This patch introduces the StyleComputer::RuleCache, which divides all of
our (author) CSS rules into buckets.
Currently, there are two buckets:
- Rules where a specific class must be present.
- All other rules.
This allows us to check a significantly smaller set of rules for each
element, since we can skip over any rule that requires a class attribute
not present on the element.
This takes the typical numer of rules tested per element on Discord from
~16000 to ~550. :^)
We can definitely improve the cache invalidation. It currently happens
too often due to media queries. And we also need to make sure we
invalidate when mutating style through CSSOM APIs.
Previously we would re-run the entire CSS selector machinery for each
property resolved. Instead of doing that, we now resolve a final set of
custom property key/value pairs at the start of the cascade.
- Replace "auto" with "auto const" where appropriate.
- Remove an unused struct.
- Make sort_matching_rules() a file-local static function.
- Remove some unnecessary includes.
This isn't perfect (especially the global object situation in
activate_event_handler), but I believe it's in a much more complete
state now :^)
This fixes the issue of crashing in prepare_for_ordinary_call with the
`i < m_size` crash, as it now uses the IDL callback functions which
requires the Environment Settings Object. The environment settings
object for the callback is fetched at the time the callback is created,
for example, WrapperGenerator gets the incumbent settings object for
the callback at the time of wrapping. This allows us to remove passing
in ScriptExecutionContext into EventTarget's constructor.
With this, we can now drop ScriptExecutionContext.
We also pass whether the shadow goes inside or outside the element. Only
outer shadows are rendered currently, and inner ones may want to be
handled separately from them, as they will never interfere with each
other.
The pattern we've adopted for other multi-value properties is to run in
a loop like this, since that makes it easier to cater for values
appearing in different orders.
During the LengthPercentage split, I converted the individual-corner
`border-foo-bar-radius` properties to LengthPercentage but forgot
`border-radius` itself! Oops. Discord's CSS was doing `border-radius:
50%` a lot, so this cuts down on CSS parser spam.
Browser has a handy debug menu option to dump all stylesheets, so we
don't need to spam the console with this. (All the spam massively slows
down page loads.)
Style updates are lazy since late last year, so the StyleInvalidator is
actually hurting us more than it's helping by running the entire CSS
selector machine on the whole DOM for every attribute change.
Instead, simply mark the entire DOM dirty and let the lazy style update
mechanism run *once* on next event loop iteration.
Modified the test-page because FontDatabase looks for exact font-weight
matches, so requesting weight 800 in a font that only has 700, causes
it to return the default font instead. So, we ask for 700 here.
The actual fix is to improve our font-matching but I am trying not to
get distracted today. :^)
None of these require any outside metrics, which is nice! I believe the
Values-4 spec would have us simplify them down into a single value at
parse time, but that's a yak for another day.
Most of the time, we cannot resolve a `calc()` expression until we go to
use it. Since any `<length-percentage>` can legally be a `calc
()`, let's store it in `LengthPercentage` rather than make every single
user care about this distinction.
The previous static functions are now methods of their respective
CalcFoo structs, but the logic has not changed, only that they work
with CalculationResults instead of converting everything to floats.
calc() sub-expressions can return a variety of different types, which
then can be combined using the basic arithmetic operators. This class
should make that easier to deal with, instead of having to handle all
the possible combinations at each call site. :^)
We take the Layout::Node as a pointer not a reference, since later we'll
need to call these functions when resolving to `<number>` or `<integer>`
which don't use those, and we don't want to force users to pass them in
unnecessarily.
See https://www.w3.org/TR/css-values-3/#calc-type-checking
If the sub-expressions' types are incompatible, we discard the calc() as
invalid.
Had to do some minor rearranging/renaming of the Calc structs to make
the `resolve_foo_type()` templates work too.
This lets us produce valid CSS in its to_string() method, instead of
always adding commas as before. :^)
Also, finally added a Formatter for StyleValues.
Many of these will need to change in the future in order to include
features we don't yet support, and touching StyleValue.h is a great way
to have to wait for all of LibWeb to rebuild. I'm hoping this saves me
time in the long run. :^)
Raw whitespace is not allowed inside a name, but escaped whitespace is,
for example `\9`, which is the tab character.
This stops yakzz.com from crashing the Browser, since it was using `\9`
in various places as a hack to only apply those properties to IE8/9.
CSS has rules about automatic blockification or inlinification of boxes
in certain circumstances.
This patch implements automatic blockification of absolutely positioned
and floating elements. This makes the smile appear on ACID2. :^)
This moves LibWeb to using the list of hidden elements from the spec.
Concretely, the following things are now explicitly marked
`display: none` in addition to before:
- elements with the `hidden` attribute
- area
- base
- basefont
- datalist
- param
- rp
The spec also wants `noframes` and `noembed` to be `display: none`,
but since support for frames and embeds doesn't exist yet, these
are omitted for now.
With this, everyone's favorite website http://45.33.8.238/ no longer
displays spans with attribute hidden. (Whitespace handling still
looks a bit off though.)
Despite looking like it was still needed, it was only used for passing
to other calls to Length::resolved() recursively. This makes the
various `foo.resolved().resolved()` calls a lot less awkward.
(Though, still quite awkward.)
I think we'd need to separate calculated lengths out to properly tidy
these calls up, but one yak at a time. :^)
