bd34572ccdf8d3a2dc148712a7f49dccfb170c31
[WebKit-https.git] / Source / WebCore / ChangeLog
1 2013-10-18  Mario Sanchez Prada  <mario.prada@samsung.com>
2
3         [ATK] Simplify implementation of atk_text_get_text
4         https://bugs.webkit.org/show_bug.cgi?id=122644
5
6         Reviewed by Chris Fleizach.
7
8         Simplified code so we only call textUnderElement() directly once
9         and only when needed. Also, moved the specific code for ColorWell
10         objects up to the beginning of that function, so we don't do any
11         additional efforts like computing text ranges in those cases.
12
13         No new tests are needed, just to make sure that the current layout
14         and unit tests are still passing, which they are.
15
16         * accessibility/atk/WebKitAccessibleInterfaceText.cpp:
17         (textForObject): Fixed a issue that got detected while working on
18         this patch, which was causing a '\n' to be artificially appended
19         at the end of text controls all the time.
20         (webkitAccessibleTextGetText): Simplified function.
21
22         * accessibility/AccessibilityRenderObject.cpp:
23         (WebCore::AccessibilityRenderObject::doAXStringForRange): Removed
24         the check that prevents from pass ranges exceeding the limits of
25         the element's text, since those will be checked anyway when
26         calling String::substring().
27
28 2013-10-18  Brendan Long  <b.long@cablelabs.com>
29
30         [GStreamer][GTK] Add GRefPtr::outPtr()
31         https://bugs.webkit.org/show_bug.cgi?id=122996
32
33         Reviewed by Philippe Normand.
34
35         No new tests because this is just simplifying existing code.
36
37         * platform/graphics/gstreamer/GRefPtrGStreamer.cpp: Add GRefPtr specialization for GstToc.
38         * platform/graphics/gstreamer/GRefPtrGStreamer.h: Same.
39         * platform/graphics/gstreamer/GStreamerGWorld.cpp: Use GRefPtr::outPtr() to simplify code.
40         (WebCore::GStreamerGWorld::enterFullscreen):
41         (WebCore::GStreamerGWorld::exitFullscreen):
42         (WebCore::GStreamerGWorld::removePlatformVideoSink):
43         * platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp: Same.
44         (WebCore::MediaPlayerPrivateGStreamer::notifyPlayerOfText):
45         (WebCore::MediaPlayerPrivateGStreamer::newTextSample):
46         (WebCore::MediaPlayerPrivateGStreamer::processTableOfContents):
47         * platform/graphics/gstreamer/MediaPlayerPrivateGStreamerBase.cpp: Same.
48         (WebCore::MediaPlayerPrivateGStreamerBase::currentVideoSinkCaps):
49         * platform/network/soup/ResourceHandleSoup.cpp: Same.
50         (WebCore::HostTLSCertificateSet::computeCertificateHash):
51
52 2013-10-17  Alexey Proskuryakov  <ap@apple.com>
53
54         Don't add an include for return type of a [Custom] function in generated bindings code
55         https://bugs.webkit.org/show_bug.cgi?id=122972
56
57         Reviewed by Sam Weinig.
58
59         * bindings/scripts/CodeGeneratorJS.pm: (GenerateImplementation): Don't.
60
61 2013-10-17  Andreas Kling  <akling@apple.com>
62
63         CTTE: RenderMathMLSpace always has a MathMLTextElement.
64         <https://webkit.org/b/122992>
65
66         The renderer is never anonymous and always has a corresponding
67         MathMLTextElement. Overload element() with a tighter return type.
68
69         Also marked the class FINAL and made most methods private.
70
71         Reviewed by Antti Koivisto.
72
73         * rendering/mathml/RenderMathMLSpace.cpp:
74         (WebCore::RenderMathMLSpace::RenderMathMLSpace):
75         (WebCore::RenderMathMLSpace::updateFromElement):
76         * rendering/mathml/RenderMathMLSpace.h:
77
78 2013-10-17  Tim Horton  <timothy_horton@apple.com>
79
80         PlatformCALayer constructor should take layer type as an argument
81         https://bugs.webkit.org/show_bug.cgi?id=122915
82
83         Reviewed by Anders Carlsson.
84
85         No new tests, just restoring old code.
86
87         Un-do part of the patch for bug 122915; we can't early-return
88         in the constructor, there's still more work to do.
89
90         * platform/graphics/ca/mac/PlatformCALayerMac.mm:
91         (PlatformCALayerMac::PlatformCALayerMac):
92
93 2013-10-17  Alexandru Chiculita  <achicu@adobe.com>
94
95         Web Inspector: [CSS Regions] Crash when highlighting a node of a flow with no regions
96         https://bugs.webkit.org/show_bug.cgi?id=122993
97
98         Reviewed by Joseph Pecoraro.
99
100         Test: inspector-protocol/dom/highlight-flow-with-no-region.html
101
102         Even if a named flow has no regions the content of the flow will still have renderer objects created.
103         Removed the assumption that all renderers inside a RenderFlowThread will always have an enclosing RenderRegion.
104
105         * inspector/InspectorOverlay.cpp:
106         (WebCore::buildObjectForRendererFragments):
107         (WebCore::InspectorOverlay::buildObjectForHighlightedNode):
108
109 2013-10-17  Andreas Kling  <akling@apple.com>
110
111         CTTE: RenderMathMLOperator always has a MathMLElement.
112         <https://webkit.org/b/122988>
113
114         Reviewed by Antti Koivisto.
115
116         The renderer is never anonymous and always has a corresponding
117         MathMLElement. Overload element() with a tighter return type.
118
119         Also marked the class FINAL and made most methods private.
120
121 2013-10-17  Nico Weber  <thakis@chromium.org>
122
123         Fix three bugs in the equals() implementations for css gradients.
124         https://bugs.webkit.org/show_bug.cgi?id=122987
125
126         Reviewed by Andreas Kling.
127
128         1. Linear gradients were considered equal if the first gradient has no x and y
129            position and the second has no x but does have y.
130         2. Same as 1, for radial gradients. (This doesn't happen in practice as
131            CSSParser::parseRadialGradient rejects such input, so no test for this case.)
132         3. Radial gradients without x and y position weren't considered equal even if
133            they were.
134
135         * css/CSSGradientValue.cpp:
136         (WebCore::CSSLinearGradientValue::equals):
137         (WebCore::CSSRadialGradientValue::equals):
138
139 2013-10-17  Antoine Quint  <graouts@apple.com>
140
141         Web Inspector: allow front-end to trigger the system beep sound to signal an error
142         https://bugs.webkit.org/show_bug.cgi?id=122955
143
144         Reviewed by Timothy Hatcher.
145
146         New beep() method exposed on InspectorFrontendHost calling into WebCore's systemBeep().
147
148         * inspector/InspectorFrontendHost.cpp:
149         (WebCore::InspectorFrontendHost::beep):
150         * inspector/InspectorFrontendHost.h:
151         * inspector/InspectorFrontendHost.idl:
152
153 2013-10-17  Anders Carlsson  <andersca@apple.com>
154
155         Remove PlatformCAAnimation::supportsValueFunction()
156         https://bugs.webkit.org/show_bug.cgi?id=122990
157
158         Reviewed by Tim Horton.
159
160         PlatformCAAnimation::supportsValueFunction always returns true now, so there's no need for it to exist anymore.
161
162         * platform/graphics/ca/GraphicsLayerCA.cpp:
163         (WebCore::GraphicsLayerCA::createTransformAnimationsFromKeyframes):
164         * platform/graphics/ca/PlatformCAAnimation.h:
165         * platform/graphics/ca/mac/PlatformCAAnimationMac.mm:
166         * platform/graphics/ca/win/PlatformCAAnimationWin.cpp:
167
168 2013-10-17  Andreas Kling  <akling@apple.com>
169
170         CTTE: RenderMathMLFenced always has a MathMLInlineContainerElement.
171         <https://webkit.org/b/122986>
172
173         This renderer is never anonymous and always has a corresponding
174         MathMLInlineContainerElement. Overload element() with a tighter
175         return type.
176
177         Also marked the class FINAL and made most methods private.
178
179         Reviewed by Anders Carlsson.
180
181 2013-10-17  Myles C. Maxfield  <mmaxfield@apple.com>
182
183         Comment AffineTransform::xScale() and yScale() to make their meanings clearer
184         https://bugs.webkit.org/show_bug.cgi?id=122981
185
186         Reviewed by Simon Fraser.
187
188         * platform/graphics/transforms/AffineTransform.h:
189
190 2013-10-17  Vivek Galatage  <vivek.vg@samsung.com>
191
192         Remove unnecessary check for RenderLayer and rename ensureLayer to createLayer in RenderLayerModelObject.
193         https://bugs.webkit.org/show_bug.cgi?id=122928
194
195         Reviewed by Darin Adler.
196
197         No new tests due to code refactoring.
198
199         RenderLayerModelObject::styleDidChange invokes the ensureLayer() only in case of !layer().
200         Again checking for layer existence would be deemed redundant in ensureLayer.
201         Replacing it with an ASSERT(!m_layer) and also renaming it to createLayer.
202
203         Blink review URL: https://codereview.chromium.org/27246003/
204
205         * rendering/RenderLayerModelObject.cpp:
206         (WebCore::RenderLayerModelObject::createLayer):
207         (WebCore::RenderLayerModelObject::styleDidChange):
208         * rendering/RenderLayerModelObject.h:
209
210 2013-10-17  Tim Horton  <timothy_horton@apple.com>
211
212         Remove PlatformCALayerMac workaround for <rdar://problem/7390716>
213         https://bugs.webkit.org/show_bug.cgi?id=122983
214
215         Reviewed by Simon Fraser.
216
217         Remove a workaround for a bug fixed in Lion.
218
219         * platform/graphics/ca/mac/PlatformCALayerMac.mm:
220         (PlatformCALayerMac::setSublayers):
221         (PlatformCALayerMac::removeAllSublayers):
222         (PlatformCALayerMac::adoptSublayers):
223
224 2013-10-17  Robert Hogan  <robert@webkit.org>
225
226         "border-collapse: collapse;" for table removes part of its border (was: Border disappears when close to some elements)
227         https://bugs.webkit.org/show_bug.cgi?id=8914
228
229         Reviewed by David Hyatt.
230
231         The table section's side of a collapsed border won't get painted if there are no cells there to paint it. So instead
232         of relying solely on cells to paint the collapsed border paint the appropriate section of the border if there's no
233         cell to take care of it.
