[GTK] Generate API documentation for GObject DOM bindings
[WebKit-https.git] / Source / WebCore / ChangeLog
1 2013-10-18  Carlos Garcia Campos  <cgarcia@igalia.com>
2
3         [GTK] Generate API documentation for GObject DOM bindings
4         https://bugs.webkit.org/show_bug.cgi?id=121538
5
6         Reviewed by Gustavo Noronha Silva.
7
8         * bindings/gobject/GNUmakefile.am: Add a explicit rule for all
9         .symbols file making them depend on the corresponding header file,
10         since the .symbols file is generated by the bindings generator.
11
12 2013-10-18  Mario Sanchez Prada  <mario.prada@samsung.com>
13
14         [ATK] Simplify implementation of atk_text_get_text
15         https://bugs.webkit.org/show_bug.cgi?id=122644
16
17         Reviewed by Chris Fleizach.
18
19         Simplified code so we only call textUnderElement() directly once
20         and only when needed. Also, moved the specific code for ColorWell
21         objects up to the beginning of that function, so we don't do any
22         additional efforts like computing text ranges in those cases.
23
24         No new tests are needed, just to make sure that the current layout
25         and unit tests are still passing, which they are.
26
27         * accessibility/atk/WebKitAccessibleInterfaceText.cpp:
28         (textForObject): Fixed a issue that got detected while working on
29         this patch, which was causing a '\n' to be artificially appended
30         at the end of text controls all the time.
31         (webkitAccessibleTextGetText): Simplified function.
32
33         * accessibility/AccessibilityRenderObject.cpp:
34         (WebCore::AccessibilityRenderObject::doAXStringForRange): Removed
35         the check that prevents from pass ranges exceeding the limits of
36         the element's text, since those will be checked anyway when
37         calling String::substring().
38
39 2013-10-18  Brendan Long  <b.long@cablelabs.com>
40
41         [GStreamer][GTK] Add GRefPtr::outPtr()
42         https://bugs.webkit.org/show_bug.cgi?id=122996
43
44         Reviewed by Philippe Normand.
45
46         No new tests because this is just simplifying existing code.
47
48         * platform/graphics/gstreamer/GRefPtrGStreamer.cpp: Add GRefPtr specialization for GstToc.
49         * platform/graphics/gstreamer/GRefPtrGStreamer.h: Same.
50         * platform/graphics/gstreamer/GStreamerGWorld.cpp: Use GRefPtr::outPtr() to simplify code.
51         (WebCore::GStreamerGWorld::enterFullscreen):
52         (WebCore::GStreamerGWorld::exitFullscreen):
53         (WebCore::GStreamerGWorld::removePlatformVideoSink):
54         * platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp: Same.
55         (WebCore::MediaPlayerPrivateGStreamer::notifyPlayerOfText):
56         (WebCore::MediaPlayerPrivateGStreamer::newTextSample):
57         (WebCore::MediaPlayerPrivateGStreamer::processTableOfContents):
58         * platform/graphics/gstreamer/MediaPlayerPrivateGStreamerBase.cpp: Same.
59         (WebCore::MediaPlayerPrivateGStreamerBase::currentVideoSinkCaps):
60         * platform/network/soup/ResourceHandleSoup.cpp: Same.
61         (WebCore::HostTLSCertificateSet::computeCertificateHash):
62
63 2013-10-17  Alexey Proskuryakov  <ap@apple.com>
64
65         Don't add an include for return type of a [Custom] function in generated bindings code
66         https://bugs.webkit.org/show_bug.cgi?id=122972
67
68         Reviewed by Sam Weinig.
69
70         * bindings/scripts/CodeGeneratorJS.pm: (GenerateImplementation): Don't.
71
72 2013-10-17  Andreas Kling  <akling@apple.com>
73
74         CTTE: RenderMathMLSpace always has a MathMLTextElement.
75         <https://webkit.org/b/122992>
76
77         The renderer is never anonymous and always has a corresponding
78         MathMLTextElement. Overload element() with a tighter return type.
79
80         Also marked the class FINAL and made most methods private.
81
82         Reviewed by Antti Koivisto.
83
84         * rendering/mathml/RenderMathMLSpace.cpp:
85         (WebCore::RenderMathMLSpace::RenderMathMLSpace):
86         (WebCore::RenderMathMLSpace::updateFromElement):
87         * rendering/mathml/RenderMathMLSpace.h:
88
89 2013-10-17  Tim Horton  <timothy_horton@apple.com>
90
91         PlatformCALayer constructor should take layer type as an argument
92         https://bugs.webkit.org/show_bug.cgi?id=122915
93
94         Reviewed by Anders Carlsson.
95
96         No new tests, just restoring old code.
97
98         Un-do part of the patch for bug 122915; we can't early-return
99         in the constructor, there's still more work to do.
100
101         * platform/graphics/ca/mac/PlatformCALayerMac.mm:
102         (PlatformCALayerMac::PlatformCALayerMac):
103
104 2013-10-17  Alexandru Chiculita  <achicu@adobe.com>
105
106         Web Inspector: [CSS Regions] Crash when highlighting a node of a flow with no regions
107         https://bugs.webkit.org/show_bug.cgi?id=122993
108
109         Reviewed by Joseph Pecoraro.
110
111         Test: inspector-protocol/dom/highlight-flow-with-no-region.html
112
113         Even if a named flow has no regions the content of the flow will still have renderer objects created.
114         Removed the assumption that all renderers inside a RenderFlowThread will always have an enclosing RenderRegion.
115
116         * inspector/InspectorOverlay.cpp:
117         (WebCore::buildObjectForRendererFragments):
118         (WebCore::InspectorOverlay::buildObjectForHighlightedNode):
119
120 2013-10-17  Andreas Kling  <akling@apple.com>
121
122         CTTE: RenderMathMLOperator always has a MathMLElement.
123         <https://webkit.org/b/122988>
124
125         Reviewed by Antti Koivisto.
126
127         The renderer is never anonymous and always has a corresponding
128         MathMLElement. Overload element() with a tighter return type.
129
130         Also marked the class FINAL and made most methods private.
131
132 2013-10-17  Nico Weber  <thakis@chromium.org>
133
134         Fix three bugs in the equals() implementations for css gradients.
135         https://bugs.webkit.org/show_bug.cgi?id=122987
136
137         Reviewed by Andreas Kling.
138
139         1. Linear gradients were considered equal if the first gradient has no x and y
140            position and the second has no x but does have y.
141         2. Same as 1, for radial gradients. (This doesn't happen in practice as
142            CSSParser::parseRadialGradient rejects such input, so no test for this case.)
143         3. Radial gradients without x and y position weren't considered equal even if
144            they were.
145
146         * css/CSSGradientValue.cpp:
147         (WebCore::CSSLinearGradientValue::equals):
148         (WebCore::CSSRadialGradientValue::equals):
149
150 2013-10-17  Antoine Quint  <graouts@apple.com>
151
152         Web Inspector: allow front-end to trigger the system beep sound to signal an error
153         https://bugs.webkit.org/show_bug.cgi?id=122955
154
155         Reviewed by Timothy Hatcher.
156
157         New beep() method exposed on InspectorFrontendHost calling into WebCore's systemBeep().
158
159         * inspector/InspectorFrontendHost.cpp:
160         (WebCore::InspectorFrontendHost::beep):
161         * inspector/InspectorFrontendHost.h:
162         * inspector/InspectorFrontendHost.idl:
163
164 2013-10-17  Anders Carlsson  <andersca@apple.com>
165
166         Remove PlatformCAAnimation::supportsValueFunction()
167         https://bugs.webkit.org/show_bug.cgi?id=122990
168
169         Reviewed by Tim Horton.
170
171         PlatformCAAnimation::supportsValueFunction always returns true now, so there's no need for it to exist anymore.
172
173         * platform/graphics/ca/GraphicsLayerCA.cpp:
174         (WebCore::GraphicsLayerCA::createTransformAnimationsFromKeyframes):
175         * platform/graphics/ca/PlatformCAAnimation.h:
176         * platform/graphics/ca/mac/PlatformCAAnimationMac.mm:
177         * platform/graphics/ca/win/PlatformCAAnimationWin.cpp:
178
179 2013-10-17  Andreas Kling  <akling@apple.com>
180
181         CTTE: RenderMathMLFenced always has a MathMLInlineContainerElement.
182         <https://webkit.org/b/122986>
183
184         This renderer is never anonymous and always has a corresponding
185         MathMLInlineContainerElement. Overload element() with a tighter
186         return type.
187
188         Also marked the class FINAL and made most methods private.
189
190         Reviewed by Anders Carlsson.
191
192 2013-10-17  Myles C. Maxfield  <mmaxfield@apple.com>
193
194         Comment AffineTransform::xScale() and yScale() to make their meanings clearer
195         https://bugs.webkit.org/show_bug.cgi?id=122981
196
197         Reviewed by Simon Fraser.
198
199         * platform/graphics/transforms/AffineTransform.h:
200
201 2013-10-17  Vivek Galatage  <vivek.vg@samsung.com>
202
203         Remove unnecessary check for RenderLayer and rename ensureLayer to createLayer in RenderLayerModelObject.
204         https://bugs.webkit.org/show_bug.cgi?id=122928
205
206         Reviewed by Darin Adler.
207
208         No new tests due to code refactoring.
209
210         RenderLayerModelObject::styleDidChange invokes the ensureLayer() only in case of !layer().
211         Again checking for layer existence would be deemed redundant in ensureLayer.
212         Replacing it with an ASSERT(!m_layer) and also renaming it to createLayer.
213
214         Blink review URL: https://codereview.chromium.org/27246003/
215
216         * rendering/RenderLayerModelObject.cpp:
217         (WebCore::RenderLayerModelObject::createLayer):
218         (WebCore::RenderLayerModelObject::styleDidChange):
219         * rendering/RenderLayerModelObject.h:
220
221 2013-10-17  Tim Horton  <timothy_horton@apple.com>
222
223         Remove PlatformCALayerMac workaround for <rdar://problem/7390716>
224         https://bugs.webkit.org/show_bug.cgi?id=122983
225
226         Reviewed by Simon Fraser.
227
228         Remove a workaround for a bug fixed in Lion.
229
230         * platform/graphics/ca/mac/PlatformCALayerMac.mm:
231         (PlatformCALayerMac::setSublayers):
232         (PlatformCALayerMac::removeAllSublayers):
233         (PlatformCALayerMac::adoptSublayers):
234
235 2013-10-17  Robert Hogan  <robert@webkit.org>
236
237         "border-collapse: collapse;" for table removes part of its border (was: Border disappears when close to some elements)
238         https://bugs.webkit.org/show_bug.cgi?id=8914
239
240         Reviewed by David Hyatt.
241
242         The table section's side of a collapsed border won't get painted if there are no cells there to paint it. So instead
243         of relying solely on cells to paint the collapsed border paint the appropriate section of the border if there's no
244         cell to take care of it.
