Timeline
Feb 1, 2008:
- 4:19 PM Changeset in webkit [29929] by
-
- 3 edits1 add in trunk/WebCore
Reviewed by Darin's rubberstamp
Move the SecurityOriginHash out into a separate header as it will soon be used in different places
- WebCore.xcodeproj/project.pbxproj:
- platform/SecurityOriginHash.h: Added. (WebCore::SecurityOriginHash::hash): (WebCore::SecurityOriginHash::equal): (WebCore::SecurityOriginTraits::deletedValue): (WebCore::SecurityOriginTraits::emptyValue):
- storage/DatabaseTracker.cpp:
- 4:02 PM Changeset in webkit [29928] by
-
- 7 edits1 move5 adds in trunk/LayoutTests
Reviewed by Dave Hyatt.
- address the remaining tests listed under <rdar://problem/5075790> Layout tests that depend on certain fonts don't match Mac results
- fast/text/cg-fallback-bolding.html: Changed to use Ahem instead of Geneva.
- fast/text/international/bidi-fallback-font-weight.html: Made Mac-only.
- platform/mac/fast/text/cg-fallback-bolding-expected.checksum: Updated.
- platform/mac/fast/text/cg-fallback-bolding-expected.png: Updated
- platform/mac/fast/text/cg-fallback-bolding-expected.txt: Updated.
- platform/mac/fast/text/international/bidi-fallback-font-weight.html: Copied from LayoutTests/fast/text/international/bidi-fallback-font-weight.html.
- platform/win/Skipped: Removed passing tests and categorized remaining failures.
- platform/win/svg/W3C-SVG-1.1: Added.
- platform/win/svg/W3C-SVG-1.1/resources: Added.
- platform/win/svg/W3C-SVG-1.1/resources/Mac-compatible-font-fallback.css: Added.
- platform/win/svg/W3C-SVG-1.1/resources/run-webkit-tests-epilogue.html: Added.
- platform/win/svg/W3C-SVG-1.1/resources/run-webkit-tests-prologue.html: Added.
- svg/text/text-tselect-02-f.svg: Added @font-face rule to use Lucida Grande for Hebrew.
- 2:18 PM Changeset in webkit [29927] by
-
- 2 edits in trunk/WebCore
Reviewed by Tim.
- fixed mistake in the JSLock fix that caused an assertion every time
- bindings/js/JSCustomSQLTransactionCallback.cpp: (WebCore::JSCustomSQLTransactionCallback::JSCustomSQLTransactionCallback): Added JSLock. (WebCore::unprotectOnMainThread): Ditto.
- 1:51 PM Changeset in webkit [29926] by
-
- 5 edits in trunk/WebCore
Reviewed by Antti and Darin.
Change Text::createWithLengthLimit to take a UChar pointer instead of a string. This
lets us avoid making a copy of the buffer in TextDocument.cpp.
- dom/Text.cpp: (WebCore::Text::createWithLengthLimit):
- dom/Text.h:
- html/HTMLParser.cpp: (WebCore::HTMLParser::parseToken):
- loader/TextDocument.cpp: (WebCore::TextTokenizer::write):
- 1:35 PM Changeset in webkit [29925] by
-
- 2 edits in trunk/WebCore
Reviewed by Adam.
- platform/network/mac/ResourceHandleMac.mm: (WebCore::ResourceHandle::loadsBlocked): Fix grammar and word typo.
- 1:33 PM Changeset in webkit [29924] by
-
- 2 edits in trunk/WebCore
Reviewed by Adam.
- fix <rdar://problem/4527931> showModalDialog calls from onload functions fail (Aspect)
No automated regression test because showModalDialog doesn't work in DumpRenderTree.
