1. 13 May, 2008 5 commits
    • ap@webkit.org's avatar
      Reviewed by John Sullivan. · c6c08514
      ap@webkit.org authored
              <rdar://problem/4852077> REGRESSION: The content of the HTML file which is saved as Thai
              (ISO 8859-11) disappear.
      
              <rdar://problem/5842597> Thai ISO-8859-11 should be upgraded to Windows-874
      
              Test: fast/encoding/char-decoding.html
      
              * platform/text/TextCodecICU.cpp: (WebCore::TextCodecICU::registerExtendedEncodingNames):
              Register ISO-8859-11, which doesn't have a standard IANA name, despite following
              the usual 8859 naming scheme. Also, made TIS-620 map to windows-874-2000 explicitly.
      
      
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@33384 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      c6c08514
    • ap@webkit.org's avatar
      Reviewed by Eric Seidel. · a1b6f102
      ap@webkit.org authored
              https://bugs.webkit.org/show_bug.cgi?id=18681
              <rdar://problem/5888130> WebKit should not remove BOM characters from content.
      
              We were only trying to match Firefox, and it doesn't do this any more.
      
              Tests: fast/encoding/bom-in-content.html
                     fast/encoding/bom-in-content-utf16.html
      
              * platform/text/TextDecoder.cpp: (WebCore::TextDecoder::checkForBOM): Skip the BOM if it's
              at the start of input stream.
      
              * platform/text/TextCodec.cpp:
              * platform/text/TextCodec.h:
              * platform/text/TextCodecICU.cpp:
              (WebCore::TextCodecICU::decode):
              * platform/text/TextCodecUTF16.cpp:
              (WebCore::TextCodecUTF16::decode):
              * platform/text/mac/TextCodecMac.cpp:
              (WebCore::TextCodecMac::decode):
              Don't remove the BOM.
      
      
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@33380 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      a1b6f102
    • ap@webkit.org's avatar
      Reviewed by Dan Bernstein. · d45dad02
      ap@webkit.org authored
              Add tests verifying that we don't mix up some similar, but incompatible encodings.
      
              * fast/encoding/char-decoding-expected.txt:
              * fast/encoding/char-decoding.html:
      
      
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@33379 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      d45dad02
    • andersca@apple.com's avatar
      WebCore: · ec2dd2c3
      andersca@apple.com authored
      2008-05-13  Anders Carlsson  <andersca@apple.com>
      
              Reviewed by Darin.
      
              Don't crash when a document loader is destroyed while an initial caching attempt
              is in progress.
      
              * loader/appcache/ApplicationCacheGroup.cpp:
              (WebCore::ApplicationCacheGroup::~ApplicationCacheGroup):
              Stop loading.
              
              (WebCore::ApplicationCacheGroup::stopLoading):
              New method that stops a cache update.
              
              (WebCore::ApplicationCacheGroup::documentLoaderDestroyed):
              Delete ourselves here.
              
              * loader/appcache/ApplicationCacheGroup.h:
      
      LayoutTests:
      
      2008-05-13  Anders Carlsson  <andersca@apple.com>
      
              Reviewed by Darin.
      
              Add testcase.
              
              * http/tests/appcache/navigating-away-while-cache-attempt-in-progress-expected.txt: Added.
              * http/tests/appcache/navigating-away-while-cache-attempt-in-progress.html: Added.
              * http/tests/appcache/resources/navigating-away-while-cache-attempt-in-progress.manifest: Added.
              * http/tests/appcache/resources/slow-resource.php: Added.
      
      
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@33378 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      ec2dd2c3
    • ap@webkit.org's avatar
      Reviewed by Darin. · 6e8259dd
      ap@webkit.org authored
              <rdar://problem/5841351> Turkish ISO 8859-9 encoding should be treated as windows-1254
      
              Match IE and upgrade ISO-8859-9 to windows-1254, which is its strict superset.
      
              Test: fast/encoding/char-decoding.html
      
              * platform/text/TextCodecICU.cpp:
              (WebCore::TextCodecICU::registerExtendedEncodingNames): Register the standard name
              unconditionally. Previously, we registered windows-949-2000 as a special case that lacked 
              an IANA standard name, and got away with not registering GBK because it happened
              to be reported by ICU normally earlier than any encoding upgraded to it.
      
      
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@33377 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      6e8259dd
  2. 12 May, 2008 5 commits
    • aroben@apple.com's avatar
      Fix a MSVC warning in PluginViewWin · d54c7262
      aroben@apple.com authored
      WebCore:
      
              Fix a MSVC warning in PluginViewWin
      
              Reviewed by Anders Carlsson.
      
              Test: http/tests/plugins/post-url-file.html
      
              * plugins/win/PluginViewWin.cpp:
              (WebCore::PluginView::handlePostReadFile): Added a missing return
              statement.
      
      WebKitTools:
      
              Support for testing NPN_PostURL
      
              Reviewed by Anders Carlsson.
      
