Context Navigation

← Previous Period
Next Period →

Timeline

10/24/07:

23:38 Changeset [27022] by eseidel

2007-10-24 Eric Seidel <eric@webkit.org>

Reviewed by Maciej.

Add a JSGlobalObject class and remove the InterpreterMap
http://bugs.webkit.org/show_bug.cgi?id=15681

This required making JSCallbackObject a template class to allow for
JSGlobalObjects with JSCallbackObject functionality.

SunSpider claims this was a 0.5% speedup.

API/JSCallbackObject.cpp: (KJS::):
API/JSCallbackObject.h:
API/JSCallbackObjectFunctions.h: Copied from API/JSCallbackObject.cpp. (KJS::::JSCallbackObject): (KJS::::init): (KJS::::~JSCallbackObject): (KJS::::initializeIfNeeded): (KJS::::className): (KJS::::getOwnPropertySlot): (KJS::::put): (KJS::::deleteProperty): (KJS::::implementsConstruct): (KJS::::construct): (KJS::::implementsHasInstance): (KJS::::hasInstance): (KJS::::implementsCall): (KJS::::callAsFunction): (KJS::::getPropertyNames): (KJS::::toNumber): (KJS::::toString): (KJS::::setPrivate): (KJS::::getPrivate): (KJS::::inherits): (KJS::::cachedValueGetter): (KJS::::staticValueGetter): (KJS::::staticFunctionGetter): (KJS::::callbackGetter):
API/JSClassRef.cpp: (OpaqueJSClass::prototype):
API/JSContextRef.cpp: (JSGlobalContextCreate):
API/JSObjectRef.cpp: (JSObjectMake): (JSObjectGetPrivate): (JSObjectSetPrivate):
API/JSValueRef.cpp: (JSValueIsObjectOfClass):
JavaScriptCore.exp:
JavaScriptCore.xcodeproj/project.pbxproj:
bindings/c/c_utility.cpp: (KJS::Bindings::convertValueToNPVariant):
bindings/jni/jni_jsobject.cpp:
bindings/objc/objc_utility.mm: (KJS::Bindings::convertValueToObjcValue):
kjs/Context.cpp: (KJS::Context::Context):
kjs/ExecState.cpp: (KJS::ExecState::lexicalInterpreter):
kjs/JSGlobalObject.h: Added. (KJS::JSGlobalObject::JSGlobalObject): (KJS::JSGlobalObject::isGlobalObject): (KJS::JSGlobalObject::interpreter): (KJS::JSGlobalObject::setInterpreter):
kjs/array_instance.cpp:
kjs/context.h:
kjs/function.cpp: (KJS::FunctionImp::callAsFunction): (KJS::GlobalFuncImp::callAsFunction):
kjs/interpreter.cpp: (KJS::Interpreter::Interpreter): (KJS::Interpreter::init): (KJS::Interpreter::~Interpreter): (KJS::Interpreter::globalObject): (KJS::Interpreter::initGlobalObject): (KJS::Interpreter::evaluate):
kjs/interpreter.h:
kjs/lookup.h: (KJS::cacheGlobalObject):
kjs/object.h: (KJS::JSObject::isGlobalObject):
kjs/testkjs.cpp:

23:26 JavaScript performance improvement ideas edited by eric@webkit.org

(diff)

22:38 JavaScript performance improvement ideas edited by eric@webkit.org

(diff)

22:33 Changeset [27021] by staikos

unbreak Qt build

22:23 Changeset [27020] by aroben

Fix <rdar://5549919> Add font database initialization code to WebCore and call it from WebKit

WebCore:

Add font database initialization code to WebCore

Reviewed by Ada.

WebCore.vcproj/WebCore.vcproj: Added FontDatabase.{cpp,h}.
platform/win/FontDatabase.cpp: Added. (WebCore::systemFontsDirectory): (WebCore::fontsPlistPath): (WebCore::systemHasFontsNewerThanFontsPlist): (WebCore::readFontPlist): (WebCore::populateFontDatabaseFromPlist): (WebCore::populateFontDatabaseFromFileSystem): (WebCore::writeFontDatabaseToPlist): (WebCore::populateFontDatabase): This is the only function callable from outside this file. It populates the font database once, either from the fonts plist, or from the filesystem (and then saves a new fonts plist).
platform/win/FontDatabase.h: Added.

WebKitTools:

Remove now-unnecessary call to InitializeCoreGraphics

WebKit takes care of this now.

Reviewed by Ada.

DumpRenderTree/win/DumpRenderTree.cpp: (initialize):

WebKit/win:

Fix <rdar://5549919> Initialize the font database before any font code is invoked

We initialize the database in two places:

When instantiating WebKitClassFactory, which is guaranteed to happen before any WebView is instantiated.
When making a WebCore::Font in WebKitGraphics.

This ensures that the font database will be populated before any font
code is invoked. We rely on WebCore to only populate the database
once.

Reviewed by Ada.

WebKitClassFactory.cpp: (WebKitClassFactory::WebKitClassFactory): Populate the font database.
WebKitGraphics.cpp: (makeFont): Ditto.

22:22 Changeset [27019] by aroben

Add [I]WebTextRenderer

WebKitTools:

Use WebTextRenderer in DRT

Reviewed by Ada.

DumpRenderTree/win/DumpRenderTree.cpp: (initialize): Use WebTextRenderer instead of using InitializeCoreGraphics/AddFontResourceEx.

WebKit/win:

Add [I]WebTextRenderer

Right now this class can only be used to add private fonts for use by
the running process, but will eventually be the home of the
WebKitGraphics functions.

Reviewed by Ada.

ForEachCoClass.h: Added WebTextRenderer.
Interfaces/WebKit.idl: Ditto.
WebKit.vcproj/Interfaces.vcproj: Ditto.
WebKit.vcproj/WebKit.vcproj: Ditto.
WebKit.vcproj/WebKitGUID.vcproj: Ditto.
WebKitClassFactory.cpp: Ditto.
WebTextRenderer.cpp: Added. (WebTextRenderer::createInstance): (WebTextRenderer::WebTextRenderer): (WebTextRenderer::~WebTextRenderer): (WebTextRenderer::QueryInterface): (WebTextRenderer::AddRef): (WebTextRenderer::Release): (WebTextRenderer::registerPrivateFont):
WebTextRenderer.h: Added.

22:22 Changeset [27018] by aroben

Use FOR_EACH_COCLASS in WebKitClassFactory

Reviewed by Ada.

WebError.cpp: (WebError::createInstance): Added an overload that takes no arguments to make the macro used in WebKitClassFactory work.
WebError.h:
WebKitClassFactory.cpp: (WebKitClassFactory::CreateInstance): Use FOR_EACH_COCLASS.

22:22 Changeset [27017] by aroben

Put FOR_EACH_COCLASS macro into its own file and export it

The macro used to be called FOR_EACH_CLASS and lived in WebKitDLL.cpp.
This way we will be able to use the macro in more places that care
about all WebKit's COM classes.

Reviewed by Ada.

ForEachCoClass.h: Added.
WebKit.vcproj/WebKit.vcproj: Copy ForEachCoClass.h to WebKitOutputDir, and added it to the project.
WebKitDLL.cpp: Updated for macro rename, and changed to #undef the macros we pass to FOR_EACH_COCLASS after we're done with them.

22:20 Changeset [27016] by aroben

Added some font-related functions needed for <rdar://5549919>

Reviewed by Ada.

win/include/WebKitSystemInterface/WebKitSystemInterface.h:
win/lib/WebKitSystemInterface.lib:
win/lib/WebKitSystemInterface_debug.lib:

22:01 Changeset [27015] by mrowe@apple.com

2007-10-24 Sven Herzberg <sven@imendio.com>

Reviewed by Mark Rowe.

Fixes http://bugs.webkit.org/show_bug.cgi?id=15614
Bug 15614: [GTK] qmake based backends don't build on OS X

Scripts/build-webkit: set QMAKESPEC correctly if building the QT or GTK backend on a Mac
Scripts/webkitdirs.pm: extracted the darwin-test from isOSX() into isDarwin() to make it reusable in other places (like the workaround- for-prebuilt-qmake in build-webkit)

20:51 Changeset [27014] by ddkilzer

Refurbish update-webkit script.

Reviewed by Adam.

Scripts/update-webkit: Add -hhelp switch and usage statement. Check result of

GetOptions() call. Fix -qquiet switch to be passed to svn command properly. Use

multi-argument version of system() for flexibility and security. Check for existence
of Internal directory using -d test instead of -x.

20:50 Changeset [27013] by ddkilzer

Minor clean-up of prepare-ChangeLog script.

Reviewed by Adam.

Scripts/prepare-ChangeLog: Keep list of updated ChangeLog files in an array instead of a string. Use multi-argument versions of open() and system() for flexibility and security.

19:40 Changeset [27012] by adachan

<rdar://problem/5552221> REGRESSION(310A24-ToT): Shortcut key disable. (15604)

Reviewed by Adam.

WebView.cpp: (WebView::handleEditingKeyboardEvent): don't handle system key events as text input (WebView::keyDown): only remove WM_SYSCHAR message from the queue if we handle it. For WM_SYSCHAR message that we don't handle, let it stay in the queue and return false to let windows handle it.

