Context Navigation

← Previous Period
Next Period →

Timeline

Jul 6, 2007:

11:29 PM Changeset in webkit [24087] by mjs

3 edits
3 adds in trunk

LayoutTests:

Reviewed by Oliver.

test case for <rdar://problem/5133420> ASSERT in WebDocumentLoaderMac::decreaseLoadCount() un-discarding Gmail message

fast/loader/early-load-cancel-expected.txt: Added.
fast/loader/early-load-cancel.html: Added.
fast/loader/resources/early-load-cancel-inner.html: Added.

WebCore:

Reviewed by Oliver.

fixed <rdar://problem/5133420> ASSERT in WebDocumentLoaderMac::decreaseLoadCount() un-discarding Gmail message

test case: fast/loader/early-load-cancel.html

loader/ResourceLoader.cpp: (WebCore::ResourceLoader::didCancel): Don't send cancel-related client callbacks if we never even sent willSendRequest because we were deferred.

9:28 PM Changeset in webkit [24086] by bdash

2 edits in trunk/WebKitTools

2007-07-07 Mark Rowe <mrowe@apple.com>

Reviewed by Oliver.

DumpRenderTree/DumpRenderTree.m: (convertWebResourceResponseToDictionary): Fix leak of two NSMutableString's introduced in r24076.

9:01 PM Changeset in webkit [24085] by oliver

13 edits in trunk

WebCore:

Reviewed by Maciej.

Pulled a number of methods from WebKit and WebCoreFrameBridge into
C++ objects.

WebCore.exp:
WebCore.xcodeproj/project.pbxproj:
editing/Editor.cpp: (WebCore::Editor::replaceMarkedText):

From -[WebCoreFrameBridge replaceMarkedTextWithText:]

editing/Editor.h:
page/Frame.cpp: (WebCore::Frame::setMarkedTextRange):

Pulled platform independent logic from Mac specific implementation

(WebCore::Frame::selectRangeInMarkedText):

Converted from -[WebHTMLView _selectRangeInMarkedText:]

page/Frame.h:
page/mac/FrameMac.mm: (WebCore::Frame::setMarkedTextRange):

Now only does Mac specific logic when manipulating initial
NSAttributedString

page/mac/WebCoreFrameBridge.h:
page/mac/WebCoreFrameBridge.mm:

Removed -[WebCoreFrameBridge replaceMarkedTextWithText:]

WebKit:

Reviewed by Maciej.

<rdar://problem/5318756> Need to refactor IM/Marked text code to share logic with windows.

Moved a number of methods from WebHTMLView into WebCore.
Replaced bridge methods with calls directly into WebCore objects.

WebKit.xcodeproj/project.pbxproj:
WebView/WebHTMLView.mm: (-[WebHTMLView setMarkedText:selectedRange:]): Now calls WebCore object methods directly, rather than calling via the bridge.

7:37 PM Changeset in webkit [24084] by bdash

7 edits
4 adds in trunk

2007-07-06 Mitz Pettel <mitz@webkit.org>

Reviewed by Darin.

fix http://bugs.webkit.org/show_bug.cgi?id=14545 REGRESSION (r21854-r21869): Repro crash in RenderBlock::updateFirstLetter @ nola.com/rose/

Test: fast/css/first-letter-capitalized.html

rendering/RenderBlock.cpp: (WebCore::RenderBlock::updateFirstLetter): Moved the call to setFirstLetter further down so that the first letter is not destroyed if a text transform is applied to the remaining text fragment when it is add to the render tree.
rendering/RenderText.cpp: (WebCore::RenderText::previousCharacter): Factored out of setTextInternal. (WebCore::RenderText::setTextInternal): Changed to call previousCharacter().
rendering/RenderText.h:
rendering/RenderTextFragment.cpp: (WebCore::RenderTextFragment::previousCharacter): Override to return the character in the original string before the start of the fragment.
rendering/RenderTextFragment.h:

2007-07-06 Mitz Pettel <mitz@webkit.org>

Reviewed by Darin.

test for http://bugs.webkit.org/show_bug.cgi?id=14545 REGRESSION (r21854-r21869): Repro crash in RenderBlock::updateFirstLetter @ nola.com/rose/

fast/css/first-letter-capitalized-expected.checksum: Added.
fast/css/first-letter-capitalized-expected.png: Added.
fast/css/first-letter-capitalized-expected.txt: Added.
fast/css/first-letter-capitalized.html: Added.

7:35 PM Changeset in webkit [24083] by bdash

2 edits in trunk/WebCore

2007-07-06 Mitz Pettel <mitz@webkit.org>

Reviewed by Beth.

fix http://bugs.webkit.org/show_bug.cgi?id=14536 Unnecessary layout done when mousing down in text field

No test possible because there is no change in functionality.

rendering/RenderTextControl.cpp: (WebCore::RenderTextControl::setStyle): Reset the height and the width in the old style to avoid getting a layout hint as a result of having mutated the old style during layout. (WebCore::RenderTextControl::layout): Update children's layout if an inner block's dimensions should change. This need for layout was previously masked by the bug.

6:02 PM Changeset in webkit [24082] by darin

2 edits in trunk/LayoutTests

mac/leopard/Skipped: Updated bug numbers for some bugs that were duplicates.

5:27 PM Changeset in webkit [24081] by beidson

2 edits in trunk/WebCore

Fix builds that use IconDatabaseNone.cpp - better, this time

loader/icon/IconDatabaseNone.cpp: (WebCore::IconDatabase::isOpen):

5:20 PM Changeset in webkit [24080] by beidson

2 edits in trunk/WebCore

Fix builds that use IconDatabaseNone.cpp

loader/icon/IconDatabaseNone.cpp: (WebCore::IconDatabase::databasePath):

4:48 PM Changeset in webkit [24079] by thatcher

3 edits in trunk/LayoutTests

Reviewed by Adam.

