GitLab

This is the first in a series of coordinate transformation cleanups.
Chiefly, I'd like to include pinch coordinate transforms in the FrameView
contentsToWindow type methods. In this patch I've looked just at "RootView"
There shouldn't be any functional change, this is all disambiguation and
renaming.

I looked at all callers of FrameView's:

    IntPoint rootViewToContents(const IntPoint&) const;
    IntPoint contentsToRootView(const IntPoint&) const;
    IntRect rootViewToContents(const IntRect&) const;
    IntRect contentsToRootView(const IntRect&) const;

Currently these methods for converting to/from RootView and Window in
FrameView do the same transformation, so it seems they've been used
interchangeably. They no longer mean the same thing, RootView is the root
FrameView which doesn't include scale and the pinch viewport offset.
Window (which I'll be renaming to viewport shortly) is the visible
viewport so it should include the pinch transformation. I've changed the
rootView version to window where the caller really means "viewport".
Also renamed "view" in coordinate cases to "frame" as I think it is
less ambiguous. e.g a call to contentsToRootView that actually wants
coordinates in the viewport space (i.e. DIP screen coordinates on Android,
DIP window coordinates on desktops) is now contentsToWindow and will soon
be contentsToViewport. A call to contentsToRootView that actually wants
coordinates in page coordinates (i.e. CSS pixels relative to the root
document) is now contentsToRootFrame.

There were external APIs that got or set coordinates in the RootView.
Again, these sometimes really meant viewport so I've renamed those as
appropriate as well. I've also made a few other coordinate related
cleanups along the way:
openPagePopup wasn't using the element rect param so I removed it.
I've renamed parameters like 'rect' and 'point' as I see them to make clear
what their coordinate space is.

BUG=371902

Review URL: https://codereview.chromium.org/919423002

git-svn-id: svn://svn.chromium.org/blink/trunk@190693 bbb929c8-8fbe-4397-9dbb-9b2b20218538

http://src.chromium.org/viewvc/blink?view=revision&revision=190682

BUG=453586
TBR=wangxianzhu@chromium.org

Review URL: https://codereview.chromium.org/950083002

git-svn-id: svn://svn.chromium.org/blink/trunk@190692 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Blink Gardening.

BUG=461042
TEST=
TBR=zmo@chromium.org
NOTRY=true

Review URL: https://codereview.chromium.org/952703002

git-svn-id: svn://svn.chromium.org/blink/trunk@190691 bbb929c8-8fbe-4397-9dbb-9b2b20218538

BUG=450612
TBR=jchaffraix@chromium.org

Review URL: https://codereview.chromium.org/948063002

git-svn-id: svn://svn.chromium.org/blink/trunk@190690 bbb929c8-8fbe-4397-9dbb-9b2b20218538

PageScaleFactor isn't an InternalSetting so it doesn't make sense to reset it
here. The PageScaleFactor is reset in Internals::resetToConsistentState so this
should have no effect on current tests as these two are always called together.

BUG=

Review URL: https://codereview.chromium.org/945363006

git-svn-id: svn://svn.chromium.org/blink/trunk@190689 bbb929c8-8fbe-4397-9dbb-9b2b20218538

None of the implementations do anything with this argument.

R=dsinclair@chromium.org

Review URL: https://codereview.chromium.org/952593003

git-svn-id: svn://svn.chromium.org/blink/trunk@190688 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Blink gardening.

BUG=461041
TEST=
TBR=zmo@chromium.org
NOTRY=true

Review URL: https://codereview.chromium.org/932243003

git-svn-id: svn://svn.chromium.org/blink/trunk@190687 bbb929c8-8fbe-4397-9dbb-9b2b20218538

This patch implements elementsFromPoint from the following spec:
http://www.w3.org/TR/cssom-view/#extensions-to-the-document-interface

There are two changes in this patch:
1) HitTestResult's rect-based results have been converted into "list-
based" results. This required extracting out a mask for hit testing
with list-based results in HitTestRequest and making sure all
rect-based tests also pass HitTestRequest::ListBased when hit testing
(this is asserted in several places).