18:19 Changeset [27011] by aroben

Touch all files that include *.lut.h to fix Windows builds

18:16 Changeset [27010] by bdash

2007-10-24 Mark Rowe <mrowe@apple.com>

Gtk build fix. Track WebCore changes in r27004.

Api/webkitgtkglobal.cpp:

17:58 Changeset [27009] by justing

WebCore:

Reviewed by Darin Adler.

Correcting the fix for:
<rdar://problem/5544856>
REGRESSION: After typing 2-byte text, undo only undoes one keystroke at a time

Made removal of the previous composition part of the current Undo step in the
case where the new composition is the empty string, too.

editing/Editor.cpp: (WebCore::Editor::confirmComposition): Call the new TypingCommand::deleteSelection, which either has the currently open typing command delete the current selection, or opens a new typing command (of type DeleteSelection) if one is not already open. (WebCore::Editor::setComposition): Ditto.
editing/TypingCommand.cpp: (WebCore::TypingCommand::deleteSelection): Added. (WebCore::TypingCommand::doApply): Handle DeleteSelection. (WebCore::TypingCommand::deleteKeyPressed): Clarified which deleteSelection is called. (WebCore::TypingCommand::forwardDeleteKeyPressed): Ditto. (WebCore::TypingCommand::preservesTypingStyle): Handle DeleteSelection.
editing/TypingCommand.h:

LayoutTests:

Reviewed by Darin.

<rdar://problem/5544856>
REGRESSION: After typing 2-byte text, undo only undoes one keystroke at a time

platform/mac/editing/input/text-input-controller-expected.txt: The delete we were using previously to remove the old composition makes fewer calls to the editing delegate.

17:30 Changeset [27008] by weinig

Build fix.

WebCore.vcproj/WebCore.vcproj:

17:10 Changeset [27007] by thatcher

Build fix.

16:55 Changeset [27006] by thatcher

WebKit:

Reviewed by Mark Rowe.

<rdar://problem/5069711> OpenSource version of libWebKitSystemInterface.a is Tiger only, causes issues if used on Leopard

Use the WebKitSystemInterface that matches the system version.

Configurations/DebugRelease.xcconfig:
WebKit.xcodeproj/project.pbxproj:

WebKitLibraries:

Reviewed by Mark Rowe.

<rdar://problem/5069711> OpenSource version of libWebKitSystemInterface.a is Tiger only, causes issues if used on Leopard

Add system specific versions of WebKitSystemInterface.

libWebKitSystemInterface.a: Removed.
libWebKitSystemInterfaceLeopard.a: Added.
libWebKitSystemInterfaceTiger.a: Added.

16:29 Changeset [27005] by aliceli1

WebCore:

Reviewed by Oliver.

Fix <rdar://5410959> editing/selection/drag-to-contenteditable-iframe.html fails on Windows

page/win/EventHandlerWin.cpp: (WebCore::EventHandler::passMouseMoveEventToSubframe): Some mouse move events are actually drags, which on mac return early from this function. Adding the same logic to its Windows equivalent.

WebKitTools:

Reviewed by Oliver.

Fix <rdar://5410959> editing/selection/drag-to-contenteditable-iframe.html fails on Windows

Move the call to replaySavedEvents from doDragDrop to doMouseMove because we don't want to
replay the saved events when we're still processing the mousedown that starts the drag

DumpRenderTree/win/EventSender.cpp: (doMouseMove):
DumpRenderTree/win/UIDelegate.cpp: (UIDelegate::doDragDrop):

LayoutTests:

Removing fixed test

platform/win/Skipped:

16:02 Changeset [27004] by beidson

WebCore:

Reviewed by Anders

<rdar://5554130> DatabaseTracker.o has a global initializer

Since DatabaseTracker is a singleton, it makes much more sense to keep the database path as a member variable.
Now constructing the shared DatabaseTracker no longer implicitly opens it - It is only opened when the databases
path is set.

WebCore.vcproj/WebCore.vcproj: Copy WebCore/storage headers for WebKit build

storage/Database.cpp: (WebCore::Database::~Database): Remove bogus assertion

storage/DatabaseTracker.cpp: (WebCore::DatabaseTracker::DatabaseTracker): Does nothing now! Move this code to openTrackerDatabase (WebCore::DatabaseTracker::setDatabasePath): Set the member variable, also closing/opening the database if needed (WebCore::DatabaseTracker::databasePath): (WebCore::DatabaseTracker::openTrackerDatabase): To open and validate the Databases db, moved from the c'tor (WebCore::DatabaseTracker::fullPathForDatabase): Return the member variable
storage/DatabaseTracker.h:

WebKit/win:

Reviewed by Anders

Windows portion of <rdar://5554130>

Slowly introduce Windows WebKit portion of the Database API that sets the
on-disk location for databases

WebDatabaseManager.cpp: Added. (WebKitSetWebDatabasesPathIfNecessary):
WebDatabaseManager.h: Added.

WebKit.vcproj/WebKit.vcproj:

WebView.cpp: (WebView::initWithFrame): Call WebKitSetWebDatabasesPathIfNecessary()

WebKit:

Reviewed by Anders

<rdar://problem/5554130> DatabaseTracker.o has a global initializer

Misc/WebDatabaseManager.mm: (WebKitSetWebDatabasesPathIfNecessary): Call the member function instead of a static one

16:00 Changeset [27003] by kmccullo

WebKitTools:

Reviewed by Sam, Steve and Darin.

Stubbed out the WebScriptDebugListener functionality in the Server Connection class to prepare it for receiving those callbacks.
Also I changed the instantiation of DebuggerClient, DebuggerDocument and the ServerConnection to not need to be initialized with a server name, since that is not the way we connect to the WebKit server.

Drosera/DebuggerDocument.cpp: Fixed a bug where I was always logging no exception. Now it only loggs when there is an exception (DebuggerDocument::willExecuteStatement): (DebuggerDocument::didEnterCallFrame): (DebuggerDocument::willLeaveCallFrame): (DebuggerDocument::windowScriptObjectAvailable): (DebuggerDocument::callFunctionOnObject):
Drosera/win/DebuggerClient.cpp: No longer needs the ServerConnection to be instantiated with a server's name. (DebuggerClient::DebuggerClient): (DebuggerClient::didFinishLoadForFrame):
Drosera/win/DebuggerClient.h: Removed unsued variable.
Drosera/win/Drosera.cpp: Client no longer needs to be initialized with a server name. (Drosera::init): (Drosera::initServer):
Drosera/win/Drosera.h: No longer need the ServerConnection to be instantiated with a server's name.
Drosera/win/ServerConnection.cpp: (ServerConnection::ServerConnection): No longer needs a server name (ServerConnection::~ServerConnection): Only release the global context if there is one. (ServerConnection::serverConnectionDidDie): Stub for IWebScriptDebugListener (ServerConnection::QueryInterface): ditto (ServerConnection::AddRef): ditto (ServerConnection::Release): ditto (ServerConnection::didLoadMainResourceForDataSource): ditto (ServerConnection::didParseSource): ditto (ServerConnection::failedToParseSource): ditto (ServerConnection::didEnterCallFrame): ditto (ServerConnection::willExecuteStatement): ditto (ServerConnection::willLeaveCallFrame): ditto (ServerConnection::exceptionWasRaised): ditto
Drosera/win/ServerConnection.h: Stubbed out the IWebScriptDebugListener functions, and this class no longer needs to be instantiated with a server name.

win:

Reviewed by Sam, Steve and Darin.

Added stubs for what will be neede to let Drosera attach to the WebKit process and debug it.

Interfaces/IWebScriptCallFrame.idl: Added.
Interfaces/IWebScriptDebugListener.idl: Added.
Interfaces/WebKit.idl:
WebKit.vcproj/Interfaces.vcproj:
WebKit.vcproj/WebKitGUID.vcproj:

15:10 Changeset [27002] by eseidel

2007-10-24 Eric Seidel <eric@webkit.org>

Build fix for Gtk, no review.

kjs/collector.cpp: #include "context.h"

14:31 Changeset [27001] by eseidel

2007-10-24 Eric Seidel <eric@webkit.org>

Reviewed by Maciej.

Stop checking isOutOfMemory after every allocation, instead let the collector
notify all ExecStates if we ever hit this rare condition.

SunSpider claims this was a 2.2% speedup.

kjs/collector.cpp: (KJS::Collector::collect): (KJS::Collector::reportOutOfMemoryToAllInterpreters):
kjs/collector.h:
kjs/nodes.cpp: (KJS::TryNode::execute):

13:33 Changeset [27000] by hyatt

Don't try to gap fill transformed selections. Fix the invalidation of selection to use the clip to visible content
code path so that it works with multi-column layouts and transforms.

Reviewed by aroben

rendering/RenderBlock.cpp: (WebCore::RenderBlock::isSelectionRoot): (WebCore::RenderBlock::fillSelectionGaps):
rendering/RenderView.cpp: (WebCore::RenderView::setSelection):

13:25 Changeset [26999] by eseidel

2007-10-24 Eric Seidel <eric@webkit.org>

No review, I want to tickle the CIA bot to test changes,
and took this as an opportunity to clean up some old change logs.

