42d9d57269d6b521f625ae8c821f6c40eeb3c315
[WebKit-https.git] / Source / WebCore / ChangeLog
1 2013-10-18  ChangSeok Oh  <changseok.oh@collabora.com>
2
3         Unreviewed build fix for --no-svg option.
4         m_svgStyle of RenderStyle is guarded with the ENABLE_SVG flag.
5
6         * rendering/style/RenderStyle.cpp:
7         (WebCore::RenderStyle::RenderStyle):
8
9 2013-10-17  Brady Eidson  <beidson@apple.com>
10
11         Cleanup the Modules group in the WebCore.xcodeproj
12         https://bugs.webkit.org/show_bug.cgi?id=123009
13
14         Rubberstamped by Antti Koivisto.
15
16         * WebCore.xcodeproj/project.pbxproj:
17
18 2013-10-18  Denis Nomiyama  <d.nomiyama@samsung.com>
19
20         [ATK] Fix invalid signal to set objects to an unknown state "layout-complete"
21         https://bugs.webkit.org/show_bug.cgi?id=122970
22
23         Reviewed by Mario Sanchez Prada.
24
25         Removed an invalid signal to set objects to an unknown state
26         layout-complete. This signal was originally generated to notify DRT
27         and WKTR in case of page load complete.
28         It was replaced by ATK:AtkDocument:load-complete, which is already sent
29         by AXObjectCache::frameLoadingEventPlatformNotification().
30
31         There is no new test since the changes are covered by existing ones
32         (e.g. accessibility/loading-iframe-sends-notification.html).
33
34         * accessibility/atk/AXObjectCacheAtk.cpp:
35         (WebCore::AXObjectCache::postPlatformNotification): Removed an invalid
36         signal to set objects to an unknown state layout-complete.
37
38 2013-10-18  Carlos Garcia Campos  <cgarcia@igalia.com>
39
40         [GTK] Generate API documentation for GObject DOM bindings
41         https://bugs.webkit.org/show_bug.cgi?id=121538
42
43         Reviewed by Gustavo Noronha Silva.
44
45         * bindings/gobject/GNUmakefile.am: Add a explicit rule for all
46         .symbols file making them depend on the corresponding header file,
47         since the .symbols file is generated by the bindings generator.
48
49 2013-10-18  Mario Sanchez Prada  <mario.prada@samsung.com>
50
51         [ATK] Simplify implementation of atk_text_get_text
52         https://bugs.webkit.org/show_bug.cgi?id=122644
53
54         Reviewed by Chris Fleizach.
55
56         Simplified code so we only call textUnderElement() directly once
57         and only when needed. Also, moved the specific code for ColorWell
58         objects up to the beginning of that function, so we don't do any
59         additional efforts like computing text ranges in those cases.
60
61         No new tests are needed, just to make sure that the current layout
62         and unit tests are still passing, which they are.
63
64         * accessibility/atk/WebKitAccessibleInterfaceText.cpp:
65         (textForObject): Fixed a issue that got detected while working on
66         this patch, which was causing a '\n' to be artificially appended
67         at the end of text controls all the time.
68         (webkitAccessibleTextGetText): Simplified function.
69
70         * accessibility/AccessibilityRenderObject.cpp:
71         (WebCore::AccessibilityRenderObject::doAXStringForRange): Removed
72         the check that prevents from pass ranges exceeding the limits of
73         the element's text, since those will be checked anyway when
74         calling String::substring().
75
76 2013-10-18  Brendan Long  <b.long@cablelabs.com>
77
78         [GStreamer][GTK] Add GRefPtr::outPtr()
79         https://bugs.webkit.org/show_bug.cgi?id=122996
80
81         Reviewed by Philippe Normand.
82
83         No new tests because this is just simplifying existing code.
84
85         * platform/graphics/gstreamer/GRefPtrGStreamer.cpp: Add GRefPtr specialization for GstToc.
86         * platform/graphics/gstreamer/GRefPtrGStreamer.h: Same.
87         * platform/graphics/gstreamer/GStreamerGWorld.cpp: Use GRefPtr::outPtr() to simplify code.
88         (WebCore::GStreamerGWorld::enterFullscreen):
89         (WebCore::GStreamerGWorld::exitFullscreen):
90         (WebCore::GStreamerGWorld::removePlatformVideoSink):
91         * platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp: Same.
92         (WebCore::MediaPlayerPrivateGStreamer::notifyPlayerOfText):
93         (WebCore::MediaPlayerPrivateGStreamer::newTextSample):
94         (WebCore::MediaPlayerPrivateGStreamer::processTableOfContents):
95         * platform/graphics/gstreamer/MediaPlayerPrivateGStreamerBase.cpp: Same.
96         (WebCore::MediaPlayerPrivateGStreamerBase::currentVideoSinkCaps):
97         * platform/network/soup/ResourceHandleSoup.cpp: Same.
98         (WebCore::HostTLSCertificateSet::computeCertificateHash):
99
100 2013-10-17  Alexey Proskuryakov  <ap@apple.com>
101
102         Don't add an include for return type of a [Custom] function in generated bindings code
103         https://bugs.webkit.org/show_bug.cgi?id=122972
104
105         Reviewed by Sam Weinig.
106
107         * bindings/scripts/CodeGeneratorJS.pm: (GenerateImplementation): Don't.
108
109 2013-10-17  Andreas Kling  <akling@apple.com>
110
111         CTTE: RenderMathMLSpace always has a MathMLTextElement.
112         <https://webkit.org/b/122992>
113
114         The renderer is never anonymous and always has a corresponding
115         MathMLTextElement. Overload element() with a tighter return type.
116
117         Also marked the class FINAL and made most methods private.
118
119         Reviewed by Antti Koivisto.
120
121         * rendering/mathml/RenderMathMLSpace.cpp:
122         (WebCore::RenderMathMLSpace::RenderMathMLSpace):
123         (WebCore::RenderMathMLSpace::updateFromElement):
124         * rendering/mathml/RenderMathMLSpace.h:
125
126 2013-10-17  Tim Horton  <timothy_horton@apple.com>
127
128         PlatformCALayer constructor should take layer type as an argument
129         https://bugs.webkit.org/show_bug.cgi?id=122915
130
131         Reviewed by Anders Carlsson.
132
133         No new tests, just restoring old code.
134
135         Un-do part of the patch for bug 122915; we can't early-return
136         in the constructor, there's still more work to do.
137
138         * platform/graphics/ca/mac/PlatformCALayerMac.mm:
139         (PlatformCALayerMac::PlatformCALayerMac):
140
141 2013-10-17  Alexandru Chiculita  <achicu@adobe.com>
142
143         Web Inspector: [CSS Regions] Crash when highlighting a node of a flow with no regions
144         https://bugs.webkit.org/show_bug.cgi?id=122993
145
146         Reviewed by Joseph Pecoraro.
147
148         Test: inspector-protocol/dom/highlight-flow-with-no-region.html
149
150         Even if a named flow has no regions the content of the flow will still have renderer objects created.
151         Removed the assumption that all renderers inside a RenderFlowThread will always have an enclosing RenderRegion.
152
153         * inspector/InspectorOverlay.cpp:
154         (WebCore::buildObjectForRendererFragments):
155         (WebCore::InspectorOverlay::buildObjectForHighlightedNode):
156
157 2013-10-17  Andreas Kling  <akling@apple.com>
158
159         CTTE: RenderMathMLOperator always has a MathMLElement.
160         <https://webkit.org/b/122988>
161
162         Reviewed by Antti Koivisto.
163
164         The renderer is never anonymous and always has a corresponding
165         MathMLElement. Overload element() with a tighter return type.
166
167         Also marked the class FINAL and made most methods private.
168
169 2013-10-17  Nico Weber  <thakis@chromium.org>
170
171         Fix three bugs in the equals() implementations for css gradients.
172         https://bugs.webkit.org/show_bug.cgi?id=122987
173
174         Reviewed by Andreas Kling.
175
176         1. Linear gradients were considered equal if the first gradient has no x and y
177            position and the second has no x but does have y.
178         2. Same as 1, for radial gradients. (This doesn't happen in practice as
179            CSSParser::parseRadialGradient rejects such input, so no test for this case.)
180         3. Radial gradients without x and y position weren't considered equal even if
181            they were.
182
183         * css/CSSGradientValue.cpp:
184         (WebCore::CSSLinearGradientValue::equals):
185         (WebCore::CSSRadialGradientValue::equals):
186
187 2013-10-17  Antoine Quint  <graouts@apple.com>
188
189         Web Inspector: allow front-end to trigger the system beep sound to signal an error
190         https://bugs.webkit.org/show_bug.cgi?id=122955
191
192         Reviewed by Timothy Hatcher.
193
194         New beep() method exposed on InspectorFrontendHost calling into WebCore's systemBeep().
195
196         * inspector/InspectorFrontendHost.cpp:
197         (WebCore::InspectorFrontendHost::beep):
198         * inspector/InspectorFrontendHost.h:
199         * inspector/InspectorFrontendHost.idl:
200
201 2013-10-17  Anders Carlsson  <andersca@apple.com>
202
203         Remove PlatformCAAnimation::supportsValueFunction()
204         https://bugs.webkit.org/show_bug.cgi?id=122990
205
206         Reviewed by Tim Horton.
207
208         PlatformCAAnimation::supportsValueFunction always returns true now, so there's no need for it to exist anymore.
209
210         * platform/graphics/ca/GraphicsLayerCA.cpp:
211         (WebCore::GraphicsLayerCA::createTransformAnimationsFromKeyframes):
212         * platform/graphics/ca/PlatformCAAnimation.h:
213         * platform/graphics/ca/mac/PlatformCAAnimationMac.mm:
214         * platform/graphics/ca/win/PlatformCAAnimationWin.cpp:
215
216 2013-10-17  Andreas Kling  <akling@apple.com>
217
218         CTTE: RenderMathMLFenced always has a MathMLInlineContainerElement.
219         <https://webkit.org/b/122986>
220
221         This renderer is never anonymous and always has a corresponding
222         MathMLInlineContainerElement. Overload element() with a tighter
223         return type.
224
225         Also marked the class FINAL and made most methods private.
226
227         Reviewed by Anders Carlsson.
228
229 2013-10-17  Myles C. Maxfield  <mmaxfield@apple.com>
230
231         Comment AffineTransform::xScale() and yScale() to make their meanings clearer
232         https://bugs.webkit.org/show_bug.cgi?id=122981
233
234         Reviewed by Simon Fraser.
235
236         * platform/graphics/transforms/AffineTransform.h:
237
238 2013-10-17  Vivek Galatage  <vivek.vg@samsung.com>
239
240         Remove unnecessary check for RenderLayer and rename ensureLayer to createLayer in RenderLayerModelObject.
241         https://bugs.webkit.org/show_bug.cgi?id=122928
242
243         Reviewed by Darin Adler.
244
245         No new tests due to code refactoring.
246
247         RenderLayerModelObject::styleDidChange invokes the ensureLayer() only in case of !layer().
248         Again checking for layer existence would be deemed redundant in ensureLayer.
249         Replacing it with an ASSERT(!m_layer) and also renaming it to createLayer.
250
251         Blink review URL: https://codereview.chromium.org/27246003/
252
253         * rendering/RenderLayerModelObject.cpp:
254         (WebCore::RenderLayerModelObject::createLayer):
255         (WebCore::RenderLayerModelObject::styleDidChange):
256         * rendering/RenderLayerModelObject.h:
257
258 2013-10-17  Tim Horton  <timothy_horton@apple.com>
259
260         Remove PlatformCALayerMac workaround for <rdar://problem/7390716>
261         https://bugs.webkit.org/show_bug.cgi?id=122983
262
263         Reviewed by Simon Fraser.
264
265         Remove a workaround for a bug fixed in Lion.
266
267         * platform/graphics/ca/mac/PlatformCALayerMac.mm:
268         (PlatformCALayerMac::setSublayers):
269         (PlatformCALayerMac::removeAllSublayers):
270         (PlatformCALayerMac::adoptSublayers):
271
272 2013-10-17  Robert Hogan  <robert@webkit.org>
273
274         "border-collapse: collapse;" for table removes part of its border (was: Border disappears when close to some elements)
275         https://bugs.webkit.org/show_bug.cgi?id=8914
276
277         Reviewed by David Hyatt.
278
279         The table section's side of a collapsed border won't get painted if there are no cells there to paint it. So instead
280         of relying solely on cells to paint the collapsed border paint the appropriate section of the border if there's no
281         cell to take care of it.
