1. 12 Jun, 2013 35 commits
    • commit-queue@webkit.org's avatar
      [atk] Replace deprecated call to atk_document_get_locale() in DumpRenderTree · 120f565e
      commit-queue@webkit.org authored
      https://bugs.webkit.org/show_bug.cgi?id=115647
      
      Patch by Eduardo Lima Mitev <elima@igalia.com> on 2013-06-12
      Reviewed by Martin Robinson.
      
      Source/WebCore:
      
      Override the get_object_locale() method of WebkitAccessibleWrapperAtk's internal
      AtkObject, to include custom implementations for AtkDocument and AtkText objects,
      taking the logic as-is from AtkDocument::get_document_locale() and DumpRenderTree's
      AccessibilityUIElementAtk::language(), respectively.
      
      Apart from improving encapsulation, this avoids calling deprecated get_document_locale()
      method.
      
      No new functionality, no new tests.
      
      * accessibility/atk/WebKitAccessibleInterfaceDocument.cpp:
      (webkitAccessibleDocumentInterfaceInit): Chains implementation of
      AtkDocument::get_document_locale() to AtkObject::get_object_locale().
      * accessibility/atk/WebKitAccessibleWrapperAtk.cpp:
      (webkitAccessibleGetObjectLocale): Add implementation of locale resolution for
      objects of type AtkDocument and AtkText.
      (webkitAccessibleClassInit): Override AtkObject::get_object_locale() method.
      
      Tools:
      
      Locale resolution is moved to WebKitAccessibleWrapperAtk using
      AtkObject::get_object_locale() API. Now, implementation of
      AccessibilityUIElement::language() in both DumpRenderTree and WebKitTestRunner can
      be leveraged to get_object_locale() of AtkObject.
      
      Apart from improving encapsulation, this avoids calling deprecated get_document_locale()
      method.
      
      No new functionality, no new tests.
      
      * DumpRenderTree/atk/AccessibilityUIElementAtk.cpp:
      (AccessibilityUIElement::language): Leverage locale resolution to
      AtkObject::get_object_locale().
      * WebKitTestRunner/InjectedBundle/atk/AccessibilityUIElementAtk.cpp:
      (WTR::AccessibilityUIElement::language): Leverage locale resolution to
      AtkObject::get_object_locale().
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@151524 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      120f565e
    • hugo.lima@openbossa.org's avatar
      Remove last bits of GYP from build-webkit and update-webkit scripts · 6fbadb36
      hugo.lima@openbossa.org authored
      https://bugs.webkit.org/show_bug.cgi?id=117557
      
      Reviewed by Ryosuke Niwa.
      
      * Scripts/build-webkit:
      * Scripts/update-webkit:
      
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@151523 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      6fbadb36
    • zandobersek@gmail.com's avatar
      [GTK] Move more build targets for source code that's free of layer violations into libPlatform · 60be53da
      zandobersek@gmail.com authored
      https://bugs.webkit.org/show_bug.cgi?id=115936
      
      Reviewed by Martin Robinson.
      
      Source/Platform: 
      
      * GNUmakefile.am: Add additional Source/WebCore/platform subdirectories to the list of the directories
      that are specific to the platform layer but still located in WebCore. Add Enchant and Freetype dependencies'
      CFLAGS to the list of libPlatform CPPFLAGS. Add GStreamer dependency's CFLAGS to the list of libPlatformGtk
      CPPFLAGS.
      
      Source/WebCore: 
      
      No new tests - no new functionality.
      
      * GNUmakefile.list.am: Move more build targets under the platform_sources and platformgtk_sources listing.
      These build targets represent source files that are already free of platform layer violations and are as such
      ready to be pushed down from the WebCore layer into the Platform layer.
      
      Source/WebKit2: 
      
      * GNUmakefile.am: Add GStreamer dependency's CFLAGS to the list of libPlatformGtk2 CPPFLAGS.
      
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@151522 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      60be53da
    • hugo.lima@openbossa.org's avatar
      Remove forgotten unused Tools/Script/make-gypi script · 3bf132b7
      hugo.lima@openbossa.org authored
      https://bugs.webkit.org/show_bug.cgi?id=117558
      
      Reviewed by Ryosuke Niwa.
      
      This script isn't used since the GYP removal from WebKit.
      
      * Scripts/make-gypi: Removed.
      
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@151521 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      3bf132b7
    • commit-queue@webkit.org's avatar
      "Open Video in New Window" doesn't make sense in fullscreen · 05cdf156
      commit-queue@webkit.org authored
      https://bugs.webkit.org/show_bug.cgi?id=117556
      
      Patch by Ruth Fong <ruth_fong@apple.com> on 2013-06-12
      Reviewed by Beth Dakin.
      
      Source/WebCore:
      
      No new tests needed.
      
      * WebCore.exp.in: Added WebKit2 hook for mediaIsInFullscreen() method.
      
      Source/WebKit2:
      
      Adds method mediaIsInFullscreen() that checks if a media element is in
      fullscreen mode, which is needed to decide when to display the
      "Open Video in New Window" context menu item.
      
      * WebProcess/InjectedBundle/API/c/WKBundleHitTestResult.cpp:
      (WKBundleHitTestResultMediaIsInFullscreen): Added.
      Hooks into new InjectedBundleHitTestResult::mediaIsInFullscreen() method.
      * WebProcess/InjectedBundle/API/c/WKBundleHitTestResult.h:
      * WebProcess/InjectedBundle/InjectedBundleHitTestResult.cpp:
      (WebKit::InjectedBundleHitTestResult::mediaIsInFullscreen): Added.
      Checks if media element is in fullscreen.
      * WebProcess/InjectedBundle/InjectedBundleHitTestResult.h:
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@151520 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      05cdf156
    • rniwa@webkit.org's avatar
      Accessibility code assumes an area element's parent is a map element · f902007d
      rniwa@webkit.org authored
      https://bugs.webkit.org/show_bug.cgi?id=117496
      
      Reviewed by Chris Fleizach.
      