<rdar://problem/5313506> Some attributed string layout tests have an extra space character on Leopard

Leopard serializes empty attribute dictionaries as "{ }", turn these into "{}" to match Tiger.

fast/text/attributed-substring-from-range-001.html:
mac/leopard/Skipped: Don't skip this test anymore.

4:44 PM Changeset in webkit [24078] by thatcher

1 edit in trunk/WebKitTools/ChangeLog

Fix typo.

4:43 PM Changeset in webkit [24077] by sullivan

8 edits in trunk

WebCore:

Reviewed by Brady

WebCore part of fix for: <rdar://problem/5310739> Time Machine shouldn't back up WebKit's icon database files

loader/icon/IconDatabase.h:
loader/icon/IconDatabase.cpp: (WebCore::IconDatabase::isOpen): made this method const (necessary to make IconDatabase::databasePath const) (WebCore::IconDatabase::databasePath): new method, returns SQLDatabase's path if open, otherwise empty string

loader/icon/SQLDatabase.cpp: (WebCore::SQLDatabase::open): don't append a null character to the m_path ivar. The appended null character was making the path() method return a String with a trailing null, which is bad. However, there were no callers to the path() method until now. Instead of appending a null character to the ivar, use charactersWithNullTermination() instead of characters() when opening the database.

loader/icon/SQLDatabase.h: (WebCore::SQLDatabase::isOpen): made this method const (necessary to make IconDatabase::databasePath const) (WebCore::SQLDatabase::path): made this method const (not required for these changes, but seemed worth doing at the same time)

WebCore.exp: added symbol for IconDatabase::databasePath

WebKit:

Reviewed by Brady

WebKit part of fix for: <rdar://problem/5310739> Time Machine shouldn't back up WebKit's icon database files

Misc/WebIconDatabase.mm: (-[WebIconDatabase init]): Use CSBackupSetItemExcluded to tell Time Machine not to back up the icon database file

4:40 PM Changeset in webkit [24076] by andersca

15 edits in trunk

LayoutTests:

Reviewed by David Kilzer.

<rdar://problem/5313502>
Many webarchive tests fail due to different NSURLResponse serialization on Leopard

mac/leopard/Skipped:
webarchive/archive-with-unencoded-url-expected.txt:
webarchive/test-body-background-expected.txt:
webarchive/test-duplicate-resources-expected.txt:
webarchive/test-frameset-expected.txt:
webarchive/test-img-src-expected.txt:
webarchive/test-input-src-expected.txt:
webarchive/test-link-href-expected.txt:
webarchive/test-object-data-expected.txt:
webarchive/test-script-src-expected.txt:
webarchive/test-table-background-expected.txt:
webarchive/test-td-background-expected.txt:

WebKitTools:

Reviewed by David Kilzer.

<rdar://problem/5313502>
Many webarchive tests fail due to different NSURLResponse serialization on Leopard

Don't dump the serialized form of NSURLResponse. Instead, create a dictionary with the
response's attributes. Also change the JavaScript MIME type to be "text/javascript".

DumpRenderTree/DumpRenderTree.m: (convertMIMEType): (convertWebResourceDataToString): (convertWebResourceResponseToDictionary): (serializeWebArchiveToXML):

4:22 PM Changeset in webkit [24075] by thatcher

5 edits in trunk

LayoutTests:

Reviewed by Oliver.

<rdar://problem/5313506> Some attributed string layout tests have an extra space character on Leopard

Don't dump the attributed string's description, this isn't helpfull for the test. Just log
the attribute names at index 0, which we expect to be nothing, to make sure the string
has no attributes.

fast/forms/attributed-strings-expected.txt:
fast/forms/attributed-strings.html:

WebKitTools:

Reviewed by Oliver.

Initlize the NSMutableAttributedString directly instead of creating
a NSAttributedString first.

DumpRenderTree/TextInputController.m: (-[TextInputController attributedStringWithString:]):

3:44 PM Changeset in webkit [24074] by staikos

1 edit in trunk/WebCore/WebCore.pro

missed in previous checkin

3:36 PM Changeset in webkit [24073] by staikos

2 edits
1 add in trunk/WebCore

Move locatizations into another file.

2:54 PM Changeset in webkit [24072] by ggaren

3 edits in trunk/WebCore

Reviewed by Anders Carlsson.

A little more FrameLoader cleanup while working on <rdar://problem/5126396>.

Moved url() and URL() methods next to each other to highlight their subtlety.

loader/FrameLoader.cpp: (WebCore::FrameLoader::url):
loader/FrameLoader.h:

2:37 PM Changeset in webkit [24071] by andersca

16 edits in trunk

LayoutTests:

Reviewed by David Kilzer.

Update tests.

webarchive/archive-empty-frame-dom-expected.txt:
webarchive/archive-empty-frame-source-expected.txt:
webarchive/archive-with-unencoded-url-expected.txt:
webarchive/test-body-background-expected.txt:
webarchive/test-duplicate-resources-expected.txt:
webarchive/test-frameset-expected.txt:
webarchive/test-img-src-expected.txt:
webarchive/test-input-src-expected.txt:
webarchive/test-link-href-expected.txt:
webarchive/test-object-data-expected.txt:
webarchive/test-script-src-expected.txt:
webarchive/test-table-background-expected.txt:
webarchive/test-td-background-expected.txt:

WebKitTools:

Reviewed by David Kilzer.

Replace "Apple Computer" with "Apple" in the DTD declaration.

DumpRenderTree/DumpRenderTree.m: (serializeWebArchiveToXML):

2:30 PM Changeset in webkit [24070] by ggaren

4 edits in trunk/WebCore

Reviewed by Anders Carlsson.

A little FrameLoader cleanup while working on <rdar://problem/5126396>.

Renamed endIfNotLoading() to endIfNotLoadingMainResource() because "loading"
and "loadingMainResource" have distinct meanings and this function
honors the latter.