282
283         Tests: fast/table/paint-section-borders-without-cells-rtl.html
284                fast/table/paint-section-borders-without-cells-vertical-lr-rtl.html
285                fast/table/paint-section-borders-without-cells-vertical-lr.html
286                fast/table/paint-section-borders-without-cells-vertical-rl.html
287                fast/table/paint-section-borders-without-cells.html
288
289         * rendering/RenderTableSection.cpp:
290         (WebCore::RenderTableSection::paintRowGroupBorder):
291         (WebCore::RenderTableSection::offsetLeftForRowGroupBorder):
292         (WebCore::RenderTableSection::offsetTopForRowGroupBorder):
293         (WebCore::RenderTableSection::verticalRowGroupBorderHeight):
294         (WebCore::RenderTableSection::horizontalRowGroupBorderWidth):
295         (WebCore::RenderTableSection::paintRowGroupBorderIfRequired):
296         (WebCore::physicalBorderForDirection):
297         (WebCore::RenderTableSection::paintObject):
298         * rendering/RenderTableSection.h:
299
300 2013-10-17  Andreas Kling  <akling@apple.com>
301
302         StyleRuleFoo::mutableProperties() should return a reference.
303         <https://webkit.org/b/122962>
304
305         The mutableProperties() functions always return objects, so make
306         them return MutableStylePropertySet&.
307
308         Also tweaked the StyleRuleCSSStyleDeclaration constructor to take
309         references to both the properties and the owner rule since both
310         are required.
311
312         Reviewed by Antti Koivisto.
313
314 2013-10-17  Hans Muller  <hmuller@adobe.com>
315
316         [CSS Shapes] Improve the performance of image valued shapes with large shape-margins
317         https://bugs.webkit.org/show_bug.cgi?id=122613
318
319         Reviewed by Andreas Kling.
320
321         The cost of computing the shape-margin boundary of an image-valued shape-outside
322         is now proportional to (2 * shape-margin + image.height) rather than
323         (2 * shape-margin * image.height). The performance improvement comes from skipping
324         sequences of rounded-rectangle intervals that will not contribute to the final
325         result. Each non-empty row in the original image contributes one rounded-rectangle
326         whose corner radius is shape-margin, height is 2 * shape-margin, and width is
327         2 * shape-margin plus the width of the limits of the intervals on the row.
328
329         Renamed private method RasterShape::getIntervals() to intervalsAt() to be a little
330         more consistent with WebKit naming conventions.
331
332         There are no new tests since is just an internal refactoring.
333
334         * rendering/shapes/RasterShape.cpp:
335         (WebCore::MarginIntervalGenerator::set): Changed the x1,x2 parameters to an IntShapeInterval.
336         (WebCore::RasterShapeIntervals::contains): Refactor for the getIntervals() => intervalsAt() rename.
337         (WebCore::RasterShapeIntervals::getIntervalX1Values): Ditto.
338         (WebCore::RasterShapeIntervals::getIncludedIntervals): Ditto.
339         (WebCore::RasterShapeIntervals::getExcludedIntervals): Ditto.
340         (WebCore::RasterShapeIntervals::computeShapeMarginIntervals): Performance tuning.
341         * rendering/shapes/RasterShape.h:
342         (WebCore::RasterShapeIntervals::intervalsAt): Renamed getIntervals().
343         (WebCore::RasterShapeIntervals::limitIntervalAt): Return the min/max limits of the intervals at Y.
344         * rendering/shapes/ShapeInterval.h:
345         (WebCore::ShapeInterval::isEmpty): Added.
346
347 2013-10-15  Philippe Normand  <pnormand@igalia.com>
348
349         [GTK] Add URLMediaStream in the build
350         https://bugs.webkit.org/show_bug.cgi?id=122833
351
352         Reviewed by Carlos Garcia Campos.
353
354         * GNUmakefile.am: Add mediastream/gstreamer in include directories list.
355         * GNUmakefile.list.am: Add DOMURLMediaStream files in the build.
356
357 2013-10-17  Andreas Kling  <akling@apple.com>
358
359         Use PassRef for constructing StylePropertySets.
360         <https://webkit.org/b/122948>
361
362         Make functions that construct StylePropertySets return PassRef
363         instead of PassRefPtr. Since they never return null, this gets rid
364         of the extra branch in ~PassRefPtr everywhere.
365
366         Also StyleRule* classes now hold a Ref<StylePropertySet>, codifying
367         the fact that they always have a property set.
368
369         Reviewed by Antti Koivisto.
370
371 2013-10-17  Andreas Kling  <akling@apple.com>
372
373         DataRef<T> should use Ref<T> internally.
374         <https://webkit.org/b/122953>
375
376         DataRef is used to hold RenderStyle substructures, and due to the
377         way style inheritance is implemented, DataRef will always point to
378         a live object.
379
380         Codify this by making DataRef::m_data a Ref, and making all methods
381         that create substructure objects return PassRef.
382
383         Reviewed by Antti Koivisto.
384
385 2013-10-17  Mihnea Ovidenie  <mihnea@adobe.com>
386
387         [CSS Regions] Anonymous nested regions
388         https://bugs.webkit.org/show_bug.cgi?id=119135
389
390         Reviewed by David Hyatt.
391
392         Tests: fast/regions/table-caption-as-region.html
393                fast/regions/table-cell-as-region.html
394
395         This patch allows any non-replaced block to behave like a region. When an element is styled with the
396         -webkit-flow-from property, instead of making the renderer a RenderRegion, we let the original
397         renderer be created the same way and we add a region as an anonymous child for the renderer.
398         The anonymous block child, modeled by the new RenderNamedFlowFragment class, will be responsible
399         for the fragmentation of the named flow thread content.
400
401         A RenderBlockFlow object will keep a reference to a RenderNamedFlowFragment(RenderRegion) inside its
402         rare data structures.
403
404         Contains code contributed by Catalin Badea.
405  
406         * CMakeLists.txt:
407         * GNUmakefile.list.am:
408         * WebCore.vcxproj/WebCore.vcxproj:
409         * WebCore.xcodeproj/project.pbxproj:
410         * dom/Element.cpp: Changed to take the anonymous region into account.
411         (WebCore::Element::renderRegion):
412         (WebCore::Element::webkitGetRegionFlowRanges):
413         * dom/WebKitNamedFlow.cpp: Ditto.
414         (WebCore::WebKitNamedFlow::firstEmptyRegionIndex):
415         (WebCore::WebKitNamedFlow::getRegionsByContent):
416         (WebCore::WebKitNamedFlow::getRegions):
417         * inspector/InspectorOverlay.cpp: Take into account the new model for regions, with an anonymous region inside a block.
418         (WebCore::buildObjectForRegionHighlight):
419         (WebCore::buildObjectForElementInfo):
420         * rendering/RenderBlock.cpp:
421         (WebCore::RenderBlock::computeShapeSize): For a render named flow fragment, there is no need to recompute the shape inside
422         we can take it from the parent.
423         (WebCore::RenderBlock::renderName): Make the block that contains a render named flow fragment (region) report RenderRegion.
424         A future patch that will change this will need to rebase a lot of tests.
425         * rendering/RenderBlockFlow.cpp:
426         (WebCore::RenderBlockFlow::insertedIntoTree): Create the anonymous region if needed (change of -webkit-flow-from determines Node reattach).
427         (WebCore::RenderBlockFlow::willBeDestroyed): Clean-up the anonymous region if necessary.
428         (WebCore::RenderBlockFlow::clearFloats): Small style change to make sure that check-webkit-style reports 0 failures on RenderBlockFlow.cpp.
429         (WebCore::RenderBlockFlow::layoutBlock):
430         (WebCore::RenderBlockFlow::styleDidChange): Update the style of the anonymous region too.
431         (WebCore::RenderBlockFlow::createRenderNamedFlowFragmentIfNeeded): Helper function to create the anonymous region
432         and to add it as a child to the block.
433         (WebCore::RenderBlockFlow::canHaveChildren):
434         (WebCore::RenderBlockFlow::canHaveGeneratedChildren):
435         (WebCore::RenderBlockFlow::namedFlowFragmentNeedsUpdate): Force a layout of the anonymous region if the
436         parent block has percentage height (similar to RenderBlock::updateBlockChildDirtyBitsBeforeLayout)
437         (WebCore::RenderBlockFlow::updateLogicalHeight): Update the logical height of anonymous region when the height of parent is updated.
438         (WebCore::RenderBlockFlow::setRenderNamedFlowFragment):
439         * rendering/RenderBlockFlow.h:
440         (WebCore::RenderBlockFlow::RenderBlockFlowRareData::RenderBlockFlowRareData):
441         (WebCore::RenderBlockFlow::renderNamedFlowFragment):
442         * rendering/RenderElement.cpp:
443         (WebCore::RenderElement::createFor): Remove the direct creation of RenderRegion objects since they will be
444         created as anonymous children of block flow objects.
445         * rendering/RenderElement.h:
446         (WebCore::RenderElement::generatingElement): Account for anonymous region if necessary.
447         * rendering/RenderFlowThread.cpp:
448         (WebCore::RenderFlowThread::adjustedPositionRelativeToOffsetParent): Take anonymous region into account
449         and use the anonymous region parent offset/border.
450         * rendering/RenderLayer.cpp:
451         (WebCore::RenderLayer::shouldBeNormalFlowOnly): Check for style instead of isRenderRegion since
452         the parent of the anonymous region will get the layer.
453         * rendering/RenderListItem.cpp:
454         (WebCore::RenderListItem::insertedIntoTree): Call RenderBlockFlow method instead.
455         * rendering/RenderNamedFlowFragment.cpp: Added. Model the behaviour of the anonymous region.
456         Has RenderRegion as a base class.
457         (WebCore::RenderNamedFlowFragment::RenderNamedFlowFragment):
458         (WebCore::RenderNamedFlowFragment::~RenderNamedFlowFragment):
459         (WebCore::RenderNamedFlowFragment::setStyleForNamedFlowFragment):
460         (WebCore::RenderNamedFlowFragment::styleDidChange):
461         (WebCore::RenderNamedFlowFragment::shouldHaveAutoLogicalHeight):
462         (WebCore::RenderNamedFlowFragment::maxPageLogicalHeight):
463         * rendering/RenderNamedFlowFragment.h: Added.