2) Document.elementsFromPoint has been implemented in a straightforward
way on top of #1. The test results match IE11 with one exception: IE11
does not ignore pointer-events none but this patch does (implied, but
not specified explicitly in the spec). Until
https://www.w3.org/Bugs/Public/show_bug.cgi?id=27829 is fixed, this
does not work on ShadowRoots, though the patch has been implemented
in TreeScope instead of Document to make adding ShadowRoot support trivial.

BUG=448529

Review URL: https://codereview.chromium.org/869813003

git-svn-id: svn://svn.chromium.org/blink/trunk@190686 bbb929c8-8fbe-4397-9dbb-9b2b20218538

These methods are not used anymore.

Review URL: https://codereview.chromium.org/947213002

git-svn-id: svn://svn.chromium.org/blink/trunk@190685 bbb929c8-8fbe-4397-9dbb-9b2b20218538

An AudioBufferSourceNode that is looping should loop forever unless a
grain duration is explicitly given.  In particular, start(0, 0) should
still loop forever.

New test added, and updated one test in
audiobuffersource-loop-comprehensive to make it clear that the looping
source with duration stops at the right time.

Manual test added for handling the case where the AudioBufferSourceNode buffer is assigned after start() has been called. This is difficult to test properly using an offline context.


BUG=457099
TEST=audiobuffersource-loop-grain-no-duration.html

Review URL: https://codereview.chromium.org/912803005

git-svn-id: svn://svn.chromium.org/blink/trunk@190684 bbb929c8-8fbe-4397-9dbb-9b2b20218538

BUG=450612
TBR=jchaffraix@chromium.org

Review URL: https://codereview.chromium.org/951583002

git-svn-id: svn://svn.chromium.org/blink/trunk@190683 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Revert of Revert of Clip incremental invalidation rects for RenderBox (patchset #4 id:160001 of https://codereview.chromium.org/899813004/)

Reason for revert:
Suspected the original CL cause of the bug, but proved not.

Original issue's description:
> Revert of Clip incremental invalidation rects for RenderBox (patchset #3 id:40001 of https://codereview.chromium.org/548523005/)
> 
> Reason for revert:
> Haven't looked into the issue deeply, but in the future we won't care about the invalidation rects, so we can just revert this kind of optimizations.
> 
> BUG=453586
> 
> Original issue's description:
> > Clip incremental invalidation rects for RenderBox
> > 
> > RenderBox::incrementallyInvalidatePaint invalidates according to
> > border-box delta, not considering the bounds. If the bounds are
> > clipped by ancestor clipping, the incrementally invalidated rect
> > may exceed the ancestor clipping.
> > 
> > Clip the incremental invalidation rects by oldBounds and newBounds
> > to avoid invalidation on unrelated areas.
> > 
> > BUG=411420
> > TEST=fast/repaint/resize-with-border-clipped.html
> > 
> > Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=182081
> 
> TBR=jchaffraix@chromium.org,dsinclair@chromium.org,esprehn@chromium.org,leviw@chromium.org
> BUG=411420
> 
> Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=189569

TBR=jchaffraix@chromium.org,dsinclair@chromium.org,esprehn@chromium.org,leviw@chromium.org
BUG=453586,456176

Review URL: https://codereview.chromium.org/948513004

git-svn-id: svn://svn.chromium.org/blink/trunk@190682 bbb929c8-8fbe-4397-9dbb-9b2b20218538

If the owning renderer gets resized, the scrollbar may be repositioned. Don't
rely on other reasons for invalidation (background, borders, child content
getting relaid out, etc.) to accidentally get the area covered by the
scrollbars invalidated, since there may very well be no other reason.

BUG=460434
R=jchaffraix@chromium.org,leviw@chromium.org

Review URL: https://codereview.chromium.org/942963002

git-svn-id: svn://svn.chromium.org/blink/trunk@190681 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Invalidation sets which are empty stems from selector features only found
in the rightmost compound selector, and can only cause invalidation of
elements for which they are scheduled. When adding these sets, just mark
the element with a LocalStyleChange and do not schedule a style
invalidation for them.