245
246         Tests: fast/table/paint-section-borders-without-cells-rtl.html
247                fast/table/paint-section-borders-without-cells-vertical-lr-rtl.html
248                fast/table/paint-section-borders-without-cells-vertical-lr.html
249                fast/table/paint-section-borders-without-cells-vertical-rl.html
250                fast/table/paint-section-borders-without-cells.html
251
252         * rendering/RenderTableSection.cpp:
253         (WebCore::RenderTableSection::paintRowGroupBorder):
254         (WebCore::RenderTableSection::offsetLeftForRowGroupBorder):
255         (WebCore::RenderTableSection::offsetTopForRowGroupBorder):
256         (WebCore::RenderTableSection::verticalRowGroupBorderHeight):
257         (WebCore::RenderTableSection::horizontalRowGroupBorderWidth):
258         (WebCore::RenderTableSection::paintRowGroupBorderIfRequired):
259         (WebCore::physicalBorderForDirection):
260         (WebCore::RenderTableSection::paintObject):
261         * rendering/RenderTableSection.h:
262
263 2013-10-17  Andreas Kling  <akling@apple.com>
264
265         StyleRuleFoo::mutableProperties() should return a reference.
266         <https://webkit.org/b/122962>
267
268         The mutableProperties() functions always return objects, so make
269         them return MutableStylePropertySet&.
270
271         Also tweaked the StyleRuleCSSStyleDeclaration constructor to take
272         references to both the properties and the owner rule since both
273         are required.
274
275         Reviewed by Antti Koivisto.
276
277 2013-10-17  Hans Muller  <hmuller@adobe.com>
278
279         [CSS Shapes] Improve the performance of image valued shapes with large shape-margins
280         https://bugs.webkit.org/show_bug.cgi?id=122613
281
282         Reviewed by Andreas Kling.
283
284         The cost of computing the shape-margin boundary of an image-valued shape-outside
285         is now proportional to (2 * shape-margin + image.height) rather than
286         (2 * shape-margin * image.height). The performance improvement comes from skipping
287         sequences of rounded-rectangle intervals that will not contribute to the final
288         result. Each non-empty row in the original image contributes one rounded-rectangle
289         whose corner radius is shape-margin, height is 2 * shape-margin, and width is
290         2 * shape-margin plus the width of the limits of the intervals on the row.
291
292         Renamed private method RasterShape::getIntervals() to intervalsAt() to be a little
293         more consistent with WebKit naming conventions.
294
295         There are no new tests since is just an internal refactoring.
296
297         * rendering/shapes/RasterShape.cpp:
298         (WebCore::MarginIntervalGenerator::set): Changed the x1,x2 parameters to an IntShapeInterval.
299         (WebCore::RasterShapeIntervals::contains): Refactor for the getIntervals() => intervalsAt() rename.
300         (WebCore::RasterShapeIntervals::getIntervalX1Values): Ditto.
301         (WebCore::RasterShapeIntervals::getIncludedIntervals): Ditto.
302         (WebCore::RasterShapeIntervals::getExcludedIntervals): Ditto.
303         (WebCore::RasterShapeIntervals::computeShapeMarginIntervals): Performance tuning.
304         * rendering/shapes/RasterShape.h:
305         (WebCore::RasterShapeIntervals::intervalsAt): Renamed getIntervals().
306         (WebCore::RasterShapeIntervals::limitIntervalAt): Return the min/max limits of the intervals at Y.
307         * rendering/shapes/ShapeInterval.h:
308         (WebCore::ShapeInterval::isEmpty): Added.
309
310 2013-10-15  Philippe Normand  <pnormand@igalia.com>
311
312         [GTK] Add URLMediaStream in the build
313         https://bugs.webkit.org/show_bug.cgi?id=122833
314
315         Reviewed by Carlos Garcia Campos.
316
317         * GNUmakefile.am: Add mediastream/gstreamer in include directories list.
318         * GNUmakefile.list.am: Add DOMURLMediaStream files in the build.
319
320 2013-10-17  Andreas Kling  <akling@apple.com>
321
322         Use PassRef for constructing StylePropertySets.
323         <https://webkit.org/b/122948>
324
325         Make functions that construct StylePropertySets return PassRef
326         instead of PassRefPtr. Since they never return null, this gets rid
327         of the extra branch in ~PassRefPtr everywhere.
328
329         Also StyleRule* classes now hold a Ref<StylePropertySet>, codifying
330         the fact that they always have a property set.
331
332         Reviewed by Antti Koivisto.
333
334 2013-10-17  Andreas Kling  <akling@apple.com>
335
336         DataRef<T> should use Ref<T> internally.
337         <https://webkit.org/b/122953>
338
339         DataRef is used to hold RenderStyle substructures, and due to the
340         way style inheritance is implemented, DataRef will always point to
341         a live object.
342
343         Codify this by making DataRef::m_data a Ref, and making all methods
344         that create substructure objects return PassRef.
345
346         Reviewed by Antti Koivisto.
347
348 2013-10-17  Mihnea Ovidenie  <mihnea@adobe.com>
349
350         [CSS Regions] Anonymous nested regions
351         https://bugs.webkit.org/show_bug.cgi?id=119135
352
353         Reviewed by David Hyatt.
354
355         Tests: fast/regions/table-caption-as-region.html
356                fast/regions/table-cell-as-region.html
357
358         This patch allows any non-replaced block to behave like a region. When an element is styled with the
359         -webkit-flow-from property, instead of making the renderer a RenderRegion, we let the original
360         renderer be created the same way and we add a region as an anonymous child for the renderer.
361         The anonymous block child, modeled by the new RenderNamedFlowFragment class, will be responsible
362         for the fragmentation of the named flow thread content.
363
364         A RenderBlockFlow object will keep a reference to a RenderNamedFlowFragment(RenderRegion) inside its
365         rare data structures.
366
367         Contains code contributed by Catalin Badea.
368  
369         * CMakeLists.txt:
370         * GNUmakefile.list.am:
371         * WebCore.vcxproj/WebCore.vcxproj:
372         * WebCore.xcodeproj/project.pbxproj:
373         * dom/Element.cpp: Changed to take the anonymous region into account.
374         (WebCore::Element::renderRegion):
375         (WebCore::Element::webkitGetRegionFlowRanges):
376         * dom/WebKitNamedFlow.cpp: Ditto.
377         (WebCore::WebKitNamedFlow::firstEmptyRegionIndex):
378         (WebCore::WebKitNamedFlow::getRegionsByContent):
379         (WebCore::WebKitNamedFlow::getRegions):
380         * inspector/InspectorOverlay.cpp: Take into account the new model for regions, with an anonymous region inside a block.
381         (WebCore::buildObjectForRegionHighlight):
382         (WebCore::buildObjectForElementInfo):
383         * rendering/RenderBlock.cpp:
384         (WebCore::RenderBlock::computeShapeSize): For a render named flow fragment, there is no need to recompute the shape inside
385         we can take it from the parent.
386         (WebCore::RenderBlock::renderName): Make the block that contains a render named flow fragment (region) report RenderRegion.
387         A future patch that will change this will need to rebase a lot of tests.
388         * rendering/RenderBlockFlow.cpp:
389         (WebCore::RenderBlockFlow::insertedIntoTree): Create the anonymous region if needed (change of -webkit-flow-from determines Node reattach).
390         (WebCore::RenderBlockFlow::willBeDestroyed): Clean-up the anonymous region if necessary.
391         (WebCore::RenderBlockFlow::clearFloats): Small style change to make sure that check-webkit-style reports 0 failures on RenderBlockFlow.cpp.
392         (WebCore::RenderBlockFlow::layoutBlock):
393         (WebCore::RenderBlockFlow::styleDidChange): Update the style of the anonymous region too.
394         (WebCore::RenderBlockFlow::createRenderNamedFlowFragmentIfNeeded): Helper function to create the anonymous region
395         and to add it as a child to the block.
396         (WebCore::RenderBlockFlow::canHaveChildren):
397         (WebCore::RenderBlockFlow::canHaveGeneratedChildren):
398         (WebCore::RenderBlockFlow::namedFlowFragmentNeedsUpdate): Force a layout of the anonymous region if the
399         parent block has percentage height (similar to RenderBlock::updateBlockChildDirtyBitsBeforeLayout)
400         (WebCore::RenderBlockFlow::updateLogicalHeight): Update the logical height of anonymous region when the height of parent is updated.
401         (WebCore::RenderBlockFlow::setRenderNamedFlowFragment):
402         * rendering/RenderBlockFlow.h:
403         (WebCore::RenderBlockFlow::RenderBlockFlowRareData::RenderBlockFlowRareData):
404         (WebCore::RenderBlockFlow::renderNamedFlowFragment):
405         * rendering/RenderElement.cpp:
406         (WebCore::RenderElement::createFor): Remove the direct creation of RenderRegion objects since they will be
407         created as anonymous children of block flow objects.
408         * rendering/RenderElement.h:
409         (WebCore::RenderElement::generatingElement): Account for anonymous region if necessary.
410         * rendering/RenderFlowThread.cpp:
411         (WebCore::RenderFlowThread::adjustedPositionRelativeToOffsetParent): Take anonymous region into account
412         and use the anonymous region parent offset/border.
413         * rendering/RenderLayer.cpp:
414         (WebCore::RenderLayer::shouldBeNormalFlowOnly): Check for style instead of isRenderRegion since
415         the parent of the anonymous region will get the layer.
416         * rendering/RenderListItem.cpp:
417         (WebCore::RenderListItem::insertedIntoTree): Call RenderBlockFlow method instead.
418         * rendering/RenderNamedFlowFragment.cpp: Added. Model the behaviour of the anonymous region.
419         Has RenderRegion as a base class.
420         (WebCore::RenderNamedFlowFragment::RenderNamedFlowFragment):
421         (WebCore::RenderNamedFlowFragment::~RenderNamedFlowFragment):
422         (WebCore::RenderNamedFlowFragment::setStyleForNamedFlowFragment):
423         (WebCore::RenderNamedFlowFragment::styleDidChange):
424         (WebCore::RenderNamedFlowFragment::shouldHaveAutoLogicalHeight):
425         (WebCore::RenderNamedFlowFragment::maxPageLogicalHeight):
426         * rendering/RenderNamedFlowFragment.h: Added.
427         (WebCore::RenderNamedFlowFragment::isPseudoElementRegion):
428         (WebCore::RenderNamedFlowFragment::renderName):
429         (WebCore::toRenderNamedFlowFragment):
430         * rendering/RenderObject.cpp:
431         (WebCore::RenderObject::isRenderNamedFlowFragmentContainer):
432         * rendering/RenderObject.h:
433         (WebCore::RenderObject::isRenderNamedFlowFragment):
434         * rendering/RenderRegion.h:
435         * rendering/RenderTableCaption.cpp: Call RenderBlockFlow method instead.