      Source/WebCore: 
      
      We can't make such an assumption. Scripts can insert any element between area and map elements.
      
      Merge https://chromium.googlesource.com/chromium/blink/+/b6f486284f08c52904701c93e1ec0b7d6e76af9f.
      
      Test: accessibility/image-map-with-indirect-area-crash.html
      
      * accessibility/AccessibilityRenderObject.cpp:
      (WebCore::AccessibilityRenderObject::accessibilityImageMapHitTest):
      
      LayoutTests: 
      
      Add a regression test from https://chromium.googlesource.com/chromium/blink/+/b6f486284f08c52904701c93e1ec0b7d6e76af9f.
      
      * accessibility/image-map-with-indirect-area-crash-expected.txt: Added.
      * accessibility/image-map-with-indirect-area-crash.html: Added.
      
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@151519 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      f902007d
    • robert@webkit.org's avatar
      Source/WebCore: Whitespace between inlines with nowrap and a shrink-to-fit... · c522345e
      robert@webkit.org authored
      Source/WebCore: Whitespace between inlines with nowrap and a shrink-to-fit parent gets a line-break when it shouldn't
      https://bugs.webkit.org/show_bug.cgi?id=117370
      
      Reviewed by David Hyatt.
      
      Test: fast/text/whitespace/inline-whitespace-wrapping-8.html
      
      A no-wrap inline shouldn't include trailing space when deciding whether it fits on a line.
      Likewise when we finish iterating through the objects on a line we should clear our linebreak
      if the only thing that prevents us fitting on the line is our collapsed trailing whitespace.
      
      Removing the trailing space from this measurement means we need to watch out for potential
      breaks between no-wrap inlines, in particular if we leave a no-wrap inline ignoring spaces
      and enter an autowrap inline then we need to mark the beginning of the autowrap inline
      as a potential linebreak. The test fast/text/whitespace/inline-whitespace-wrapping-5.html
      is an example of such a case.
      
      * rendering/RenderBlockLineLayout.cpp:
      (WebCore::LineWidth::LineWidth):
      (WebCore::LineWidth::fitsOnLine):
      (WebCore::LineWidth::fitsOnLineExcludingTrailingWhitespace):
      (WebCore::LineWidth::fitsOnLineExcludingTrailingCollapsedWhitespace):
      (WebCore::LineWidth::setTrailingWhitespaceWidth):
      (WebCore::RenderBlock::LineBreaker::nextSegmentBreak):
      
      LayoutTests: Whitespace between inlines with nowrap and a shrink-to-fit parent line-break incorrectly
      https://bugs.webkit.org/show_bug.cgi?id=117370
      
      Reviewed by David Hyatt.
      
      * fast/text/whitespace/inline-whitespace-wrapping-8-expected.html: Added.
      * fast/text/whitespace/inline-whitespace-wrapping-8.html: Added.
      
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@151518 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      c522345e
    • commit-queue@webkit.org's avatar
      [CSS Shapes] rectangle and inset-rectangle do not properly handle rx and ry · ebbe06f8
      commit-queue@webkit.org authored
      https://bugs.webkit.org/show_bug.cgi?id=116745
      
      Patch by Bem Jones-Bey <bjonesbe@adobe.com> on 2013-06-12
      Reviewed by Dirk Schulze.
      
      Source/WebCore:
      
      If ry is not supplied, it now defaults to the rx value. Also, if rx
      and ry are not supplied, they default to 0px. This also has the effect
      that the computed style for any rectangle or inset-rectangle now
      contains all six parameters.
      
      If rx > width/2 then it is clamped to width/2, and if ry > height/2,
      then it is clamped to height/2. This happens at layout time because
      given mixed units and relative units, that is the only time this
      determination can be made.
      
      Tests: fast/exclusions/shape-inside/shape-inside-rounded-rectangle-large-radius.html
             fast/exclusions/shape-outside-floats/shape-outside-floats-rounded-rectangle-large-radius.html
      
      * css/BasicShapeFunctions.cpp:
      (WebCore::valueForBasicShape): Remove checks for undefined, since rx
          and ry cannot be undefined in BasicShapes anymore.
      (WebCore::basicShapeForValue): If radii are undefined in CSS, set
          default values in the BasicShape, per the spec.
      * rendering/ExclusionShape.cpp:
      (WebCore::ensureRadiiDoNotOverlap): Utility method to ensure radii
          don't overlap.
      (WebCore::ExclusionShape::createExclusionShape): Remove defaulting to
          zero, clamp radii per the spec.
      * rendering/style/BasicShapes.cpp:
      (WebCore::BasicShapeRectangle::path): Remove defaulting.
      (WebCore::BasicShapeRectangle::blend): Ditto.
      (WebCore::BasicShapeInsetRectangle::path): Ditto.
      (WebCore::BasicShapeInsetRectangle::blend): Ditto.
      * rendering/style/BasicShapes.h:
      (WebCore::BasicShapeRectangle::BasicShapeRectangle): Remove defaulting
          to undefined for radii.
      (WebCore::BasicShapeRectangle::setCornerRadiusX): Add assert to
          prevent undefined.
             (WebCore::BasicShapeRectangle::setCornerRadiusY): Add assert to
          prevent undefined.
      (WebCore::BasicShapeInsetRectangle::BasicShapeInsetRectangle): Ditto.
      (WebCore::BasicShapeInsetRectangle::setCornerRadiusX): Add assert to
          prevent undefined.
             (WebCore::BasicShapeInsetRectangle::setCornerRadiusY): Add assert to
          prevent undefined.
      
