Timeline
Sep 30, 2007:
- 6:54 PM Changeset in webkit [25813] by
-
- 5 edits in trunk/WebCore
Reviewed by Maciej.
- tone down the assertion I asked Harrison to include in his fix for <rdar://problem/5511128>; it's OK to re-ref and deref the document as long as you do so after the children are done being destroyed
No effect on release builds. Assertion change only.
Besides the changes listed below, renamed m_hasDeleted flag to
m_deletionHasBegun.
- dom/ContainerNode.cpp: (WebCore::ContainerNode::removeAllChildren): Added code to set the m_deletionHasBegun flag and some assertions that test its state.
- dom/Document.h: Removed m_hasDeleted -- we now use m_deletionHasBegun in the base class TreeShared.
- dom/Document.cpp: (WebCore::Document::Document): Removed initialization of m_hasDeleted. (WebCore::Document::removedLastRef): Added code to clear m_inRemovedLastRefFunction if we end up deciding not to delete this.
- platform/Shared.h: (WebCore::TreeShared::TreeShared): Added m_deletionHasBegun in addition to m_inRemovedLastRefFunction (formerly named m_hasRemovedLastRef). (WebCore::TreeShared::~TreeShared): Assert that m_deletionHasBegun is true. (WebCore::TreeShared::ref): Assert neither flag is true. (WebCore::TreeShared::deref): Ditto. (WebCore::TreeShared::hasOneRef): Ditto. (WebCore::TreeShared::removedLastRef): Made private. Added code to set m_hasDeleted to true. Also removed cast; since this class template has a virtual destructor, we don't need to cast before calling delete.
- 2:08 PM Web Inspector edited by
- Point them to the right plist on Windows. (diff)
- 1:13 PM Changeset in webkit [25812] by
-
- 4 edits in trunk
Qt build fix - Mac OS X specific
- 9:03 AM Changeset in webkit [25811] by
-
- 3 edits2 adds in branches/feature-branch
2007-09-30 Julien <julien.chaffraix@gmail.com>
Reviewed by aroben & eseidel.
getAllResponseHeaders() separates each header with a CR LF according to the specification
Test: http/tests/xmlhttprequest/xmlhttprequest-crlf-getAllResponseHeader.html
- xml/XMLHttpRequest.cpp: (WebCore::XMLHttpRequest::getAllResponseHeaders):
Sep 29, 2007:
- 6:58 PM Changeset in webkit [25810] by
-
- 3 edits in trunk/WebKit/win
Clean-up in preparation for <rdar://5505062> WebKit's version-independent ProgIDs contain version numbers
Reviewed by Steve.
- ProgIDMacros.h: Stringify the passed in class name.
- WebKitDLL.cpp: Added a FOR_EACH_CLASS macro to ensure that our class lists stay in the right order, and updated uses of the *_PROGID macros for to take the stringification into account.
- 9:28 AM Changeset in webkit [25809] by
-
- 3 edits10 adds in branches/feature-branch
2007-09-29 Eric Seidel <eric@webkit.org>
Reviewed by Alexey Proskuryakov.
SVG Subresources will not be saved when creating WebArchives
http://bugs.webkit.org/show_bug.cgi?id=15280
Implement _subresourceURLs methods for more SVGElement types
more such methods will be needed as we add support for other
external references (such as use, mpath, tref, etc.)
- DOM/WebDOMOperations.mm: (-[DOMSVGScriptElement _subresourceURLs]): added. (-[DOMSVGCursorElement _subresourceURLs]): added. (-[DOMSVGFEImageElement _subresourceURLs]): added.
- 8:34 AM Changeset in webkit [25808] by
-
- 4 edits11 deletes in trunk
Bakefiles:
2007-09-29 Holger Hans Peter Freyther <zecke@selfish.org>
Reviewed by Mark.
-Fix http://bugs.webkit.org/show_bug.cgi?id=13226.
Remove Bakefiles from svn.
- Bakefiles/Bakefiles.bkgen: Removed.
- Bakefiles/ChangeLog: Removed.
- Bakefiles/Readme.txt: Removed.
- Bakefiles/presets.bkl: Removed.