464         (WebCore::RenderNamedFlowFragment::isPseudoElementRegion):
465         (WebCore::RenderNamedFlowFragment::renderName):
466         (WebCore::toRenderNamedFlowFragment):
467         * rendering/RenderObject.cpp:
468         (WebCore::RenderObject::isRenderNamedFlowFragmentContainer):
469         * rendering/RenderObject.h:
470         (WebCore::RenderObject::isRenderNamedFlowFragment):
471         * rendering/RenderRegion.h:
472         * rendering/RenderTableCaption.cpp: Call RenderBlockFlow method instead.
473         (WebCore::RenderTableCaption::insertedIntoTree):
474         * rendering/RenderTreeAsText.cpp:
475         (WebCore::write):
476         (WebCore::writeRenderRegionList):
477         * rendering/shapes/ShapeInfo.h:
478         (WebCore::ShapeInfo::shapeSize):
479         * style/StyleResolveTree.cpp:
480         (WebCore::Style::elementInsideRegionNeedsRenderer):
481
482 2013-10-17  Krzysztof Czech  <k.czech@samsung.com>
483
484         [EFL] Properly expose tables in accessibility
485         https://bugs.webkit.org/show_bug.cgi?id=122894
486
487         Reviewed by Chris Fleizach.
488
489         All tables should exposed as tables.
490
491         * accessibility/AccessibilityTable.cpp:
492         (WebCore::AccessibilityTable::isTableExposableThroughAccessibility):
493         (WebCore::AccessibilityTable::addChildren):
494         * accessibility/AccessibilityTableColumn.cpp:
495         (WebCore::AccessibilityTableColumn::computeAccessibilityIsIgnored):
496         * accessibility/AccessibilityTableHeaderContainer.cpp:
497         (WebCore::AccessibilityTableHeaderContainer::computeAccessibilityIsIgnored):
498
499 2013-10-17  Brendan Long  <b.long@cablelabs.com>
500
501         [GStreamer] Too many arguments for format in WebKitWebAudioSourceGStreamer.cpp
502         https://bugs.webkit.org/show_bug.cgi?id=122932
503
504         Reviewed by Philippe Normand.
505
506         No new tests because this just fixes a build warning.
507
508         * platform/audio/gstreamer/WebKitWebAudioSourceGStreamer.cpp:
509         (webKitWebAudioSrcLoop): Add another %s for the second part of the pad name.
510
511 2013-10-16  Tim Horton  <timothy_horton@apple.com>
512
513         Attempt to fix the Windows build after http://trac.webkit.org/changeset/157547.
514
515         The relevant conversion doesn't happen implicitly.
516
517         * platform/graphics/win/MediaPlayerPrivateFullscreenWindow.cpp:
518         (WebCore::MediaPlayerPrivateFullscreenWindow::setRootChildLayer):
519         (WebCore::MediaPlayerPrivateFullscreenWindow::wndProc):
520
521 2013-10-16  Tim Horton  <timothy_horton@apple.com>
522
523         Attempt to fix the Windows build after http://trac.webkit.org/changeset/157547.
524
525         Apparently some Windows code uses setFrame and I missed it.
526
527         * platform/graphics/win/MediaPlayerPrivateFullscreenWindow.cpp:
528         (WebCore::MediaPlayerPrivateFullscreenWindow::setRootChildLayer):
529         (WebCore::MediaPlayerPrivateFullscreenWindow::wndProc):
530
531 2013-10-16  Tim Horton  <timothy_horton@apple.com>
532
533         Remote Layer Tree: Complete support for simple layer properties
534         https://bugs.webkit.org/show_bug.cgi?id=122933
535
536         Reviewed by Anders Carlsson.
537
538         No new tests, not yet testable.
539
540         * platform/graphics/ca/GraphicsLayerCA.cpp:
541         (WebCore::GraphicsLayerCA::setName):
542         Don't dump the CALayer pointer if we own a PlatformCALayerRemote.
543
544         (WebCore::GraphicsLayerCA::recursiveCommitChanges):
545         Fix the visible tile wash (my fault!), and make it use setPosition and
546         setBounds instead of setFrame; while more convenient, it is the only
547         caller of setFrame, so we'll remove it.
548
549         * platform/graphics/ca/PlatformCALayer.h:
550         (WebCore::PlatformCALayer::isRemote): Added.
551
552         * platform/graphics/ca/mac/PlatformCALayerMac.h:
553         * platform/graphics/ca/mac/PlatformCALayerMac.mm:
554         (nullActionsDictionary):
555         (toCAFilterType):
556         (PlatformCALayerMac::synchronouslyDisplayTilesInRect):
557         (PlatformCALayerMac::playerLayer):
558         Remove setFrame, fix some pointer sides.
559
560         * platform/graphics/ca/win/PlatformCALayerWin.cpp:
561         * platform/graphics/ca/win/PlatformCALayerWin.h:
562         Remove setFrame.
563
564 2013-10-16  Andreas Kling  <akling@apple.com>
565
566         Take RenderObjects out of the arena.
567         <https://webkit.org/b/122895>
568
569         Reviewed by Antti Koivisto.
570
571         Stop arena-allocating renderers so we can move forward on improving
572         render tree memory management. This will also allow rendering code
573         to take advantage of malloc optimizations.
574
575         Line boxes and BiDi runs remain in the arena for now.
576
577 2013-10-16  Roger Fong  <roger_fong@apple.com>
578
579         [Windows] Speculative fix for test, media/video-canvas-drawing-output.html.
580
581         * platform/graphics/avfoundation/cf/MediaPlayerPrivateAVFoundationCF.cpp:
582         (WebCore::AVFWrapper::createImageForTimeInRect):
583
584 2013-10-16  Tim Horton  <timothy_horton@apple.com>
585
586         PlatformCALayer constructor should take layer type as an argument
587         https://bugs.webkit.org/show_bug.cgi?id=122915
588
589         Reviewed by Simon Fraser.
590
591         No new tests, just a minor refactoring.
592
593         * platform/graphics/ca/PlatformCALayer.h:
594         (WebCore::PlatformCALayer::PlatformCALayer):
595         Add a LayerType argument.
596
597         * platform/graphics/ca/mac/PlatformCALayerMac.mm:
598         (PlatformCALayerMac::PlatformCALayerMac):
599         * platform/graphics/ca/win/PlatformCALayerWin.cpp:
600         (PlatformCALayerWin::PlatformCALayerWin):
601         Use the new LayerType argument, and early-return in the case where we
602         are wrapping a custom PlatformLayer.
603         Drive-by un-indent the switch in the Mac version.
604
605 2013-10-15  Brady Eidson  <beidson@apple.com>
606
607         Flesh out the DatabaseProcess (and launch it!)
608         https://bugs.webkit.org/show_bug.cgi?id=122884
609
610         Reviewed by Tim Horton.
611
612         * English.lproj/Localizable.strings: Add a localizable string.
613         * WebCore.xcodeproj/project.pbxproj: Export some more headers.
614
615 2013-10-16  Tim Horton  <timothy_horton@apple.com>
616
617         RemoteLayerTree: Add support for more layer properties and transform layers
618         https://bugs.webkit.org/show_bug.cgi?id=122906
619
620         Reviewed by Anders Carlsson.
621
622         No new tests, this code is not yet testable.
623
624         * WebCore.exp.in:
625         Export some TextStream and TransformationMatrix stuff.
626
627 2013-10-16  KyungTae Kim  <ktf.kim@samsung.com>
628
629         During editing, merge inline style with overriding other author styles
630         https://bugs.webkit.org/show_bug.cgi?id=122874
631
632         Reviewed by Ryosuke Niwa.
633
634         Inline styles need to override other author styles even on DoNotOverrideValues mode.
635         So, merge and override inline styles to other author styles before merging them to m_mutableStyle.
636
637         Test: editing/deleting/merge-div-with-inline-style.html
638
639         * editing/EditingStyle.cpp:
640         (WebCore::EditingStyle::mergeInlineAndImplicitStyleOfElement):
641
642 2013-10-16  peavo@outlook.com  <peavo@outlook.com>
643
644         Emphasis marks has wrong color.
645         https://bugs.webkit.org/show_bug.cgi?id=122829
646
647         Reviewed by Antti Koivisto.
648
649         Tests: fast/text/text-emphasis.html.
650                fast/text/text-emphasis-expected.html.
651
652         Emphasis color should be set as fill color, not stroke color.
653
654         * rendering/TextPaintStyle.cpp:
655         (WebCore::updateGraphicsContext):
656         * rendering/TextPaintStyle.h:
657
658 2013-10-16  Antti Koivisto  <antti@apple.com>
659
660         Move code for finding rendered character offset to RenderTextLineBoxes
661         https://bugs.webkit.org/show_bug.cgi?id=122892
662
663         Reviewed by Andreas Kling.
664
665         * rendering/RenderText.cpp:
666         (WebCore::RenderText::countRenderedCharacterOffsets):
667         (WebCore::RenderText::containsRenderedCharacterOffset):
668         
669             Renamed for consistency.
670
671         * rendering/RenderTextLineBoxes.cpp:
672         (WebCore::RenderTextLineBoxes::countCharacterOffsets):
673         
674             This used to be Position::renderedPosition.
675
676 2013-10-16  Andreas Kling  <akling@apple.com>
677
678         RenderElement::removeChild() should take child as a reference.
679         <https://webkit.org/b/122888>
680
681         We can't remove a child without a child to remove.
682
683         Reviewed by Antti Koivisto.
684
685 2013-10-16  Antti Koivisto  <antti@apple.com>
686
687         Move test for contained caret offset to RenderTextLineBoxes
688         https://bugs.webkit.org/show_bug.cgi?id=122887
689
690         Reviewed by Andreas Kling.
691
692         * dom/Position.cpp:
693         (WebCore::Position::renderedOffset):
694         (WebCore::Position::isCandidate):
695         
696             Remove isRenderedText, call RenderText::containsCaretOffset instead.
697
698         (WebCore::Position::isRenderedCharacter):
699         (WebCore::Position::rendersInDifferentPosition):
700         * dom/Position.h:
701         * dom/PositionIterator.cpp:
702         (WebCore::PositionIterator::isCandidate):
703         * rendering/InlineTextBox.cpp:
704         * rendering/InlineTextBox.h:
705         * rendering/RenderText.cpp:
706         (WebCore::RenderText::containsCharacterOffset):
707         (WebCore::RenderText::containsCaretOffset):
708         * rendering/RenderText.h:
709         * rendering/RenderTextLineBoxes.cpp:
710         (WebCore::RenderTextLineBoxes::containsOffset):
711         
712             Combined implementations of Position::isRenderedCharacter and Position::isRenderedText.
713
714         * rendering/RenderTextLineBoxes.h:
715         
716             Remove containsCaretOffset(), functionality is now in RenderTextLineBoxes::containsOffset.
717
718 2013-10-16  Andreas Kling  <akling@apple.com>
719
720         RenderElement::isChildAllowed() should take const references.
721         <https://webkit.org/b/122870>
722
723         Reviewed by Anders Carlsson.
724
725         The isChildAllowed() functions expect non-null values to be passed,
726         so enforce this at compile-time.
