GitLab

Currently, first letter renderers are created and destroyed through the updateFirstLetters() call in
RenderBlock. This has, historically, been problematic as we may miss places where the call
was required, leading to accessing invalid memory.

This CL converts the first letter code to use a PseudoElement for the first-letter instead of
creating classes during layout.

With the PseudoElement implementation the two tests which are currently in
LayoutTests/fast/css/first-letter-removed-added.html work as expected.

BUG=391288

Review URL: https://codereview.chromium.org/571603003

git-svn-id: svn://svn.chromium.org/blink/trunk@183913 bbb929c8-8fbe-4397-9dbb-9b2b20218538

To be compliant to the WHATWG Encoding spec (living standard), we should
treat ISO-8859-1 and windows-1252 synonymously. Moreover, windows-1252
is the canonical encoding name.

Update TextCodecLatin1 to list only aliases specified in the encoding
spec.

See https://codereview.chromium.org/598383002 (ICU update) and
    https://codereview.chromium.org/649413002 (roll ICU with necessary
    test/pref/UI changes in Chrome).

BUG=412053
TEST=fast/encoding/*
TEST=http/tests/misc/char*
R=jochen@chromium.org, jsbell@chromium.org

Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=183781

Review URL: https://codereview.chromium.org/655083002

git-svn-id: svn://svn.chromium.org/blink/trunk@183912 bbb929c8-8fbe-4397-9dbb-9b2b20218538

NOTRY=True
TBR=kouhei,eae
BUG=397902,422050

Review URL: https://codereview.chromium.org/663863002

git-svn-id: svn://svn.chromium.org/blink/trunk@183911 bbb929c8-8fbe-4397-9dbb-9b2b20218538

When a render object changes location, we should force paint
invalidation checking for all descendants on the same backing,
including positioned ones. However, previously we only force
invalidation checking for non-positioned descendants.

Now let RenderBlock override
RenderObject::invalidatePaintOfSubtreesIfNeeded() so that it
can use the childPaintInvalidationState, which has been forced
checking of invaldation, for the positioned descendants.

BUG=422812
TEST=fast/repaint/positioned-great-grandparent-change-location.html

Review URL: https://codereview.chromium.org/662483002

git-svn-id: svn://svn.chromium.org/blink/trunk@183910 bbb929c8-8fbe-4397-9dbb-9b2b20218538

So as to be able to trace its PendingScript part object.

R=haraken
BUG=

Review URL: https://codereview.chromium.org/660233002

git-svn-id: svn://svn.chromium.org/blink/trunk@183909 bbb929c8-8fbe-4397-9dbb-9b2b20218538

The plan is to create a LocalFrame that is loosely tied to an existing RemoteFrame but is not really attached to the FrameTree. The LocalFrame will begin loading, and when it commits, it will be swapped into where the RemotFrame was. This adds sufficient logic to actually support that swap (and to maintain the intermediate semi-attached state while the LocalFrame is provisional).

BUG=422583

Review URL: https://codereview.chromium.org/643333002

git-svn-id: svn://svn.chromium.org/blink/trunk@183908 bbb929c8-8fbe-4397-9dbb-9b2b20218538

BUG=424648
TBR=yoichio@chromium.org
NOTRY=true

Review URL: https://codereview.chromium.org/665643002

git-svn-id: svn://svn.chromium.org/blink/trunk@183907 bbb929c8-8fbe-4397-9dbb-9b2b20218538

- add graphics for chrome on top of viewport rectange;
- make viewport border light-grey, let its width scale along with
    contents scale.
- change Viewport checkbox to Chrome (always show viewport)
- (drive by) properly handle potential image/texture creation errors
- (drive by) disable user selection on Layers3DView in CSS as annoying.

