Bug 82847 – Add mechanism for mapping from StyleRules back to fully constructed CSSStyleRules

• Status: RESOLVED
• Resolution: FIXED
• Component: CSS

[Antti Koivisto]

Inspector is using CSSStyleSelector to calculate the CSS rules matched by a given element and expects to be able to walk the parent chain. After 82728 the stylesheet object tree won't have parent pointers and we are going to need another mechanism to support this.

[Antti Koivisto]

Created attachment 135012 [details]
patch

[Andreas Kling]

Comment on attachment 135012 [details]
patch

View in context: https://bugs.webkit.org/attachment.cgi?id=135012&action=review

> Source/WebCore/ChangeLog:3
> +        Add mechanism for mapping from StyleRules back to fully constructed CSSStyleRules 

NO BUG NUMBER R- BRO

[Antti Koivisto]

Created attachment 135013 [details]
with bug url

[WebKit Review Bot]

Attachment 135013 [details] did not pass style-queue:

Failed to run "['Tools/Scripts/check-webkit-style', '--diff-files', u'Source/WebCore/ChangeLog', u'Source/WebCor..." exit_code: 1
Source/WebCore/css/CSSStyleSelector.cpp:2927:  Weird number of spaces at line-start.  Are you using a 4-space indent?  [whitespace/indent] [3]
Total errors found: 1 in 5 files


If any of these errors are false positives, please file a bug against check-webkit-style.

[Andreas Kling]

Comment on attachment 135013 [details]
with bug url

View in context: https://bugs.webkit.org/attachment.cgi?id=135013&action=review

> Source/WebCore/css/CSSStyleSelector.cpp:292
> +static CSSStyleSheet* fullscreenStyleSheet;

fullScreenStylesheet;

> Source/WebCore/css/CSSStyleSelector.h:495
> +    HashMap<StyleRule*, RefPtr<CSSStyleRule> >  m_styleRuleToCSSOMWrapperMap;

Style: Two spaces after >.

[Andreas Kling]

Comment on attachment 135013 [details]
with bug url

TD;DR

[Sam Weinig]

Would it make sense to just switch the inspector to using the internal representation?

[Antti Koivisto]

(In reply to comment #7)
> Would it make sense to just switch the inspector to using the internal representation?

Sure but that's a separate refactoring. It also doesn't solve this particular problem.

[Andreas Kling]

Comment on attachment 135013 [details]
with bug url

View in context: https://bugs.webkit.org/attachment.cgi?id=135013&action=review

r=me. Let's do this thing.

> Source/WebCore/css/CSSStyleSelector.cpp:2905
> +    const Vector<RefPtr<StyleSheet> >& styleSheets = document->styleSheets()->vector();
> +    for (unsigned i = 0; i < styleSheets.size(); ++i) {

You could use the iteration helpers of StyleSheetList directly here instead of grabbing the Vector. NABD.

> Source/WebCore/css/CSSStyleSelector.h:247
> +    CSSStyleRule* ensureFullCSSOMWrapperForStyleRule(StyleRule*);

As you mentioned, this could do with a scarier name, e.g ensureFullCSSOMWrapperForInspectorWhileUsingTerabytesOfMemory().

[Antti Koivisto]

http://trac.webkit.org/changeset/112858