Normalize top controls offset to (0, 1), Blink-side.

Normalize top controls offset to (0, 1), Blink-side.

Since top controls height can now be changed, storing top controls
offset as an absolute value causes problems.  Normalizing its stored
format to (0, 1) means we we don't need to adjust it when the height
changes.

On the Blink side, Blink now needs to hear about the top controls
height, not just the "layout height" (effect on WebViewImpl size, which
can be 0), so that it can apply the absolute value to FrameView.

The new boolean m_topControlsOffsetIsNormalized is just a temporary hack
to allow this patch to be a no-op until the Chromium-side lands.  I'll
delete it in the same patch as WebWidget::setTopControlsLayoutHeight().

See also Chromium-side patch: https://codereview.chromium.org/901813002

BUG=430635
Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=189737

[aelias_OOO_until_Jul13, 2015-02-05 06:15:15]

[email protected] changed reviewers:
+ [email protected]

[aelias_OOO_until_Jul13, 2015-02-05 06:16:54]

Hi bokan@, PTAL.  This patch is to land first of the two.  Then I'll come back
and clean some of the cruft here as a third patch.

[bokan, 2015-02-06 08:25:24]

lgtm

https://codereview.chromium.org/882683003/diff/1/Source/web/WebViewImpl.h
File Source/web/WebViewImpl.h (right):

https://codereview.chromium.org/882683003/diff/1/Source/web/WebViewImpl.h#new...
Source/web/WebViewImpl.h:765: // as of the last Resize event.
It'd be more clear to say here (and in WebView.h) that the "WebView was shrunk
by the top controls...".

[aelias_OOO_until_Jul13, 2015-02-06 08:42:21]

[email protected] changed reviewers:
+ [email protected]

[aelias_OOO_until_Jul13, 2015-02-06 08:42:58]

Adding rbyers@ for public/web OWNERS.

[aelias_OOO_until_Jul13, 2015-02-06 20:32:21]

New patchsets have been uploaded after l-g-t-m from [email protected]

[aelias_OOO_until_Jul13, 2015-02-06 20:32:38]

https://codereview.chromium.org/882683003/diff/1/Source/web/WebViewImpl.h
File Source/web/WebViewImpl.h (right):

https://codereview.chromium.org/882683003/diff/1/Source/web/WebViewImpl.h#new...
Source/web/WebViewImpl.h:765: // as of the last Resize event.
On 2015/02/06 08:25:23, bokan wrote:
> It'd be more clear to say here (and in WebView.h) that the "WebView was shrunk
> by the top controls...". 

Done.

[Rick Byers, 2015-02-07 01:19:16]

On 2015/02/06 20:32:38, aelias wrote:
> https://codereview.chromium.org/882683003/diff/1/Source/web/WebViewImpl.h
> File Source/web/WebViewImpl.h (right):
> 
>
https://codereview.chromium.org/882683003/diff/1/Source/web/WebViewImpl.h#new...
> Source/web/WebViewImpl.h:765: // as of the last Resize event.
> On 2015/02/06 08:25:23, bokan wrote:
> > It'd be more clear to say here (and in WebView.h) that the "WebView was
shrunk
> > by the top controls...". 
> 
> Done.

What's your testing plan here?

[aelias_OOO_until_Jul13, 2015-02-07 02:41:03]

There were ~6 existing top controls related tests that well covered this
codepath.  I changed them to use the normalized format (was going to need to do
that anyway upon deleting the boolean).  I also enabled one test that was
disabled, and made them run on all platforms by replacing an OS(ANDROID)
codepath with one based on settings()->mainFrameResizesAreOrientationChanges(),
and removing an unnecessary non-Android assertion (a reasonable-ish setting to
use given that the Android-specific codepath in question relates to main frame
sizing).

[Rick Byers, 2015-02-07 03:13:52]

Thanks for the tests, and removing the #ifdef (yay for test coverage on linux of
clank features!).

LGTM with nit

https://codereview.chromium.org/882683003/diff/40001/public/web/WebWidget.h
File public/web/WebWidget.h (right):

https://codereview.chromium.org/882683003/diff/40001/public/web/WebWidget.h#n...
public/web/WebWidget.h:277: // Notification about the top controls height.  If
the boolean is true, then
nit: it's not clear that height here is a normalized value.  Perhaps change the
param to 'heightFraction' and clarify in the comment.

[aelias_OOO_until_Jul13, 2015-02-07 03:31:59]

New patchsets have been uploaded after l-g-t-m from [email protected]

[aelias_OOO_until_Jul13, 2015-02-07 03:32:08]

https://codereview.chromium.org/882683003/diff/40001/public/web/WebWidget.h
File public/web/WebWidget.h (right):

https://codereview.chromium.org/882683003/diff/40001/public/web/WebWidget.h#n...
public/web/WebWidget.h:277: // Notification about the top controls height.  If
the boolean is true, then
On 2015/02/07 03:13:52, Rick Byers wrote:
> nit: it's not clear that height here is a normalized value.  Perhaps change
the
> param to 'heightFraction' and clarify in the comment.

Height is still an absolute value here.  The one that's normalized is in
applyViewportDeltas.  I just renamed that one to topControlsShownRatioDelta to
clarify it.

(I needed to change this height method to add a boolean so that Blink is capable
of denormalizing the value without the layout size being changed.)

[Rick Byers, 2015-02-07 04:23:32]

On 2015/02/07 03:32:08, aelias wrote:
> https://codereview.chromium.org/882683003/diff/40001/public/web/WebWidget.h
> File public/web/WebWidget.h (right):
> 
>
https://codereview.chromium.org/882683003/diff/40001/public/web/WebWidget.h#n...
> public/web/WebWidget.h:277: // Notification about the top controls height.  If
> the boolean is true, then
> On 2015/02/07 03:13:52, Rick Byers wrote:
> > nit: it's not clear that height here is a normalized value.  Perhaps change
> the
> > param to 'heightFraction' and clarify in the comment.
> 
> Height is still an absolute value here.  The one that's normalized is in
> applyViewportDeltas.  I just renamed that one to topControlsShownRatioDelta to
> clarify it.
> 
> (I needed to change this height method to add a boolean so that Blink is
capable
> of denormalizing the value without the layout size being changed.)

Ah, that makes more sense - I missed the division there.  Thanks.  Still LGTM.

[aelias_OOO_until_Jul13, 2015-02-07 04:40:48]

The CQ bit was checked by [email protected]

[commit-bot: I haz the power, 2015-02-07 04:41:22]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/882683003/60001

[commit-bot: I haz the power, 2015-02-07 05:14:37]

Committed patchset #4 (id:60001) as
https://src.chromium.org/viewvc/blink?view=rev&revision=189737