- platform/network/mac/ResourceHandleMac.mm: (WebCore::CallbackGuard::CallbackGuard): Added. (WebCore::CallbackGuard::~CallbackGuard): Added. (WebCore::ResourceHandle::supportsBufferedData): Simplified to take advantage of how static initialization works in C++. (WebCore::ResourceHandle::loadsBlocked): Always return false on Leopard. The problem in NSURLConnection that created the need to block loads is fixed in Leoaprd. This is the bug fix. (-[WebCoreResourceHandleAsDelegate connection:willSendRequest:redirectResponse:]): Use CallbackGuard instead of directly incrementing the count; allows us to omit the code entirely on Leopard. (-[WebCoreResourceHandleAsDelegate connection:didReceiveAuthenticationChallenge:]): Ditto. (-[WebCoreResourceHandleAsDelegate connection:didCancelAuthenticationChallenge:]): Ditto. (-[WebCoreResourceHandleAsDelegate connection:didReceiveResponse:]): Ditto. (-[WebCoreResourceHandleAsDelegate connection:didReceiveData:lengthReceived:]): Ditto. (-[WebCoreResourceHandleAsDelegate connection:willStopBufferingData:]): Ditto. (-[WebCoreResourceHandleAsDelegate connectionDidFinishLoading:]): Ditto. (-[WebCoreResourceHandleAsDelegate connection:didFailWithError:]): Ditto. (-[WebCoreResourceHandleAsDelegate connection:willCacheResponse:]): Ditto. Fixes a problem where this could leave inNSURLConnectionCallback set permanently in one of the code paths; this would break showModalDialog from that point on. This problem doesn't affect Safari.
- 12:30 PM Changeset in webkit [29923] by
-
- 2 edits in trunk/WebCore
- platform/network/mac/ResourceHandleMac.mm: (WebCore::ResourceHandle::loadsBlocked): Roll out accidentally checked in file.
- 12:28 PM Changeset in webkit [29922] by
-
- 19 edits in trunk
WebCore:
Reviewed by Geoff and Brady.
- fix <rdar://problem/5680469> Each database worker thread takes a JSLock, which slows down all JavaScript execution until the thread terminates
- bindings/js/JSCustomSQLTransactionCallback.cpp: (WebCore::JSCustomSQLTransactionCallback::JSCustomSQLTransactionCallback): Explicitly gcProtect, since we aren't using ProtectedPtr any more. (WebCore::unprotectOnMainThread): Added. (WebCore::JSCustomSQLTransactionCallback::~JSCustomSQLTransactionCallback): Use unprotectOnMainThread rather than doing a gcUnprotect here.
- bindings/js/JSCustomSQLTransactionCallback.h: Changed m_callback from a ProtectedPtr to a plain old JSObject*; we gcProtect by hand now.
- loader/icon/IconDatabase.cpp: (WebCore::IconDatabase::notifyPendingLoadDecisionsOnMainThread): Renamed. Changed to takea a void* with the icon database pointer. (WebCore::IconDatabase::notifyPendingLoadDecisions): Renamed. (WebCore::IconDatabase::performURLImport): Updated name, and pass the icon database pointer in.
- loader/icon/IconDatabase.h: Renamed and added void* parameter.
- platform/Threading.h: Changed callOnMainThread to take a function with a single parameter rather than a function with no parameters. Added a typedef.
- platform/gtk/ThreadingGtk.cpp: (WebCore::callFunctionOnMainThread): Changed to use a structure with both a function pointer and a context pointer, and delete it after calling. (WebCore::callOnMainThread): Changed to create the structure with both the function and context pointer.
- platform/mac/Threading.mm: (-[WebCoreFunctionWrapper initWithFunction:context:]): Added context. (-[WebCoreFunctionWrapper invoke]): Updated to do both function and context. Renamed from _call. (WebCore::callOnMainThread): Updated to take both function and context.
- platform/qt/ThreadingQt.cpp: (WebCore::PerformFunctionEvent::PerformFunctionEvent): Added context. (WebCore::PerformFunctionEvent::invoke): Added. Calls function with context pointer. (WebCore::MainThreadInvoker::event): Changed to call invoke. (WebCore::callOnMainThread): Updated to take both function and context.
- platform/win/ThreadingWin.cpp: Added FunctionWithContext. (WebCore::callFunctionsOnMainThread): Changed to pass call function with context. (WebCore::callOnMainThread): Updated to queue both function and context.
- platform/wx/ThreadingWx.cpp: (WebCore::callOnMainThread): Updated to take both function and context.
- storage/Database.cpp: (WebCore::Database::scheduleTransactionCallback): Pass 0 for context. (WebCore::Database::deliverAllPendingCallbacks): Changed to take ignored context pointer.
- storage/Database.h: Added void* parameter to deliverAllPendingCallbacks.
- storage/DatabaseTracker.cpp: (WebCore::DatabaseTracker::scheduleForNotification): Pass 0 for context. (WebCore::DatabaseTracker::notifyDatabasesChanged): Changed to take ignored context pointer.
- storage/DatabaseTracker.h: Added void* parameter to notifyDatabasesChanged.
WebKit/win:
- fix <rdar://problem/5680469> Each database worker thread takes a JSLock, which slows down all JavaScript execution until the thread terminates
- WebIconDatabase.cpp: (WebIconDatabase::scheduleNotificationDelivery): Updated for new callOnMainThread. (WebIconDatabase::deliverNotifications): Ditto.