ChangeLog-2005-08-23: update email address
ChangeLog-2006-12-31: fix spelling mistakes

11:45 Changeset [26998] by mitz

WebCore:

Reviewed by Oliver Hunt.

add "(anonymous)" to the renderName of anonymous table cells, rows and sections

rendering/RenderTableCell.h: (WebCore::RenderTableCell::renderName):
rendering/RenderTableRow.h: (WebCore::RenderTableRow::renderName):
rendering/RenderTableSection.h: (WebCore::RenderTableSection::renderName):

LayoutTests:

Reviewed by Oliver Hunt.

updated results to reflect which table parts are anonymous

11:41 Changeset [26997] by thatcher

Reviewed by Adam.

Bug 15476: DOM tree fully expands some elements when arrowing up
http://bugs.webkit.org/show_bug.cgi?id=15476

Add a new expandTreeElementsWhenArrowing option to TreeOutline. This option
is used only for the sidebar and search results.

page/inspector/inspector.js: Opt-in for the sidebar and search results.
page/inspector/treeoutline.js: Add expandTreeElementsWhenArrowing and pass this option to traverseNextTreeElement and traversePreviousTreeElement.

11:30 Changeset [26996] by bdash

2007-10-24 Mark Rowe <mrowe@apple.com>

Gtk build fix.

kjs/identifier.h: Remove extra qualification.

11:04 Changeset [26995] by ggaren

Reviewed by Sam Weinig.

Disable ALWAYS_INLINE in debug builds, since it drives the debugger
crazy.

wtf/AlwaysInline.h:

11:02 Changeset [26994] by ggaren

Reviewed by Sam Weinig.

Inlined the fast path for creating an Identifier from an Identifier.

This is a .4% speedup on SunSpider overall, but as big as a 2.5%
speedup on certain individual tests. 65% of the Identifiers creating
by SunSpider are already Identifiers.

(The main reason I'm making this change is that it resolves a large
regression in a patch I haven't checked in yet.)

JavaScriptCore.exp:
kjs/identifier.cpp: (KJS::Identifier::addSlowCase):
kjs/identifier.h: (KJS::Identifier::Identifier::add):

10:58 Changeset [26993] by andersca

Fix ChangeLog.

10:56 Changeset [26992] by andersca

Reviewed by Geoff and Mitz.

<rdar://problem/5493833>
REGRESSION: ~5MB of image data leaked @ cuteoverload.com (often seen while browsing other sites, too)

WebCore.xcodeproj/project.pbxproj:
bindings/js/kjs_binding.cpp: (KJS::ScriptInterpreter::markDOMNodesForDocument): If an image element that is currently loading an image is not in the document, it should still be marked.

bindings/js/kjs_html.cpp: (WebCore::ImageConstructorImp::construct): Force the document wrapper to be created.

html/HTMLImageElement.h: (WebCore::HTMLImageElement::haveFiredLoadEvent): New method which calls down to the image loader.

html/HTMLImageLoader.cpp: (WebCore::HTMLImageLoader::HTMLImageLoader): (WebCore::HTMLImageLoader::~HTMLImageLoader): (WebCore::HTMLImageLoader::setLoadingImage): (WebCore::HTMLImageLoader::dispatchLoadEvent): Remove code that's not needed anymore.

html/HTMLImageLoader.h: (WebCore::HTMLImageLoader::haveFiredLoadEvent): Make this public.

10:27 Changeset [26991] by aroben

Move Windows safe file creation code into WebCore from WebPreferences

WebCore:

Move Windows safe file creation code into WebCore from WebPreferences

Reviewed by Brady.

platform/FileSystem.h:
platform/win/FileSystemWin.cpp: (WebCore::safeCreateFile): Moved here from WebKit/win/WebPreferences.cpp.

WebKit/win:

Move safe file creation code to WebCore

Reviewed by Brady.

WebPreferences.cpp: (preferencesPath): Made into a static helper function. (WebPreferences::save): Now calls WebCore's safeCreateFile function. (WebPreferences::load): Uses String/CString to handle the UTF-8 conversion.
WebPreferences.h: Removed preferencesPath and safeCreateFileWithData.

10:27 Changeset [26990] by aroben

Reduce code duplication by using WebCore's FileSystem functions

WebCore:

Add methods to FileSystemWin to get some user profile directories

These directories are used to hold things like preferences, caches,
etc.

Reviewed by Brady.

platform/FileSystem.h: Added new method declarations for Windows only.
platform/win/FileSystemWin.cpp: (WebCore::bundleName): Added. (WebCore::storageDirectory): Added. (WebCore::cachedStorageDirectory): Added. (WebCore::localUserSpecificStorageDirectory): Added. Returns the directory where WebKit should store any user-specific data that should stay local to the current machine (i.e., shouldn't be stored in a roaming profile). (WebCore::roamingUserSpecificStorageDirectory): Added. Returns the directory where WebKit should store any user-specific data that should move with the user from machine to machine (i.e., should be stored in a roaming profile).

WebKit/win:

Reduce code duplication by using WebCore's FileSystem functions

Reviewed by Brady.

WebIconDatabase.cpp: Removed a now-unused function and a fixed FIXME. (WebIconDatabase::init): Changed to use FileSystem functions.
WebPreferences.cpp: (WebPreferences::preferencesPath): Ditto.

06:28 Changeset [26989] by bdash

2007-10-24 Mark Rowe <mrowe@apple.com>

Fix incorrect test results landed with r26986.

http/tests/xmlhttprequest/xmlhttprequest-InvalidStateException-getAllRequestHeaders-expected.txt:
http/tests/xmlhttprequest/xmlhttprequest-InvalidStateException-getRequestHeader-expected.txt:

06:21 Changeset [26988] by lars

update some test results for the Qt port.

06:06 Changeset [26987] by alp

2007-10-24 Alp Toker <alp@atoker.com>

Reviewed by Mark Rowe.

http://bugs.webkit.org/show_bug.cgi?id=15659
InlineTextBox does not setStrokeStyle() as needed

GraphicsContextCairo workaround to support a GraphicsContextCG quirk:

Save and restore the StrokeStyle manually.

platform/graphics/cairo/GraphicsContextCairo.cpp: (WebCore::GraphicsContext::drawLineForText):

05:49 Changeset [26986] by bdash

2007-10-24 Julien Chaffraix <julien.chaffraix@gmail.com>

Reviewed by Darin.

Patch for http://bugs.webkit.org/show_bug.cgi?id=15356
Bug 15356: getResponseHeader and getAllResponseHeaders do not throw exceptions

getAllResponseHeaders and getResponseHeader throws INVALID_STATE_ERR exception as specified in the specification

Add the check to field-name production in getResponseHeader (call to isValidToken) and returns an empty string if the header is not valid (Opera and IE behaviour)

Updated XMLHttpRequest::responseMIMEType() to keep its behaviour

Tests: http/tests/xmlhttprequest/xmlhttprequest-InvalidStateException-getAllRequestHeaders.html

http/tests/xmlhttprequest/xmlhttprequest-InvalidStateException-getRequestHeader.html
http/tests/xmlhttprequest/xmlhttprequest-invalidHeader-getRequestHeader.html

bindings/js/JSXMLHttpRequest.cpp: (KJS::JSXMLHttpRequestPrototypeFunction::callAsFunction):
xml/XMLHttpRequest.cpp: (WebCore::XMLHttpRequest::getAllResponseHeaders): (WebCore::XMLHttpRequest::getResponseHeader): (WebCore::XMLHttpRequest::responseMIMEType):
xml/XMLHttpRequest.h:

2007-10-24 Julien Chaffraix <julien.chaffraix@gmail.com>

Reviewed by Darin.

Tests for http://bugs.webkit.org/show_bug.cgi?id=15356
getResponseHeader and getAllResponseHeaders do not throw exceptions

http/tests/xmlhttprequest/xmlhttprequest-InvalidStateException-getAllRequestHeaders-expected.txt: Added.
http/tests/xmlhttprequest/xmlhttprequest-InvalidStateException-getAllRequestHeaders.html: Added.
http/tests/xmlhttprequest/xmlhttprequest-InvalidStateException-getRequestHeader-expected.txt: Added.
http/tests/xmlhttprequest/xmlhttprequest-InvalidStateException-getRequestHeader.html: Added.
http/tests/xmlhttprequest/xmlhttprequest-invalidHeader-getRequestHeader-expected.txt: Added.
http/tests/xmlhttprequest/xmlhttprequest-invalidHeader-getRequestHeader.html: Added.

05:33 Changeset [26985] by lars

these tests don't crash anymore :)

05:33 Changeset [26984] by lars

Stop crashing on fast/events/frame-tab-focus.html the keyEvent can be 0.

Signed-off-by: Lars Knoll <lars@trolltech.com>

05:33 Changeset [26983] by lars

implement most of DragData.

05:33 Changeset [26982] by lars

add a layoutTestController.dumpSelectionRect() dummy. It only has an effect on mac pixel tests anyway. Makes us pass another 5 test cases.