      LayoutTests:
      
      Update tests to check for the rx and ry conditions.
      
      * fast/exclusions/parsing/parsing-shape-inside-expected.txt: New results.
      * fast/exclusions/parsing/parsing-shape-outside-expected.txt: New results.
      * fast/exclusions/parsing/script-tests/parsing-shape-inside.js: Update
          to reflect that the computed value now always has rx and ry.
      * fast/exclusions/parsing/script-tests/parsing-shape-outside.js: Ditto.
      * fast/exclusions/parsing/script-tests/parsing-test-utils.js: Ditto.
      * fast/exclusions/resources/rounded-rectangle.js:
      (genLeftRightRoundedRectFloatShapeOutsideRefTest): Factor out left and
          right into separate methods so that they can be used in tests that
          only have left floats or only right floats.
      (genLeftRoundedRectFloatShapeOutsideRefTest): Ditto.
      (genRightRoundedRectFloatShapeOutsideRefTest): Ditto.
      * fast/exclusions/shape-inside/shape-inside-animation-expected.txt: New results.
      * fast/exclusions/shape-inside/shape-inside-animation.html: Change to
          reflect new computed style value.
      * fast/exclusions/shape-inside/shape-inside-rounded-rectangle-large-radius-expected.html: Added.
      * fast/exclusions/shape-inside/shape-inside-rounded-rectangle-large-radius.html: Added.
      * fast/exclusions/shape-outside-floats/shape-outside-floats-rounded-rectangle-large-radius-expected.html: Added.
      * fast/exclusions/shape-outside-floats/shape-outside-floats-rounded-rectangle-large-radius.html: Added.
          Note that the expected test here uses shape-inside, because we
          really are just testing that the behavior is the same as a
          shape-inside with rx and ry set to 50%, and it turned out to be
          very hard to construct an expected result that didn't use
          shape-inside and wasn't off by a pixel.
      * css3/masking/clip-path-animation.html: Update expected value.
      * css3/masking/clip-path-animation-expected.txt: New results.
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@151517 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      ebbe06f8
    • bfulgham@apple.com's avatar
      [Windows] Support Title UI Element Accessibility Attribute · 1b60f890
      bfulgham@apple.com authored
      https://bugs.webkit.org/show_bug.cgi?id=117530
      
      Reviewed by Anders Carlsson.
      
      Source/WebCore: 
      
      Fixes: accessibility/list-item-role.html
             accessibility/hidden-legend.html
      
      * accessibility/win/AccessibilityObjectWrapperWin.cpp:
      (WebCore::AccessibilityObjectWrapper::accessibilityAttributeValue): Extend to support access
      for the Title UI Element object.
      * accessibility/win/AccessibilityObjectWrapperWin.h: Revise interface to use Variant output,
      which is more in keeping with the Mac's 'id' method signature.
      
      Source/WebKit/win: 
      
      * AccessibleBase.cpp:
      (AccessibleBase::attributeValue): Extend to handle Title UI Element.
      * AccessibleBase.h: Revise signature to use a variant argument.
      * Interfaces/AccessibleComparable.idl:  Revise interface to use variant
      
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@151516 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      1b60f890
    • timothy@apple.com's avatar
      Add new Web Inspector images for a blog post. · c2134315
      timothy@apple.com authored
      * blog-files/new-inspector/activity-viewer.png: Added.
      * blog-files/new-inspector/codemirror.png: Added.
      * blog-files/new-inspector/console.png: Added.
      * blog-files/new-inspector/hero-labeled.png: Added.
      * blog-files/new-inspector/hero.png: Added.
      * blog-files/new-inspector/layer-borders.png: Added.
      * blog-files/new-inspector/layers.png: Added.
      * blog-files/new-inspector/navigation-bar.png: Added.
      * blog-files/new-inspector/open-sign.png: Added.
      * blog-files/new-inspector/reasons.png: Added.
      * blog-files/new-inspector/save.png: Added.
      * blog-files/new-inspector/source-map.png: Added.
      * blog-files/new-inspector/styles.png: Added.
      * blog-files/new-inspector/toolbar-context-menu.png: Added.
      * blog-files/new-inspector/toolbar-docked.png: Added.
      * blog-files/new-inspector/toolbar.png: Added.
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@151515 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      c2134315
    • leoyang@rim.com's avatar
      [BlackBerry] Use RefPtr for HTMLInputElement inside CredentialTransformData · f7286af0
      leoyang@rim.com authored
      https://bugs.webkit.org/show_bug.cgi?id=117553
      
      Reviewed by Carlos Garcia Campos.
      
      In CredentialTransformData we should use RefPtr for
      m_userNameElement, m_passwordElement and m_oldPasswordElement
      because otherwise the elements could go away when the form
      elements get destroyed.
      
      Also add *const* for the parameter of CredentialTransformData::findPasswordFormFields().
      
      No functionalities changed no new tests.
      
      * WebCoreSupport/CredentialTransformData.cpp:
      (WebCore::CredentialTransformData::findPasswordFormFields):
      (WebCore::CredentialTransformData::locateSpecificPasswords):
      * WebCoreSupport/CredentialTransformData.h:
      
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@151514 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      f7286af0
    • commit-queue@webkit.org's avatar
      [CSS Shaders] Animations and transitions should use validated custom programs · 028b821a
      commit-queue@webkit.org authored
      https://bugs.webkit.org/show_bug.cgi?id=117550
      
      Run filters used for animations and transitions through RenderLayer::computeFilterOperations
      in order to validate custom filter operations.
      
