add 'manipulation' support to CSS touch-action parsing.

add 'manipulation' support to CSS touch-action parsing.
'manipulation' is added in latest editors draft
https://dvcs.w3.org/hg/pointerevents/raw-file/tip/pointerEvents.html#the-touch-action-css-property

BUG=349016
Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=168752

[gnana, 2014-01-24 16:17:07]

Please have a look.

[Rick Byers, 2014-01-24 17:03:26]

I'd love to be able to do this!  But unfortunately it's unlikely to meet the
blink new API guidelines: http://www.chromium.org/blink#new-features since there
is no standard for it.

My plan has been to wait until we've shipped the standard touch-action support,
then talk to the other browser vendors to see if I can get any interest in
extending the standard somehow.  Unfortunately there are IP issues in play here
- eg. anything around gestures (like "double tap") is explicitly out of scope
for the PointerEvents WG: http://www.w3.org/2012/pointerevents/charter/.

So, although the code change is easy, the bureaucracy here is a pain...

[gnana, 2014-01-24 18:04:45]

On 2014/01/24 17:03:26, Rick Byers wrote:
> I'd love to be able to do this!  But unfortunately it's unlikely to meet the
> blink new API guidelines: http://www.chromium.org/blink#new-features since
there
> is no standard for it.
> 
> My plan has been to wait until we've shipped the standard touch-action
support,
> then talk to the other browser vendors to see if I can get any interest in
> extending the standard somehow.  Unfortunately there are IP issues in play
here
> - eg. anything around gestures (like "double tap") is explicitly out of scope
> for the PointerEvents WG: http://www.w3.org/2012/pointerevents/charter/.
> 
> So, although the code change is easy, the bureaucracy here is a pain...

Thanks Rick.
I completely understand and agree.

[Rick Byers, 2014-03-04 16:31:31]

On 2014/01/24 18:04:45, gnana wrote:
> On 2014/01/24 17:03:26, Rick Byers wrote:
> > I'd love to be able to do this!  But unfortunately it's unlikely to meet the
> > blink new API guidelines: http://www.chromium.org/blink#new-features since
> there
> > is no standard for it.
> > 
> > My plan has been to wait until we've shipped the standard touch-action
> support,
> > then talk to the other browser vendors to see if I can get any interest in
> > extending the standard somehow.  Unfortunately there are IP issues in play
> here
> > - eg. anything around gestures (like "double tap") is explicitly out of
scope
> > for the PointerEvents WG: http://www.w3.org/2012/pointerevents/charter/.
> > 
> > So, although the code change is easy, the bureaucracy here is a pain...
> 
> Thanks Rick.
> I completely understand and agree.

Hey Gnana, the working group just agreed to my proposal to add 'manipulation' to
the specification!  I think this will give us almost all of the benefit of
controlling double-tap without having to mention the gesture explicitly (which
is out of scope for the standard).  I.e. 'touch-action: manipulation' and
'touch-action: auto' are identical except that 'auto' includes double-tap-zoom.

So I'd like to get 'manipulation' into our initial M35 release (i.e. in the next
two weeks - time is short!).  Are you interested in updating this CL to replace
'double-tap-zoom' with 'manipulation'?  If you're busy I'm happy to do it, but
you did most of the work here already :-)

[gnana, 2014-03-04 16:34:18]

On 2014/03/04 16:31:31, Rick Byers wrote:
> On 2014/01/24 18:04:45, gnana wrote:
> > On 2014/01/24 17:03:26, Rick Byers wrote:
> > > I'd love to be able to do this!  But unfortunately it's unlikely to meet
the
> > > blink new API guidelines: http://www.chromium.org/blink#new-features since
> > there
> > > is no standard for it.
> > > 
> > > My plan has been to wait until we've shipped the standard touch-action
> > support,
> > > then talk to the other browser vendors to see if I can get any interest in
> > > extending the standard somehow.  Unfortunately there are IP issues in play
> > here
> > > - eg. anything around gestures (like "double tap") is explicitly out of
> scope
> > > for the PointerEvents WG: http://www.w3.org/2012/pointerevents/charter/.
> > > 
> > > So, although the code change is easy, the bureaucracy here is a pain...
> > 
> > Thanks Rick.
> > I completely understand and agree.
> 
> Hey Gnana, the working group just agreed to my proposal to add 'manipulation'
to
> the specification!  I think this will give us almost all of the benefit of
> controlling double-tap without having to mention the gesture explicitly (which
> is out of scope for the standard).  I.e. 'touch-action: manipulation' and
> 'touch-action: auto' are identical except that 'auto' includes
double-tap-zoom.
> 
> So I'd like to get 'manipulation' into our initial M35 release (i.e. in the
next
> two weeks - time is short!).  Are you interested in updating this CL to
replace
> 'double-tap-zoom' with 'manipulation'?  If you're busy I'm happy to do it, but
> you did most of the work here already :-)

Sure Rick, i will take this up.