05:33 Changeset [26981] by lars

we're passing most editing tests now. Commit them.

05:33 Changeset [26980] by lars

add another command.

05:32 Changeset [26979] by lars

convert the text to be text only, and fix it up a little so it doesn't depend on specific font metrics anymore.

05:32 Changeset [26978] by lars

remove some notImplemented() warnings.

05:32 Changeset [26977] by lars

add empty eventSender.clearKillRing(), and map some of the special keys used in eventSender.keyDown() to things that work on Qt/X11.

05:32 Changeset [26976] by lars

remove a stupid notImplemented() warning and implement PasteBoad::clear().

05:32 Changeset [26975] by lars

allow paste from DOM so we pass more test cases.

05:32 Changeset [26974] by lars

implement Pasteboard::documentFragment(), and fix a mem leak.

05:32 Changeset [26973] by lars

implement eventSender.keyDown().

05:32 Changeset [26972] by lars

some changes to the way JS values are converted to Qt values in the script bindings. Added support for converting JS arrays into QStringList's.

05:32 Changeset [26971] by lars

Simplify the PlatformKeyEvent constructor. No need to have an extra boolean for isKeyUp in there, as the QKeyEvent has the information.

05:32 Changeset [26970] by lars

some smaller fixes to the editing support in DRT. Makes another few tests pass.

05:32 Changeset [26969] by lars

explicitly add some tests that currently cause DRT to hang.

05:32 Changeset [26968] by lars

first bit of implementation for the textinputcontroller.

05:32 Changeset [26967] by lars

implemented support for most editing shortcuts to make contentEditable usable.

05:32 Changeset [26966] by lars

no need to call setIsActive ourselves on the frame, as the focus controller does it for us.

05:32 Changeset [26965] by lars

make the man webpage believe it has focus, so the editing tests work correctly.

05:32 Changeset [26964] by lars

implement layoutTestController.dumpEditingCallbacks() correctly.

05:32 Changeset [26963] by lars

use the new QTextBoundaryFinder class in Qt.

05:32 Changeset [26962] by lars

Implement support for testing editing.

01:53 Changeset [26961] by oliver

Remove old relation method, replace with specialised LessThan and lessThenEq functions for a 0.5-0.6% improvement in SunSpider

Reviewed by Darin

01:50 Changeset [26960] by eseidel

2007-10-24 Eric Seidel <eric@webkit.org>

Reviewed by darin.

kjs/nodes.h: (KJS::ImmediateNumberNode::): Fix ASSERT correctness (and debug build!)

01:21 Changeset [26959] by darin

Reviewed by Eric.

kjs/object.cpp: (KJS::JSObject::defaultValue): Get rid of a little Identifier ref/deref for what SunSpider claims is a 0.4% speedup.

01:14 Changeset [26958] by darin

Reviewed by Maciej.

separate out the code to create a hash table the first time from the code to rehash

SunSpider claims this was a 0.7% speedup.

kjs/property_map.cpp: (KJS::PropertyMap::expand): Changed to call either createTable or rehash. (KJS::PropertyMap::createTable): Added. For the case where we had no table. (KJS::PropertyMap::rehash): Removed code needed only in the case where we had no table.
kjs/property_map.h: Added createTable.

01:06 Changeset [26957] by eseidel

2007-10-24 Eric Seidel <eric@webkit.org>

Reviewed by darin.

Add ImmediateNumberNode to hold a JSValue* instead of a double for numbers
which can be represented by JSImmediate.

kjs/grammar.y:
kjs/nodes.cpp: (KJS::NumberNode::evaluate): (KJS::ImmediateNumberNode::evaluate):
kjs/nodes.h: (KJS::Node::): (KJS::ImmediateNumberNode::):
kjs/nodes2string.cpp: (ImmediateNumberNode::streamTo):

01:03 Changeset [26956] by darin

JavaScriptCore:

Reviewed by Maciej.

http://bugs.webkit.org/show_bug.cgi?id=15657 change static hash tables to use powers of two for speed

Seems to give 0.7% SunSpider speedup.

kjs/create_hash_table: Updated to generate new format.
kjs/lookup.cpp: (KJS::keysMatch): Took out unneeded typecast. (KJS::findEntry): Updated to expect table type 3 -- changed the printf to a plain old assert. Replaced the modulus with a bit mask. (KJS::Lookup::findEntry): Get the hash directly, since we know identifiers already have computed their hash -- saves a branch. (KJS::Lookup::find): Ditto.
kjs/lookup.h: Changed attr from 2-byte value to one-byte value. Replaced hashSize with hashSizeMask.

WebCore:

Reviewed by Maciej.

http://bugs.webkit.org/show_bug.cgi?id=15657 change static hash tables to use powers of two for speed

bindings/scripts/CodeGeneratorJS.pm: Updated to generate new format.

00:59 Changeset [26955] by mjs

Reviewed by Darin.

remove KJS_CHECKEXCEPTIONs in places where exceptions can't happen for 0.6% SunSpider speedup

kjs/nodes.cpp: (KJS::DoWhileNode::execute): (KJS::WhileNode::execute): (KJS::ForNode::execute): (KJS::ForInNode::execute): (KJS::SourceElementsNode::execute):

00:41 Changeset [26954] by darin

Reviewed by Eric.

sunspider: Open the profile in shark after completing a run.
sunspider-compare-results: Tiny perl tweaks.

00:40 Changeset [26953] by hausmann

Fix pathByAppendingComponent for the Qt port.

10/23/07:

23:52 Changeset [26952] by hyatt

Make repainting work with transforms.

Reviewed by John Sullivan

rendering/RenderBox.cpp: (WebCore::RenderBox::computeAbsoluteRepaintRect):

23:41 Changeset [26951] by darin

Ignore files generated by the make-hosted script.

23:19 Changeset [26950] by darin

Reviewed by Maciej.

kjs/JSImmediate.h: (KJS::JSImmediate::getUInt32): Changed an && to an & for a 1% gain in SunSpider.

21:40 Changeset [26949] by alp

2007-10-23 Alp Toker <alp@atoker.com>

Reviewed by Maciej.

http://bugs.webkit.org/show_bug.cgi?id=14412
[GDK] Clipboard support

Initial clipboard implementation for the GTK+ port.

platform/gtk/ClipboardGtk.cpp: (WebCore::Editor::newGeneralClipboard):
platform/gtk/PasteboardGtk.cpp: (WebCore::Pasteboard::generalPasteboard): (WebCore::Pasteboard::Pasteboard): (WebCore::Pasteboard::~Pasteboard): (WebCore::Pasteboard::writeSelection): (WebCore::Pasteboard::writeImage): (WebCore::Pasteboard::clear): (WebCore::Pasteboard::plainText):
platform/gtk/TemporaryLinkStubs.cpp:

21:35 Changeset [26948] by thatcher

Reviewed by Adam.

Fixes the Network panel so that it refreshes as needed.

page/inspector/ConsolePanel.js: Call the base class show() and hide() first.
page/inspector/NetworkPanel.js: Add refreshNeeded and refreshIfNeeded back. Ditto.

20:47 Changeset [26947] by weinig

Check for null BSTR that can be passed to the UIDelegate methods
from javascript null and undefined. Fixes a failing test case on
Windows (fast/dom/Window/alert-undefined.html)

Reviewed by Eric Seidel.

DumpRenderTree/win/UIDelegate.cpp: (UIDelegate::runJavaScriptAlertPanelWithMessage): (UIDelegate::runJavaScriptConfirmPanelWithMessage): (UIDelegate::runJavaScriptTextInputPanelWithPrompt): (UIDelegate::webViewAddMessageToConsole):

19:15 Changeset [26946] by mitz

Rolled out r26941 because it of layout test failures.

19:01 Changeset [26945] by oliver

Reduce branching in implementations of some operator implementations, yielding 1.3% boost to SunSpider.

Reviewed by Maciej

kjs/nodes.cpp: (KJS::MultNode::evaluate): (KJS::DivNode::evaluate): (KJS::ModNode::evaluate): (KJS::add): (KJS::sub): (KJS::AddNode::evaluate): (KJS::SubNode::evaluate): (KJS::valueForReadModifyAssignment):
kjs/operations.cpp:
kjs/operations.h:

18:59 Changeset [26944] by justing

Fixing an accidental change made in r25484.

editing/pasteboard/quirks-mode-br-1.html:
platform/mac/editing/pasteboard/quirks-mode-br-1-expected.checksum:
platform/mac/editing/pasteboard/quirks-mode-br-1-expected.txt:

18:56 Changeset [26943] by justing

Reviewed by Oliver Hunt.

<rdar://problem/5544856>
REGRESSION: After typing 2-byte text, undo only undoes one keystroke at a time

platform/mac/editing/input/firstrectforcharacterrange-styled-expected.txt:
platform/mac/editing/input/text-input-controller-expected.txt:
platform/mac/editing/input/wrapped-line-char-rect-expected.txt:

18:53 Changeset [26942] by justing

Reviewed by Oliver Hunt.