              * DumpRenderTree/TestNetscapePlugIn.subproj/PluginObject.cpp:
              (toCString): Added this helper function.
              (testPostURLFile): Added. Writes the passed-in content to the
              passed-in file and calls NPN_PostURL with the passed-in URL and
              target.
              (pluginInvoke): Added a case for testPostURLFile.
              * DumpRenderTree/win/LayoutTestControllerWin.cpp:
              (followShortcuts): Changed to allow paths that don't yet exist.
      
      LayoutTests:
      
              Test that NPN_PostURL works correctly
      
              Reviewed by Anders Carlsson.
      
              * http/tests/plugins/post-url-file-ex...
      d54c7262
    • andersca@apple.com's avatar
      WebCore: · 0e6a6ea0
      andersca@apple.com authored
      2008-05-12  Anders Carlsson  <andersca@apple.com>
      
              Reviewed by Adam.
      
              Handle the case when a resource with the same URL as the manifest is listed in the manifest.
      
              * loader/appcache/ApplicationCacheGroup.cpp:
              (WebCore::ApplicationCacheGroup::addEntry):
      
      LayoutTests:
      
      2008-05-12  Anders Carlsson  <andersca@apple.com>
      
              Reviewed by Adam.
      
              Add test with a manifest file that lists itself.
              
              * http/tests/appcache/manifest-containing-itself-expected.txt: Added.
              * http/tests/appcache/manifest-containing-itself.html: Added.
              * http/tests/appcache/resources/manifest-containing-itself.manifest: Added.
      
      
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@33067 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      0e6a6ea0
    • andersca@apple.com's avatar
      WebCore: · 7b7d28be
      andersca@apple.com authored
      2008-05-12  Anders Carlsson  <andersca@apple.com>
      
              Reviewed by Adam.
      
              Handle empty manifest files without crashing.
              
              * loader/appcache/ApplicationCacheGroup.cpp:
              (WebCore::ApplicationCacheGroup::startLoadingEntry):
      
      LayoutTests:
      
      2008-05-12  Anders Carlsson  <andersca@apple.com>
      
              Reviewed by Adam.
      
              * http/tests/appcache/empty-manifest-expected.txt: Added.
              * http/tests/appcache/empty-manifest.html: Added.
              * http/tests/appcache/resources/empty.manifest: Added.
      
      
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@33061 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      7b7d28be
    • andersca@apple.com's avatar
      2008-05-12 Anders Carlsson <andersca@apple.com> · 893e60b4
      andersca@apple.com authored
              Update the expected result.
              
              * http/tests/appcache/simple-expected.txt:
      
      
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@33059 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      893e60b4
    • andersca@apple.com's avatar
      WebCore: · bbe3968b
      andersca@apple.com authored
      2008-05-12  Anders Carlsson  <andersca@apple.com>
      
              Reviewed by Adam.
      
              Make it possible to perform synchronous loads from the application cache.
              
              * loader/DocumentLoader.cpp:
              (WebCore::DocumentLoader::shouldLoadResourceFromApplicationCache):
              Factor out code from scheduleApplicationCacheLoad in its own method.
              
              (WebCore::DocumentLoader::scheduleApplicationCacheLoad):
              Call shouldLoadResourceFromApplicationCache here instead.
              
              * loader/DocumentLoader.h:
              
              * loader/FrameLoader.cpp:
              (WebCore::FrameLoader::loadResourceSynchronously):
              Call shouldLoadResourceFromApplicationCache.
      
      WebKitTools:
      
      2008-05-12  Anders Carlsson  <andersca@apple.com>
      
              Reviewed by Adam.
      
              Add support for testing application caches.
              
              * DumpRenderTree/mac/DumpRenderTree.mm:
              (dumpRenderTree):
              Empty the cache.
              
              (resetWebViewToConsistentStateBeforeTesting):
              Turn on support for the application cache.
      
      LayoutTests:
      
      2008-05-12  Anders Carlsson  <andersca@apple.com>
      
              Reviewed by Adam.
      
              Add simple appcache test.
      
              * http/conf/mime.types:
              * http/tests/appcache: Added.
              * http/tests/appcache/resources: Added.
              * http/tests/appcache/resources/not-in-cache.txt: Added.
              * http/tests/appcache/resources/simple.manifest: Added.
              * http/tests/appcache/resources/simple.txt: Added.
              * http/tests/appcache/simple-expected.txt: Added.
              * http/tests/appcache/simple.html: Added.
              * platform/gtk/Skipped:
              * platform/qt/Skipped:
              * platform/win/Skipped:
      
      
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@33057 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      bbe3968b
  3. 09 May, 2008 2 commits
    • weinig@apple.com's avatar
      WebCore: · 751ee9fb
      weinig@apple.com authored
      2008-05-09  Sam Weinig  <sam@webkit.org>
      
              Reviewed by Mark Rowe.
      