436         (WebCore::RenderTableCaption::insertedIntoTree):
437         * rendering/RenderTreeAsText.cpp:
438         (WebCore::write):
439         (WebCore::writeRenderRegionList):
440         * rendering/shapes/ShapeInfo.h:
441         (WebCore::ShapeInfo::shapeSize):
442         * style/StyleResolveTree.cpp:
443         (WebCore::Style::elementInsideRegionNeedsRenderer):
444
445 2013-10-17  Krzysztof Czech  <k.czech@samsung.com>
446
447         [EFL] Properly expose tables in accessibility
448         https://bugs.webkit.org/show_bug.cgi?id=122894
449
450         Reviewed by Chris Fleizach.
451
452         All tables should exposed as tables.
453
454         * accessibility/AccessibilityTable.cpp:
455         (WebCore::AccessibilityTable::isTableExposableThroughAccessibility):
456         (WebCore::AccessibilityTable::addChildren):
457         * accessibility/AccessibilityTableColumn.cpp:
458         (WebCore::AccessibilityTableColumn::computeAccessibilityIsIgnored):
459         * accessibility/AccessibilityTableHeaderContainer.cpp:
460         (WebCore::AccessibilityTableHeaderContainer::computeAccessibilityIsIgnored):
461
462 2013-10-17  Brendan Long  <b.long@cablelabs.com>
463
464         [GStreamer] Too many arguments for format in WebKitWebAudioSourceGStreamer.cpp
465         https://bugs.webkit.org/show_bug.cgi?id=122932
466
467         Reviewed by Philippe Normand.
468
469         No new tests because this just fixes a build warning.
470
471         * platform/audio/gstreamer/WebKitWebAudioSourceGStreamer.cpp:
472         (webKitWebAudioSrcLoop): Add another %s for the second part of the pad name.
473
474 2013-10-16  Tim Horton  <timothy_horton@apple.com>
475
476         Attempt to fix the Windows build after http://trac.webkit.org/changeset/157547.
477
478         The relevant conversion doesn't happen implicitly.
479
480         * platform/graphics/win/MediaPlayerPrivateFullscreenWindow.cpp:
481         (WebCore::MediaPlayerPrivateFullscreenWindow::setRootChildLayer):
482         (WebCore::MediaPlayerPrivateFullscreenWindow::wndProc):
483
484 2013-10-16  Tim Horton  <timothy_horton@apple.com>
485
486         Attempt to fix the Windows build after http://trac.webkit.org/changeset/157547.
487
488         Apparently some Windows code uses setFrame and I missed it.
489
490         * platform/graphics/win/MediaPlayerPrivateFullscreenWindow.cpp:
491         (WebCore::MediaPlayerPrivateFullscreenWindow::setRootChildLayer):
492         (WebCore::MediaPlayerPrivateFullscreenWindow::wndProc):
493
494 2013-10-16  Tim Horton  <timothy_horton@apple.com>
495
496         Remote Layer Tree: Complete support for simple layer properties
497         https://bugs.webkit.org/show_bug.cgi?id=122933
498
499         Reviewed by Anders Carlsson.
500
501         No new tests, not yet testable.
502
503         * platform/graphics/ca/GraphicsLayerCA.cpp:
504         (WebCore::GraphicsLayerCA::setName):
505         Don't dump the CALayer pointer if we own a PlatformCALayerRemote.
506
507         (WebCore::GraphicsLayerCA::recursiveCommitChanges):
508         Fix the visible tile wash (my fault!), and make it use setPosition and
509         setBounds instead of setFrame; while more convenient, it is the only
510         caller of setFrame, so we'll remove it.
511
512         * platform/graphics/ca/PlatformCALayer.h:
513         (WebCore::PlatformCALayer::isRemote): Added.
514
515         * platform/graphics/ca/mac/PlatformCALayerMac.h:
516         * platform/graphics/ca/mac/PlatformCALayerMac.mm:
517         (nullActionsDictionary):
518         (toCAFilterType):
519         (PlatformCALayerMac::synchronouslyDisplayTilesInRect):
520         (PlatformCALayerMac::playerLayer):
521         Remove setFrame, fix some pointer sides.
522
523         * platform/graphics/ca/win/PlatformCALayerWin.cpp:
524         * platform/graphics/ca/win/PlatformCALayerWin.h:
525         Remove setFrame.
526
527 2013-10-16  Andreas Kling  <akling@apple.com>
528
529         Take RenderObjects out of the arena.
530         <https://webkit.org/b/122895>
531
532         Reviewed by Antti Koivisto.
533
534         Stop arena-allocating renderers so we can move forward on improving
535         render tree memory management. This will also allow rendering code
536         to take advantage of malloc optimizations.
537
538         Line boxes and BiDi runs remain in the arena for now.
539
540 2013-10-16  Roger Fong  <roger_fong@apple.com>
541
542         [Windows] Speculative fix for test, media/video-canvas-drawing-output.html.
543
544         * platform/graphics/avfoundation/cf/MediaPlayerPrivateAVFoundationCF.cpp:
545         (WebCore::AVFWrapper::createImageForTimeInRect):
546
547 2013-10-16  Tim Horton  <timothy_horton@apple.com>
548
549         PlatformCALayer constructor should take layer type as an argument
550         https://bugs.webkit.org/show_bug.cgi?id=122915
551
552         Reviewed by Simon Fraser.
553
554         No new tests, just a minor refactoring.
555
556         * platform/graphics/ca/PlatformCALayer.h:
557         (WebCore::PlatformCALayer::PlatformCALayer):
558         Add a LayerType argument.
559
560         * platform/graphics/ca/mac/PlatformCALayerMac.mm:
561         (PlatformCALayerMac::PlatformCALayerMac):
562         * platform/graphics/ca/win/PlatformCALayerWin.cpp:
563         (PlatformCALayerWin::PlatformCALayerWin):
564         Use the new LayerType argument, and early-return in the case where we
565         are wrapping a custom PlatformLayer.
566         Drive-by un-indent the switch in the Mac version.
567
568 2013-10-15  Brady Eidson  <beidson@apple.com>
569
570         Flesh out the DatabaseProcess (and launch it!)
571         https://bugs.webkit.org/show_bug.cgi?id=122884
572
573         Reviewed by Tim Horton.
574
575         * English.lproj/Localizable.strings: Add a localizable string.
576         * WebCore.xcodeproj/project.pbxproj: Export some more headers.
577
578 2013-10-16  Tim Horton  <timothy_horton@apple.com>
579
580         RemoteLayerTree: Add support for more layer properties and transform layers
581         https://bugs.webkit.org/show_bug.cgi?id=122906
582
583         Reviewed by Anders Carlsson.
584
585         No new tests, this code is not yet testable.
586
587         * WebCore.exp.in:
588         Export some TextStream and TransformationMatrix stuff.
589
590 2013-10-16  KyungTae Kim  <ktf.kim@samsung.com>
591
592         During editing, merge inline style with overriding other author styles
593         https://bugs.webkit.org/show_bug.cgi?id=122874
594
595         Reviewed by Ryosuke Niwa.
596
597         Inline styles need to override other author styles even on DoNotOverrideValues mode.
598         So, merge and override inline styles to other author styles before merging them to m_mutableStyle.
599
600         Test: editing/deleting/merge-div-with-inline-style.html
601
602         * editing/EditingStyle.cpp:
603         (WebCore::EditingStyle::mergeInlineAndImplicitStyleOfElement):
604
605 2013-10-16  peavo@outlook.com  <peavo@outlook.com>
606
607         Emphasis marks has wrong color.
608         https://bugs.webkit.org/show_bug.cgi?id=122829
609
610         Reviewed by Antti Koivisto.
611
612         Tests: fast/text/text-emphasis.html.
613                fast/text/text-emphasis-expected.html.
614
615         Emphasis color should be set as fill color, not stroke color.
616
617         * rendering/TextPaintStyle.cpp:
618         (WebCore::updateGraphicsContext):
619         * rendering/TextPaintStyle.h:
620
621 2013-10-16  Antti Koivisto  <antti@apple.com>
622
623         Move code for finding rendered character offset to RenderTextLineBoxes
624         https://bugs.webkit.org/show_bug.cgi?id=122892
625
626         Reviewed by Andreas Kling.
627
628         * rendering/RenderText.cpp:
629         (WebCore::RenderText::countRenderedCharacterOffsets):
630         (WebCore::RenderText::containsRenderedCharacterOffset):
631         
632             Renamed for consistency.
633
634         * rendering/RenderTextLineBoxes.cpp:
635         (WebCore::RenderTextLineBoxes::countCharacterOffsets):
636         
637             This used to be Position::renderedPosition.
638
639 2013-10-16  Andreas Kling  <akling@apple.com>
640
641         RenderElement::removeChild() should take child as a reference.
642         <https://webkit.org/b/122888>
643
644         We can't remove a child without a child to remove.
645
646         Reviewed by Antti Koivisto.
647
648 2013-10-16  Antti Koivisto  <antti@apple.com>
649
650         Move test for contained caret offset to RenderTextLineBoxes
651         https://bugs.webkit.org/show_bug.cgi?id=122887
652
653         Reviewed by Andreas Kling.
654
655         * dom/Position.cpp:
656         (WebCore::Position::renderedOffset):
657         (WebCore::Position::isCandidate):
658         
659             Remove isRenderedText, call RenderText::containsCaretOffset instead.
660
661         (WebCore::Position::isRenderedCharacter):
662         (WebCore::Position::rendersInDifferentPosition):
663         * dom/Position.h:
664         * dom/PositionIterator.cpp:
665         (WebCore::PositionIterator::isCandidate):
666         * rendering/InlineTextBox.cpp:
667         * rendering/InlineTextBox.h:
668         * rendering/RenderText.cpp:
669         (WebCore::RenderText::containsCharacterOffset):
670         (WebCore::RenderText::containsCaretOffset):
671         * rendering/RenderText.h:
672         * rendering/RenderTextLineBoxes.cpp:
673         (WebCore::RenderTextLineBoxes::containsOffset):
674         
675             Combined implementations of Position::isRenderedCharacter and Position::isRenderedText.
676
677         * rendering/RenderTextLineBoxes.h:
678         
679             Remove containsCaretOffset(), functionality is now in RenderTextLineBoxes::containsOffset.
680
681 2013-10-16  Andreas Kling  <akling@apple.com>
682
683         RenderElement::isChildAllowed() should take const references.
684         <https://webkit.org/b/122870>
685
686         Reviewed by Anders Carlsson.
687
688         The isChildAllowed() functions expect non-null values to be passed,
689         so enforce this at compile-time.
690
691         Reordered some checks to do bit tests before virtual calls.
692
693 2013-10-15  Philippe Normand  <pnormand@igalia.com>
694
695         [GStreamer] video info unset if upstream doesn't query allocation
696         https://bugs.webkit.org/show_bug.cgi?id=122834
697
698         Reviewed by Gustavo Noronha Silva.
699
700         * platform/graphics/gstreamer/VideoSinkGStreamer.cpp:
701         (webkitVideoSinkRender): If the sink didn't process any allocation
702         query then use the configured source pad caps and don't rely on
703         invalid video info.
704
705 2013-10-16  Philippe Normand  <pnormand@igalia.com>
706
707         [GStreamer] move Logging.h include to GStreamerUtilities.h
708         https://bugs.webkit.org/show_bug.cgi?id=122886
709
710         Reviewed by Gustavo Noronha Silva.
711
712         Include Logging.h from GStreamerUtilities.h so the modules using
713         the LOG_MEDIA macros don't need to bother, especially for Debug builds.