<rdar://problem/5544856>
REGRESSION: After typing 2-byte text, undo only undoes one keystroke at a time

No layout test for now because I'm having trouble getting Undo during a layout
test to only undo the last Undo step, and I want to write a test that fails
without the code change.

editing/Editor.cpp: (WebCore::Editor::confirmComposition): When we replace the previous composition, delete it with a sub-command of the command used to insert the new composition, instead of with a separate command. No new code was added because insertText already deletes the current selection. (WebCore::Editor::setComposition): Ditto.

18:51 Changeset [26941] by mitz

Reviewed by Eric Seidel.

fix http://bugs.webkit.org/show_bug.cgi?id=15405 ASSERTION FAILED: d->m_view && !d->m_view->needsLayout() in Frame::Paint

Calling updateWidget() during attach() led to arbitrary (plugin and resource load delegate)
code execution under attach(). The fix is to use the mechanism that's already in place for
deferring updateWidget() until after layout.

html/HTMLEmbedElement.cpp: (WebCore::HTMLEmbedElement::attach): Replaced call to updateWidget() with call to updateWidgetSoon()
html/HTMLObjectElement.cpp: (WebCore::HTMLObjectElement::attach): Ditto.
manual-tests/paint-during-plugin-attach.html: Added.
rendering/RenderPartObject.cpp: (WebCore::RenderPartObject::updateWidgetSoon): Added this function that schedules the updateWidget() call for after the next layout.
rendering/RenderPartObject.h:

18:36 Changeset [26940] by weinig

Make the WebNotificationCenter work with null (wildcard) and specific
observed objects, matching NSNotificationCenter.

Removes the ObserverKey, ObserverHash, and ObserverKeyTraits as we now hash against the notification name only and check the object on notification posting.
Use OwnPtr for the WebNotificationCenterPrivate member variable.

Reviewed by Adam Roben.

WebNotificationCenter.cpp: (WebNotificationCenter::WebNotificationCenter): (WebNotificationCenter::~WebNotificationCenter): (WebNotificationCenter::postNotificationInternal): (WebNotificationCenter::addObserver): (WebNotificationCenter::postNotification): (WebNotificationCenter::postNotificationName): (WebNotificationCenter::removeObserver):
WebNotificationCenter.h:

18:33 Changeset [26939] by sfalken

New tag.

18:33 Changeset [26938] by sfalken

Bump versions for submit

18:30 Changeset [26937] by aroben

Fix the behavior of pathByAppendingComponent when path is empty

We now use the Windows Shell API function PathAppendW instead of
rolling our own broken implementation.

Reviewed by Anders.

platform/win/FileSystemWin.cpp: (WebCore::pathByAppendingComponent):

17:56 Changeset [26936] by mjs

Reviewed by Darin.

Don't print a massive pile of setenvs from tools that automatically build testkjs.

Scripts/build-testkjs:

17:40 Changeset [26935] by mjs

SunSpider:

Reviewed by Darin.

Don't hardcode my path to testkjs

sunspider-compare-results:

WebKitTools:

Reviewed by Darin.

add wrapper that finds the right copy of testkjs

Scripts/sunspider-compare-results: Added.

17:12 Changeset [26934] by adele

WebCore:

Reviewed by Darin.

Fix for <rdar://problem/5543228> CrashTracer: [USER] 2 crashes in Safari at com.apple.WebCore: WebCore::HTMLInputElement::defaultEventHandler + 872

Test: fast/forms/textfield-onchange-deletion.html

html/HTMLInputElement.cpp: (WebCore::HTMLInputElement::defaultEventHandler): Refetch the renderer since arbitrary JS code run during onchange can do anything, including destroying it.

LayoutTests:

Reviewed by Darin.

Test for <rdar://problem/5543228> CrashTracer: [USER] 2 crashes in Safari at com.apple.WebCore: WebCore::HTMLInputElement::defaultEventHandler + 872

fast/forms/textfield-onchange-deletion-expected.txt: Added.
fast/forms/textfield-onchange-deletion.html: Added.

17:08 Changeset [26933] by thatcher

Reviewed by Sam Weinig.

Made seperate files for the various classes in ResourcePanel.js.
Broke up ResourcePanel.js and created seperate panel classes for different resource types.
Moved View code down to the Panel base-class.
Reduced code duplication in DatabasePanel.js by sub-classing from the new ResourcePanel.

page/inspector/DatabasePanel.js:
page/inspector/DocumentPanel.js: Added.
page/inspector/FontPanel.js: Added.
page/inspector/ImagePanel.js: Added.
page/inspector/Panel.js: Added.
page/inspector/PropertiesSection.js: Added.
page/inspector/Resource.js:
page/inspector/ResourcePanel.js:
page/inspector/SidebarPane.js: Added.
page/inspector/SourcePanel.js: Added.
page/inspector/inspector.css:
page/inspector/inspector.html:
page/inspector/inspector.js:

16:47 Changeset [26932] by andersca

Correct the version #if check.

platform/sql/SQLDatabase.cpp: (WebCore::SQLDatabase::authorizerFunction):

16:40 Changeset [26931] by sfalken

Reviewed by Anders.

<rdar://5548217>: [NTS] Java 6 update 3 crashes Safari when loading a
java page

It is possible to load Mozilla's Java plugin instead of our own, which
can lead to a crash. The fix is to prefer plugins in our own Plugins
directory when searching for a plugin.

plugins/win/PluginDatabaseWin.cpp: (WebCore::safariPluginsPath): Return the path to our own Plugins directory (WebCore::PluginDatabaseWin::defaultPluginPaths): Use new method safariPluginsPath() (WebCore::PluginDatabaseWin::pluginForMIMEType): If the plugin's path is our Plugins path, return this plugin. Otherwise, continue scanning the list of plugins (WebCore::PluginDatabaseWin::pluginForExtension): Same
plugins/win/PluginPackageWin.h: (WebCore::PluginPackageWin::path): Return this plugin's path

16:30 Changeset [26930] by sfalken

Fix build.

plugins/win/PluginViewWin.cpp:

16:16 Changeset [26929] by honeycutt

2007-10-23 Jon Honeycutt <jhoneycutt@apple.com>

Reviewed by Anders.

<rdar://5548217>: [NTS] Java 6 update 3 crashes Safari when loading a
java page

It is possible to load Mozilla's Java plugin instead of our own, which
can lead to a crash. The fix is to prefer plugins in our own Plugins
directory when searching for a plugin.

plugins/win/PluginDatabaseWin.cpp: (WebCore::safariPluginsPath): Return the path to our own Plugins directory (WebCore::PluginDatabaseWin::defaultPluginPaths): Use new method safariPluginsPath() (WebCore::PluginDatabaseWin::pluginForMIMEType): If the plugin's path is our Plugins path, return this plugin. Otherwise, continue scanning the list of plugins (WebCore::PluginDatabaseWin::pluginForExtension): Same
plugins/win/PluginPackageWin.h: (WebCore::PluginPackageWin::path): Return this plugin's path

16:14 Changeset [26928] by adachan

<rdar://problem/5244261> SafariWin ignores cookie policy setting "never" in the preferences
Custom WebPreferences (not the shared WebPreferences) could override the cookie accept
policy setting on the default cookie storage. To fix that, I added a new method in
WebView called updateGlobalSettingsFromPreferences() to handle updating the global pref
options such as cookie accept policy, and it's only called to update changes from
the shared WebPreferences.

Use CLSID_WebPreferences and remove IID_WebPreferences.

Reviewed by Darin and Adam.

WebPreferences.cpp: (WebPreferences::QueryInterface):
WebPreferences.h:
WebView.cpp: (WebView::updateWebCoreSettingsFromPreferences): (WebView::updateGlobalSettingsFromPreferences): (WebView::updateSettingsFromPreferences): (WebView::developerExtrasEnabled): (WebView::initWithFrame): (WebView::onNotify):
WebView.h:

16:11 Changeset [26927] by kmccullo

Reviewed by Sam.

After talking with Steve I now see that the WebKit server must be running for DCOM to create the distributed objects, which makes sense and currently WebKit only allows one instance to be running at a time which avoids accidentally connecting to the wrong server
In light of this I have removed the code for the attach box and NotificationServer and known server names, since they are all extranious now.

Drosera/win/Drosera.cpp: Removed notification classes and known server names. (_tWinMain): Uses init instead of initUI. (droseraWndProc): No longer creats an attach dialog box. (Drosera::Drosera): Does the OleInitialize so COM is ready to go and it's not manditory to call init before doing COM stuff. (Drosera::init): calls initUI and will call attach when the functionality exists. (Drosera::initUI): Has changed very little. (Drosera::attach): Changed the signature to reflect that we no longer need the dictionary of known server names.
Drosera/win/Drosera.h: Removed notification classes and known server names. Also renamed and moved some functions.
Drosera/win/Drosera.vcproj/Drosera.rc: Removed the Attach box.
Drosera/win/Drosera.vcproj/Drosera.vcproj: Moved resource.h from the headers to the resources folder.
Drosera/win/resource.h: Removed the Attach box.

15:27 Changeset [26926] by mjs

Reviewed by Darin.

Add a compare script to compare results, and improve formatting of standard results a bit.

resources/sunspider-analyze-results.js:
resources/sunspider-compare-results.js: Added.
sunspider-compare-results: Added.