(WebCore::FrameLoader::endIfNotLoadingMainResource): Protect after the
early return to avoid ref count churn.

loader/FrameLoader.h: (WebCore::FrameLoader::isLoadingMainResource): Inlined this function.

2:06 PM Changeset in webkit [24069] by staikos

15 edits in trunk

Patch from Adam Treat to convert QWebFrame to a QFrame from a QAbstractScrollArea.

1:52 PM Changeset in webkit [24068] by andersca

3 edits
2 adds in trunk

LayoutTests:

Reviewed by Mitz.

<rdar://problem/5280532>
REGRESSION: Can't access nested embed by document.name if object is not immediate parent

plugins/embed-inside-object-expected.txt: Added.
plugins/embed-inside-object.html: Added.

WebCore:

Reviewed by Mitz.

<rdar://problem/5280532>
REGRESSION: Can't access nested embed by document.name if object is not immediate parent

Traverse up the tree looking for an object element.

html/HTMLEmbedElement.cpp: (WebCore::HTMLEmbedElement::getInstance):

1:29 PM Changeset in webkit [24067] by aroben

2 edits in trunk/LayoutTests

Add another Leopard failure

mac/leopard/Skipped:

11:11 AM Changeset in webkit [24066] by aroben

2 edits in trunk/LayoutTests

Add another crasher to the Skipped file

Also moved a test that had a bug filed about it out of the section of
unfiled failures.

win/Skipped:

11:00 AM Changeset in webkit [24065] by andersca

4 edits in trunk/LayoutTests

Reviewed by Alexey.

<rdar://problem/5313518> fast/events/iframe-object-onload.html shows events out of order on Leopard

Fix test so that it doesn't matter in what order the frames/objects are loaded. This is done by keeping
the load logs in an array and sort it before dumping it.

fast/events/iframe-object-onload-expected.txt:
fast/events/iframe-object-onload.html:
mac/leopard/Skipped:

10:28 AM Changeset in webkit [24064] by staikos

2 edits in trunk/WebCore

Don't clobber variables that are already set, resulting in a broken build.

10:08 AM Changeset in webkit [24063] by andersca

6 edits
4 deletes in trunk/LayoutTests

Reviewed by Alexey.

<rdar://problem/5313512> fast/frames/frame-set-same-{location,src}.html sometimes hang on Leopard

Make these tests more robust against frames loading in a different order. Also, turn the tests
into text tests and remove them from the Leopard skipped list.

fast/frames/frame-set-same-location-expected.checksum: Removed.
fast/frames/frame-set-same-location-expected.png: Removed.
fast/frames/frame-set-same-location-expected.txt:
fast/frames/frame-set-same-location.html:
fast/frames/frame-set-same-src-expected.checksum: Removed.
fast/frames/frame-set-same-src-expected.png: Removed.
fast/frames/frame-set-same-src-expected.txt:
fast/frames/frame-set-same-src.html:
mac/leopard/Skipped:

9:54 AM Changeset in webkit [24062] by staikos

2 edits in trunk/WebKitTools

Add --qmakearg=foo to make it possible to configure the build without editing
the .pro files each time.

7:56 AM Changeset in webkit [24061] by bdash

2 edits in trunk/WebKit

2007-07-07 Mark Rowe <mrowe@apple.com>

Build fix. Update WebDynamicScrollBarsView.h to include methods added and used in r24060.

WebView/WebDynamicScrollBarsView.h:

7:40 AM Changeset in webkit [24060] by thatcher

3 edits in trunk/WebKit

Reviewed by John.

<rdar://problem/5236277> REGRESSION-9A458: SPI for setting scroll bar behavior doesn't work

Calling setHorizontalScrollingMode: calls updateScrollers before returning, this will cause
WebCore to reset the scrolling mode based on the CSS overflow rules. So the setAlwaysShowHorizontalScroller:
and setAlwaysShowVerticalScroller: methods needed a way to lock the scrolling mode before
calling updateScrollers.

WebView/WebDynamicScrollBarsView.m: (-[WebDynamicScrollBarsView updateScrollers]): (-[WebDynamicScrollBarsView setHorizontalScrollingMode:]): (-[WebDynamicScrollBarsView setHorizontalScrollingMode:andLock:]): (-[WebDynamicScrollBarsView setVerticalScrollingMode:]): (-[WebDynamicScrollBarsView setVerticalScrollingMode:andLock:]): (-[WebDynamicScrollBarsView setScrollingMode:]): (-[WebDynamicScrollBarsView setScrollingMode:andLock:]):
WebView/WebView.mm: (-[WebView setAlwaysShowVerticalScroller:]): (-[WebView setAlwaysShowHorizontalScroller:]):

5:09 AM Changeset in webkit [24059] by mjs

12 edits in trunk

JavaScriptCore:

Reviewed by Antti.

<rdar://problem/5311093> JavaScriptCore fails to build with strict-aliasing warnings

Configurations/Base.xcconfig: Re-enable -Wstrict-aliasing
bindings/jni/jni_utility.cpp: (KJS::Bindings::getJNIEnv): Type-pun via a union instead of a pointer cast.
wtf/HashMap.h: (WTF::): Instead of doing type-punned assignments via pointer cast, do one of three things: (1) assign directly w/o cast if storage type matches real type; (2) assign using cast via union if type does not need reffing; (3) copy with memcpy and ref/deref manually if type needs reffing. This is ok peref-wise because memcpy of a constant length gets optomized. HashTraits are now expected to make ref()/deref() take the storage type, not the true type.
wtf/HashSet.h: (WTF::): Same basic idea.
wtf/HashTable.h: (WTF::): Added Assigner template for use by HashMap/HashSet. Change RefCounter to call ref() and deref() via storage type, avoiding the need to type-pun. (WTF::RefCounter::ref): ditto (WTF::RefCounter::deref): ditto
wtf/HashTraits.h: (WTF::): Change ref() and deref() for RefPtr HashTraits to take the storage type; cast via union to pointer type.
wtf/FastMalloc.cpp: (WTF::TCMalloc_PageHeap::init): Changed from constructor to init function so this can go in a union. (WTF::): redefine pageheap macro in terms of getPageHeap(). (WTF::getPageHeap): new inline function, helper for pageheap macro. This hides the cast in a union. (WTF::TCMalloc_ThreadCache::InitModule): Call init() instead of using placement new to initialize page heap.
wtf/TCPageMap.h: (TCMalloc_PageMap1::init): Changed from constructor to init function. (TCMalloc_PageMap2::init): ditto (TCMalloc_PageMap3::init): ditto

WebCore:

Reviewed by Antti.