Similarly, mark with a SubtreeStyleChange and clear all scheduled
invalidation sets for an element which gets a wholeSubtreeInvalid set
scheduled.

This fixes two sub-optimal behaviors in the code.

1. We avoid running style invalidation, hence traversing the DOM tree an
   extra time, if we only have local/subtree changes scheduled.
2. We avoid traversing subtrees checking against an empty invalidation set
   that will never match.

Note that we will still have to run the style invalidation step to clear
style invalidation bits on elements if there were invalidation sets
scheduled prior to a SubtreeStyleChange that nuked the entry in the
m_pendingInvalidationMap.

Added missing check for m_customPseudoInvalid in isEmpty and resetting it
in setWholeSubtreeInvalid() was added.

Since rules of type ".x {}" will not lead to a style invalidation pass
anymore, ".x .dummy {}" type rules have been added to the inspector tests
when testing style invalidation tracking.

Review URL: https://codereview.chromium.org/946993002

git-svn-id: svn://svn.chromium.org/blink/trunk@190680 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Source/core/page/FocusType was moved to
public/WebFocusType.h in https://codereview.chromium.org/852083002
This updates the gypi file to reflect that change and to avoid
a warning in the msvs-ninja gyp generator.

Note: there is no gypi file that tracks the files in public/platform.

R=fsamuel@chromium.org,jochen@chromium.org

BUG=448716,147644,459058

Review URL: https://codereview.chromium.org/950833002

git-svn-id: svn://svn.chromium.org/blink/trunk@190679 bbb929c8-8fbe-4397-9dbb-9b2b20218538

BUG=26723

Review URL: https://codereview.chromium.org/915543003

git-svn-id: svn://svn.chromium.org/blink/trunk@190676 bbb929c8-8fbe-4397-9dbb-9b2b20218538

This relands https://blink.lc/blink/commit/?id=5668429, which was reverted in
https://blink.lc/blink/commit/?id=a1303a8.

The Mac trackpad regression is addressed by
https://blink.lc/blink/commit/?id=021397d.

BUG=575,417782

Review URL: https://codereview.chromium.org/882813005

git-svn-id: svn://svn.chromium.org/blink/trunk@190675 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Previously these shared a display item type with 2D affine transform items.
This resolves a FIXME in Transform3DDisplayItem.h.

Review URL: https://codereview.chromium.org/946423002

git-svn-id: svn://svn.chromium.org/blink/trunk@190674 bbb929c8-8fbe-4397-9dbb-9b2b20218538

The change may have caused a bad crasher on Canary.

> Make RenderObject::style() return a const object
> 
> The current code was sloppy and allowed accidental mutation
> of RenderStyle outside of style recalc. While this change
> doesn't fix that class of bugs (as we introduce some
> mutable accessors), it greatly reduces the risk of
> them being introduced by mistake.
> 
> In order to make the change work, 2 fields on RenderStyle had
> to be made mutable (unique and explicitInheritance) because
> they are propagated up the tree during style recalc.
> 
> Also StyleSelectorState had to be changed to store the override
> parent style independently of the regular parent style. That's
> required as RefPtr modifies the style by increasing the refCount
> and we have to hold onto the override as it can be a free style.
> 
> Review URL: https://codereview.chromium.org/889563002

TBR=jchaffraix@chromium.org
BUG=460820

Review URL: https://codereview.chromium.org/948053002

git-svn-id: svn://svn.chromium.org/blink/trunk@190673 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Source/platform/graphics/test/MockDiscardablePixelRef.h was removed in
https://codereview.chromium.org/485833005 and this removes the file
from blink_platform.gypi as well to avoid warnings in the msvs-ninja
gyp generator.