- Bakefiles/update-file-lists.py: Removed.
JavaScriptCore:
2007-09-29 Holger Hans Peter Freyther <zecke@selfish.org>
Reviewed by Mark.
-Fix http://bugs.webkit.org/show_bug.cgi?id=13226.
Remove Bakefiles from svn.
- JavaScriptCoreSources.bkl: Removed.
- jscore.bkl: Removed.
WebCore:
2007-09-29 Holger Hans Peter Freyther <zecke@selfish.org>
Reviewed by Mark.
-Fix http://bugs.webkit.org/show_bug.cgi?id=13226.
Remove Bakefiles from svn.
- Projects/gdk/webcore-gdk.bkl: Removed.
- Projects/wx/webcore-wx.bkl: Removed.
- WebCoreSources.bkl: Removed.
- webcore-base.bkl: Removed.
WebKitTools
2007-09-29 Holger Hans Peter Freyther <zecke@selfish.org>
Reviewed by Mark.
-Fix http://bugs.webkit.org/show_bug.cgi?id=13226.
Remove Bakefiles from svn.
- GtkLauncher/Bakefiles.bkgen: Removed.
- GtkLauncher/ENV: Removed.
- GtkLauncher/gdklauncher.bkl: Removed.
- GtkLauncher/mk: Removed.
- Scripts/regenerate-makefiles: Removed.
- 8:30 AM Changeset in webkit [25807] by
-
- 2 edits in trunk/WebCore
2007-09-28 Jan Michael Alonzo <jmalonzo@unpluggable.com>
Reviewed by Mark.
- platform/gtk/RenderThemeGtk.cpp:
(WebCore::RenderThemeGtk::determineState):
- Apply state if control is readonly
- Added state GTK_STATE_SELECTED of object is checked
- Apply GTK_STATE_ACTIVE if RenderObject isFocused()
(WebCore::RenderThemeGtk::paintTextField):
(WebCore::RenderThemeGtk::gtkEntry):
- Implemented theme-aware text field based on gtk/gtkentry.c implementation
- 8:16 AM Changeset in webkit [25806] by
-
- 2 edits in trunk/WebCore
2007-09-29 Holger Hans Peter Freyther <zecke@selfish.org>
Reviewed by Eric.
Build fix for Mac Gtk+ and Qt builds. In revision 25798 changing
XSLTProcessor.cpp was forgotten.
- xml/XSLTProcessor.cpp: Use PLATFORM(MAC) instead of PLATFORM(DARWIN)
- 8:15 AM Changeset in webkit [25805] by
-
- 3 edits3 adds in branches/feature-branch
2007-09-29 Eric Seidel <eric@webkit.org>
Reviewed by Alexey Proskuryakov.
WebArchives do not embed stylesheets referenced by xml-stylesheeet
http://bugs.webkit.org/show_bug.cgi?id=15320
- DOM/WebDOMOperations.mm: (-[DOMProcessingInstruction _stylesheetURL]): needed to access [[self sheet] href] (-[DOMProcessingInstruction _subresourceURLs]): call and return _stylesheetURL
- 1:13 AM Changeset in webkit [25804] by
-
- 5 edits in branches/feature-branch
Reviewed by Adam.
http://bugs.webkit.org/show_bug.cgi?id=13472
Misparsing date in javascript leads to year value of -1
http://bugs.webkit.org/show_bug.cgi?id=14176
Some date values not handled consistently with IE/Firefox
Allow an optional comma between month and year, and year and time.
Sep 28, 2007:
- 8:58 PM Changeset in webkit [25803] by
-
- 2 edits in trunk/WebCore
Fix <rdar://4965715> Horizontal scrollwheel scrolling moves in opposite direction on Vista
Reviewed by Sam.
No regression test possible.
- platform/win/WheelEventWin.cpp: (WebCore::PlatformWheelEvent::PlatformWheelEvent): Use a negative delta when scrolling right to match EventHandler's expectations (which are based on AppKit).
- 8:26 PM Changeset in webkit [25802] by
-
- 3 edits2 adds in branches/feature-branch
2007-09-28 Eric Seidel <eric@webkit.org>
Reviewed by Sam Wenig.
