15 years agoLayoutTests:
mjs [Wed, 18 Jul 2007 09:07:30 +0000 (09:07 +0000)]

        Reviewed by Oliver.

        - test for <rdar://problem/5340603> REGRESSION: javascript being written to the screen in Safari

        * fast/tokenizer/badscript-expected.txt: Added.
        * fast/tokenizer/badscript.html: Added.


        Reviewed by Oliver.

        - fixed <rdar://problem/5340603> REGRESSION: javascript being written to the screen in Safari

        Test Case: fast/tokenizer/badscript.html

        * html/HTMLTokenizer.cpp:
        (WebCore::HTMLTokenizer::parseTag): Don't apply our self-closing
        <script> quirk in cases where the / appears in a mangled attribtue

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24405 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Maciej Stachowiak.
tristan [Wed, 18 Jul 2007 08:20:28 +0000 (08:20 +0000)]
    Reviewed by Maciej Stachowiak.

        <rdar://problem/5341334> Alt-clicking a link in Safari does not register original URL info with gatekeeper
        <rdar://problem/5342570> REGRESSION: A hang occurs when attempting to open a attached file from a .Mac web mail message

        * WebCoreSupport/WebFrameLoaderClient.h:
        Added a new method, setOriginalURLForDownload, for both download()
        and startDownload() to share.

        * WebCoreSupport/WebFrameLoaderClient.mm:
        Moved all the gatekeeper code into setOriginalURLForDownload.

        Same code that was in download() but moved into a single place for both
        startDownload and download to use.  Also returned the boolean logic
        for detecting _wasUserGesture back to it's previous state, since my prior change
        was incorrect.  Additionally I found a loop index bug with backListCount and fixed it.

        Calls out to setOriginalURLForDownload after a download is created.

        * WebView/WebView.mm:
        (-[WebView _downloadURL:]):
        Changed to return the WebDownload object created inside _downloadURL

        * WebView/WebViewInternal.h:
        Changed _downloadURL to return the WebDownload it creates.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24404 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoCleanliness is next to ... well, it's a Good Thing
bdash [Wed, 18 Jul 2007 06:36:58 +0000 (06:36 +0000)]
Cleanliness is next to ... well, it's a Good Thing

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24403 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Blind Windows build fix from r24395
beidson [Wed, 18 Jul 2007 06:33:00 +0000 (06:33 +0000)]
    Blind Windows build fix from r24395

        * WebView.cpp:
        (WebView::initWithFrame): BSTR, not BString

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24402 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Hyatt.
weinig [Wed, 18 Jul 2007 06:27:57 +0000 (06:27 +0000)]
    Reviewed by Hyatt.

        Don't set link targets to "_blank" on middle-click; rely on the policy
        delegate to do the right thing.

        * html/HTMLAnchorElement.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24401 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago2007-07-18 Mark Rowe <mrowe@apple.com>
bdash [Wed, 18 Jul 2007 06:20:34 +0000 (06:20 +0000)]
2007-07-18  Mark Rowe  <mrowe@apple.com>

        Update path to the Drosera Xcode project now that it has moved into a subdirectory.

        * BuildSlaveSupport/build-launcher-app:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24400 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Maciej.
oliver [Wed, 18 Jul 2007 06:01:24 +0000 (06:01 +0000)]
    Reviewed by Maciej.

        Fix for <rdar://problem/4799967> Safari and other WebKit
        apps sometimes crash in WebCore::FontData::platformInit

        Always retain/release our font handle, instead of only retaining
        in the presence of GC.

        * platform/mac/FontPlatformData.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24399 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoLayoutTests:
weinig [Wed, 18 Jul 2007 05:40:54 +0000 (05:40 +0000)]

        Reviewed by Geoff Garen.

        Test for <rdar://problem/5333957>

        * http/tests/security/cross-frame-access-protocol.html: Fix a few typos.
        * http/tests/security/cross-frame-access-selection-expected.txt: Added.
        * http/tests/security/cross-frame-access-selection.html: Added.
        * http/tests/security/resources/foreign-domain-iframe-for-selection-test.html: Added.


        Reviewed by Geoff Garen.

        Fix for <rdar://problem/5333957>

        Original patch by Feng Qian.

        - Clear the selection object when disconnecting a frame.

        Test: http/tests/security/cross-frame-access-selection.html

        * loader/FrameLoader.cpp:
        * page/DOMWindow.cpp:
        * page/DOMWindow.h:
        * page/Frame.cpp:
        * page/Frame.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24398 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoLayoutTests:
oliver [Wed, 18 Jul 2007 04:44:18 +0000 (04:44 +0000)]

        Reviewed by Sam.

        Correct layout test.

        * fast/events/keydown-keypress-preventDefault-expected.txt:
        * fast/events/keydown-keypress-preventDefault.html:


        Reviewed by Sam.

        Fix for <rdar://problem/5154518> REGRESSION: Hitting return does
        nothing in ruby terminal at http://tryruby.hobix.com/

        Calling event.preventDefault() during a keydown event should not
        prevent the keypress event from firing.

        * WebCore.xcodeproj/project.pbxproj:
        * page/EventHandler.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24397 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoLayoutTests:
andersca [Wed, 18 Jul 2007 03:46:10 +0000 (03:46 +0000)]

        Reviewed by Geoff.

        REGRESSION: Comments fields on ircimages not showing up in nightly

        * fast/frames/iframe-set-inner-html-expected.txt: Added.
        * fast/frames/iframe-set-inner-html.html: Added.


        Reviewed by Geoff.

        REGRESSION: Comments fields on ircimages not showing up in nightly

        Postpone setting the frame name to just right before we request the frame. Otherwise,
        two calls to FrameTree::uniqueChildName will return the same frame name and in that case only
        one Frame object will actually be created and multiple iframe elements will refer to it.

        * html/HTMLFrameElementBase.cpp:
        * html/HTMLFrameElementBase.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24396 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoWebCore: 2007-07-17 Brady Eidson <beidson@apple.com>
beidson [Wed, 18 Jul 2007 02:39:03 +0000 (02:39 +0000)]
WebCore: 2007-07-17  Brady Eidson  <beidson@apple.com>

        Reviewed by Steve

        Required for <rdar://problem/4516185> - FTP Directory Listings

        * WebCore.vcproj/WebCore.vcproj: Added SharedBufferWin

        * platform/SharedBuffer.cpp: Windows doesn't get the default empty resource anymore
        * platform/win/SharedBufferWin.cpp: Added.

WebKit/win: 2007-07-17  Brady Eidson  <beidson@apple.com>

        Reviewed by Steve

        <rdar://problem/4516185> - FTP Directory Listings

        * Interfaces/IWebUIDelegate.idl: Add IWebUIDelegate2 method for getting the path to the template

        * WebView.cpp:
          (WebView::initWithFrame): Set the preference for the template path in WebCore when the first
          WebView is initialized

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24395 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoJavaScriptCore:
darin [Wed, 18 Jul 2007 02:25:38 +0000 (02:25 +0000)]

        Reviewed by Darin, Maciej, and Adam.

        Fixes <http://bugs.webkit.org/show_bug.cgi?id=9697>,
              the failure of ecma/GlobalObject/,
              the failure of ecma/LexicalConventions/7.7.3-1.js,
              and most of the failures of tests in ecma/TypeConversion/9.3.1-3.js.

        Bug 9697: parseInt results may be inaccurate for numbers greater than 2^53

        This patch also fixes similar issues in the lexer and UString::toDouble().

        * kjs/function.cpp:
        * kjs/function.h:
        * kjs/lexer.cpp:
        * kjs/ustring.cpp:
        * tests/mozilla/expected.html:


        Reviewed by Darin.

        Added tests for:

        Bug 9697: parseInt results may be inaccurate for numbers greater than 2^53

        * fast/js/numeric-conversion-expected.txt: Added.
        * fast/js/numeric-conversion.html: Added.
        * fast/js/resources/numeric-conversion.js: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24394 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Geoff.
thatcher [Tue, 17 Jul 2007 23:39:54 +0000 (23:39 +0000)]
    Reviewed by Geoff.

        <rdar://problem/5336267> loadData:MIMEType:textEncodingName:baseURL: doesn't like relative URLs

        Get the absoluteURL from any user supplied NSURL before we pass it down to WebCore.

        * WebView/WebFrame.mm:
        (-[WebFrame _loadURL:referrer:intoChild:]):
        (-[WebFrame _loadData:MIMEType:textEncodingName:baseURL:unreachableURL:]):
        * WebView/WebView.mm:
        (-[WebView userAgentForURL:]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24392 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoRemove WebKit.LP64.exp from te target membership.
thatcher [Tue, 17 Jul 2007 23:28:09 +0000 (23:28 +0000)]
Remove WebKit.LP64.exp from te target membership.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24391 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago2007-07-17 Ada Chan <adachan@apple.com>
adachan [Tue, 17 Jul 2007 22:08:07 +0000 (22:08 +0000)]
2007-07-17  Ada Chan  <adachan@apple.com>

        Rubbet-stamped by Adam.

        Fixed build.

        * Interfaces/IWebFramePrivate.idl:
        * WebContextMenuClient.cpp:
        * WebFrame.cpp:
        * WebFrame.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24389 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoFix the build.
treat [Tue, 17 Jul 2007 22:03:05 +0000 (22:03 +0000)]
Fix the build.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24388 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoMade the recently added functions static so that they'll have internal linkage.
justing [Tue, 17 Jul 2007 21:55:05 +0000 (21:55 +0000)]
Made the recently added functions static so that they'll have internal linkage.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24387 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago2007-07-17 Holger Hans Peter Freyther <zecke@selfish.org>
zecke [Tue, 17 Jul 2007 21:28:08 +0000 (21:28 +0000)]
2007-07-17  Holger Hans Peter Freyther  <zecke@selfish.org>

        Blind build fix for Qt after r24366 by adding the additional
        WebCore::ResourceRequest& parameter to the download method.