<rdar://problem/5311093> JavaScriptCore fails to build with strict-aliasing warnings

platform/StringHash.h: (WTF::): Adapt to newer way to do storage types.

4:42 AM Changeset in webkit [24058] by bdash

2 edits in trunk/WebCore

2007-07-06 Bernhard Rosenkraenzer <bero@arklinux.org>

Reviewed by Sam.

Fix for http://bugs.webkit.org/show_bug.cgi?id=13985
Bug 13985: WebCore/rendering/RenderTable.cpp fails to compile with gcc 4.2.x

rendering/RenderTable.cpp: (WebKit::RenderTable::calcBorderLeft) (WebKit::RenderTable::calcBorderRight)

3:31 AM Changeset in webkit [24057] by bdash

2 edits
3 adds in trunk/WebCore

2007-07-06 Maxime Britto <mbritto@pleyo.com>

Reviewed by Maciej.

Fix http://bugs.webkit.org/show_bug.cgi?id=13038

Bug 13038: ASSERTION FAILED: item->target().isEmpty()

m_mainFrame->tree()->find(item->target()) == m_mainFrame

Modify an ASSERT in goToItem related to going back/forward in history. We don't need to check the frameset if the target is _blank.
The attached test case is a manual one since it relies on having pages opened in tabs in Safari.

manual-tests/goBack-blank-tab-page.html: Added.
manual-tests/resources/before-go-back.html: Added.
manual-tests/resources/will-go-back.html: Added.
page/Page.cpp: (WebCore::Page::goToItem):

3:21 AM Changeset in webkit [24056] by bdash

5 edits in trunk/LayoutTests

2007-07-06 Mark Rowe <mrowe@apple.com>

Unreviewed. Update test results.

fast/encoding/utf-32-big-endian-bom-expected.txt:
fast/encoding/utf-32-big-endian-nobom-expected.txt:
fast/encoding/utf-32-little-endian-bom-expected.txt:
fast/encoding/utf-32-little-endian-nobom-expected.txt:

3:06 AM Changeset in webkit [24055] by bdash

4 edits
2 adds in trunk

2007-07-06 Mitz Pettel <mitz@webkit.org>

Reviewed by Maciej.

fix http://bugs.webkit.org/show_bug.cgi?id=14400 Assertion failure (SHOULD NEVER BE REACHED) going back on YouTube

Test: fast/forms/button-state-restore.html

dom/Document.h: (WebCore::Document::isFormElementRegistered): Added.
html/HTMLGenericFormElement.cpp: (WebCore::HTMLGenericFormElement::closeRenderer): Restore saved form element state only if the element is registered with the document to have its state saved.

2007-07-06 Mitz Pettel <mitz@webkit.org>

Reviewed by Maciej.

test for http://bugs.webkit.org/show_bug.cgi?id=14400 Assertion failure (SHOULD NEVER BE REACHED) going back on YouTube

fast/forms/button-state-restore-expected.txt: Added.
fast/forms/button-state-restore.html: Added.

3:05 AM Changeset in webkit [24054] by bdash

3 edits
4 adds in trunk

2007-07-06 Mitz Pettel <mitz@webkit.org>

Reviewed by Maciej.

fix http://bugs.webkit.org/show_bug.cgi?id=14487 Incomplete repaint of background and border of boxes with -webkit-border-fit:lines

Test: fast/repaint/border-fit-lines.html

rendering/RenderObject.cpp: (WebCore::RenderObject::repaintAfterLayoutIfNeeded): Do a full repaint if -webkit-border-fit:lines was specified.

2007-07-06 Mitz Pettel <mitz@webkit.org>

Reviewed by Maciej.

repaint test for http://bugs.webkit.org/show_bug.cgi?id=14487 Incomplete repaint of background and border of boxes with -webkit-border-fit:lines

fast/repaint/border-fit-lines-expected.checksum: Added.
fast/repaint/border-fit-lines-expected.png: Added.
fast/repaint/border-fit-lines-expected.txt: Added.
fast/repaint/border-fit-lines.html: Added.

3:03 AM Changeset in webkit [24053] by bdash

3 edits
3 adds in trunk

2007-07-06 Tony Chang <idealisms@gmail.com>

Reviewed by Maciej.

http://bugs.webkit.org/show_bug.cgi?id=14516
Bug 14516: crash loading multipart/x-mixed-replace data on windows safari

NULL check m_resourceData to prevent a crash in Safari on Windows.

loader/ResourceLoader.cpp: (WebCore::ResourceLoader::clearResourceData):

2007-07-06 Tony Chang <idealisms@gmail.com>

Reviewed by Maciej.

http://bugs.webkit.org/show_bug.cgi?id=14516
Bug 14516: crash loading multipart/x-mixed-replace data on windows safari

http/tests/multipart/resources/multipart-nodashes.php: Added.
http/tests/multipart/win-boundary-crash-expected.txt: Added.
http/tests/multipart/win-boundary-crash.html: Added.

3:00 AM Changeset in webkit [24052] by bdash

7 edits
8 adds in trunk

2007-07-06 Jungshik Shin <jungshik.shin@gmail.com>

Reviewed by Alexey.