ASSERT when dragging <svg:image> with non-null prefix
http://bugs.webkit.org/show_bug.cgi?id=15308
- bindings/objc/DOM.mm: (WebCore::lookupElementClass): added (WebCore::elementClass): use new lookupElementClass function
- 5:38 PM Changeset in webkit [25801] by
-
- 4 edits in trunk
Versioning.
- 5:37 PM Changeset in webkit [25800] by
-
- 1 copy in tags/Safari-4523.10
New tag.
- 5:36 PM Changeset in webkit [25799] by
-
- 1 copy in tags/Safari-5523.10
New tag.
- 2:56 PM Changeset in webkit [25798] by
-
- 3 edits in trunk/WebCore
Reviewed by Mark Rowe
Build fix for Mac Gtk+ and Qt builds.
- xml/XSLStyleSheet.cpp: Use PLATFORM(MAC) instead of PLATFORM(DARWIN)
- xml/XSLTExtensions.cpp: Ditto.
- 2:49 PM Changeset in webkit [25797] by
-
- 6 edits in trunk/WebCore
Reviewed by Darin Adler.
<rdar://problem/5511128> Crash closing or reloading this SVG
- dom/ContainerNode.cpp: (WebCore::dispatchChildInsertionEvents): (WebCore::dispatchChildRemovalEvents): Use DocPtr instead of RefPtr, since these events are dispatched when the Document is being being torn down by removedLastRef().
- dom/DocPtr.h: (WebCore::DocPtr::DocPtr): Fix longstanding typo in template so that the DocPtr(DocPtr) constructor can be used.
- dom/Document.cpp: (WebCore::Document::Document): (WebCore::Document::removedLastRef):
- dom/Document.h: (WebCore::Document::selfOnlyRef): (WebCore::Document::selfOnlyDeref):
- platform/Shared.h: (WebCore::Shared::Shared): (WebCore::Shared::ref): (WebCore::Shared::deref): (WebCore::Shared::hasOneRef): (WebCore::TreeShared::TreeShared): (WebCore::TreeShared::ref): (WebCore::TreeShared::deref): (WebCore::TreeShared::hasOneRef): (WebCore::TreeShared::refCount): Add debug-only checks for a document being ref-counted while being deleted.
- 11:31 AM Changeset in webkit [25796] by
-
- 8 edits2 adds in trunk
WebCore:
Reviewed by Darin.
- <rdar://5261371> Nothing downloaded when exporting bookmarks from iGoogle web history
- Implemented IWebHTTPURLResponse::allHeaderFields so that if the content disposition is "attachment" we will download the file instead of display it. Also implemented some missing functionality.
- platform/network/ResourceResponse.cpp: (WebCore::ResourceResponse::isAttachment):
- platform/network/ResourceResponse.h:
win:
Reviewed by Darin.
- <rdar://5261371> Nothing downloaded when exporting bookmarks from iGoogle web history
- Implemented IWebHTTPURLResponse::allHeaderFields so that if the content disposition is "attachment" we will download the file instead of display it. Also implemented some missing functionality.
- HTTPHeaderPropertyBag.cpp: Added. (HTTPHeaderPropertyBag::HTTPHeaderPropertyBag): (HTTPHeaderPropertyBag::createInstance): (HTTPHeaderPropertyBag::setResponse): (HTTPHeaderPropertyBag::response): (HTTPHeaderPropertyBag::QueryInterface): (HTTPHeaderPropertyBag::AddRef): (HTTPHeaderPropertyBag::Release): (ConvertFromLPCOLESTR): (ConvertToVariant): (ConvertFromVariant): (HTTPHeaderPropertyBag::Read): (HTTPHeaderPropertyBag::Write):
- HTTPHeaderPropertyBag.h: Added.
- Interfaces/IWebHTTPURLResponse.idl:
- WebKit.vcproj/WebKit.vcproj:
- WebURLResponse.cpp: (WebURLResponse::allHeaderFields): (WebURLResponse::isAttachment):
- WebURLResponse.h:
- 8:37 AM Changeset in webkit [25795] by
-
- 3 edits in trunk/WebCore
2007-09-26 Holger Hans Peter Freyther <zecke@selfish.org>
Reviewed by Anders.