              Fix for https://bugs.webkit.org/show_bug.cgi?id=18958
              NULL pointer dereference in NamedAttrMap::setNamedItem
      
              Test: fast/dom/NamedNodeMap-setNamedItem-crash.html
      
              * dom/NamedAttrMap.cpp:
              (WebCore::NamedAttrMap::setNamedItem): Null check the argument.
      
      LayoutTests:
      
      2008-05-09  Sam Weinig  <sam@webkit.org>
      
              Reviewed by Mark Rowe.
      
              Test for https://bugs.webkit.org/show_bug.cgi?id=18958
              NULL pointer dereference in NamedAttrMap::setNamedItem
      
              * fast/dom/NamedNodeMap-setNamedItem-crash-expected.txt: Added.
              * fast/dom/NamedNodeMap-setNamedItem-crash.html: Added.
      
      
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@33023 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      751ee9fb
    • weinig@apple.com's avatar
      WebCore: · c092b7ef
      weinig@apple.com authored
      2008-05-09  Adam Barth  <abarth-webkit@adambarth.com>
      
              Reviewed by Sam Weinig.
      
              https://bugs.webkit.org/show_bug.cgi?id=18771
      
              Make postMessage generate an event on the window instead of the
              document.
      
              Test: http/tests/messaging/cross-domain-message-event-dispatch.html
      
              * dom/EventTargetNode.cpp:
              (WebCore::EventTargetNode::dispatchWindowEvent):
              * dom/EventTargetNode.h:
              * page/DOMWindow.cpp:
              (WebCore::DOMWindow::postMessageTimerFired):
      
      LayoutTests:
      
      2008-05-09  Adam Barth  <abarth-webkit@adambarth.com>
      
              Reviewed by Sam Weinig.
      
              Update tests to listen for message events on the window instead of
              the document.  Also, added a test that the event is sent to the
              window and not to the document.
      
              * http/tests/messaging/cross-domain-message-event-dispatch-expected.txt: Added.
              * http/tests/messaging/cross-domain-message-event-dispatch.html: Added.
              * http/tests/messaging/cross-domain-message-send.html:
              * http/tests/messaging/resources/cross-domain-message-receive.html:
              * http/tests/security/cross-frame-access-delete.html:
              * http/tests/security/cross-frame-access-history-put.html:
              * http/tests/security/cross-frame-access-location-put.html:
              * http/tests/security/postMessage/delivery-order.html:
              * http/tests/security/postMessage/invalid-origin-throws-exception.html:
              * http/tests/security/postMessage/javascript-page-still-sends-origin.html:
              * http/tests/security/postMessage/origin-unaffected-by-base-tag.html:
              * http/tests/security/postMessage/origin-unaffected-by-document-domain.html:
              * http/tests/security/postMessage/resources/post-message-listener.html:
              * http/tests/security/postMessage/target-origin.html:
              * http/tests/security/resources/cross-frame-iframe-for-delete-test.html:
              * http/tests/security/resources/cross-frame-iframe-for-history-put-test.html:
              * http/tests/security/resources/cross-frame-iframe-for-location-put-test.html:
              * http/tests/security/xss-eval.html:
      
      
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@33006 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      c092b7ef
  4. 08 May, 2008 2 commits
    • mitz@apple.com's avatar
      WebCore: · c2266d1a
      mitz@apple.com authored
              Reviewed by Mark Rowe.
      
              - fix https://bugs.webkit.org/show_bug.cgi?id=18818
                <rdar://problem/5901544> REGRESSION (3.1.1-TOT): Character order (float:left ordered after the first letter)
      
              Test: fast/css/first-letter-float-after-float.html
      
              * rendering/RenderBlock.cpp:
              (WebCore::RenderBlock::updateFirstLetter): Changed to add the first
              letter container right before the rest of the text it is taken from,
              instead of at the beginning of the block.
      
      LayoutTests:
      
              Reviewed by Mark Rowe.
      
              - test and updated results for https://bugs.webkit.org/show_bug.cgi?id=18818
                <rdar://problem/5901544> REGRESSION (3.1.1-TOT): Character order (float:left ordered after the first letter)
      
              * ChangeLog:
              * fast/css/first-letter-float-after-float.html: Added.
              * platform/mac-leopard/fast/css/first-letter-skip-out-of-flow-expected.checksum:
              * platform/mac-leopard/fast/css/first-letter-skip-out-of-flow-expected.png:
              * platform/mac/fast/css/first-letter-float-after-float-expected.checksum: Added.
              * platform/mac/fast/css/first-letter-float-after-float-expected.png: Added.
              * platform/mac/fast/css/first-letter-float-after-float-expected.txt: Added.
              * platform/mac/fast/css/first-letter-skip-out-of-flow-expected.txt:
      
      
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@32992 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      c2266d1a
    • hausmann@webkit.org's avatar
      2008-05-08 Simon Hausmann <hausmann@webkit.org> · 80cd2543
      hausmann@webkit.org authored
              Reviewed and found by Holger.
      