727
728         Reordered some checks to do bit tests before virtual calls.
729
730 2013-10-15  Philippe Normand  <pnormand@igalia.com>
731
732         [GStreamer] video info unset if upstream doesn't query allocation
733         https://bugs.webkit.org/show_bug.cgi?id=122834
734
735         Reviewed by Gustavo Noronha Silva.
736
737         * platform/graphics/gstreamer/VideoSinkGStreamer.cpp:
738         (webkitVideoSinkRender): If the sink didn't process any allocation
739         query then use the configured source pad caps and don't rely on
740         invalid video info.
741
742 2013-10-16  Philippe Normand  <pnormand@igalia.com>
743
744         [GStreamer] move Logging.h include to GStreamerUtilities.h
745         https://bugs.webkit.org/show_bug.cgi?id=122886
746
747         Reviewed by Gustavo Noronha Silva.
748
749         Include Logging.h from GStreamerUtilities.h so the modules using
750         the LOG_MEDIA macros don't need to bother, especially for Debug builds.
751
752         * platform/graphics/gstreamer/GStreamerUtilities.h:
753         * platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp:
754         * platform/graphics/gstreamer/MediaPlayerPrivateGStreamerBase.cpp:
755
756 2013-10-16  Mark Rowe  <mrowe@apple.com>
757
758         Fix the build after r157478.
759
760         Rubber-stamped by Tim Horton.
761
762         Due to the way WebCore.exp.in is used, it can't be used to export a differing set of symbols
763         for different architectures. We often work around this by tweaking code slightly to avoid
764         needing to export different symbols. However, in this case the symbol name itself encodes an
765         architecture-specific detail and there's no clear way to avoid the requirement to export it.
766
767         To deal with this case we turn to ld's support for wildcards in the symbol export list.
768
769         * WebCore.exp.in: Use wildcards in place of the number that represents by how much "this"
770         should be adjusted when calling through the vtable thunk. Also sort the remainder of the file.
771         * make-export-file-generator: Don't attempt to verify symbol names that contain wildcard characters.
772
773 2013-10-15  Tim Horton  <timothy_horton@apple.com>
774
775         Two more exports for 32-bit build fix.
776         The duality of CGFloat means that we use a different
777         version of getRGBA (and the cast operator)
778         on 32-bit systems.
779
780         * WebCore.exp.in:
781
782 2013-10-15  Tim Horton  <timothy_horton@apple.com>
783
784         Another shot at a build fix; apparently these
785         do need exporting for some reason, but are different
786         on 32-bit.
787
788         * WebCore.exp.in:
789
790 2013-10-15  Alexey Proskuryakov  <ap@apple.com>
791
792         GenerateIsReachable=ImplContext is confusing
793         https://bugs.webkit.org/show_bug.cgi?id=122864
794
795         Reviewed by Geoffrey Garen.
796
797         Renamed to ImplWebGLRenderingContext.
798
799         * bindings/scripts/CodeGeneratorJS.pm:
800         (GenerateImplementation):
801         * bindings/scripts/IDLAttributes.txt:
802         * html/canvas/EXTDrawBuffers.idl:
803         * html/canvas/EXTTextureFilterAnisotropic.idl:
804         * html/canvas/OESElementIndexUint.idl:
805         * html/canvas/OESStandardDerivatives.idl:
806         * html/canvas/OESTextureFloat.idl:
807         * html/canvas/OESTextureHalfFloat.idl:
808         * html/canvas/OESVertexArrayObject.idl:
809         * html/canvas/WebGLCompressedTextureATC.idl:
810         * html/canvas/WebGLCompressedTexturePVRTC.idl:
811         * html/canvas/WebGLCompressedTextureS3TC.idl:
812         * html/canvas/WebGLDebugRendererInfo.idl:
813         * html/canvas/WebGLDebugShaders.idl:
814         * html/canvas/WebGLDepthTexture.idl:
815         * html/canvas/WebGLLoseContext.idl:
816
817         * WebCore.xcodeproj/project.pbxproj: While at it, added OESElementIndexUint.idl
818         to Xcode project.
819
820 2013-10-15  Tim Horton  <timothy_horton@apple.com>
821
822         Unreviewed build fix; I don't know how to export.
823         This may not help.
824
825         * WebCore.exp.in:
826
827 2013-10-15  Dean Jackson  <dino@apple.com>
828
829         Add ENABLE_WEB_ANIMATIONS flag
830         https://bugs.webkit.org/show_bug.cgi?id=122871
831
832         Reviewed by Tim Horton.
833
834         Eventually might be http://dev.w3.org/fxtf/web-animations/
835         but this is just engine-internal work at the moment.
836
837         * Configurations/FeatureDefines.xcconfig:
838
839 2013-10-15  Tim Horton  <timothy_horton@apple.com>
840
841         Add a PlatformCALayer subclass that proxies its property changes across the process boundary
842         https://bugs.webkit.org/show_bug.cgi?id=122773
843
844         Reviewed by Anders Carlsson.
845
846         No new tests, the new drawing area is not yet testable.
847
848         * WebCore.exp.in:
849         Export lots of GraphicsLayerCA stuff so we can inherit from it in WebKit2.
850
851         * WebCore.xcodeproj/project.pbxproj:
852         Make PlatformCAFilters.h a private header.
853
854         * platform/graphics/GraphicsLayer.h:
855         (WebCore::GraphicsLayer::initialize):
856         * platform/graphics/ca/GraphicsLayerCA.cpp:
857         (WebCore::GraphicsLayer::create):
858         (WebCore::GraphicsLayerCA::GraphicsLayerCA):
859         (WebCore::GraphicsLayerCA::initialize):
860         * platform/graphics/ca/GraphicsLayerCA.h:
861         Defer creation of the main PlatformCALayer until just after the constructor is finished
862         so that GraphicsLayerCA subclasses can successfully override createPlatformCALayer.
863
864         * platform/graphics/ca/PlatformCALayer.h:
865         (WebCore::PlatformCALayer::platformLayer):
866         Make platformLayer virtual so that subclasses which don't have PlatformLayers can override.
867
868 2013-10-14  Ryosuke Niwa  <rniwa@webkit.org>
869
870         REGRESSION: Crash in XMLDocumentParser::startElementNs
871         https://bugs.webkit.org/show_bug.cgi?id=122817
872
873         Reviewed by Darin Adler.
874
875         Exit early in startElementNs when listeners and handlers of synchronous events such as load event
876         removes the inserted node inside parserAppendChild.
877
878         Test: fast/parser/xhtml-synchronous-detach-crash.html
879
880         * xml/parser/XMLDocumentParserLibxml2.cpp:
881         (WebCore::XMLDocumentParser::startElementNs):
882
883 2013-10-15  Joseph Pecoraro  <pecoraro@apple.com>
884
885         Web Inspector: Remove old frontend localizedStrings.js
886         https://bugs.webkit.org/show_bug.cgi?id=122846
887
888         Reviewed by Timothy Hatcher.
889
890         * Configurations/WebCore.xcconfig:
891         * English.lproj/localizedStrings.js: Removed.
892         * WebCore.xcodeproj/project.pbxproj:
893         Remove the file and references to it. We no longer need to exclude
894         localizedString.js from some builds.
895
896 2013-10-15  Joseph Pecoraro  <pecoraro@apple.com>
897
898         Web Inspector: Remove Windows old front-end related code
899         https://bugs.webkit.org/show_bug.cgi?id=122845
900
901         Reviewed by Brent Fulgham.
902
903         * WebCore.vcxproj/copyWebCoreResourceFiles.cmd:
904
905 2013-10-15  Morten Stenshorne  <mstensho@opera.com>
906
907         Add support for the column-fill property
908         https://bugs.webkit.org/show_bug.cgi?id=117693
909
910         Reviewed by David Hyatt.
911
912         This is only supported in the (new) region based multicol implementation.
913
914         With column-fill support, a lot of multicol tests needed an update.
915         The old implementation behaved as if column-fill were 'auto', but the
916         initial value is 'balance', so now we need to be explicit about that.
917         For auto-height tests it doesn't really matter - such multicols are always
918         balanced anyway.
919
920         Tests: fast/multicol/newmulticol/fixed-height-fill-auto.html
921                fast/multicol/newmulticol/fixed-height-fill-balance.html
922
923         * css/CSSComputedStyleDeclaration.cpp:
924         (WebCore::ComputedStyleExtractor::propertyValue):
925         * css/CSSParser.cpp:
926         (WebCore::isValidKeywordPropertyAndValue):
927         (WebCore::isKeywordPropertyID):
928         (WebCore::CSSParser::parseValue):
929         * css/CSSPrimitiveValueMappings.h:
930         (WebCore::CSSPrimitiveValue::CSSPrimitiveValue):
931         (WebCore::CSSPrimitiveValue::operator ColumnFill):
932         * css/CSSPropertyNames.in:
933         * css/CSSValueKeywords.in:
934         * css/DeprecatedStyleBuilder.cpp:
935         (WebCore::DeprecatedStyleBuilder::DeprecatedStyleBuilder):
936         * rendering/RenderMultiColumnBlock.h:
937         * rendering/style/RenderStyle.h:
938         * rendering/style/RenderStyleConstants.h:
939         * rendering/style/StyleMultiColData.cpp:
940         (WebCore::StyleMultiColData::StyleMultiColData):
941         (WebCore::StyleMultiColData::operator==):
942         * rendering/style/StyleMultiColData.h:
943
944 2013-10-15  Andreas Kling  <akling@apple.com>
945
946         Skip unnecessary null check in RenderText::textLength().
947         <https://webkit.org/b/122841>
948
949         Reviewed by Antti Koivisto.
950
951         RenderText will never have a null String in m_text, so textLength()
952         can grab at the StringImpl directly, avoiding a null check.
953
954 2013-10-15  Andreas Kling  <akling@apple.com>
955
956         FontGenericFamilies should not be ref-counted.
957         <https://webkit.org/b/122835>
958
959         Reviewed by Anders Carlsson.
960
961         FontGenericFamilies is singly-owned by Settings.
962
963 2013-10-15  Zoltan Horvath  <zoltan@webkit.org>
964
965         [CSS Shapes] Move RenderBlock::layoutShapeInsideInfo into RenderBlock.cpp
966         http://bugs.webkit.org/show_bug.cgi?id=122843
967
968         Reviewed by Oliver Hunt.
969
970         Historically, layoutShapeInsideInfo was a static function in RenderBlockLineLayout, then it has changed to be a member of RenderBlock,
971         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
972         an unnecessary CSS_SHAPES #ifdef as well from RenderBlock.cpp.
973
974         No new tests, no behavior change.
975
976         * rendering/RenderBlock.cpp:
977         (WebCore::RenderBlock::markShapeInsideDescendantsForLayout):
978         (WebCore::RenderBlock::layoutShapeInsideInfo):
979         * rendering/RenderBlockLineLayout.cpp:
980
981 2013-10-15  peavo@outlook.com  <peavo@outlook.com>
982
983         [WinCairo] Build fails.
984         https://bugs.webkit.org/show_bug.cgi?id=122830
985
986         Reviewed by Brent Fulgham.
987
988         * platform/network/NetworkStorageSessionStub.cpp:
989         (WebCore::NetworkStorageSession::createPrivateBrowsingSession): Update to new return type.
