[webkit-changes] cvs commit:
WebCore/layout-tests/fast/css-generated-content 013-expected.txt
013.html
Dave Hyatt
hyatt at apple.com
Fri Jun 10 19:00:36 PDT 2005
Yes, that is a better fix.
dave
On Jun 10, 2005, at 3:18 AM, Allan Sandfeld Jensen wrote:
> On Friday 10 June 2005 08:47, David wrote:
>
>> Index: cssstyleselector.cpp
>> ===================================================================
>> RCS file: /cvs/root/WebCore/khtml/css/cssstyleselector.cpp,v
>> retrieving revision 1.183
>> retrieving revision 1.184
>> diff -u -r1.183 -r1.184
>> --- cssstyleselector.cpp 11 May 2005 05:49:34 -0000 1.183
>> +++ cssstyleselector.cpp 10 Jun 2005 06:47:36 -0000 1.184
>> @@ -3042,8 +3042,12 @@
>> // just the hardcoded HTML set. Can a
>> namespace be
>> specified for // an attr(foo)?
>> int attrID = element->getDocument()->attrId(0,
>> val->getStringValue().implementation(), false); - if
>> (attrID)
>> -
>> style->setContent(element->getAttribute(attrID).implementation(),
>> i != 0);
>> + if (attrID) {
>> + DOMStringImpl* v =
>> element->getAttribute(attrID).implementation();
>> + if
>> (!v)
>> + v = DOMStringImpl::empty(); // Don't
>> allow v to
>> be null, since we want to concatenate this string even if it's
>> empty. +
>> style->setContent(v, i != 0);
>> + }
>> }
>> else if (val->primitiveType()
>> ==CSSPrimitiveValue::CSS_URI)
>> {
>>
> Wouldn't be better solved by allowing to concatenate to 0 content in
> render_style.cpp?
>
> Index: render_style.cpp
> ===================================================================
> --- render_style.cpp (revision 423202)
> +++ render_style.cpp (working copy)
> @@ -668,10 +668,7 @@
> lastContent = lastContent->_nextContent;
>
> bool reuseContent = !add;
> - if (add) {
> - if (!lastContent)
> - return; // Something's wrong. We had no previous
> content, and we
> should have.
> -
> + if (add && lastContent) {
> if (lastContent->_contentType == CONTENT_TEXT) {
> // We can augment the existing string and share this
> ContentData
> node.
> DOMStringImpl* oldStr = lastContent->_content.text;
>
> `Allan
> _______________________________________________
> Do not post admin requests to the list. They will be ignored.
> Webcore-dev mailing list (Webcore-dev at lists.apple.com)
> Help/Unsubscribe/Update your Subscription:
> http://lists.apple.com/mailman/options/webcore-dev/hyatt%40apple.com
>
> This email sent to hyatt at apple.com
>
More information about the webkit-changes
mailing list