Add UTF-32 encoding support http://bugs.webkit.org/show_bug.cgi?id=13415

Test:

fast/encoding/utf-32-big-endian-bom.html
fast/encoding/utf-32-big-endian-nobom.xml
fast/encoding/utf-32-little-endian-bom.html
fast/encoding/utf-32-little-endian-nobom.xml

loader/TextResourceDecoder.cpp: (WebCore::TextResourceDecoder::checkForBOM): (WebCore::TextResourceDecoder::checkForHeadCharset):
platform/TextDecoder.cpp: (WebCore::TextDecoder::checkForBOM):
platform/TextDecoder.h:
platform/TextEncoding.cpp: (WebCore::UTF32BigEndianEncoding): (WebCore::UTF32LittleEndianEncoding):
platform/TextEncoding.h:

2007-07-06 Jungshik Shin <jungshik.shin@gmail.com>

Reviewed by Alexey.

test for http://bugs.webkit.org/show_bug.cgi?id=13415

fast/encoding/utf-32-big-endian-bom-expected.txt: Added.
fast/encoding/utf-32-big-endian-bom.html: Added.
fast/encoding/utf-32-big-endian-nobom-expected.txt: Added.
fast/encoding/utf-32-big-endian-nobom.xml: Added.
fast/encoding/utf-32-little-endian-bom-expected.txt: Added.
fast/encoding/utf-32-little-endian-bom.html: Added.
fast/encoding/utf-32-little-endian-nobom-expected.txt: Added.
fast/encoding/utf-32-little-endian-nobom.xml: Added.

2:56 AM Changeset in webkit [24051] by bdash

3 edits in trunk/WebCore

2007-07-06 Holger Hans Peter Freyther <zecke@selfish.org>

Reviewed by Maciej.

Start implementening RenderThemeGdk/Gtk using GtkStyle. The approach is
to allocate one native control for each widget type. And use margins, spacing,
style properties from the Widget. And for drawing use GtkStyle on the rect supplied.

This should allow nice integration and be relative resource friendly. Mozilla is using
a similiar approach. They seem to share the GdkPixmap among several operations and this
will be an obvious candidate for improvements. Create a normally big enough GdkPixmap and
use that for drawing and then set clip regions for cairo to make sure to not go beyond the
width and height.

The next steps will be to introduce states for the widgets, add more widgets, handle focus,
and various other style attributes, honor the size and margin and use that to propagate this
back to RenderStyle*.

platform/gdk/RenderThemeGdk.cpp: (WebCore::RenderThemeGdk::RenderThemeGdk): (WebCore::RenderThemeGdk::setCheckboxSize): (WebCore::RenderThemeGdk::paintCheckbox): (WebCore::RenderThemeGdk::setRadioSize): (WebCore::RenderThemeGdk::paintRadio): (WebCore::RenderThemeGdk::paintButton): (WebCore::RenderThemeGdk::copyToContext): (WebCore::RenderThemeGdk::gtkButton): (WebCore::RenderThemeGdk::gtkCheckbox): (WebCore::RenderThemeGdk::gtkRadioButton): (WebCore::RenderThemeGdk::gtkWindowContainer):
platform/gdk/RenderThemeGdk.h:

2:52 AM Changeset in webkit [24050] by bdash

2 edits in trunk/WebKit

2007-07-06 Mitz Pettel <mitz@webkit.org>

Reviewed by Maciej.

fix http://bugs.webkit.org/show_bug.cgi?id=10267 Can't scroll page downwards with scroll wheel, when pointer is on top of non-scrolling iframe

WebView/WebDynamicScrollBarsView.m: (-[WebDynamicScrollBarsView scrollWheel:]): Override the superclass implementation to forward the wheel event to the next responder if this view does not allow scrolling in the event's direction.

2:11 AM Changeset in webkit [24049] by rwlbuis

1 edit in trunk/WebCore/html/CanvasGradient.cpp

Fix build.

2:00 AM Changeset in webkit [24048] by rwlbuis

3 edits
4 adds in trunk

Reviewed by Adam.

http://bugs.webkit.org/show_bug.cgi?id=14442
Adding a stop with value 1.0 to a gradient that has already been used has no effect

Implement start and end stop handling differently so they are not
in the color stop array.

1:44 AM Changeset in webkit [24047] by aroben

7 edits in trunk

Fix <rdar://problem/5301994> Garbled text still showing on some sites

WebCore:

Fix <rdar://problem/5301994> Garbled text still showing on some sites

Replace calls to GetGlyphIndices with wkGetGlyphs.

Reviewed by Alice.

platform/win/GlyphPageTreeNodeWin.cpp: (WebCore::GlyphPage::fill): Use wkGetGlyphs.
platform/win/FontDataWin.cpp: (WebCore::FontData::containsCharacters):

WebKitLibraries:

Update WebKitSystemInterface.{h,lib} for <rdar://problem/5301994>

Reviewed by Alice.

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

12:41 AM Changeset in webkit [24046] by pewtermoose

1 edit in trunk/WebKitSite/ChangeLog

Forgot ChangeLog description

12:37 AM Changeset in webkit [24045] by pewtermoose

2 edits in trunk/WebKitSite

Reviewed by Adam Roben.

building/build.html:

12:01 AM Changeset in webkit [24044] by pewtermoose

2 edits in trunk/WebCore

Not reviewed - Windows build fix.

bridge/win/FrameWin.cpp: (WebCore::imageFromSelection):

Jul 5, 2007:

11:14 PM Changeset in webkit [24043] by staikos

2 edits in trunk/WebKitTools

Add optional --qmake= and --qt arguments to force Qt builds and force with a
specific qmake binary

10:12 PM Changeset in webkit [24042] by staikos

5 edits in trunk

Switch USE(ICONDATABASE) to ENABLE(ICONDATABASE)

8:49 PM Changeset in webkit [24041] by sullivan

18 edits in trunk

WebCore:

Reviewed by Adam

WebCore part of fix for <rdar://problem/5315033>

rendering/RenderObject.h: replaced PaintRestrictionSelectionOnlyWhiteText with PaintRestrictionSelectionOnlyBlackText (WebCore::RenderObject::PaintInfo::PaintInfo): replaced forceWhiteText with forceBlackText

rendering/RenderObject.cpp: (WebCore::RenderObject::addLineBoxRects): added a useSelectionHeight parameter, which defaults to false; this method is unimplemented at this level so this has no effect.

rendering/RenderText.h:
rendering/RenderText.cpp: (WebCore::RenderText::addLineBoxRects): added a useSelectionHeight parameter, which defaults to false. If true, use the vertical position and height returned by selectionRect rather than replacing it with the more tight-fitting position and height of the line box.

dom/Range.h:
dom/Range.cpp: (WebCore::Range::addLineBoxRects): added a useSelectionHeight parameter, which defaults to false. Passes this along to RenderObject.

page/Frame.h: renamed parameter to selectionImage
page/Frame.cpp: (WebCore::Frame::selectionTextRects): pass true for addLineBoxRects new useSelectionHeight parameter, so the rects returned by this new method will match the height that the selection would be painted here (WebCore::Frame::paint): updated for white->black name changes

page/mac/FrameMac.mm: (WebCore::Frame::selectionImage): updated for white->black name changes

rendering/InlineFlowBox.cpp: (WebCore::InlineFlowBox::paintTextDecorations): updated for white->black name changes

rendering/RenderLayer.cpp: (WebCore::RenderLayer::paintLayer): updated for white->black name changes

rendering/InlineTextBox.cpp: (WebCore::InlineTextBox::paint): updated for white->black name changes, and actually use black instead of white. This was the goal of all the name changes.

WebKit:

Reviewed by Adam

WebKit part of fix for <rdar://problem/5315033>

WebView/WebDocumentPrivate.h: new selectionImageForcingBlackText: method. selectionImageForcingWhiteText: is no longer used and was never in an official release of WebKit, so it could be removed, except that doing so would cause trouble for people using nightly WebKit with Safari 3.0 beta. So I left it in, but made it just force black text instead of white text, which will look different for those nightly WebKit/Safari 3.0 beta people but not break anything.

Misc/WebSearchableTextView.m: (-[WebSearchableTextView selectionImageForcingBlackText:]): new unimplemented protocol method for this obsolete class

WebView/WebHTMLView.mm: (-[WebHTMLView selectionImageForcingBlackText:]): calls through to WebCore the way selectionImageForcingWhiteText: used to (-[WebHTMLView selectionImageForcingWhiteText:]): now just calls selectionImageForcingBlackText:, thus not working as you would expect from the name

WebView/WebPDFView.mm: (-[WebPDFView selectionImageForcingBlackText:]): guts of old selectionImageForcingWhiteText:, but with black substituted for white (-[WebPDFView selectionImageForcingWhiteText:]): now just calls selectionImageForcingBlackText:, thus not working as you would expect from the name

8:33 PM Changeset in webkit [24040] by aroben

1 edit in trunk/LayoutTests/win/Skipped

Forgot the file extension

8:31 PM Changeset in webkit [24039] by aroben

2 edits in trunk/LayoutTests

Updated the Windows Skipped file

win/Skipped:

7:28 PM Changeset in webkit [24038] by andersca

2 edits in trunk/WebCore

Reviewed by Oliver.

Fix layout test failures.

loader/FrameLoader.cpp: (WebCore::FrameLoader::commitProvisionalLoad): Don't make a representation if we're creating the initial document.

6:58 PM Changeset in webkit [24037] by aroben

4 edits
4 adds
4 deletes in trunk

Fix Bug 14388: Input in text fields limited to 1024 characters

WebCore:

2007-07-05 Alp Toker <alp.toker@collabora.co.uk>

Fix Bug 14388: Input in text fields limited to 1024 characters
http://bugs.webkit.org/show_bug.cgi?id=14388

Slightly modified by Adam Roben.

Reviewed by Adele.

Test: fast/forms/input-length.html

html/HTMLInputElement.cpp: Bump limit to 524288. (WebCore::HTMLInputElement::init): (WebCore::HTMLInputElement::parseMappedAttribute):
html/HTMLInputElement.h:

LayoutTests:

Added two tests for Bug 14388: Input in text fields limited to 1024 characters

fast/forms/input-maxlength-2.html is obsoleted by
fast/forms/input-maxlength.html.

Reviewed by Adele.

fast/forms/input-implicit-length-limit-expected.txt: Added.
fast/forms/input-implicit-length-limit.html: Added.
fast/forms/input-maxlength-2-expected.checksum: Removed.
fast/forms/input-maxlength-2-expected.png: Removed.
fast/forms/input-maxlength-2-expected.txt: Removed.
fast/forms/input-maxlength-2.html: Removed.
fast/forms/input-maxlength-expected.txt: Added.
fast/forms/input-maxlength.html: Added.

6:58 PM Changeset in webkit [24036] by aroben

2 edits in trunk/LayoutTests

Add another hanging test to the Leopard Skipped file

Rubberstamped by Adele.

mac/leopard/Skipped:

6:14 PM Changeset in webkit [24035] by kevino

1 edit in branches/wx-port-alpha/trunk/WebCore/platform/wx/RenderThemeWx.cpp

Selected text under MSW should be white too.

6:02 PM Changeset in webkit [24034] by justing

1 edit in trunk/LayoutTests/fast/frames/iframe-double-attach-expected.txt

Updating expected results for r24027.

5:56 PM Changeset in webkit [24033] by kevino

3 edits in branches/wx-port-alpha/trunk/WebKit/wx

Get RunScript working and add a way to test it in the demo.

5:40 PM Changeset in webkit [24032] by kevino

1 edit in branches/wx-port-alpha/trunk/WebKit/wx/bindings/python/webview.i

Add checks to ensure that you won't get a crash when trying to create a wxWebView or wxWebFrame without a running wx app instance.