BUG=

Review URL: https://codereview.chromium.org/661753002

git-svn-id: svn://svn.chromium.org/blink/trunk@183906 bbb929c8-8fbe-4397-9dbb-9b2b20218538

This change adds code to save and replay the canvas state stack when
transferring from a display list to a raste canvas

BUG=423913
TEST=fast/canvas/canvas-unballanced-save.html

Review URL: https://codereview.chromium.org/661763003

git-svn-id: svn://svn.chromium.org/blink/trunk@183904 bbb929c8-8fbe-4397-9dbb-9b2b20218538

http://src.chromium.org/viewvc/blink?view=revision&revision=183890

BUG=421380
TBR=jchaffraix@chromium.org

Review URL: https://codereview.chromium.org/663783003

git-svn-id: svn://svn.chromium.org/blink/trunk@183903 bbb929c8-8fbe-4397-9dbb-9b2b20218538

beginLayer()/endLayer() behaves like save()/restore() for only CTM and clip states.
SkCanvas::saveLayer() behaves the same as save(), but in addition it allocates an offscreen buffer.
It causes the inconsistency for restoring states managed by between SkCanvas and SkPaint.

For example,
    context.translate(1, 0);
    context.setCompositeOperation(CompositeDestinationIn);

    context.beginLayer(1, CompositeSourceIn);
    context.translate(-1, 0);
    context.setCompositeOperation(CompositeSourceOver);
    context.endLayer();

    // At this moment, CTM is translate(1, 0) while current composite operator is CompositeSourceOver.

It's because CTM is managed by SkCanvas, while current composite operator is managed by GraphicsContext.

BUG=423414

Review URL: https://codereview.chromium.org/651243002

git-svn-id: svn://svn.chromium.org/blink/trunk@183902 bbb929c8-8fbe-4397-9dbb-9b2b20218538

The exception handling was added in r177449, but then was accidentally
removed via r183686. This patch brings the test and functionality back.

R=vsevik

Review URL: https://codereview.chromium.org/661983003

git-svn-id: svn://svn.chromium.org/blink/trunk@183901 bbb929c8-8fbe-4397-9dbb-9b2b20218538

This setting will not longer be used after:
https://codereview.chromium.org/618013003/

It will be landed after that change.

BUG=398485

Review URL: https://codereview.chromium.org/614173003

git-svn-id: svn://svn.chromium.org/blink/trunk@183900 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Add WebRuntimeFeatures::enableCSSViewport() method that should substitute the unfortunately named 'WebRuntimeFeatures::setCSSViewportEnabled()'.

BUG=424459

Review URL: https://codereview.chromium.org/662003002

git-svn-id: svn://svn.chromium.org/blink/trunk@183899 bbb929c8-8fbe-4397-9dbb-9b2b20218538

This adds UseCounters for SVG DOM, to see how much usage there is.

SVG DOM support is enabled by default, but it will be possible to disable it with a commandline flag (--disable-svg1dom), see https://codereview.chromium.org/577943002.

BUG=415074

Review URL: https://codereview.chromium.org/474013004

git-svn-id: svn://svn.chromium.org/blink/trunk@183898 bbb929c8-8fbe-4397-9dbb-9b2b20218538

These forward references are not needed anymore. After
this patch, Source/modules has no unneeded forward references
any more.

Review URL: https://codereview.chromium.org/651503002

git-svn-id: svn://svn.chromium.org/blink/trunk@183897 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Updated spec, IDL files, and all the *Credential variants.

BUG=400674

Review URL: https://codereview.chromium.org/661973002

git-svn-id: svn://svn.chromium.org/blink/trunk@183896 bbb929c8-8fbe-4397-9dbb-9b2b20218538

As per recent spec changes, remove create and get and replace it with open.

CLs in this three-legged-patch:

1. Add open to browser - https://codereview.chromium.org/664433003/
*2. Add open (remove get/create) to Blink - https://codereview.chromium.org/638023003
3. Remove get/create from browser - https://codereview.chromium.org/663503002/

BUG=423942

Review URL: https://codereview.chromium.org/638023003

git-svn-id: svn://svn.chromium.org/blink/trunk@183895 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Revert of [SVG] Promote PathOpsSVGClipping to stable. (patchset #1 id:1 of https://codereview.chromium.org/446583002/)

Reason for revert:
We've received valuable feedback from exposure to stable. Dialing back while the issues are being worked on.