- WebIconDatabase.h: Ditto.
- 11:42 AM Changeset in webkit [29921] by
-
- 2 edits in trunk/WebCore
2008-02-01 Jon Honeycutt <jhoneycutt@apple.com>
Rubber-stamped by Ollie.
- plugins/win/PluginPackageWin.cpp: (WebCore::PluginPackageWin::load): Fix two function pointers
- 11:39 AM Changeset in webkit [29920] by
-
- 3 edits2 moves2 adds3 deletes in trunk/LayoutTests
Reviewed by Alexey Proskuryakov.
- change fast/xsl/xslt-text to dump as text
- fast/xsl/resources/xslt-text.xml: Copied from LayoutTests/fast/xsl/xslt-text.xml.
- fast/xsl/resources/xslt-text.xsl: Copied from LayoutTests/fast/xsl/xslt-text.xsl.
- fast/xsl/xslt-processor.html:
- fast/xsl/xslt-text-expected.txt: Added.
- fast/xsl/xslt-text.html: Added.
- fast/xsl/xslt-text.xml: Removed.
- fast/xsl/xslt-text.xsl: Removed.
- platform/mac/fast/xsl/xslt-text-expected.checksum: Removed.
- platform/mac/fast/xsl/xslt-text-expected.png: Removed.
- platform/mac/fast/xsl/xslt-text-expected.txt: Removed.
- platform/win/Skipped:
- 11:35 AM Changeset in webkit [29919] by
-
- 2 edits in trunk/WebCore
Reviewed by John Sullivan, Oliver Hunt.
Fixed <rdar://problem/5688039> REGRESSION (r29428): Weather widget
fails to load due to reliance on "var location" quirk
Added a Dashboard quirk.
- bindings/js/kjs_window.cpp: (KJS::Window::put): In DB backwards compatibility mode, make assignment to window.location in top-level windows shadow the DOM location API instead of performing a navigation. This makes a reasonable amount of sense, because widgets can't navigate their top-level frames, anyway.
- 11:21 AM Changeset in webkit [29918] by
-
- 13 edits4 adds in trunk
WebCore:
Fix for bug 11387, CSS3 :empty selector is not dynamic.
Add support for dynamically updating elements whose style is affected by :empty when the child count changes.
Reviewed by olliej, mitzpettel
fast/css/empty-pseudo-class.html
- css/CSSStyleSelector.cpp: (WebCore::CSSStyleSelector::locateSharedStyle): (WebCore::CSSStyleSelector::checkOneSelector):
- dom/Element.cpp: (WebCore::Element::childrenChanged):
- dom/Element.h:
- rendering/RenderStyle.h: (WebCore::RenderStyle::affectedByEmpty): (WebCore::RenderStyle::emptyState): (WebCore::RenderStyle::setEmptyState):
LayoutTests:
Add a new layout test for dynamic :empty selectors.
Reviewed by olliej
- fast/css/empty-pseudo-class.html: Added.
- platform/mac/fast/css/empty-pseudo-class-expected.checksum: Added.
- platform/mac/fast/css/empty-pseudo-class-expected.png: Added.
- platform/mac/fast/css/empty-pseudo-class-expected.txt: Added.
- 10:38 AM Changeset in webkit [29917] by
-
- 1 edit in trunk/WebKitLibraries/ChangeLog
Fix typo.
- 10:36 AM Changeset in webkit [29916] by
-
- 2 edits in trunk/WebKitLibraries
<rdar://problem/5717523> Don't set DEP opt-in flag (data execution prevention) since it is incompaible with the video plugin used on CNN.com
Rubber-stamped by John Honeycutt.
- win/tools/vsprops/common.vsprops:
- 10:06 AM Changeset in webkit [29915] by
-
- 2 edits in trunk/WebCore
Reviewed by Darin Adler.
- switch line breaking from using Carbon Unicode Utilities to using ICU on Leopard <http://bugs.webkit.org/show_bug.cgi?id=4628>
fast/text/international/thai-line-breaks.html results do not change.
- rendering/break_lines.cpp: (WebCore::nextBreakablePosition): Changed to use Carbon only on Tiger.
- 9:45 AM Changeset in webkit [29914] by
-
- 2 edits in trunk/WebCore
2008-02-01 Rodney Dawes <dobey@wayofthemonkey.com>
Reviewed by Darin.