234
235         Tests: fast/table/paint-section-borders-without-cells-rtl.html
236                fast/table/paint-section-borders-without-cells-vertical-lr-rtl.html
237                fast/table/paint-section-borders-without-cells-vertical-lr.html
238                fast/table/paint-section-borders-without-cells-vertical-rl.html
239                fast/table/paint-section-borders-without-cells.html
240
241         * rendering/RenderTableSection.cpp:
242         (WebCore::RenderTableSection::paintRowGroupBorder):
243         (WebCore::RenderTableSection::offsetLeftForRowGroupBorder):
244         (WebCore::RenderTableSection::offsetTopForRowGroupBorder):
245         (WebCore::RenderTableSection::verticalRowGroupBorderHeight):
246         (WebCore::RenderTableSection::horizontalRowGroupBorderWidth):
247         (WebCore::RenderTableSection::paintRowGroupBorderIfRequired):
248         (WebCore::physicalBorderForDirection):
249         (WebCore::RenderTableSection::paintObject):
250         * rendering/RenderTableSection.h:
251
252 2013-10-17  Andreas Kling  <akling@apple.com>
253
254         StyleRuleFoo::mutableProperties() should return a reference.
255         <https://webkit.org/b/122962>
256
257         The mutableProperties() functions always return objects, so make
258         them return MutableStylePropertySet&.
259
260         Also tweaked the StyleRuleCSSStyleDeclaration constructor to take
261         references to both the properties and the owner rule since both
262         are required.
263
264         Reviewed by Antti Koivisto.
265
266 2013-10-17  Hans Muller  <hmuller@adobe.com>
267
268         [CSS Shapes] Improve the performance of image valued shapes with large shape-margins
269         https://bugs.webkit.org/show_bug.cgi?id=122613
270
271         Reviewed by Andreas Kling.
272
273         The cost of computing the shape-margin boundary of an image-valued shape-outside
274         is now proportional to (2 * shape-margin + image.height) rather than
275         (2 * shape-margin * image.height). The performance improvement comes from skipping
276         sequences of rounded-rectangle intervals that will not contribute to the final
277         result. Each non-empty row in the original image contributes one rounded-rectangle
278         whose corner radius is shape-margin, height is 2 * shape-margin, and width is
279         2 * shape-margin plus the width of the limits of the intervals on the row.
280
281         Renamed private method RasterShape::getIntervals() to intervalsAt() to be a little
282         more consistent with WebKit naming conventions.
283
284         There are no new tests since is just an internal refactoring.
285
286         * rendering/shapes/RasterShape.cpp:
287         (WebCore::MarginIntervalGenerator::set): Changed the x1,x2 parameters to an IntShapeInterval.
288         (WebCore::RasterShapeIntervals::contains): Refactor for the getIntervals() => intervalsAt() rename.
289         (WebCore::RasterShapeIntervals::getIntervalX1Values): Ditto.
290         (WebCore::RasterShapeIntervals::getIncludedIntervals): Ditto.
291         (WebCore::RasterShapeIntervals::getExcludedIntervals): Ditto.
292         (WebCore::RasterShapeIntervals::computeShapeMarginIntervals): Performance tuning.
293         * rendering/shapes/RasterShape.h:
294         (WebCore::RasterShapeIntervals::intervalsAt): Renamed getIntervals().
295         (WebCore::RasterShapeIntervals::limitIntervalAt): Return the min/max limits of the intervals at Y.
296         * rendering/shapes/ShapeInterval.h:
297         (WebCore::ShapeInterval::isEmpty): Added.
298
299 2013-10-15  Philippe Normand  <pnormand@igalia.com>
300
301         [GTK] Add URLMediaStream in the build
302         https://bugs.webkit.org/show_bug.cgi?id=122833
303
304         Reviewed by Carlos Garcia Campos.
305
306         * GNUmakefile.am: Add mediastream/gstreamer in include directories list.
307         * GNUmakefile.list.am: Add DOMURLMediaStream files in the build.
308
309 2013-10-17  Andreas Kling  <akling@apple.com>
310
311         Use PassRef for constructing StylePropertySets.
312         <https://webkit.org/b/122948>
313
314         Make functions that construct StylePropertySets return PassRef
315         instead of PassRefPtr. Since they never return null, this gets rid
316         of the extra branch in ~PassRefPtr everywhere.
317
318         Also StyleRule* classes now hold a Ref<StylePropertySet>, codifying
319         the fact that they always have a property set.
320
321         Reviewed by Antti Koivisto.
322
323 2013-10-17  Andreas Kling  <akling@apple.com>
324
325         DataRef<T> should use Ref<T> internally.
326         <https://webkit.org/b/122953>
327
328         DataRef is used to hold RenderStyle substructures, and due to the
329         way style inheritance is implemented, DataRef will always point to
330         a live object.
331
332         Codify this by making DataRef::m_data a Ref, and making all methods
333         that create substructure objects return PassRef.
334
335         Reviewed by Antti Koivisto.
336
337 2013-10-17  Mihnea Ovidenie  <mihnea@adobe.com>
338
339         [CSS Regions] Anonymous nested regions
340         https://bugs.webkit.org/show_bug.cgi?id=119135
341
342         Reviewed by David Hyatt.
343
344         Tests: fast/regions/table-caption-as-region.html
345                fast/regions/table-cell-as-region.html
346
347         This patch allows any non-replaced block to behave like a region. When an element is styled with the
348         -webkit-flow-from property, instead of making the renderer a RenderRegion, we let the original
349         renderer be created the same way and we add a region as an anonymous child for the renderer.
350         The anonymous block child, modeled by the new RenderNamedFlowFragment class, will be responsible
351         for the fragmentation of the named flow thread content.
352
353         A RenderBlockFlow object will keep a reference to a RenderNamedFlowFragment(RenderRegion) inside its
354         rare data structures.
355
356         Contains code contributed by Catalin Badea.
357  
358         * CMakeLists.txt:
359         * GNUmakefile.list.am:
360         * WebCore.vcxproj/WebCore.vcxproj:
361         * WebCore.xcodeproj/project.pbxproj:
362         * dom/Element.cpp: Changed to take the anonymous region into account.
363         (WebCore::Element::renderRegion):
364         (WebCore::Element::webkitGetRegionFlowRanges):
365         * dom/WebKitNamedFlow.cpp: Ditto.
366         (WebCore::WebKitNamedFlow::firstEmptyRegionIndex):
367         (WebCore::WebKitNamedFlow::getRegionsByContent):
368         (WebCore::WebKitNamedFlow::getRegions):
369         * inspector/InspectorOverlay.cpp: Take into account the new model for regions, with an anonymous region inside a block.
370         (WebCore::buildObjectForRegionHighlight):
371         (WebCore::buildObjectForElementInfo):
372         * rendering/RenderBlock.cpp:
373         (WebCore::RenderBlock::computeShapeSize): For a render named flow fragment, there is no need to recompute the shape inside
374         we can take it from the parent.
375         (WebCore::RenderBlock::renderName): Make the block that contains a render named flow fragment (region) report RenderRegion.
376         A future patch that will change this will need to rebase a lot of tests.
377         * rendering/RenderBlockFlow.cpp:
378         (WebCore::RenderBlockFlow::insertedIntoTree): Create the anonymous region if needed (change of -webkit-flow-from determines Node reattach).
379         (WebCore::RenderBlockFlow::willBeDestroyed): Clean-up the anonymous region if necessary.
380         (WebCore::RenderBlockFlow::clearFloats): Small style change to make sure that check-webkit-style reports 0 failures on RenderBlockFlow.cpp.
381         (WebCore::RenderBlockFlow::layoutBlock):
382         (WebCore::RenderBlockFlow::styleDidChange): Update the style of the anonymous region too.
383         (WebCore::RenderBlockFlow::createRenderNamedFlowFragmentIfNeeded): Helper function to create the anonymous region
384         and to add it as a child to the block.
385         (WebCore::RenderBlockFlow::canHaveChildren):
386         (WebCore::RenderBlockFlow::canHaveGeneratedChildren):
387         (WebCore::RenderBlockFlow::namedFlowFragmentNeedsUpdate): Force a layout of the anonymous region if the
388         parent block has percentage height (similar to RenderBlock::updateBlockChildDirtyBitsBeforeLayout)
389         (WebCore::RenderBlockFlow::updateLogicalHeight): Update the logical height of anonymous region when the height of parent is updated.
390         (WebCore::RenderBlockFlow::setRenderNamedFlowFragment):
391         * rendering/RenderBlockFlow.h:
392         (WebCore::RenderBlockFlow::RenderBlockFlowRareData::RenderBlockFlowRareData):
393         (WebCore::RenderBlockFlow::renderNamedFlowFragment):
394         * rendering/RenderElement.cpp:
395         (WebCore::RenderElement::createFor): Remove the direct creation of RenderRegion objects since they will be
396         created as anonymous children of block flow objects.
397         * rendering/RenderElement.h:
398         (WebCore::RenderElement::generatingElement): Account for anonymous region if necessary.
399         * rendering/RenderFlowThread.cpp:
400         (WebCore::RenderFlowThread::adjustedPositionRelativeToOffsetParent): Take anonymous region into account
401         and use the anonymous region parent offset/border.
402         * rendering/RenderLayer.cpp:
403         (WebCore::RenderLayer::shouldBeNormalFlowOnly): Check for style instead of isRenderRegion since
404         the parent of the anonymous region will get the layer.
405         * rendering/RenderListItem.cpp:
406         (WebCore::RenderListItem::insertedIntoTree): Call RenderBlockFlow method instead.
407         * rendering/RenderNamedFlowFragment.cpp: Added. Model the behaviour of the anonymous region.
408         Has RenderRegion as a base class.
409         (WebCore::RenderNamedFlowFragment::RenderNamedFlowFragment):
410         (WebCore::RenderNamedFlowFragment::~RenderNamedFlowFragment):
411         (WebCore::RenderNamedFlowFragment::setStyleForNamedFlowFragment):
412         (WebCore::RenderNamedFlowFragment::styleDidChange):
413         (WebCore::RenderNamedFlowFragment::shouldHaveAutoLogicalHeight):
414         (WebCore::RenderNamedFlowFragment::maxPageLogicalHeight):
415         * rendering/RenderNamedFlowFragment.h: Added.
416         (WebCore::RenderNamedFlowFragment::isPseudoElementRegion):
417         (WebCore::RenderNamedFlowFragment::renderName):
418         (WebCore::toRenderNamedFlowFragment):
419         * rendering/RenderObject.cpp:
420         (WebCore::RenderObject::isRenderNamedFlowFragmentContainer):
421         * rendering/RenderObject.h:
422         (WebCore::RenderObject::isRenderNamedFlowFragment):
423         * rendering/RenderRegion.h:
424         * rendering/RenderTableCaption.cpp: Call RenderBlockFlow method instead.
425         (WebCore::RenderTableCaption::insertedIntoTree):
426         * rendering/RenderTreeAsText.cpp:
427         (WebCore::write):
428         (WebCore::writeRenderRegionList):
429         * rendering/shapes/ShapeInfo.h:
430         (WebCore::ShapeInfo::shapeSize):
431         * style/StyleResolveTree.cpp:
432         (WebCore::Style::elementInsideRegionNeedsRenderer):
433
434 2013-10-17  Krzysztof Czech  <k.czech@samsung.com>
435
436         [EFL] Properly expose tables in accessibility
437         https://bugs.webkit.org/show_bug.cgi?id=122894
438
439         Reviewed by Chris Fleizach.
440
441         All tables should exposed as tables.
442
443         * accessibility/AccessibilityTable.cpp:
444         (WebCore::AccessibilityTable::isTableExposableThroughAccessibility):
445         (WebCore::AccessibilityTable::addChildren):
446         * accessibility/AccessibilityTableColumn.cpp:
447         (WebCore::AccessibilityTableColumn::computeAccessibilityIsIgnored):
448         * accessibility/AccessibilityTableHeaderContainer.cpp:
449         (WebCore::AccessibilityTableHeaderContainer::computeAccessibilityIsIgnored):
450
451 2013-10-17  Brendan Long  <b.long@cablelabs.com>
452
453         [GStreamer] Too many arguments for format in WebKitWebAudioSourceGStreamer.cpp
454         https://bugs.webkit.org/show_bug.cgi?id=122932
455
456         Reviewed by Philippe Normand.
457
458         No new tests because this just fixes a build warning.
459
460         * platform/audio/gstreamer/WebKitWebAudioSourceGStreamer.cpp:
461         (webKitWebAudioSrcLoop): Add another %s for the second part of the pad name.