              Include .css in the list of extensions for text/css. Fixes
              fast/loader/local-css-allowed-in-strict-mode.html.
      
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@32974 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      80cd2543
  5. 07 May, 2008 8 commits
    • slewis@apple.com's avatar
      2008-05-07 Stephanie Lewis <slewis@apple.com> · f2aaa5af
      slewis@apple.com authored
              
              fix messed up result
      
              * fast/events/onunload-not-on-body-expected.txt:
      
      
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@32969 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      f2aaa5af
    • slewis@apple.com's avatar
      2008-05-07 Stephanie Lewis <slewis@apple.com> · 7c543d56
      slewis@apple.com authored
              Reviewed by Maciej.
      
              remove non-functioning code.  If onunload should be fired from these nodes then
              a WindowEventListener needs to be set.  However, Firefox and IE also do not
              fire onunload events for these cases.
      
              Test: fast/events/onunload-not-on-body.html 
              
              * html/HTMLFrameElementBase.cpp:
              (WebCore::HTMLFrameElementBase::parseMappedAttribute):
              * html/HTMLObjectElement.cpp
              (WebCore::HTMLObjectElement::parseMappedAttribute):
      
      
              Test onunload attr on a frame does not call onunload, but that an onunload attr on the body of
              a frame does.  Also test that onunloads on object elements are not called.
      
              * fast/events/onunload-not-on-body-expected.txt: Added.
              * fast/events/onunload-not-on-body.html: Added.
      
      
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@32967 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      7c543d56
    • mitz@apple.com's avatar
      WebCore: · a801ef1d
      mitz@apple.com authored
              Reviewed by John Sullivan.
      
              - fix https://bugs.webkit.org/show_bug.cgi?id=18909
                <rdar://problem/5914165> REGRESSION (r31872-r31878): Viewed photos not closing completely at indycar.com
      
              Test: fast/dynamic/positioned-movement-with-positioned-children.html
      
              * rendering/RenderBlock.cpp:
              (WebCore::RenderBlock::layoutOnlyPositionedObjects): Cover the case of
              a positioned object that has only moved and has only positioned children
              that have changed.
              * rendering/RenderObject.h:
              (WebCore::RenderObject::needsPositionedMovementLayout): Added this
              accessor.
      
      LayoutTests:
      
              Reviewed by John Sullivan.
      
              - test for https://bugs.webkit.org/show_bug.cgi?id=18909
                <rdar://problem/5914165> REGRESSION (r31872-r31878): Viewed photos not closing completely at indycar.com
      
              * fast/dynamic/positioned-movement-with-positioned-children.html: Added.
              * platform/mac/fast/dynamic/positioned-movement-with-positioned-children-expected.checksum: Added.
              * platform/mac/fast/dynamic/positioned-movement-with-positioned-children-expected.png: Added.
              * platform/mac/fast/dynamic/positioned-movement-with-positioned-children-expected.txt: Added.
      
      
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@32966 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      a801ef1d
    • justin.garcia@apple.com's avatar
      WebCore: · e269419a
      justin.garcia@apple.com authored
      2008-05-07  Justin Garcia  <justin.garcia@apple.com>
      
              Reviewed by John Sullivan.
      
              <rdar://problem/5666354> Crashes in Mail at WebCore::RemoveNodeCommand::doApply
              
              When a single tab or series of tabs was copied, we weren't putting them into a 
              tab span.  On Paste, we would be given a text node with a single tab or series
              of tabs in it, and we would crash when inserting it at the start of a block
              (where that kind of text becomes completely unrendered).
      
              * editing/ReplaceSelectionCommand.cpp:
              (WebCore::ReplaceSelectionCommand::removeUnrenderedTextNodesAtEnds): Clean
              up correctly and early return if we removed everything that was inserted.
              * editing/markup.cpp:
              (WebCore::createMarkup):  Include the tab span when the only thing copied
              was a tab or a series of tabs.
      
      LayoutTests:
      
      2008-05-07  Justin Garcia  <justin.garcia@apple.com>
      
              Reviewed by John Sullivan.
      
              <rdar://problem/5666354> Crashes in Mail at WebCore::RemoveNodeCommand::doApply
      
              * editing/pasteboard/5761530-1-expected.txt: Added.
              * editing/pasteboard/5761530-1.html: Added.
              * editing/pasteboard/5761530-2-expected.txt: Added.
              * editing/pasteboard/5761530-2.html: Added.
      
      
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@32963 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      e269419a
    • ddkilzer@apple.com's avatar
      Added radar to changelog for r32959. · e2bc9c53
      ddkilzer@apple.com authored
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@32961 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      e2bc9c53
    • ddkilzer@apple.com's avatar
      WebCore: · 6e957d7c
      ddkilzer@apple.com authored
              Bug 18900: Password field has focus but can't type text (i.rememberthemilk.com)
              <https://bugs.webkit.org/show_bug.cgi?id=18900>
      
              Original patch by Aaron Golden and Tim Omernick.  Reviewed by Adele.
      
