Add test for style responsive animation behaviour

Add test for style responsive animation behaviour

This adds a test helper for checking that animations correctly update
themselves in response to style changes.
This is intended to aid testing the InvalidatableStyleInterpolation
refactor.

BUG=437696
Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=197419

[alancutter (OOO until 2018), 2015-06-18 06:00:02]

alancutter@chromium.org changed reviewers:
+ shans@chromium.org

[shans, 2015-06-18 06:26:02]

https://codereview.chromium.org/1181253006/diff/1/LayoutTests/animations/resp...
File LayoutTests/animations/responsive/resources/responsive-test.js (right):

https://codereview.chromium.org/1181253006/diff/1/LayoutTests/animations/resp...
LayoutTests/animations/responsive/resources/responsive-test.js:53: }
for (var j = i + 1; ...

https://codereview.chromium.org/1181253006/diff/1/LayoutTests/animations/resp...
LayoutTests/animations/responsive/resources/responsive-test.js:59:
console.assert(beforeExpect[k].at == afterExpect[k].at);
It's really not very obvious from the input data structure that this is a
requirement. I think it would be clearer to treat 'k' as the key to lists of
expectations.

https://codereview.chromium.org/1181253006/diff/1/LayoutTests/animations/resp...
LayoutTests/animations/responsive/resources/responsive-test.js:87: function
setInlineStyle(target, style) {
This seems predicated on specifying style dictionaries for parentStyle and
targetStyle. Do you anticipate it will be likely people will want to set more
than one property?

https://codereview.chromium.org/1181253006/diff/1/LayoutTests/animations/resp...
LayoutTests/animations/responsive/resources/responsive-test.js:183: }
why are these separate loops?

[alancutter (OOO until 2018), 2015-06-18 07:11:33]

https://codereview.chromium.org/1181253006/diff/1/LayoutTests/animations/resp...
File LayoutTests/animations/responsive/resources/responsive-test.js (right):

https://codereview.chromium.org/1181253006/diff/1/LayoutTests/animations/resp...
LayoutTests/animations/responsive/resources/responsive-test.js:53: }
On 2015/06/18 at 06:26:02, shans wrote:
> for (var j = i + 1; ...

"var j = i" was a debugging tweak that got accidentally uploaded, reuploaded
with "var j = 0".

https://codereview.chromium.org/1181253006/diff/1/LayoutTests/animations/resp...
LayoutTests/animations/responsive/resources/responsive-test.js:59:
console.assert(beforeExpect[k].at == afterExpect[k].at);
On 2015/06/18 at 06:26:02, shans wrote:
> It's really not very obvious from the input data structure that this is a
requirement. I think it would be clearer to treat 'k' as the key to lists of
expectations.

On second thought there's no reason to require this, the destination
expectations are the only things checked in each state transition test. Removed
the assert.

https://codereview.chromium.org/1181253006/diff/1/LayoutTests/animations/resp...
LayoutTests/animations/responsive/resources/responsive-test.js:87: function
setInlineStyle(target, style) {
On 2015/06/18 at 06:26:02, shans wrote:
> This seems predicated on specifying style dictionaries for parentStyle and
targetStyle. Do you anticipate it will be likely people will want to set more
than one property?

I can't think of any tests that would need more than one property, thankfully
our CSS property interdependencies aren't too crazy. I went with dictionaries
because they seem to be cleanest syntactically to both specify and output in
test descriptions.

https://codereview.chromium.org/1181253006/diff/1/LayoutTests/animations/resp...
LayoutTests/animations/responsive/resources/responsive-test.js:183: }
On 2015/06/18 at 06:26:02, shans wrote:
> why are these separate loops?

The above comment describes. I batch the style modifications separate from the
getComputedStyle() calls to avoid doing a ton of style recalcs. This was an
issue with interpolation-test.js where tests would timeout (on a Z620!) because
they took too long to compute.

[alancutter (OOO until 2018), 2015-06-18 23:49:26]

Removed setInlineStyle, PTAL.

[shans, 2015-06-18 23:53:46]

lgtm

[alancutter (OOO until 2018), 2015-06-18 23:57:48]

The CQ bit was checked by alancutter@chromium.org

[alancutter (OOO until 2018), 2015-06-18 23:57:49]

The patchset sent to the CQ was uploaded after l-g-t-m from shans@chromium.org
Link to the patchset: https://codereview.chromium.org/1181253006/#ps80001
(title: "Renamed transition to stateTransition to avoid confusion with CSS
Transitions")

[commit-bot: I haz the power, 2015-06-18 23:58:28]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1181253006/80001

[darktears, 2015-06-18 23:58:38]

I'm currently out of the office in vacations until Monday 6th of July. I will
not be checking my email regularly, expect delays in my answers.

For Crosswalk related topics please contact either
<kenneth.r.christiansen@intel.com> or <francesco.balestrieri@intel.com>

For Crosswalk infrastructure please contact <raphael.kubo.da.costa@intel.com>

[commit-bot: I haz the power, 2015-06-19 02:09:58]

Committed patchset #5 (id:80001) as
https://src.chromium.org/viewvc/blink?view=rev&revision=197419