Update npfunctions.h to export the proper function signatures for UNIX and add
the appropriate #if around the Windows versions
- plugins/npfunctions.h:
- 6:32 AM UsingWine edited by
- fixed link to nightly builds (diff)
- 5:20 AM UsingWine edited by
- rearranged scripts into separate ones per program (diff)
- 4:57 AM UsingWine created by
- added notes on running Safari with custom webkit under wine
- 1:35 AM Changeset in webkit [29913] by
-
- 12 edits3 adds in trunk
Reviewed by Oliver.
Add getSVGDocument to frame and iframe.
Test: svg/custom/frame-getSVGDocument.html
- html/HTMLEmbedElement.cpp:
- html/HTMLEmbedElement.h:
- html/HTMLFrameElement.idl:
- html/HTMLFrameOwnerElement.cpp: (WebCore::HTMLFrameOwnerElement::getSVGDocument):
- html/HTMLFrameOwnerElement.h:
- html/HTMLIFrameElement.idl:
- html/HTMLObjectElement.cpp:
- html/HTMLObjectElement.h:
- 1:14 AM Changeset in webkit [29912] by
-
- 4 edits in trunk
Versioning.
- 1:13 AM Changeset in webkit [29911] by
-
- 1 copy in tags/Safari-5525.7
New tag.
- 12:00 AM Changeset in webkit [29910] by
-
- 5 edits in trunk/WebCore
Reviewed by darin.
Beat CSSParser with the RefPtr stick. Hopefully squashing any leaks in the process.
(Deploy RefPtr and PassRefPtr throughout CSSParser)
http://bugs.webkit.org/show_bug.cgi?id=17108
No functional changes, thus no tests.
- css/CSSParser.cpp: (WebCore::CSSParser::parseColor): (WebCore::CSSParser::parseValue): (WebCore::CSSParser::addBackgroundValue): (WebCore::CSSParser::parseBackgroundShorthand): (WebCore::CSSParser::addTransitionValue): (WebCore::CSSParser::parseTransitionShorthand): (WebCore::CSSParser::parseBackgroundPosition): (WebCore::CSSParser::parseBackgroundProperty): (WebCore::CSSParser::parseTransitionProperty): (WebCore::CSSParser::parseFontFaceSrc): (WebCore::ShadowParseContext::ShadowParseContext): (WebCore::CSSParser::parseShadow): (WebCore::BorderImageParseContext::BorderImageParseContext): (WebCore::BorderImageParseContext::commitBorderImage):
- css/CSSParser.h:
- css/SVGCSSParser.cpp: (WebCore::CSSParser::parseSVGValue):
Jan 31, 2008:
- 11:01 PM Changeset in webkit [29909] by
-
- 4 edits1 move in trunk
2008-01-31 Alp Toker <alp@atoker.com>
Reviewed by Mark Rowe.
Fix the pkg-config file to follow GTK+ package naming and versioning
conventions.
Remove unneeded dependency listings.
- 10:08 PM Changeset in webkit [29908] by
-
- 2 edits in trunk/WebCore
Add platform/Locker.h to vcproj
RS=Brady
- WebCore.vcproj/WebCore.vcproj:
- 8:31 PM Changeset in webkit [29907] by
-
- 2 edits in trunk/WebCore
Build fix
- WebCore.xcodeproj/project.pbxproj: Header -> WebKit
- 8:02 PM Changeset in webkit [29906] by
-
- 3 edits1 add in trunk/WebCore
Reviewed by Maciej
Abstracted the concept of a "Locker" as upcoming work will rely on it.
- platform/Locker.h: Added. Template class to "lock()" and "unlock()" some arbitrary object (WebCore::Locker::Locker): Lock the object (WebCore::Locker::~Locker): Unlock it
- platform/Threading.h:
class MutexLockeris nowtypedef Locker<Mutex> - WebCore.xcodeproj/project.pbxproj:
- 5:53 PM Changeset in webkit [29905] by
-
- 15 edits10 deletes in trunk
WebCore:
<rdar://problem/5716912> REGRESSION (r29816): Can't navigate back to
Gmail Inbox from Gmail Compose page
Rolling out the change that caused the regression.