              Test: fast/forms/textfield-to-password-on-focus.html
      
              * html/HTMLInputElement.cpp:
              (WebCore::HTMLInputElement::setInputType): If the current node has
              focus, call updateFocusAppearance() to make sure its state is correct
              after changing its type.
      
      LayoutTests:
      
              Bug 18900: Password field has focus but can't type text (i.rememberthemilk.com)
              <https://bugs.webkit.org/show_bug.cgi?id=18900>
      
              Reviewed by Adele.
      
              * fast/forms/textfield-to-password-on-focus-expected.txt: Added.
              * fast/forms/textfield-to-password-on-focus.html: Added.
      
      
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@32959 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      6e957d7c
    • hausmann@webkit.org's avatar
      2008-05-07 Tor Arne Vestbø <tavestbo@trolltech.com> · a1814878
      hausmann@webkit.org authored
              Reviewed by Simon
      
              Add ~250 tests that fail to the skipped list so we have a sane baseline.
      
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@32957 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      a1814878
    • mitz@apple.com's avatar
      WebCore: · 0676eebf
      mitz@apple.com authored
              Reviewed by Darin Adler.
      
              - fix <rdar://problem/5914544> Crash in layoutInlineChildren()
      
              Test: fast/block/float/float-on-zero-height-line.html
      
              * rendering/RenderBlock.cpp:
              (WebCore::RenderBlock::markLinesDirtyInVerticalRange):
      
      LayoutTests:
      
              Reviewed by Darin Adler.
      
              - test for <rdar://problem/5914544> Crash in layoutInlineChildren()
      
              * fast/block/float/float-on-zero-height-line.html: Added.
              * platform/mac/fast/block/float/float-on-zero-height-line-expected.checksum: Added.
              * platform/mac/fast/block/float/float-on-zero-height-line-expected.png: Added.
              * platform/mac/fast/block/float/float-on-zero-height-line-expected.txt: Added.
      
      
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@32945 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      0676eebf
  6. 06 May, 2008 1 commit
    • weinig@apple.com's avatar
      WebCore: · a184230c
      weinig@apple.com authored
      2008-05-06  Adam Barth  <abarth-webkit@adambarth.com>
      
              Reviewed by Sam Weinig.
      
              https://bugs.webkit.org/show_bug.cgi?id=18725
              Implement asynchronous postMessage.
              MessageEvent no longer bubbles as per r1237 in the HTML 5 working draft.
      
              Collin Jackson <collinj-webkit@collinjackson.com> also contributed to this patch.
      
              Test: http/tests/security/postMessage/delivery-order.html
      
              * dom/MessageEvent.cpp:
              (WebCore::MessageEvent::MessageEvent):
              * page/DOMWindow.cpp:
              (WebCore::PostMessageTimer::PostMessageTimer):
              (WebCore::PostMessageTimer::event):
              (WebCore::PostMessageTimer::targetOrigin):
              (WebCore::PostMessageTimer::fired):
              (WebCore::DOMWindow::postMessage):
              (WebCore::DOMWindow::postMessageTimerFired):
              * page/DOMWindow.h:
              * page/DOMWindow.idl:
      
      LayoutTests:
      
      2008-05-06  Adam Barth  <abarth-webkit@adambarth.com>
      
              Reviewed by Sam Weinig.
      
              https://bugs.webkit.org/show_bug.cgi?id=18725
              Update tests for asynchronous postMessage.
      
              Collin Jackson <collinj-webkit@collinjackson.com> also contributed to this patch.
      
              * http/tests/messaging/cross-domain-message-send.html:
              * http/tests/messaging/resources/cross-domain-message-receive.html:
              * http/tests/security/cross-frame-access-delete.html:
              * http/tests/security/cross-frame-access-history-put.html:
              * http/tests/security/cross-frame-access-location-put.html:
              * http/tests/security/postMessage/delivery-order-expected.txt: Added.
              * http/tests/security/postMessage/delivery-order.html: Added.
              * http/tests/security/postMessage/invalid-origin-throws-exception-expected.txt:
              * http/tests/security/postMessage/invalid-origin-throws-exception.html:
              * http/tests/security/postMessage/origin-unaffected-by-base-tag-expected.txt:
              * http/tests/security/postMessage/origin-unaffected-by-base-tag.html:
              * http/tests/security/postMessage/origin-unaffected-by-document-domain-expected.txt:
              * http/tests/security/postMessage/origin-unaffected-by-document-domain.html:
              * http/tests/security/postMessage/resources/javascript-post-message-sender.html:
              * http/tests/security/postMessage/resources/post-message-listener.html:
              * http/tests/security/postMessage/resources/recv.js: Added.
              * http/tests/security/postMessage/target-origin-expected.txt:
              * http/tests/security/postMessage/target-origin.html:
              * http/tests/security/resources/cross-frame-iframe-for-delete-test.html:
              * http/tests/security/resources/cross-frame-iframe-for-history-put-test.html:
              * http/tests/security/resources/cross-frame-iframe-for-location-put-test.html:
              * http/tests/security/resources/xss-eval3.html:
              * http/tests/security/xss-eval.html:
      