Original issue's description:
> [SVG] Promote PathOpsSVGClipping to stable.
> 
> The feature has been in experimental state for more than a month with
> no known issues.
> 
> BUG=263911
> R=schenney@chromium.org, fs@opera.com, pdr@chromium.org
> 
> Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=179529

TBR=fs@opera.com,pdr@chromium.org,schenney@chromium.org
NOTREECHECKS=true
NOTRY=true
BUG=263911

Review URL: https://codereview.chromium.org/645023003

git-svn-id: svn://svn.chromium.org/blink/trunk@183894 bbb929c8-8fbe-4397-9dbb-9b2b20218538

JSFrame events are added in TimelineFlameChart and their selfTime is never calculated. These events are never added into the model as they are used only to draw the flame chart. When such event is selected we calculate its selfTime on the fly as duration - sum of children times.

BUG=424507

Review URL: https://codereview.chromium.org/645673004

git-svn-id: svn://svn.chromium.org/blink/trunk@183893 bbb929c8-8fbe-4397-9dbb-9b2b20218538

This patch replaces the following document.XXXX functions with window.XXXX custom ones:
- createDocumentFragment;
- createElement;
- createElementWithClass;
- createTextNode.

BUG=424188

Review URL: https://codereview.chromium.org/662793002

git-svn-id: svn://svn.chromium.org/blink/trunk@183892 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Improves code readabilty.
Also use >> for > > in template

BUG=none
TEST=no layout test failures

Review URL: https://codereview.chromium.org/637223008

git-svn-id: svn://svn.chromium.org/blink/trunk@183891 bbb929c8-8fbe-4397-9dbb-9b2b20218538

https://codereview.chromium.org/620553009 moved
the selection to invalidate-after-compositing-updates,
except that it lied and we still do some invalidations
on the spot. This is needed as we don't store the old
selection information so we can't generate the old
selection invalidation rect later.

This change just generates the invalidation on the spot
for block selection gaps (like we do for normal selection).

The long-term fix will be to store the old rect but it requires
more thoughts and we need some quick fix for this nasty issue.

BUG=421380

Review URL: https://codereview.chromium.org/659983002

git-svn-id: svn://svn.chromium.org/blink/trunk@183890 bbb929c8-8fbe-4397-9dbb-9b2b20218538

- properly annotated TracingModel.Event fields to let compiler catch the problem;
- check target is not null before using it.

BUG=424503

Review URL: https://codereview.chromium.org/658363002

git-svn-id: svn://svn.chromium.org/blink/trunk@183889 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Enables posting of idle tasks, via the blink scheduler, which notify V8 of idle
time when it can perform GC work.

BUG=421880

Review URL: https://codereview.chromium.org/639773007

git-svn-id: svn://svn.chromium.org/blink/trunk@183888 bbb929c8-8fbe-4397-9dbb-9b2b20218538

FileWriter calls onwrite as notification about the progress of writing.
So If we call write again from the onwrite event handler FileWriter fails to do that.
Actually we need to use onwriteend callback.

BUG=423843

Review URL: https://codereview.chromium.org/661063002

git-svn-id: svn://svn.chromium.org/blink/trunk@183886 bbb929c8-8fbe-4397-9dbb-9b2b20218538

BUG=424171

Review URL: https://codereview.chromium.org/651783004

git-svn-id: svn://svn.chromium.org/blink/trunk@183885 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Modifies the idle time scheduling such that idle tasks which are posted during one
idle period will not become pending until the next idle period.  This enables idle
tasks to repost themselves at the end of their execution without causing a flood of
idle task execution in the remaining deadline period.

BUG=421880

Review URL: https://codereview.chromium.org/640053003

git-svn-id: svn://svn.chromium.org/blink/trunk@183884 bbb929c8-8fbe-4397-9dbb-9b2b20218538

If a video is too narrow to display the full controls, including the cast
button, on the control bar, then use an overlay cast button instead.

BUG=423753

Review URL: https://codereview.chromium.org/658173002

git-svn-id: svn://svn.chromium.org/blink/trunk@183883 bbb929c8-8fbe-4397-9dbb-9b2b20218538

BUG=420022

Review URL: https://codereview.chromium.org/661643005

git-svn-id: svn://svn.chromium.org/blink/trunk@183882 bbb929c8-8fbe-4397-9dbb-9b2b20218538

As of today, the "format script" button is initialized as hidden and
will update its state in response to editorSelected event. However,
in case the event is sent prior to button construction, it should be
initialized with current selected editor.