        * WebCoreSupport/FrameLoaderClientQt.cpp:
        * WebCoreSupport/FrameLoaderClientQt.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24385 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Tim Hatcher
beidson [Tue, 17 Jul 2007 21:27:27 +0000 (21:27 +0000)]
    Reviewed by Tim Hatcher

        Tweak the class names we give to the icon table cell to make it easier to style

        * loader/FTPDirectoryDocument.cpp:
          (WebCore::FTPDirectoryTokenizer::appendEntry): The the table cell a "IconCell" class name and
          either a "FileType" or "DirectoryType" class name depending on the type

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24384 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoLayoutTests:
ggaren [Tue, 17 Jul 2007 21:27:01 +0000 (21:27 +0000)]

        Reviewed by Kevin Decker.

        Layout test that demonstrates why an old ASSERT in WebCore was wrong.

        * ChangeLog:
        * fast/history/go-back-to-changed-name-expected.txt: Added.
        * fast/history/go-back-to-changed-name.html: Added.
        * fast/history/resources/go-back-to-changed-name2.html: Added.
        * fast/history/resources/go-back-to-changed-name3.html: Added.
        * http/tests/security/cross-frame-access-data-url-expected.txt: Added.
        * http/tests/security/cross-frame-access-data-url.html: Added.


        Reviewed by NOBODY.

        Build fix after r24366 add an additional ResourceRequest parameter to
        the download method.

        * loader/gdk/FrameLoaderClientGdk.cpp:
        * loader/gdk/FrameLoaderClientGdk.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24383 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago2007-07-17 Holger Hans Peter Freyther <zecke@selfish.org>
zecke [Tue, 17 Jul 2007 21:22:40 +0000 (21:22 +0000)]
2007-07-17  Holger Hans Peter Freyther  <zecke@selfish.org>

        Reviewed by NOBODY.

        Build fix after r24366 add an additional ResourceRequest parameter to
        the download method.

        * loader/gdk/FrameLoaderClientGdk.cpp:
        * loader/gdk/FrameLoaderClientGdk.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24382 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoFix Bug 14324: Cannot remove/customize the "Inspect Element" contextual menu item
aroben [Tue, 17 Jul 2007 21:17:25 +0000 (21:17 +0000)]
Fix Bug 14324: Cannot remove/customize the "Inspect Element" contextual menu item


        Fix Bug 14324: Cannot remove/customize the "Inspect Element" contextual menu item

        Reviewed by Tim.

        No regression test possible.

        * page/ContextMenuController.cpp:
        (WebCore::ContextMenuController::handleContextMenuEvent): Add the
        Inspect Element item to the menu before passing it to the
        ContextMenuClient so that the UI delegate has a chance to
        modify/remove the item.


        Fix Bug 14324: Cannot remove/customize the "Inspect Element" contextual menu item

        Only clients linking against new versions of WebKit will see the item.
        I've maintained our behavior for old clients of not including the
        Inspect Element item in the menu items passed to the UI delegate.

        Reviewed by Tim.

        * Misc/WebKitVersionChecks.h: Added a new constant.
        * WebCoreSupport/WebContextMenuClient.mm:
        (isPreInspectElementTagClient): Added.
        (fixMenusToSendToOldClients): Return an array of items that should be
        appended to the menu received from the delegate.
        (fixMenusReceivedFromOldClients): Append the saved items to the array.
        (WebContextMenuClient::getCustomMenuFromDefaultItems): Retain/release
        the saved items.


        Fix Bug 14324: Cannot remove/customize the "Inspect Element" contextual menu item

        If we detect that we're running against the Safari 3 Beta, we add back
        the Inspect Element menu item after passing it off to the delegate
        because Safari's UI delegate will remove it.

        Reviewed by Tim.

        * WebContextMenuClient.cpp:
        (isPreInspectElementTagSafari): Added.
        (fixMenuReceivedFromOldSafari): Added.
        (WebContextMenuClient::getCustomMenuFromDefaultItems): Call
        fixMenuReceivedFromOldSafari before returning the new menu.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24381 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoRemove ContextMenuClient::shouldIncludeInspectElementItem
aroben [Tue, 17 Jul 2007 21:17:14 +0000 (21:17 +0000)]
Remove ContextMenuClient::shouldIncludeInspectElementItem


        Remove ContextMenuClient::shouldIncludeInspectElementItem

        Reviewed by Tim.

        No regression test possible.

        * page/ContextMenuClient.h:
        * page/ContextMenuController.cpp:
        (WebCore::ContextMenuController::handleContextMenuEvent): Use Settings
        to determine whether to include the Inspect Element item.
        * page/gdk/ContextMenuClientGdk.cpp:
        * page/gdk/ContextMenuClientGdk.h:
        * platform/graphics/svg/SVGImageEmptyClients.h:


        Remove WebContextMenuClient::shouldIncludeInspectElementItem

        Reviewed by Tim.

        * WebCoreSupport/WebContextMenuClient.h:
        * WebCoreSupport/WebContextMenuClient.mm:


        Remove WebContextMenuClient::shouldIncludeInspectElementItem

        Reviewed by Tim.

        * WebContextMenuClient.cpp:
        * WebContextMenuClient.h:


        Remove ContextMenuClientQt::shouldIncludeInspectElementItem

        Reviewed by Tim.

        * WebCoreSupport/ContextMenuClientQt.cpp:
        * WebCoreSupport/ContextMenuClientQt.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24380 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoInitialize ContextMenuController::m_page
aroben [Tue, 17 Jul 2007 21:17:00 +0000 (21:17 +0000)]
Initialize ContextMenuController::m_page

Somehow this variable had never been initialized before.

Reviewed by Tim.

* page/ContextMenuController.cpp:
* page/ContextMenuController.h:
* page/Page.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24379 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoAdd/initialize Settings::developerExtrasEnabled
aroben [Tue, 17 Jul 2007 21:16:21 +0000 (21:16 +0000)]
Add/initialize Settings::developerExtrasEnabled


        Add Settings::developerExtrasEnabled

        Reviewed by Tim.

        * WebCore.exp:
        * page/Settings.cpp:
        * page/Settings.h:


        Initialize Settings::developerExtrasEnabled

        Reviewed by Tim.

        * WebView/WebView.mm:
        (-[WebView _updateWebCoreSettingsFromPreferences:]):


        Initialize Settings::developerExtrasEnabled

        Reviewed by Tim.

        * WebView.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24378 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Build fix.
ggaren [Tue, 17 Jul 2007 21:13:20 +0000 (21:13 +0000)]
    Build fix.

        * WebContextMenuClient.cpp:
        (WebContextMenuClient::searchWithGoogle): Pass false for lockHistory
        like we do elsewhere.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24377 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Antti Koivisto.
ggaren [Tue, 17 Jul 2007 21:10:17 +0000 (21:10 +0000)]
    Reviewed by Antti Koivisto.

        Added watchdog timer to waitUntilDone to prevent a run-away test from
        hanging the test harness.

        * DumpRenderTree/DumpRenderTree.m:
        (-[LayoutTestController waitUntilDone]):
        (-[LayoutTestController waitUntilDoneWatchdogFired]):
        * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24376 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoLayoutTests:
justing [Tue, 17 Jul 2007 21:05:21 +0000 (21:05 +0000)]

        Reviewed by Antti.

        Whitespace can't be removed when editing text pasted into from web page

        * editing/deleting/5206311-1-expected.checksum: Added.
        * editing/deleting/5206311-1-expected.png: Added.
        * editing/deleting/5206311-1-expected.txt: Added.
        * editing/deleting/5206311-1.html: Added.
        * editing/deleting/5206311-2-expected.checksum: Added.
        * editing/deleting/5206311-2-expected.png: Added.
        * editing/deleting/5206311-2-expected.txt: Added.
        * editing/deleting/5206311-2.html: Added.


        Reviewed by Antti.

        Whitespace can't be removed when editing text pasted into from web page

        The whitespace comes from table rows that aren't being removed after they are
        emptied out during a deletion.

        * editing/DeleteSelectionCommand.cpp:
        (WebCore::DeleteSelectionCommand::initializePositionData): Remember the table
        rows that contained the start and end of the selection.
        Remove table rows that come before the end of the selection
        that was removed (except for the row that contained the start of the selection
        and any before that).
        Remove table rows that are come after the start of the selection
        that was removed (except for the row that contained the end of the selection
        and any after that).
        (WebCore::DeleteSelectionCommand::doApply): Call the new function.
        * editing/DeleteSelectionCommand.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24375 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Adam.
kmccullo [Tue, 17 Jul 2007 21:03:21 +0000 (21:03 +0000)]
    Reviewed by Adam.

        - Implemented cross-platform code for functions calling into the JavaScript.

        * Drosera/DebuggerDocument.cpp:
        * Drosera/DebuggerDocument.h:
        * Drosera/mac/DebuggerDocumentMac.mm:
        (-[DebuggerDocumentMac pause:]):
        (-[DebuggerDocumentMac resume:]):
        (-[DebuggerDocumentMac stepInto:]):
        (-[DebuggerDocumentMac stepOver:]):
        (-[DebuggerDocumentMac stepOut:]):
        (-[DebuggerDocumentMac showConsole:]):
        (-[DebuggerDocumentMac closeCurrentFile:]):
        (-[DebuggerDocumentMac webView:didLoadMainResourceForDataSource:]):
        (-[DebuggerDocumentMac webView:didParseSource:baseLineNumber:fromURL:sourceId:forWebFrame:]):
        (-[DebuggerDocumentMac webView:didEnterCallFrame:sourceId:line:forWebFrame:]):
        (-[DebuggerDocumentMac webView:willExecuteStatement:sourceId:line:forWebFrame:]):
        (-[DebuggerDocumentMac webView:willLeaveCallFrame:sourceId:line:forWebFrame:]):
        (-[DebuggerDocumentMac webView:exceptionWasRaised:sourceId:line:forWebFrame:]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24374 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Build fix.
ggaren [Tue, 17 Jul 2007 20:57:49 +0000 (20:57 +0000)]
    Build fix.