990
991 2013-10-15  Andreas Kling  <akling@apple.com>
992
993         FileIconLoader should not be ref-counted.
994         <https://webkit.org/b/122827>
995
996         FileIconLoader is singly-owned by FileInputType.
997
998         Reviewed by Antti Koivisto.
999
1000 2013-10-15  Andreas Kling  <akling@apple.com>
1001
1002         RenderText should cache RenderStyle in locals more.
1003         <https://webkit.org/b/122823>
1004
1005         Reviewed by Antti Koivisto.
1006
1007         Now that fetching the RenderStyle has to go through the parent,
1008         we should avoid unnecessary loads by caching style() in a local.
1009
1010 2013-10-15  Csaba Osztrogon√°c  <ossy@webkit.org>
1011
1012         Move PlatformCertificateInfo to WebCore and make the ResourceResponse primitives work in terms of that platform agnostic object
1013         https://bugs.webkit.org/show_bug.cgi?id=118520
1014
1015         Reviewed by Anders Carlsson.
1016
1017         Added PlatformCertificateInfo getter and setter to ResourceErrorBase
1018         and ResourceResponseBase so that getting and setting certificates
1019         becomes cross platform.
1020
1021         Changed the existing platform specific certificate getters and setters
1022         of ResourceError and ResourceResponse to use the
1023         PlatformCertificateInfo member.
1024
1025         Original patch by Kwang Yul Seo  <skyul@company100.net>
1026
1027         The following things were fixed by Csaba Osztrogon√°c:
1028         - trivial conflicts resolved in WebCore.exp.in, project.pbxproj, NetworkResourceLoader.cpp,
1029           WebPageProxy.h, WebFrameLoaderClient.cpp and PlatformEfl.cmake.
1030         - trivial fix in the new AsynchronousNetworkLoaderClient.cpp
1031         - style fixed (NULL -> 0 and smaller indentation in WebCoreArgumentCodersSoup.cpp)
1032         - obsolete change removed from NetworkResourceLoader.cpp
1033         - unneeded typo removed from ResourceHandleSoup.cpp
1034         - Fixed the if guard of including RetainPtr.h in PlatformCertificateInfo.h to make Windows build happy.
1035         - add PlatformCertificateInfoCFNet.cpp with empty constructor and destructor to make Windows build happy.
1036         - resolved conflict in Source/WebCore/WebCore.xcodeproj/project.pbxproj after r156488
1037         - renamed KURL to URL after r156550.
1038         - removed obsolete Qt changes
1039         - added back initializers for m_soupFlags
1040         - included gio/gio.h instead of libsoup/soup.h in PlatformCertificateInfo.h
1041         - removed unused class forward declarations and includes
1042
1043         * GNUmakefile.list.am:
1044         * PlatformEfl.cmake:
1045         * Target.pri:
1046         * WebCore.exp.in:
1047         * WebCore.vcxproj/WebCore.vcxproj:
1048         * WebCore.vcxproj/WebCore.vcxproj.filters:
1049         * WebCore.xcodeproj/project.pbxproj:
1050         * platform/network/PlatformCertificateInfo.h: Renamed from Source/WebKit2/Shared/soup/PlatformCertificateInfo.h.
1051         (WebCore::PlatformCertificateInfo::certificate):
1052         (WebCore::PlatformCertificateInfo::setCertificate):
1053         (WebCore::PlatformCertificateInfo::tlsErrors):
1054         (WebCore::PlatformCertificateInfo::setTLSErrors):
1055         * platform/network/ResourceErrorBase.h:
1056         (WebCore::ResourceErrorBase::platformCertificateInfo):
1057         (WebCore::ResourceErrorBase::setPlatformCertificateInfo):
1058         * platform/network/ResourceResponseBase.h:
1059         (WebCore::ResourceResponseBase::platformCertificateInfo):
1060         (WebCore::ResourceResponseBase::setPlatformCertificateInfo):
1061         * platform/network/cf/ResourceResponse.h:
1062         * platform/network/mac/PlatformCertificateInfoMac.mm: Renamed from Source/WebKit2/Shared/mac/PlatformCertificateInfo.mm.
1063         (WebCore::PlatformCertificateInfo::PlatformCertificateInfo):
1064         (WebCore::PlatformCertificateInfo::~PlatformCertificateInfo):
1065         (WebCore::PlatformCertificateInfo::setCertificateChain):
1066         (WebCore::PlatformCertificateInfo::certificateChain):
1067         (WebCore::PlatformCertificateInfo::dump):
1068         * platform/network/mac/ResourceResponseMac.mm:
1069         (WebCore::ResourceResponse::setCertificateChain):
1070         (WebCore::ResourceResponse::certificateChain):
1071         * platform/network/soup/PlatformCertificateInfoSoup.cpp:
1072         (WebCore::PlatformCertificateInfo::PlatformCertificateInfo):
1073         (WebCore::PlatformCertificateInfo::~PlatformCertificateInfo):
1074         * platform/network/soup/ResourceError.h:
1075         (WebCore::ResourceError::ResourceError):
1076         (WebCore::ResourceError::tlsErrors):
1077         (WebCore::ResourceError::setTLSErrors):
1078         (WebCore::ResourceError::certificate):
1079         (WebCore::ResourceError::setCertificate):
1080         * platform/network/soup/ResourceErrorSoup.cpp:
1081         (WebCore::ResourceError::platformCopy):
1082         * platform/network/soup/ResourceResponse.h:
1083         (WebCore::ResourceResponse::ResourceResponse):
1084         (WebCore::ResourceResponse::soupMessageCertificate):
1085         (WebCore::ResourceResponse::setSoupMessageCertificate):
1086         (WebCore::ResourceResponse::soupMessageTLSErrors):
1087         (WebCore::ResourceResponse::setSoupMessageTLSErrors):
1088         * platform/network/soup/ResourceResponseSoup.cpp:
1089         (WebCore::ResourceResponse::toSoupMessage):
1090         (WebCore::ResourceResponse::updateFromSoupMessage):
1091
1092 2013-10-15  Ryosuke Niwa  <rniwa@webkit.org>
1093
1094         Remove redundant Document::getElementById
1095         https://bugs.webkit.org/show_bug.cgi?id=122813
1096
1097         Reviewed by Andreas Kling.
1098
1099         Merge https://chromium.googlesource.com/chromium/blink/+/4e8f1c5316415614b84370c602beae4a1008299f
1100
1101         This function simply calls virtual TreeScope::getElementById and Document inherits from TreeScope.
1102
1103         * WebCore.exp.in:
1104         * dom/Document.cpp:
1105         * dom/Document.h:
1106
1107 2013-10-14  Santosh Mahto  <santosh.ma@samsung.com>
1108
1109         in safari,the background-color of input[type="search"] can't work
1110         https://bugs.webkit.org/show_bug.cgi?id=119967
1111
1112         Reviewed by Ryosuke Niwa.
1113
1114         When input type=search is styled with css background property then
1115         it does not change the background-color of field. Its happening becasue
1116         search field is not counted as styled control. Thus theme ignores the 
1117         css background property. With this patch search field is also counted as 
1118         styled control so background property reflects on search field.
1119
1120         Test: fast/forms/search/search-field-background-color.html
1121
1122         * rendering/RenderTheme.cpp:
1123         (WebCore::RenderTheme::isControlStyled):Now search field is also
1124         a styled control.
1125
1126 2013-10-14  Ryosuke Niwa  <rniwa@webkit.org>
1127
1128         EventPath::updateTouchLists traverses through EventPath thrice
1129         https://bugs.webkit.org/show_bug.cgi?id=122804
1130
1131         Reviewed by Benjamin Poulain.
1132
1133         Instead of traversing through EventPath for each TouchList, traverse through TouchList for every EventContext.
1134         This paves our way to have one-pass traversal over EventPath, and evetually to remove EventContext altogether.
1135
1136         This change should also improve the cache hit rate since all Touch objects tend to be allocated at the same time
1137         but this performance improvement is probably not observable.
1138
1139         * dom/EventContext.h:
1140         * dom/EventDispatcher.cpp:
1141         (WebCore::EventRelatedNodeResolver::EventRelatedNodeResolver): Added a new constructor that takes Touch and
1142         and TouchListType. We need to store these two values in order to update EventContext later.
1143         (WebCore::EventRelatedNodeResolver::touch): Added,
1144         (WebCore::EventRelatedNodeResolver::touchListType): Added.
1145         (WebCore::addRelatedNodeResolversForTouchList): Extracted from updateTouchListsInEventPath.
1146         (WebCore::EventPath::updateTouchLists): Moved the loop over m_path here. Notice that the outer loop iterates
1147         over m_path instead of touchList as done in updateTouchListsInEventPath. The inner loop goes through resolvers
1148         and adds Touch objects each EventContext as needed.
1149
1150 2013-10-14  Alexey Proskuryakov  <ap@apple.com>
1151
1152         Don't generate a wasteful isObservable check in isReachableFromOpaqueRoots
1153         https://bugs.webkit.org/show_bug.cgi?id=122802
1154
1155         Reviewed by Mark Hahnenberg.
1156
1157         * bindings/scripts/CodeGeneratorJS.pm: (GenerateImplementation): Don't.
1158
1159         * bindings/scripts/test/JS/JSTestActiveDOMObject.cpp:
1160         * bindings/scripts/test/JS/JSTestCustomNamedGetter.cpp:
1161         * bindings/scripts/test/JS/JSTestEventConstructor.cpp:
1162         * bindings/scripts/test/JS/JSTestEventTarget.cpp:
1163         * bindings/scripts/test/JS/JSTestException.cpp:
1164         * bindings/scripts/test/JS/JSTestInterface.cpp:
1165         * bindings/scripts/test/JS/JSTestMediaQueryListListener.cpp:
1166         * bindings/scripts/test/JS/JSTestNamedConstructor.cpp:
1167         * bindings/scripts/test/JS/JSTestObj.cpp:
1168         * bindings/scripts/test/JS/JSTestOverloadedConstructors.cpp:
1169         * bindings/scripts/test/JS/JSTestSerializedScriptValueInterface.cpp:
1170         * bindings/scripts/test/JS/JSTestTypedefs.cpp:
1171         * bindings/scripts/test/JS/JSattribute.cpp:
1172         * bindings/scripts/test/JS/JSreadonly.cpp:
1173         Updated results.
1174
1175 2013-10-14  Samuel White  <samuel_white@apple.com>
1176
1177         AX: fieldset should have GroupRole and legend should be description.
1178         https://bugs.webkit.org/show_bug.cgi?id=122534
1179
1180         Reviewed by Chris Fleizach.
1181
1182         Changes fieldset to derive AXDescription from legend if one is available. Added
1183         convenience method to AccessibilityObject to fetch element if available.
1184
1185         Test: accessibility/fieldset-element.html
1186
1187         * accessibility/AccessibilityNodeObject.cpp:
1188         (WebCore::AccessibilityNodeObject::canHaveChildren):
1189         (WebCore::AccessibilityNodeObject::alternativeText):
1190         * accessibility/AccessibilityObject.cpp:
1191         (WebCore::AccessibilityObject::element):
1192         (WebCore::AccessibilityObject::isARIAHidden):
1193         (WebCore::AccessibilityObject::isDOMHidden):
1194         (WebCore::AccessibilityObject::defaultObjectInclusion):
1195         * accessibility/AccessibilityObject.h:
1196         (WebCore::AccessibilityObject::isHidden):
1197         * accessibility/AccessibilitySlider.cpp:
1198         (WebCore::AccessibilitySlider::getAttribute):
1199         (WebCore::AccessibilitySlider::valueForRange):
1200         (WebCore::AccessibilitySlider::maxValueForRange):
1201         (WebCore::AccessibilitySlider::minValueForRange):
1202         (WebCore::AccessibilitySlider::setValue):
1203         (WebCore::AccessibilitySlider::inputElement):
1204         * accessibility/AccessibilitySlider.h:
1205         * accessibility/mac/AccessibilityObjectMac.mm:
1206         (WebCore::AccessibilityObject::accessibilityPlatformIncludesObject):
1207         * html/HTMLFieldSetElement.cpp:
1208         (WebCore::HTMLFieldSetElement::legend):
1209         * html/HTMLFieldSetElement.h:
1210
1211 2013-10-14  Roger Fong  <roger_fong@apple.com>
1212
1213         Windows select element doesn't draw RTL properly.
1214         https://bugs.webkit.org/show_bug.cgi?id=122785.