- WebCore.base.exp:
- bindings/js/JSHTMLDocumentCustom.cpp: (WebCore::JSHTMLDocument::open):
- dom/DOMImplementation.cpp: (WebCore::DOMImplementation::createHTMLDocument):
- dom/Document.cpp: (WebCore::Document::recalcStyle): (WebCore::Document::open): (WebCore::Document::write): (WebCore::Document::clear):
- dom/Document.h:
- history/HistoryItem.cpp: (WebCore::HistoryItem::HistoryItem):
- history/HistoryItem.h:
- loader/FrameLoader.cpp: (WebCore::FrameLoader::didExplicitOpen): (WebCore::FrameLoader::load): (WebCore::FrameLoader::reloadAllowingStaleData): (WebCore::FrameLoader::reload): (WebCore::FrameLoader::tokenizerProcessedData): (WebCore::FrameLoader::post): (WebCore::FrameLoader::continueLoadAfterNewWindowPolicy): (WebCore::FrameLoader::shouldTreatURLAsSameAsCurrent): (WebCore::FrameLoader::loadItem):
- loader/FrameLoader.h:
- platform/text/CharacterNames.h:
- xml/DOMParser.cpp: (WebCore::DOMParser::parseFromString):
- xml/XMLHttpRequest.cpp: (WebCore::XMLHttpRequest::getResponseXML):
- xml/XSLTProcessor.cpp: (WebCore::XSLTProcessor::createDocumentFromSource):
LayoutTests:
<rdar://problem/5716912> REGRESSION (r29816): Can't navigate back to
Gmail Inbox from Gmail Compose page
Rolling out the change that caused the regression.
- http/tests/navigation/document-open-adds-history-item-expected.txt: Removed.
- http/tests/navigation/document-open-adds-history-item.html: Removed.
- http/tests/navigation/document-open-delayed-adds-history-item-expected.txt: Removed.
- http/tests/navigation/document-open-delayed-adds-history-item.html: Removed.
- http/tests/navigation/document-open-new-window-adds-history-item-expected.txt: Removed.
- http/tests/navigation/document-open-new-window-adds-history-item.html: Removed.
- http/tests/navigation/document-open-replace-no-history-item-expected.txt: Removed.
- http/tests/navigation/document-open-replace-no-history-item.html: Removed.
- http/tests/navigation/resources/document-open-page-2.html: Removed.
- http/tests/navigation/resources/document-open.js: Removed.
- 5:03 PM Changeset in webkit [29904] by
-
- 2 edits in trunk/LayoutTests
- re-categorize three skipped tests
- platform/win/Skipped:
- 5:01 PM Changeset in webkit [29903] by
-
- 2 edits in trunk/WebKit/win
Fix <rdar://5713172> Focus lost from Gmail message body after toggling
tabs
This regressed in r29583. Prior to r29583, the code path that resulted
in calling focusController->setFocusedFrame(0) was never getting hit.
r29583 changed the code to adhere to the comment above the code and
actually call setFocusedFrame(0) when losing focus to another part of
the window. It turns out this is wrong behavior (and not what
WebKit/mac does). The only time we should clear the focused frame is
when we tab out of the WebView, and that's already taken care of by
FocusController::advanceFocus. The code to clear the focused frame in
WebView was written before FocusController::advanceFocus, so it was
correct at one point but no longer is.
We now never call setFocusedFrame(0) when the WebView receives a
WM_KILLFOCUS message. Instead we just call setFocused(false) on the
focused frame so that blur events will be sent. This matches what
WebKit/mac does.
Reviewed by Darin and Adele.
- WebView.cpp: (WebViewWndProc): Also reworded a couple of comments to make them clearer.
- 4:50 PM Changeset in webkit [29902] by
-
- 2 edits in trunk/WebCore
Reviewed by Tim Hatcher.
Bug 14960: Cannot copy text in Web Inspector Network panel
http://bugs.webkit.org/show_bug.cgi?id=14960
- page/inspector/inspector.css:
- 4:48 PM Changeset in webkit [29901] by
-
- 2 edits in trunk/WebCore
Reviewed by Adam Roben.
Bug 14514: Add full URL view on mouseOver in the network page
http://bugs.webkit.org/show_bug.cgi?id=14514
<rdar://problem/5712841>
- page/inspector/inspector.js:
- 4:35 PM Changeset in webkit [29900] by
-
- 5 edits in trunk/LayoutTests
Reviewed by Geoffrey Garen.