462
463 2013-10-16  Tim Horton  <timothy_horton@apple.com>
464
465         Attempt to fix the Windows build after http://trac.webkit.org/changeset/157547.
466
467         The relevant conversion doesn't happen implicitly.
468
469         * platform/graphics/win/MediaPlayerPrivateFullscreenWindow.cpp:
470         (WebCore::MediaPlayerPrivateFullscreenWindow::setRootChildLayer):
471         (WebCore::MediaPlayerPrivateFullscreenWindow::wndProc):
472
473 2013-10-16  Tim Horton  <timothy_horton@apple.com>
474
475         Attempt to fix the Windows build after http://trac.webkit.org/changeset/157547.
476
477         Apparently some Windows code uses setFrame and I missed it.
478
479         * platform/graphics/win/MediaPlayerPrivateFullscreenWindow.cpp:
480         (WebCore::MediaPlayerPrivateFullscreenWindow::setRootChildLayer):
481         (WebCore::MediaPlayerPrivateFullscreenWindow::wndProc):
482
483 2013-10-16  Tim Horton  <timothy_horton@apple.com>
484
485         Remote Layer Tree: Complete support for simple layer properties
486         https://bugs.webkit.org/show_bug.cgi?id=122933
487
488         Reviewed by Anders Carlsson.
489
490         No new tests, not yet testable.
491
492         * platform/graphics/ca/GraphicsLayerCA.cpp:
493         (WebCore::GraphicsLayerCA::setName):
494         Don't dump the CALayer pointer if we own a PlatformCALayerRemote.
495
496         (WebCore::GraphicsLayerCA::recursiveCommitChanges):
497         Fix the visible tile wash (my fault!), and make it use setPosition and
498         setBounds instead of setFrame; while more convenient, it is the only
499         caller of setFrame, so we'll remove it.
500
501         * platform/graphics/ca/PlatformCALayer.h:
502         (WebCore::PlatformCALayer::isRemote): Added.
503
504         * platform/graphics/ca/mac/PlatformCALayerMac.h:
505         * platform/graphics/ca/mac/PlatformCALayerMac.mm:
506         (nullActionsDictionary):
507         (toCAFilterType):
508         (PlatformCALayerMac::synchronouslyDisplayTilesInRect):
509         (PlatformCALayerMac::playerLayer):
510         Remove setFrame, fix some pointer sides.
511
512         * platform/graphics/ca/win/PlatformCALayerWin.cpp:
513         * platform/graphics/ca/win/PlatformCALayerWin.h:
514         Remove setFrame.
515
516 2013-10-16  Andreas Kling  <akling@apple.com>
517
518         Take RenderObjects out of the arena.
519         <https://webkit.org/b/122895>
520
521         Reviewed by Antti Koivisto.
522
523         Stop arena-allocating renderers so we can move forward on improving
524         render tree memory management. This will also allow rendering code
525         to take advantage of malloc optimizations.
526
527         Line boxes and BiDi runs remain in the arena for now.
528
529 2013-10-16  Roger Fong  <roger_fong@apple.com>
530
531         [Windows] Speculative fix for test, media/video-canvas-drawing-output.html.
532
533         * platform/graphics/avfoundation/cf/MediaPlayerPrivateAVFoundationCF.cpp:
534         (WebCore::AVFWrapper::createImageForTimeInRect):
535
536 2013-10-16  Tim Horton  <timothy_horton@apple.com>
537
538         PlatformCALayer constructor should take layer type as an argument
539         https://bugs.webkit.org/show_bug.cgi?id=122915
540
541         Reviewed by Simon Fraser.
542
543         No new tests, just a minor refactoring.
544
545         * platform/graphics/ca/PlatformCALayer.h:
546         (WebCore::PlatformCALayer::PlatformCALayer):
547         Add a LayerType argument.
548
549         * platform/graphics/ca/mac/PlatformCALayerMac.mm:
550         (PlatformCALayerMac::PlatformCALayerMac):
551         * platform/graphics/ca/win/PlatformCALayerWin.cpp:
552         (PlatformCALayerWin::PlatformCALayerWin):
553         Use the new LayerType argument, and early-return in the case where we
554         are wrapping a custom PlatformLayer.
555         Drive-by un-indent the switch in the Mac version.
556
557 2013-10-15  Brady Eidson  <beidson@apple.com>
558
559         Flesh out the DatabaseProcess (and launch it!)
560         https://bugs.webkit.org/show_bug.cgi?id=122884
561
562         Reviewed by Tim Horton.
563
564         * English.lproj/Localizable.strings: Add a localizable string.
565         * WebCore.xcodeproj/project.pbxproj: Export some more headers.
566
567 2013-10-16  Tim Horton  <timothy_horton@apple.com>
568
569         RemoteLayerTree: Add support for more layer properties and transform layers
570         https://bugs.webkit.org/show_bug.cgi?id=122906
571
572         Reviewed by Anders Carlsson.
573
574         No new tests, this code is not yet testable.
575
576         * WebCore.exp.in:
577         Export some TextStream and TransformationMatrix stuff.
578
579 2013-10-16  KyungTae Kim  <ktf.kim@samsung.com>
580
581         During editing, merge inline style with overriding other author styles
582         https://bugs.webkit.org/show_bug.cgi?id=122874
583
584         Reviewed by Ryosuke Niwa.
585
586         Inline styles need to override other author styles even on DoNotOverrideValues mode.
587         So, merge and override inline styles to other author styles before merging them to m_mutableStyle.
588
589         Test: editing/deleting/merge-div-with-inline-style.html
590
591         * editing/EditingStyle.cpp:
592         (WebCore::EditingStyle::mergeInlineAndImplicitStyleOfElement):
593
594 2013-10-16  peavo@outlook.com  <peavo@outlook.com>
595
596         Emphasis marks has wrong color.
597         https://bugs.webkit.org/show_bug.cgi?id=122829
598
599         Reviewed by Antti Koivisto.
600
601         Tests: fast/text/text-emphasis.html.
602                fast/text/text-emphasis-expected.html.
603
604         Emphasis color should be set as fill color, not stroke color.
605
606         * rendering/TextPaintStyle.cpp:
607         (WebCore::updateGraphicsContext):
608         * rendering/TextPaintStyle.h:
609
610 2013-10-16  Antti Koivisto  <antti@apple.com>
611
612         Move code for finding rendered character offset to RenderTextLineBoxes
613         https://bugs.webkit.org/show_bug.cgi?id=122892
614
615         Reviewed by Andreas Kling.
616
617         * rendering/RenderText.cpp:
618         (WebCore::RenderText::countRenderedCharacterOffsets):
619         (WebCore::RenderText::containsRenderedCharacterOffset):
620         
621             Renamed for consistency.
622
623         * rendering/RenderTextLineBoxes.cpp:
624         (WebCore::RenderTextLineBoxes::countCharacterOffsets):
625         
626             This used to be Position::renderedPosition.
627
628 2013-10-16  Andreas Kling  <akling@apple.com>
629
630         RenderElement::removeChild() should take child as a reference.
631         <https://webkit.org/b/122888>
632
633         We can't remove a child without a child to remove.
634
635         Reviewed by Antti Koivisto.
636
637 2013-10-16  Antti Koivisto  <antti@apple.com>
638
639         Move test for contained caret offset to RenderTextLineBoxes
640         https://bugs.webkit.org/show_bug.cgi?id=122887
641
642         Reviewed by Andreas Kling.
643
644         * dom/Position.cpp:
645         (WebCore::Position::renderedOffset):
646         (WebCore::Position::isCandidate):
647         
648             Remove isRenderedText, call RenderText::containsCaretOffset instead.
649
650         (WebCore::Position::isRenderedCharacter):
651         (WebCore::Position::rendersInDifferentPosition):
652         * dom/Position.h:
653         * dom/PositionIterator.cpp:
654         (WebCore::PositionIterator::isCandidate):
655         * rendering/InlineTextBox.cpp:
656         * rendering/InlineTextBox.h:
657         * rendering/RenderText.cpp:
658         (WebCore::RenderText::containsCharacterOffset):
659         (WebCore::RenderText::containsCaretOffset):
660         * rendering/RenderText.h:
661         * rendering/RenderTextLineBoxes.cpp:
662         (WebCore::RenderTextLineBoxes::containsOffset):
663         
664             Combined implementations of Position::isRenderedCharacter and Position::isRenderedText.
665
666         * rendering/RenderTextLineBoxes.h:
667         
668             Remove containsCaretOffset(), functionality is now in RenderTextLineBoxes::containsOffset.
669
670 2013-10-16  Andreas Kling  <akling@apple.com>
671
672         RenderElement::isChildAllowed() should take const references.
673         <https://webkit.org/b/122870>
674
675         Reviewed by Anders Carlsson.
676
677         The isChildAllowed() functions expect non-null values to be passed,
678         so enforce this at compile-time.
679
680         Reordered some checks to do bit tests before virtual calls.
681
682 2013-10-15  Philippe Normand  <pnormand@igalia.com>
683
684         [GStreamer] video info unset if upstream doesn't query allocation
685         https://bugs.webkit.org/show_bug.cgi?id=122834
686
687         Reviewed by Gustavo Noronha Silva.
688
689         * platform/graphics/gstreamer/VideoSinkGStreamer.cpp:
690         (webkitVideoSinkRender): If the sink didn't process any allocation
691         query then use the configured source pad caps and don't rely on
692         invalid video info.
693
694 2013-10-16  Philippe Normand  <pnormand@igalia.com>
695
696         [GStreamer] move Logging.h include to GStreamerUtilities.h
697         https://bugs.webkit.org/show_bug.cgi?id=122886
698
699         Reviewed by Gustavo Noronha Silva.
700
701         Include Logging.h from GStreamerUtilities.h so the modules using
702         the LOG_MEDIA macros don't need to bother, especially for Debug builds.
703
704         * platform/graphics/gstreamer/GStreamerUtilities.h:
705         * platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp:
706         * platform/graphics/gstreamer/MediaPlayerPrivateGStreamerBase.cpp:
707
708 2013-10-16  Mark Rowe  <mrowe@apple.com>
709
710         Fix the build after r157478.
711
712         Rubber-stamped by Tim Horton.
713
714         Due to the way WebCore.exp.in is used, it can't be used to export a differing set of symbols
715         for different architectures. We often work around this by tweaking code slightly to avoid
716         needing to export different symbols. However, in this case the symbol name itself encodes an
717         architecture-specific detail and there's no clear way to avoid the requirement to export it.