5:34 PM Changeset in webkit [24031] by kevino

1 edit in branches/wx-port-alpha/trunk/WebCore/platform/graphics/wx/ColorWx.cpp

Fix wxColor->Color conversion.

4:12 PM Changeset in webkit [24030] by andersca

2 edits in trunk/WebCore

Build fix.

loader/FrameLoader.cpp: (WebCore::FrameLoader::init):

4:04 PM Changeset in webkit [24029] by ggaren

2 edits in trunk/LayoutTests

Forgot to add this file for the test for <rdar://problem/5303566> REGRESSION:
XMLHttpRequest.responseXML. Darin added a dummy file before -- here's the
real one.

fast/loader/resources/plist.app:

3:56 PM Changeset in webkit [24028] by weinig

8 edits in trunk/WebCore

Reviewed by Geoff.

Add focus, blur and close functions to DOMWindow implementation class
and move bindings from Window to JSDOMWindow.

bindings/js/JSDOMWindowCustom.cpp: (WebCore::JSDOMWindow::customGetOwnPropertySlot):
bindings/js/kjs_window.cpp: (KJS::Window::getOwnPropertySlot): (KJS::WindowFunc::callAsFunction):
bindings/js/kjs_window.h: (KJS::Window::):
bindings/scripts/CodeGeneratorJS.pm: Move prototype function declaration from the implementation to the header.
page/DOMWindow.cpp: (WebCore::DOMWindow::focus): (WebCore::DOMWindow::blur): (WebCore::DOMWindow::close):
page/DOMWindow.h:
page/DOMWindow.idl:

3:52 PM Changeset in webkit [24027] by justing

5 edits
4 adds in trunk

LayoutTests:

Reviewed by Harrison.

<rdar://problem/5279521> REGRESSION: In Mail, a crash occurs at WebCore::Range::startPosition() when attempting to apply alignment to selected image/text in message body

Demonstrates bug:

editing/style/5279521-expected.checksum: Added.
editing/style/5279521-expected.png: Added.
editing/style/5279521-expected.txt: Added.
editing/style/5279521.html: Added.

Fixed:

fast/forms/radio-check-click-and-drag-expected.txt:
fast/forms/textarea-type-spaces-expected.txt:

WebCore:

Reviewed by Harrison.

<rdar://problem/5279521> REGRESSION: In Mail, a crash occurs at WebCore::Range::startPosition() when attempting to apply alignment to selected image/text in message body

editing/TextIterator.cpp: (WebCore::TextIterator::exitNode): Previously emitted content shouldn't have to come from a text node in order to emit a '\n' as we exit block containers.

3:50 PM Changeset in webkit [24026] by andersca

2 edits in trunk/WebCore

Reviewed by Geoff.

<rdar://problem/5315284> REGRESSION: Leopard Launch Time - plugins are being accessed at startup.

Set a response that has a MIME type of "text/html" on the document loader so WebKit won't refresh the
plugin database looking for a plug-in that can handle the document's MIME type (which will be null anyway).

loader/FrameLoader.cpp: (WebCore::FrameLoader::init):

3:34 PM Changeset in webkit [24025] by andersca

2 edits in trunk/WebCore

Reviewed by Oliver.

<rdar://problem/5106788>
WebKit loads user style sheets too late as seen in Mail (breaks DataDetectors; Mail flashes wrong font)

Let the document know when the user style sheet is loading so that it try not to display anything
until the user style sheet has finished loading.

page/Frame.cpp: (WebCore::UserStyleSheetLoader::UserStyleSheetLoader): (WebCore::UserStyleSheetLoader::~UserStyleSheetLoader): (WebCore::UserStyleSheetLoader::setCSSStyleSheet): (WebCore::Frame::setUserStyleSheetLocation):

2:02 PM Changeset in webkit [24024] by aroben

2 edits in trunk/WebKit/win

Fix a leak in WebView::setToolTip

Reviewed by Beth.

WebView.cpp:

(WebView::setToolTip):

1:43 PM Changeset in webkit [24023] by thatcher

2 edits in trunk/WebKit

Reviewed by John.

<rdar://problem/5314993> Shiira 2.1 throws an exception open a new window: -window: unrecognized selector

Add an empty implementation of this method to prevent Shiira from throwing an exception.
Also log that this method is obsolete and the class will be removed.

WebInspector/WebInspector.mm: (-window):

1:16 PM Changeset in webkit [24022] by aroben

4 edits in trunk/WebKit/win

Fix Bug 14143: Tooltips not displayed on Windows
http://bugs.webkit.org/show_bug.cgi?id=14143
<rdar://problem/4719799>

Reviewed by Oliver.

WebChromeClient.cpp:

(WebChromeClient::setToolTip): Call up to WebView.

WebView.cpp:

(WebView::WebView): Initialize m_toolTipHwnd member.
(WebView::initWithFrame): Set up the tool tip window.
(initCommonControls): Added.
(WebView::initializeToolTipWindow): Added.
(WebView::setToolTip): Set the tool tip text and enable/disable the
tool tip.

WebView.h: Added/updated declarations.

1:11 PM Applications using WebKit edited by jalkut@red-sweater.com

Update MarsEdit info with info and link pertinent to Red Sweater … (diff)

11:09 AM Changeset in webkit [24021] by weinig

4 edits in trunk/WebCore

Reviewed by Darin.