      Patch by Ralph Thomas <ralpht@gmail.com> on 2013-06-12
      Reviewed by Alexandru Chiculita.
      
      No new tests as no platform can run custom filters with accelerated animations yet.
      
      * rendering/RenderLayerBacking.cpp:
      (WebCore::RenderLayerBacking::startAnimation):
      (WebCore::RenderLayerBacking::startTransition):
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@151513 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      028b821a
    • commit-queue@webkit.org's avatar
      Allow for toggling fullscreen on <video> elements · 0f4ee83b
      commit-queue@webkit.org authored
      https://bugs.webkit.org/show_bug.cgi?id=117220
      
      Patch by Ruth Fong <ruth_fong@apple.com> on 2013-06-12
      Reviewed by Dean Jackson.
      
      Source/WebCore:
      
      This patch adds the ability for fullscreen
      context menu item on <video> elements to switch between "Enter Fullscreen"
      and "Exit Fullscreen" and behave appropriately.
      
      No new tests. media/context-menu-action.html,
      which has been disabled by bug 116651, is used to test context menus.
      
      * English.lproj/Localizable.strings: Add "Exit Fullscreen" string.
      * html/HTMLMediaElement.cpp:
      (WebCore::HTMLMediaElement::toggleFullscreenState): Added to appropriately enter/exit fullscreen.
      * html/HTMLMediaElement.h:
      * page/ContextMenuController.cpp:
      * platform/ContextMenuItem.h:
      * platform/LocalizedStrings.cpp:
      * platform/LocalizedStrings.h:
      Updated to rename variables more appropriately to reflect the toggle-ability of video fullscreen.
      * rendering/HitTestResult.cpp:
      (WebCore::HitTestResult::mediaIsInFullscreen): Added to check if an element
      was a media element in fullscreen.
      (WebCore::HitTestResult::toggleMediaFullscreenState): Added to hook into
      HTMLMediaElement::toggleFullscreenState.
      * rendering/HitTestResult.h:
      
      Source/WebKit2:
      
      * Shared/API/c/WKContextMenuItemTypes.h:
      * Shared/API/c/WKSharedAPICast.h:
      Added variables to support the toggle-ability of the fullscreen
      video context menu item.
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@151512 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      0f4ee83b
    • ossy@webkit.org's avatar
      [ARM] Unreviewed gardening, unskip now passing tests and try to paint the bot green. · 24c0bd0a
      ossy@webkit.org authored
      * platform/qt-arm/TestExpectations:
      * platform/qt-arm/svg/W3C-SVG-1.1/paths-data-05-t-expected.txt: Removed.
      * platform/qt-arm/svg/W3C-SVG-1.1/paths-data-09-t-expected.txt: Removed.
      * platform/qt-arm/svg/batik/text/textOnPath3-expected.txt: Removed.
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@151511 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      24c0bd0a
    • sergio@webkit.org's avatar
      Skipping {}, () and [] blocks while error recovering in CSS · 8828148a
      sergio@webkit.org authored
      https://bugs.webkit.org/show_bug.cgi?id=116071
      
      Reviewed by Darin Adler.
      
      From Blink r150201 and r150755 by <serya@chromium.org>
      
      Source/WebCore:
      
      Test: fast/css/parsing-expr-error-recovery.html
      
      The CSS parser should properly recover from invalid {}, () and []
      blocks skipping them instead of discarding the whole declaration
      as invalid. This merge is actually made of two different changes
      from Blink, the original one that fixes the bug and another one
      which refactors a bit the code making it more legible.
      
      * css/CSSGrammar.y.in:
      
      LayoutTests:
      
      * fast/css/parsing-expr-error-recovery-expected.txt: Added.
      * fast/css/parsing-expr-error-recovery.html: Added.
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@151510 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      8828148a
    • commit-queue@webkit.org's avatar
      JSExport header documentation substitutes "semicolon" for "colon" · 4ea719da
      commit-queue@webkit.org authored
      https://bugs.webkit.org/show_bug.cgi?id=117552
      
      Patch by Conrad Shultz <conrad_shultz@apple.com> on 2013-06-12
      Reviewed by Mark Hahnenberg.
      
      * API/JSExport.h:
      Fix a couple typos.
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@151509 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      4ea719da
    • carlosgc@webkit.org's avatar
      [GTK] Invalidate the ProcessLauncher when the process is terminated before it... · d55b9e9f
      carlosgc@webkit.org authored
      [GTK] Invalidate the ProcessLauncher when the process is terminated before it has finished launching
      https://bugs.webkit.org/show_bug.cgi?id=117412
      
      Reviewed by Anders Carlsson.
      
      * UIProcess/Launcher/gtk/ProcessLauncherGtk.cpp:
      (WebKit::ProcessLauncher::terminateProcess): If process is still
      launching just invalidate the launcher. Reset the process
      identifier after killing the process.
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@151508 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      d55b9e9f
    • commit-queue@webkit.org's avatar
      [BlackBerry] Remove dead WebDOM code · 46a1370f
      commit-queue@webkit.org authored
      https://bugs.webkit.org/show_bug.cgi?id=113370
      
      Patch by Alberto Garcia <agarcia@igalia.com> on 2013-06-12
      Reviewed by Anders Carlsson.
      
      BlackBerry PR 347565
      Internally reviewed by Charles Wei.
      