718
719         To deal with this case we turn to ld's support for wildcards in the symbol export list.
720
721         * WebCore.exp.in: Use wildcards in place of the number that represents by how much "this"
722         should be adjusted when calling through the vtable thunk. Also sort the remainder of the file.
723         * make-export-file-generator: Don't attempt to verify symbol names that contain wildcard characters.
724
725 2013-10-15  Tim Horton  <timothy_horton@apple.com>
726
727         Two more exports for 32-bit build fix.
728         The duality of CGFloat means that we use a different
729         version of getRGBA (and the cast operator)
730         on 32-bit systems.
731
732         * WebCore.exp.in:
733
734 2013-10-15  Tim Horton  <timothy_horton@apple.com>
735
736         Another shot at a build fix; apparently these
737         do need exporting for some reason, but are different
738         on 32-bit.
739
740         * WebCore.exp.in:
741
742 2013-10-15  Alexey Proskuryakov  <ap@apple.com>
743
744         GenerateIsReachable=ImplContext is confusing
745         https://bugs.webkit.org/show_bug.cgi?id=122864
746
747         Reviewed by Geoffrey Garen.
748
749         Renamed to ImplWebGLRenderingContext.
750
751         * bindings/scripts/CodeGeneratorJS.pm:
752         (GenerateImplementation):
753         * bindings/scripts/IDLAttributes.txt:
754         * html/canvas/EXTDrawBuffers.idl:
755         * html/canvas/EXTTextureFilterAnisotropic.idl:
756         * html/canvas/OESElementIndexUint.idl:
757         * html/canvas/OESStandardDerivatives.idl:
758         * html/canvas/OESTextureFloat.idl:
759         * html/canvas/OESTextureHalfFloat.idl:
760         * html/canvas/OESVertexArrayObject.idl:
761         * html/canvas/WebGLCompressedTextureATC.idl:
762         * html/canvas/WebGLCompressedTexturePVRTC.idl:
763         * html/canvas/WebGLCompressedTextureS3TC.idl:
764         * html/canvas/WebGLDebugRendererInfo.idl:
765         * html/canvas/WebGLDebugShaders.idl:
766         * html/canvas/WebGLDepthTexture.idl:
767         * html/canvas/WebGLLoseContext.idl:
768
769         * WebCore.xcodeproj/project.pbxproj: While at it, added OESElementIndexUint.idl
770         to Xcode project.
771
772 2013-10-15  Tim Horton  <timothy_horton@apple.com>
773
774         Unreviewed build fix; I don't know how to export.
775         This may not help.
776
777         * WebCore.exp.in:
778
779 2013-10-15  Dean Jackson  <dino@apple.com>
780
781         Add ENABLE_WEB_ANIMATIONS flag
782         https://bugs.webkit.org/show_bug.cgi?id=122871
783
784         Reviewed by Tim Horton.
785
786         Eventually might be http://dev.w3.org/fxtf/web-animations/
787         but this is just engine-internal work at the moment.
788
789         * Configurations/FeatureDefines.xcconfig:
790
791 2013-10-15  Tim Horton  <timothy_horton@apple.com>
792
793         Add a PlatformCALayer subclass that proxies its property changes across the process boundary
794         https://bugs.webkit.org/show_bug.cgi?id=122773
795
796         Reviewed by Anders Carlsson.
797
798         No new tests, the new drawing area is not yet testable.
799
800         * WebCore.exp.in:
801         Export lots of GraphicsLayerCA stuff so we can inherit from it in WebKit2.
802
803         * WebCore.xcodeproj/project.pbxproj:
804         Make PlatformCAFilters.h a private header.
805
806         * platform/graphics/GraphicsLayer.h:
807         (WebCore::GraphicsLayer::initialize):
808         * platform/graphics/ca/GraphicsLayerCA.cpp:
809         (WebCore::GraphicsLayer::create):
810         (WebCore::GraphicsLayerCA::GraphicsLayerCA):
811         (WebCore::GraphicsLayerCA::initialize):
812         * platform/graphics/ca/GraphicsLayerCA.h:
813         Defer creation of the main PlatformCALayer until just after the constructor is finished
814         so that GraphicsLayerCA subclasses can successfully override createPlatformCALayer.
815
816         * platform/graphics/ca/PlatformCALayer.h:
817         (WebCore::PlatformCALayer::platformLayer):
818         Make platformLayer virtual so that subclasses which don't have PlatformLayers can override.
819
820 2013-10-14  Ryosuke Niwa  <rniwa@webkit.org>
821
822         REGRESSION: Crash in XMLDocumentParser::startElementNs
823         https://bugs.webkit.org/show_bug.cgi?id=122817
824
825         Reviewed by Darin Adler.
826
827         Exit early in startElementNs when listeners and handlers of synchronous events such as load event
828         removes the inserted node inside parserAppendChild.
829
830         Test: fast/parser/xhtml-synchronous-detach-crash.html
831
832         * xml/parser/XMLDocumentParserLibxml2.cpp:
833         (WebCore::XMLDocumentParser::startElementNs):
834
835 2013-10-15  Joseph Pecoraro  <pecoraro@apple.com>
836
837         Web Inspector: Remove old frontend localizedStrings.js
838         https://bugs.webkit.org/show_bug.cgi?id=122846
839
840         Reviewed by Timothy Hatcher.
841
842         * Configurations/WebCore.xcconfig:
843         * English.lproj/localizedStrings.js: Removed.
844         * WebCore.xcodeproj/project.pbxproj:
845         Remove the file and references to it. We no longer need to exclude
846         localizedString.js from some builds.
847
848 2013-10-15  Joseph Pecoraro  <pecoraro@apple.com>
849
850         Web Inspector: Remove Windows old front-end related code
851         https://bugs.webkit.org/show_bug.cgi?id=122845
852
853         Reviewed by Brent Fulgham.
854
855         * WebCore.vcxproj/copyWebCoreResourceFiles.cmd:
856
857 2013-10-15  Morten Stenshorne  <mstensho@opera.com>
858
859         Add support for the column-fill property
860         https://bugs.webkit.org/show_bug.cgi?id=117693
861
862         Reviewed by David Hyatt.
863
864         This is only supported in the (new) region based multicol implementation.
865
866         With column-fill support, a lot of multicol tests needed an update.
867         The old implementation behaved as if column-fill were 'auto', but the
868         initial value is 'balance', so now we need to be explicit about that.
869         For auto-height tests it doesn't really matter - such multicols are always
870         balanced anyway.
871
872         Tests: fast/multicol/newmulticol/fixed-height-fill-auto.html
873                fast/multicol/newmulticol/fixed-height-fill-balance.html
874
875         * css/CSSComputedStyleDeclaration.cpp:
876         (WebCore::ComputedStyleExtractor::propertyValue):
877         * css/CSSParser.cpp:
878         (WebCore::isValidKeywordPropertyAndValue):
879         (WebCore::isKeywordPropertyID):
880         (WebCore::CSSParser::parseValue):
881         * css/CSSPrimitiveValueMappings.h:
882         (WebCore::CSSPrimitiveValue::CSSPrimitiveValue):
883         (WebCore::CSSPrimitiveValue::operator ColumnFill):
884         * css/CSSPropertyNames.in:
885         * css/CSSValueKeywords.in:
886         * css/DeprecatedStyleBuilder.cpp:
887         (WebCore::DeprecatedStyleBuilder::DeprecatedStyleBuilder):
888         * rendering/RenderMultiColumnBlock.h:
889         * rendering/style/RenderStyle.h:
890         * rendering/style/RenderStyleConstants.h:
891         * rendering/style/StyleMultiColData.cpp:
892         (WebCore::StyleMultiColData::StyleMultiColData):
893         (WebCore::StyleMultiColData::operator==):
894         * rendering/style/StyleMultiColData.h:
895
896 2013-10-15  Andreas Kling  <akling@apple.com>
897
898         Skip unnecessary null check in RenderText::textLength().
899         <https://webkit.org/b/122841>
900
901         Reviewed by Antti Koivisto.
902
903         RenderText will never have a null String in m_text, so textLength()
904         can grab at the StringImpl directly, avoiding a null check.
905
906 2013-10-15  Andreas Kling  <akling@apple.com>
907
908         FontGenericFamilies should not be ref-counted.
909         <https://webkit.org/b/122835>
910
911         Reviewed by Anders Carlsson.
912
913         FontGenericFamilies is singly-owned by Settings.
914
915 2013-10-15  Zoltan Horvath  <zoltan@webkit.org>
916
917         [CSS Shapes] Move RenderBlock::layoutShapeInsideInfo into RenderBlock.cpp
918         http://bugs.webkit.org/show_bug.cgi?id=122843
919
920         Reviewed by Oliver Hunt.
921
922         Historically, layoutShapeInsideInfo was a static function in RenderBlockLineLayout, then it has changed to be a member of RenderBlock,
923         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
924         an unnecessary CSS_SHAPES #ifdef as well from RenderBlock.cpp.
925
926         No new tests, no behavior change.
927
928         * rendering/RenderBlock.cpp:
929         (WebCore::RenderBlock::markShapeInsideDescendantsForLayout):
930         (WebCore::RenderBlock::layoutShapeInsideInfo):
931         * rendering/RenderBlockLineLayout.cpp:
932
933 2013-10-15  peavo@outlook.com  <peavo@outlook.com>
934
935         [WinCairo] Build fails.
936         https://bugs.webkit.org/show_bug.cgi?id=122830
937
938         Reviewed by Brent Fulgham.
939
940         * platform/network/NetworkStorageSessionStub.cpp:
941         (WebCore::NetworkStorageSession::createPrivateBrowsingSession): Update to new return type.
942
943 2013-10-15  Andreas Kling  <akling@apple.com>
944
945         FileIconLoader should not be ref-counted.
946         <https://webkit.org/b/122827>
947
948         FileIconLoader is singly-owned by FileInputType.
949
950         Reviewed by Antti Koivisto.
951
952 2013-10-15  Andreas Kling  <akling@apple.com>
953
954         RenderText should cache RenderStyle in locals more.
955         <https://webkit.org/b/122823>
956
957         Reviewed by Antti Koivisto.
958
959         Now that fetching the RenderStyle has to go through the parent,
960         we should avoid unnecessary loads by caching style() in a local.
961
962 2013-10-15  Csaba Osztrogon√°c  <ossy@webkit.org>
963
964         Move PlatformCertificateInfo to WebCore and make the ResourceResponse primitives work in terms of that platform agnostic object
965         https://bugs.webkit.org/show_bug.cgi?id=118520
966
967         Reviewed by Anders Carlsson.
968
969         Added PlatformCertificateInfo getter and setter to ResourceErrorBase
970         and ResourceResponseBase so that getting and setting certificates
971         becomes cross platform.
