servo/components/style/gecko
Bobby Holley e65b1be07b Revert restyled_previous_sibling_element tracking, and separate child preprocessing.
I realized that I fixed this issue incorrectly when the test failed before. Our design
document specifies that restyle hints must be expanded by the parent before traversing
children, so that we can properly apply LaterSiblings restyle hints. This includes
parents that do not themselves need processing (StylingMode::Traverse).

So we need to preprocess children even in the case where we don't restyle the parent.
On the flip side, we do in fact know whether a child needs processing before enqueuing
it, so we can skip the conservative visit I added before.

MozReview-Commit-ID: AEiRzdsN0h5
2016-11-24 17:07:40 -08:00
..
generated Move most of geckolib into style::gecko 2016-09-26 09:07:17 +02:00
context.rs Bug 1317016 - Basic infrastructure for RestyleHint-driven traversal. 2016-11-24 17:07:38 -08:00
conversions.rs Add stylo FFI for CSSStyleRule 2016-11-24 10:28:38 +11:00
data.rs Rename selector_matching.rs to stylist.rs 2016-11-20 15:29:09 +01:00
mod.rs Rename selector_impl.rs to selector_parser.rs 2016-11-20 15:30:37 +01:00
restyle_damage.rs Bug 1317016 - Basic infrastructure for RestyleHint-driven traversal. 2016-11-24 17:07:38 -08:00
selector_parser.rs Remove SelectorImpl aliases 2016-11-22 14:50:22 +01:00
snapshot.rs Bug 1317016 - Basic infrastructure for RestyleHint-driven traversal. 2016-11-24 17:07:38 -08:00
snapshot_helpers.rs Move most of geckolib into style::gecko 2016-09-26 09:07:17 +02:00
traversal.rs Revert restyled_previous_sibling_element tracking, and separate child preprocessing. 2016-11-24 17:07:40 -08:00
values.rs Prefer Either<A, B> for LengthOrNumber 2016-11-18 09:14:09 +05:30
wrapper.rs Bug 1317016 - Basic infrastructure for RestyleHint-driven traversal. 2016-11-24 17:07:38 -08:00