        * plugins/win/PluginViewWin.cpp: Pass false for lockHistory like we
        do elsewhere.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24373 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago2007-07-17 Holger Hans Peter Freyther <zecke@selfish.org>
zecke [Tue, 17 Jul 2007 20:51:53 +0000 (20:51 +0000)]
2007-07-17  Holger Hans Peter Freyther  <zecke@selfish.org>

        Reviewed by Maciej.

        Implement the callback responsible for handling HTTP headers. This
        callback gets called for every header.
        We will add these headers to our ResourceResponse and on the 'end-of-header'
        indicator we will update the ResourceResponse and dispatch it.

        This patch adds various set methods to ResourceResponse. This improves
        the readability of the headerCallback and avoids storing a CURL handle inside
        the ResourceResponse which would be needed to implement ResourceResponse::doUpdateResourceResponse

        Add a destructor for ResourceHandleManager which would free the resources and remove
        a unused variable.

        * platform/network/ResourceHandleInternal.h:
        * platform/network/ResourceResponse.cpp:
        * platform/network/ResourceResponse.h:
        * platform/network/curl/ResourceHandleCurl.cpp:
        * platform/network/curl/ResourceHandleManager.cpp:
        * platform/network/curl/ResourceHandleManager.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24372 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoLayoutTests:
weinig [Tue, 17 Jul 2007 20:31:58 +0000 (20:31 +0000)]

        Reviewed by Adam Roben.

        * fast/dom/Window/alert-undefined-expected.txt: Added.
        * fast/dom/Window/alert-undefined.html: Added.


        Reviewed by Adam Roben.

        Convert JS undefined to the empty string when calling alert, confirm, and prompt.

        Test: fast/dom/Window/alert-undefined.html

        * page/DOMWindow.idl:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24371 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Build fix.
ggaren [Tue, 17 Jul 2007 18:59:48 +0000 (18:59 +0000)]
    Build fix.

        * WebView/WebFrame.mm:
        (-[WebFrame _loadURL:referrer:intoChild:]):
        * WebView/WebFramePrivate.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24369 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoAdding /VWebKit.LP64.exp to the project.
thatcher [Tue, 17 Jul 2007 18:57:18 +0000 (18:57 +0000)]
Adding /VWebKit.LP64.exp to the project.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24368 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoLayoutTests:
ggaren [Tue, 17 Jul 2007 18:40:22 +0000 (18:40 +0000)]

        Reviewed by Kevin McCulough.

        * http/tests/navigation/multiple-back-forward-entries-expected.txt:
        Updated results to match new behavior, which is correct (multiple
        history entries should not be created).


        Reviewed by Kevin McCulough.

        Fixed failing layout test.

        Renamed "FrameLoadTypeInternal" to "FrameLoadTypeRedirectWithLockedHistory".
        This is almost certainly not a correct name for the original intent
        of this load type, but it's a correct name for its current behavior.

        When history is supposed to be locked, and no current history item
        exists, only add a new history item if you're the top-level frame.

        This rule, combined with the rule that setting iframe.src should not
        create a history item, is pretty hackish. On the other hand, it
        matches WinIE and doesn't cause any test regressions. So I'm OK with
        it as a short-term solution until the solution described in
        <rdar://problem/5339292> can be implemented.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24367 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Maciej Stachowiak.
tristan [Tue, 17 Jul 2007 18:27:04 +0000 (18:27 +0000)]
    Reviewed by Maciej Stachowiak.

        <rdar://problem/5294691> Source of file is misrepresented if downloaded by typing in URL in Safari address bar

        * loader/FrameLoaderClient.h:
        * loader/MainResourceLoader.cpp:
        Added an extra request parameter, initialRequest, to the download() method
        by passing up the resource handle's request.

        * platform/graphics/svg/SVGImageEmptyClients.h:
        Added an extra parameter to the empty SVG download method.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24366 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Maciej Stachowiak.
tristan [Tue, 17 Jul 2007 18:24:36 +0000 (18:24 +0000)]
    Reviewed by Maciej Stachowiak.

        <rdar://problem/5294691> Source of file is misrepresented if downloaded by typing in URL in Safari address bar

        * WebCoreSupport/WebFrameLoaderClient.h:
        * WebCoreSupport/WebFrameLoaderClient.mm:
        Revised code to check the initial request's referrer before assuming it has a history to check.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24365 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by home-bradee.
antti [Tue, 17 Jul 2007 18:06:29 +0000 (18:06 +0000)]
    Reviewed by home-bradee.

        Icon database uses too much memory

        XRaying Safari startup memory consumption revealed that icon database is eating quite
        a bit of RAM if Icon.db is large (which it probably is if it has been in use for a while,
        mine used for getting figures below was 2.6MB).

        Note that the wins are less impressive with smaller Icon.db.

        This patch addresses three separate issues

        - SQLite fails to free the memory used by temporary tables. Icon database uses a temporary table
          on startup for pruning unused page urls. This wastes around 1MB. Addressed by rewriting
          pruning so it does not need a temporary table. The new method is also quite a bit faster speeding
          up Safari launch time by around 100ms
        - SQLite has it's own memory cache limited by default to 3MB. Icon database does not really need that much.
          Dropped the cache size to 300kB saving ~1MB on startup.
          Smaller cache slows down startup by ~30ms (more than compensated by faster pruning above)
        - Don't populate m_pageURLToIconURLMap with all urls from database on startup, instead let it get populated
          when urls are accessed (user opens history menu for example). This shouldn't have any real performance impact
          as the accesses are icon loads that need to hit the database anyway. This saves ~700kB.

        All in all with this Icon.db these changes reduce allocated memory by around 2.7MB on startup. Release build
        Safari RPRVT (empty start page) goes from 12.4MB to 10.4MB (TCMalloc pooling probably explaining why the win
        looks bit smaller here).

        * loader/icon/IconDatabase.cpp:
        * loader/icon/IconDatabase.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24364 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoLayoutTests:
darin [Tue, 17 Jul 2007 17:48:32 +0000 (17:48 +0000)]

        Reviewed by Maciej.
        Landed by Darin.

        - test for http://bugs.webkit.org/show_bug.cgi?id=14569
          Repro crash due to saved state not being cleared when select and textarea elements are adopted by another document
          and for http://bugs.webkit.org/show_bug.cgi?id=13701
          REGRESSION (r21431): Reproducible crash resulting from calling adoptNode on a password field

        * fast/forms/saved-state-adoptNode-crash-expected.txt: Added.
        * fast/forms/saved-state-adoptNode-crash.html: Added.


        Reviewed by Mitz.

        - fix http://bugs.webkit.org/show_bug.cgi?id=14569
          Repro crash due to saved state not being cleared when select and textarea elements
          are adopted by another document

        Test: fast/forms/saved-state-adoptNode-crash.html

        * dom/Document.h:
        (WebCore::Document::registerFormElementWithState): Updated parameter type to be
        (WebCore::Document::unregisterFormElementWithState): Ditto.
        Also removed isFormElementRegistered function and changed the type of the
        m_formElementsWithState ListHashSet.
        * dom/Document.cpp: (WebCore::Document::formElementsState): Updated to use new type,
        for set contents and to use new saveState() function and only save state for elements
        that return true.

        * html/HTMLGenericFormElement.h: Added HTMLFormControlElementWithState class,
        derived from HTMLGenericFormElement, and moved down all the behavior related to
        saving form control element state. Also removed the unnecessary default parameter
        to the HTMLGenericFormElement constructor.
        * html/HTMLGenericFormElement.cpp:
        (WebCore::HTMLFormControlElementWithState::HTMLFormControlElementWithState): Added.
        Calls registerFormElementWithState.
        (WebCore::HTMLFormControlElementWithState::~HTMLFormControlElementWithState): Added.
        Calls unregisterFormElementWithState.
        (WebCore::HTMLFormControlElementWithState::willMoveToNewOwnerDocument): Added.
        Calls unregisterFormElementWithState.
        (WebCore::HTMLFormControlElementWithState::didMoveToNewOwnerDocument): Added.
        Calls registerFormElementWithState.
        (WebCore::HTMLFormControlElementWithState::closeRenderer): Moved to this class from
        the base class.

        * html/HTMLInputElement.h: Updated the base class name and overrode saveState instead
        of stateValue.
        * html/HTMLInputElement.cpp: Updated the base class name.
        (WebCore::HTMLInputElement::init): Removed the call to registerFormElementWithState.
        (WebCore::HTMLInputElement::~HTMLInputElement): Removed the call to
        unregisterFormElementWithState. Also changed the code to only call the
        unregisterForDidRestoreFromCacheCallback function for type PASSWORD (consistency
        and efficiency).
        (WebCore::HTMLInputElement::setInputType): Removed calls to registerFormElementWithState
        and unregisterFormElementWithState; we now let the element stay registered and simply
        don't save or restore state for password fields.
        (WebCore::HTMLInputElement::saveState): Updated function name and signature to replace
        the old stateValue function. Returns false for password fields.
        (WebCore::HTMLInputElement::willMoveToNewOwnerDocument): Removed the call to
        (WebCore::HTMLInputElement::didMoveToNewOwnerDocument): Removed the call to

        * html/HTMLSelectElement.h: Updated the base class name and overrode saveState instead
        of stateValue.
        * html/HTMLSelectElement.cpp: Updated the base class name.
        (WebCore::HTMLSelectElement::HTMLSelectElement): Removed the call to
        (WebCore::HTMLSelectElement::~HTMLSelectElement): Removed the call to
        (WebCore::HTMLSelectElement::saveState): Updated function name and signature to
        replace the old stateValue function.
        (WebCore::HTMLSelectElement::setLength): Updated to use document() instead of
        ownerDocument(); the latter is just a slower virtual-function way of doing the
        same thing in all cases except for Document and XPathNamespace.