Implement PopupMenu using GtkMenu.
- platform/PopupMenu.h:
- platform/gtk/PopupMenuGtk.cpp: (WebCore::PopupMenu::PopupMenu): (WebCore::PopupMenu::~PopupMenu): (WebCore::PopupMenu::show): (WebCore::PopupMenu::hide): (WebCore::PopupMenu::updateFromElement): (WebCore::PopupMenu::itemWritingDirectionIsNatural): (WebCore::PopupMenu::menuItemActivated): (WebCore::PopupMenu::menuUnmapped): (WebCore::PopupMenu::menuPositionFunction): (WebCore::PopupMenu::menuRemoveItem):
- 8:36 AM Changeset in webkit [25794] by
-
- 2 edits in trunk/WebCore
2007-09-23 Holger Hans Peter Freyther <zecke@selfish.org>
Rubber stamped by Mark.
Obey the Coding-Style and move the asterisk to the
left to be part of the type.
- platform/gtk/RenderThemeGtk.h:
- 6:45 AM Changeset in webkit [25793] by
-
- 9 edits in trunk/LayoutTests
Reviewed by bdash.
Use frames instead windows for these tests. Closing windows reliably is difficult in DRT.
- http/tests/security/resources/has-custom-property.html:
- http/tests/security/resources/no-custom-property.html:
- http/tests/security/window-properties-clear-domain-expected.txt:
- http/tests/security/window-properties-clear-domain.html:
- http/tests/security/window-properties-clear-port-expected.txt:
- http/tests/security/window-properties-clear-port.html:
- http/tests/security/window-properties-pass-expected.txt:
- http/tests/security/window-properties-pass.html:
Sep 27, 2007:
- 10:56 PM Changeset in webkit [25792] by
-
- 2 edits in trunk/WebKit
Reviewed by Ollie
- fixed <rdar://problem/5408186> REGRESSION (5522-5523.9): Safari leaks every browser window
The leak started occurring when we removed the code to clear the delegates and the host window
from Safari as part of the fix for 5479443. But it turns out that Safari code was masking a
bug here in WebView: setHostWindow:nil needs to be called before setting _private->closed to
YES, or it will do nothing at all, causing a world leak due to a circular reference between
the window and the WebView.
I toyed with a more complex fix, but this is the simplest one that retains the fix for 5479443
while otherwise restoring the code order to be as close as possible to what it was before
5479443 was fixed.
- WebView/WebView.mm: (-[WebView _close]): Moved the call that sets _private->closed to YES to be after the code that clears the delegates and the host window. Added a comment about this order.
- 10:50 PM Changeset in webkit [25791] by
-
- 1 edit1 add in trunk/JavaScriptCore/JavaScriptCore.vcproj
Internal build fix
- 10:20 PM Changeset in webkit [25790] by
-
- 1 edit in trunk/JavaScriptCore/JavaScriptCore.vcproj/JavaScriptCore.make
Internal build fix
- 6:28 PM Changeset in webkit [25789] by
-
- 4 edits in trunk/LayoutTests
Reviewed by NOBODY.
Update tests to hopefully fix occasional failures
- fast/dom/Window/window-early-properties.html:
- http/tests/security/resources/has-custom-property.html:
- http/tests/security/resources/no-custom-property.html:
- 6:01 PM Changeset in webkit [25788] by
-
- 12 edits in trunk
2007-09-27 Mitz Pettel <mitz@webkit.org>
Reviewed by Dave Hyatt.
- fix http://bugs.webkit.org/show_bug.cgi?id=15262 REGRESSION: Margin-top:18px is pushing down the multiple select list box outside the div.
Reflected in existing tests.
- css/html4.css: Changed listboxes to use border-box sizing, to match Firefox, WinIE and apparently also Safari 2.
2007-09-27 Mitz Pettel <mitz@webkit.org>
Reviewed by Dave Hyatt.
- updated tests and results for http://bugs.webkit.org/show_bug.cgi?id=15262 REGRESSION: Margin-top:18px is pushing down the multiple select list box outside the div.