14:25 Changeset [26925] by oliver

Separating all of the simple binary expression nodes into multiple classes

Reviewed by Maciej

Separating all of the simple (eg. non-read-modify-write) binary operators into separate classes in preparation for further JS optimisations.

Happily this produces a 0.8% to 1.0% performance increase in SunSpider with no further work.

12:29 Changeset [26924] by alp

2007-10-23 Jasper Bryant-Greene <m@ni.ac.nz>

Reviewed by Darin.

http://bugs.webkit.org/show_bug.cgi?id=15058
Precedence fault in KeyEventGdk causes shift, control, alt etc to work incorrectly

Corrected precendence fault which was causing Shift, Alt, Control and
Meta to behave incorrectly on GTK.

Coding style fix by Alp.

platform/gtk/KeyEventGtk.cpp: (WebCore::PlatformKeyboardEvent::PlatformKeyboardEvent):

12:05 BuildingGtk edited by iwkse@claws-mail.org

(diff)

11:25 Changeset [26923] by aroben

Add some SVG tests to the Windows Skipped file

Reviewed by Oliver.

platform/win/Skipped:

11:11 Changeset [26922] by weinig

Reviewed by Tim Hatcher.

Add resizing of the DOM view right sidebar. This allows us to actually see the values now.

page/inspector/ResourcePanel.js:
page/inspector/inspector.css:
page/inspector/inspector.js:
page/inspector/utilities.js:

08:52 Changeset [26921] by hyatt

Get basic hit testing right for transforms.

Reviewed by ollliej

platform/graphics/AffineTransform.cpp: (WebCore::AffineTransform::mapPoint):
platform/graphics/AffineTransform.h:
rendering/InlineBox.cpp: (WebCore::InlineBox::nodeAtPoint):
rendering/RenderBlock.cpp: (WebCore::RenderBlock::nodeAtPoint):
rendering/RenderLayer.cpp: (WebCore::RenderLayer::paintLayer): (WebCore::RenderLayer::hitTest): (WebCore::RenderLayer::hitTestLayer):
rendering/RenderLayer.h:
rendering/RenderObject.cpp: (WebCore::RenderObject::hitTest):
rendering/RenderObject.h:

08:49 Changeset [26920] by darin

removed numeric results so the test is more platform-independent

fast/js/math-expected.txt: Updated.
fast/js/resources/math.js: Updated.

07:12 Changeset [26919] by darin

JavaScriptCore:

Reviewed by Maciej.

fix http://bugs.webkit.org/show_bug.cgi?id=15639 fix Math.abs(0), Math.ceil(-0), and Math.floor(-0)

Test: fast/js/math.html

kjs/math_object.cpp: (MathFuncImp::callAsFunction): Fix abs to look at the sign bit. Add a special case for values in the range between -0 and -1 and a special case for ceil and for -0 for floor.

LayoutTests:

Reviewed by Maciej.

test for http://bugs.webkit.org/show_bug.cgi?id=15639 Math functions

fast/js/math-expected.txt: Added.
fast/js/math.html: Added.
fast/js/resources/math.js: Added.

05:11 Changeset [26918] by bdash

2007-10-23 Mark Rowe <mrowe@apple.com>

Build fix for Eric's build fix in r26916.

MigrateHeaders.make:

02:44 Changeset [26917] by eseidel

2007-10-23 Eric Seidel <eric@webkit.org>

Reviewed by Mark Rowe.

Fix leaks seen on TOT from new font-face code.

ksvg2/svg/SVGFontFaceElement.cpp: (WebCore::SVGFontFaceElement::rebuildFontFace): use a stack allocated CSSProperty instead of mallocing

01:33 Changeset [26916] by eseidel

2007-10-23 Eric Seidel <eric@webkit.org>

Build fix only, no review.

Sacrifice three virgin hours upon the altar of the heathen XCode gods
and pray that our build troubles are finally over. :(

DerivedSources.make: add DOMSVGException.h
WebCore.xcodeproj/project.pbxproj: Remove DOMSVG*Interal.h files from "Copy Generated Headers" phase. Why? Who knows.

00:07 Changeset [26915] by mjs

Reviewed by Darin.

resources/sunspider-analyze-results.js: Do the error range properly, using the t-distribution instead of 1.96 (which was based on the normal distribution).

sunspider: Print results in a way that will be friendlier to the soon-to-come compare mode.

00:00 Changeset [26914] by darin

Reviewed by Eric.

streamline exception handling code for a >1% speed-up of SunSpider

kjs/nodes.cpp: Changed macros to use functions for everything that's not part of normal execution. We'll take function call overhead when propagating an exception or out of memory. (KJS::createOutOfMemoryCompletion): Added. (KJS::substitute): Use append instead of the relatively inefficient + operator. (KJS::Node::rethrowException): Added.
kjs/nodes.h: Added rethrowException.

10/22/07:

23:45 Changeset [26913] by sfalken

Fix build.

plugins/win/PluginViewWin.cpp:

23:44 Changeset [26912] by darin

Reviewed by Maciej.

fix http://bugs.webkit.org/show_bug.cgi?id=15636 some JavaScriptCore regression tests are failing due to numeric conversion

This should restore correctness and make speed better too, restoring some
of the optimization we lost in my last check-in.

kjs/JSImmediate.h: (KJS::JSImmediate::getTruncatedInt32): Added. Uses the range checking idiom I used in my patch yesterday. (KJS::JSImmediate::getTruncatedUInt32): Ditto.

kjs/internal.h: Removed getInt32 and added getTruncatedInt/UInt32.
kjs/internal.cpp: (KJS::NumberImp::getUInt32): Changed to always use double, since I can't find a way to write this more efficiently for float. (KJS::NumberImp::getTruncatedInt32): Added. (KJS::NumberImp::getTruncatedUInt32): Added.

kjs/value.h: Removed getInt32 and added getTruncatedInt/UInt32. (KJS::JSValue::getUInt32): (KJS::JSValue::getTruncatedInt32): Added. (KJS::JSValue::getTruncatedUInt32): Added. (KJS::JSValue::toInt32): Changed getInt32 call to getTruncatedInt32. (KJS::JSValue::toUInt32): Changed getUInt32 call to getTruncatedUInt32.
kjs/value.cpp: (KJS::JSCell::getTruncatedInt32): Added. (KJS::JSCell::getTruncatedUInt32): Added. (KJS::JSValue::toInteger): Changed getUInt32 call to getTruncatedInt32. (KJS::JSValue::toInt32SlowCase): Removed extra getInt32 call I accidentally had left in here. (KJS::JSValue::toUInt32SlowCase): Ditto. (KJS::JSValue::toUInt16): Changed getUInt32 call to getTruncatedUInt32.

JavaScriptCore.exp: Updated.

23:28 Changeset [26911] by eseidel

2007-10-22 Eric Seidel <eric@webkit.org>

Build fix only, no review.

WebCore.xcodeproj/project.pbxproj: re-add all DOMSVG headers to copy-files phase to make sure.

23:22 Changeset [26910] by eseidel

2007-10-22 Eric Seidel <eric@webkit.org>

Build fix only, no review.

WebCore.xcodeproj/project.pbxproj: make sure DOMSVGAnimateElement.h is copied into headers.

23:07 Changeset [26909] by eseidel

Touch SVGAnimateElement.idl to force to bots to rebuild *SVGAnimateElement.*

21:19 Changeset [26908] by mjs

Reviewed by Review.

add more tests (probably the final set for now)

Already balanced. Added date, regexp, control flow, and a few
more string and object/array access tests.

tests/LIST:
tests/access-fannkuch.js: Added.
tests/access-nbody.js: Added.
tests/controlflow-recursive.js: Added.
tests/date-format-tofte.js: Added.
tests/date-format-xparb.js: Added.
tests/regexp-dna.js: Added.
tests/string-validate-input.js: Added.

20:49 Changeset [26907] by eseidel

2007-10-22 Eric Seidel <eric@webkit.org>

Reviewed by Maciej.

Fix build by properly exposing SVG font-face dom bindings.

DerivedSources.make:
WebCore.vcproj/WebCore.vcproj: add font-face files to build
WebCore.xcodeproj/project.pbxproj: add missing DOM bindings files
bindings/js/JSSVGElementWrapperFactory.cpp:
bindings/objc/DOM.mm: (WebCore::createElementClassMap):
bindings/objc/DOMSVG.h:
ksvg2/svg/SVGDefinitionSrcElement.idl: inherit from SVGElement

19:44 Changeset [26906] by bdash

2007-10-22 Mark Rowe <mrowe@apple.com>

Gtk build fix.

WebCore.pro:

19:30 Changeset [26905] by darin

a first cut at fixing the Qt and GTK builds

WebCore.pro: Add new .idl and .cpp files.
DerivedSources.make: Re-sorted.

18:02 JavaScript performance improvement ideas edited by mjs@apple.com

(diff)

17:50 JavaScript performance improvement ideas edited by ggaren@apple.com

(diff)

17:38 Changeset [26904] by eseidel

2007-10-22 Eric Seidel <eric@webkit.org>

Reviewed by hyatt.