- change the font used in fast/text/monospace-width-cache.html again to a font that contains the U+2500 character in Tiger too
- fast/text/monospace-width-cache.html:
- platform/mac/fast/text/monospace-width-cache-expected.checksum:
- platform/mac/fast/text/monospace-width-cache-expected.png:
- platform/mac/fast/text/monospace-width-cache-expected.txt:
- 4:00 PM Changeset in webkit [29899] by
-
- 4 edits in trunk
Fix r29898 ChangeLog entries: rubber-stamped, not reviewed
- 3:51 PM Changeset in webkit [29898] by
-
- 13 edits18 copies1 move in trunk
2008-01-31 Alp Toker <alp@atoker.com>
Reviewed by Adam Roben.
http://bugs.webkit.org/show_bug.cgi?id=17006
[GTK] Header path should be webkit/webkit.h
Move the GTK+ API sources as needed and update the build systems.
- 3:09 PM Changeset in webkit [29897] by
-
- 2 edits in trunk/WebCore
Reviewed by Geoff Garen.
No need to get the prototype just to compare compare property names.
- bindings/js/JSHistoryCustom.cpp: (WebCore::JSHistory::customGetOwnPropertySlot):
- 1:37 PM Changeset in webkit [29896] by
-
- 6 edits in trunk
WebCore:
Fix non-mac builds.
- GNUmakefile.am:
- WebCore.pro:
- WebCore.vcproj/WebCore.vcproj:
- WebCoreSources.bkl:
- 1:06 PM Changeset in webkit [29895] by
-
- 3 edits in trunk/LayoutTests/platform/mac/fast/lists
- 1:04 PM Changeset in webkit [29894] by
-
- 3 edits in trunk/LayoutTests/platform/mac/tables/mozilla/bugs
- 1:03 PM Changeset in webkit [29893] by
-
- 4 adds in trunk/LayoutTests
- 1:03 PM Changeset in webkit [29892] by
-
- 2 edits in trunk/WebCore
Fix for bug 17101.
List items need to ignore the line box shrinking quirk, since IE and Firefox both do.
Reviewed by Eric
fast/lists/list-item-line-height.html
- rendering/InlineFlowBox.h: (WebCore::InlineFlowBox::InlineFlowBox):
- 12:11 PM Changeset in webkit [29891] by
-
- 26 edits in trunk/LayoutTests
Reviewed by Adam Roben.
- change several tests that use Mac-only fonts to use fonts that are available on both Windows and Mac OS X
- fast/text/monospace-width-cache.html:
- platform/mac/fast/text/monospace-width-cache-expected.checksum:
- platform/mac/fast/text/monospace-width-cache-expected.png:
- platform/mac/fast/text/monospace-width-cache-expected.txt:
- platform/mac/tables/mozilla/bugs/bug10633-expected.checksum:
- platform/mac/tables/mozilla/bugs/bug10633-expected.png:
- platform/mac/tables/mozilla/bugs/bug10633-expected.txt:
- platform/mac/tables/mozilla/bugs/bug113424-expected.checksum:
- platform/mac/tables/mozilla/bugs/bug113424-expected.png:
- platform/mac/tables/mozilla/bugs/bug113424-expected.txt:
- platform/mac/tables/mozilla/bugs/bug23151-expected.checksum:
- platform/mac/tables/mozilla/bugs/bug23151-expected.png:
- platform/mac/tables/mozilla/bugs/bug23151-expected.txt:
- platform/mac/tables/mozilla/core/bloomberg-expected.checksum:
- platform/mac/tables/mozilla/core/bloomberg-expected.png:
- platform/mac/tables/mozilla/core/bloomberg-expected.txt:
- platform/mac/tables/mozilla_expected_failures/bugs/bug11331-expected.checksum:
- platform/mac/tables/mozilla_expected_failures/bugs/bug11331-expected.png:
- platform/mac/tables/mozilla_expected_failures/bugs/bug11331-expected.txt:
- platform/win/Skipped:
- tables/mozilla/bugs/bug10633.html:
- tables/mozilla/bugs/bug113424.html:
- tables/mozilla/bugs/bug23151.html:
- tables/mozilla/core/bloomberg.html:
- tables/mozilla_expected_failures/bugs/bug11331.html:
- 11:52 AM Changeset in webkit [29890] by
-
- 15 edits2 moves8 adds in trunk
WebCore:
Reviewed by Darin Adler.
Fix for <rdar://problem/5708993> Mutability of the History object
- Don't allow cross-domain get access to any of the history objects properties except the back(), forward() and go() methods.
- Don't allow cross-domain put access to any of the history objects properties.
- Don't allow cross-domain enumeration of the History or Location objects.