- fast/forms/listbox-clip-expected.checksum:
- fast/forms/listbox-clip-expected.png:
- fast/forms/listbox-clip-expected.txt:
- fast/forms/listbox-selection-2.html:
- fast/forms/select-item-background-clip.html:
- fast/forms/select-list-box-with-height-expected.checksum:
- fast/forms/select-list-box-with-height-expected.png:
- fast/forms/select-list-box-with-height-expected.txt:
- fast/text/drawBidiText.html:
- 5:06 PM Changeset in webkit [25787] by
-
- 1 edit in trunk/WebKit/ChangeLog
Tweaked ChangeLog comment..
- 5:04 PM Changeset in webkit [25786] by
-
- 2 edits1 add in trunk/JavaScriptCore
Rubber stamped by John Sullivan.
<rdar://problem/5493093>
- JavaScriptCore.order: Added.
- JavaScriptCore.xcodeproj/project.pbxproj: We're changing from using an order file built by another team to using one we actually check into our project repository. Linker settings for Symbol Ordering Flags have been updated accordingly.
- 5:03 PM Changeset in webkit [25785] by
-
- 2 edits1 add in trunk/WebCore
<rdar://problem/5493093>
Rubber stamped by John Sullivan.
- WebCore.order: Added.
- WebCore.xcodeproj/project.pbxproj: We're changing from using an order file built by another team to using one we actually check into our project repository. Linker settings for Symbol Ordering Flags have been updated accordingly.
- 5:01 PM Changeset in webkit [25784] by
-
- 3 edits in trunk/WebKit/win
<rdar://problem/5507481> Added a delegate method to inform the client
app that WebView has painted.
Reviewed by Steve.
- Interfaces/IWebUIDelegatePrivate.idl:
- WebView.cpp: (WebView::updateBackingStore):
- 4:56 PM Changeset in webkit [25783] by
-
- 8 edits10 adds in trunk
WebCore:
Reviewed by Geoff.
Fix for <rdar://problem/5499125>
REGRESSION (r21367): After launching Kidzui beta, a webview frame fails to appear in its main window
If window object properties were modified when it had initial empty document (synchronously after
window.open() for example) those modifications were lost when the real document was switched in.
Match Firefox behavior where window properties are not cleared if the inital document and the loaded one
have matching security domains.
Tests: fast/dom/Window/window-early-properties.html
http/tests/security/window-properties-clear-domain.html
http/tests/security/window-properties-clear-port.html
http/tests/security/window-properties-pass.html
- WebCore.exp:
- loader/FrameLoader.cpp: (WebCore::FrameLoader::FrameLoader): (WebCore::FrameLoader::init): (WebCore::FrameLoader::clear): (WebCore::FrameLoader::isSecureTransition): (WebCore::FrameLoader::begin):
- loader/FrameLoader.h: If we are transitioning from initial empty document to the final one, do a domain security check between old security policy URL and new URL. If that passes don't clear script proxy and script objects.
- bindings/js/kjs_proxy.cpp: (WebCore::KJSProxy::updateDocumentWrapper):
- bindings/js/kjs_proxy.h:
- page/Frame.cpp: (WebCore::Frame::setDocument): Since we don't always clear window properties anymore, we need to update the document property to point to the newly created one.
LayoutTests:
Reviewed by Geoff.
Tests for <rdar://problem/5499125>
REGRESSION (r21359-21368): After launching Kidzui beta, a webview frame fails to appear in its main window
- fast/dom/Window/window-early-properties-expected.txt: Added.
- fast/dom/Window/window-early-properties.html: Added.
- http/tests/security/resources/has-custom-property.html: Added.
- http/tests/security/resources/no-custom-property.html: Added.
- http/tests/security/window-properties-clear-domain-expected.txt: Added.
- http/tests/security/window-properties-clear-domain.html: Added.
- http/tests/security/window-properties-clear-port-expected.txt: Added.
- http/tests/security/window-properties-clear-port.html: Added.
- http/tests/security/window-properties-pass-expected.txt: Added.
- http/tests/security/window-properties-pass.html: Added.