714
715         * platform/graphics/gstreamer/GStreamerUtilities.h:
716         * platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp:
717         * platform/graphics/gstreamer/MediaPlayerPrivateGStreamerBase.cpp:
718
719 2013-10-16  Mark Rowe  <mrowe@apple.com>
720
721         Fix the build after r157478.
722
723         Rubber-stamped by Tim Horton.
724
725         Due to the way WebCore.exp.in is used, it can't be used to export a differing set of symbols
726         for different architectures. We often work around this by tweaking code slightly to avoid
727         needing to export different symbols. However, in this case the symbol name itself encodes an
728         architecture-specific detail and there's no clear way to avoid the requirement to export it.
729
730         To deal with this case we turn to ld's support for wildcards in the symbol export list.
731
732         * WebCore.exp.in: Use wildcards in place of the number that represents by how much "this"
733         should be adjusted when calling through the vtable thunk. Also sort the remainder of the file.
734         * make-export-file-generator: Don't attempt to verify symbol names that contain wildcard characters.
735
736 2013-10-15  Tim Horton  <timothy_horton@apple.com>
737
738         Two more exports for 32-bit build fix.
739         The duality of CGFloat means that we use a different
740         version of getRGBA (and the cast operator)
741         on 32-bit systems.
742
743         * WebCore.exp.in:
744
745 2013-10-15  Tim Horton  <timothy_horton@apple.com>
746
747         Another shot at a build fix; apparently these
748         do need exporting for some reason, but are different
749         on 32-bit.
750
751         * WebCore.exp.in:
752
753 2013-10-15  Alexey Proskuryakov  <ap@apple.com>
754
755         GenerateIsReachable=ImplContext is confusing
756         https://bugs.webkit.org/show_bug.cgi?id=122864
757
758         Reviewed by Geoffrey Garen.
759
760         Renamed to ImplWebGLRenderingContext.
761
762         * bindings/scripts/CodeGeneratorJS.pm:
763         (GenerateImplementation):
764         * bindings/scripts/IDLAttributes.txt:
765         * html/canvas/EXTDrawBuffers.idl:
766         * html/canvas/EXTTextureFilterAnisotropic.idl:
767         * html/canvas/OESElementIndexUint.idl:
768         * html/canvas/OESStandardDerivatives.idl:
769         * html/canvas/OESTextureFloat.idl:
770         * html/canvas/OESTextureHalfFloat.idl:
771         * html/canvas/OESVertexArrayObject.idl:
772         * html/canvas/WebGLCompressedTextureATC.idl:
773         * html/canvas/WebGLCompressedTexturePVRTC.idl:
774         * html/canvas/WebGLCompressedTextureS3TC.idl:
775         * html/canvas/WebGLDebugRendererInfo.idl:
776         * html/canvas/WebGLDebugShaders.idl:
777         * html/canvas/WebGLDepthTexture.idl:
778         * html/canvas/WebGLLoseContext.idl:
779
780         * WebCore.xcodeproj/project.pbxproj: While at it, added OESElementIndexUint.idl
781         to Xcode project.
782
783 2013-10-15  Tim Horton  <timothy_horton@apple.com>
784
785         Unreviewed build fix; I don't know how to export.
786         This may not help.
787
788         * WebCore.exp.in:
789
790 2013-10-15  Dean Jackson  <dino@apple.com>
791
792         Add ENABLE_WEB_ANIMATIONS flag
793         https://bugs.webkit.org/show_bug.cgi?id=122871
794
795         Reviewed by Tim Horton.
796
797         Eventually might be http://dev.w3.org/fxtf/web-animations/
798         but this is just engine-internal work at the moment.
799
800         * Configurations/FeatureDefines.xcconfig:
801
802 2013-10-15  Tim Horton  <timothy_horton@apple.com>
803
804         Add a PlatformCALayer subclass that proxies its property changes across the process boundary
805         https://bugs.webkit.org/show_bug.cgi?id=122773
806
807         Reviewed by Anders Carlsson.
808
809         No new tests, the new drawing area is not yet testable.
810
811         * WebCore.exp.in:
812         Export lots of GraphicsLayerCA stuff so we can inherit from it in WebKit2.
813
814         * WebCore.xcodeproj/project.pbxproj:
815         Make PlatformCAFilters.h a private header.
816
817         * platform/graphics/GraphicsLayer.h:
818         (WebCore::GraphicsLayer::initialize):
819         * platform/graphics/ca/GraphicsLayerCA.cpp:
820         (WebCore::GraphicsLayer::create):
821         (WebCore::GraphicsLayerCA::GraphicsLayerCA):
822         (WebCore::GraphicsLayerCA::initialize):
823         * platform/graphics/ca/GraphicsLayerCA.h:
824         Defer creation of the main PlatformCALayer until just after the constructor is finished
825         so that GraphicsLayerCA subclasses can successfully override createPlatformCALayer.
826
827         * platform/graphics/ca/PlatformCALayer.h:
828         (WebCore::PlatformCALayer::platformLayer):
829         Make platformLayer virtual so that subclasses which don't have PlatformLayers can override.
830
831 2013-10-14  Ryosuke Niwa  <rniwa@webkit.org>
832
833         REGRESSION: Crash in XMLDocumentParser::startElementNs
834         https://bugs.webkit.org/show_bug.cgi?id=122817
835
836         Reviewed by Darin Adler.
837
838         Exit early in startElementNs when listeners and handlers of synchronous events such as load event
839         removes the inserted node inside parserAppendChild.
840
841         Test: fast/parser/xhtml-synchronous-detach-crash.html
842
843         * xml/parser/XMLDocumentParserLibxml2.cpp:
844         (WebCore::XMLDocumentParser::startElementNs):
845
846 2013-10-15  Joseph Pecoraro  <pecoraro@apple.com>
847
848         Web Inspector: Remove old frontend localizedStrings.js
849         https://bugs.webkit.org/show_bug.cgi?id=122846
850
851         Reviewed by Timothy Hatcher.
852
853         * Configurations/WebCore.xcconfig:
854         * English.lproj/localizedStrings.js: Removed.
855         * WebCore.xcodeproj/project.pbxproj:
856         Remove the file and references to it. We no longer need to exclude
857         localizedString.js from some builds.
858
859 2013-10-15  Joseph Pecoraro  <pecoraro@apple.com>
860
861         Web Inspector: Remove Windows old front-end related code
862         https://bugs.webkit.org/show_bug.cgi?id=122845
863
864         Reviewed by Brent Fulgham.
865
866         * WebCore.vcxproj/copyWebCoreResourceFiles.cmd:
867
868 2013-10-15  Morten Stenshorne  <mstensho@opera.com>
869
870         Add support for the column-fill property
871         https://bugs.webkit.org/show_bug.cgi?id=117693
872
873         Reviewed by David Hyatt.
874
875         This is only supported in the (new) region based multicol implementation.
876
877         With column-fill support, a lot of multicol tests needed an update.
878         The old implementation behaved as if column-fill were 'auto', but the
879         initial value is 'balance', so now we need to be explicit about that.
880         For auto-height tests it doesn't really matter - such multicols are always
881         balanced anyway.
882
883         Tests: fast/multicol/newmulticol/fixed-height-fill-auto.html
884                fast/multicol/newmulticol/fixed-height-fill-balance.html
885
886         * css/CSSComputedStyleDeclaration.cpp:
887         (WebCore::ComputedStyleExtractor::propertyValue):
888         * css/CSSParser.cpp:
889         (WebCore::isValidKeywordPropertyAndValue):
890         (WebCore::isKeywordPropertyID):
891         (WebCore::CSSParser::parseValue):
892         * css/CSSPrimitiveValueMappings.h:
893         (WebCore::CSSPrimitiveValue::CSSPrimitiveValue):
894         (WebCore::CSSPrimitiveValue::operator ColumnFill):
895         * css/CSSPropertyNames.in:
896         * css/CSSValueKeywords.in:
897         * css/DeprecatedStyleBuilder.cpp:
898         (WebCore::DeprecatedStyleBuilder::DeprecatedStyleBuilder):
899         * rendering/RenderMultiColumnBlock.h:
900         * rendering/style/RenderStyle.h:
901         * rendering/style/RenderStyleConstants.h:
902         * rendering/style/StyleMultiColData.cpp:
903         (WebCore::StyleMultiColData::StyleMultiColData):
904         (WebCore::StyleMultiColData::operator==):
905         * rendering/style/StyleMultiColData.h:
906
907 2013-10-15  Andreas Kling  <akling@apple.com>
908
909         Skip unnecessary null check in RenderText::textLength().
910         <https://webkit.org/b/122841>
911
912         Reviewed by Antti Koivisto.
913
914         RenderText will never have a null String in m_text, so textLength()
915         can grab at the StringImpl directly, avoiding a null check.
916
917 2013-10-15  Andreas Kling  <akling@apple.com>
918
919         FontGenericFamilies should not be ref-counted.
920         <https://webkit.org/b/122835>
921
922         Reviewed by Anders Carlsson.
923
924         FontGenericFamilies is singly-owned by Settings.
925
926 2013-10-15  Zoltan Horvath  <zoltan@webkit.org>
927
928         [CSS Shapes] Move RenderBlock::layoutShapeInsideInfo into RenderBlock.cpp
929         http://bugs.webkit.org/show_bug.cgi?id=122843
930
931         Reviewed by Oliver Hunt.
932
933         Historically, layoutShapeInsideInfo was a static function in RenderBlockLineLayout, then it has changed to be a member of RenderBlock,
934         but at that time it hasn't been moved to RenderBlock.cpp. This patch moves it into RenderBlock.cpp next to the Shapes functions. I removed
935         an unnecessary CSS_SHAPES #ifdef as well from RenderBlock.cpp.
936
937         No new tests, no behavior change.
938
939         * rendering/RenderBlock.cpp:
940         (WebCore::RenderBlock::markShapeInsideDescendantsForLayout):
941         (WebCore::RenderBlock::layoutShapeInsideInfo):
942         * rendering/RenderBlockLineLayout.cpp:
943
944 2013-10-15  peavo@outlook.com  <peavo@outlook.com>
945
946         [WinCairo] Build fails.
947         https://bugs.webkit.org/show_bug.cgi?id=122830
948
949         Reviewed by Brent Fulgham.
950
951         * platform/network/NetworkStorageSessionStub.cpp:
952         (WebCore::NetworkStorageSession::createPrivateBrowsingSession): Update to new return type.
953
954 2013-10-15  Andreas Kling  <akling@apple.com>
955
956         FileIconLoader should not be ref-counted.
957         <https://webkit.org/b/122827>
958
959         FileIconLoader is singly-owned by FileInputType.
960
961         Reviewed by Antti Koivisto.
962
963 2013-10-15  Andreas Kling  <akling@apple.com>
964
965         RenderText should cache RenderStyle in locals more.
966         <https://webkit.org/b/122823>
967
968         Reviewed by Antti Koivisto.
969
970         Now that fetching the RenderStyle has to go through the parent,
971         we should avoid unnecessary loads by caching style() in a local.
972
973 2013-10-15  Csaba Osztrogon√°c  <ossy@webkit.org>
974
975         Move PlatformCertificateInfo to WebCore and make the ResourceResponse primitives work in terms of that platform agnostic object
976         https://bugs.webkit.org/show_bug.cgi?id=118520
977
978         Reviewed by Anders Carlsson.
979
980         Added PlatformCertificateInfo getter and setter to ResourceErrorBase
981         and ResourceResponseBase so that getting and setting certificates
982         becomes cross platform.