1215
1216         Reviewed by Brent Fulgham.
1217
1218         Covered by fast/text/international/pop-up-button-text-alignment-and-direction.html.
1219
1220         Problems include the popup items not drawing on the right hand side and 
1221         not respecting the direction or the directional override styling of the option.
1222         The selected element (drawn in the actual select element) also doesn't respect 
1223         the style settings of the selected menu option.
1224
1225         * platform/win/PopupMenuWin.cpp:
1226         (WebCore::PopupMenuWin::paint):
1227
1228 2013-10-14  Roger Fong  <roger_fong@apple.com>
1229
1230         [Windows] Unreviewed build fix.
1231
1232         * WebCore.vcxproj/WebCoreCommon.props:
1233
1234 2013-10-14  Ryosuke Niwa  <rniwa@webkit.org>
1235
1236         Crash in WebCore::BidiResolver<WebCore::InlineIterator, WebCore::BidiRun>::createBidiRunsForLine
1237         https://bugs.webkit.org/show_bug.cgi?id=122776
1238
1239         Reviewed by Darin Adler.
1240
1241         Merge https://chromium.googlesource.com/chromium/blink/+/aca89bc4d984705a1f94b623dae0ab03e239a248
1242
1243         Fix modification of whitespace endpoints to not assume it's operating on RenderTexts
1244
1245         During line layout, we use midpoints to identify RenderObjects, or parts of
1246         RenderObjects, that don't need InlineBoxes, usually because of collapsed whitespace.
1247
1248         Prior to actually creating BidiRuns (the precursor to InlineBoxes), we use
1249         checkMidpoints to fix up our lineMidpointState to handle the case where we start
1250         ignoring spaces in our line, but don't stop until somewhere on the following line.
1251         Previously, this function assumed that the final midpoint (called an endpoint)
1252         was a RenderText, but this assumption is wrong if we have a beginning midpoint
1253         created by shouldSkipWhitespaceAfterStartObject (which handles inlines and list
1254         markers) and no endpoint on that line. In that case, we'd instead adjust the
1255         position backwards on the beginning midpoint, which would cause us to fail to
1256         create an InlineBox for the inline or list marker. In the new test added, this
1257         would actually trigger a crash due to an assumption when visually re-ordering
1258         BidiRuns that a non-empty line would actually contain at least one such run.
1259
1260         Test: fast/text/whitespace/whitespace-and-margin-wrap-after-list-marker-crash.html
1261
1262         * rendering/RenderBlockLineLayout.cpp:
1263         (WebCore::checkMidpoints):
1264
1265 2013-10-14  Ryosuke Niwa  <rniwa@webkit.org>
1266
1267         Assertion failure in Range::processContentsBetweenOffsets
1268         https://bugs.webkit.org/show_bug.cgi?id=122777
1269
1270         Reviewed by Darin Adler.
1271
1272         Merge https://chromium.googlesource.com/chromium/blink/+/c15de182774c7859c20d97126eb844ae97b792a4
1273
1274         This patch changes ASSERT statements for checking |endOffset| inbound in Range::processContentsBetweenOffsets()
1275         to limit |endOffset|. This is necessary when DOMNodeRemovedFromDocument event handler splits text nodes,
1276         Range::insertNode() on text node, in the range calling Range::deleteContents().
1277
1278         Test: fast/dom/Range/range-delete-contents-mutation-event-crash.html
1279
1280         * dom/Range.cpp:
1281         (WebCore::Range::processContentsBetweenOffsets):
1282
1283 2013-10-14  Alexey Proskuryakov  <ap@apple.com>
1284
1285         Add an empty window.crypto.webkitSubtle
1286         https://bugs.webkit.org/show_bug.cgi?id=122778
1287
1288         Reviewed by Mark Hahnenberg.
1289
1290         Tests: security/crypto-subtle-gc-2.html
1291                security/crypto-subtle-gc-3.html
1292                security/crypto-subtle-gc.html
1293
1294         * DerivedSources.make: Process SubtleCrypto.idl.
1295
1296         * crypto: Added.
1297         * WebCore.xcodeproj/project.pbxproj:
1298         * CMakeLists.txt:
1299         * DerivedSources.make:
1300         * GNUmakefile.am:
1301         * GNUmakefile.list.am:
1302         * WebCore.vcxproj/WebCore.vcxproj.filters:
1303
1304         * bindings/js/JSSubtleCryptoCustom.cpp: Added. Empty for now, but we'll certainly
1305         need custom bindings code here.
1306
1307         * crypto/SubtleCrypto.cpp: Added.
1308         (WebCore::SubtleCrypto::SubtleCrypto):
1309         (WebCore::SubtleCrypto::document):
1310         * crypto/SubtleCrypto.h: Added.        
1311         * crypto/SubtleCrypto.idl: Added.
1312         * page/Crypto.cpp:
1313         (WebCore::Crypto::subtle):
1314         * page/Crypto.h:
1315         * page/Crypto.idl:
1316
1317 2013-10-14  Nick Diego Yamane  <nick.yamane@openbossa.org>
1318
1319         Remove GestureEvent leftovers from WebCore
1320         <https://webkit.org/b/122780>
1321
1322         Reviewed by Anders Carlsson.
1323
1324         - Removed some remaining references to PlatformGestureEvent, supposed to
1325           be removed by r157316
1326         - TOUCH_ADJUSTMENT should be reworked after GestureEvent feature
1327           removal
1328
1329         * page/EventHandler.cpp:
1330         (WebCore::EventHandler::bestZoomableAreaForTouchPoint):
1331         * page/EventHandler.h:
1332         * platform/PlatformEvent.h:
1333         * platform/ScrollAnimatorNone.cpp:
1334         * platform/ScrollableArea.h:
1335
1336 2013-10-14  Nick Diego Yamane  <nick.yamane@openbossa.org>
1337
1338         Build fix after r157366
1339         http://bugs.webkit.org/show_bug.cgi?id=122783
1340
1341         When TOUCH_AJUSTMENT is enabled build fails due to some
1342         refactors in TextRender functions.
1343
1344         Reviewed by Anders Carlsson.
1345
1346         * page/TouchAdjustment.cpp:
1347         (WebCore::TouchAdjustment::appendContextSubtargetsForNode):
1348
1349 2013-10-14  Alexandru Chiculita  <achicu@adobe.com>
1350
1351         The content of the DOM panel for iframes is not updated until the "onload" event
1352         https://bugs.webkit.org/show_bug.cgi?id=122653
1353
1354         Reviewed by Darin Adler.
1355
1356         Test: http/tests/inspector-protocol/loading-iframe-document-node.html
1357
1358         Renamed InspectorDOMAgent::loadEventFired to InspectorDOMAgent::didCommitLoad and moved the call site
1359         from InspectorInstrumentation::loadEventFiredImpl to InspectorInstrumentation::didCommitLoadImpl.
1360         This is to make sure that it will invalidate the content of the iframe as soon as the frame navigates
1361         to a different page. This way the new node can be retrieved as soon as the page has some content, and
1362         not just when the page is fully loaded.
1363
1364         * inspector/InspectorDOMAgent.cpp:
1365         (WebCore::InspectorDOMAgent::didCommitLoad): Renamed from loadEventFired, as it is now called from
1366         didCommitLoadImpl instead.
1367         (WebCore::InspectorDOMAgent::frameDocumentUpdated): Updated comment to point to the new function name.
1368         * inspector/InspectorDOMAgent.h:
1369         * inspector/InspectorInstrumentation.cpp:
1370         (WebCore::InspectorInstrumentation::loadEventFiredImpl): Removed call do InspectorDOMAgent.loadEventFired.
1371         (WebCore::InspectorInstrumentation::didCommitLoadImpl): Added call to InspectorDOMAgent.didCommitLoad.
1372
1373 2013-10-14  Roger Fong  <roger_fong@apple.com>
1374
1375         https://bugs.webkit.org/show_bug.cgi?id=122774.
1376         <rdar://problem/6138855>.
1377
1378         Reviewed by Brent Fulgham.
1379
1380         Add a field to keep track of hovered over index.
1381         Use index to determine whether or not to use the existing selected index on the mouse down event.
1382
1383         * platform/win/PopupMenuWin.cpp:
1384         (WebCore::PopupMenuWin::PopupMenuWin):
1385         (WebCore::PopupMenuWin::show):
1386         (WebCore::PopupMenuWin::wndProc):
1387         * platform/win/PopupMenuWin.h:
1388
1389 2013-10-14  Tim Horton  <timothy_horton@apple.com>
1390
1391         Virtualize PlatformCALayer
1392         https://bugs.webkit.org/show_bug.cgi?id=122672
1393
1394         Reviewed by Anders Carlsson.
1395
1396         No new tests, just a refactoring.
1397
1398         * WebCore.exp.in:
1399         setGeometryFlipped is on PlatformCALayerMac now.
1400
1401         * WebCore.vcxproj/WebCore.vcxproj:
1402         * WebCore.vcxproj/WebCore.vcxproj.filters:
1403         Add PlatformCALayer.cpp, PlatformCALayerWin.h, and let VS do its
1404         thing with some other files.
1405
1406         * WebCore.xcodeproj/project.pbxproj:
1407         Add PlatformCALayer.cpp and PlatformCALayerMac.h.
1408
1409         * platform/graphics/avfoundation/cf/MediaPlayerPrivateAVFoundationCF.cpp:
1410         (WebCore::AVFWrapper::platformLayer):
1411         Make a PlatformCALayerWin explicitly.
1412
1413         * platform/graphics/ca/GraphicsLayerCA.cpp:
1414         (WebCore::GraphicsLayerCA::createPlatformCALayer):
1415         Added. Decide whether to make a PlatformCALayer{Mac, Win} based on the platform.
1416         Later, we will decide between other subclasses based on other things.
1417
1418         (WebCore::GraphicsLayerCA::filtersCanBeComposited):
1419         Do the same thing for filtersCanBeComposited.
1420
1421         (WebCore::GraphicsLayerCA::GraphicsLayerCA):
1422         (WebCore::GraphicsLayerCA::setContentsToSolidColor):
1423         (WebCore::GraphicsLayerCA::setContentsToMedia):
1424         (WebCore::GraphicsLayerCA::setContentsToCanvas):
1425         (WebCore::GraphicsLayerCA::recursiveCommitChanges):
1426         (WebCore::GraphicsLayerCA::ensureStructuralLayer):
1427         (WebCore::GraphicsLayerCA::updateContentsImage):
1428         (WebCore::GraphicsLayerCA::updateContentsRects):
1429         (WebCore::GraphicsLayerCA::swapFromOrToTiledLayer):
1430         Use createPlatformCALayer instead of PlatformCALayer::create.
1431
1432         * platform/graphics/ca/mac/PlatformCAFiltersMac.mm:
1433         For now, use the PlatformCALayerMac version of filtersCanBeComposited,
1434         since this code is heavily tied to having CALayers in the Web process.