        * html/HTMLTextAreaElement.h: Update the base class name, overrode saveState instead
        of stateValue, and removed the now-unnecessary explicit destructor.
        * html/HTMLTextAreaElement.cpp: Updated the base class name.
        (WebCore::HTMLTextAreaElement::HTMLTextAreaElement): Removed the call to
        (WebCore::HTMLTextAreaElement::saveState): Updated function name and signature to replace
        the old stateValue function.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24363 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Darin.
weinig [Tue, 17 Jul 2007 17:46:39 +0000 (17:46 +0000)]
    Reviewed by Darin.

        - Make HTMLFrameOwnerElement::contentWindow() return a DOMWindow instead of a Frame
        - Remove the unnecessary hack in HTMLFrameElement.idl and HTMLIFrameElement.idl that
          converted from Frame to Window now that it is not necessary.
        - Remove now unused [NullCheck] extended attribute from CodeGeneratorJS.pm.

        * bindings/scripts/CodeGeneratorJS.pm:
        * html/HTMLDocument.idl:
        * html/HTMLFrameElement.idl:
        * html/HTMLFrameOwnerElement.cpp:
        * html/HTMLFrameOwnerElement.h:
        * html/HTMLIFrameElement.idl:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24362 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago * http/tests/navigation/multiple-back-forward-entries-expected.txt: Updated...
darin [Tue, 17 Jul 2007 16:55:22 +0000 (16:55 +0000)]
    * http/tests/navigation/multiple-back-forward-entries-expected.txt: Updated results to
        match what I see locally and what we see on the buildbot. Geoff, please double-check that
        these new results are correct.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24361 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Adam.
oliver [Tue, 17 Jul 2007 07:19:59 +0000 (07:19 +0000)]
    Reviewed by Adam.

        Fix for <rdar://problem/5339416> Candidate window does not
                appear in the correct location for japanese IME

        * WebView.cpp:
          We now use an exclusion zone so that the composition window never
          overlaps the composition string
          Correctly handle character offsets marked regions

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24357 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Forgot to do the delete half of the file move. Oops.
ggaren [Tue, 17 Jul 2007 07:02:25 +0000 (07:02 +0000)]
    Forgot to do the delete half of the file move. Oops.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24356 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Fixed a bunch of layout test failures I just caused.
ggaren [Tue, 17 Jul 2007 06:45:32 +0000 (06:45 +0000)]
    Fixed a bunch of layout test failures I just caused.

        Made the tests I just added into http test to make their results

        Also checked in the result change I forgot to check in last time.

        * fast/history/location-assign-adds-history-item-expected.txt: Removed.
        * fast/history/location-assign-adds-history-item.html: Removed.
        * fast/history/location-href-set-adds-history-item-expected.txt: Removed.
        * fast/history/location-href-set-adds-history-item.html: Removed.
        * fast/history/location-replace-adds-history-item-expected.txt: Removed.
        * fast/history/location-replace-adds-history-item.html: Removed.
        * fast/history/location-set-adds-history-item-expected.txt: Removed.
        * fast/history/location-set-adds-history-item.html: Removed.
        * fast/history/window-open-adds-history-item-expected.txt: Removed.
        * fast/history/window-open-adds-history-item.html: Removed.
        * fast/history/window-open-adds-history-item2-expected.txt: Removed.
        * fast/history/window-open-adds-history-item2.html: Removed.
        * http/tests/navigation/location-assign-adds-history-item-expected.txt: Copied from fast/history/location-assign-adds-history-item-expected.txt.
        * http/tests/navigation/location-assign-adds-history-item.html: Copied from fast/history/location-assign-adds-history-item.html.
        * http/tests/navigation/location-href-set-adds-history-item-expected.txt: Copied from fast/history/location-href-set-adds-history-item-expected.txt.
        * http/tests/navigation/location-href-set-adds-history-item.html: Copied from fast/history/location-href-set-adds-history-item.html.
        * http/tests/navigation/location-replace-adds-history-item-expected.txt: Copied from fast/history/location-replace-adds-history-item-expected.txt.
        * http/tests/navigation/location-replace-adds-history-item.html: Copied from fast/history/location-replace-adds-history-item.html.
        * http/tests/navigation/location-set-adds-history-item-expected.txt: Copied from fast/history/location-set-adds-history-item-expected.txt.
        * http/tests/navigation/location-set-adds-history-item.html: Copied from fast/history/location-set-adds-history-item.html.
        * http/tests/navigation/redirect-load-no-form-restoration-expected.txt:
        * http/tests/navigation/window-open-adds-history-item-expected.txt: Copied from fast/history/window-open-adds-history-item-expected.txt.
        * http/tests/navigation/window-open-adds-history-item.html: Copied from fast/history/window-open-adds-history-item.html.
        * http/tests/navigation/window-open-adds-history-item2-expected.txt: Copied from fast/history/window-open-adds-history-item2-expected.txt.
        * http/tests/navigation/window-open-adds-history-item2.html: Copied from fast/history/window-open-adds-history-item2.html.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24355 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Adam Roben.
weinig [Tue, 17 Jul 2007 06:42:03 +0000 (06:42 +0000)]
    Reviewed by Adam Roben.

        Move self, opener, parent, top, window, and frames from kjs_window
        to DOMWindow and auto-generate the bindings for them.

        * bindings/js/kjs_window.cpp:
        * bindings/js/kjs_window.h:
        * page/DOMWindow.cpp:
        * page/DOMWindow.h:
        * page/DOMWindow.idl:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24354 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoLayoutTests:
ggaren [Tue, 17 Jul 2007 06:17:50 +0000 (06:17 +0000)]

        Reviewed by Sam Weinig.

        Tests for <rdar://problem/5334483> REGRESSION: JavaScript-induced loads
        not added to back/forward list

        * fast/history/location-assign-adds-history-item-expected.txt: Added.
        * fast/history/location-assign-adds-history-item.html: Added.
        * fast/history/location-href-set-adds-history-item-expected.txt: Added.
        * fast/history/location-href-set-adds-history-item.html: Added.
        * fast/history/location-replace-adds-history-item-expected.txt: Added.
        * fast/history/location-replace-adds-history-item.html: Added.
        * fast/history/location-set-adds-history-item-expected.txt: Added.
        * fast/history/location-set-adds-history-item.html: Added.
        * fast/history/window-open-adds-history-item-expected.txt: Added.
        * fast/history/window-open-adds-history-item.html: Added.
        * fast/history/window-open-adds-history-item2-expected.txt: Added.
        * fast/history/window-open-adds-history-item2.html: Added.
        * http/tests/navigation/redirect-load-no-form-restoration-expected.txt:
        Updated results. Adding a history entry here is correct behavior.


        Reviewed by Sam Weinig.

        Fixed <rdar://problem/5334483> REGRESSION: JavaScript-induced
        window.open loads not added to back/forward list

        I did an audit of our history rules in loading and tried to establish
        some sane uniformity.

        The uniform rule is:
            - HTTP redirects and HTTP redirects simulated by <meta http-equiv>
            add a history item if and only if the redirect takes > 1 second.
            - Other navigations, including JavaScript navigations, always
            add a history item, except for location.replace navigations.

        In the future, we'll want to refine the second case to be more like the
        first. I've filed <rdar://problem/5339292> about that.

        * bindings/js/JSHTMLDocumentCustom.cpp:
        (WebCore::JSHTMLDocument::setLocation): Don't pass 'true' for
        userGesture unconditionally. userGesture is used to determine popup
        blocking, not history item creation.

        * bindings/js/kjs_window.cpp: Pass 'false' for lockHistory in all loads
        except location.replace, which intends to lock history.

        * loader/FrameLoader.cpp: Distinguish between lockHistory and
        userGesture. The former determines whether a new history item gets
        created. The latter determines whether JavaScript can open popup
        windows. Start passing these variables in functions that used to
        swallow or conflate them.

        (WebCore::FrameLoader::requestFrame): Pass 'true' for lockHistory here
        because that's usually correct when setting the 'src' attribute of a
        child frame, and we want to avoid regressing <rdar://problem/4921797>.

        (WebCore::FrameLoader::load): Use the lockHistory variable to determine
        whether to start a history-creating load. Using userGesture for this
        purpose is wrong, as explained above.

        * loader/FrameLoader.h: Renamed one variant of scheduleRedirection to
        scheduleHTTPRedirection because the behavior there of measuring elapsed
        time is specific to the HTTP redirection case.

        * page/ContextMenuController.cpp:
        (WebCore::ContextMenuController::contextMenuItemSelected): lockHistory
        can always be false here because this navigation is never the result of
        a redirection.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24353 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Maciej.
oliver [Tue, 17 Jul 2007 05:28:43 +0000 (05:28 +0000)]
    Reviewed by Maciej.

        Fix for <rdar://problem/5334818> Support IME reconversion in windows

        Also includes a small amount of IME refactoring.

        * WebView.cpp:
        * WebView.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24352 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Adam and Maciej.
weinig [Tue, 17 Jul 2007 05:06:26 +0000 (05:06 +0000)]
    Reviewed by Adam and Maciej.

        Move the frameElement attribute from kjs_window to DOMWindow and
        auto-generate the bindings for it.