- 4:53 PM Changeset in webkit [25782] by
-
- 2 edits1 add in trunk/WebKit
Rubber stamped by Darin.
<rdar://problem/5493093>
- WebKit.order: Added.
- WebKit.xcodeproj/project.pbxproj: Altered build settings to use the new order file project instead of AppleInternal.
- 4:18 PM Changeset in webkit [25781] by
-
- 6 edits in trunk/WebCore
Reviewed by Darin.
<rdar://problem/5493093>
- WebCore.xcodeproj/project.pbxproj: Removed libxslt from statically linked frameworks.
- platform/mac/SoftLinking.h: Formatting tweakes, added SOFT_LINK_LIBRARY, and if'def for PLATFORM(DARWIN)
- xml/XSLStyleSheet.cpp: Soft-linked libxslt.
- xml/XSLTExtensions.cpp: Removed the call to xsltGenericError, and added a FIXME comment stating that it might be helpful to push any errors from xmlXPathNewValueTree up to the Javascript Console.
- xml/XSLTProcessor.cpp: Soft-linked libxslt.
- 3:55 PM Changeset in webkit [25780] by
-
- 2 edits in trunk/WebKitTools
Reviewed by John.
- Updating leopard leak list to refelct bugs that have been fixed and no longer leak.
- Scripts/run-webkit-tests:
- 3:06 PM Changeset in webkit [25779] by
-
- 2 edits in trunk/WebCore
Reviewed by Brady.
Fix for <rdar://problem/5453350> REGRESSION (r18541, Tiger only): Refreshing a cached https page does not reload CSS and JavaScript resources
http://bugs.webkit.org/show_bug.cgi?id=15121
- loader/FrameLoader.cpp: (WebCore::FrameLoader::addExtraFieldsToRequest): Now sets the cache policy to ReloadIgnoringCacheData. Without setting the cache policy, the default policy in this case was ReturnCacheDataElseLoad. If we're in a situation where the resource is in the cache, we want to make sure to avoid loading from the cache in this case.
- 2:26 PM Changeset in webkit [25778] by
-
- 1 edit in trunk/WebCore/ChangeLog
Correction, NSURLConnection not CFURLConnection.
- 2:24 PM Changeset in webkit [25777] by
-
- 2 edits in trunk/WebCore
Reviewed by Darin.
<rdar://problem/5199546> CrashTracer: [REGRESSION] 8216 crashes in Safari at com.apple.WebCore: WebCore::ResourceHandle::client const + 6
On Tiger CFURLConnection can sometimes call the connection:willCacheResponse: delegate method on
a secondary thread instead of the main thread. This was never an issue before, since the implementation
of this method was very simple and thread safe.
The fix is to block during connection:willCacheResponse: and perform the work on the main thread.
We need to block since this delegate method needs to return a result. If we are already on the
main thread (which sometimes we are), the method does nothing different. If we are on a secondary
thread, we make a mutable dictionary to hold all the arguments and later the result object.
Then performSelectorOnMainThread:withObject: using the _callConnectionWillCacheResponseWithInfo:
selector. This new method just pulls the arguments out for the dictionary and calls the real
connection:willCacheResponse: delegate method (this time on the main thread). The result is stored
in the dictionary, and when performSelectorOnMainThread:withObject: finishes the result is pulled
out and returned.
- platform/network/mac/ResourceHandleMac.mm: (-[WebCoreResourceHandleAsDelegate _callConnectionWillCacheResponseWithInfo:]): (-[WebCoreResourceHandleAsDelegate connection:willCacheResponse:]):
- 9:57 AM Changeset in webkit [25776] by
-
- 2 edits in trunk/WebKitLibraries
Build fix. Ran update-webkitsysteminterface script on Tiger, because
the resulting binary differs when built on Tiger.
- libWebKitSystemInterface.a:
- 9:18 AM Changeset in webkit [25775] by
-
- 3 edits in trunk/WebKitLibraries
Update WebKitSYstemInterface for @font-face changes.
- WebKitSystemInterface.h:
- libWebKitSystemInterface.a:
- 9:10 AM Changeset in webkit [25774] by
-
- 4 edits in trunk/WebCore
Update to reflect the minimal WebKitSystemInterface changes required
to support @font-face.