      .:
      
      * Source/cmake/OptionsBlackBerry.cmake:
      
      Source/WebCore:
      
      * PlatformBlackBerry.cmake:
      
      Source/WebKit/blackberry:
      
      * Api/WebPage.cpp:
      * Api/WebPage.h:
      * WebCoreSupport/AboutDataEnableFeatures.in:
      * WebCoreSupport/FrameLoaderClientBlackBerry.cpp:
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@151507 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      46a1370f
    • sergio@webkit.org's avatar
      [css exclusions] fast/css/variables/var-inside-shape.html is failing · 7e502adf
      sergio@webkit.org authored
      https://bugs.webkit.org/show_bug.cgi?id=116059
      
      Reviewed by Alexandru Chiculita.
      
      Some shapes were incorrectly defined, there was a "rectange"
      instead of a "rectangle" and also the number of parameters and
      their types was wrong. Apart from that, both the CSS exclusions
      and CSS variables settings were not enabled at runtime.
      
      Some extra text was also added to the divs in order to verify that
      the CSS exclusions are working properly.
      
      * fast/css/variables/var-inside-shape-expected.html:
      * fast/css/variables/var-inside-shape.html:
      * platform/efl/TestExpectations: Removed var-inside-shape.html.
      * platform/gtk/TestExpectations: Ditto.
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@151506 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      7e502adf
    • dbates@webkit.org's avatar
      window.find() case-insensitive search doesn't match diacritical marks · 5de2972f
      dbates@webkit.org authored
      https://bugs.webkit.org/show_bug.cgi?id=117353
      <rdar://problem/8535227>
      
      Patch by Daniel Bates <dabates@apple.com> on 2013-06-12
      Reviewed by Darin Adler.
      
      Source/WebCore:
      
      Currently, a case-insensitive window.find() search will return true for a word w
      (e.g. café) with diacritical characters when w appears on the page without
      diacritical marks (e.g. cafe). This leads to bad results, especially when
      searching for a non-English word (e.g. qué) where the presence of an accented
      character(s) may change the meaning of the word. Instead window.find() should
      return false when case-insensitively matching a word with diacritical marks to
      the same word without diacritical marks.
      
      Test: fast/text/find-diacritical-mark.html
      
      * editing/TextIterator.cpp:
      (WebCore::SearchBuffer::SearchBuffer):
      
      LayoutTests:
      
      Added test to use window.find() to find a search string in text, where either
      the search string or text (or both) contain diacritical characters.
      
      Additionally, updated tests LayoutTests/fast/text/find-soft-hyphen.html
      and LayoutTests/editing/execCommand/findString-diacriticals.html based on the
      new behavior of window.find(). Moved function testFind() from LayoutTests/fast/text/script-tests/find-backwards.js
      to its own file, LayoutTests/fast/text/resources/test-find.js, so that this function can
      be used by both tests LayoutTests/fast/text/find-{backwards, diacritical-mark}.html.
      
      * LayoutTests/editing/execCommand/findString-diacriticals.html: Update test based on the new
      behavior of window.find().
      * fast/text/find-backwards.html: Include external JavaScript file test-find.js.
      * fast/text/find-case-folding-expected.txt: Rebase result; expected failure.
      * fast/text/find-case-folding.html: Add comment to see bug #117548 with regards to expected failure.
      * fast/text/find-diacritical-mark-expected.txt: Added.
      * fast/text/find-diacritical-mark.html: Copied from LayoutTests/fast/text/find-backwards.html.
      * fast/text/find-soft-hyphen-expected.txt: Rebase result.
      * fast/text/resources/test-find.js: Added.
      (testFind):
      * fast/text/script-tests/find-backwards.js: Moved function testFind() from here to
      file fast/text/resources/test-find.js and modified it to take an additional argument to
      toggle search case sensitivity.
      * fast/text/script-tests/find-diacritical-mark.js: Added.
      * fast/text/script-tests/find-soft-hyphen.js: Update test based on the new behavior of window.find().
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@151505 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      5de2972f
    • commit-queue@webkit.org's avatar
      [BlackBerry] Smarter algorithm to determine the backingstore rect · 2b9d59d9
      commit-queue@webkit.org authored
      https://bugs.webkit.org/show_bug.cgi?id=117451
      
      JIRA115644
      https://jira.bbqnx.net/browse/BRWSR-7028
      
      Patch by Jakob Petsovits <jpetsovits@blackberry.com> on 2013-06-12
      Reviewed by Rob Buis.
      
      So far, the backingstore tile geometry allocation was
      pretty straightforward: We would start off from the
      current viewport and append all available tiles into
      the current scrolling direction from there.
      
      This will usually work well enough, but has the downside
      of discarding all the tiles in the opposite direction.
      Also, tiles very close to the viewport will often get
      discarded even if the user only scrolls very slowly.
      
      This patch completely revamps the algorithm for
      determining where the backingstore should be positioned.
      
      The general idea is that we construct a "desired rect"
      based on the viewport and inflate it into all four
      directions according to the current scroll momentum.
      This rectangle will be similarly large as a backingstore
      tile geometry rectangle might be, by using the
      approximate number of pixels that are available in the
      given number of tiles.
      
      The proportions for extending the rectangle from the
      viewport are influenced by different factors, including
      scroll momentum, viewport ratio, available space in the
      overall contents rectangle, and natural bias for the
      "down" direction.
      
      In practice, this results in a backingstore that is
      roughly evenly distributed around the viewport when no
      movement is happening, and will gradually narrow down
      and extend into the scroll direction at a higher momentum.
      
      The final tile geometry is constructed by trying fit
      the tiles into the desired rect in a way that maximizes
      the area of its intersection. There are a few parameters
      that can be tweaked, the ones in this patch seem to
      handle most cases well enough to minimize checkerboarding.
      
      As an additional bonus, a rectangle-based tiling strategy
      can more easily be adopted for accelerated compositing,
      which currently operates on a simpler algorithm that also
      inflates the viewport but does not take scrolling into
      account.
      