1435
1436         * platform/graphics/ca/GraphicsLayerCA.h:
1437         Include PlatformCALayer.h here so we can get the LayerType enum.
1438         (NOTE-to-be-removed: if there's a better way to do this, I'm open to
1439         suggestions; I couldn't puzzle out nested 'enum class' stuff).
1440
1441         Add the createPlatformCALayers.
1442
1443         * platform/graphics/ca/PlatformCAAnimation.h:
1444         Friend the subclasses too.
1445
1446         * platform/graphics/ca/PlatformCALayer.cpp: Added.
1447         (WebCore::PlatformCALayer::~PlatformCALayer):
1448         Pull the shared part of the PlatformCALayer destructor out.
1449
1450         * platform/graphics/ca/PlatformCALayer.h:
1451         (WebCore::PlatformCALayer::platformLayer):
1452         (WebCore::PlatformCALayer::setOwner):
1453         (WebCore::PlatformCALayer::PlatformCALayer):
1454         Virtualize all the things. Move platform specific members to their new subclasses.
1455
1456         * platform/graphics/ca/mac/PlatformCALayerMac.h: Added.
1457         * platform/graphics/ca/mac/PlatformCALayerMac.mm:
1458         Move the PLATFORM(MAC) implementations from PlatformCALayer to PlatformCALayerMac.
1459
1460         (PlatformCALayer::platformCALayer):
1461         The platformCALayer lookup function is static on PlatformCALayer, so it can't
1462         be moved to the subclasses. It might be a good idea in the future to move towards
1463         a platform-independent mechanism for looking up PlatformCALayers from PlatformLayers,
1464         and to avoid needing to do this as often as we do now.
1465
1466         * platform/graphics/ca/win/CACFLayerTreeHost.cpp:
1467         (WebCore::CACFLayerTreeHost::CACFLayerTreeHost):
1468         * platform/graphics/ca/win/PlatformCALayerWin.cpp:
1469         Move the PLATFORM(WIN) implementations from PlatformCALayer to PlatformCALayerWin.
1470
1471         (PlatformCALayerWin::create):
1472         (PlatformCALayer::platformCALayer):
1473
1474         * platform/graphics/ca/win/PlatformCALayerWin.h: Added.
1475
1476         * platform/graphics/win/GraphicsContext3DWin.cpp:
1477         (WebCore::GraphicsContext3D::GraphicsContext3D):
1478         * platform/graphics/win/MediaPlayerPrivateQuickTimeVisualContext.cpp:
1479         (WebCore::MediaPlayerPrivateQuickTimeVisualContext::createLayerForMovie):
1480         Include PlatformCALayerWin.h and explicitly make PlatformCALayerWins here.
1481
1482 2013-10-14  Hugo Parente Lima  <hugo.lima@openbossa.org>
1483
1484         [cmake] MediaControlsApple is used only by Efl port and is on CMakeLists.txt
1485         https://bugs.webkit.org/show_bug.cgi?id=122772
1486
1487         Reviewed by Anders Carlsson.
1488
1489         * CMakeLists.txt: Removed MediaControlsApple.cpp
1490         * PlatformEfl.cmake: Added MediaControlsApple.cpp
1491
1492 2013-10-14  Alexey Proskuryakov  <ap@apple.com>
1493
1494         window.crypto doesn't preserve custom properties
1495         https://bugs.webkit.org/show_bug.cgi?id=122770
1496
1497         Reviewed by Mark Hahnenberg.
1498
1499         Test: security/crypto-gc.html
1500
1501         Generate isReachableFromOpaqueRoots that makes Crypto live as long as the document
1502         lives (because that's when it's observable through window object).
1503
1504         * page/Crypto.cpp:
1505         (WebCore::Crypto::Crypto):
1506         (WebCore::Crypto::~Crypto):
1507         (WebCore::Crypto::document):
1508         * page/Crypto.h:
1509         (WebCore::Crypto::create):
1510         Made Crypto a ContextDestructionObserver, so that it can report its document to bindings.
1511         Removed ScriptWrappable, because it seems to have served no purpose in this class.
1512
1513         * page/Crypto.idl: Added GenerateIsReachable. Removed ImplementationLacksVTable,
1514         because the class now has a vtable, and can be checked for bindings integrity.
1515
1516         * page/DOMWindow.cpp: (WebCore::DOMWindow::crypto): Pass a document when creating
1517         crypto.
1518
1519 2013-10-14  Andreas Kling  <akling@apple.com>
1520
1521         CTTE: NamedNodeMap always has a corresponding Element.
1522         <https://webkit.org/b/122769>
1523
1524         Reviewed by Anders Carlsson.
1525
1526         Made NamedNodeMap::m_element a reference and remove an assertion
1527         that it's never null.
1528
1529 2013-10-14  Andreas Kling  <akling@apple.com>
1530
1531         REGRESSION(r157408): Crashes in RenderFullScreen::wrapRenderer().
1532
1533         Unreviewed crash fix for these two tests:
1534
1535         - fullscreen/full-screen-restrictions.html
1536         - fullscreen/empty-anonymous-block-continuation-crash.html
1537
1538         * rendering/RenderFullScreen.cpp:
1539         (RenderFullScreen::wrapRenderer):
1540
1541             Get the RenderArena from Document like we did before this patch.
1542
1543 2013-10-14  Hans Muller  <hmuller@adobe.com>
1544
1545         [CSS Shapes] Image valued shape-outside shapes should update the layout after the image has been loaded
1546         https://bugs.webkit.org/show_bug.cgi?id=122340
1547
1548         Reviewed by Simon Fraser.
1549
1550         Ensure that the an image-valued shape-outside layout is updated after the image has
1551         been loaded.
1552
1553         Test: http/tests/css/css-image-valued-shape.html
1554
1555         * rendering/RenderBlock.cpp:
1556         (WebCore::RenderBlock::imageChanged): Added code for the shape-outside case.
1557         (WebCore::RenderBlock::updateShapeInsideInfoAfterStyleChange): Ditto.
1558         * rendering/RenderElement.cpp:
1559         (WebCore::RenderElement::~RenderElement): Ditto.
1560         (WebCore::RenderElement::setStyle): Ditto.
1561
1562 2013-10-14  Andreas Kling  <akling@apple.com>
1563
1564         Remove some silly null checks in Element/NamedNodeMap.
1565         <https://webkit.org/b/122767>
1566
1567         Reviewed by Darin Adler.
1568
1569         Make shouldIgnoreAttributeCase() take a const Element&, exposing
1570         some unnecessary null checks.
1571
1572 2013-10-14  Brent Fulgham  <bfulgham@apple.com>
1573
1574         [Win] Build fix after r122737.
1575
1576         * dom/Node.h: Add explicit WebCore namespace to macro definition to work around
1577         Visual Studio bug.
1578
1579 2013-10-14  Andreas Kling  <akling@apple.com>
1580
1581         Pass Document directly to anonymous renderer constructors.
1582         <https://webkit.org/b/122752>
1583
1584         Reviewed by Antti Koivisto.
1585
1586         Added separate constructors for creating anonymous renderers that
1587         take a Document& instead of a null Element*/Text*.
1588
1589         Removed setDocumentForAnonymous() and all createAnonymous() helpers.
1590         ...and RenderObject::m_node is now a Node&, wohoo!
1591
1592 2013-10-13  Sam Weinig  <sam@webkit.org>
1593
1594         CTTE: Add more node conversion helpers
1595         https://bugs.webkit.org/show_bug.cgi?id=122737
1596
1597         Reviewed by Darin Adler.
1598
1599         - Factor NODE_TYPE_CASTS into TYPE_CASTS_BASE(ToClassName, FromClassName)
1600           to allow for DOCUMENT_TYPE_CASTS.
1601         - Replace more static_casts<>.
1602
1603 2013-10-14  Zan Dobersek  <zdobersek@igalia.com>
1604
1605         Reintroduce PassRefPtr<Event> copy in ScopedEventQueue::dispatchEvent
1606         https://bugs.webkit.org/show_bug.cgi?id=122742
1607
1608         Reviewed by Alexey Proskuryakov.
1609
1610         This is a follow-up to r157219 which introduced a workaround for the GCC's quirky behavior that
1611         was resulting in crashes due to the PassRefPtr<Event> object passed to EventDispatcher::dispatchEvent
1612         being copied and nullified first before retrieving the EventTarget of the Event object wrapped in that
1613         PassRefPtr.
1614
1615         The implementation is now adjusted to first retrieve the pointer to the Event's EventTarget and store
1616         it in a local variable. That variable is then passed as the first parameter to EventDispatcher::dispatchEvent,
1617         and the PassRefPtr<Event> passed directly as the second parameter. Previously the pointer of that PassRefPtr
1618         object was passed in, with a new PassRefPtr being created which would increase the reference count of the
1619         ref-counted object. Passing in the original PassRefPtr avoids the unnecessary reference count increase by creating
1620         a copy. That still nullifies the original PassRefPtr, but that's not a problem anymore.
1621
1622         * dom/ScopedEventQueue.cpp:
1623         (WebCore::ScopedEventQueue::dispatchEvent):
1624
1625 2013-10-14  Bear Travis  <betravis@adobe.com>
1626
1627         [CSS Shapes] Shape-Margin should be animatable
1628         https://bugs.webkit.org/show_bug.cgi?id=122524
1629
1630         Reviewed by Darin Adler.
1631
1632         Mark content for relayout after shape-margin changes, and add shape-margin
1633         to the list of animatable properties.
1634
1635         Tests: fast/shapes/shape-outside-floats/shape-outside-dynamic-shape-margin.html
1636                fast/shapes/shape-outside-floats/shape-outside-shape-margin-animation.html
1637
1638         * page/animation/CSSPropertyAnimation.cpp:
1639         (WebCore::CSSPropertyAnimationWrapperMap::CSSPropertyAnimationWrapperMap): Add
1640         shape-margin to the map of animatable CSS properties.
1641         * rendering/RenderBox.cpp:
1642         (WebCore::RenderBox::updateShapeOutsideInfoAfterStyleChange): Compare shape-margins,
1643         and mark dependent content for relayout if they have changed.
1644
1645 2013-10-14  Arvid Nilsson  <anilsson@blackberry.com>
1646
1647         Don't crash after OpenGL robustness reset
1648         https://bugs.webkit.org/show_bug.cgi?id=122750
1649
1650         Reviewed by George Staikos.
1651
1652         JIRA 517132.
1653         Just log the incident and pretend like nothing happened.
1654
1655         No new tests, we don't have repeatable steps to reproduce a robustness
1656         reset.
1657
1658         * platform/graphics/blackberry/LayerRenderer.cpp:
1659         (WebCore::LayerRenderer::makeContextCurrent):
1660
1661 2013-10-14  Krzysztof Czech  <k.czech@samsung.com>
1662
1663         [EFL] Buildfix after r157393
1664         https://bugs.webkit.org/show_bug.cgi?id=122749
1665
1666         Reviewed by Andreas Kling.
1667
1668         Buildfix with error enumeration value 'CSS_FR' not handled in switch.
1669
1670         * css/CSSCalculationValue.cpp:
1671         (WebCore::hasDoubleValue):
1672
1673 2013-10-14  Krzysztof Czech  <k.czech@samsung.com>
1674
1675         [EFL] Present replaced objects with 0xFFFC character
1676         https://bugs.webkit.org/show_bug.cgi?id=122744
1677
1678         Reviewed by Mario Sanchez Prada.
1679
1680         Replaced elements should be emitted in GTK/EFL and
1681         marked their presence with the replacement character.