Implement <font-face> and friends for SVG.
http://bugs.webkit.org/show_bug.cgi?id=10652

DerivedSources.make:
WebCore.xcodeproj/project.pbxproj:
bindings/js/JSSVGElementWrapperFactory.cpp:
dom/Document.cpp: (WebCore::Document::mappedElementSheet): (WebCore::Document::recalcStyleSelector):
dom/Document.h:
ksvg2/scripts/make_names.pl:
ksvg2/svg/SVGDefinitionSrcElement.cpp: Added. (WebCore::SVGDefinitionSrcElement::SVGDefinitionSrcElement): (WebCore::SVGDefinitionSrcElement::~SVGDefinitionSrcElement): (WebCore::SVGDefinitionSrcElement::childrenChanged):
ksvg2/svg/SVGDefinitionSrcElement.h: Added.
ksvg2/svg/SVGDefinitionSrcElement.idl: Added.
ksvg2/svg/SVGFontFaceElement.cpp: Added. (WebCore::SVGFontFaceElement::SVGFontFaceElement): (WebCore::SVGFontFaceElement::~SVGFontFaceElement): (WebCore::cssPropertyIdForName): (WebCore::mapAttributeToCSSProperty): (WebCore::cssPropertyIdForSVGAttributeName): (WebCore::SVGFontFaceElement::parseMappedAttribute): (WebCore::SVGFontFaceElement::rebuildFontFace): (WebCore::SVGFontFaceElement::childrenChanged):
ksvg2/svg/SVGFontFaceElement.h: Added.
ksvg2/svg/SVGFontFaceElement.idl: Added.
ksvg2/svg/SVGFontFaceFormatElement.cpp: Added. (WebCore::SVGFontFaceFormatElement::SVGFontFaceFormatElement): (WebCore::SVGFontFaceFormatElement::~SVGFontFaceFormatElement): (WebCore::SVGFontFaceFormatElement::childrenChanged):
ksvg2/svg/SVGFontFaceFormatElement.h: Added.
ksvg2/svg/SVGFontFaceFormatElement.idl: Added.
ksvg2/svg/SVGFontFaceNameElement.cpp: Added. (WebCore::SVGFontFaceNameElement::SVGFontFaceNameElement): (WebCore::SVGFontFaceNameElement::~SVGFontFaceNameElement): (WebCore::SVGFontFaceNameElement::srcValue):
ksvg2/svg/SVGFontFaceNameElement.h: Added.
ksvg2/svg/SVGFontFaceNameElement.idl: Added.
ksvg2/svg/SVGFontFaceSrcElement.cpp: Added. (WebCore::SVGFontFaceSrcElement::SVGFontFaceSrcElement): (WebCore::SVGFontFaceSrcElement::~SVGFontFaceSrcElement): (WebCore::SVGFontFaceSrcElement::srcValue): (WebCore::SVGFontFaceSrcElement::childrenChanged):
ksvg2/svg/SVGFontFaceSrcElement.h: Added.
ksvg2/svg/SVGFontFaceSrcElement.idl: Added.
ksvg2/svg/SVGFontFaceUriElement.cpp: Added. (WebCore::SVGFontFaceUriElement::SVGFontFaceUriElement): (WebCore::SVGFontFaceUriElement::~SVGFontFaceUriElement): (WebCore::SVGFontFaceUriElement::srcValue): (WebCore::SVGFontFaceUriElement::childrenChanged):
ksvg2/svg/SVGFontFaceUriElement.h: Added.
ksvg2/svg/SVGFontFaceUriElement.idl: Added.
ksvg2/svg/svgtags.in:

17:33 Changeset [26903] by aroben

Windows build fix

Reviewed by NOBODY.

Scripts/build-dumprendertree: Fix path to DumpRenderTree.sln.

17:28 Changeset [26902] by sfalken

New tag.

17:28 Changeset [26901] by sfalken

Bump versions for submit

17:26 Changeset [26900] by sfalken

Merged fix from r26898.

16:36 Changeset [26899] by darin

JavaScriptCore:

Reviewed by Geoff.

fix http://bugs.webkit.org/show_bug.cgi?id=15632 js1_5/Array/array-001.js test failing

One of the JavaScriptCore tests was failing; it failed because of
my change to NumberImp::getUInt32. The incorrect code I copied was
from JSImmediate::getUInt32, and was a pre-existing bug.

This patch fixes correctness, but will surely slow down SunSpider.
We may be able to code this tighter and get the speed back.

kjs/JSImmediate.h: (KJS::JSImmediate::getInt32): Renamed from toInt32 to more accurately reflect the fact that this function only returns true if the value is accurate (no fractional part, etc.). Changed code so that it returns false when the value has a fraction. (KJS::JSImmediate::getUInt32): Ditto.

kjs/internal.cpp: (KJS::NumberImp::getInt32): Changed code so that it returns false when the value has a fraction. Restores the old behavior. (KJS::NumberImp::getUInt32): Ditto.

kjs/value.h: (KJS::JSValue::getInt32): Updated for name change. (KJS::JSValue::getUInt32): Ditto. (KJS::JSValue::toInt32): Ditto. (KJS::JSValue::toUInt32): Ditto.

LayoutTests:

Reviewed by Geoff.

tests for http://bugs.webkit.org/show_bug.cgi?id=15632

Added tests for cases where you use something that looks like an array
index, but it has a fractional part.

fast/js/kde/resources/Array.js: Added tests.
fast/js/kde/Array-expected.txt: Updated.

16:29 Changeset [26898] by honeycutt

2007-10-22 Jon Honeycutt <jhoneycutt@apple.com>

Reviewed by Anders.

<rdar://5548217>: [NTS] Java 6 update 3 crashes Safari when loading a
java page

We purposefully do not destroy our Java VM when its reference count
reaches 0, but we were unloading our JavaPlugin.dll when its reference
count reached 0, which lost the reference to the VM. This led to our
process trying to create a new VM the next time a page using Java was
loaded, and the JNI spec states that a single process is not allowed to
create more than one VM. The fix is to avoid unloading the Java plugin
via our PluginQuirkDontUnloadPlugin.

plugins/win/PluginViewWin.cpp: (WebCore::PluginViewWin::determineQuirks):

15:18 Changeset [26897] by darin

Reviewed by Brady.

fix crash seen when running JavaScriptCore tests

kjs/array_instance.cpp: (KJS::ArrayInstance::mark): Copy and paste error: I accidentally had code here that was making a copy of the HashMap -- that's illegal inside a mark function and was unnecessary. The other callsite was modifying the map as it iterated it, but this function is not.

14:32 Changeset [26896] by hyatt

Add results for new layout tests for transforms.

14:32 Changeset [26895] by hyatt

Add two new layout tests for transforms.

14:29 Changeset [26894] by hyatt

Fix for bug 15624, make transforms work properly with opacity.

Make sure a unitless 0 is allowed as an angle argument to rotation/skew.

Reviewed by Mitz Pettel

fast/transforms/transforms-with-opacity.html
fast/transforms/skew-with-unitless-zero.html

css/CSSParser.cpp: (WebCore::CSSParser::validUnit):
rendering/RenderLayer.cpp: (WebCore::transparencyClipBox): (WebCore::RenderLayer::beginTransparencyLayers): (WebCore::RenderLayer::paintLayer): (WebCore::RenderLayer::calculateClipRects):
rendering/RenderLayer.h: (WebCore::RenderLayer::transform):

14:26 Changeset [26893] by mjs

Reviewed by Oliver.

Avoid moving floats into integer registers in jsNumber() for 3% speedup on SunSpider http://bugs.webkit.org/show_bug.cgi?id=15627

kjs/JSImmediate.h: (KJS::JSImmediate::fromDouble): Avoid moving floats to integer registers since this is very slow.

11:39 Changeset [26892] by darin

Reviewed by Eric Seidel.

http://bugs.webkit.org/show_bug.cgi?id=15617 improve speed of integer conversions

Makes SunSpider 6% faster.

kjs/JSImmediate.h: Added toInt32 and toUInt32, with separate versions for 32-bit and 64-bit.
kjs/value.h: (KJS::JSValue::getUInt32): Call JSImmediate::toUInt32.

kjs/internal.h: Added getInt32.
kjs/internal.cpp: (KJS::NumberImp::getInt32): Added. (KJS::NumberImp::getUInt32): Replaced with more-optimal implementation stolen from JSValue.

kjs/value.h: (KJS::jsNumber): Marked ALWAYS_INLINE, because this wasn't getting inlined. (KJS::JSValue::getInt32): Added. (KJS::JSValue::getUInt32): Changed to call the new JSImmediate::toUInt32 to avoid converting from float to double. (KJS::JSValue::toInt32): Made inline, separated out the slow case. (KJS::JSValue::toUInt32): Ditto.
kjs/value.cpp: (KJS::JSCell::getInt32): Added. (KJS::JSValue::toInt32SlowCase): Renamed from toInt32. Changed to use the new getInt32. Added a faster case for in-range numbers. (KJS::JSValue::toUInt32SlowCase): Ditto. (KJS::JSValue::toUInt16): Added a faster case for in-range numbers.

