https://bugs.webkit.org/show_bug.cgi?id=35991
Would be faster to not use a thread specific to implement StringImpl::empty()

Reviewed by Darin Adler, Geoffrey Garen, Maciej Stachowiak.

JavaScriptCore:

Change JSC::UStringImpl's implementation of empty() match to match StringImpl's new implementation
(use a static defined within the empty() method), and change the interface to match too (return
a pointer not a reference).

~0% performance impact (possible minor progression from moving empty() from .h to .cpp).

* JavaScriptCore.exp:
* runtime/Identifier.cpp:
(JSC::Identifier::add):
(JSC::Identifier::addSlowCase):
* runtime/PropertyNameArray.cpp:
(JSC::PropertyNameArray::add):
* runtime/UString.cpp:
(JSC::initializeUString):
(JSC::UString::UString):
* runtime/UStringImpl.cpp:
(JSC::UStringImpl::empty):
(JSC::UStringImpl::create):
* runtime/UStringImpl.h:
(JSC::UStringImpl::adopt):
(JSC::UStringImpl::createUninitialized):
(JSC::UStringImpl::tryCreateUninitialized):

WebCore:

Copy JavaScriptCore in making 'static' strings threadsafe, make the empty string a static,
shared by all threads.

~2% progression on Dromaeo DOM core & JS lib tests.

* platform/ThreadGlobalData.cpp:
(WebCore::ThreadGlobalData::ThreadGlobalData):
(WebCore::ThreadGlobalData::~ThreadGlobalData):
* platform/ThreadGlobalData.h:
(WebCore::ThreadGlobalData::eventNames):
* platform/text/StringImpl.cpp:
(WebCore::StringImpl::StringImpl):
(WebCore::StringImpl::empty):
* platform/text/StringImpl.h:
(WebCore::StringImpl::deref):
(WebCore::StringImpl::hasOneRef):

WebCore: Adding support for the optional creation callback that could be
passed to openDatabase().

Reviewed by Adam Barth.

Tests: storage/open-database-creation-callback.html
storage/open-database-creation-callback-isolated-world.html

https://bugs.webkit.org/show_bug.cgi?id=34726

* Android.jscbindings.mk
* Android.v8bindings.mk
* GNUmakefile.am:
* WebCore.gypi:
* WebCore.pro:
* WebCore.vcproj/WebCore.vcproj:
* WebCore.xcodeproj/project.pbxproj:
* bindings/js/JSDOMWindowCustom.cpp:
(WebCore::JSDOMWindow::openDatabase):
* bindings/js/JSDatabaseCallback.cpp: Added.
(WebCore::JSDatabaseCallback::JSDatabaseCallback):
(WebCore::JSDatabaseCallback::~JSDatabaseCallback):
(WebCore::JSDatabaseCallback::handleEvent):
* bindings/js/JSDatabaseCallback.h: Added.
(WebCore::JSDatabaseCallback::create):
* bindings/v8/custom/V8DOMWindowCustom.cpp:
(WebCore::V8DOMWindow::openDatabaseCallback):
* bindings/v8/custom/V8DatabaseCallback.cpp: Added.
(WebCore::V8DatabaseCallback::V8DatabaseCallback):
(WebCore::V8DatabaseCallback::~V8DatabaseCallback):
(WebCore::V8DatabaseCallback::handleEvent):
* bindings/v8/custom/V8DatabaseCallback.h: Added.
(WebCore::V8DatabaseCallback::create):
* page/DOMWindow.cpp:
(WebCore::DOMWindow::openDatabase):
* page/DOMWindow.h:
* page/DOMWindow.idl:
* storage/Database.cpp:
(WebCore::DatabaseCreationCallbackTask::create):
(WebCore::DatabaseCreationCallbackTask::performTask):
(WebCore::DatabaseCreationCallbackTask::DatabaseCreationCallbackTask):
(WebCore::Database::openDatabase):
(WebCore::Database::Database):
(WebCore::Database::performOpenAndVerify):
(WebCore::Database::performCreationCallback):
* storage/Database.h:
(WebCore::Database::isNew):
* storage/DatabaseCallback.h: Added.
(WebCore::DatabaseCallback::~DatabaseCallback):
* workers/WorkerContext.cpp:
(WebCore::WorkerContext::openDatabase):
* workers/WorkerContext.h:

LayoutTests: Adding a test to test the creation callback passed to
openDatabase() and another one to make sure it is executed in the
correct world.

Reviewed by Adam Barth.