983
984         Changed the existing platform specific certificate getters and setters
985         of ResourceError and ResourceResponse to use the
986         PlatformCertificateInfo member.
987
988         Original patch by Kwang Yul Seo  <skyul@company100.net>
989
990         The following things were fixed by Csaba Osztrogon√°c:
991         - trivial conflicts resolved in WebCore.exp.in, project.pbxproj, NetworkResourceLoader.cpp,
992           WebPageProxy.h, WebFrameLoaderClient.cpp and PlatformEfl.cmake.
993         - trivial fix in the new AsynchronousNetworkLoaderClient.cpp
994         - style fixed (NULL -> 0 and smaller indentation in WebCoreArgumentCodersSoup.cpp)
995         - obsolete change removed from NetworkResourceLoader.cpp
996         - unneeded typo removed from ResourceHandleSoup.cpp
997         - Fixed the if guard of including RetainPtr.h in PlatformCertificateInfo.h to make Windows build happy.
998         - add PlatformCertificateInfoCFNet.cpp with empty constructor and destructor to make Windows build happy.
999         - resolved conflict in Source/WebCore/WebCore.xcodeproj/project.pbxproj after r156488
1000         - renamed KURL to URL after r156550.
1001         - removed obsolete Qt changes
1002         - added back initializers for m_soupFlags
1003         - included gio/gio.h instead of libsoup/soup.h in PlatformCertificateInfo.h
1004         - removed unused class forward declarations and includes
1005
1006         * GNUmakefile.list.am:
1007         * PlatformEfl.cmake:
1008         * Target.pri:
1009         * WebCore.exp.in:
1010         * WebCore.vcxproj/WebCore.vcxproj:
1011         * WebCore.vcxproj/WebCore.vcxproj.filters:
1012         * WebCore.xcodeproj/project.pbxproj:
1013         * platform/network/PlatformCertificateInfo.h: Renamed from Source/WebKit2/Shared/soup/PlatformCertificateInfo.h.
1014         (WebCore::PlatformCertificateInfo::certificate):
1015         (WebCore::PlatformCertificateInfo::setCertificate):
1016         (WebCore::PlatformCertificateInfo::tlsErrors):
1017         (WebCore::PlatformCertificateInfo::setTLSErrors):
1018         * platform/network/ResourceErrorBase.h:
1019         (WebCore::ResourceErrorBase::platformCertificateInfo):
1020         (WebCore::ResourceErrorBase::setPlatformCertificateInfo):
1021         * platform/network/ResourceResponseBase.h:
1022         (WebCore::ResourceResponseBase::platformCertificateInfo):
1023         (WebCore::ResourceResponseBase::setPlatformCertificateInfo):
1024         * platform/network/cf/ResourceResponse.h:
1025         * platform/network/mac/PlatformCertificateInfoMac.mm: Renamed from Source/WebKit2/Shared/mac/PlatformCertificateInfo.mm.
1026         (WebCore::PlatformCertificateInfo::PlatformCertificateInfo):
1027         (WebCore::PlatformCertificateInfo::~PlatformCertificateInfo):
1028         (WebCore::PlatformCertificateInfo::setCertificateChain):
1029         (WebCore::PlatformCertificateInfo::certificateChain):
1030         (WebCore::PlatformCertificateInfo::dump):
1031         * platform/network/mac/ResourceResponseMac.mm:
1032         (WebCore::ResourceResponse::setCertificateChain):
1033         (WebCore::ResourceResponse::certificateChain):
1034         * platform/network/soup/PlatformCertificateInfoSoup.cpp:
1035         (WebCore::PlatformCertificateInfo::PlatformCertificateInfo):
1036         (WebCore::PlatformCertificateInfo::~PlatformCertificateInfo):
1037         * platform/network/soup/ResourceError.h:
1038         (WebCore::ResourceError::ResourceError):
1039         (WebCore::ResourceError::tlsErrors):
1040         (WebCore::ResourceError::setTLSErrors):
1041         (WebCore::ResourceError::certificate):
1042         (WebCore::ResourceError::setCertificate):
1043         * platform/network/soup/ResourceErrorSoup.cpp:
1044         (WebCore::ResourceError::platformCopy):
1045         * platform/network/soup/ResourceResponse.h:
1046         (WebCore::ResourceResponse::ResourceResponse):
1047         (WebCore::ResourceResponse::soupMessageCertificate):
1048         (WebCore::ResourceResponse::setSoupMessageCertificate):
1049         (WebCore::ResourceResponse::soupMessageTLSErrors):
1050         (WebCore::ResourceResponse::setSoupMessageTLSErrors):
1051         * platform/network/soup/ResourceResponseSoup.cpp:
1052         (WebCore::ResourceResponse::toSoupMessage):
1053         (WebCore::ResourceResponse::updateFromSoupMessage):
1054
1055 2013-10-15  Ryosuke Niwa  <rniwa@webkit.org>
1056
1057         Remove redundant Document::getElementById
1058         https://bugs.webkit.org/show_bug.cgi?id=122813
1059
1060         Reviewed by Andreas Kling.
1061
1062         Merge https://chromium.googlesource.com/chromium/blink/+/4e8f1c5316415614b84370c602beae4a1008299f
1063
1064         This function simply calls virtual TreeScope::getElementById and Document inherits from TreeScope.
1065
1066         * WebCore.exp.in:
1067         * dom/Document.cpp:
1068         * dom/Document.h:
1069
1070 2013-10-14  Santosh Mahto  <santosh.ma@samsung.com>
1071
1072         in safari,the background-color of input[type="search"] can't work
1073         https://bugs.webkit.org/show_bug.cgi?id=119967
1074
1075         Reviewed by Ryosuke Niwa.
1076
1077         When input type=search is styled with css background property then
1078         it does not change the background-color of field. Its happening becasue
1079         search field is not counted as styled control. Thus theme ignores the 
1080         css background property. With this patch search field is also counted as 
1081         styled control so background property reflects on search field.
1082
1083         Test: fast/forms/search/search-field-background-color.html
1084
1085         * rendering/RenderTheme.cpp:
1086         (WebCore::RenderTheme::isControlStyled):Now search field is also
1087         a styled control.
1088
1089 2013-10-14  Ryosuke Niwa  <rniwa@webkit.org>
1090
1091         EventPath::updateTouchLists traverses through EventPath thrice
1092         https://bugs.webkit.org/show_bug.cgi?id=122804
1093
1094         Reviewed by Benjamin Poulain.
1095
1096         Instead of traversing through EventPath for each TouchList, traverse through TouchList for every EventContext.
1097         This paves our way to have one-pass traversal over EventPath, and evetually to remove EventContext altogether.
1098
1099         This change should also improve the cache hit rate since all Touch objects tend to be allocated at the same time
1100         but this performance improvement is probably not observable.
1101
1102         * dom/EventContext.h:
1103         * dom/EventDispatcher.cpp:
1104         (WebCore::EventRelatedNodeResolver::EventRelatedNodeResolver): Added a new constructor that takes Touch and
1105         and TouchListType. We need to store these two values in order to update EventContext later.
1106         (WebCore::EventRelatedNodeResolver::touch): Added,
1107         (WebCore::EventRelatedNodeResolver::touchListType): Added.
1108         (WebCore::addRelatedNodeResolversForTouchList): Extracted from updateTouchListsInEventPath.
1109         (WebCore::EventPath::updateTouchLists): Moved the loop over m_path here. Notice that the outer loop iterates
1110         over m_path instead of touchList as done in updateTouchListsInEventPath. The inner loop goes through resolvers
1111         and adds Touch objects each EventContext as needed.
1112
1113 2013-10-14  Alexey Proskuryakov  <ap@apple.com>
1114
1115         Don't generate a wasteful isObservable check in isReachableFromOpaqueRoots
1116         https://bugs.webkit.org/show_bug.cgi?id=122802
1117
1118         Reviewed by Mark Hahnenberg.
1119
1120         * bindings/scripts/CodeGeneratorJS.pm: (GenerateImplementation): Don't.
1121
1122         * bindings/scripts/test/JS/JSTestActiveDOMObject.cpp:
1123         * bindings/scripts/test/JS/JSTestCustomNamedGetter.cpp:
1124         * bindings/scripts/test/JS/JSTestEventConstructor.cpp:
1125         * bindings/scripts/test/JS/JSTestEventTarget.cpp:
1126         * bindings/scripts/test/JS/JSTestException.cpp:
1127         * bindings/scripts/test/JS/JSTestInterface.cpp:
1128         * bindings/scripts/test/JS/JSTestMediaQueryListListener.cpp:
1129         * bindings/scripts/test/JS/JSTestNamedConstructor.cpp:
1130         * bindings/scripts/test/JS/JSTestObj.cpp:
1131         * bindings/scripts/test/JS/JSTestOverloadedConstructors.cpp:
1132         * bindings/scripts/test/JS/JSTestSerializedScriptValueInterface.cpp:
1133         * bindings/scripts/test/JS/JSTestTypedefs.cpp:
1134         * bindings/scripts/test/JS/JSattribute.cpp:
1135         * bindings/scripts/test/JS/JSreadonly.cpp:
1136         Updated results.
1137
1138 2013-10-14  Samuel White  <samuel_white@apple.com>
1139
1140         AX: fieldset should have GroupRole and legend should be description.
1141         https://bugs.webkit.org/show_bug.cgi?id=122534
1142
1143         Reviewed by Chris Fleizach.
1144
1145         Changes fieldset to derive AXDescription from legend if one is available. Added
1146         convenience method to AccessibilityObject to fetch element if available.
1147
1148         Test: accessibility/fieldset-element.html
1149
1150         * accessibility/AccessibilityNodeObject.cpp:
1151         (WebCore::AccessibilityNodeObject::canHaveChildren):
1152         (WebCore::AccessibilityNodeObject::alternativeText):
1153         * accessibility/AccessibilityObject.cpp:
1154         (WebCore::AccessibilityObject::element):
1155         (WebCore::AccessibilityObject::isARIAHidden):
1156         (WebCore::AccessibilityObject::isDOMHidden):
1157         (WebCore::AccessibilityObject::defaultObjectInclusion):
1158         * accessibility/AccessibilityObject.h:
1159         (WebCore::AccessibilityObject::isHidden):
1160         * accessibility/AccessibilitySlider.cpp:
1161         (WebCore::AccessibilitySlider::getAttribute):
1162         (WebCore::AccessibilitySlider::valueForRange):
1163         (WebCore::AccessibilitySlider::maxValueForRange):
1164         (WebCore::AccessibilitySlider::minValueForRange):
1165         (WebCore::AccessibilitySlider::setValue):
1166         (WebCore::AccessibilitySlider::inputElement):
1167         * accessibility/AccessibilitySlider.h:
1168         * accessibility/mac/AccessibilityObjectMac.mm:
1169         (WebCore::AccessibilityObject::accessibilityPlatformIncludesObject):
1170         * html/HTMLFieldSetElement.cpp:
1171         (WebCore::HTMLFieldSetElement::legend):
1172         * html/HTMLFieldSetElement.h:
1173
1174 2013-10-14  Roger Fong  <roger_fong@apple.com>
1175
1176         Windows select element doesn't draw RTL properly.
1177         https://bugs.webkit.org/show_bug.cgi?id=122785.