JavaScriptCore.exp: Updated for changes.

11:02 Changeset [26891] by kmccullo

Removed a leak that was fixed.

Scripts/run-webkit-tests:

10:48 Changeset [26890] by kmccullo

Reviewed by Darin.

Changed the Client so that the DebuggerDocument now own the ServerConnection. This simplifies ownership and cleanup.

Drosera/win/DebuggerClient.cpp: The DebuggerDocument now owns the ServerConnection. (DebuggerClient::initWithServerName): (DebuggerClient::didFinishLoadForFrame):
Drosera/win/DebuggerClient.h: DebuggerDocument now owns the ServerConnection.
Drosera/win/Drosera.cpp: Moved some WebFrame initialization logic to until after we have a server which we are attached to. (Drosera::initUI): (Drosera::attach):
Drosera/win/Drosera.h: Removed two needless pointers I forgot to take out previously.

10:43 Changeset [26889] by aroben

Windows build fix

JavaScriptCore:

Windows build fix

Reviewed by NOBODY.

JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.vcproj: Turn off warning about implicit conversion to bool.

WebCore:

Windows build fix

Reviewed by NOBODY.

WebCore.vcproj/WebCore.vcproj: Copy header files from platform/sql.
page/Page.cpp: Touched to force the header files to be copied.

10:14 Changeset [26888] by zimmermann

Reviewed by Simon.

Move textPath related SVGChar data in it's own structure SVGCharOnPath.
Store a pointer to that datastructure inside SVGChar, instead of 4 floats & one bool.

Saves space for the common case. And allows me to add more data (another float) to support
glyph-orientation-* on textPath later on...

09:29 QtWebKit edited by hausmann@webkit.org

(diff)

09:21 BuildingQtOnWindows edited by hausmann@webkit.org

(diff)

09:01 Changeset [26887] by hausmann

When running build-testkjs make sure it's called through the perl interpreter, to fix the build for Qt/Windows.

08:06 Changeset [26886] by alp

2007-10-22 Alp Toker <alp@atoker.com>

Reviewed by Nikolas Zimmermann.

Implement more GraphicsContextCairo stubs.

Remove a hack "to work around no current point bug" that was breaking
canvas tests.

Fix warnings.

platform/graphics/cairo/CairoPath.h: (WebCore::CairoPath::CairoPath):
platform/graphics/cairo/GraphicsContextCairo.cpp: (WebCore::GraphicsContext::addInnerRoundedRectClip): (WebCore::GraphicsContext::strokeRect):
platform/graphics/cairo/PathCairo.cpp: (WebCore::Path::isEmpty):
platform/graphics/svg/cairo/SVGPaintServerGradientCairo.cpp: (WebCore::SVGPaintServerGradient::setup):
platform/graphics/svg/cairo/SVGPaintServerSolidCairo.cpp: (WebCore::SVGPaintServerSolid::setup):

07:54 Changeset [26885] by hausmann

Wrap WEXITSTATUS with a little exitStatus() helper function that falls back to the use of $returnvalue >> 8 on Windows to determine the exit status on platforms without WEXITSTATUS.

07:53 Changeset [26884] by hausmann

Fix compilation from a clean build with the database feature disabled.
JSCustomVersionChangeCallback.cpp doesn't actually need the SQL header file, just ScriptInterpreter.

06:56 Changeset [26883] by bdash

2007-10-22 Mark Rowe <mrowe@apple.com>

Gtk build fix.

kjs/array_instance.cpp:

06:46 Changeset [26882] by darin

fast/js/kde/resources/Array.js: Added tests to cover missing value behavior (not the same as undefined values in arrays). This matches the ECMA JavaScript specification, but doesn't exactly match Firefox, because Firefox incorrectly inserts undefined values rather than missing values in array literals with extra commas.
fast/js/kde/Array-expected.txt: Updated with results.

06:35 Changeset [26881] by darin

JavaScriptCore:

Reviewed by Maciej.

http://bugs.webkit.org/show_bug.cgi?id=15606 make cut-off for sparse vs. dense arrays smarter for speed with large arrays

Makes the morph test in SunSpider 26% faster, and the overall
benchmark 3% faster.

This also fixes some small problems we had with the distinction
between nonexistent and undefined values in arrays.

kjs/array_instance.h: Tweaked formatting and naming.
kjs/array_instance.cpp: Copied from kjs/array_object.cpp. (KJS::storageSize): Added. Computes the size of the storage given a vector length. (KJS::increasedVectorLength): Added. Implements the rule for resizing the vector. (KJS::isDenseEnoughForVector): Added. (KJS::ArrayInstance::ArrayInstance): Initialize the new fields. (KJS::ArrayInstance::~ArrayInstance): Since m_storage is now never 0, delete it. (KJS::ArrayInstance::getItem): Updated for name changes. (KJS::ArrayInstance::lengthGetter): Ditto. (KJS::ArrayInstance::inlineGetOwnPropertySlot): Added. Allows both versions of getOwnPropertySlot to share more code. (KJS::ArrayInstance::getOwnPropertySlot): Just refactored, no code change. (KJS::ArrayInstance::put): Added logic for extending the vector as long as the array is dense enough. Also keep m_numValuesInVector up to date. (KJS::ArrayInstance::deleteProperty): Added code to keep m_numValuesInVector up to date. (KJS::ArrayInstance::getPropertyNames): Fixed bug where this would omit names for array indices with undefined values. (KJS::ArrayInstance::increaseVectorLength): Renamed from resizeStorage. Also simplified to only handle getting larger. (KJS::ArrayInstance::setLength): Added code to update m_numValuesInVector, to zero out the unused part of the vector and to delete the map if it's no longer needed. (KJS::ArrayInstance::mark): Tweaked formatting. (KJS::compareByStringForQSort): Ditto. (KJS::ArrayInstance::sort): Ditto. (KJS::CompareWithCompareFunctionArguments::CompareWithCompareFunctionArguments): Ditto. (KJS::compareWithCompareFunctionForQSort): Ditto. (KJS::ArrayInstance::compactForSorting): Fixed bug where this would turn undefined values into nonexistent values in some cases.

kjs/array_object.h: Removed MAX_ARRAY_INDEX.
kjs/array_object.cpp: Removed ArrayInstance. Moved to a separate file.

JavaScriptCore.pri: Added array_instance.cpp.
JavaScriptCore.xcodeproj/project.pbxproj: Ditto.
kjs/AllInOneFile.cpp: Ditto.

LayoutTests:

fast/js/kde/resources/Array.js: Added tests to cover missing value behavior (not the same as undefined values in arrays). This matches the ECMA JavaScript specification, but doesn't exactly match Firefox.
fast/js/kde/Array-expected.txt: Updated with results.

05:56 Changeset [26880] by andrew

JavaScriptCore:

Reviewed by Mark Rowe.

Fix for local database support after r26879
Ensure that ENABLE_DATABASE and ENABLE_ICONDATABASE are correctly set

Configurations/JavaScriptCore.xcconfig:

WebCore:

Reviewed by Mark Rowe.

Fix for local database support after r26879
Ensure that ENABLE_DATABASE and ENABLE_ICONDATABASE are correctly set

Configurations/WebCore.xcconfig:
WebCore.pro:
WebCore.vcproj/build-generated-files.sh:

WebKit:

Reviewed by Mark Rowe.

Fix for local database support after r26879
Ensure that ENABLE_DATABASE and ENABLE_ICONDATABASE are correctly set

Configurations/WebKit.xcconfig:

WebKitTools:

Reviewed by Mark Rowe.

Fix for local database support after r26879
Ensure that ENABLE_DATABASE and ENABLE_ICONDATABASE are correctly set

Scripts/build-webkit:

04:39 Changeset [26879] by hausmann

Build fix for the non-qmake builds.

04:28 Changeset [26878] by hausmann

Disable the database feature in the qmake build for now.

04:22 Changeset [26877] by hausmann

Make disabling the database feature (ENABLE_DATABASE=0) work by
placing various #ifdefs into the code and making the compilation of
some files optional.

03:41 Changeset [26876] by alp

2007-10-22 Alp Toker <alp@atoker.com>

Reviewed by Mark Rowe.

http://bugs.webkit.org/show_bug.cgi?id=15611
[GTK] Text selection behaviour different in Debug and Release builds

http://bugs.webkit.org/show_bug.cgi?id=15578
[GTK] Text editor caret does not blink

Never allow control to reach the end of non-void functions.

Return more sensible values, or in some cases, nulls.

02:35 Changeset [26875] by hausmann

Implement ResourceHandle::loadResourceSynchronously using ResourceHandle and a special ResourceHandleClient. This approach has the possible danger of reentrancy.

Signed-off-by: Simon Hausmann <hausmann@kde.org>

02:33 Changeset [26874] by hausmann

Do not build testkjs as an application bundle. This is

needed for run-javascriptcore-tests on OSX.

Also, based on r26633, allow to test the WebKit/Qt port on OSX.
Set DYLD_LIBRARY_PATH if it was set in the environment. It must be set

as we do not have -rpath on OSX.

Signed-off-by: Simon Hausmann <hausmann@kde.org>

02:24 Changeset [26873] by hausmann