⚠ Archived content — this site is no longer maintained.   Current WebKit documentation is at docs.webkit.org.

Timeline



Dec 25, 2005:

10:28 AM Changeset in webkit [11765] by ggaren
  • 1 edit in trunk/WebCore/ChangeLog

ChangeLog fixup

3:05 AM Changeset in webkit [11764] by mjs
  • 5 edits in trunk/WebKit

Reviewed by Geoff.

  • WebView.subproj/WebDataSource.m: (-[WebDataSource _archiveWithMarkupString:nodes:]): Assert that the data source is committed, doesn't make sense to archive otherwise. (-[WebDataSource _subframeArchivesWithCurrentState:]): New helper method. (-[WebDataSource _archiveWithCurrentState:]): Assert that the data source is committed. Use the helper. (-[WebDataSource _setWebView:]): Comment that we won't tell subframes that defers callback changed. (-[WebDataSource _startLoading]): Remove stray space. (-[WebDataSource _stopLoading]): Only handle local _stopLoading business. Cound on WebFrame to tell subframes to stop loading. Fold in _stopLoadingInternal and remove _recursiveStopLoading. (-[WebDataSource _startLoading:]): Clean up an assert slightly. (-[WebDataSource _setTitle:]): Remove stray spaces. (-[WebDataSource _defersCallbacksChanged]): Don't call subframes. WebFrame can do that. (-[WebDataSource isLoading]): Move checking of subframes down to WebFrame. (-[WebDataSource webArchive]): Return nil if the data source is not yet committed. It makes no sense to archive a provisional data source since it has no data yet. (-[WebDataSource addSubresource:]): Replace an assert with an early return, ASSERT is not an appropriate way to flag problems with the argument of a public method.
  • WebView.subproj/WebDataSourcePrivate.h:
  • WebView.subproj/WebFrame.m: (-[WebFrame _setLoadType:]): Remove stray space. (-[WebFrame _checkLoadCompleteForThisFrame]): Stop loading subframes manually and add a FIXME about confusingness of stopping loading here. (-[WebFrame _defersCallbacksChanged]): Tell our subframes. (-[WebFrame _addChild:]): Remove stray space. (-[WebFrame _stopLoadingSubframes]): New helper method. (-[WebFrame _subframeIsLoading]): New helper method, code moved from WebDataSource. (-[WebFrame stopLoading]): Tell subframes to stop loading. (-[WebFrame reload]): Remove extra braces.
  • WebView.subproj/WebFrameInternal.h:
1:22 AM Changeset in webkit [11763] by mjs
  • 7 edits in trunk/JavaScriptCore

Reviewed by Eric and Dave Hyatt.

This is a somewhat cheesy change. Having to use PassRefPtr_Ref creates ambiguities
in assignment and copy construction. And this makes life way easier and removes
the need for pass(). It is not really correct, but we pretty much never need a real
const PassRefPtr, and this takes care of things for PassRefPtr temporaries.

  • kjs/identifier.cpp: (KJS::Identifier::add): No more need for pass()
  • kjs/property_map.cpp: (KJS::PropertyMap::addSparseArrayPropertiesToReferenceList): No more need for pass()
  • kjs/ustring.cpp: (KJS::UString::Rep::create): Use adoptRef (KJS::UString::UString): No more need for pass (KJS::UString::append): No more need for pass (KJS::UString::substr): No more need for pass
  • kxmlcore/PassRefPtr.h: made m_ptr mutable (ugh) (KXMLCore::PassRefPtr::PassRefPtr): Take a const PassRefPtr reference (KXMLCore::PassRefPtr::release): Made this a const method (ugh) (KXMLCore::PassRefPtr::operator=): clean up appropriately (KXMLCore::adoptRef): Added this to use instead of PassRefPtr<T>::adopt, I think it makes the behavior more clear and it is less verbose. (KXMLCore::static_pointer_cast): use adoptRef (KXMLCore::const_pointer_cast): use adoptRef
  • kxmlcore/RefPtr.h: (KXMLCore::RefPtr::RefPtr): take const PassRefPtr& (KXMLCore::PassRefPtr::operator=): take const PassRefPtr&
12:11 AM Changeset in webkit [11762] by eseidel
  • 2 edits in trunk/JavaScriptCore