1178
1179         Reviewed by Brent Fulgham.
1180
1181         Covered by fast/text/international/pop-up-button-text-alignment-and-direction.html.
1182
1183         Problems include the popup items not drawing on the right hand side and 
1184         not respecting the direction or the directional override styling of the option.
1185         The selected element (drawn in the actual select element) also doesn't respect 
1186         the style settings of the selected menu option.
1187
1188         * platform/win/PopupMenuWin.cpp:
1189         (WebCore::PopupMenuWin::paint):
1190
1191 2013-10-14  Roger Fong  <roger_fong@apple.com>
1192
1193         [Windows] Unreviewed build fix.
1194
1195         * WebCore.vcxproj/WebCoreCommon.props:
1196
1197 2013-10-14  Ryosuke Niwa  <rniwa@webkit.org>
1198
1199         Crash in WebCore::BidiResolver<WebCore::InlineIterator, WebCore::BidiRun>::createBidiRunsForLine
1200         https://bugs.webkit.org/show_bug.cgi?id=122776
1201
1202         Reviewed by Darin Adler.
1203
1204         Merge https://chromium.googlesource.com/chromium/blink/+/aca89bc4d984705a1f94b623dae0ab03e239a248
1205
1206         Fix modification of whitespace endpoints to not assume it's operating on RenderTexts
1207
1208         During line layout, we use midpoints to identify RenderObjects, or parts of
1209         RenderObjects, that don't need InlineBoxes, usually because of collapsed whitespace.
1210
1211         Prior to actually creating BidiRuns (the precursor to InlineBoxes), we use
1212         checkMidpoints to fix up our lineMidpointState to handle the case where we start
1213         ignoring spaces in our line, but don't stop until somewhere on the following line.
1214         Previously, this function assumed that the final midpoint (called an endpoint)
1215         was a RenderText, but this assumption is wrong if we have a beginning midpoint
1216         created by shouldSkipWhitespaceAfterStartObject (which handles inlines and list
1217         markers) and no endpoint on that line. In that case, we'd instead adjust the
1218         position backwards on the beginning midpoint, which would cause us to fail to
1219         create an InlineBox for the inline or list marker. In the new test added, this
1220         would actually trigger a crash due to an assumption when visually re-ordering
1221         BidiRuns that a non-empty line would actually contain at least one such run.
1222
1223         Test: fast/text/whitespace/whitespace-and-margin-wrap-after-list-marker-crash.html
1224
1225         * rendering/RenderBlockLineLayout.cpp:
1226         (WebCore::checkMidpoints):
1227
1228 2013-10-14  Ryosuke Niwa  <rniwa@webkit.org>
1229
1230         Assertion failure in Range::processContentsBetweenOffsets
1231         https://bugs.webkit.org/show_bug.cgi?id=122777
1232
1233         Reviewed by Darin Adler.
1234
1235         Merge https://chromium.googlesource.com/chromium/blink/+/c15de182774c7859c20d97126eb844ae97b792a4
1236
1237         This patch changes ASSERT statements for checking |endOffset| inbound in Range::processContentsBetweenOffsets()
1238         to limit |endOffset|. This is necessary when DOMNodeRemovedFromDocument event handler splits text nodes,
1239         Range::insertNode() on text node, in the range calling Range::deleteContents().
1240
1241         Test: fast/dom/Range/range-delete-contents-mutation-event-crash.html
1242
1243         * dom/Range.cpp:
1244         (WebCore::Range::processContentsBetweenOffsets):
1245
1246 2013-10-14  Alexey Proskuryakov  <ap@apple.com>
1247
1248         Add an empty window.crypto.webkitSubtle
1249         https://bugs.webkit.org/show_bug.cgi?id=122778
1250
1251         Reviewed by Mark Hahnenberg.
1252
1253         Tests: security/crypto-subtle-gc-2.html
1254                security/crypto-subtle-gc-3.html
1255                security/crypto-subtle-gc.html
1256
1257         * DerivedSources.make: Process SubtleCrypto.idl.
1258
1259         * crypto: Added.
1260         * WebCore.xcodeproj/project.pbxproj:
1261         * CMakeLists.txt:
1262         * DerivedSources.make:
1263         * GNUmakefile.am:
1264         * GNUmakefile.list.am:
1265         * WebCore.vcxproj/WebCore.vcxproj.filters:
1266
1267         * bindings/js/JSSubtleCryptoCustom.cpp: Added. Empty for now, but we'll certainly
1268         need custom bindings code here.
1269
1270         * crypto/SubtleCrypto.cpp: Added.
1271         (WebCore::SubtleCrypto::SubtleCrypto):
1272         (WebCore::SubtleCrypto::document):
1273         * crypto/SubtleCrypto.h: Added.        
1274         * crypto/SubtleCrypto.idl: Added.
1275         * page/Crypto.cpp:
1276         (WebCore::Crypto::subtle):
1277         * page/Crypto.h:
1278         * page/Crypto.idl:
1279
1280 2013-10-14  Nick Diego Yamane  <nick.yamane@openbossa.org>
1281
1282         Remove GestureEvent leftovers from WebCore
1283         <https://webkit.org/b/122780>
1284
1285         Reviewed by Anders Carlsson.
1286
1287         - Removed some remaining references to PlatformGestureEvent, supposed to
1288           be removed by r157316
1289         - TOUCH_ADJUSTMENT should be reworked after GestureEvent feature
1290           removal
1291
1292         * page/EventHandler.cpp:
1293         (WebCore::EventHandler::bestZoomableAreaForTouchPoint):
1294         * page/EventHandler.h:
1295         * platform/PlatformEvent.h:
1296         * platform/ScrollAnimatorNone.cpp:
1297         * platform/ScrollableArea.h:
1298
1299 2013-10-14  Nick Diego Yamane  <nick.yamane@openbossa.org>
1300
1301         Build fix after r157366
1302         http://bugs.webkit.org/show_bug.cgi?id=122783
1303
1304         When TOUCH_AJUSTMENT is enabled build fails due to some
1305         refactors in TextRender functions.
1306
1307         Reviewed by Anders Carlsson.
1308
1309         * page/TouchAdjustment.cpp:
1310         (WebCore::TouchAdjustment::appendContextSubtargetsForNode):
1311
1312 2013-10-14  Alexandru Chiculita  <achicu@adobe.com>
1313
1314         The content of the DOM panel for iframes is not updated until the "onload" event
1315         https://bugs.webkit.org/show_bug.cgi?id=122653
1316
1317         Reviewed by Darin Adler.
1318
1319         Test: http/tests/inspector-protocol/loading-iframe-document-node.html
1320
1321         Renamed InspectorDOMAgent::loadEventFired to InspectorDOMAgent::didCommitLoad and moved the call site
1322         from InspectorInstrumentation::loadEventFiredImpl to InspectorInstrumentation::didCommitLoadImpl.
1323         This is to make sure that it will invalidate the content of the iframe as soon as the frame navigates
1324         to a different page. This way the new node can be retrieved as soon as the page has some content, and
1325         not just when the page is fully loaded.
1326
1327         * inspector/InspectorDOMAgent.cpp:
1328         (WebCore::InspectorDOMAgent::didCommitLoad): Renamed from loadEventFired, as it is now called from
1329         didCommitLoadImpl instead.
1330         (WebCore::InspectorDOMAgent::frameDocumentUpdated): Updated comment to point to the new function name.
1331         * inspector/InspectorDOMAgent.h:
1332         * inspector/InspectorInstrumentation.cpp:
1333         (WebCore::InspectorInstrumentation::loadEventFiredImpl): Removed call do InspectorDOMAgent.loadEventFired.
1334         (WebCore::InspectorInstrumentation::didCommitLoadImpl): Added call to InspectorDOMAgent.didCommitLoad.
1335
1336 2013-10-14  Roger Fong  <roger_fong@apple.com>
1337
1338         https://bugs.webkit.org/show_bug.cgi?id=122774.
1339         <rdar://problem/6138855>.
1340
1341         Reviewed by Brent Fulgham.
1342
1343         Add a field to keep track of hovered over index.
1344         Use index to determine whether or not to use the existing selected index on the mouse down event.
1345
1346         * platform/win/PopupMenuWin.cpp:
1347         (WebCore::PopupMenuWin::PopupMenuWin):
1348         (WebCore::PopupMenuWin::show):
1349         (WebCore::PopupMenuWin::wndProc):
1350         * platform/win/PopupMenuWin.h:
1351
1352 2013-10-14  Tim Horton  <timothy_horton@apple.com>
1353
1354         Virtualize PlatformCALayer
1355         https://bugs.webkit.org/show_bug.cgi?id=122672
1356
1357         Reviewed by Anders Carlsson.
1358
1359         No new tests, just a refactoring.
1360
1361         * WebCore.exp.in:
1362         setGeometryFlipped is on PlatformCALayerMac now.
1363
1364         * WebCore.vcxproj/WebCore.vcxproj:
1365         * WebCore.vcxproj/WebCore.vcxproj.filters:
1366         Add PlatformCALayer.cpp, PlatformCALayerWin.h, and let VS do its
1367         thing with some other files.
1368
1369         * WebCore.xcodeproj/project.pbxproj:
1370         Add PlatformCALayer.cpp and PlatformCALayerMac.h.
1371
1372         * platform/graphics/avfoundation/cf/MediaPlayerPrivateAVFoundationCF.cpp:
1373         (WebCore::AVFWrapper::platformLayer):
1374         Make a PlatformCALayerWin explicitly.
1375
1376         * platform/graphics/ca/GraphicsLayerCA.cpp:
1377         (WebCore::GraphicsLayerCA::createPlatformCALayer):
1378         Added. Decide whether to make a PlatformCALayer{Mac, Win} based on the platform.
1379         Later, we will decide between other subclasses based on other things.
1380
1381         (WebCore::GraphicsLayerCA::filtersCanBeComposited):
1382         Do the same thing for filtersCanBeComposited.
1383
1384         (WebCore::GraphicsLayerCA::GraphicsLayerCA):
1385         (WebCore::GraphicsLayerCA::setContentsToSolidColor):
1386         (WebCore::GraphicsLayerCA::setContentsToMedia):
1387         (WebCore::GraphicsLayerCA::setContentsToCanvas):
1388         (WebCore::GraphicsLayerCA::recursiveCommitChanges):
1389         (WebCore::GraphicsLayerCA::ensureStructuralLayer):
1390         (WebCore::GraphicsLayerCA::updateContentsImage):
1391         (WebCore::GraphicsLayerCA::updateContentsRects):
1392         (WebCore::GraphicsLayerCA::swapFromOrToTiledLayer):
1393         Use createPlatformCALayer instead of PlatformCALayer::create.
1394
1395         * platform/graphics/ca/mac/PlatformCAFiltersMac.mm:
1396         For now, use the PlatformCALayerMac version of filtersCanBeComposited,
1397         since this code is heavily tied to having CALayers in the Web process.