1682
1683         * accessibility/AccessibilityObject.cpp:
1684         (WebCore::AccessibilityObject::textIteratorBehaviorForTextRange):
1685
1686 2013-09-16  Sergio Villar Senin  <svillar@igalia.com>
1687
1688         [CSS Grid Layout] Implement support for <flex>
1689         https://bugs.webkit.org/show_bug.cgi?id=115362
1690
1691         Reviewed by Andreas Kling.
1692
1693         From Blink r149134, r149480, r149532, r150287 and r156127 by <jchaffraix@chromium.org>
1694         From Blink r157820 by <svillar@igalia.com>
1695
1696         Added support for flexible lengths ('fr' unit) in CSS Grid Layout
1697         code. This requires the addition of GridLength class to
1698         encapsulate the knowledge of <flex> to the grid layout code.
1699
1700         Also updated the algorithm that computes the layout. It increases
1701         the value of 1fr based on the grid tracks' usedBreath to fraction
1702         ratio (called normalizedFractionValue). This enables increasing
1703         the fraction value while updating the available space to account
1704         for processed grid tracks. The algorithm stops when we run out of
1705         grid tracks or available space (one grid item has an intrinsic
1706         size too big). This matches the specs to the letter for the known
1707         available space case (both the unknown case and the interaction
1708         with 'span' are left out of this patch).
1709
1710         Tests: fast/css-grid-layout/flex-and-minmax-content-resolution-columns.html
1711                fast/css-grid-layout/flex-and-minmax-content-resolution-rows.html
1712                fast/css-grid-layout/flex-content-resolution-columns.html
1713                fast/css-grid-layout/flex-content-resolution-rows.html
1714
1715         * GNUmakefile.list.am: Added GridLength.h to the build system.
1716         * Target.pri: Ditto.
1717         * WebCore.vcxproj/WebCore.vcxproj: Ditto.
1718         * WebCore.vcxproj/WebCore.vcxproj.filters: Ditto.
1719         * WebCore.xcodeproj/project.pbxproj: Ditto.
1720         * css/CSSComputedStyleDeclaration.cpp:
1721         (WebCore::valueForGridTrackBreadth): Replace Length by GridLength.
1722         * css/CSSGrammar.y.in: Added FR support.
1723         * css/CSSParser.cpp: Ditto.
1724         (WebCore::CSSParser::parseGridBreadth):
1725         (WebCore::CSSParser::detectNumberToken):
1726         * css/CSSParserValues.cpp: Added FR support.
1727         (WebCore::CSSParserValue::createCSSValue):
1728         * css/CSSParserValues.h:
1729         (WebCore::CSSParserString::operator[]):
1730         (WebCore::CSSParserString::equalIgnoringCase):
1731         * css/CSSPrimitiveValue.cpp: Added FR support.
1732         (WebCore::isValidCSSUnitTypeForDoubleConversion):
1733         (WebCore::CSSPrimitiveValue::cleanup):
1734         (WebCore::CSSPrimitiveValue::customCSSText):
1735         (WebCore::CSSPrimitiveValue::cloneForCSSOM):
1736         (WebCore::CSSPrimitiveValue::equals):
1737         * css/CSSPrimitiveValue.h: Added a couple of missing const.
1738         (WebCore::CSSPrimitiveValue::isFlex):
1739         * css/StyleResolver.cpp: Added FR support.
1740         (WebCore::createGridTrackBreadth):
1741         (WebCore::createGridTrackSize):
1742         * rendering/RenderGrid.cpp:
1743         (WebCore::GridTrackForNormalization::GridTrackForNormalization):
1744         New helper struct to ease the computation of track breadths with
1745         flexible lengths.
1746         (WebCore::GridTrackForNormalization::operator=):
1747         (WebCore::RenderGrid::computePreferredTrackWidth): Replaced Length by GridLength.
1748         (WebCore::RenderGrid::computedUsedBreadthOfGridTracks): Grow grid lines
1749         having a fraction as the MaxTrackSizingFunction.
1750         (WebCore::RenderGrid::computeUsedBreadthOfMinLength): Replaced Length by GridLength.
1751         (WebCore::RenderGrid::computeUsedBreadthOfMaxLength): Ditto.
1752         (WebCore::sortByGridNormalizedFlexValue):
1753         (WebCore::RenderGrid::computeNormalizedFractionBreadth): Increase
1754         the fraction value while updating the available space to account
1755         for processed grid tracks.
1756         (WebCore::RenderGrid::resolveContentBasedTrackSizingFunctions):
1757         (WebCore::RenderGrid::distributeSpaceToTracks): Never shrink track sizes.
1758         (WebCore::RenderGrid::tracksAreWiderThanMinTrackBreadth):
1759         * rendering/RenderGrid.h:
1760         * rendering/style/GridLength.h: Added.
1761         (WebCore::GridLength::GridLength):
1762         (WebCore::GridLength::isLength):
1763         (WebCore::GridLength::isFlex):
1764         (WebCore::GridLength::length):
1765         (WebCore::GridLength::flex):
1766         (WebCore::GridLength::setFlex):
1767         (WebCore::GridLength::operator==):
1768         * rendering/style/GridTrackSize.h: Replaced Length by GridLength.
1769         (WebCore::GridTrackSize::length):
1770         (WebCore::GridTrackSize::setLength):
1771         (WebCore::GridTrackSize::minTrackBreadth):
1772         (WebCore::GridTrackSize::maxTrackBreadth):
1773         (WebCore::GridTrackSize::setMinMax):
1774         (WebCore::GridTrackSize::hasMinOrMaxContentMinTrackBreadth):
1775         (WebCore::GridTrackSize::hasMaxContentMinTrackBreadth):
1776         (WebCore::GridTrackSize::hasMinOrMaxContentMaxTrackBreadth):
1777         (WebCore::GridTrackSize::hasMaxContentMaxTrackBreadth):
1778
1779 2013-10-14  peavo@outlook.com  <peavo@outlook.com>
1780
1781         Broken text rendering when input field has selection.
1782         https://bugs.webkit.org/show_bug.cgi?id=122716
1783
1784         Reviewed by Antti Koivisto.
1785
1786         Tests: fast/text/text-rendering-with-input-selection.html.
1787                fast/text/text-rendering-with-input-selection-expected.html.
1788
1789         * rendering/InlineTextBox.cpp:
1790         (WebCore::InlineTextBox::paint): Check that text has selection.
1791
1792 2013-10-14  Zalan Bujtas  <zalan@apple.com>
1793
1794         Unexpected word wrapping for wrapped content then raw content.
1795         https://bugs.webkit.org/show_bug.cgi?id=121130
1796
1797         Reviewed by Antti Koivisto.
1798
1799         When deciding whether a line is considered empty, we need to check if the current
1800         object is fully responsible for the currently uncommitted width. It helps differentiating
1801         <span></span><span>abcd</span> from <span>a</span><span>bcd</span>, where in the first
1802         case when we hit the second <span> the line is still considered empty, as opposed to the
1803         second example.
1804
1805         Test: fast/css/unexpected-word-wrapping-with-non-empty-spans.html
1806
1807         * rendering/RenderBlockLineLayout.cpp:
1808         (WebCore::LineBreaker::nextSegmentBreak):
1809
1810 2013-10-14  Andreas Kling  <akling@apple.com>
1811
1812         Be more efficient about passing RenderStyle to attachRenderTree().
1813         <https://webkit.org/b/122743>
1814
1815         Reviewed by Antti Koivisto.
1816
1817         Have attachRenderTree() and createRendererTreeIfNeeded() pass the
1818         RenderStyle in a PassRefPtr to avoid churning the ref count.
1819
1820 2013-10-14  Sergio Villar Senin  <svillar@igalia.com>
1821
1822         [CSS Grid Layout] 2 span positions are not resolved correctly
1823         https://bugs.webkit.org/show_bug.cgi?id=119717
1824
1825         Reviewed by Andreas Kling.
1826
1827         From Blink r155397 by <jchaffraix@chromium.org>
1828
1829         Test: fast/css-grid-layout/grid-item-bad-resolution-double-span.html
1830
1831         Two opposite 'span' or 'auto' positions should be resolved using
1832         the auto placement algorithm. We were only checking for the 'auto'
1833         case. This also covers the case of other positions that, according
1834         to the spec, should be treated as 'auto'.
1835
1836         * rendering/RenderGrid.cpp:
1837         (WebCore::RenderGrid::resolveGridPositionsFromStyle):
1838
1839 2013-10-13  Andreas Kling  <akling@apple.com>
1840
1841         Use RenderElement instead of RenderObject in more places.
1842         <https://webkit.org/b/122734>
1843
1844         Reviewed by Antti Koivisto.
1845
1846         Convert some sites to use RenderElement (or type inference) instead
1847         of RenderObject for less branchy code.
1848
1849 2013-10-13  Darin Adler  <darin@apple.com>
1850
1851         Deprecate or remove deleteAllValues functions; there are only a few call sites left
1852         https://bugs.webkit.org/show_bug.cgi?id=122738
1853
1854         Reviewed by Anders Carlsson.
1855
1856         * platform/blackberry/CookieMap.cpp:
1857         (WebCore::CookieMap::deleteAllCookiesAndDomains):
1858         * platform/network/blackberry/rss/RSSParserBase.cpp:
1859         (WebCore::RSSFeed::clear):
1860         * platform/win/WCDataObject.cpp:
1861         (WebCore::WCDataObject::~WCDataObject):
1862         Renamed deleteAllValues to deprecatedDeleteAllValues.
1863
1864 2013-10-13  Sam Weinig  <sam@webkit.org>
1865
1866         Merge NODE_TYPE_CASTS and ELEMENT_TYPE_CASTS
1867         https://bugs.webkit.org/show_bug.cgi?id=122735
1868
1869         Reviewed by Antti Koivisto.
1870
1871         NODE_TYPE_CASTS and ELEMENT_TYPE_CASTS are identical. Let them become one
1872         with one another.
1873
1874 2013-10-13  Andreas Kling  <akling@apple.com>
1875
1876         Uncrashify Document::head() too. (Why am I even awake?)
1877
1878 2013-10-13  Andreas Kling  <akling@apple.com>
1879
1880         REGRESSION(r157381): Make Document::body() crash less when there is no documentElement.
1881
1882         Unreviewed.
1883
1884 2013-10-13  Darin Adler  <darin@apple.com>
1885
1886         Rewrite Document::body and Document::head in modern style, way clearer and shorter
1887         https://bugs.webkit.org/show_bug.cgi?id=122717
1888
1889         Reviewed by Andreas Kling.
1890
1891         * dom/Document.cpp:
1892         (WebCore::Document::body): Use iterator to make this way easier to read.
1893         (WebCore::Document::head): Ditto.
1894
1895         * html/HTMLTagNames.in: Added generateTypeHelpers for body and head.
1896
1897 == Rolled over to ChangeLog-2013-10-13 ==