BUG=424089
R=vsevik

Review URL: https://codereview.chromium.org/658293003

git-svn-id: svn://svn.chromium.org/blink/trunk@183881 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Settings checkbox UI state was changed from outside, that confused control.

BUG=402406

Review URL: https://codereview.chromium.org/640633007

git-svn-id: svn://svn.chromium.org/blink/trunk@183880 bbb929c8-8fbe-4397-9dbb-9b2b20218538

When CSSPropertyParser::parseValue fails, we sometimes have already
yielded property/value pairs (e.g. when parsing shorthands). In this
case we roll back any yielded values. This patch moves the logic for this
to CSSPropertyParser since this is a better fit for it and makes the
parseValue interface more sensible.

BUG=404023

Review URL: https://codereview.chromium.org/645033004

git-svn-id: svn://svn.chromium.org/blink/trunk@183879 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Chromium side fix is here: https://codereview.chromium.org/660973002

BUG=424464

Review URL: https://codereview.chromium.org/663723002

git-svn-id: svn://svn.chromium.org/blink/trunk@183878 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Revert of Use count strings as keyframes/animation-name identifiers (patchset #2 id:20001 of https://codereview.chromium.org/651123003/)

Reason for revert:
Seems to cause crash: 
https://code.google.com/p/chromium/issues/detail?id=424520

Original issue's description:
> Use count strings as keyframes/animation-name identifiers
> 
> Strings are not valid keyframes/animation-name identifiers. This change
> tracks the occurrence of these strings in the wild to provide data
> on the impact dropping support for this would have.
> 
> BUG=423424
> 
> Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=183856

TBR=timloh@chromium.org,dstockwell@chromium.org,alancutter@chromium.org
NOTREECHECKS=true
NOTRY=true
BUG=423424

Review URL: https://codereview.chromium.org/666433002

git-svn-id: svn://svn.chromium.org/blink/trunk@183877 bbb929c8-8fbe-4397-9dbb-9b2b20218538

BUG=422841
TBR=eseidel

Review URL: https://codereview.chromium.org/650103003

git-svn-id: svn://svn.chromium.org/blink/trunk@183876 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Start supporting style invalidation for selectors containing ::content,
but still use SubtreeStyleChange on distributed children of the content
element.

If you have a selector ".class ::content .target", we will not invalidate
the whole subtree for the element matching ".class", but do style
invalidation for elements matching ".target". We will not cross the
insertion point during invalidation at this point, but merely mark
content elements found during style invalidation with SubtreeStyleChange.

That may be done in a separate patch, but it requires the style
invalidator to have invalidation sets scheduled for elements inside a
shadow tree affect style invalidation of its host subtree.

The addition of :host-context as boundary-crossing and insertion-point-
crossing is in preparation for supporting style invalidation for the
:host-context selector.

This CL is split out of [1].
This CL relies on [2].

[1] https://codereview.chromium.org/639433002/
[2] https://codereview.chromium.org/637273003/

R=esprehn@chromium.org,chrishtr@chromium.org
BUG=424123

Review URL: https://codereview.chromium.org/652223002

git-svn-id: svn://svn.chromium.org/blink/trunk@183875 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Editable element check is not taken care of disabled and readonly elements
after getting the hittest result. Hence causing this issue. In this patch we
are checking whether the element is readonly or disabled and not allowing
paste popup to show.

BUG=

Review URL: https://codereview.chromium.org/246203004

git-svn-id: svn://svn.chromium.org/blink/trunk@183874 bbb929c8-8fbe-4397-9dbb-9b2b20218538

... but use proper layering instead.

PendingScript holds a Resource, and many places in the code are using both the
PendingScript and the underlying Resource. For example, ScriptRunner holds a
collection of PendingScripts, and ScriptLoaders then listen to the underlying
Resources. This is confusing and error prone.

The Resource - ScriptLoader - PendingScript - Element love quadrangle is messy
in general.  Both ScriptLoader and PendingScript refer to the Resource. The way
to get ScriptLoader from PendingScript is via Element which PendingScript also
holds. But there's no way to get PendingScript from ScriptLoader.

This CL makes ScriptRunner access the ScriptLoaders directly and moves the
PendingScript to ScriptLoader. In addition, it makes ScriptRunner not access the
Resource of the ScriptLoader directly.