1398
1399         * platform/graphics/ca/GraphicsLayerCA.h:
1400         Include PlatformCALayer.h here so we can get the LayerType enum.
1401         (NOTE-to-be-removed: if there's a better way to do this, I'm open to
1402         suggestions; I couldn't puzzle out nested 'enum class' stuff).
1403
1404         Add the createPlatformCALayers.
1405
1406         * platform/graphics/ca/PlatformCAAnimation.h:
1407         Friend the subclasses too.
1408
1409         * platform/graphics/ca/PlatformCALayer.cpp: Added.
1410         (WebCore::PlatformCALayer::~PlatformCALayer):
1411         Pull the shared part of the PlatformCALayer destructor out.
1412
1413         * platform/graphics/ca/PlatformCALayer.h:
1414         (WebCore::PlatformCALayer::platformLayer):
1415         (WebCore::PlatformCALayer::setOwner):
1416         (WebCore::PlatformCALayer::PlatformCALayer):
1417         Virtualize all the things. Move platform specific members to their new subclasses.
1418
1419         * platform/graphics/ca/mac/PlatformCALayerMac.h: Added.
1420         * platform/graphics/ca/mac/PlatformCALayerMac.mm:
1421         Move the PLATFORM(MAC) implementations from PlatformCALayer to PlatformCALayerMac.
1422
1423         (PlatformCALayer::platformCALayer):
1424         The platformCALayer lookup function is static on PlatformCALayer, so it can't
1425         be moved to the subclasses. It might be a good idea in the future to move towards
1426         a platform-independent mechanism for looking up PlatformCALayers from PlatformLayers,
1427         and to avoid needing to do this as often as we do now.
1428
1429         * platform/graphics/ca/win/CACFLayerTreeHost.cpp:
1430         (WebCore::CACFLayerTreeHost::CACFLayerTreeHost):
1431         * platform/graphics/ca/win/PlatformCALayerWin.cpp:
1432         Move the PLATFORM(WIN) implementations from PlatformCALayer to PlatformCALayerWin.
1433
1434         (PlatformCALayerWin::create):
1435         (PlatformCALayer::platformCALayer):
1436
1437         * platform/graphics/ca/win/PlatformCALayerWin.h: Added.
1438
1439         * platform/graphics/win/GraphicsContext3DWin.cpp:
1440         (WebCore::GraphicsContext3D::GraphicsContext3D):
1441         * platform/graphics/win/MediaPlayerPrivateQuickTimeVisualContext.cpp:
1442         (WebCore::MediaPlayerPrivateQuickTimeVisualContext::createLayerForMovie):
1443         Include PlatformCALayerWin.h and explicitly make PlatformCALayerWins here.
1444
1445 2013-10-14  Hugo Parente Lima  <hugo.lima@openbossa.org>
1446
1447         [cmake] MediaControlsApple is used only by Efl port and is on CMakeLists.txt
1448         https://bugs.webkit.org/show_bug.cgi?id=122772
1449
1450         Reviewed by Anders Carlsson.
1451
1452         * CMakeLists.txt: Removed MediaControlsApple.cpp
1453         * PlatformEfl.cmake: Added MediaControlsApple.cpp
1454
1455 2013-10-14  Alexey Proskuryakov  <ap@apple.com>
1456
1457         window.crypto doesn't preserve custom properties
1458         https://bugs.webkit.org/show_bug.cgi?id=122770
1459
1460         Reviewed by Mark Hahnenberg.
1461
1462         Test: security/crypto-gc.html
1463
1464         Generate isReachableFromOpaqueRoots that makes Crypto live as long as the document
1465         lives (because that's when it's observable through window object).
1466
1467         * page/Crypto.cpp:
1468         (WebCore::Crypto::Crypto):
1469         (WebCore::Crypto::~Crypto):
1470         (WebCore::Crypto::document):
1471         * page/Crypto.h:
1472         (WebCore::Crypto::create):
1473         Made Crypto a ContextDestructionObserver, so that it can report its document to bindings.
1474         Removed ScriptWrappable, because it seems to have served no purpose in this class.
1475
1476         * page/Crypto.idl: Added GenerateIsReachable. Removed ImplementationLacksVTable,
1477         because the class now has a vtable, and can be checked for bindings integrity.
1478
1479         * page/DOMWindow.cpp: (WebCore::DOMWindow::crypto): Pass a document when creating
1480         crypto.
1481
1482 2013-10-14  Andreas Kling  <akling@apple.com>
1483
1484         CTTE: NamedNodeMap always has a corresponding Element.
1485         <https://webkit.org/b/122769>
1486
1487         Reviewed by Anders Carlsson.
1488
1489         Made NamedNodeMap::m_element a reference and remove an assertion
1490         that it's never null.
1491
1492 2013-10-14  Andreas Kling  <akling@apple.com>
1493
1494         REGRESSION(r157408): Crashes in RenderFullScreen::wrapRenderer().
1495
1496         Unreviewed crash fix for these two tests:
1497
1498         - fullscreen/full-screen-restrictions.html
1499         - fullscreen/empty-anonymous-block-continuation-crash.html
1500
1501         * rendering/RenderFullScreen.cpp:
1502         (RenderFullScreen::wrapRenderer):
1503
1504             Get the RenderArena from Document like we did before this patch.
1505
1506 2013-10-14  Hans Muller  <hmuller@adobe.com>
1507
1508         [CSS Shapes] Image valued shape-outside shapes should update the layout after the image has been loaded
1509         https://bugs.webkit.org/show_bug.cgi?id=122340
1510
1511         Reviewed by Simon Fraser.
1512
1513         Ensure that the an image-valued shape-outside layout is updated after the image has
1514         been loaded.
1515
1516         Test: http/tests/css/css-image-valued-shape.html
1517
1518         * rendering/RenderBlock.cpp:
1519         (WebCore::RenderBlock::imageChanged): Added code for the shape-outside case.
1520         (WebCore::RenderBlock::updateShapeInsideInfoAfterStyleChange): Ditto.
1521         * rendering/RenderElement.cpp:
1522         (WebCore::RenderElement::~RenderElement): Ditto.
1523         (WebCore::RenderElement::setStyle): Ditto.
1524
1525 2013-10-14  Andreas Kling  <akling@apple.com>
1526
1527         Remove some silly null checks in Element/NamedNodeMap.
1528         <https://webkit.org/b/122767>
1529
1530         Reviewed by Darin Adler.
1531
1532         Make shouldIgnoreAttributeCase() take a const Element&, exposing
1533         some unnecessary null checks.
1534
1535 2013-10-14  Brent Fulgham  <bfulgham@apple.com>
1536
1537         [Win] Build fix after r122737.
1538
1539         * dom/Node.h: Add explicit WebCore namespace to macro definition to work around
1540         Visual Studio bug.
1541
1542 2013-10-14  Andreas Kling  <akling@apple.com>
1543
1544         Pass Document directly to anonymous renderer constructors.
1545         <https://webkit.org/b/122752>
1546
1547         Reviewed by Antti Koivisto.
1548
1549         Added separate constructors for creating anonymous renderers that
1550         take a Document& instead of a null Element*/Text*.
1551
1552         Removed setDocumentForAnonymous() and all createAnonymous() helpers.
1553         ...and RenderObject::m_node is now a Node&, wohoo!
1554
1555 2013-10-13  Sam Weinig  <sam@webkit.org>
1556
1557         CTTE: Add more node conversion helpers
1558         https://bugs.webkit.org/show_bug.cgi?id=122737
1559
1560         Reviewed by Darin Adler.
1561
1562         - Factor NODE_TYPE_CASTS into TYPE_CASTS_BASE(ToClassName, FromClassName)
1563           to allow for DOCUMENT_TYPE_CASTS.
1564         - Replace more static_casts<>.
1565
1566 2013-10-14  Zan Dobersek  <zdobersek@igalia.com>
1567
1568         Reintroduce PassRefPtr<Event> copy in ScopedEventQueue::dispatchEvent
1569         https://bugs.webkit.org/show_bug.cgi?id=122742
1570
1571         Reviewed by Alexey Proskuryakov.
1572
1573         This is a follow-up to r157219 which introduced a workaround for the GCC's quirky behavior that
1574         was resulting in crashes due to the PassRefPtr<Event> object passed to EventDispatcher::dispatchEvent
1575         being copied and nullified first before retrieving the EventTarget of the Event object wrapped in that
1576         PassRefPtr.
1577
1578         The implementation is now adjusted to first retrieve the pointer to the Event's EventTarget and store
1579         it in a local variable. That variable is then passed as the first parameter to EventDispatcher::dispatchEvent,
1580         and the PassRefPtr<Event> passed directly as the second parameter. Previously the pointer of that PassRefPtr
1581         object was passed in, with a new PassRefPtr being created which would increase the reference count of the
1582         ref-counted object. Passing in the original PassRefPtr avoids the unnecessary reference count increase by creating
1583         a copy. That still nullifies the original PassRefPtr, but that's not a problem anymore.
1584
1585         * dom/ScopedEventQueue.cpp:
1586         (WebCore::ScopedEventQueue::dispatchEvent):
1587
1588 2013-10-14  Bear Travis  <betravis@adobe.com>
1589
1590         [CSS Shapes] Shape-Margin should be animatable
1591         https://bugs.webkit.org/show_bug.cgi?id=122524
1592
1593         Reviewed by Darin Adler.
1594
1595         Mark content for relayout after shape-margin changes, and add shape-margin
1596         to the list of animatable properties.
1597
1598         Tests: fast/shapes/shape-outside-floats/shape-outside-dynamic-shape-margin.html
1599                fast/shapes/shape-outside-floats/shape-outside-shape-margin-animation.html
1600
1601         * page/animation/CSSPropertyAnimation.cpp:
1602         (WebCore::CSSPropertyAnimationWrapperMap::CSSPropertyAnimationWrapperMap): Add
1603         shape-margin to the map of animatable CSS properties.
1604         * rendering/RenderBox.cpp:
1605         (WebCore::RenderBox::updateShapeOutsideInfoAfterStyleChange): Compare shape-margins,
1606         and mark dependent content for relayout if they have changed.
1607
1608 2013-10-14  Arvid Nilsson  <anilsson@blackberry.com>
1609
1610         Don't crash after OpenGL robustness reset
1611         https://bugs.webkit.org/show_bug.cgi?id=122750
1612
1613         Reviewed by George Staikos.
1614
1615         JIRA 517132.
1616         Just log the incident and pretend like nothing happened.
1617
1618         No new tests, we don't have repeatable steps to reproduce a robustness
1619         reset.
1620
1621         * platform/graphics/blackberry/LayerRenderer.cpp:
1622         (WebCore::LayerRenderer::makeContextCurrent):
1623
1624 2013-10-14  Krzysztof Czech  <k.czech@samsung.com>
1625
1626         [EFL] Buildfix after r157393
1627         https://bugs.webkit.org/show_bug.cgi?id=122749
1628
1629         Reviewed by Andreas Kling.
1630
1631         Buildfix with error enumeration value 'CSS_FR' not handled in switch.
1632
1633         * css/CSSCalculationValue.cpp:
1634         (WebCore::hasDoubleValue):
1635
1636 2013-10-14  Krzysztof Czech  <k.czech@samsung.com>
1637
1638         [EFL] Present replaced objects with 0xFFFC character
1639         https://bugs.webkit.org/show_bug.cgi?id=122744
1640
1641         Reviewed by Mario Sanchez Prada.
1642
1643         Replaced elements should be emitted in GTK/EFL and
1644         marked their presence with the replacement character.