[Rick Byers, 2014-03-04 17:58:59]

On 2014/03/04 16:34:18, gnana wrote:
> On 2014/03/04 16:31:31, Rick Byers wrote:
> > On 2014/01/24 18:04:45, gnana wrote:
> > > On 2014/01/24 17:03:26, Rick Byers wrote:
> > > > I'd love to be able to do this!  But unfortunately it's unlikely to meet
> the
> > > > blink new API guidelines: http://www.chromium.org/blink#new-features
since
> > > there
> > > > is no standard for it.
> > > > 
> > > > My plan has been to wait until we've shipped the standard touch-action
> > > support,
> > > > then talk to the other browser vendors to see if I can get any interest
in
> > > > extending the standard somehow.  Unfortunately there are IP issues in
play
> > > here
> > > > - eg. anything around gestures (like "double tap") is explicitly out of
> > scope
> > > > for the PointerEvents WG: http://www.w3.org/2012/pointerevents/charter/.
> > > > 
> > > > So, although the code change is easy, the bureaucracy here is a pain...
> > > 
> > > Thanks Rick.
> > > I completely understand and agree.
> > 
> > Hey Gnana, the working group just agreed to my proposal to add
'manipulation'
> to
> > the specification!  I think this will give us almost all of the benefit of
> > controlling double-tap without having to mention the gesture explicitly
(which
> > is out of scope for the standard).  I.e. 'touch-action: manipulation' and
> > 'touch-action: auto' are identical except that 'auto' includes
> double-tap-zoom.
> > 
> > So I'd like to get 'manipulation' into our initial M35 release (i.e. in the
> next
> > two weeks - time is short!).  Are you interested in updating this CL to
> replace
> > 'double-tap-zoom' with 'manipulation'?  If you're busy I'm happy to do it,
but
> > you did most of the work here already :-)
> 
> Sure Rick, i will take this up.

Thanks!

I realized it's a little more involved then a simple swap of double-tap for
manipulation, but should still be straight forward.  To avoid complicating
EventHandler::intersectTouchAction I think it'll be simplest to add a
'pinch-zoom' flag internally, and then parse 'manipulation' as 'PAN-X | PAN-Y |
PINCH-ZOOM'.  I think that way you won't have to touch intersectTouchAction at
all (pinch-zoom is just one more flag).  You'll want a test that verifies, eg:

 <div style='touch-action: pan-x'>
  <div style='touch-action: manipulation' expected-action='pan-x'></div>
 </div>

[gnana, 2014-03-05 10:32:03]

Please have a look.

[Rick Byers, 2014-03-05 14:41:12]

Looks good, thank you!  Just a couple small suggestions.

https://codereview.chromium.org/134153016/diff/90001/LayoutTests/fast/css/tou...
File LayoutTests/fast/css/touch-action-parsing.html (right):

https://codereview.chromium.org/134153016/diff/90001/LayoutTests/fast/css/tou...
LayoutTests/fast/css/touch-action-parsing.html:24: <div class="test"
id="explicit-manipulation" style="touch-action: manipulation;"
expected="manipulation"></div>
please add a test that verifies a combination like 'manipulation pan-x' is
invalid.

https://codereview.chromium.org/134153016/diff/90001/Source/core/css/CSSCompu...
File Source/core/css/CSSComputedStyleDeclaration.cpp (right):

https://codereview.chromium.org/134153016/diff/90001/Source/core/css/CSSCompu...
Source/core/css/CSSComputedStyleDeclaration.cpp:1446: if (touchAction &
TouchActionZoom)
This code should ASSERT on invalid combinations like 'TouchActionZoom' on it's
own (rather than silently converting it to 'manipulation').

Perhaps put 'if (touchAction == TouchActionZoom | TouchActionPanX |
TouchActionPanY)' first and put the remaining PanX/PanY cases inside an else
block?

https://codereview.chromium.org/134153016/diff/90001/Source/web/tests/data/to...
File Source/web/tests/data/touch-action-pan.html (right):

https://codereview.chromium.org/134153016/diff/90001/Source/web/tests/data/to...
Source/web/tests/data/touch-action-pan.html:57: <div class='ta-panx'>
please also add a simple test for 'manipulation' on it's own (or manipulation
intersected with 'auto') - just to verify that's it's possible to ever get
manipulation as the expected action.

https://codereview.chromium.org/134153016/diff/90001/public/web/WebTouchAction.h
File public/web/WebTouchAction.h (right):

https://codereview.chromium.org/134153016/diff/90001/public/web/WebTouchActio...
public/web/WebTouchAction.h:42: WebTouchActionZoom = 0x8,
Since this is intended to control pinch-zoom but NOT double-tap-zoom I think we
should be more explicit in the name.  Maybe WebTouchActionPinchZoom?