This CL also cancels the hack from https://bugs.webkit.org/show_bug.cgi?id=92211
(r139942) where it was speculated that a crash can be caused by ScriptLoader
getting the notifyFinished notification twice because it doesn't unsubscribe yet
when it gets the first one. The speculative fix was to made it nullify
ScriptLoader::m_resource and check for the nullity later. This is pretty
surprising, since the obvious solution would've been to just unsubscribe early
enough (which is done by this CL).

Note a behavioral change: It's OK to call Resource::addClient even if the
Resource is finished, and in that case, the client will be notified
immediately. But PendingScript used to ASSERT that the Resource is not
finished. This CL removes that restriction (the underlying Resource already
handles adding clients to finished resources + made ScriptStreamer handle it
too).

Review URL: https://codereview.chromium.org/656113002

git-svn-id: svn://svn.chromium.org/blink/trunk@183873 bbb929c8-8fbe-4397-9dbb-9b2b20218538

Intend to replace the following typical for loop:

for (Node* child = node.firstChild(); child; child = child->nextSibling()) {
    ....
}

with range-based for loop of C++11:

for (Node& child : NodeTraversal::childrenOf(node)) {
    ....
}

Less verbose and improves readability, especially if we could close our eyes on NodeTraversal prefix.


The following iterators are introduced to address common use cases in our code base:

1. NodeTraversal::childrenOf(const Node&)

  This could replace:
    for (Node* child = node.firstChild(); child; child = child->nextSibling()) { ... }
  with:
    for (Node& child : NodeTraversal::childrenOf(node)) { ... }

2. NodeTraversal::descendantsOf(const Node&)

  This could replace:
    for (Node* node = root.firstChild(); node; node = NodeTraversal::next(*node, &root) { ... }
  with:
    for (Node& node : NodeTraversal::descendantsOf(root)) { ... }

3. NodeTraversal::inclusiveDescendantsOf(const Node&)

  This could replace:
    for (Node* node = &root; node; node = NodeTraversal::next(*node, &root) { ... }
  with:
    for (Node& node : NodeTraversal::inclusiveDescendantsOf(root)) { ... }

4. NodeTraversal::fromNext(const Node&)

  This could replace:
    for (Node* node = NodeTraversal::next(start); node; node = NodeTraversal::next(*node) { ... }
  with:
    for (Node& node : NodeTraversal::fromNext(start)) { ... }


There are also typed template versions for other element types:

5. Traversal<ElementType>::childrenOf(const Node&)

  This could replace:
    for (ElementType* child = Traversal<ElementType>::firstChild(node); child; child = Traversal<ElementType>::nextSibling(*child)) { ... }
  with:
    for (ElementType& child : Traversal<ElementType>::childrenOf(node)) { ... }

6. Traversal<ElementType>::descendantsOf(const Node&)

  This could replace:
    for (ElementType* node = Traversal<ElementType>::firstWithin(root); node; node = Traversal<ElementType>::next(*node, &root)) { ... }
  with:
    for (ElementType& node : Traversal<ElementType>::descendantsOf(root)) { ... }

7. Traversal<ElementType>::inclusiveDescendantsOf(const ElementType&)

  This could replace:
    for (ElementType* node = &root; node; node = Traversal<ElementType>::next(*node, &root)) { ... }
  with:
    for (ElementType& node : Traversal<ElementType>::inclusiveDescendantsOf(root)) { ... }

8. Traversal<ElementType>::fromNext(const Node&)

  This could replace:
    for (ElementType* node = Traversal<ElementType>::next(start); node; node = Traversal<ElementType>::next(*node) { ... }
  with:
    for (ElementType& node : Traversal<ElementType>::fromNext(start)) { ... }


New iterators should be zero overhead because compilers can make them all inlined in a release build.

As the first showcase, I let some files use this range-based for loops. This is not comprehensive so that this patch wouldn't get larger.
Other changes will be done in another patch.

We might want to have similar things for other classes, such as RenderObject, which can also be done in another patch.

BUG=none
TEST=no layout test failure

Review URL: https://codereview.chromium.org/642973003

git-svn-id: svn://svn.chromium.org/blink/trunk@183872 bbb929c8-8fbe-4397-9dbb-9b2b20218538