972
973         Changed the existing platform specific certificate getters and setters
974         of ResourceError and ResourceResponse to use the
975         PlatformCertificateInfo member.
976
977         Original patch by Kwang Yul Seo  <skyul@company100.net>
978
979         The following things were fixed by Csaba Osztrogon√°c:
980         - trivial conflicts resolved in WebCore.exp.in, project.pbxproj, NetworkResourceLoader.cpp,
981           WebPageProxy.h, WebFrameLoaderClient.cpp and PlatformEfl.cmake.
982         - trivial fix in the new AsynchronousNetworkLoaderClient.cpp
983         - style fixed (NULL -> 0 and smaller indentation in WebCoreArgumentCodersSoup.cpp)
984         - obsolete change removed from NetworkResourceLoader.cpp
985         - unneeded typo removed from ResourceHandleSoup.cpp
986         - Fixed the if guard of including RetainPtr.h in PlatformCertificateInfo.h to make Windows build happy.
987         - add PlatformCertificateInfoCFNet.cpp with empty constructor and destructor to make Windows build happy.
988         - resolved conflict in Source/WebCore/WebCore.xcodeproj/project.pbxproj after r156488
989         - renamed KURL to URL after r156550.
990         - removed obsolete Qt changes
991         - added back initializers for m_soupFlags
992         - included gio/gio.h instead of libsoup/soup.h in PlatformCertificateInfo.h
993         - removed unused class forward declarations and includes
994
995         * GNUmakefile.list.am:
996         * PlatformEfl.cmake:
997         * Target.pri:
998         * WebCore.exp.in:
999         * WebCore.vcxproj/WebCore.vcxproj:
1000         * WebCore.vcxproj/WebCore.vcxproj.filters:
1001         * WebCore.xcodeproj/project.pbxproj:
1002         * platform/network/PlatformCertificateInfo.h: Renamed from Source/WebKit2/Shared/soup/PlatformCertificateInfo.h.
1003         (WebCore::PlatformCertificateInfo::certificate):
1004         (WebCore::PlatformCertificateInfo::setCertificate):
1005         (WebCore::PlatformCertificateInfo::tlsErrors):
1006         (WebCore::PlatformCertificateInfo::setTLSErrors):
1007         * platform/network/ResourceErrorBase.h:
1008         (WebCore::ResourceErrorBase::platformCertificateInfo):
1009         (WebCore::ResourceErrorBase::setPlatformCertificateInfo):
1010         * platform/network/ResourceResponseBase.h:
1011         (WebCore::ResourceResponseBase::platformCertificateInfo):
1012         (WebCore::ResourceResponseBase::setPlatformCertificateInfo):
1013         * platform/network/cf/ResourceResponse.h:
1014         * platform/network/mac/PlatformCertificateInfoMac.mm: Renamed from Source/WebKit2/Shared/mac/PlatformCertificateInfo.mm.
1015         (WebCore::PlatformCertificateInfo::PlatformCertificateInfo):
1016         (WebCore::PlatformCertificateInfo::~PlatformCertificateInfo):
1017         (WebCore::PlatformCertificateInfo::setCertificateChain):
1018         (WebCore::PlatformCertificateInfo::certificateChain):
1019         (WebCore::PlatformCertificateInfo::dump):
1020         * platform/network/mac/ResourceResponseMac.mm:
1021         (WebCore::ResourceResponse::setCertificateChain):
1022         (WebCore::ResourceResponse::certificateChain):
1023         * platform/network/soup/PlatformCertificateInfoSoup.cpp:
1024         (WebCore::PlatformCertificateInfo::PlatformCertificateInfo):
1025         (WebCore::PlatformCertificateInfo::~PlatformCertificateInfo):
1026         * platform/network/soup/ResourceError.h:
1027         (WebCore::ResourceError::ResourceError):
1028         (WebCore::ResourceError::tlsErrors):
1029         (WebCore::ResourceError::setTLSErrors):
1030         (WebCore::ResourceError::certificate):
1031         (WebCore::ResourceError::setCertificate):
1032         * platform/network/soup/ResourceErrorSoup.cpp:
1033         (WebCore::ResourceError::platformCopy):
1034         * platform/network/soup/ResourceResponse.h:
1035         (WebCore::ResourceResponse::ResourceResponse):
1036         (WebCore::ResourceResponse::soupMessageCertificate):
1037         (WebCore::ResourceResponse::setSoupMessageCertificate):
1038         (WebCore::ResourceResponse::soupMessageTLSErrors):
1039         (WebCore::ResourceResponse::setSoupMessageTLSErrors):
1040         * platform/network/soup/ResourceResponseSoup.cpp:
1041         (WebCore::ResourceResponse::toSoupMessage):
1042         (WebCore::ResourceResponse::updateFromSoupMessage):
1043
1044 2013-10-15  Ryosuke Niwa  <rniwa@webkit.org>
1045
1046         Remove redundant Document::getElementById
1047         https://bugs.webkit.org/show_bug.cgi?id=122813
1048
1049         Reviewed by Andreas Kling.
1050
1051         Merge https://chromium.googlesource.com/chromium/blink/+/4e8f1c5316415614b84370c602beae4a1008299f
1052
1053         This function simply calls virtual TreeScope::getElementById and Document inherits from TreeScope.
1054
1055         * WebCore.exp.in:
1056         * dom/Document.cpp:
1057         * dom/Document.h:
1058
1059 2013-10-14  Santosh Mahto  <santosh.ma@samsung.com>
1060
1061         in safari,the background-color of input[type="search"] can't work
1062         https://bugs.webkit.org/show_bug.cgi?id=119967
1063
1064         Reviewed by Ryosuke Niwa.
1065
1066         When input type=search is styled with css background property then
1067         it does not change the background-color of field. Its happening becasue
1068         search field is not counted as styled control. Thus theme ignores the 
1069         css background property. With this patch search field is also counted as 
1070         styled control so background property reflects on search field.
1071
1072         Test: fast/forms/search/search-field-background-color.html
1073
1074         * rendering/RenderTheme.cpp:
1075         (WebCore::RenderTheme::isControlStyled):Now search field is also
1076         a styled control.
1077
1078 2013-10-14  Ryosuke Niwa  <rniwa@webkit.org>
1079
1080         EventPath::updateTouchLists traverses through EventPath thrice
1081         https://bugs.webkit.org/show_bug.cgi?id=122804
1082
1083         Reviewed by Benjamin Poulain.
1084
1085         Instead of traversing through EventPath for each TouchList, traverse through TouchList for every EventContext.
1086         This paves our way to have one-pass traversal over EventPath, and evetually to remove EventContext altogether.
1087
1088         This change should also improve the cache hit rate since all Touch objects tend to be allocated at the same time
1089         but this performance improvement is probably not observable.
1090
1091         * dom/EventContext.h:
1092         * dom/EventDispatcher.cpp:
1093         (WebCore::EventRelatedNodeResolver::EventRelatedNodeResolver): Added a new constructor that takes Touch and
1094         and TouchListType. We need to store these two values in order to update EventContext later.
1095         (WebCore::EventRelatedNodeResolver::touch): Added,
1096         (WebCore::EventRelatedNodeResolver::touchListType): Added.
1097         (WebCore::addRelatedNodeResolversForTouchList): Extracted from updateTouchListsInEventPath.
1098         (WebCore::EventPath::updateTouchLists): Moved the loop over m_path here. Notice that the outer loop iterates
1099         over m_path instead of touchList as done in updateTouchListsInEventPath. The inner loop goes through resolvers
1100         and adds Touch objects each EventContext as needed.
1101
1102 2013-10-14  Alexey Proskuryakov  <ap@apple.com>
1103
1104         Don't generate a wasteful isObservable check in isReachableFromOpaqueRoots
1105         https://bugs.webkit.org/show_bug.cgi?id=122802
1106
1107         Reviewed by Mark Hahnenberg.
1108
1109         * bindings/scripts/CodeGeneratorJS.pm: (GenerateImplementation): Don't.
1110
1111         * bindings/scripts/test/JS/JSTestActiveDOMObject.cpp:
1112         * bindings/scripts/test/JS/JSTestCustomNamedGetter.cpp:
1113         * bindings/scripts/test/JS/JSTestEventConstructor.cpp:
1114         * bindings/scripts/test/JS/JSTestEventTarget.cpp:
1115         * bindings/scripts/test/JS/JSTestException.cpp:
1116         * bindings/scripts/test/JS/JSTestInterface.cpp:
1117         * bindings/scripts/test/JS/JSTestMediaQueryListListener.cpp:
1118         * bindings/scripts/test/JS/JSTestNamedConstructor.cpp:
1119         * bindings/scripts/test/JS/JSTestObj.cpp:
1120         * bindings/scripts/test/JS/JSTestOverloadedConstructors.cpp:
1121         * bindings/scripts/test/JS/JSTestSerializedScriptValueInterface.cpp:
1122         * bindings/scripts/test/JS/JSTestTypedefs.cpp:
1123         * bindings/scripts/test/JS/JSattribute.cpp:
1124         * bindings/scripts/test/JS/JSreadonly.cpp:
1125         Updated results.
1126
1127 2013-10-14  Samuel White  <samuel_white@apple.com>
1128
1129         AX: fieldset should have GroupRole and legend should be description.
1130         https://bugs.webkit.org/show_bug.cgi?id=122534
1131
1132         Reviewed by Chris Fleizach.
1133
1134         Changes fieldset to derive AXDescription from legend if one is available. Added
1135         convenience method to AccessibilityObject to fetch element if available.
1136
1137         Test: accessibility/fieldset-element.html
1138
1139         * accessibility/AccessibilityNodeObject.cpp:
1140         (WebCore::AccessibilityNodeObject::canHaveChildren):
1141         (WebCore::AccessibilityNodeObject::alternativeText):
1142         * accessibility/AccessibilityObject.cpp:
1143         (WebCore::AccessibilityObject::element):
1144         (WebCore::AccessibilityObject::isARIAHidden):
1145         (WebCore::AccessibilityObject::isDOMHidden):
1146         (WebCore::AccessibilityObject::defaultObjectInclusion):
1147         * accessibility/AccessibilityObject.h:
1148         (WebCore::AccessibilityObject::isHidden):
1149         * accessibility/AccessibilitySlider.cpp:
1150         (WebCore::AccessibilitySlider::getAttribute):
1151         (WebCore::AccessibilitySlider::valueForRange):
1152         (WebCore::AccessibilitySlider::maxValueForRange):
1153         (WebCore::AccessibilitySlider::minValueForRange):
1154         (WebCore::AccessibilitySlider::setValue):
1155         (WebCore::AccessibilitySlider::inputElement):
1156         * accessibility/AccessibilitySlider.h:
1157         * accessibility/mac/AccessibilityObjectMac.mm:
1158         (WebCore::AccessibilityObject::accessibilityPlatformIncludesObject):
1159         * html/HTMLFieldSetElement.cpp:
1160         (WebCore::HTMLFieldSetElement::legend):
1161         * html/HTMLFieldSetElement.h:
1162
1163 2013-10-14  Roger Fong  <roger_fong@apple.com>
1164
1165         Windows select element doesn't draw RTL properly.
1166         https://bugs.webkit.org/show_bug.cgi?id=122785.