BUG=459058
R=senorblanco@chromium.org,hclam@chromium.org

Review URL: https://codereview.chromium.org/952443003

git-svn-id: svn://svn.chromium.org/blink/trunk@190671 bbb929c8-8fbe-4397-9dbb-9b2b20218538

BUG=450612
TBR=jchaffraix@chromium.org, dmazzoni@chromium.org,

Review URL: https://codereview.chromium.org/939273002

git-svn-id: svn://svn.chromium.org/blink/trunk@190670 bbb929c8-8fbe-4397-9dbb-9b2b20218538

BUG=450977

Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=190594

Review URL: https://codereview.chromium.org/899923004

git-svn-id: svn://svn.chromium.org/blink/trunk@190668 bbb929c8-8fbe-4397-9dbb-9b2b20218538

This method only operates on the TextTrack that it gets as an argument, so
the code can be moved to the caller without changing behavior. The
HTMLMediaElement will still be notified via the mode-change (if one takes
place.)

BUG=321654

Review URL: https://codereview.chromium.org/924273004

git-svn-id: svn://svn.chromium.org/blink/trunk@190667 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Straightforward migration of rendering/RenderBoxModelObject.

BUG=450612

Review URL: https://codereview.chromium.org/944923004

git-svn-id: svn://svn.chromium.org/blink/trunk@190666 bbb929c8-8fbe-4397-9dbb-9b2b20218538

BUG=450612
TBR=jchaffraix@chromium.org, dmazzoni@chromium.org

Review URL: https://codereview.chromium.org/942583002

git-svn-id: svn://svn.chromium.org/blink/trunk@190665 bbb929c8-8fbe-4397-9dbb-9b2b20218538

BUG=460633
TBR=vollick@chromium.org

Review URL: https://codereview.chromium.org/945423003

git-svn-id: svn://svn.chromium.org/blink/trunk@190664 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Source\core\testing\PartialPrivateScriptTest.h was removed in
https://codereview.chromium.org/519703002 and this removed the file
from the gypi file as well. This avoids warnings in the
msvs-ninja gyp generator.

R=haraken@chromium.org,jl@opera.com

BUG=341031,459058

Review URL: https://codereview.chromium.org/925043004

git-svn-id: svn://svn.chromium.org/blink/trunk@190663 bbb929c8-8fbe-4397-9dbb-9b2b20218538

BUG=450977

Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=190568

Review URL: https://codereview.chromium.org/915703002

git-svn-id: svn://svn.chromium.org/blink/trunk@190662 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Revert of Teach ScriptRunner how to yield and post on loading task queue (patchset #10 id:260001 of https://codereview.chromium.org/866273005/)

Reason for revert:
Looks like this was causing oilpan leaks

Original issue's description:
> Teach ScriptRunner how to yield
> 
> Previously ScriptRunner could execute an arbitrary number of scripts
> leading to janks.  Now ScriptRunner polls shouldYieldForHighPriorityWork
> before running each script.
> 
> In addition the task to execute the scripts is now posted on the
> loading task queue.
> 
> BUG=456777
> 
> Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=190648

TBR=skyostil@chromium.org,marja@chromium.org,sigbjornf@opera.com
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=456777

Review URL: https://codereview.chromium.org/936493003

git-svn-id: svn://svn.chromium.org/blink/trunk@190661 bbb929c8-8fbe-4397-9dbb-9b2b20218538

BUG=460885
TBR=tkent@chromium.org

Review URL: https://codereview.chromium.org/943303004

git-svn-id: svn://svn.chromium.org/blink/trunk@190660 bbb929c8-8fbe-4397-9dbb-9b2b20218538

The latest editor's draft introduces MIDIPortState state attribute in
MIDIPort. To support the attribute, update the blink API to use enum
for representing MIDIPortState instead of bool isActive flag.

Spec: http://webaudio.github.io/web-midi-api/#idl-def-MIDIPort

This is the final step of three to update the MIDIAccessor blink API.

BUG=279097
TEST=build with chromium; content_unittests; LayoutTest/webmidi

Review URL: https://codereview.chromium.org/950643002

git-svn-id: svn://svn.chromium.org/blink/trunk@190658 bbb929c8-8fbe-4397-9dbb-9b2b20218538

When a LoadableTextTrack receives cues from the loader, it adds them to
the cue list and notifies the CueTimeline. This is the only place where
this kind of TextTrack state is directly handled by subclasses of
TextTrack. Moving this to TextTrack we can "lock down" state - cue list
and timeline - some more, and have updates due to list-mutations in a
single place (there's currently still updates in HTMLMediaElement though.)
This also makes LoadableTextTrack look more like a provider of cues.
Transport of cues needs to be updated to use a wider type. Also make use
of Vector::swap in VTTParser while changing the signatures.

