When @media rules change match state, avoid treating that as a
whole-document stylesheet change. Record the rules under the query
that became effective or ineffective, build the normal stylesheet
invalidation set from those rules, and invalidate only matching
elements. Keep broad invalidation for rule kinds whose effects are not
selector-targetable.
Preserve cascade-wide fallout by reusing stylesheet-change shadow
effect analysis for broad invalidations inside shadow roots. This keeps
:host and slotted content current when active rules in the same shadow
scope can match there.
Also report an imported stylesheet's owner rule when the imported
sheet's own media gate changes. Layered @import rules can affect layer
ordering even when the imported sheet contributes no rules, so they
need the same broad invalidation treatment as other cascade-wide rules.
Add viewport resize coverage for media query breakpoints, broad
shadow-root media invalidation, and empty layered imports whose media
gate changes layer order.
Previously we would always use the window's viewport which was incorrect
if we were within an iframe.
This is likely applicable to all uses of
`Length::ResolutionContext::for_window`.
Keep track of which CSSRule owns a CSSRuleList, and then use that to
produce a stack of RuleContexts for the CSS Parser to use.
There are certainly other places we should do this!
The spec algorithm changed at some point to support nested declarations,
but I only just noticed. The subtest regression is one we were passing
incorrectly.
Resulting in a massive rename across almost everywhere! Alongside the
namespace change, we now have the following names:
* JS::NonnullGCPtr -> GC::Ref
* JS::GCPtr -> GC::Ptr
* JS::HeapFunction -> GC::Function
* JS::CellImpl -> GC::Cell
* JS::Handle -> GC::Root