1167
1168         Reviewed by Brent Fulgham.
1169
1170         Covered by fast/text/international/pop-up-button-text-alignment-and-direction.html.
1171
1172         Problems include the popup items not drawing on the right hand side and 
1173         not respecting the direction or the directional override styling of the option.
1174         The selected element (drawn in the actual select element) also doesn't respect 
1175         the style settings of the selected menu option.
1176
1177         * platform/win/PopupMenuWin.cpp:
1178         (WebCore::PopupMenuWin::paint):
1179
1180 2013-10-14  Roger Fong  <roger_fong@apple.com>
1181
1182         [Windows] Unreviewed build fix.
1183
1184         * WebCore.vcxproj/WebCoreCommon.props:
1185
1186 2013-10-14  Ryosuke Niwa  <rniwa@webkit.org>
1187
1188         Crash in WebCore::BidiResolver<WebCore::InlineIterator, WebCore::BidiRun>::createBidiRunsForLine
1189         https://bugs.webkit.org/show_bug.cgi?id=122776
1190
1191         Reviewed by Darin Adler.
1192
1193         Merge https://chromium.googlesource.com/chromium/blink/+/aca89bc4d984705a1f94b623dae0ab03e239a248
1194
1195         Fix modification of whitespace endpoints to not assume it's operating on RenderTexts
1196
1197         During line layout, we use midpoints to identify RenderObjects, or parts of
1198         RenderObjects, that don't need InlineBoxes, usually because of collapsed whitespace.
1199
1200         Prior to actually creating BidiRuns (the precursor to InlineBoxes), we use
1201         checkMidpoints to fix up our lineMidpointState to handle the case where we start
1202         ignoring spaces in our line, but don't stop until somewhere on the following line.
1203         Previously, this function assumed that the final midpoint (called an endpoint)
1204         was a RenderText, but this assumption is wrong if we have a beginning midpoint
1205         created by shouldSkipWhitespaceAfterStartObject (which handles inlines and list
1206         markers) and no endpoint on that line. In that case, we'd instead adjust the
1207         position backwards on the beginning midpoint, which would cause us to fail to
1208         create an InlineBox for the inline or list marker. In the new test added, this
1209         would actually trigger a crash due to an assumption when visually re-ordering
1210         BidiRuns that a non-empty line would actually contain at least one such run.
1211
1212         Test: fast/text/whitespace/whitespace-and-margin-wrap-after-list-marker-crash.html
1213
1214         * rendering/RenderBlockLineLayout.cpp:
1215         (WebCore::checkMidpoints):
1216
1217 2013-10-14  Ryosuke Niwa  <rniwa@webkit.org>
1218
1219         Assertion failure in Range::processContentsBetweenOffsets
1220         https://bugs.webkit.org/show_bug.cgi?id=122777
1221
1222         Reviewed by Darin Adler.
1223
1224         Merge https://chromium.googlesource.com/chromium/blink/+/c15de182774c7859c20d97126eb844ae97b792a4
1225
1226         This patch changes ASSERT statements for checking |endOffset| inbound in Range::processContentsBetweenOffsets()
1227         to limit |endOffset|. This is necessary when DOMNodeRemovedFromDocument event handler splits text nodes,
1228         Range::insertNode() on text node, in the range calling Range::deleteContents().
1229
1230         Test: fast/dom/Range/range-delete-contents-mutation-event-crash.html
1231
1232         * dom/Range.cpp:
1233         (WebCore::Range::processContentsBetweenOffsets):
1234
1235 2013-10-14  Alexey Proskuryakov  <ap@apple.com>
1236
1237         Add an empty window.crypto.webkitSubtle
1238         https://bugs.webkit.org/show_bug.cgi?id=122778
1239
1240         Reviewed by Mark Hahnenberg.
1241
1242         Tests: security/crypto-subtle-gc-2.html
1243                security/crypto-subtle-gc-3.html
1244                security/crypto-subtle-gc.html
1245
1246         * DerivedSources.make: Process SubtleCrypto.idl.
1247
1248         * crypto: Added.
1249         * WebCore.xcodeproj/project.pbxproj:
1250         * CMakeLists.txt:
1251         * DerivedSources.make:
1252         * GNUmakefile.am:
1253         * GNUmakefile.list.am:
1254         * WebCore.vcxproj/WebCore.vcxproj.filters:
1255
1256         * bindings/js/JSSubtleCryptoCustom.cpp: Added. Empty for now, but we'll certainly
1257         need custom bindings code here.
1258
1259         * crypto/SubtleCrypto.cpp: Added.
1260         (WebCore::SubtleCrypto::SubtleCrypto):
1261         (WebCore::SubtleCrypto::document):
1262         * crypto/SubtleCrypto.h: Added.        
1263         * crypto/SubtleCrypto.idl: Added.
1264         * page/Crypto.cpp:
1265         (WebCore::Crypto::subtle):
1266         * page/Crypto.h:
1267         * page/Crypto.idl:
1268
1269 2013-10-14  Nick Diego Yamane  <nick.yamane@openbossa.org>
1270
1271         Remove GestureEvent leftovers from WebCore
1272         <https://webkit.org/b/122780>
1273
1274         Reviewed by Anders Carlsson.
1275
1276         - Removed some remaining references to PlatformGestureEvent, supposed to
1277           be removed by r157316
1278         - TOUCH_ADJUSTMENT should be reworked after GestureEvent feature
1279           removal
1280
1281         * page/EventHandler.cpp:
1282         (WebCore::EventHandler::bestZoomableAreaForTouchPoint):
1283         * page/EventHandler.h:
1284         * platform/PlatformEvent.h:
1285         * platform/ScrollAnimatorNone.cpp:
1286         * platform/ScrollableArea.h:
1287
1288 2013-10-14  Nick Diego Yamane  <nick.yamane@openbossa.org>
1289
1290         Build fix after r157366
1291         http://bugs.webkit.org/show_bug.cgi?id=122783
1292
1293         When TOUCH_AJUSTMENT is enabled build fails due to some
1294         refactors in TextRender functions.
1295
1296         Reviewed by Anders Carlsson.
1297
1298         * page/TouchAdjustment.cpp:
1299         (WebCore::TouchAdjustment::appendContextSubtargetsForNode):
1300
1301 2013-10-14  Alexandru Chiculita  <achicu@adobe.com>
1302
1303         The content of the DOM panel for iframes is not updated until the "onload" event
1304         https://bugs.webkit.org/show_bug.cgi?id=122653
1305
1306         Reviewed by Darin Adler.
1307
1308         Test: http/tests/inspector-protocol/loading-iframe-document-node.html
1309
1310         Renamed InspectorDOMAgent::loadEventFired to InspectorDOMAgent::didCommitLoad and moved the call site
1311         from InspectorInstrumentation::loadEventFiredImpl to InspectorInstrumentation::didCommitLoadImpl.
1312         This is to make sure that it will invalidate the content of the iframe as soon as the frame navigates
1313         to a different page. This way the new node can be retrieved as soon as the page has some content, and
1314         not just when the page is fully loaded.
1315
1316         * inspector/InspectorDOMAgent.cpp:
1317         (WebCore::InspectorDOMAgent::didCommitLoad): Renamed from loadEventFired, as it is now called from
1318         didCommitLoadImpl instead.
1319         (WebCore::InspectorDOMAgent::frameDocumentUpdated): Updated comment to point to the new function name.
1320         * inspector/InspectorDOMAgent.h:
1321         * inspector/InspectorInstrumentation.cpp:
1322         (WebCore::InspectorInstrumentation::loadEventFiredImpl): Removed call do InspectorDOMAgent.loadEventFired.
1323         (WebCore::InspectorInstrumentation::didCommitLoadImpl): Added call to InspectorDOMAgent.didCommitLoad.
1324
1325 2013-10-14  Roger Fong  <roger_fong@apple.com>
1326
1327         https://bugs.webkit.org/show_bug.cgi?id=122774.
1328         <rdar://problem/6138855>.
1329
1330         Reviewed by Brent Fulgham.
1331
1332         Add a field to keep track of hovered over index.
1333         Use index to determine whether or not to use the existing selected index on the mouse down event.
1334
1335         * platform/win/PopupMenuWin.cpp:
1336         (WebCore::PopupMenuWin::PopupMenuWin):
1337         (WebCore::PopupMenuWin::show):
1338         (WebCore::PopupMenuWin::wndProc):
1339         * platform/win/PopupMenuWin.h:
1340
1341 2013-10-14  Tim Horton  <timothy_horton@apple.com>
1342
1343         Virtualize PlatformCALayer
1344         https://bugs.webkit.org/show_bug.cgi?id=122672
1345
1346         Reviewed by Anders Carlsson.
1347
1348         No new tests, just a refactoring.
1349
1350         * WebCore.exp.in:
1351         setGeometryFlipped is on PlatformCALayerMac now.
1352
1353         * WebCore.vcxproj/WebCore.vcxproj:
1354         * WebCore.vcxproj/WebCore.vcxproj.filters:
1355         Add PlatformCALayer.cpp, PlatformCALayerWin.h, and let VS do its
1356         thing with some other files.
1357
1358         * WebCore.xcodeproj/project.pbxproj:
1359         Add PlatformCALayer.cpp and PlatformCALayerMac.h.
1360
1361         * platform/graphics/avfoundation/cf/MediaPlayerPrivateAVFoundationCF.cpp:
1362         (WebCore::AVFWrapper::platformLayer):
1363         Make a PlatformCALayerWin explicitly.
1364
1365         * platform/graphics/ca/GraphicsLayerCA.cpp:
1366         (WebCore::GraphicsLayerCA::createPlatformCALayer):
1367         Added. Decide whether to make a PlatformCALayer{Mac, Win} based on the platform.
1368         Later, we will decide between other subclasses based on other things.
1369
1370         (WebCore::GraphicsLayerCA::filtersCanBeComposited):
1371         Do the same thing for filtersCanBeComposited.
1372
1373         (WebCore::GraphicsLayerCA::GraphicsLayerCA):
1374         (WebCore::GraphicsLayerCA::setContentsToSolidColor):
1375         (WebCore::GraphicsLayerCA::setContentsToMedia):
1376         (WebCore::GraphicsLayerCA::setContentsToCanvas):
1377         (WebCore::GraphicsLayerCA::recursiveCommitChanges):
1378         (WebCore::GraphicsLayerCA::ensureStructuralLayer):
1379         (WebCore::GraphicsLayerCA::updateContentsImage):
1380         (WebCore::GraphicsLayerCA::updateContentsRects):
1381         (WebCore::GraphicsLayerCA::swapFromOrToTiledLayer):
1382         Use createPlatformCALayer instead of PlatformCALayer::create.
1383
1384         * platform/graphics/ca/mac/PlatformCAFiltersMac.mm:
1385         For now, use the PlatformCALayerMac version of filtersCanBeComposited,
1386         since this code is heavily tied to having CALayers in the Web process.
1387
1388         * platform/graphics/ca/GraphicsLayerCA.h:
1389         Include PlatformCALayer.h here so we can get the LayerType enum.
1390         (NOTE-to-be-removed: if there's a better way to do this, I'm open to
1391         suggestions; I couldn't puzzle out nested 'enum class' stuff).