        * bindings/js/kjs_window.cpp:
        * bindings/js/kjs_window.h:
        * bindings/scripts/CodeGeneratorJS.pm:
        * page/DOMWindow.cpp:
        * page/DOMWindow.h:
        * page/DOMWindow.idl:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24351 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Adam.
weinig [Tue, 17 Jul 2007 04:12:02 +0000 (04:12 +0000)]
    Reviewed by Adam.

        Move more functions from kjs_window to DOMWindow and auto-generate
        the bindings for them.

        * bindings/js/kjs_window.cpp:
        * bindings/js/kjs_window.h:
        * page/DOMWindow.cpp:
        * page/DOMWindow.h:
        * page/DOMWindow.idl:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24350 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago2007-07-17 Mark Rowe <mrowe@apple.com>
bdash [Tue, 17 Jul 2007 04:04:44 +0000 (04:04 +0000)]
2007-07-17  Mark Rowe  <mrowe@apple.com>

        Gdk build fix.

        * platform/gdk/ChromeClientGdk.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24348 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoWebCore:
beidson [Tue, 17 Jul 2007 03:47:04 +0000 (03:47 +0000)]

        Reviewed by Adam

        Begin the arduous task of localizing FTP directory listings while removing a global initializer!

        * loader/FTPDirectoryDocument.cpp:

        * page/mac/WebCoreViewFactory.h:

        * platform/LocalizedStrings.h:

        * platform/gdk/TemporaryLinkStubs.cpp:

        * platform/mac/LocalizedStringsMac.mm:

        * platform/qt/Localizations.cpp:


        Reviewed by Adam

        Begin the arduous task of localizing FTP directory listings while removing a global initializer!

        * English.lproj/Localizable.strings:
        * WebCoreSupport/WebViewFactory.mm:
        (-[WebViewFactory unknownFileSizeText]):


        Reviewed by Adam

        Begin the arduous task of localizing FTP directory listings while removing a global initializer!

        * English.lproj/Localizable.strings:
        * WebCoreLocalizedStrings.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24347 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoLayoutTests:
weinig [Tue, 17 Jul 2007 02:40:25 +0000 (02:40 +0000)]

        Reviewed by Oliver.

        Remove more redundant isSafeScript checks.

        * http/tests/security/cross-frame-access-custom-expected.txt:
        * http/tests/security/cross-frame-access-get-expected.txt:
        * http/tests/security/cross-frame-access-name-getter-expected.txt:


        Reviewed by Oliver.

        Remove more redundant isSafeScript checks.

        * bindings/js/kjs_window.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24346 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoFix spelling
weinig [Tue, 17 Jul 2007 02:17:47 +0000 (02:17 +0000)]
Fix spelling

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24345 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoJavaScriptCore:
weinig [Tue, 17 Jul 2007 02:01:37 +0000 (02:01 +0000)]

        Reviewed by Oliver.

        Turn of -Wshorten-64-to-32 warning for 64-bit builds.

        * Configurations/Base.xcconfig:


        Reviewed by Oliver.

        Turn of -Wshorten-64-to-32 warning for 64-bit builds.

        * Configurations/Base.xcconfig:


        Reviewed by Oliver.

        Turn of -Wshorten-64-to-32 warning for 64-bit builds.

        * Configurations/Base.xcconfig:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24344 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Maciej.
oliver [Tue, 17 Jul 2007 01:40:23 +0000 (01:40 +0000)]
    Reviewed by Maciej.

        Fix for <rdar://problem/5334826> Chinese IME composition window does not appear in the correct location

        Respect IME requests for character position.

        * WebView.cpp:
        * WebView.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24343 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoFix ChangeLog order
weinig [Tue, 17 Jul 2007 01:12:42 +0000 (01:12 +0000)]
Fix ChangeLog order

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24342 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Oliver.
andersca [Tue, 17 Jul 2007 01:07:33 +0000 (01:07 +0000)]
    Reviewed by Oliver.

        ASSERT(cs->accessCount() > 0) when instantiating widget templates in DashCode

        If the cache is disabled, accessCount will always be 0. Account for this in the assert.

        * html/HTMLTokenizer.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24341 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Darin
beidson [Tue, 17 Jul 2007 00:51:39 +0000 (00:51 +0000)]
    Reviewed by Darin

        Append FTP row entries to the first TBODY instead of the TABLE element itself,
        as many TABLE elements end up with an implicit TBODY and that's where most people would
        expect it to be added.

        * html/HTMLTableElement.h:
        (WebCore::HTMLTableElement::firstTBody): Added this accessor

        * loader/FTPDirectoryDocument.cpp:
        (WebCore::FTPDirectoryTokenizer::appendEntry): Add to the first tbody if it exists

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24340 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Darin.
kmccullo [Tue, 17 Jul 2007 00:44:23 +0000 (00:44 +0000)]
    Reviewed by Darin.

        - Continued x-platform modifications.

        * Drosera/win/DebuggerDocumentWin.cpp: Copied from Drosera/win/DebuggerObjectCallbacks.cpp.
        * Drosera/win/DebuggerDocumentWin.h: Copied from Drosera/win/DebuggerObjectCallbacks.h.
        * Drosera/win/DebuggerObjectCallbacks.cpp: Removed.
        * Drosera/win/DebuggerObjectCallbacks.h: Removed.
        * Drosera/win/Drosera.cpp:
        * Drosera/win/Drosera.vcproj/Drosera.vcproj:
        * Drosera/win/stdafx.cpp: Removed.
        * Drosera/win/stdafx.h: Removed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24339 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Geoff Garen.
weinig [Tue, 17 Jul 2007 00:43:04 +0000 (00:43 +0000)]
    Reviewed by Geoff Garen.

        Rename cross-frame-access.html to cross-frame-access-get.html

        * http/tests/security/cross-frame-access-expected.txt: Removed.
        * http/tests/security/cross-frame-access-get-expected.txt: Copied from http/tests/security/cross-frame-access-expected.txt.
        * http/tests/security/cross-frame-access-get.html: Copied from http/tests/security/cross-frame-access.html.
        * http/tests/security/cross-frame-access.html: Removed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24338 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Geoff Garen.
weinig [Tue, 17 Jul 2007 00:19:52 +0000 (00:19 +0000)]
    Reviewed by Geoff Garen.

        Move cross frame access put tests into cross-frame-access-put.html.

        * http/tests/security/cross-frame-access-custom-expected.txt:
        * http/tests/security/cross-frame-access-custom.html:
        * http/tests/security/cross-frame-access-expected.txt:
        * http/tests/security/cross-frame-access-first-time-expected.txt:
        * http/tests/security/cross-frame-access-first-time.html:
        * http/tests/security/cross-frame-access-frames.html:
        * http/tests/security/cross-frame-access-history.html:
        * http/tests/security/cross-frame-access-location-expected.txt:
        * http/tests/security/cross-frame-access-location.html:
        * http/tests/security/cross-frame-access-name-getter-expected.txt:
        * http/tests/security/cross-frame-access-name-getter.html:
        * http/tests/security/cross-frame-access-put-expected.txt:
        * http/tests/security/cross-frame-access-put.html:
        * http/tests/security/cross-frame-access.html:
        * http/tests/security/resources/cross-frame-iframe-for-get-test.html: Copied from http/tests/security/resources/cross-frame-iframe.html.
        * http/tests/security/resources/cross-frame-iframe-for-put-test.html:
        * http/tests/security/resources/cross-frame-iframe.html: Removed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24337 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoCorrectly check for input fields and text areas.
justing [Mon, 16 Jul 2007 23:21:11 +0000 (23:21 +0000)]
Correctly check for input fields and text areas.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24336 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Updated results for Tiger and filed:
adele [Mon, 16 Jul 2007 23:08:02 +0000 (23:08 +0000)]
    Updated results for Tiger and filed:
        <rdar://problem/5338680> fast/forms/hidden-listbox and fast/forms/hidden-input-file are missing empty RenderText on Leopard

        * fast/forms/hidden-input-file-expected.txt:
        * fast/forms/hidden-listbox-expected.txt:
        * mac/leopard/Skipped:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24335 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoLayoutTests:
justing [Mon, 16 Jul 2007 22:31:10 +0000 (22:31 +0000)]

        Reviewed by Adele.

        <rdar://problem/5240265> REGRESSION: Clicking a <button> clears the selection in contenteditable areas

        * editing/selection/5240265-expected.checksum: Added.
        * editing/selection/5240265-expected.png: Added.
        * editing/selection/5240265-expected.txt: Added.
        * editing/selection/5240265.html: Added.


        Reviewed by Adele.

        REGRESSION: Clicking a <button> clears the selection in contenteditable areas

        * page/FocusController.cpp:
        (WebCore::clearSelectionIfNeeded): Don't clear the selection if
        the mouse press that caused the focus shift occurred in
        a -webkit-user-select: ignore region.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24334 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago2007-07-16 Charles Gaudette <charles@gaudette-net.com>
pewtermoose [Mon, 16 Jul 2007 22:22:27 +0000 (22:22 +0000)]
2007-07-16  Charles Gaudette  <charles@gaudette-net.com>

        Reviewed by Niko.

        - http://bugs.webkit.org/show_bug.cgi?id=14628
          broken link on webkit.org home page - "KHTML" points to http://khtml.info/

        * index.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24333 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Added a manual test to test calling window.print on a subframe.
hyatt [Mon, 16 Jul 2007 22:09:05 +0000 (22:09 +0000)]
    Added a manual test to test calling window.print on a subframe.

        Reviewed by Darin.