Also, fixing fast/frames/sandboxed-iframe-storage.html: calling
openDatabase() without at least 4 parameters should always fail,
because the database name, version, description and estimated size
are not optional according to the spec.

https://bugs.webkit.org/show_bug.cgi?id=34726

* fast/frames/resources/sandboxed-iframe-storage-allowed.html:
* fast/frames/resources/sandboxed-iframe-storage-disallowed.html:
* fast/frames/sandboxed-iframe-storage-expected.txt:
* platform/qt/Skipped:
* storage/open-database-creation-callback-expected.txt: Added.
* storage/open-database-creation-callback.html: Added.
* storage/open-database-creation-callback-isolated-world-expected.txt: Added.
* storage/open-database-creation-callback-isolated-world.html: Added.

Implementing DOMFormData class.
https://bugs.webkit.org/show_bug.cgi?id=35707

Reviewed by Dmitry Titov.

This patch only addresses the implementation of DOMFormData class and
moves the FormData construction logic from HTMLFormElement::createFormData
to FormData::create() so that it can be used by both HTMLFormElement
and XMLHttpRequest.

The DOMFormData IDL interface will be exposed in another patch and the
test will be added then.

* Android.mk:
* GNUmakefile.am:
* WebCore.gypi:
* WebCore.pro:
* WebCore.vcproj/WebCore.vcproj:
* WebCore.xcodeproj/project.pbxproj:
* html/DOMFormData.cpp: Added.
* html/DOMFormData.h: Added.
* html/FormDataList.h:
(WebCore::FormDataList::encoding):
* html/HTMLFormElement.cpp:
(WebCore::HTMLFormElement::createFormData):
(WebCore::HTMLFormElement::submit):
* html/HTMLFormElement.h:
* platform/network/FormData.cpp:
(WebCore::FormData::create):
(WebCore::FormData::createMultiPart):
(WebCore::FormData::appendDOMFormData):
* platform/network/FormData.h:
(WebCore::FormData::boundary):
* xml/XMLHttpRequest.cpp:
(WebCore::XMLHttpRequest::send):
* xml/XMLHttpRequest.h:

Skipping test on GTK and WIN. No Review.

VoiceOver can navigate to hidden content in widget (WAI-ARIA)
https://bugs.webkit.org/show_bug.cgi?id=35986

Thought this test might pass on the other platforms, but no.

* platform/gtk/Skipped:
* platform/win/Skipped:

Not reviewed, fixing Snow Leopard build.

* wtf/mac/MainThreadMac.mm: Forgot 'static' for a new local function.
(WTF::postTimer):

Windows build fix: added new symbol.

* JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.def:

Windows build fix: removed old symbol.

* JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.def:

Refactored fastCheckConsistency to match some review comments:
- renamed fastCheckConsistency to fastMallocSize, and changed ValueCheck
to ASSERT that a pointer's fastMallocSize is not 0.
- implemented a version of fastMallocSize for tcmalloc.

Reviewed by Alexey Proskuryakov, Darin Adler, and Sam Weinig.

Also moved some pre-existing code around to avoid a problem related to
mismatched #define/#undef of malloc/free in this source file.

* JavaScriptCore.exp:
* wtf/FastMalloc.cpp:
(WTF::fastMallocSize): Renamed. Fixed indentation.

(WTF::TCMalloc_PageHeap::scavenge): Removed an incorrect ASSERT that
got in the way of testing the tcmalloc implementation. (More information
on why this ASSERT is incorrect is in <rdar://problem/7165917>.)

(WTF::TCMallocStats::fastMallocSize): Implemented for tcmalloc.

* wtf/FastMalloc.h: Updated for rename.

* wtf/ValueCheck.h:
(WTF::): Moved the ASSERT that used to be in fastCheckConsistency here.

2010-03-10 Eric Uhrhane <ericu@chromium.org>

Reviewed by David Levin.

The build fix for my patch on bug #35763 wasn't quite right--it removed
the call entirely, instead of replacing it with the new API. This adds
the call to Database::setIsAvailable.

https://bugs.webkit.org/show_bug.cgi?id=35763

* WebView.cpp: Added a call to Database::setIsAvailable where change 55667 removed the old Settings API call <http://trac.webkit.org/changeset/55667>.
(WebView::notifyPreferencesChanged):

Roll out the prepare-ChangeLog part of r55870

This change wasn't needed (prepare-ChangeLog calls svn-create-patch
when it's asked to print out diffs for the user) and was screwing up
its ability to find changed function names.

* Scripts/prepare-ChangeLog:
(diffCommand): Change the options we pass to svn-diff back to their
pre-r55870 form.