Tests: http/tests/security/cross-frame-access-history-get-override.html
http/tests/security/cross-frame-access-history-get.html
http/tests/security/cross-frame-access-history-put.html
- WebCore.xcodeproj/project.pbxproj:
- bindings/js/JSDOMWindowCustom.cpp: Remove unnessary KJS::'s (WebCore::JSDOMWindow::customGetOwnPropertySlot): (WebCore::JSDOMWindow::customPut): (WebCore::JSDOMWindow::getPropertyNames): Moved implementation from KJS::Window now that the declaration is autogenerated using the new CustomGetPropertyNames. (WebCore::JSDOMWindow::postMessage):
- bindings/js/JSHistoryCustom.cpp: Added. (WebCore::allowsAccessFromFrame): (WebCore::JSHistory::customGetOwnPropertySlot): Only allow getting the declared functions back(), forward() and go() from cross-domain. Deny all other gets. (WebCore::JSHistory::customPut): Don't allow putting cross-domain. (WebCore::JSHistory::getPropertyNames): Don't allow enumeration cross-domain.
- bindings/js/JSLocation.cpp: (WebCore::allowsAccessFromFrame): (WebCore::JSLocation::getPropertyNames): Don't allow enumeration cross-domain.
- bindings/js/JSLocation.h:
- bindings/js/kjs_window.cpp:
- bindings/js/kjs_window.h:
- bindings/scripts/CodeGeneratorJS.pm: Add support for new CustomGetPropertNames extended attribute and changed the logic of CustomPutFunction to create an overrided put() function even if no read-write properties exist.
- page/DOMWindow.idl: Added CustomGetPropertNames
- page/History.idl: Added CustomGetPropertNames
LayoutTests:
Reviewed by Darin Adler.
Tests for <rdar://problem/5708993> Mutability of the History object
- http/tests/security/cross-frame-access-enumeration-expected.txt:
- http/tests/security/cross-frame-access-enumeration.html:
- http/tests/security/cross-frame-access-history-expected.txt: Removed.
- http/tests/security/cross-frame-access-history-get-expected.txt: Renamed from LayoutTests/http/tests/security/cross-frame-access-history-expected.txt.
- http/tests/security/cross-frame-access-history-get-override-expected.txt: Added.
- http/tests/security/cross-frame-access-history-get-override.html: Added.
- http/tests/security/cross-frame-access-history-get.html: Renamed from LayoutTests/http/tests/security/cross-frame-access-history.html.
- http/tests/security/cross-frame-access-history-put-expected.txt: Added.
- http/tests/security/cross-frame-access-history-put.html: Added.
- http/tests/security/cross-frame-access-history.html: Removed.
- http/tests/security/resources/cross-frame-access.js:
- http/tests/security/resources/cross-frame-iframe-for-enumeration-test.html:
- http/tests/security/resources/cross-frame-iframe-for-history-get-override-test.html: Added.
- http/tests/security/resources/cross-frame-iframe-for-history-get-test.html: Added.
- http/tests/security/resources/cross-frame-iframe-for-history-put-test.html: Added.
- 11:01 AM Changeset in webkit [29889] by
-
- 5 edits in trunk
WebCore:
Reviewed by Darin Adler.
<rdar://problem/5708115> REGRESSION: Words selected with a double click and copied won't paste into Mail
- page/mac/WebCoreFrameBridge.h: Re-exposed smartInsertForString:, it's used by a WebKit method used by Mail.
WebKit/mac:
Reviewed by Darin Adler.
<rdar://problem/5708115> REGRESSION: Words selected with a double click and copied won't paste into Mail
- WebView/WebHTMLView.mm: (-[WebHTMLView _smartInsertForString:replacingRange:beforeString:afterString:]): Brought this back, it's used by Mail. (-[WebHTMLView _canSmartReplaceWithPasteboard:]): This WebInternal method is also used by Mail. Moved to WebPrivate.
- WebView/WebHTMLViewPrivate.h: Expose two methods that Mail uses here, so that we don't accidently remove them in the future.
- 10:50 AM Changeset in webkit [29888] by
-
- 3 edits in trunk/WebCore
Add line box drawing code to InspectorController::drawNodeHighlight
This makes drawNodeHighlight a complete replacement for the painting
code in WebKit/mac/WebNodeHighlightView.mm, and also brings line box
rects to Windows for the first time.
Reviewed by Darin.
- dom/Node.h: Changed isSVGElement to always exist, but to only be virtual when ENABLE(SVG) is true. This way you can always call node->isSVGElement() without checking ENABLE(SVG).
- page/InspectorController.cpp: (WebCore::InspectorController::drawNodeHighlight): Ported line box rect code from the Mac implementation in WebNodeHighlightView.mm.
