[Webkit-unassigned] [Bug 88606] [Shadow DOM] Needs @host rule for ShadowDOM styling

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=88606





--- Comment #9 from Dimitri Glazkov (Google) <dglazkov at chromium.org>  2012-07-31 11:29:09 PST ---
(From update of attachment 155477)
View in context: https://bugs.webkit.org/attachment.cgi?id=155477&action=review

There are several patches in here, to be landed separately. It's fine to work on this as one patch for now, however.

ProTip: don't fill out the ChangeLog until you're ready to start landing. Otherwise, you'll end up rewriting it many times :)

> Source/WebCore/html/HTMLStyleElement.cpp:155
> +    if (hasHostRule())
> +        if (ShadowRoot* sr = shadowRoot())
> +            if (Element* host = sr->host()) {
> +                host->registerScopedHTMLStyleChild();
> +                isHostRuleRegistered = true;
> +            }

Wild to see this here. Why can't we do this when we're adding rules in StyleResolver, like all other machinery in CSS does?

> Source/WebCore/html/HTMLStyleElement.cpp:185
> +        if (m_scopedStyleRegistrationState == RegisteredInShadowRootWithHostRule || m_scopedStyleRegistrationState == RegisteredAsScopedWithHostRule)
> +            if (ShadowRoot* sr = shadowRoot())
> +                if (Element* host = sr->host())
> +                    host->unregisterScopedHTMLStyleChild();

You won't need any of these when you move @host accounting in styleresolver.

> Source/WebCore/html/HTMLStyleElement.cpp:200
> +bool HTMLStyleElement::hasHostRule() const
> +{
> +    if (CSSStyleSheet* styleSheet = const_cast<HTMLStyleElement*>(this)->sheet())
> +        return styleSheet->contents()->hasHostRule();
> +    return false;
> +}

Or these.

> Source/WebCore/html/shadow/HTMLShadowElement.h:56
> +    bool m_registeredWithShadowRoot;

It seems we're building an extra set of machinery for something that already exists. We already _know_ which shadow trees are rendered or not by the time we've created the composed tree, which happens before style resolution. Please ask morrita@/hayato@ for guidance here.

-- 
Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.