      
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@32922 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      a184230c
  7. 05 May, 2008 4 commits
    • ap@webkit.org's avatar
      Reviewed by Darin. · 095e1f6b
      ap@webkit.org authored
              https://bugs.webkit.org/show_bug.cgi?id=11947
              nbsps should be converted to entities in innerHTML
      
              https://bugs.webkit.org/show_bug.cgi?id=18769
              replacing &nbsp; with spaces using regexp creates inconsistent result
      
              Tests: fast/dom/innerHTML-nbsp.html
                     fast/dom/innerHTML-escaping-attribute.html
      
              * editing/markup.cpp:
              (WebCore::appendAttributeValue):
              (WebCore::escapeContentText):
              (WebCore::appendEscapedContent):
              Added U+00a0/nbsp to the list of characters to escape.
      
      
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@32879 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      095e1f6b
    • hyatt@apple.com's avatar
      WebCore: · 3cb88e1a
      hyatt@apple.com authored
      2008-05-05  David Hyatt  <hyatt@apple.com>
      
              Fix for https://bugs.webkit.org/show_bug.cgi?id=18821.  Fix some bugs in both get/PutImageData of <canvas>.
      
              Reviewed by Eric
      
              New tests added in fast/canvas/
      
              * platform/graphics/cg/ImageBufferCG.cpp:
              (WebCore::ImageBuffer::getImageData):
              (WebCore::ImageBuffer::putImageData):
      
      LayoutTests:
      
      2008-05-05  David Hyatt  <hyatt@apple.com>
      
              Fix for https://bugs.webkit.org/show_bug.cgi?id=18821, putImageData incorrect when data has alpha.
      
              Reviewed by Eric
      
              * fast/canvas/canvas-alphaImageData-behavior-expected.txt: Added.
              * fast/canvas/canvas-alphaImageData-behavior.html: Added.
              * fast/canvas/canvas-alphaImageData-behavior.js: Added.
      
      
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@32878 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      3cb88e1a
    • mitz@apple.com's avatar
      Rubber-stamped by Adam Roben. · fcc46237
      mitz@apple.com authored
              - fix test failure due to Windows' preference for Mac results over
                generic results
      
              * platform/win/editing/selection/move-left-right-expected.txt: Copied from LayoutTests/editing/selection/move-left-right-expected.txt.
      
      
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@32873 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      fcc46237
    • mitz@apple.com's avatar
      WebCore: · cd6deb7d
      mitz@apple.com authored
              Reviewed by Dave Hyatt.
      
              - fix https://bugs.webkit.org/show_bug.cgi?id=18809
                Forms with block level generated content and absolutely positioned labels break inline layout (fixed on reflow)
      
              Test: fast/block/basic/adding-near-anonymous-block.html
      
              * rendering/RenderBlock.cpp:
              (WebCore::RenderBlock::addChildToFlow): When adding a floating or
              positioned object, if it follows an anonymous block, put it
              inside the anonymous block. When adding an inline, check if it comes
              after an anonymous block and put it in the anonymous block.
      
      LayoutTests:
      
              Reviewed by Dave Hyatt.
      
              - test, updated test and updated results for https://bugs.webkit.org/show_bug.cgi?id=18809
                Forms with block level generated content and absolutely positioned labels break inline layout (fixed on reflow)
      
              * fast/block/basic/adding-near-anonymous-block.html: Added.
              * media/video-controls-rendering.html: Changed to work around
              https://bugs.webkit.org/show_bug.cgi?id=18857 which now affects this
              test in its original form.
              * platform/mac/css1/box_properties/clear-expected.txt:
              * platform/mac/fast/block/basic/adding-near-anonymous-block-expected.checksum: Added.
              * platform/mac/fast/block/basic/adding-near-anonymous-block-expected.png: Added.
              * platform/mac/fast/block/basic/adding-near-anonymous-block-expected.txt: Added.
              * platform/mac/fast/forms/input-align-image-expected.txt:
              * platform/mac/media/audio-controls-rendering-expected.txt:
              * platform/mac/media/video-controls-rendering-expected.txt:
              * platform/mac/tables/mozilla/bugs/bug51140-expected.txt:
      
      
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@32872 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      cd6deb7d
  8. 04 May, 2008 4 commits
  9. 03 May, 2008 2 commits
    • weinig@apple.com's avatar
      2008-05-03 Sam Weinig <sam@webkit.org> · 2769a26c
      weinig@apple.com authored
              Reviewed by Dan Bernstein.
      
              Add tests that show our behavior when using closures from a global object
              whose frame has been navigated.
      