Reviewed by john
- platform/mac/FontDataMac.mm: (WebCore::FontData::platformInit): (WebCore::FontData::platformWidthForGlyph):
- platform/mac/WebCoreSystemInterface.h:
- platform/mac/WebCoreSystemInterface.mm:
Sep 26, 2007:
- 9:01 PM Changeset in webkit [25773] by
-
- 2 edits in trunk/LayoutTests
Remove a test that was probably only intermittently failing
Reviewed by NOBODY.
- platform/win/Skipped:
- 6:30 PM Changeset in webkit [25772] by
-
- 2 edits in trunk/WebKitQt
2007-09-26 Mark Rowe <mrowe@apple.com>
Qt build fix.
- WebCoreSupport/FrameLoaderClientQt.cpp: (WebCore::FrameLoaderClientQt::objectContentType): Check for empty URL instead of invalid URL.
- 6:29 PM Changeset in webkit [25771] by
-
- 2 edits in trunk/WebKit/gtk
2007-09-26 Mark Rowe <mrowe@apple.com>
Gtk build fix.
- WebCoreSupport/FrameLoaderClientGtk.cpp: Check for empty URL instead of invalid URL.
- 6:24 PM Changeset in webkit [25770] by
-
- 3 edits in trunk/WebCore
- fix Windows build
- platform/win/ClipboardWin.cpp: (WebCore::writeURL): Add a check for empty URL. Remove the check for invalid URL. (WebCore::ClipboardWin::setData): Remove the check for invalid URL.
- plugins/win/PluginViewWin.cpp: (WebCore::PluginViewWin::start): Check for empty URL instead of invalid URL. (WebCore::PluginViewWin::load): Same here.
- 5:56 PM Changeset in webkit [25769] by
-
- 6 edits in trunk/WebCore
Reviewed by John Sullivan.
- fix <rdar://problem/5491054> REGRESSION (Mail, plain-text only): Nothing happens when you click on rdar://<num>&<num> links (or AOL links in Safari)
Remove the non-useful concept of a "malformed"/"invalid" URL.
There are URLs we can parse, and others we can't, but that's not sufficient to
determine if we should try to work with the URL. It's entirely possible that
a so-called "malformed" URL will work just fine if it's passed to the right
software.
- platform/KURL.h: Removed isMalformed() and isValid().
- loader/Cache.cpp: (WebCore::Cache::requestResource): Removed unneeded check if the URL is valid. But do check for an empty URL just to guarantee we don't trip up with a null string. It's possible we can remove this empty URL check later, but it's less risky to leave the empty string behavior alone for now.
- loader/FrameLoader.cpp: (WebCore::FrameLoader::urlSelected): Removed unneeded check if the URL is valid. Back on 2007-07-08, we realized that we needed to allow empty URLs. But we also need to allow other URLs here. This is the code path from the Mail case. (WebCore::FrameLoader::submitForm): Replaced URL validity check with a check for an empty URL (same reasoning as for requestResource above).
- page/InspectorController.cpp: (WebCore::InspectorResource::type):
- rendering/RenderObject.cpp: (WebCore::RenderObject::addPDFURLRect): Removed unneeded check for an invalid URL. In both of these cases it's definitely safe to allow event an empty URL.
- 4:26 PM Changeset in webkit [25768] by
-
- 4 edits in trunk/WebKit
Reviewed by Adele Peterson.
Fixed <rdar://problem/5507476> Promote cache model SPI to API
Promoted cache model SPI to API. This was just a move, with some small
edits to the documentation (changing 'application' to 'WebView' in
some cases, since the interface is now per-WebView).
- WebView/WebPreferences.h:
- WebView/WebPreferences.m: (-[WebPreferences setCacheModel:]): (-[WebPreferences cacheModel]):
- WebView/WebPreferencesPrivate.h:
- 4:02 PM Changeset in webkit [25767] by
-
- 4 edits in branches/Safari-2.0-dashboard/WebCore
2007-08-24 Sam Weinig <sam@webkit.org>
Reviewed by Adele.