      * Api/BackingStore.cpp:
      (BlackBerry::WebKit::BackingStorePrivate::BackingStorePrivate):
      (BlackBerry::WebKit::BackingStorePrivate::expandedContentsSize):
      (WebKit):
      (BlackBerry::WebKit::BackingStorePrivate::nonOverscrolled):
      (BlackBerry::WebKit::BackingStorePrivate::enclosingTileRect):
      (BlackBerry::WebKit::BackingStorePrivate::desiredBackingStoreRect):
      (BlackBerry::WebKit::BackingStorePrivate::mergeDesiredBackingStoreRect):
      (BlackBerry::WebKit::BackingStorePrivate::largestTileRectForDesiredRect):
      (BlackBerry::WebKit::BackingStorePrivate::scrollBackingStore):
      (BlackBerry::WebKit::BackingStorePrivate::createSurfaces):
      * Api/BackingStore_p.h:
      (BackingStorePrivate):
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@151503 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      2b9d59d9
    • commit-queue@webkit.org's avatar
      [Qt][CSS Regions] fast/regions/overflow-size-change-with-stacking-context.html fails · 7301ec06
      commit-queue@webkit.org authored
      https://bugs.webkit.org/show_bug.cgi?id=97199
      
      Unreviewed, Qt Gardening.
      
      Unskip test that is now passing on the bots.
      
      Patch by Javier Fernandez <jfernandez@igalia.com> on 2013-06-12
      
      * platform/qt/TestExpectations:
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@151502 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      7301ec06
    • ch.dumez@sisa.samsung.com's avatar
      Support latest Web IDL indexed property getters · 922b14a4
      ch.dumez@sisa.samsung.com authored
      https://bugs.webkit.org/show_bug.cgi?id=117507
      
      Reviewed by Kentaro Hara.
      
      Support latest Web IDL indexed property getters:
      http://www.w3.org/TR/WebIDL/#dfn-support-indexed-properties
      
      Drop support for outdated [IndexedGetter] and [NumericIndexedGetter]
      IDL extended attributes. Since the new getter format specifies the
      return type, the bindings generator is now smart enough to know if
      the indexed getter is numeric or not on its own.
      
      Based on Blink r149564 by <kojih@chromium.org>.
      
      No new tests, no behavior change.
      
      * Modules/filesystem/EntryArray.idl:
      * Modules/filesystem/EntryArraySync.idl:
      * Modules/gamepad/GamepadList.idl:
      * Modules/mediasource/SourceBufferList.idl:
      * Modules/speech/SpeechGrammarList.idl:
      * Modules/speech/SpeechRecognitionResult.idl:
      * Modules/speech/SpeechRecognitionResultList.idl:
      * bindings/scripts/CodeGenerator.pm:
      (IsNumericType):
      * bindings/scripts/CodeGeneratorJS.pm:
      (GenerateGetOwnPropertySlotBody):
      (GenerateGetOwnPropertyDescriptorBody):
      (GetIndexedGetterFunction):
      (GenerateHeader):
      (GenerateImplementation):
      * bindings/scripts/IDLAttributes.txt:
      * bindings/scripts/test/TestEventTarget.idl:
      * bindings/scripts/test/TestTypedArray.idl:
      * css/CSSRuleList.idl:
      * css/CSSStyleDeclaration.idl:
      * css/CSSValueList.idl:
      * css/MediaList.idl:
      * css/StyleSheetList.idl:
      * css/WebKitCSSFilterValue.idl:
      * css/WebKitCSSKeyframesRule.idl:
      * css/WebKitCSSTransformValue.idl:
      * dom/ClientRectList.idl:
      * dom/DOMNamedFlowCollection.idl:
      * dom/DOMStringList.idl:
      * dom/DataTransferItemList.idl:
      * dom/NamedNodeMap.idl:
      * dom/NodeList.idl:
      * dom/PropertyNodeList.idl:
      * dom/TouchList.idl:
      * fileapi/FileList.idl:
      * html/DOMSettableTokenList.idl:
      * html/DOMTokenList.idl:
      * html/HTMLAllCollection.idl:
      * html/HTMLCollection.idl:
      * html/HTMLFormControlsCollection.idl:
      * html/HTMLFormElement.idl:
      * html/HTMLPropertiesCollection.idl:
      * html/HTMLSelectElement.idl:
      * html/RadioNodeList.idl:
      * html/canvas/Float32Array.idl:
      * html/canvas/Float64Array.idl:
      * html/canvas/Int16Array.idl:
      * html/canvas/Int32Array.idl:
      * html/canvas/Int8Array.idl:
      * html/canvas/Uint16Array.idl:
      * html/canvas/Uint32Array.idl:
      * html/canvas/Uint8Array.idl:
      * html/canvas/Uint8ClampedArray.idl:
      * html/track/AudioTrackList.idl:
      * html/track/TextTrackCueList.idl:
      * html/track/TextTrackList.idl:
      * html/track/TextTrackRegionList.idl:
      * html/track/VideoTrackList.idl:
      * page/PerformanceEntryList.idl:
      * page/SpeechInputResultList.idl:
      * plugins/DOMMimeTypeArray.idl:
      * plugins/DOMPlugin.idl:
      * plugins/DOMPluginArray.idl:
      
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@151499 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      922b14a4
    • commit-queue@webkit.org's avatar
      [Coordinated Graphics] Remove unused headers in CoordinatedLayerTreeHostProxy · 78e9efd9
      commit-queue@webkit.org authored
      https://bugs.webkit.org/show_bug.cgi?id=117539
      
      Patch by Jae Hyun Park <jae.park@company100.net> on 2013-06-12
      Reviewed by Noam Rosenthal.
      