              Related to https://bugs.webkit.org/show_bug.cgi?id=17249.
      
              * fast/dom/Window/closure-access-after-navigation-iframe-expected.txt: Added.
              * fast/dom/Window/closure-access-after-navigation-iframe.html: Added.
              * fast/dom/Window/closure-access-after-navigation-window-expected.txt: Added.
              * fast/dom/Window/closure-access-after-navigation-window.html: Added.
              * fast/dom/Window/dom-access-from-closure-iframe-expected.txt: Added.
              * fast/dom/Window/dom-access-from-closure-iframe.html: Added.
              * fast/dom/Window/dom-access-from-closure-window-expected.txt: Added.
              * fast/dom/Window/dom-access-from-closure-window.html: Added.
              * fast/dom/Window/resources/closure-access-after-navigation-iframe-child-1.html: Added.
              * fast/dom/Window/resources/closure-access-after-navigation-iframe-child-2.html: Added.
              * fast/dom/Window/resources/closure-access-after-navigation-window-child-1.html: Added.
              * fast/dom/Window/resources/closure-access-after-navigation-window-child-2.html: Added.
              * fast/dom/Window/resources/dom-access-from-closure-iframe-child.html: Added.
              * fast/dom/Window/resources/dom-access-from-closure-window-child.html: Added.
              * fast/dom/Window/resources/notify-opener-done.html: Added.
              * fast/dom/Window/resources/notify-parent-done.html: Added.
      
      
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@32845 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      2769a26c
    • rwlbuis@webkit.org's avatar
      Reviewed by Eric. · 654496db
      rwlbuis@webkit.org authored
              https://bugs.webkit.org/show_bug.cgi?id=18568
              background: currentColor fails
      
              Implement currentColor from CSS3 color module.
      
              Tests: fast/css/background-currentcolor.html
      
      
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@32836 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      654496db
  10. 02 May, 2008 2 commits
    • mitz@apple.com's avatar
      WebCore: · b74d5ba6
      mitz@apple.com authored
              Reviewed by John Sullivan.
      
              - render text shadows with zero offset, as the shadow can be seen behind
                the text if the text is translucent
      
              Test: fast/text/shadow-no-blur.html
      
              * platform/graphics/mac/FontMac.mm:
              (WebCore::Font::drawComplexText): Removed the isEmpty() condition on the
              shadow offset. Since isEmpty() is also true for sizes that have one or
              more non-positive components, this also ensures that subpixel
              antialiasing of the text is maintained with shadows in all directions.
              (WebCore::Font::drawGlyphs): Ditto.
              * platform/graphics/win/FontCGWin.cpp:
              (WebCore::Font::drawGlyphs): Ditto.
      
      LayoutTests:
      
              Reviewed by John Sullivan.
      
              - add pixel test for text shadows with zero blur and various offsets,
                including zero offset
      
              * fast/text/shadow-no-blur.html: Added.
              * platform/mac/fast/text/shadow-no-blur-expected.checksum: Added.
              * platform/mac/fast/text/shadow-no-blur-expected.png: Added.
              * platform/mac/fast/text/shadow-no-blur-expected.txt: Added.
      
      
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@32817 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      b74d5ba6
    • hausmann@webkit.org's avatar
      2008-05-02 Tor Arne Vestbø <tavestbo@trolltech.com> · 7b9e3957
      hausmann@webkit.org authored
      Updated Qt layout-tests to match changes in font metrics.
      
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@32816 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      7b9e3957
  11. 01 May, 2008 5 commits
    • weinig@apple.com's avatar
      2008-05-01 Sam Weinig <sam@webkit.org> · 50d9efc7
      weinig@apple.com authored
              Add missing result.
      
              * fast/canvas/canvas-path-with-inf-nan-dimensions-expected.txt: Added.
      
      
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@32796 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      50d9efc7
    • weinig@apple.com's avatar
      WebCore: · 796309ca
      weinig@apple.com authored
      2008-05-01  Sam Weinig  <sam@webkit.org>
      
              Reviewed by Geoffrey Garen.
      
              Fixes:
                - https://bugs.webkit.org/show_bug.cgi?id=17249
                  Incorrect lexical scope after navigation leads to UXSS
                  <rdar://problem/5738497>
      
                - https://bugs.webkit.org/show_bug.cgi?id=16824
                  Script authorization should follow lexical (not dynamic) scope
                  <rdar://problem/5683032>
      
              This patch changes us to perform same-origin checks based on the lexical global object) 
              rather than dynamic global object, which is now possible we don't re-use the window on 
              navigations, but rather switch in a new one and re-use the outer shell.  This is both
              more secure and conforms with the HTML5 specification.  Now that all the checks are
              done based on the lexical global object, we can remove the SecurityOrigin::Reason
              concept, as it was only around to work around an ebay.com bug that required the check to
              be done that way.
      
