This also fixes a bug where task IDs were being deallocated from the
wrong IDAllocator. I don't know if it was actually possible to cause any
real trouble with that mistake, nor do I know how to write a test for
it, but this makes the bug go away.
To avoid expensive lookups, we now cache a weak pointer from document to
the last known node navigable. Before using the cache, we validate that
the document is still the navigable's active document.
We were saving to source declarations for *every* property, even though
we only ever looked it up for animation-name.
This patch gets rid of the per-property source pointer and we now keep
a single pointer to the animation-name source only.
This shrinks StyleProperties from 6512 bytes to 4368 bytes per instance.
Navigables are re-used for navigations within the same tab. Its current
ownership of the cursor position is a bit ad-hoc, so nothing in the spec
indicates when to reset the cursor, nor do we manually do so. So when a
cursor update happens on one page, that cursor is retained on the next
page.
Instead, let's have the document own the cursor. Each navigation results
in a new document, thus we don't need to worry about resetting cursors.
This also makes many of the callsites feel nicer. We were previously
often going from the node, to the document, to the navigable, to the
cursor. This patch removes the navigable hop.
We explicitly stopped visting the map of documents to console clients in
commit 44659f2f2a to avoid keeping the
document alive. However, if nothing else visits the console clients, we
may set the top-level console client to a client that has been garbage
collected.
So instead of storing this map, just store the console client on the
document itself. This will allow the document to visit its client.
Previously, pseudo-elements had their style computed while the layout
tree was being built. Instead, do so inside Element::recompute_style(),
using the same invalidation mechanism that the element itself uses.
This also has the effect of invalidating the layout much less often.
Pseudo-elements' style is only computed while building the layout tree.
This meant that previously, they would not have their style recomputed
in some cases. (Such as when :hover is applied to an ancestor.)
Now, when recomputing an element's style, we also return a full
invalidation if one or more pseudo-elements would exist either before or
after style recomputation.
This heuristic produces some false positives, but no false negatives.
Because pseudo-elements' style is computed during layout building, any
computation done here is then thrown away. So this approach minimises
the amount of wasted style computation. Plus it's simple, until we have
data on what approach would be faster.
This fixes the Acid2 nose becoming blue when the .nose div is hovered.
This removes some ambiguity about what the return value should be if
the index is out of range.
Previously, we would sometimes return a JS null, and other times a JS
undefined.
It will also let us fold together the checks for whether an index is a
supported property index, followed by getting the value just afterwards.
These control the state of CSS counters.
Parsing code for `reversed(counter-name)` is implemented, but disabled
for now until we are able to resolve values for those.
This represents each element's set of CSS counters.
https://drafts.csswg.org/css-lists-3/#css-counters-set
Counters are resolved while building the tree. Most elements will not
have any counters to keep track of, so as an optimization, we don't
create a CountersSet object until the element actually needs one.
In order to properly support counters on pseudo-elements, the
CountersSet needs to go somewhere else. However, my experiments with
placing it on the Layout::Node kept hitting a wall. For now, this is
fairly simple at least.
We now ensure that `Node::is_character_data()` returns true for all
nodes of type character data.
Previously, calling `Node::length()` on `CDataSection` or
`ProcessingInstruction` nodes would return an incorrect value.
We were mistakenly executing the current node's script instead of the
document's pending parsing-blocking script.
This caused ~1000 WPT tests to time out, since we never ended up firing
a load event for XHTML pages that load multiple external scripts.
Previously, when creating a HTML element with
`document.createElementNS()` we would convert the given local name to
lowercase before deciding which element type to return. We now no
longer perform this lower case conversion, so if an uppercase local
name is provided, an element of type `HTMLUnknownElement` will be
returned. This aligns our implementation with the specification.
The :host family of pseudo class selectors select the shadow host
element when matching against a rule from within the element's shadow
tree.
This is a bit convoluted due to the fact that the document-level
StyleComputer keeps track of *all* style rules, and not just the
document-level ones.
In the future, we should refactor style storage so that shadow roots
have their own style scope, and we can simplify a lot of this.
Before this change, removing a style element from inside a shadow tree
would cause it to be unregistered with the document-level list of sheets
instead of the shadow-root-level list.
This would eventually lead to a verification failure if someone tried to
update the text contents of that style element, since it was still in
the shadow-root-level list, but now with a null owner element.
Fixes a crash on https://www.swedbank.se/
Previously, if a document had any element with a name attribute that
was set to the empty string, then `document.getElementsByName("")` and
`element.getElementsByName("")` would return a collection including
those elements.
Previously, if a document had an element whose id was the empty string,
then `document.getElementById("")` and `element.getElementById("")`
would return that element.
We had a const and non-const version of this function, with slightly
different behavior (oops!)
This patch consolidates the implementations and keeps only the correct
behavior in there.
Fixes an issue where comments were not collapsible on Hacker News.
GCC 14 emits a warning when an always succeeding `dynamic_cast`'s return
value is compared to NULL inside the `AK::is<T>(U)` template when `T` ==
`U`.
While warning on tautological `is` calls seems useful, it's a bit
awkward when it comes from a function template where the cast may fail
in some instantiation. There is a GCC bug open for it:
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=115664
Work around the warning by performing the algorithm on the base type
(`EventTarget`), with a wrapper that casts it to the more specialized
input type.