Move JSDOMWindow overriding getOwnPropertySlot logic from KJS::Window
up into JSDOMWindow. No change in functionality so no testcase required.

bindings/js/JSDOMWindowCustom.cpp: (WebCore::JSDOMWindow::customGetOwnPropertySlot):
bindings/js/kjs_window.cpp: (KJS::Window::getOwnPropertySlot):
bindings/js/kjs_window.h:

10:51 AM Changeset in webkit [24020] by darin

3 edits
1 copy
2 adds in trunk/LayoutTests

tweaked IDN test and checked in both Tiger results and Leopard

fast/encoding/idn-security.html: Changed test to only test lowercase "a" since the behavior for uppercase letters has changed in the new ICU. Also added code to detect the older ICU and expect different results.

fast/encoding/idn-security-expected.txt: Regenerated results.

mac/leopard/fast/encoding/idn-security-expected.txt: Copied from fast/encoding/idn-security-expected.txt. The modified.

10:02 AM Changeset in webkit [24019] by weinig

2 edits in trunk/WebCore

Reviewed and Landed by Sam.

fix bug 14465 for window.showModalDialog

bindings/js/kjs_window.cpp: (KJS::parseModalDialogFeatures): (KJS::showModalDialog):

9:14 AM Changeset in webkit [24018] by weinig

2 edits
1 add in trunk/WebCore

Reviewed by Mitz.

Patch for http://bugs.webkit.org/show_bug.cgi?id=14465
window.open() uses the literal "undefined" as the URL

bindings/js/kjs_window.cpp: (KJS::WindowFunc::callAsFunction): add checks for undefined and null values when setting the url.
manual-tests/open-url-undefined.html: Added.

6:51 AM Changeset in webkit [24017] by bdash

4 edits in trunk/WebCore

2007-07-05 Holger Hans Peter Freyther <zecke@selfish.org>

Reviewed by Rob Buis.

Rework FontData and FontPlatformData for the Gdk work to make valgrind
happy.

Move destroying of FontConfig and Cairo objects held in FontPlatformData
to FontData::platformDestroy. The ownership of these objects is held by
FontData. This is following the Mac and Qt port at least.

Currently it is possible that FontPlatformData will not create cairo resources
in this case at least make sure the variables are initialized to 0.
Initialize x,y of the cairo_glyph_t with zero to avoid extents.x_advance depending
on subtracing with not defined values.

FontPlatformData::hash. Using FontDescription is a bad idea as we don't have a clue
about how it is packed, if holes are present, etc. The FcPattern and the m_fontMatrix
should be good enough to describe the Font. All other objects are created using these
variables

Replace C-style casts with C++ casts, in this case reinterpret_cast.

platform/gdk/FontDataGdk.cpp: (WebCore::FontData::platformInit): Don't assert (WebCore::FontData::platformDestroy): Take over ownership (WebCore::FontData::platformWidthForGlyph): initialize (WebCore::FontData::setFont): be paranoid and assert
platform/gdk/FontPlatformData.h: (WebCore::FontPlatformData::FontPlatformData): Initialize everything (WebCore::FontPlatformData::hash): make valgrind happy and don't use m_fontDescription
platform/gdk/FontPlatformDataGdk.cpp: (WebCore::FontPlatformData::FontPlatformData): (WebCore::FontPlatformData::~FontPlatformData): FontData::platformDestroy will do this (WebCore::FontPlatformData::list): Don't use C-style casts. (WebCore::FontPlatformData::setFont): m_fontFace, m_fontMatrix, m_options might not be initialized (WebCore::FontPlatformData::operator==): Don't use C-style casts.

6:47 AM Changeset in webkit [24016] by bdash

4 edits in trunk/WebCore

2007-07-05 Alp Toker <alp.toker@collabora.co.uk>

Reviewed by Mark Rowe.

http://bugs.webkit.org/show_bug.cgi?id=14445
GraphicsContextCairo code style cleanup

Follow the cr naming convention for cairo_t*.
Remove trailing whitespace.
Replace C casts with C++ casts.
Avoid needless casts.

platform/graphics/cairo/GraphicsContextCairo.cpp: (WebCore::GraphicsContextPlatformPrivate::GraphicsContextPlatformPrivate): (WebCore::GraphicsContextPlatformPrivate::~GraphicsContextPlatformPrivate): (WebCore::GraphicsContext::GraphicsContext): (WebCore::GraphicsContext::platformContext): (WebCore::GraphicsContext::savePlatformState): (WebCore::GraphicsContext::restorePlatformState): (WebCore::GraphicsContext::drawRect): (WebCore::adjustLineToPixelBoundaries): (WebCore::GraphicsContext::drawLine): (WebCore::GraphicsContext::drawEllipse): (WebCore::GraphicsContext::strokeArc): Save/restore the graphics state. (WebCore::GraphicsContext::drawConvexPolygon): (WebCore::GraphicsContext::fillRect): (WebCore::GraphicsContext::clip): (WebCore::GraphicsContext::drawFocusRing): (WebCore::GraphicsContext::roundToDevicePixels): (WebCore::GraphicsContext::translate): (WebCore::GraphicsContext::origin): (WebCore::GraphicsContext::setPlatformStrokeThickness): (WebCore::GraphicsContext::setPlatformStrokeStyle): (WebCore::GraphicsContext::setPlatformFont): (WebCore::GraphicsContext::concatCTM): (WebCore::GraphicsContext::addInnerRoundedRectClip): (WebCore::GraphicsContext::beginTransparencyLayer): (WebCore::GraphicsContext::endTransparencyLayer): (WebCore::GraphicsContext::clearRect): (WebCore::GraphicsContext::setLineCap): (WebCore::GraphicsContext::setLineJoin): (WebCore::GraphicsContext::setMiterLimit): (WebCore::GraphicsContext::setCompositeOperation): (WebCore::GraphicsContext::beginPath): (WebCore::GraphicsContext::addPath): (WebCore::GraphicsContext::rotate): (WebCore::GraphicsContext::scale):
platform/graphics/cairo/ImageBufferCairo.cpp: (WebCore::ImageBuffer::ImageBuffer):
platform/graphics/cairo/ImageCairo.cpp: Avoid the confusing ctxt/context naming scheme. (WebCore::BitmapImage::draw): (WebCore::Image::drawPattern):

6:45 AM Changeset in webkit [24015] by bdash