Bug #: none
Submitted by: eseidel
Reviewed by: mjs

Unbreak HashTableConstIterator++ by returning const_iterator

  • kxmlcore/HashTable.h: (KXMLCore::HashTableConstIterator::operator++): use const_iterator

Dec 24, 2005:

11:52 PM Changeset in webkit [11761] by eseidel
  • 2 edits in trunk/JavaScriptCore

Bug #: none
Submitted by: eseidel
Reviewed by: mjs

Un-break HashTable copy constructor.

  • kxmlcore/HashTable.h: (KXMLCore::::HashTable): use const_iterator instead
9:11 PM Changeset in webkit [11760] by eseidel
  • 1 edit
    2 adds in trunk/LayoutTests

Bug #: 6236
Submitted by: mitz
Reviewed by: eseidel

  • fast/forms/domstring-replace-crash-expected.txt: Added.
  • fast/forms/domstring-replace-crash.html: Added.
9:11 PM Changeset in webkit [11759] by eseidel
  • 2 edits in trunk/WebCore

Bug #: 6326
Submitted by: mitz
Reviewed by: eseidel

Test: fast/forms/domstring-replace-crash.html

  • khtml/dom/dom_string.h: (DOMString::replace): Added check for null m_impl.
2:45 PM Changeset in webkit [11758] by mjs
  • 3 edits in trunk/WebCore

Reviewed by Eric.

  • khtml/ecma/kjs_html.cpp: (KJS::HTMLElement::accessors): (KJS::HTMLElement::getValueProperty): (KJS::HTMLElement::putValueProperty):
  • khtml/ecma/kjs_html.h:
2:06 PM Changeset in webkit [11757] by mjs
  • 2 edits in trunk/JavaScriptCore

Reviewed by Eric.

  • kxmlcore/HashMapPtrSpec.h: (KXMLCore::HashMap): In all methods, explicitly cast all pointers to void * before passing to internal implementation. Use C-style casts instead of new-style casts, because the real solution would require a combo of reinterpret_cast anc const_cast.
2:00 PM Changeset in webkit [11756] by ggaren
  • 4 edits in trunk/WebCore

Reviewed by eric.

​http://bugzilla.opendarwin.org/show_bug.cgi?id=6231

  • Fixed assertion failure in QString::mid caused by my last checkin. parseWindowFeatuers() sometimes called QString::mid() with a start index past the end of the string but a length of 0. There were two problems there. (1) parseWindowString() shouldn't have gotten past the end of the string. (2) QString had range truncating code that was borked. So I fixed both.
  • khtml/ecma/kjs_window.cpp: (KJS::parseWindowFeatures): One of the "past the end" tests was > instead of >=.
  • kwq/KWQString.mm: (QString::mid): Truncate length *before* deciding whether to return early. Also, compare start and data._length before computing data._length - start, because the error case we're trying to catch will almost certainly cause unsigned overflow.
  • manual-tests/window-open-features-parsing.html: Added trailing ',' to uber-test to catch this failure, too.
2:02 AM Changeset in webkit [11755] by ap
  • 4 edits in trunk/WebCore

Reviewed by Maciej.

  • khtml/kwq/KWQSlot.mm: (KWQSlot::KWQSlot): Changed the parameter type of slotTextChanged to const DOMString &
  • khtml/rendering/render_form.h:
  • khtml/rendering/render_form.cpp: (RenderLineEdit::RenderLineEdit): ditto. (RenderLineEdit::slotTextChanged): ditto. (RenderFileButton::RenderFileButton): ditto. (RenderFileButton::slotTextChanged): ditto.
Note: See TracTimeline for information about the timeline view.