Finishes what 7f6b70fafb started.
Having one part use length and another code unit length lead to crashes,
the added test ensures we don't mess that up again.
Fixes race condition introduced in eed47acb when rendering thread
accesses ScrollFrame that could be mutated in the middle of
rasterization by the main thread, resulting in broken rendering.
Fixes https://github.com/LadybirdBrowser/ladybird/issues/5553
This adds a new IDL type, Utf16DOMString. This is the same as DOMString,
except it is UTF-16. This type is temporary - we will want DOMString to
be UTF-16 by default once we've ported enough of LibWeb.
To make this support easier, some string IDL generator handling is moved
directly into `generate_to_string` from the call sites.
Skia allows you to pass a bounding rect to its saveLayer() function as
an optimization when you know that you won't paint outside those bounds.
Unfortunately, we were passing a too-small rectangle that didn't take
into account transformed descendants, etc.
For now, simply pass null instead of a bounding rect. This way, Skia
figures it out internally. It may allocate larger temporary bitmaps than
needed this way, but at least we get more correct results. I've left
re-enabling the optimization as a FIXME in the code.
This fixes unwanted clipping in various parts of the Discord UI.
Both sides of the Editing internals now have to deal with some awkward
converting between UTF-8 and UTF-16, but the upside is that it
immediately exposed an issue with the `insertText` command: instead of
dealing with code units, it was iterating over code points causing the
selection to be updated only once instead of twice. This resulted in the
final selection potentially ending up in between a surrogate pair.
Fixes#5547 (pasting/typing surrogate pairs).
Originally I added this to use it in Utf16View::ends_with(), but the
final implementation ended up a lot simpler. I chose to keep this anyway
since it mirrors Span::starts_with().
This prevents empty matches from overwriting non-empty captures in
quantified alternations. Fixes patterns like (a|a?)+ where the optional
branch would incorrectly overwrite meaningful captures with empty
strings.
This will build both aarch64 and x86_64 flatpaks nightly. The test jobs
I did with the default GitHub runners took ~1h30m for aarch64 and
~2h30m for x86_64, but this will hopefully improve over time as the
cache is built up. And be better time-wise on the blacksmith runners.
Disk space is a large concern with these flatpak builds. A naive script
on my local machine showed that we needed a max of 10.3 GiB of free
space for the x86_64 build, with clang as the compiler. For some reason,
building with default gcc 14 uses more disk space than that, bumping
up against the default 14 GiB free space limit guaranteed by GitHub for
their default runners.
`AnimationTimeline` visits pointers of all registered animations, so if
element is removed from DOM tree but its animations remain registered in
timeline, then `Animation` and owner `Element` will be kept alive until
`AnimationTimeline` is destroyed.
flatpak-builder doesn't respect .gitignore when creating its local build
directory, so we need to explicitly skip potentially large ignored
directories to avoid bloating the flatpak build directory during builds.
This behavior is part of the cyclic percentage contribution logic from
CSS-SIZING-3 which explicitly only applies to non-replaced boxes.
This fixes an issue on Discord where buttons in the settings UI were
cropped to a narrower width than intended.
Fixes#3572
This uses a `foo>bar` notation in the `valid-identifiers` field of
Properties.json, to say "replace `foo` with `bar`".
The motivation here is to avoid calling `parse_css_value_for_property()`
inside the per-property switch in `parse_css_value()`. Eventually we'll
need to be able to call that switch from
`parse_css_value_for_properties()` so that shorthands can make use of
any bespoke parsing code to parse their longhands.