1392
1393         Add the createPlatformCALayers.
1394
1395         * platform/graphics/ca/PlatformCAAnimation.h:
1396         Friend the subclasses too.
1397
1398         * platform/graphics/ca/PlatformCALayer.cpp: Added.
1399         (WebCore::PlatformCALayer::~PlatformCALayer):
1400         Pull the shared part of the PlatformCALayer destructor out.
1401
1402         * platform/graphics/ca/PlatformCALayer.h:
1403         (WebCore::PlatformCALayer::platformLayer):
1404         (WebCore::PlatformCALayer::setOwner):
1405         (WebCore::PlatformCALayer::PlatformCALayer):
1406         Virtualize all the things. Move platform specific members to their new subclasses.
1407
1408         * platform/graphics/ca/mac/PlatformCALayerMac.h: Added.
1409         * platform/graphics/ca/mac/PlatformCALayerMac.mm:
1410         Move the PLATFORM(MAC) implementations from PlatformCALayer to PlatformCALayerMac.
1411
1412         (PlatformCALayer::platformCALayer):
1413         The platformCALayer lookup function is static on PlatformCALayer, so it can't
1414         be moved to the subclasses. It might be a good idea in the future to move towards
1415         a platform-independent mechanism for looking up PlatformCALayers from PlatformLayers,
1416         and to avoid needing to do this as often as we do now.
1417
1418         * platform/graphics/ca/win/CACFLayerTreeHost.cpp:
1419         (WebCore::CACFLayerTreeHost::CACFLayerTreeHost):
1420         * platform/graphics/ca/win/PlatformCALayerWin.cpp:
1421         Move the PLATFORM(WIN) implementations from PlatformCALayer to PlatformCALayerWin.
1422
1423         (PlatformCALayerWin::create):
1424         (PlatformCALayer::platformCALayer):
1425
1426         * platform/graphics/ca/win/PlatformCALayerWin.h: Added.
1427
1428         * platform/graphics/win/GraphicsContext3DWin.cpp:
1429         (WebCore::GraphicsContext3D::GraphicsContext3D):
1430         * platform/graphics/win/MediaPlayerPrivateQuickTimeVisualContext.cpp:
1431         (WebCore::MediaPlayerPrivateQuickTimeVisualContext::createLayerForMovie):
1432         Include PlatformCALayerWin.h and explicitly make PlatformCALayerWins here.
1433
1434 2013-10-14  Hugo Parente Lima  <hugo.lima@openbossa.org>
1435
1436         [cmake] MediaControlsApple is used only by Efl port and is on CMakeLists.txt
1437         https://bugs.webkit.org/show_bug.cgi?id=122772
1438
1439         Reviewed by Anders Carlsson.
1440
1441         * CMakeLists.txt: Removed MediaControlsApple.cpp
1442         * PlatformEfl.cmake: Added MediaControlsApple.cpp
1443
1444 2013-10-14  Alexey Proskuryakov  <ap@apple.com>
1445
1446         window.crypto doesn't preserve custom properties
1447         https://bugs.webkit.org/show_bug.cgi?id=122770
1448
1449         Reviewed by Mark Hahnenberg.
1450
1451         Test: security/crypto-gc.html
1452
1453         Generate isReachableFromOpaqueRoots that makes Crypto live as long as the document
1454         lives (because that's when it's observable through window object).
1455
1456         * page/Crypto.cpp:
1457         (WebCore::Crypto::Crypto):
1458         (WebCore::Crypto::~Crypto):
1459         (WebCore::Crypto::document):
1460         * page/Crypto.h:
1461         (WebCore::Crypto::create):
1462         Made Crypto a ContextDestructionObserver, so that it can report its document to bindings.
1463         Removed ScriptWrappable, because it seems to have served no purpose in this class.
1464
1465         * page/Crypto.idl: Added GenerateIsReachable. Removed ImplementationLacksVTable,
1466         because the class now has a vtable, and can be checked for bindings integrity.
1467
1468         * page/DOMWindow.cpp: (WebCore::DOMWindow::crypto): Pass a document when creating
1469         crypto.
1470
1471 2013-10-14  Andreas Kling  <akling@apple.com>
1472
1473         CTTE: NamedNodeMap always has a corresponding Element.
1474         <https://webkit.org/b/122769>
1475
1476         Reviewed by Anders Carlsson.
1477
1478         Made NamedNodeMap::m_element a reference and remove an assertion
1479         that it's never null.
1480
1481 2013-10-14  Andreas Kling  <akling@apple.com>
1482
1483         REGRESSION(r157408): Crashes in RenderFullScreen::wrapRenderer().
1484
1485         Unreviewed crash fix for these two tests:
1486
1487         - fullscreen/full-screen-restrictions.html
1488         - fullscreen/empty-anonymous-block-continuation-crash.html
1489
1490         * rendering/RenderFullScreen.cpp:
1491         (RenderFullScreen::wrapRenderer):
1492
1493             Get the RenderArena from Document like we did before this patch.
1494
1495 2013-10-14  Hans Muller  <hmuller@adobe.com>
1496
1497         [CSS Shapes] Image valued shape-outside shapes should update the layout after the image has been loaded
1498         https://bugs.webkit.org/show_bug.cgi?id=122340
1499
1500         Reviewed by Simon Fraser.
1501
1502         Ensure that the an image-valued shape-outside layout is updated after the image has
1503         been loaded.
1504
1505         Test: http/tests/css/css-image-valued-shape.html
1506
1507         * rendering/RenderBlock.cpp:
1508         (WebCore::RenderBlock::imageChanged): Added code for the shape-outside case.
1509         (WebCore::RenderBlock::updateShapeInsideInfoAfterStyleChange): Ditto.
1510         * rendering/RenderElement.cpp:
1511         (WebCore::RenderElement::~RenderElement): Ditto.
1512         (WebCore::RenderElement::setStyle): Ditto.
1513
1514 2013-10-14  Andreas Kling  <akling@apple.com>
1515
1516         Remove some silly null checks in Element/NamedNodeMap.
1517         <https://webkit.org/b/122767>
1518
1519         Reviewed by Darin Adler.
1520
1521         Make shouldIgnoreAttributeCase() take a const Element&, exposing
1522         some unnecessary null checks.
1523
1524 2013-10-14  Brent Fulgham  <bfulgham@apple.com>
1525
1526         [Win] Build fix after r122737.
1527
1528         * dom/Node.h: Add explicit WebCore namespace to macro definition to work around
1529         Visual Studio bug.
1530
1531 2013-10-14  Andreas Kling  <akling@apple.com>
1532
1533         Pass Document directly to anonymous renderer constructors.
1534         <https://webkit.org/b/122752>
1535
1536         Reviewed by Antti Koivisto.
1537
1538         Added separate constructors for creating anonymous renderers that
1539         take a Document& instead of a null Element*/Text*.
1540
1541         Removed setDocumentForAnonymous() and all createAnonymous() helpers.
1542         ...and RenderObject::m_node is now a Node&, wohoo!
1543
1544 2013-10-13  Sam Weinig  <sam@webkit.org>
1545
1546         CTTE: Add more node conversion helpers
1547         https://bugs.webkit.org/show_bug.cgi?id=122737
1548
1549         Reviewed by Darin Adler.
1550
1551         - Factor NODE_TYPE_CASTS into TYPE_CASTS_BASE(ToClassName, FromClassName)
1552           to allow for DOCUMENT_TYPE_CASTS.
1553         - Replace more static_casts<>.
1554
1555 2013-10-14  Zan Dobersek  <zdobersek@igalia.com>
1556
1557         Reintroduce PassRefPtr<Event> copy in ScopedEventQueue::dispatchEvent
1558         https://bugs.webkit.org/show_bug.cgi?id=122742
1559
1560         Reviewed by Alexey Proskuryakov.
1561
1562         This is a follow-up to r157219 which introduced a workaround for the GCC's quirky behavior that
1563         was resulting in crashes due to the PassRefPtr<Event> object passed to EventDispatcher::dispatchEvent
1564         being copied and nullified first before retrieving the EventTarget of the Event object wrapped in that
1565         PassRefPtr.
1566
1567         The implementation is now adjusted to first retrieve the pointer to the Event's EventTarget and store
1568         it in a local variable. That variable is then passed as the first parameter to EventDispatcher::dispatchEvent,
1569         and the PassRefPtr<Event> passed directly as the second parameter. Previously the pointer of that PassRefPtr
1570         object was passed in, with a new PassRefPtr being created which would increase the reference count of the
1571         ref-counted object. Passing in the original PassRefPtr avoids the unnecessary reference count increase by creating
1572         a copy. That still nullifies the original PassRefPtr, but that's not a problem anymore.
1573
1574         * dom/ScopedEventQueue.cpp:
1575         (WebCore::ScopedEventQueue::dispatchEvent):
1576
1577 2013-10-14  Bear Travis  <betravis@adobe.com>
1578
1579         [CSS Shapes] Shape-Margin should be animatable
1580         https://bugs.webkit.org/show_bug.cgi?id=122524
1581
1582         Reviewed by Darin Adler.
1583
1584         Mark content for relayout after shape-margin changes, and add shape-margin
1585         to the list of animatable properties.
1586
1587         Tests: fast/shapes/shape-outside-floats/shape-outside-dynamic-shape-margin.html
1588                fast/shapes/shape-outside-floats/shape-outside-shape-margin-animation.html
1589
1590         * page/animation/CSSPropertyAnimation.cpp:
1591         (WebCore::CSSPropertyAnimationWrapperMap::CSSPropertyAnimationWrapperMap): Add
1592         shape-margin to the map of animatable CSS properties.
1593         * rendering/RenderBox.cpp:
1594         (WebCore::RenderBox::updateShapeOutsideInfoAfterStyleChange): Compare shape-margins,
1595         and mark dependent content for relayout if they have changed.
1596
1597 2013-10-14  Arvid Nilsson  <anilsson@blackberry.com>
1598
1599         Don't crash after OpenGL robustness reset
1600         https://bugs.webkit.org/show_bug.cgi?id=122750
1601
1602         Reviewed by George Staikos.
1603
1604         JIRA 517132.
1605         Just log the incident and pretend like nothing happened.
1606
1607         No new tests, we don't have repeatable steps to reproduce a robustness
1608         reset.
1609
1610         * platform/graphics/blackberry/LayerRenderer.cpp:
1611         (WebCore::LayerRenderer::makeContextCurrent):
1612
1613 2013-10-14  Krzysztof Czech  <k.czech@samsung.com>
1614
1615         [EFL] Buildfix after r157393
1616         https://bugs.webkit.org/show_bug.cgi?id=122749
1617
1618         Reviewed by Andreas Kling.
1619
1620         Buildfix with error enumeration value 'CSS_FR' not handled in switch.
1621
1622         * css/CSSCalculationValue.cpp:
1623         (WebCore::hasDoubleValue):
1624
1625 2013-10-14  Krzysztof Czech  <k.czech@samsung.com>
1626
1627         [EFL] Present replaced objects with 0xFFFC character
1628         https://bugs.webkit.org/show_bug.cgi?id=122744
1629
1630         Reviewed by Mario Sanchez Prada.
1631
1632         Replaced elements should be emitted in GTK/EFL and
1633         marked their presence with the replacement character.