      * UIProcess/CoordinatedGraphics/CoordinatedLayerTreeHostProxy.cpp:
      * UIProcess/CoordinatedGraphics/CoordinatedLayerTreeHostProxy.h:
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@151498 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      78e9efd9
    • kseo@webkit.org's avatar
      Unreviewed gardening. Unskipping passing tests. · 94d450d4
      kseo@webkit.org authored
      * platform/gtk-wk2/TestExpectations:
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@151497 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      94d450d4
    • commit-queue@webkit.org's avatar
      [GTK] MiniBrowser to automatically download "non-showable" documents when left click in link · 89c27730
      commit-queue@webkit.org authored
      https://bugs.webkit.org/show_bug.cgi?id=117533
      
      Patch by Andres Gomez <agomez@igalia.com> on 2013-06-12
      Reviewed by Carlos Garcia Campos.
      
      MiniBrowser now starts automatically the download of the documents
      that cannot be shown just by WebKit itself on link left clicking.
      
      * MiniBrowser/gtk/BrowserWindow.c:
      (webViewDecidePolicy): handled the
      WEBKIT_POLICY_DECISION_TYPE_RESPONSE decision type too.
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@151496 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      89c27730
    • ch.dumez@sisa.samsung.com's avatar
      Rename [CallWith] to [ConstructorCallWith] for IDL interfaces · 55afbdfe
      ch.dumez@sisa.samsung.com authored
      https://bugs.webkit.org/show_bug.cgi?id=117531
      
      Reviewed by Kentaro Hara.
      
      Rename [CallWith] to [ConstructorCallWith] for IDL interfaces as the
      meaning is slightly different when [CallWith] is used for IDL operations
      or attributes. This improves IDL readability a bit and makes it more
      directly understandable.
      
      Also removed [ConstructorCallWith] statements from interfaces with custom
      constructors or no constructor, as this does not have any effect.
      
      No new tests, no behavior change.
      
      * Modules/filesystem/FileWriter.idl:
      * Modules/mediasource/MediaSource.idl:
      * Modules/mediasource/SourceBufferList.idl:
      * Modules/mediastream/MediaStream.idl:
      * Modules/mediastream/RTCPeerConnection.idl:
      * Modules/notifications/Notification.idl:
      * Modules/speech/SpeechRecognition.idl:
      * Modules/speech/SpeechSynthesisUtterance.idl:
      * Modules/webaudio/OfflineAudioContext.idl:
      * Modules/websockets/WebSocket.idl:
      * bindings/scripts/CodeGeneratorJS.pm:
      (GenerateConstructorDefinition):
      * bindings/scripts/IDLAttributes.txt:
      * bindings/scripts/test/TestInterface.idl:
      * dom/MessageChannel.idl:
      * fileapi/FileReader.idl:
      * html/MediaController.idl:
      * html/track/TextTrackCue.idl:
      * page/EventSource.idl:
      * workers/SharedWorker.idl:
      * workers/Worker.idl:
      * xml/XMLHttpRequest.idl:
      
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@151495 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      55afbdfe
    • allan.jensen@digia.com's avatar
      [Qt] Set windows key-code for multimedia keys · 35b22e7e
      allan.jensen@digia.com authored
      https://bugs.webkit.org/show_bug.cgi?id=117535
      
      Reviewed by Jocelyn Turcotte.
      
      Map Qt Key-events to their defined windows keycode values.
      
      * platform/qt/PlatformKeyboardEventQt.cpp:
      (WebCore::windowsKeyCodeForKeyEvent):
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@151494 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      35b22e7e
    • sergio@webkit.org's avatar
      [Soup] Use ResourceLoadPriority · 233a34f5
      sergio@webkit.org authored
      https://bugs.webkit.org/show_bug.cgi?id=112902
      
      Reviewed by Carlos Garcia Campos.
      
      Use the ResourceLoadPriority from the ResourceRequest to set the
      correnpondent priority to each SoupMessage.
      
      No new tests required as there is no change in functionality,
      we're just enabling a new feature in the underlying network stack.
      
      * platform/network/soup/ResourceHandleSoup.cpp:
      (WebCore::createSoupMessageForHandleAndRequest):
      * platform/network/soup/ResourceRequest.h:
      (WebCore):
      (WebCore::toSoupMessagePriority): Translates ResourceLoadPriority
      to SoupMessagePriority.
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@151493 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      233a34f5
    • zandobersek@gmail.com's avatar
      [GTK] Remove the WebCoreLayer(Gtk2).a archives before regenerating them · 1978f085
      zandobersek@gmail.com authored
      https://bugs.webkit.org/show_bug.cgi?id=116723
      
      Reviewed by Carlos Garcia Campos.
      
      Source/WebKit2: 
      
      * GNUmakefile.am: When generating the WebCoreLayerGtk2.a archive from all the dependency archives, remove the
      current archive (if any). Updating an existent archive in an incremental build is not bulletproof and can lead
      to corrupt archives and subsequent build failures. On the other hand, the archive generation is not expensive
      and produces a working archive without problems even on incremental builds.
      
      Tools: 
      
      * TestWebKitAPI/GNUmakefile.am: When generating the WebCoreLayer.a archive from all the dependency archives,
      remove the current archive (if any). Updating an existent archive in an incremental build is not bulletproof
      and can lead to corrupt archives and subsequent build failures. On the other hand, the archive generation is
      not expensive and produces a working archive without problems even on incremental builds.
      
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@151492 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      1978f085
    • zandobersek@gmail.com's avatar
      Remove memoryInfoEnabled, quantizedMemoryInfoEnabled settings · ee606799
      zandobersek@gmail.com authored
      https://bugs.webkit.org/show_bug.cgi?id=117512
      
      Reviewed by Darin Adler.
      