              An important thing to note is that we currently implement a stricter than necessary policy
              and perform the same-origin check based on the currently active global object to avoid leaking
              the document in cases when the target frame is navigated before access.  This will be fixed in
              an upcoming patch.
      
              * bindings/js/JSDOMWindowBase.cpp:
              (WebCore::JSDOMWindowBase::allowsAccessFrom):
              (WebCore::JSDOMWindowBase::allowsAccessFromNoErrorMessage):
              (WebCore::JSDOMWindowBase::allowsAccessFromPrivate):
              (WebCore::JSDOMWindowBase::crossDomainAccessErrorMessage):
              (WebCore::JSDOMWindowBase::printErrorMessage):
              (WebCore::asJSDOMWindow):
              * bindings/js/JSDOMWindowBase.h:
              * html/CanvasRenderingContext2D.cpp:
              (WebCore::CanvasRenderingContext2D::checkOrigin):
              (WebCore::CanvasRenderingContext2D::createPattern):
              * loader/FrameLoader.cpp:
              (WebCore::FrameLoader::begin):
              (WebCore::FrameLoader::write):
              (WebCore::FrameLoader::setOpener):
              (WebCore::FrameLoader::shouldAllowNavigation):
              * page/DOMWindow.h:
              (WebCore::DOMWindow::setSecurityOrigin):
              (WebCore::DOMWindow::securityOrigin):
              (WebCore::DOMWindow::setURL):
              (WebCore::DOMWindow::url):
              * platform/SecurityOrigin.cpp:
              (WebCore::SecurityOrigin::canAccess):
              (WebCore::SecurityOrigin::isSecureTransitionTo):
              * platform/SecurityOrigin.h:
      
      LayoutTests:
      
      2008-05-01  Sam Weinig  <sam@webkit.org>
      
              Reviewed by Geoffrey Garen.
      
              * http/tests/security/listener/xss-JSTargetNode-onclick-addEventListener-expected.txt:
              * http/tests/security/listener/xss-JSTargetNode-onclick-shortcut-expected.txt:
              * http/tests/security/listener/xss-XMLHttpRequest-addEventListener-expected.txt:
              * http/tests/security/listener/xss-XMLHttpRequest-shortcut-expected.txt:
              * http/tests/security/listener/xss-window-onclick-addEventListener-expected.txt:
              * http/tests/security/listener/xss-window-onclick-shortcut-expected.txt:
              * http/tests/security/xss-eval-expected.txt:
      
      
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@32791 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      796309ca
    • andersca@apple.com's avatar
      2008-05-01 Anders Carlsson <andersca@apple.com> · 91a53dd5
      andersca@apple.com authored
              Reviewed by Mitz.
      
              Disable an SVG font test that crashes. 
              
              http://bugs.webkit.org/show_bug.cgi?id=18830 tracks fixing the crash.
              
              * svg/W3C-SVG-1.1/fonts-glyph-02-t.svg: Removed.
              * svg/W3C-SVG-1.1/fonts-glyph-02-t.svg-disabled: Copied from svg/W3C-SVG-1.1/fonts-glyph-02-t.svg.
      
      
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@32785 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      91a53dd5
    • mitz@apple.com's avatar
      WebCore: · 63ccccdb
      mitz@apple.com authored
      2008-05-01  Anatoli Papirovski  <apapirovski@mac.com>
      
              Reviewed by Dave Hyatt.
      
              - fix https://bugs.webkit.org/show_bug.cgi?id=18347
                Absolutely positioned image percentage width does not respect container's padding
      
              Test: fast/replaced/absolute-position-percentage-width.html
      
              * rendering/RenderBox.cpp:
              (WebCore::RenderBox::calcReplacedWidthUsing):
      
      LayoutTests:
      
      2008-05-01  Anatoli Papirovski  <apapirovski@mac.com>
      
              Reviewed by Dave Hyatt.
      
              - added a test case for https://bugs.webkit.org/show_bug.cgi?id=18347
                Absolutely positioned image percentage width does not respect container's padding
      
              * fast/replaced/absolute-position-percentage-width.html: Added.
              * platform/mac/fast/replaced/absolute-position-percentage-width-expected.checksum: Added.
              * platform/mac/fast/replaced/absolute-position-percentage-width-expected.png: Added.
              * platform/mac/fast/replaced/absolute-position-percentage-width-expected.txt: Added.
      
      
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@32784 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      63ccccdb
    • eric@webkit.org's avatar
      Rubber-stamped by Sam Weinig. · 95bb948a
      eric@webkit.org authored
              
              Mark a couple UTF16 files as binary so windows-svn doesn't mutilate them on checkout.
      
              * fast/xsl/xslt-enc16.xml: add svn:mime-type application/octet-stream
              * fast/xsl/xslt-enc16to16.xml: add svn:mime-type application/octet-stream
      
      
      git-svn-id: http://svn.webkit.org/repository/webkit/trunk@32776 268f45cc-cd09-0410-ab3c-d52691b4dbfc
      95bb948a