/1/2#newcode103
Line 103: static WebWidgetClient* getWebWidgetClient(Widget* widget)
Ditto.

http://codereview.chromium.org/350003/diff/1/3
File webkit/api/src/WebPluginContainerImpl.cpp (right):

http://codereview.chromium.org/350003/diff/1/3#newcode376
Line 376: //setCursorForPlugin(cursorInfo, parentView->frame());
Did you mean to not remove this?

[darin (slow to review), 2009-10-30 17:47:58]

http://codereview.chromium.org/350003/diff/1/2
File webkit/api/src/ChromiumBridge.cpp (right):

http://codereview.chromium.org/350003/diff/1/2#newcode86
Line 86: static ChromeClientImpl* getChromeClientImpl(Widget* widget)
On 2009/10/30 17:41:49, Dimitri Glazkov wrote:
> should be chromeClientImpl(..) according to WebKit style.

i thought about that but since it isn't a 'getter' function, i thought that
would be bad.  maybe toChromeClientImpl is better?  i think that is the
convention for methods in WebKit that convert / unwrap / etc, right?

http://codereview.chromium.org/350003/diff/1/3
File webkit/api/src/WebPluginContainerImpl.cpp (right):

http://codereview.chromium.org/350003/diff/1/3#newcode376
Line 376: //setCursorForPlugin(cursorInfo, parentView->frame());
On 2009/10/30 17:41:49, Dimitri Glazkov wrote:
> Did you mean to not remove this?

oops, thanks.  i'll remove that.

[dglazkov, 2009-10-30 18:24:17]

> i thought about that but since it isn't a 'getter' function, i thought that
> would be bad.  maybe toChromeClientImpl is better?  i think that is the
> convention for methods in WebKit that convert / unwrap / etc, right?

yep, you're right -- to* is better.