      Source/WebCore: 
      
      * page/Settings.in: Remove the memoryInfoEnabled and the quantizedMemoryInfoEnabled settings. These are not used
      anywhere anymore after r151199 which simplified and moved the MemoryInfo interface under window.internals.
      
      Source/WebKit/blackberry: 
      
      * Api/WebPage.cpp:
      (BlackBerry::WebKit::WebPagePrivate::init): Remove the call to Settings::setMemoryInfoEnabled, the setting is being removed.
      
      Source/WebKit/mac: 
      
      * WebKit.order: Remove the memoryInfoEnabled entry.
      * WebView/WebPreferenceKeysPrivate.h: Remove WebKitMemoryInfoEnabledPreferenceKey.
      * WebView/WebPreferences.mm: Remove memoryInfoEnabled, setMemoryInfoEnabled methods.
      (+[WebPreferences initialize]): Don't set the default value for the WebKitMemoryInfoEnabledPreferenceKey, the preference is being removed.
      * WebView/WebPreferencesPrivate.h: Remove memoryInfoEnabled, setMemoryInfoEnabled method declarations.
      * WebView/WebView.mm:
      (-[WebView _preferencesChanged:]): Remove the call to Settings::setMemoryInfoEnabled, the setting is being removed.
      
      Source/WebKit/win: 
      
      * Interfaces/IWebPreferencesPrivate.idl: Remove memoryInfoEnabled, setMemoryInfoEnabled entries.
      * WebPreferenceKeysPrivate.h: Remove the WebKitMemoryInfoEnabledPreferenceKey definition.
      * WebPreferences.cpp: Remove the memoryInfo, setMemoryInfoEnabled methods.
      (WebPreferences::initializeDefaultSettings): Remove the WebKitMemoryInfoEnabledPreferenceKey entry initialization.
      * WebPreferences.h: Remove the memoryInfo, setMemoryInfoEnabled method declarations.
      * WebView.cpp:
      (WebView::notifyPreferencesChanged): Remove the Settings::setMemoryInfoEnabled call, the method is being removed.
      
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@151491 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      ee606799
    • zandobersek@gmail.com's avatar
      Unreviewed GTK build fix. · 959209ec
      zandobersek@gmail.com authored
      * GNUmakefile.list.am: Adding missing build targets to the plugin process after r151480.
      
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@151490 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      959209ec
    • commit-queue@webkit.org's avatar
      [GTK] Parameter 'pseudoElement' from function... · fee2155a
      commit-queue@webkit.org authored
      [GTK] Parameter 'pseudoElement' from function 'webkit_dom_dom_window_get_computed_style' should be allowed to be NULL
      https://bugs.webkit.org/show_bug.cgi?id=117332
      
      Patch by Diego Pino Garcia <dpino@igalia.com> on 2013-06-12
      Reviewed by Xan Lopez.
      
      Source/WebCore:
      
      Add parameter 'pseudoElement' to the list of parameters that are allowed to
      be NULL.
      
      * bindings/scripts/CodeGeneratorGObject.pm:
      (GetGReturnMacro):
      
      Source/WebKit/gtk:
      
      Add test for function 'webkit_dom_dom_window_get_computed_style'.
      
      * tests/testdomdomwindow.c:
      (dom_dom_window_fixture_setup): Test setup for function
      'dom_dom_window_get_computed_style'.
      (dom_dom_window_fixture_teardown): Test teardown for function
      'dom_dom_window_get_computed_style'.
      (loadedCallback):
      (test_dom_dom_window_get_computed_style): Checks function
      'dom_dom_window_get_computed_style'.
      (main):
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@151489 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      fee2155a
    • sergio@webkit.org's avatar
      Add CSS parsing recovery to functions · 0e29e856
      sergio@webkit.org authored
      https://bugs.webkit.org/show_bug.cgi?id=117500
      
      Reviewed by Andreas Kling.
      
      From Blink r150205 by <serya@chromium.org>
      
      Source/WebCore:
      
      Add parsing recovery capabilities to functions. Errors were
      correctly detected without this change but then the whole
      declaration was invalidated. By using expr_recovery to handle them
      we can recover from those errors and go on with the parsing.
      
      * css/CSSGrammar.y.in:
      
      LayoutTests:
      
      * fast/css/parsing-error-recovery.html: added some test cases for
      parsing recovery inside functions.
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@151488 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      0e29e856
    • ch.dumez@sisa.samsung.com's avatar
      Unreviewed, rolling out r151378. · 69031adb
      ch.dumez@sisa.samsung.com authored
      http://trac.webkit.org/changeset/151378
      https://bugs.webkit.org/show_bug.cgi?id=117400
      
      Decreased IDL readability a bit.
      
      * Modules/encryptedmedia/MediaKeys.idl:
      * Modules/mediastream/RTCIceCandidate.idl:
      * Modules/mediastream/RTCPeerConnection.idl:
      * Modules/mediastream/RTCSessionDescription.idl:
      * Modules/webaudio/OfflineAudioContext.idl:
      * Modules/websockets/WebSocket.idl:
      * bindings/scripts/CodeGeneratorJS.pm:
      (GenerateConstructorDefinition):
      * bindings/scripts/IDLAttributes.txt:
      * bindings/scripts/test/TestInterface.idl:
      * bindings/scripts/test/TestNamedConstructor.idl:
      * css/WebKitCSSMatrix.idl:
      * html/HTMLOptionElement.idl:
      * page/EventSource.idl:
      * workers/SharedWorker.idl:
      * workers/Worker.idl:
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@151487 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      69031adb
  2. 11 Jun, 2013 5 commits