        * manual-tests/resources/print-subframe.html: Added.
        * manual-tests/window-print-subframe.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24332 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoWebKit/win part of <rdar://problem/5336005> Calling window.print() on a subframe...
aroben [Mon, 16 Jul 2007 22:01:35 +0000 (22:01 +0000)]
WebKit/win part of <rdar://problem/5336005> Calling window.print() on a subframe prints whole page, should only print that subframe


        WebKit/win part of <rdar://problem/5336005> Calling window.print() on a subframe prints whole page, should only print that subframe

        Reviewed by Darin.

        * Interfaces/IWebUIDelegate.idl: Rename print to printFrame to closer
        match the Mac method, and copy more of the Mac API comments.
        * WebChromeClient.cpp:
        (WebChromeClient::print): Call printFrame instead of print.


        Added a manual test to test calling window.print on a subframe.

        Reviewed by Darin.

        * manual-tests/resources/print-subframe.html: Added.
        * manual-tests/window-print-subframe.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24331 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoReplace Frame::print with Chrome::print and remove -[WebFrameBridge print]
aroben [Mon, 16 Jul 2007 22:01:23 +0000 (22:01 +0000)]
Replace Frame::print with Chrome::print and remove -[WebFrameBridge print]


        Replace Frame::print with Chrome::print

        This also adds a Frame* parameter to Chrome::print and
        ChromeClient::print so that it knows which Frame to print.

        Reviewed by Darin.

        All tests pass.

        * bindings/js/kjs_window.cpp:
        (KJS::WindowFunc::callAsFunction): Call Chrome::print.
        * editing/JSEditor.cpp: Ditto.

        * bridge/win/FrameWin.cpp: Removed Frame::print.
        * page/Frame.h: Ditto.
        * page/mac/FrameMac.mm: Ditto.
        * page/qt/FrameQt.cpp: Ditto.
        * platform/gdk/FrameGdk.cpp: Ditto.
        * page/mac/WebCoreFrameBridge.h: Removed -print.

        * page/Chrome.cpp: Added Frame* parameter to Chrome::print and pass it
        up to the ChromeClient.
        * page/Chrome.h: Ditto.
        * page/ChromeClient.h: Ditto.
        * platform/gdk/TemporaryLinkStubs.cpp: Ditto.
        * platform/graphics/svg/SVGImageEmptyClients.h: Ditto.


        Move printing from WebFrameBridge to WebChromeClient

        Reviewed by Darin.

        * WebCoreSupport/WebChromeClient.h: Updated for ChromeClient changes.
        * WebCoreSupport/WebChromeClient.mm:
        (WebChromeClient::print): Moved code from WebFrameBridge.
        * WebCoreSupport/WebFrameBridge.mm: Removed -print.


        Updated WebChromeClient for ChromeClient changes.

        Reviewed by Darin.

        * WebChromeClient.cpp:
        (WebChromeClient::print): Added a Frame* parameter.
        * WebChromeClient.h: Ditto.


        Updated ChromeClientQt for ChromeClient changes.

        Reviewed by Darin.

        * WebCoreSupport/ChromeClientQt.cpp:
        (WebCore::ChromeClientQt::print): Added a Frame* parameter.
        * WebCoreSupport/ChromeClientQt.h: Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24330 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Sam.
kmccullo [Mon, 16 Jul 2007 21:59:35 +0000 (21:59 +0000)]
    Reviewed by Sam.

        - Modified files to use cross-platform code.

        * Drosera/DebuggerDocument.cpp: Added.
        * Drosera/DebuggerDocument.h: Added.
        * Drosera/config.h: Added.
        * Drosera/mac/DebuggerApplication.mm:
        (-[DebuggerApplication attach:]):
        * Drosera/mac/DebuggerDocument.h: Removed.
        * Drosera/mac/DebuggerDocument.mm: Removed.
        * Drosera/mac/DebuggerDocumentMac.h: Copied from Drosera/mac/DebuggerDocument.h.
        * Drosera/mac/DebuggerDocumentMac.mm: Copied from Drosera/mac/DebuggerDocument.mm.
        (-[DebuggerDocumentMac initWithServerName:]):
        (-[DebuggerDocumentMac dealloc]):
        (-[DebuggerDocumentMac breakpointEditorHTML]):
        (-[DebuggerDocumentMac isPaused]):
        (-[DebuggerDocumentMac pause]):
        (-[DebuggerDocumentMac resume]):
        * Drosera/mac/Drosera.xcodeproj/project.pbxproj:
        * Drosera/win/DebuggerApplication.cpp:
        * Drosera/win/DebuggerObjectCallbacks.cpp:
        * Drosera/win/Drosera.cpp:
        * Drosera/win/DroseraPrefix.cpp: Added.
        * Drosera/win/DroseraPrefix.h: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24329 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago * StringsNotToBeLocalized.txt: Update for recent changes.
darin [Mon, 16 Jul 2007 21:55:53 +0000 (21:55 +0000)]
    * StringsNotToBeLocalized.txt: Update for recent changes.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24328 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoLayoutTests:
adele [Mon, 16 Jul 2007 21:46:58 +0000 (21:46 +0000)]

        Reviewed by Brady.

        Test for <rdar://problem/5321490> REGRESSION: visibility:hidden property doesn't hide parts of file input element

        * fast/forms/hidden-input-file-expected.checksum: Added.
        * fast/forms/hidden-input-file-expected.png: Added.
        * fast/forms/hidden-input-file-expected.txt: Added.
        * fast/forms/hidden-input-file.html: Added.
        * fast/forms/hidden-listbox-expected.checksum
        * fast/forms/hidden-listbox-expected.png
        * fast/forms/hidden-listbox-expected.txt
        * fast/forms/hidden-listbox.html


        Reviewed by Brady.

        Fix for <rdar://problem/5321490> REGRESSION: visibility:hidden property doesn't hide parts of file input element and listbox


        * rendering/RenderFileUploadControl.cpp: (WebCore::RenderFileUploadControl::paintObject):
        Only paint if the control is supposed to be visible.
        * rendering/RenderListBox.cpp: (WebCore::RenderListBox::paintObject):
        ditto.  We already had checks to prevent drawing the text that considered the option's style too, but if visibility: hidden
        was just set on the listbox then a lot of extra work was being done, and the scrollbar was drawing too.  This change fixes that.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24327 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Steve.
oliver [Mon, 16 Jul 2007 21:42:58 +0000 (21:42 +0000)]
    Reviewed by Steve.

        Fix for http://bugs.webkit.org/show_bug.cgi?id=14630

        Remove bogus assertion

        * WebView.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24326 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Darin.
weinig [Mon, 16 Jul 2007 21:34:08 +0000 (21:34 +0000)]
    Reviewed by Darin.

        Add all the window's functions to cross-frame-access-put.html test.

        * http/tests/security/cross-frame-access-put-expected.txt:
        * http/tests/security/cross-frame-access-put.html:
        * http/tests/security/resources/cross-frame-iframe-for-put-test.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24325 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Fix for Radar 5338081, text-shadow should be reset on form controls so that
hyatt [Mon, 16 Jul 2007 21:02:02 +0000 (21:02 +0000)]
    Fix for Radar 5338081, text-shadow should be reset on form controls so that
        it doesn't inherit in when set e.g., on the <body>.

        Reviewed by beth

        * css/html4.css:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24324 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoskip these crashing tests
treat [Mon, 16 Jul 2007 20:51:06 +0000 (20:51 +0000)]
skip these crashing tests

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24323 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago2007-07-16 Holger Hans Peter Freyther <zecke@selfish.org>
zecke [Mon, 16 Jul 2007 20:39:03 +0000 (20:39 +0000)]
2007-07-16  Holger Hans Peter Freyther  <zecke@selfish.org>

        Reviewed by Niko.

        FrameView, PlatformScrollbar and changes to the way we draw

        Fix the lifetime of PlatformScrollbar, use a default width and
        height and fix drawing of the PlatformScrollbar and other widgets
        the following way: FrameGdk handles the expose events of the Frame
        and will make the frame redraw and now it will draw the childrent of
        the FrameView as well. This approach has the issue of honoring the
        z-order of elements inside the RenderTree. Honoring the z-order will
        be a bit more work

        Widget can now handle Widget::setGtkWidget call where the GtkWidget
        has not yet a GdkWindow allocated. We will lazily set the GdkDrawable.

        In preparation of honoring the z-order of the RenderTree for RenderWidgets
        it is started to store native objects inside the GraphicsContext. Doing this
        nicely eliminates the need of RenderThemeGdk to do any drawing to a temporary
        GdkPixmap. This should fix themes with rounded buttons.

        ScrollView implement add- and removeChild to get a working PlatformScrollbar

        * platform/Widget.h:
        * platform/gdk/FrameGdk.cpp:
        (frame_gdk_expose_child): Will send the expose to all children
        (WebCore::FrameGdk::handleGdkEvent): Painting changes, move to mouseMoved
        * platform/gdk/PlatformScreenGdk.cpp:
        (WebCore::screenDepth): gdkDrawable
        * platform/gdk/PlatformScrollBarGdk.cpp:
        (PlatformScrollbar::PlatformScrollbar): Fix ownership
        (PlatformScrollbar::~PlatformScrollbar): Fix ownership
        (PlatformScrollbar::paint): Widget::paint will do the right thing soon
        * platform/gdk/RenderThemeGdk.cpp: No need for using a GdkPixmap, draw directly
        (WebCore::RenderThemeGdk::paintCheckbox): No need for using a GdkPixmap, draw directly
        (WebCore::RenderThemeGdk::paintRadio): No need for using a GdkPixmap, draw directly
        (WebCore::RenderThemeGdk::paintButton): No need for using a GdkPixmap, draw directly
        * platform/gdk/RenderThemeGdk.h: remove the copyContext call
        * platform/gdk/ScrollViewGdk.cpp: gdkDrawable
        (WebCore::ScrollView::updateView): gdkDrawable
        (WebCore::ScrollView::update): clear the area to fix repainting issues
        (WebCore::ScrollView::setGtkWidget): gdkDrawable
        (WebCore::ScrollView::addChild): implement
        (WebCore::ScrollView::removeChild): implement
        * platform/gdk/TemporaryLinkStubs.cpp: Not needed header removed
        * platform/gdk/WidgetGdk.cpp:
        (WebCore::Widget::gdkDrawable): Renamed from drawable
        (WebCore::Widget::setGtkWidget): use gdkDrawable
        (WebCore::Widget::setCursor): gdkDrawable
        (WebCore::Widget::show): gdkDrawable
        (WebCore::Widget::hide): gdkDrawable
        * platform/graphics/GraphicsContext.h: Allow to set the GdkDrawable, e.g. used inside a expose event
        * platform/graphics/cairo/GraphicsContextCairo.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24322 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago2007-07-16 Holger Hans Peter Freyther <zecke@selfish.org>
zecke [Mon, 16 Jul 2007 20:21:22 +0000 (20:21 +0000)]
2007-07-16  Holger Hans Peter Freyther  <zecke@selfish.org>

        Reviewed by Maciej.