1645
1646         * accessibility/AccessibilityObject.cpp:
1647         (WebCore::AccessibilityObject::textIteratorBehaviorForTextRange):
1648
1649 2013-09-16  Sergio Villar Senin  <svillar@igalia.com>
1650
1651         [CSS Grid Layout] Implement support for <flex>
1652         https://bugs.webkit.org/show_bug.cgi?id=115362
1653
1654         Reviewed by Andreas Kling.
1655
1656         From Blink r149134, r149480, r149532, r150287 and r156127 by <jchaffraix@chromium.org>
1657         From Blink r157820 by <svillar@igalia.com>
1658
1659         Added support for flexible lengths ('fr' unit) in CSS Grid Layout
1660         code. This requires the addition of GridLength class to
1661         encapsulate the knowledge of <flex> to the grid layout code.
1662
1663         Also updated the algorithm that computes the layout. It increases
1664         the value of 1fr based on the grid tracks' usedBreath to fraction
1665         ratio (called normalizedFractionValue). This enables increasing
1666         the fraction value while updating the available space to account
1667         for processed grid tracks. The algorithm stops when we run out of
1668         grid tracks or available space (one grid item has an intrinsic
1669         size too big). This matches the specs to the letter for the known
1670         available space case (both the unknown case and the interaction
1671         with 'span' are left out of this patch).
1672
1673         Tests: fast/css-grid-layout/flex-and-minmax-content-resolution-columns.html
1674                fast/css-grid-layout/flex-and-minmax-content-resolution-rows.html
1675                fast/css-grid-layout/flex-content-resolution-columns.html
1676                fast/css-grid-layout/flex-content-resolution-rows.html
1677
1678         * GNUmakefile.list.am: Added GridLength.h to the build system.
1679         * Target.pri: Ditto.
1680         * WebCore.vcxproj/WebCore.vcxproj: Ditto.
1681         * WebCore.vcxproj/WebCore.vcxproj.filters: Ditto.
1682         * WebCore.xcodeproj/project.pbxproj: Ditto.
1683         * css/CSSComputedStyleDeclaration.cpp:
1684         (WebCore::valueForGridTrackBreadth): Replace Length by GridLength.
1685         * css/CSSGrammar.y.in: Added FR support.
1686         * css/CSSParser.cpp: Ditto.
1687         (WebCore::CSSParser::parseGridBreadth):
1688         (WebCore::CSSParser::detectNumberToken):
1689         * css/CSSParserValues.cpp: Added FR support.
1690         (WebCore::CSSParserValue::createCSSValue):
1691         * css/CSSParserValues.h:
1692         (WebCore::CSSParserString::operator[]):
1693         (WebCore::CSSParserString::equalIgnoringCase):
1694         * css/CSSPrimitiveValue.cpp: Added FR support.
1695         (WebCore::isValidCSSUnitTypeForDoubleConversion):
1696         (WebCore::CSSPrimitiveValue::cleanup):
1697         (WebCore::CSSPrimitiveValue::customCSSText):
1698         (WebCore::CSSPrimitiveValue::cloneForCSSOM):
1699         (WebCore::CSSPrimitiveValue::equals):
1700         * css/CSSPrimitiveValue.h: Added a couple of missing const.
1701         (WebCore::CSSPrimitiveValue::isFlex):
1702         * css/StyleResolver.cpp: Added FR support.
1703         (WebCore::createGridTrackBreadth):
1704         (WebCore::createGridTrackSize):
1705         * rendering/RenderGrid.cpp:
1706         (WebCore::GridTrackForNormalization::GridTrackForNormalization):
1707         New helper struct to ease the computation of track breadths with
1708         flexible lengths.
1709         (WebCore::GridTrackForNormalization::operator=):
1710         (WebCore::RenderGrid::computePreferredTrackWidth): Replaced Length by GridLength.
1711         (WebCore::RenderGrid::computedUsedBreadthOfGridTracks): Grow grid lines
1712         having a fraction as the MaxTrackSizingFunction.
1713         (WebCore::RenderGrid::computeUsedBreadthOfMinLength): Replaced Length by GridLength.
1714         (WebCore::RenderGrid::computeUsedBreadthOfMaxLength): Ditto.
1715         (WebCore::sortByGridNormalizedFlexValue):
1716         (WebCore::RenderGrid::computeNormalizedFractionBreadth): Increase
1717         the fraction value while updating the available space to account
1718         for processed grid tracks.
1719         (WebCore::RenderGrid::resolveContentBasedTrackSizingFunctions):
1720         (WebCore::RenderGrid::distributeSpaceToTracks): Never shrink track sizes.
1721         (WebCore::RenderGrid::tracksAreWiderThanMinTrackBreadth):
1722         * rendering/RenderGrid.h:
1723         * rendering/style/GridLength.h: Added.
1724         (WebCore::GridLength::GridLength):
1725         (WebCore::GridLength::isLength):
1726         (WebCore::GridLength::isFlex):
1727         (WebCore::GridLength::length):
1728         (WebCore::GridLength::flex):
1729         (WebCore::GridLength::setFlex):
1730         (WebCore::GridLength::operator==):
1731         * rendering/style/GridTrackSize.h: Replaced Length by GridLength.
1732         (WebCore::GridTrackSize::length):
1733         (WebCore::GridTrackSize::setLength):
1734         (WebCore::GridTrackSize::minTrackBreadth):
1735         (WebCore::GridTrackSize::maxTrackBreadth):
1736         (WebCore::GridTrackSize::setMinMax):
1737         (WebCore::GridTrackSize::hasMinOrMaxContentMinTrackBreadth):
1738         (WebCore::GridTrackSize::hasMaxContentMinTrackBreadth):
1739         (WebCore::GridTrackSize::hasMinOrMaxContentMaxTrackBreadth):
1740         (WebCore::GridTrackSize::hasMaxContentMaxTrackBreadth):
1741
1742 2013-10-14  peavo@outlook.com  <peavo@outlook.com>
1743
1744         Broken text rendering when input field has selection.
1745         https://bugs.webkit.org/show_bug.cgi?id=122716
1746
1747         Reviewed by Antti Koivisto.
1748
1749         Tests: fast/text/text-rendering-with-input-selection.html.
1750                fast/text/text-rendering-with-input-selection-expected.html.
1751
1752         * rendering/InlineTextBox.cpp:
1753         (WebCore::InlineTextBox::paint): Check that text has selection.
1754
1755 2013-10-14  Zalan Bujtas  <zalan@apple.com>
1756
1757         Unexpected word wrapping for wrapped content then raw content.
1758         https://bugs.webkit.org/show_bug.cgi?id=121130
1759
1760         Reviewed by Antti Koivisto.
1761
1762         When deciding whether a line is considered empty, we need to check if the current
1763         object is fully responsible for the currently uncommitted width. It helps differentiating
1764         <span></span><span>abcd</span> from <span>a</span><span>bcd</span>, where in the first
1765         case when we hit the second <span> the line is still considered empty, as opposed to the
1766         second example.
1767
1768         Test: fast/css/unexpected-word-wrapping-with-non-empty-spans.html
1769
1770         * rendering/RenderBlockLineLayout.cpp:
1771         (WebCore::LineBreaker::nextSegmentBreak):
1772
1773 2013-10-14  Andreas Kling  <akling@apple.com>
1774
1775         Be more efficient about passing RenderStyle to attachRenderTree().
1776         <https://webkit.org/b/122743>
1777
1778         Reviewed by Antti Koivisto.
1779
1780         Have attachRenderTree() and createRendererTreeIfNeeded() pass the
1781         RenderStyle in a PassRefPtr to avoid churning the ref count.
1782
1783 2013-10-14  Sergio Villar Senin  <svillar@igalia.com>
1784
1785         [CSS Grid Layout] 2 span positions are not resolved correctly
1786         https://bugs.webkit.org/show_bug.cgi?id=119717
1787
1788         Reviewed by Andreas Kling.
1789
1790         From Blink r155397 by <jchaffraix@chromium.org>
1791
1792         Test: fast/css-grid-layout/grid-item-bad-resolution-double-span.html
1793
1794         Two opposite 'span' or 'auto' positions should be resolved using
1795         the auto placement algorithm. We were only checking for the 'auto'
1796         case. This also covers the case of other positions that, according
1797         to the spec, should be treated as 'auto'.
1798
1799         * rendering/RenderGrid.cpp:
1800         (WebCore::RenderGrid::resolveGridPositionsFromStyle):
1801
1802 2013-10-13  Andreas Kling  <akling@apple.com>
1803
1804         Use RenderElement instead of RenderObject in more places.
1805         <https://webkit.org/b/122734>
1806
1807         Reviewed by Antti Koivisto.
1808
1809         Convert some sites to use RenderElement (or type inference) instead
1810         of RenderObject for less branchy code.
1811
1812 2013-10-13  Darin Adler  <darin@apple.com>
1813
1814         Deprecate or remove deleteAllValues functions; there are only a few call sites left
1815         https://bugs.webkit.org/show_bug.cgi?id=122738
1816
1817         Reviewed by Anders Carlsson.
1818
1819         * platform/blackberry/CookieMap.cpp:
1820         (WebCore::CookieMap::deleteAllCookiesAndDomains):
1821         * platform/network/blackberry/rss/RSSParserBase.cpp:
1822         (WebCore::RSSFeed::clear):
1823         * platform/win/WCDataObject.cpp:
1824         (WebCore::WCDataObject::~WCDataObject):
1825         Renamed deleteAllValues to deprecatedDeleteAllValues.
1826
1827 2013-10-13  Sam Weinig  <sam@webkit.org>
1828
1829         Merge NODE_TYPE_CASTS and ELEMENT_TYPE_CASTS
1830         https://bugs.webkit.org/show_bug.cgi?id=122735
1831
1832         Reviewed by Antti Koivisto.
1833
1834         NODE_TYPE_CASTS and ELEMENT_TYPE_CASTS are identical. Let them become one
1835         with one another.
1836
1837 2013-10-13  Andreas Kling  <akling@apple.com>
1838
1839         Uncrashify Document::head() too. (Why am I even awake?)
1840
1841 2013-10-13  Andreas Kling  <akling@apple.com>
1842
1843         REGRESSION(r157381): Make Document::body() crash less when there is no documentElement.
1844
1845         Unreviewed.
1846
1847 2013-10-13  Darin Adler  <darin@apple.com>
1848
1849         Rewrite Document::body and Document::head in modern style, way clearer and shorter
1850         https://bugs.webkit.org/show_bug.cgi?id=122717
1851
1852         Reviewed by Andreas Kling.
1853
1854         * dom/Document.cpp:
1855         (WebCore::Document::body): Use iterator to make this way easier to read.
1856         (WebCore::Document::head): Ditto.
1857
1858         * html/HTMLTagNames.in: Added generateTypeHelpers for body and head.
1859
1860 == Rolled over to ChangeLog-2013-10-13 ==