1634
1635         * accessibility/AccessibilityObject.cpp:
1636         (WebCore::AccessibilityObject::textIteratorBehaviorForTextRange):
1637
1638 2013-09-16  Sergio Villar Senin  <svillar@igalia.com>
1639
1640         [CSS Grid Layout] Implement support for <flex>
1641         https://bugs.webkit.org/show_bug.cgi?id=115362
1642
1643         Reviewed by Andreas Kling.
1644
1645         From Blink r149134, r149480, r149532, r150287 and r156127 by <jchaffraix@chromium.org>
1646         From Blink r157820 by <svillar@igalia.com>
1647
1648         Added support for flexible lengths ('fr' unit) in CSS Grid Layout
1649         code. This requires the addition of GridLength class to
1650         encapsulate the knowledge of <flex> to the grid layout code.
1651
1652         Also updated the algorithm that computes the layout. It increases
1653         the value of 1fr based on the grid tracks' usedBreath to fraction
1654         ratio (called normalizedFractionValue). This enables increasing
1655         the fraction value while updating the available space to account
1656         for processed grid tracks. The algorithm stops when we run out of
1657         grid tracks or available space (one grid item has an intrinsic
1658         size too big). This matches the specs to the letter for the known
1659         available space case (both the unknown case and the interaction
1660         with 'span' are left out of this patch).
1661
1662         Tests: fast/css-grid-layout/flex-and-minmax-content-resolution-columns.html
1663                fast/css-grid-layout/flex-and-minmax-content-resolution-rows.html
1664                fast/css-grid-layout/flex-content-resolution-columns.html
1665                fast/css-grid-layout/flex-content-resolution-rows.html
1666
1667         * GNUmakefile.list.am: Added GridLength.h to the build system.
1668         * Target.pri: Ditto.
1669         * WebCore.vcxproj/WebCore.vcxproj: Ditto.
1670         * WebCore.vcxproj/WebCore.vcxproj.filters: Ditto.
1671         * WebCore.xcodeproj/project.pbxproj: Ditto.
1672         * css/CSSComputedStyleDeclaration.cpp:
1673         (WebCore::valueForGridTrackBreadth): Replace Length by GridLength.
1674         * css/CSSGrammar.y.in: Added FR support.
1675         * css/CSSParser.cpp: Ditto.
1676         (WebCore::CSSParser::parseGridBreadth):
1677         (WebCore::CSSParser::detectNumberToken):
1678         * css/CSSParserValues.cpp: Added FR support.
1679         (WebCore::CSSParserValue::createCSSValue):
1680         * css/CSSParserValues.h:
1681         (WebCore::CSSParserString::operator[]):
1682         (WebCore::CSSParserString::equalIgnoringCase):
1683         * css/CSSPrimitiveValue.cpp: Added FR support.
1684         (WebCore::isValidCSSUnitTypeForDoubleConversion):
1685         (WebCore::CSSPrimitiveValue::cleanup):
1686         (WebCore::CSSPrimitiveValue::customCSSText):
1687         (WebCore::CSSPrimitiveValue::cloneForCSSOM):
1688         (WebCore::CSSPrimitiveValue::equals):
1689         * css/CSSPrimitiveValue.h: Added a couple of missing const.
1690         (WebCore::CSSPrimitiveValue::isFlex):
1691         * css/StyleResolver.cpp: Added FR support.
1692         (WebCore::createGridTrackBreadth):
1693         (WebCore::createGridTrackSize):
1694         * rendering/RenderGrid.cpp:
1695         (WebCore::GridTrackForNormalization::GridTrackForNormalization):
1696         New helper struct to ease the computation of track breadths with
1697         flexible lengths.
1698         (WebCore::GridTrackForNormalization::operator=):
1699         (WebCore::RenderGrid::computePreferredTrackWidth): Replaced Length by GridLength.
1700         (WebCore::RenderGrid::computedUsedBreadthOfGridTracks): Grow grid lines
1701         having a fraction as the MaxTrackSizingFunction.
1702         (WebCore::RenderGrid::computeUsedBreadthOfMinLength): Replaced Length by GridLength.
1703         (WebCore::RenderGrid::computeUsedBreadthOfMaxLength): Ditto.
1704         (WebCore::sortByGridNormalizedFlexValue):
1705         (WebCore::RenderGrid::computeNormalizedFractionBreadth): Increase
1706         the fraction value while updating the available space to account
1707         for processed grid tracks.
1708         (WebCore::RenderGrid::resolveContentBasedTrackSizingFunctions):
1709         (WebCore::RenderGrid::distributeSpaceToTracks): Never shrink track sizes.
1710         (WebCore::RenderGrid::tracksAreWiderThanMinTrackBreadth):
1711         * rendering/RenderGrid.h:
1712         * rendering/style/GridLength.h: Added.
1713         (WebCore::GridLength::GridLength):
1714         (WebCore::GridLength::isLength):
1715         (WebCore::GridLength::isFlex):
1716         (WebCore::GridLength::length):
1717         (WebCore::GridLength::flex):
1718         (WebCore::GridLength::setFlex):
1719         (WebCore::GridLength::operator==):
1720         * rendering/style/GridTrackSize.h: Replaced Length by GridLength.
1721         (WebCore::GridTrackSize::length):
1722         (WebCore::GridTrackSize::setLength):
1723         (WebCore::GridTrackSize::minTrackBreadth):
1724         (WebCore::GridTrackSize::maxTrackBreadth):
1725         (WebCore::GridTrackSize::setMinMax):
1726         (WebCore::GridTrackSize::hasMinOrMaxContentMinTrackBreadth):
1727         (WebCore::GridTrackSize::hasMaxContentMinTrackBreadth):
1728         (WebCore::GridTrackSize::hasMinOrMaxContentMaxTrackBreadth):
1729         (WebCore::GridTrackSize::hasMaxContentMaxTrackBreadth):
1730
1731 2013-10-14  peavo@outlook.com  <peavo@outlook.com>
1732
1733         Broken text rendering when input field has selection.
1734         https://bugs.webkit.org/show_bug.cgi?id=122716
1735
1736         Reviewed by Antti Koivisto.
1737
1738         Tests: fast/text/text-rendering-with-input-selection.html.
1739                fast/text/text-rendering-with-input-selection-expected.html.
1740
1741         * rendering/InlineTextBox.cpp:
1742         (WebCore::InlineTextBox::paint): Check that text has selection.
1743
1744 2013-10-14  Zalan Bujtas  <zalan@apple.com>
1745
1746         Unexpected word wrapping for wrapped content then raw content.
1747         https://bugs.webkit.org/show_bug.cgi?id=121130
1748
1749         Reviewed by Antti Koivisto.
1750
1751         When deciding whether a line is considered empty, we need to check if the current
1752         object is fully responsible for the currently uncommitted width. It helps differentiating
1753         <span></span><span>abcd</span> from <span>a</span><span>bcd</span>, where in the first
1754         case when we hit the second <span> the line is still considered empty, as opposed to the
1755         second example.
1756
1757         Test: fast/css/unexpected-word-wrapping-with-non-empty-spans.html
1758
1759         * rendering/RenderBlockLineLayout.cpp:
1760         (WebCore::LineBreaker::nextSegmentBreak):
1761
1762 2013-10-14  Andreas Kling  <akling@apple.com>
1763
1764         Be more efficient about passing RenderStyle to attachRenderTree().
1765         <https://webkit.org/b/122743>
1766
1767         Reviewed by Antti Koivisto.
1768
1769         Have attachRenderTree() and createRendererTreeIfNeeded() pass the
1770         RenderStyle in a PassRefPtr to avoid churning the ref count.
1771
1772 2013-10-14  Sergio Villar Senin  <svillar@igalia.com>
1773
1774         [CSS Grid Layout] 2 span positions are not resolved correctly
1775         https://bugs.webkit.org/show_bug.cgi?id=119717
1776
1777         Reviewed by Andreas Kling.
1778
1779         From Blink r155397 by <jchaffraix@chromium.org>
1780
1781         Test: fast/css-grid-layout/grid-item-bad-resolution-double-span.html
1782
1783         Two opposite 'span' or 'auto' positions should be resolved using
1784         the auto placement algorithm. We were only checking for the 'auto'
1785         case. This also covers the case of other positions that, according
1786         to the spec, should be treated as 'auto'.
1787
1788         * rendering/RenderGrid.cpp:
1789         (WebCore::RenderGrid::resolveGridPositionsFromStyle):
1790
1791 2013-10-13  Andreas Kling  <akling@apple.com>
1792
1793         Use RenderElement instead of RenderObject in more places.
1794         <https://webkit.org/b/122734>
1795
1796         Reviewed by Antti Koivisto.
1797
1798         Convert some sites to use RenderElement (or type inference) instead
1799         of RenderObject for less branchy code.
1800
1801 2013-10-13  Darin Adler  <darin@apple.com>
1802
1803         Deprecate or remove deleteAllValues functions; there are only a few call sites left
1804         https://bugs.webkit.org/show_bug.cgi?id=122738
1805
1806         Reviewed by Anders Carlsson.
1807
1808         * platform/blackberry/CookieMap.cpp:
1809         (WebCore::CookieMap::deleteAllCookiesAndDomains):
1810         * platform/network/blackberry/rss/RSSParserBase.cpp:
1811         (WebCore::RSSFeed::clear):
1812         * platform/win/WCDataObject.cpp:
1813         (WebCore::WCDataObject::~WCDataObject):
1814         Renamed deleteAllValues to deprecatedDeleteAllValues.
1815
1816 2013-10-13  Sam Weinig  <sam@webkit.org>
1817
1818         Merge NODE_TYPE_CASTS and ELEMENT_TYPE_CASTS
1819         https://bugs.webkit.org/show_bug.cgi?id=122735
1820
1821         Reviewed by Antti Koivisto.
1822
1823         NODE_TYPE_CASTS and ELEMENT_TYPE_CASTS are identical. Let them become one
1824         with one another.
1825
1826 2013-10-13  Andreas Kling  <akling@apple.com>
1827
1828         Uncrashify Document::head() too. (Why am I even awake?)
1829
1830 2013-10-13  Andreas Kling  <akling@apple.com>
1831
1832         REGRESSION(r157381): Make Document::body() crash less when there is no documentElement.
1833
1834         Unreviewed.
1835
1836 2013-10-13  Darin Adler  <darin@apple.com>
1837
1838         Rewrite Document::body and Document::head in modern style, way clearer and shorter
1839         https://bugs.webkit.org/show_bug.cgi?id=122717
1840
1841         Reviewed by Andreas Kling.
1842
1843         * dom/Document.cpp:
1844         (WebCore::Document::body): Use iterator to make this way easier to read.
1845         (WebCore::Document::head): Ditto.
1846
1847         * html/HTMLTagNames.in: Added generateTypeHelpers for body and head.
1848
1849 == Rolled over to ChangeLog-2013-10-13 ==