- 10:47 AM Changeset in webkit [29887] by
-
- 7 edits in trunk
Let WebCore take care of the highlight drawing entirely
WebCore:
Put more knowledge about the node highlight in WebCore
InspectorController now calculates the overlay rect and node rect when
drawing the node highlight instead of having them be passed in.
InspectorController now holds onto the highlighted node so that it can
determine these rects.
Once all platforms are calling down to drawNodeHighlight instead of
drawing the highlight themselves, we can change
InspectorClient::highlight(Node*) to something like
InspectorClient::updateAndShowHighlight().
This also fixes Bug 14264: Node highlight makes it impossible to
scroll the page
<http://bugs.webkit.org/show_bug.cgi?id=14264>
<rdar://5712788>
Reviewed by Darin.
- page/InspectorController.cpp: (WebCore::InspectorController::highlight): Store the node for use in drawNodeHighlight. (WebCore::InspectorController::drawNodeHighlight): Changed to be a const instance method. Now calculates the overlay rect and node rect instead of having them passed in.
- page/InspectorController.h:
WebKit/win:
Let WebCore take care of the highlight drawing entirely
Reviewed by Darin.
- WebInspectorClient.cpp: (WebInspectorClient::highlight): We now just show our highlight window and let WebCore figure out what/where to paint. Once all ports follow suit the Node* parameter to this method should be removed entirely, and the name should probably change to updateAndShowHighlight or something similar.
- WebNodeHighlight.cpp: Renamed m_webView to m_inspectedWebViewWindow. (WebNodeHighlight::WebNodeHighlight): Now takes a WebView* parameter. (WebNodeHighlight::show): Renamed from highlight(). Now gets the WebView's HWND. Updated for member rename/removal. (WebNodeHighlight::updateWindow): Updated for member rename and for InspectorController changes.
- WebNodeHighlight.h: Added m_inspectedWebView member, removed m_rect member, renamed highlight(Node*) -> show().
- 10:44 AM Changeset in webkit [29886] by
-
- 5 edits in trunk
Move node highlight drawing code to WebCore
WebCore:
Add node highlight drawing code to InspectorController
The code came from WebKit/win/WebNodeHighlight.cpp. It's not quite as
complete as the Mac implementation (in particular, it doesn't handle
line-box rects), but it's a start.
Reviewed by Darin.
- page/InspectorController.cpp: (WebCore::InspectorController::drawNodeHighlight): Added.
- page/InspectorController.h:
WebKit/win:
Move node highlight drawing code to WebCore
Reviewed by Darin.
- WebNodeHighlight.cpp: (WebNodeHighlight::updateWindow): Call into WebCore to do the node highlight drawing.
- 10:34 AM Changeset in webkit [29885] by
-
- 4 edits4 adds in trunk
WebCore:
Reviewed by Dave Hyatt.
- fix http://bugs.webkit.org/show_bug.cgi?id=17107 <rdar://problem/5716722> REGRESSION (r29834): Article text on redhat.com magazine site appears to be painting twice
Test: fast/block/float/intruding-painted-twice.html
- rendering/RenderBlock.cpp: (WebCore::RenderBlock::layoutBlock): Pass 'false' for the new makeChildPaintOtherFloats parameter to addOverhangingFloats() because at this point we are only taking away floats from the child. (WebCore::RenderBlock::layoutBlockChildren): Pass 'true' for the new makeChildPaintOtherFloats parameter to addOverhangingFloats() iff the child was not laid out again. Only in that case, it may have overhanging floats that it does not paint because they used to be overhanging from the parent, but now they are not. (WebCore::RenderBlock::addOverhangingFloats): Refined the conditions for making the child paint the float: require that the float be a descendant of the child (the other case is when it intrudes into the child from another sibling) and that it does not have a layer (in which case it paints itself). In addition, do the check only if the caller passed 'true' for the makeChildPaintOtherFloats parameter.
- rendering/RenderBlock.h:
LayoutTests:
Reviewed by Dave Hyatt.
- pixel test for http://bugs.webkit.org/show_bug.cgi?id=17107 <rdar://problem/5716722> REGRESSION (r29834): Article text on redhat.com magazine site appears to be painting twice
- fast/block/float/intruding-painted-twice.html: Added.
- platform/mac-leopard/fast/block/float/intruding-painted-twice-expected.checksum: Added.
- platform/mac-leopard/fast/block/float/intruding-painted-twice-expected.png: Added.
- platform/mac/fast/block/float/intruding-painted-twice-expected.txt: Added.