[gnana, 2014-03-06 08:11:14]

Thanks rick,
Please have a look.

https://codereview.chromium.org/134153016/diff/90001/LayoutTests/fast/css/tou...
File LayoutTests/fast/css/touch-action-parsing.html (right):

https://codereview.chromium.org/134153016/diff/90001/LayoutTests/fast/css/tou...
LayoutTests/fast/css/touch-action-parsing.html:24: <div class="test"
id="explicit-manipulation" style="touch-action: manipulation;"
expected="manipulation"></div>
On 2014/03/05 14:41:13, Rick Byers wrote:
> please add a test that verifies a combination like 'manipulation pan-x' is
> invalid.

Done.

https://codereview.chromium.org/134153016/diff/90001/Source/core/css/CSSCompu...
File Source/core/css/CSSComputedStyleDeclaration.cpp (right):

https://codereview.chromium.org/134153016/diff/90001/Source/core/css/CSSCompu...
Source/core/css/CSSComputedStyleDeclaration.cpp:1446: if (touchAction &
TouchActionZoom)
On 2014/03/05 14:41:13, Rick Byers wrote:
> This code should ASSERT on invalid combinations like 'TouchActionZoom' on it's
> own (rather than silently converting it to 'manipulation').
> 
> Perhaps put 'if (touchAction == TouchActionZoom | TouchActionPanX |
> TouchActionPanY)' first and put the remaining PanX/PanY cases inside an else
> block?

Done.

https://codereview.chromium.org/134153016/diff/90001/Source/web/tests/data/to...
File Source/web/tests/data/touch-action-pan.html (right):

https://codereview.chromium.org/134153016/diff/90001/Source/web/tests/data/to...
Source/web/tests/data/touch-action-pan.html:57: <div class='ta-panx'>
On 2014/03/05 14:41:13, Rick Byers wrote:
> please also add a simple test for 'manipulation' on it's own (or manipulation
> intersected with 'auto') - just to verify that's it's possible to ever get
> manipulation as the expected action.

Done.

https://codereview.chromium.org/134153016/diff/90001/public/web/WebTouchAction.h
File public/web/WebTouchAction.h (right):

https://codereview.chromium.org/134153016/diff/90001/public/web/WebTouchActio...
public/web/WebTouchAction.h:42: WebTouchActionZoom = 0x8,
On 2014/03/05 14:41:13, Rick Byers wrote:
> Since this is intended to control pinch-zoom but NOT double-tap-zoom I think
we
> should be more explicit in the name.  Maybe WebTouchActionPinchZoom? 

Done.

[Rick Byers, 2014-03-06 14:12:59]

LGTM, thanks for the quick turn-around!  I'm hoping we can ship touch action
next week (one more blocking bug, and I'll send the intent-to-ship).

+Eric for OWNERS

[eseidel, 2014-03-07 21:05:39]

Is this CSS Property guarded behind a runtime flag?  Should it be?

https://code.google.com/p/chromium/codesearch#chromium/src/third_party/WebKit...

[Rick Byers, 2014-03-07 21:29:50]

On 2014/03/07 21:05:39, eseidel wrote:
> Is this CSS Property guarded behind a runtime flag?  Should it be?
> 
>
https://code.google.com/p/chromium/codesearch#chromium/src/third_party/WebKit...

Yes (unless I'm missing something) the whole property is already covered, see:

bool CSSPropertyParser::parseTouchAction(bool important)
{
  if (!RuntimeEnabledFeatures::cssTouchActionEnabled())
    return false;

[eseidel, 2014-03-07 21:34:59]

lgtm

https://codereview.chromium.org/134153016/diff/150001/Source/core/css/CSSValu...
File Source/core/css/CSSValueKeywords.in (right):

https://codereview.chromium.org/134153016/diff/150001/Source/core/css/CSSValu...
Source/core/css/CSSValueKeywords.in:1012: manipulation
I see now, this is adding a new *value* not property, the property is already
guarded.  thanks for explaining.

[commit-bot: I haz the power, 2014-03-07 21:35:12]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/gnanasekar.s@samsung.com/134153016/150001

[commit-bot: I haz the power, 2014-03-07 22:01:11]

Change committed as 168752