        Lazily update the scrollbars (e.g. on loading a page) to
        generate less expose events, which will lead to laying out less often
        and will fix http://bugs.webkit.org/show_bug.cgi?id=14020.

        * platform/gdk/ScrollViewGdk.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24321 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Adam, Sam and Tim.
kmccullo [Mon, 16 Jul 2007 18:09:35 +0000 (18:09 +0000)]
    Reviewed by Adam, Sam and Tim.

        - Moving files to prepare for cross-platform architecture.

        * Drosera/DebuggerApplication.h: Removed.
        * Drosera/DebuggerApplication.m: Removed.
        * Drosera/DebuggerDocument.h: Removed.
        * Drosera/DebuggerDocument.m: Removed.
        * Drosera/Drosera.pch: Removed.
        * Drosera/Drosera.xcodeproj: Removed.
        * Drosera/Drosera.xcodeproj/project.pbxproj: Removed.
        * Drosera/Info.plist: Removed.
        * Drosera/LauncherInfo.plist: Removed.
        * Drosera/Makefile: Removed.
        * Drosera/launcher.m: Removed.
        * Drosera/mac: Added.
        * Drosera/mac/DebuggerApplication.h: Copied from Drosera/DebuggerApplication.h.
        * Drosera/mac/DebuggerApplication.mm: Copied from Drosera/DebuggerApplication.m.
        * Drosera/mac/DebuggerDocument.h: Copied from Drosera/DebuggerDocument.h.
        * Drosera/mac/DebuggerDocument.mm: Copied from Drosera/DebuggerDocument.m.
        * Drosera/mac/Drosera.pch: Copied from Drosera/Drosera.pch.
        * Drosera/mac/Drosera.xcodeproj: Copied from Drosera/Drosera.xcodeproj.
        * Drosera/mac/Drosera.xcodeproj/project.pbxproj:
        * Drosera/mac/Info.plist: Copied from Drosera/Info.plist.
        * Drosera/mac/LauncherInfo.plist: Copied from Drosera/LauncherInfo.plist.
        * Drosera/mac/Makefile: Copied from Drosera/Makefile.
        * Drosera/mac/launcher.m: Copied from Drosera/launcher.m.
        * Drosera/mac/main.m: Copied from Drosera/main.m.
        * Drosera/main.m: Removed.
        * Makefile:
        * Makefile.Drosera: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24319 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Kevin Decker and Darin Adler
sullivan [Mon, 16 Jul 2007 17:53:01 +0000 (17:53 +0000)]
    Reviewed by Kevin Decker and Darin Adler

        - fixed <rdar://problem/5337529> Holes in Find banner overlay on PDF pages are left in wrong place after changing scale

        * WebView/WebPDFView.mm:
        (-[WebPDFView _scaleOrDisplayModeChanged:]):
        tell UI delegate that the entire PDF view has been redrawn

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24318 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Kevin McCullough.
weinig [Mon, 16 Jul 2007 17:05:34 +0000 (17:05 +0000)]
    Reviewed by Kevin McCullough.

        Breakup http/tests/security/cross-frame-access.html into multiple tests to
        make it easier to identify flaws should they arise.

        - Also factors out JS into a separate file.

        * http/tests/security/cross-frame-access-custom-expected.txt: Added.
        * http/tests/security/cross-frame-access-custom.html: Added.
        * http/tests/security/cross-frame-access-expected.txt:
        * http/tests/security/cross-frame-access-frames-expected.txt: Added.
        * http/tests/security/cross-frame-access-frames.html: Added.
        * http/tests/security/cross-frame-access-history-expected.txt: Added.
        * http/tests/security/cross-frame-access-history.html: Added.
        * http/tests/security/cross-frame-access-location-expected.txt: Added.
        * http/tests/security/cross-frame-access-location.html: Added.
        * http/tests/security/cross-frame-access-name-getter-expected.txt: Added.
        * http/tests/security/cross-frame-access-name-getter.html: Added.
        * http/tests/security/cross-frame-access.html:
        * http/tests/security/resources/cross-frame-access.js: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24317 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoForgot Skipped file
andersca [Mon, 16 Jul 2007 15:20:47 +0000 (15:20 +0000)]
Forgot Skipped file

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24314 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Remove basic-auth.html, it no longer hangs on Leopard.
andersca [Mon, 16 Jul 2007 15:18:35 +0000 (15:18 +0000)]
    Remove basic-auth.html, it no longer hangs on Leopard.

        * mac/leopard/Skipped:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24313 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago2007-07-16 Mark Rowe <mrowe@apple.com>
bdash [Mon, 16 Jul 2007 09:06:29 +0000 (09:06 +0000)]
2007-07-16  Mark Rowe  <mrowe@apple.com>

        Update expected result after r24302.

        * fast/css/empty-script-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24303 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago2007-07-16 Rob Buis <buis@kde.org>
bdash [Mon, 16 Jul 2007 08:38:41 +0000 (08:38 +0000)]
2007-07-16  Rob Buis  <buis@kde.org>

        Reviewed by Maciej.

        non empty <style> element doesn't work

        Return an empty sheet for empty <style> elements
        instead of a null sheet. Clean up StyleElement a bit.

        * dom/StyleElement.cpp:
        * dom/StyleElement.h:
        * html/HTMLStyleElement.cpp:
        * html/HTMLStyleElement.h:
        * ksvg2/svg/SVGStyleElement.cpp:
        * ksvg2/svg/SVGStyleElement.h:

2007-07-16  Rob Buis  <buis@kde.org>

        Reviewed by Maciej.

        Testcase for:
        non empty <style> element doesn't work

        * fast/css/empty-script-expected.txt: Added.
        * fast/css/empty-script.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24302 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoLayoutTests:
oliver [Mon, 16 Jul 2007 06:13:32 +0000 (06:13 +0000)]

        Reviewed by Adam.

        Update for changed behaviour now that we allow tab characters in input fields again

        * editing/inserting/typing-tab-designmode-forms-expected.checksum:
        * editing/inserting/typing-tab-designmode-forms-expected.png:
        * editing/inserting/typing-tab-designmode-forms-expected.txt:


        Reviewed by Adam.

        Fix for fix for <rdar://problem/5336602> REGRESSION: tab characters
                        cannot be entered into text fields

        Need to use UChar not char.

        * html/HTMLInputElement.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24301 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoLayoutTests:
weinig [Mon, 16 Jul 2007 02:09:02 +0000 (02:09 +0000)]

        Reviewed by Geoff Garen.

        Update results after removing redundant calls to isSafeScript()

        * http/tests/security/cross-frame-access-expected.txt:


        Reviewed by Geoff Garen.

        Remove redundant calls to isSafeScript()

        * bindings/js/JSDOMWindowCustom.cpp:
        * bindings/js/kjs_window.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24300 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Maciej Stachowiak.
ggaren [Mon, 16 Jul 2007 01:44:46 +0000 (01:44 +0000)]
    Reviewed by Maciej Stachowiak.

        Removed incorrect error log, seen while running layout tests.

        Perhaps at some point FrameLoadTypeInteral had a specific meaning
        that prevented it from being used in non-child frame, non-redirect
        situations. However, that meaning has been lost and nobody remembers
        how to get it back.

        We currently use FrameLoadTypeInternal for any load that is not
        initiated by a user gesture, so it's trivial to produce this error
        log, even though nothing has gone wrong.

        * loader/FrameLoader.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24299 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoLayoutTests:
oliver [Mon, 16 Jul 2007 01:36:49 +0000 (01:36 +0000)]

        Reviewed by Maciej.

        Testcase for <rdar://problem/5336602> REGRESSION: tab characters cannot be entered into text fields

        * fast/forms/tab-in-input-expected.txt: Added.
        * fast/forms/tab-in-input.html: Added.


        Reviewed by Maciej

        Fix for <rdar://problem/5336602> REGRESSION: tab characters cannot be entered into text fields

        constrainValue attempted to block control characters from being inserted
        by a simple less than ' ' check, which prevent tab characters from being
        inserted.  This patch adds an additional check to allow the tab character

        * html/HTMLInputElement.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24298 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Third shot at blind Qt/GDK build fix
beidson [Sat, 14 Jul 2007 23:39:03 +0000 (23:39 +0000)]
    Third shot at blind Qt/GDK build fix

        * platform/gdk/TemporaryLinkStubs.cpp:
        * platform/qt/TemporaryLinkStubs.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24295 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoRUBBERSTAMPED BY SAM WEINIG - Try to (actually) fix Qt and GDK
beidson [Sat, 14 Jul 2007 23:25:13 +0000 (23:25 +0000)]
RUBBERSTAMPED BY SAM WEINIG - Try to (actually) fix Qt and GDK

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24294 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoFix Windows build and set the pref to allow FTP directory listings in Windows nightlies
beidson [Sat, 14 Jul 2007 23:21:36 +0000 (23:21 +0000)]
Fix Windows build and set the pref to allow FTP directory listings in Windows nightlies

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24293 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Fix Tiger build.
weinig [Sat, 14 Jul 2007 22:59:57 +0000 (22:59 +0000)]
    Fix Tiger build.