Fix for <rdar://problem/5426142>
Use the EventTarget's frame when creating the EventListener.
Tests: http/tests/security/listener/xss-JSTargetNode-onclick-addEventListener.html
http/tests/security/listener/xss-JSTargetNode-onclick-shortcut.html
http/tests/security/listener/xss-window-onclick-addEventListener.html
http/tests/security/listener/xss-window-onclick-shortcut.html
- bindings/js/JSEventTargetNode.cpp: (WebCore::JSEventTargetNode::setListener): (WebCore::JSEventTargetNodePrototypeFunction::callAsFunction):
- bindings/js/kjs_window.cpp: (KJS::WindowFunc::callAsFunction):
2007-08-28 Sam Weinig <sam@webkit.org>
Reviewed by Darin.
Update fix landed in r25249 to account for XMLHttpRequest, which can also be EventTargets.
Tests: http/tests/security/listener/xss-XMLHttpRequest-addEventListener.html
http/tests/security/listener/xss-XMLHttpRequest-shortcut.html
- bindings/js/JSXMLHttpRequest.cpp: (KJS::JSXMLHttpRequest::putValueProperty): Use the Window object associated with document's frame, not the active Window. (KJS::JSXMLHttpRequestPrototypeFunction::callAsFunction): ditto
- xml/XMLHttpRequest.h: (WebCore::XMLHttpRequest::document): Expose Document member so bindings can access the correct frame.
- 2:23 PM Changeset in webkit [25766] by
-
- 2 edits in trunk/LayoutTests
Add more failing tests to the Windows Skipped file
Reviewed by NOBODY.
- platform/win/Skipped:
- 1:06 PM Changeset in webkit [25765] by
-
- 1 edit in trunk/JavaScriptCore/kjs/testkjs.cpp
Oh, Visual Studio, why don't you see when a project file has changed that the project needs to be rebuilt?
- 12:57 PM Changeset in webkit [25764] by
-
- 2 edits in trunk/JavaScriptCore
Make testkjs delay-load WebKit.dll so WebKitInitializer can work its magic
Rubberstamped by Anders.
- JavaScriptCore.vcproj/testkjs/testkjs.vcproj:
- 10:23 AM Changeset in webkit [25763] by
-
- 3 edits in trunk/WebKitTools
Make not finding an installed WebKit non-fatal
What really matters is that we are able to load WebKit and its
dependencies, not whether we think we've found an installed WebKit.
Reviewed by Sam.
- WebKitInitializer/WebKitInitializer.cpp: (getWebViewCLSID): Use new LOG_WARNING macro. (getInstalledWebKitDirectory): Ditto. (initializeWebKit): Don't die if we didn't find an installed WebKit, and use new macros.
- WebKitInitializer/WebKitInitializer.vcproj: Disabled a warning.
- 10:22 AM Changeset in webkit [25762] by
-
- 2 edits in trunk/WebKitTools
Committed this file before saving it
- WebKitInitializer/debug_internal.vsprops:
- 4:28 AM Changeset in webkit [25761] by
-
- 2 edits in trunk/WebCore
2007-09-26 George Staikos <staikos@kde.org>
Reviewed by John Sullivan.
Only highlight the applicable portion of the text run for a search
highlight. No way to make a testcase for this presently.
- rendering/InlineTextBox.cpp: (WebCore::InlineTextBox::paintTextMatchMarker):
- 12:07 AM Changeset in webkit [25760] by
-
- 3 edits1 add in trunk/WebKitTools
Load WebKit and its dependencies manually so they come from DllDirectory
I had to change WebKitInitializer to use malloc/free instead of
new/delete to avoid pulling in WebKit (and its dependencies) early
through use of fastMalloc.
Reviewed by Oliver.
- WebKitInitializer/WebKitInitializer.cpp: (getStringValue): Changed to use malloc/free. (getInstalledWebKitDirectory): Ditto. (initializeWebKit): Call SetDllDirectory first, then load all of WebKit's dependencies manually, finally followed by WebKit itself.
- WebKitInitializer/WebKitInitializer.vcproj: Added a new configuration to match our other projects.
- WebKitInitializer/debug_internal.vsprops: Added.