BUG=321654

Review URL: https://codereview.chromium.org/941973005

git-svn-id: svn://svn.chromium.org/blink/trunk@190657 bbb929c8-8fbe-4397-9dbb-9b2b20218538

We were already showing a horizontal scrollbar for -webkit-paged-x. For
-webkit-paged-y, on the other hand, it only worked when it was the viewport
that was paginated, but not when specified on a regular block.

The CSS parser currently deals with the overflow shorthand in such a way that
it's the overflow-y property that gets the -webkit-paged-foo value (while
overflow-x becomes auto). Digression: While one could say that's not quite
right when it comes to horizontal writing modes, handling shorthands based on
writing mode would be nasty...

Skipping 2 of the new tests for the old / current multicol implementation,
since they fail there. All tests pass in the new implementation.

BUG=460024
R=dsinclair@chromium.org,jchaffraix@chromium.org

Review URL: https://codereview.chromium.org/940913002

git-svn-id: svn://svn.chromium.org/blink/trunk@190656 bbb929c8-8fbe-4397-9dbb-9b2b20218538

TextTrack.addCue is specified to always remove the cue from any list it
currently resides in, but the implementation would only do that if it
was currently in a different track's list.
Also update spec. prose snippets and merge the two exception-conditions
in the removeCue-implementation. The "!m_cues || !m_cues->remove(cue)"
part should never be true unless state has been corrupted, and the spec.
only says to throw NotFoundError in any case.

BUG=460476

Review URL: https://codereview.chromium.org/944823003

git-svn-id: svn://svn.chromium.org/blink/trunk@190655 bbb929c8-8fbe-4397-9dbb-9b2b20218538

The file itself was removed in r189473.

Review URL: https://codereview.chromium.org/946373002

git-svn-id: svn://svn.chromium.org/blink/trunk@190654 bbb929c8-8fbe-4397-9dbb-9b2b20218538

In the latest editor's draft spec, MIDIAccess has only one EventHandler
onstatechange that is integrated thing of onconnect and ondisconnect.
Also, ondisconnect of MIDIPort is also renamed to onstatechange.

See the latest editor]s draft spec:
  - http://webaudio.github.io/web-midi-api/#midiaccess-interface
  - http://webaudio.github.io/web-midi-api/#idl-def-MIDIPort

This patch update these EventHandler attributes, and also add
implementation to dispatch these events.

BUG=279097

Review URL: https://codereview.chromium.org/943173002

git-svn-id: svn://svn.chromium.org/blink/trunk@190653 bbb929c8-8fbe-4397-9dbb-9b2b20218538

http/tests/xmlhttprequest/xmlhttprequest-responseXML-abort.html is flaky, but
I cannot reproduce it and don't see any cause. To understand the problem, this
CL adds two more layout tests.

R=hiroshige
BUG=415872

Review URL: https://codereview.chromium.org/950813002

git-svn-id: svn://svn.chromium.org/blink/trunk@190652 bbb929c8-8fbe-4397-9dbb-9b2b20218538

R=jl@opera.com, haraken@chromium.org

Review URL: https://codereview.chromium.org/946293003

git-svn-id: svn://svn.chromium.org/blink/trunk@190651 bbb929c8-8fbe-4397-9dbb-9b2b20218538

R=haraken
BUG=420515

Review URL: https://codereview.chromium.org/946163003

git-svn-id: svn://svn.chromium.org/blink/trunk@190650 bbb929c8-8fbe-4397-9dbb-9b2b20218538

https://dom.spec.whatwg.org/#interface-domtokenlist

BUG=460722

Review URL: https://codereview.chromium.org/947833002

git-svn-id: svn://svn.chromium.org/blink/trunk@190649 bbb929c8-8fbe-4397-9dbb-9b2b20218538