        * config.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24291 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Blind attempt at (hopefully) fixing the Qt and GDK builds
beidson [Sat, 14 Jul 2007 22:58:42 +0000 (22:58 +0000)]
    Blind attempt at (hopefully) fixing the Qt and GDK builds

        * platform/gdk/TemporaryLinkStubs.cpp:
        * platform/qt/TemporaryLinkStubs.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24290 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoJavaScriptCore:
beidson [Sat, 14 Jul 2007 22:25:31 +0000 (22:25 +0000)]

        Reviewed by Sam Weinig

        Initial check-in for <rdar://problem/3154486> - Supporting FTP directory listings in the browser

        * wtf/Platform.h: Add ENABLE_FTPDIR feature to handle building on platforms that don't have the
          proper network-layer support


        Reviewed by Sam Weinig

        Initial check-in for <rdar://problem/3154486> - Supporting FTP directory listings in the browser

        Development has taken place on both Leopard and Windows (windows enabling patch will follow shortly)
        Support will only exist on platforms whose network layers (ie, ResourceHandle) support piping the raw
        FTP Directory data into the engine.

        This is known to include Leopard, Windows, and CURL based platforms - probably others
        Known incompatible platforms include Tiger.

        The FTP listing page is built in HTML.  An external HTML document can be provided as a template for the
        directory listing.  This document is parsed, and expected to have a <table> with a certain class in it.
        As FTP directory listings come in, each entry is added to this table as a row.

        If no HTML template is provided to work with, a very basic empty document is hard coded that contains only
        this special table.

        Upgrades to the look and feel of the directory listings can take place largely in this HTML template which,
        admittedly, is conspicuously absent in this initial checkin

        * WebCore.exp: New Settings symbols for WebKit
        * WebCore.xcodeproj/project.pbxproj:

        * config.h: If BUILDING_ON_TIGER, disable FTP directory listing support

        * dom/DOMImplementation.cpp:
        (WebCore::DOMImplementation::createDocument): Create an FTPDirectoryDocument if the mime type is

        * loader/FTPDirectoryDocument.cpp: Added.
        (WebCore::FTPDirectoryTokenizer::FTPDirectoryTokenizer): Building on HTML tokenizer, this facilitates
          parsing FTP Directory listings in the engine
        (WebCore::FTPDirectoryTokenizer::appendEntry): Add a table row for a directory listing to the document
        (WebCore::FTPDirectoryTokenizer::createTDForFilename): Create the TD for the filename with the anchor in it

        (WebCore::processFilesizeString): Prettify the filesize
        (WebCore::processFileDateString): Prettify the date
        (WebCore::FTPDirectoryTokenizer::loadDocumentTemplate): Loads the HTML template that FTP directory listings
          can build on top of
        (WebCore::FTPDirectoryTokenizer::createBasicDocument): Creates a most basic document (body and one table) to
          append the rows to in case the document template is not loaded and parsed
        (WebCore::FTPDirectoryTokenizer::write): Add the FTP listing to the buffer and parse entries out of it

        (WebCore::FTPDirectoryDocument::FTPDirectoryDocument): Special document-type for FTP directory listings
        * loader/FTPDirectoryDocument.h: Added.

        * loader/FTPDirectoryParser.cpp: Added.
        (WebCore::parseOneFTPLine): Adapted from ParseFTPList.cpp from Firefox - parses most known
          FTP directory listing styles into discrete FTP directory entries
        * loader/FTPDirectoryParser.h: Added.
        (WebCore::ListState::ListState): Maintains FTP Parser state
        (WebCore::ListResult::ListResult): Represents one FTP directory listing

        * loader/FrameLoader.cpp:
        (WebCore::FrameLoader::checkContentPolicy): If the Settings say to bypass the policy check for
          "application/x-ftp-directory", skip it.  This will allow people to test this feature even if their
          browser/WebKit app would otherwise not allow it

        * page/Settings.cpp:
        (WebCore::Settings::setFTPDirectoryTemplatePath): Set the path to the FTP listing document template
        (WebCore::Settings::setForceFTPDirectoryListings): Set to bypass the policy delegate check for mime type
        * page/Settings.h:

        * platform/Logging.cpp: Add an "LogFTP" logging channel
        * platform/Logging.h:
        * platform/mac/LoggingMac.mm:

        * platform/MimeTypeRegistry.cpp:
        (WebCore::initialiseSupportedNonImageMimeTypes): Add "application/x-ftp-directory"

        * platform/SharedBuffer.cpp:
        (WebCore::SharedBuffer::createWithContentsOfFile): Stub for now
        * platform/SharedBuffer.h:

        * platform/mac/SharedBufferMac.mm:
        (WebCore::SharedBuffer::createWithContentsOfFile): Load contents of the file into an NSData, then
          wrap that NSData


        Reviewed by Sam Weinig

        Initial check-in for <rdar://problem/3154486> - Supporting FTP directory listings in the browser

        * WebView/WebPreferenceKeysPrivate.h: Added preference keys for the FTP template location, as well as to force
          FTP directory listings, bypassing the policy delegate.  This is necessary to test the new feature until browser
          policy delegate support is added.

        * WebView/WebPreferences.m:
        (-[WebPreferences _setFTPDirectoryTemplatePath:]):
        (-[WebPreferences _ftpDirectoryTemplatePath]):
        (-[WebPreferences _setForceFTPDirectoryListings:]):
        (-[WebPreferences _forceFTPDirectoryListings]):
        * WebView/WebPreferencesPrivate.h:
        * WebView/WebView.mm:
        (-[WebView _updateWebCoreSettingsFromPreferences:]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24289 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Darin.
ap [Sat, 14 Jul 2007 18:02:26 +0000 (18:02 +0000)]
    Reviewed by Darin.

        REGRESSION(13464): drivehq.com login form doesn't work

        Test: http/tests/misc/post-submit-button.html

        * html/HTMLInputElement.cpp:
        (WebCore::HTMLInputElement::appendFormData): Append a submit even if its value is empty.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24288 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago2007-07-14 Cameron Zwarich <cwzwarich@uwaterloo.ca>
bdash [Sat, 14 Jul 2007 17:04:03 +0000 (17:04 +0000)]
2007-07-14  Cameron Zwarich  <cwzwarich@uwaterloo.ca>

        Reviewed by Darin.

        Fixes http://bugs.webkit.org/show_bug.cgi?id=13517,
              http://bugs.webkit.org/show_bug.cgi?id=14237, and
              the failure of test js1_5/Scope/regress-185485.js

        Bug 13517: DOM Exception 8 in finance.aol.com sub-page
        Bug 14237: Javascript "var" statement interprets initialization in the topmost function scope

        * kjs/nodes.cpp:
        * tests/mozilla/expected.html:

2007-07-14  Cameron Zwarich  <cwzwarich@uwaterloo.ca>

        Reviewed by Darin.

        Added tests for the following bugs:

        Bug 13517: DOM Exception 8 in finance.aol.com sub-page
        Bug 14237: Javascript "var" statement interprets initialization in the topmost function scope

        * fast/js/resources/vardecl-blocks-init.js: Added.
        * fast/js/resources/vardecl-preserve-arguments.js: Updated.
        * fast/js/vardecl-blocks-init-expected.txt: Added.
        * fast/js/vardecl-blocks-init.html: Added.
        * fast/js/vardecl-preserve-arguments-expected.txt: Updated result.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24287 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago2007-07-15 Mark Rowe <mrowe@apple.com>
bdash [Sat, 14 Jul 2007 16:53:14 +0000 (16:53 +0000)]
2007-07-15  Mark Rowe  <mrowe@apple.com>

        * Scripts/generate-coverage-data: Don't fail if WebKitBuild directory does not exist.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24286 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Darin and Alexey.
oliver [Sat, 14 Jul 2007 08:21:24 +0000 (08:21 +0000)]
    Reviewed by Darin and Alexey.

        Fix for <rdar://problem/5231528> Inline input of International text (IME)

        This patch adds IME support to WebKit/win, it currently does not support
        reconversion (<rdar://problem/5334818>) and has issues with the chinese
        IMEs (<rdar://problem/5334826>)

        * WebEditorClient.cpp:
          Prevent the initial keydown for an IME from triggering a keypressed event
        * WebView.cpp:
           Dynamic loader for IME libraries

           Helper functions

           IME event handling, so far most of these are not implemented, but the bulk of functionality
           is performed the the composition event handlers
        * WebView.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24285 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Mark.
oliver [Sat, 14 Jul 2007 06:40:32 +0000 (06:40 +0000)]
    Reviewed by Mark.

        Correct expected pixel results

        * fast/forms/input-appearance-readonly-expected.checksum:
        * fast/forms/input-appearance-readonly-expected.png:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24284 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Sam.
oliver [Sat, 14 Jul 2007 06:03:39 +0000 (06:03 +0000)]
    Reviewed by Sam.

        Fix for
            <rdar://problem/5335265> REGRESSION: WebCore::DragController crashes if WebDragSourceActionDHTML cleared

        Reorder logic for drag initiation: we can always preflight the clipboard, regardless as to whether
        we are performing any dhtml drag handling

        * page/EventHandler.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@24283 268f45cc-cd09-0410-ab3c-d52691b4dbfc