This allows simplifying the code somewhat and means we can ignore
unanimated shorthands a little sooner. Furthermore, it removes the odd
inconsistency where TransitionProperty only included animatable
longhands but allowed all shorthands regardless of whether or not they
were animatable.
stylo: Bug 1335998 - Handle interpolation and accumulation of mismatched transform lists
These are the interdependent patches of Bug 1335998. We want to do interpolation and accumulation for mismatched transform lists, so introduce ComputedOperation::InterpolateMatrix and ComputedOperation::Accumulation. Both arms store the from_list and to_list, and resolve them until we have the layout information. For the Servo part, we haven't implemented how to read the transform lists in layout/fragment.rs, but I think it would be easy. (related issue #13267)
---
- [X] `./mach build -d` does not report any errors
- [X] `./mach test-tidy` does not report any errors
- [X] These changes fix [Bug 1335998](https://bugzilla.mozilla.org/show_bug.cgi?id=1335998)
- [X] There are tests for these changes
<!-- Reviewable:start -->
---
This change is [<img src="https://reviewable.io/review_button.svg" height="34" align="absmiddle" alt="Reviewable"/>](https://reviewable.io/reviews/servo/servo/17202)
<!-- Reviewable:end -->
If we set the composite operation to accumulate, we may need to accumulate two
mismatched transform lists, and then to interpolate them. In order to
accumulate two mismatched transform lists, we also need the reference
box information, so use the same idea as that for interpolation. i.e.
use AccmulateMatrix to store it temporarily, and convert it into matrix
later.
We use this arm to store the interpolated result of two mismatched
transform lists, and we resolve it until we know the reference box size
(on Gecko side). The conversion from ComputedOperation::InterpolateMatrix
to eCSSKeyword_interpolatematrix will be implemented later in this patch series.
Without this patch anim-css-strokewidth-1-by-pct-pct.svg (and possibly
others) fails because we calculate the result as 'calc(0px + 10%)' and
we don't support calc on stroke-width (yet) so the rendered result is
incorrect.
As a more thorough fix, we should make the zero-value for
LengthOrPercentageOrNumber a zero *number* (instead of a zero length)
but that won't work yet since we don't support animating between
stroke-widths with units and those that don't (see
https://bugzilla.mozilla.org/show_bug.cgi?id=1369614). Regardless of
that, we still shouldn't introduce calc in order to add a zero value
using the LengthOrPercentage type, so this change is still needed.
Although there are no specific tests for this yet if any code does try
to perform addition or accumulation with a font-stretch value it will
likely fail the bounds check in the Into trait. To avoid that this
patch provides an basic implementation of add_weighted for font-stretch
that should work correctly with portions whose sum is outside the [0.0,
1.0] range.
A specified position is now a struct made of two values of different types,
the first one being PositionComponent<X>, and the second one PositionComponent<Y>.
A position component is represented by the new enum PositionComponent<Side>,
with the three values Center, Length(LengthOrPercentage), and
Side(Side, Option<LengthOrPercentage>).
Side keywords are represented by the X and Y enums, which don't include a value
for the center keyword anymore. They are accompanied by the Side trait, which
allows us to determine whether a side keyword is "left" or "top".
This refactor simplified the parsing and serialisation code and exposed bugs in it,
where it would reject valid <position> values followed by arbitrary tokens,
and where it would fail to prefer "left" to "right" when serialising positions
in basic shapes.
This part includes making stroke-linecap, stroke-linejoin, stroke-miterlimit,
stroke-opacity, and stroke-dasharray animatable.
For properties that already implemented Interpolate trait and clone() for
glue code, we can just make them animatable by replacing the animation_value_type
with proper type name. So, set animation_value_type to 'discrete' for
stroke-linecap and stroke-linejoin. Set animation_value_type to 'ComputedValue'
for stroke-miterlimit and stroke-opacity.
As to stroke-dasharray, we need to implement Interpolate trait and glue codes
for it.
Move impl's into macro for macro-generated keyword types
Three keyword types are created through a macro but have some of their
impl's handled elsewhere. Since all impl's are the same, this commit
moves them into the macro to have them auto generated, for more concise
code.
<!-- Please describe your changes on the following line: -->
---
<!-- Thank you for contributing to Servo! Please replace each `[ ]` by `[X]` when the step is complete, and replace `__` with appropriate data: -->
- [X] `./mach build -d` does not report any errors
- [X] `./mach test-tidy` does not report any errors
- [X] These changes fix#16604 (github issue number if applicable).
<!-- Either: -->
- [ ] There are tests for these changes OR
- [X] These changes do not require tests because this doesn't add any new features, just alters `impl` locations
<!-- Also, please make sure that "Allow edits from maintainers" checkbox is checked, so that we can help you if you get stuck somewhere along the way.-->
<!-- Pull requests that do not address these steps are welcome, but they will require additional verification as part of the review process. -->
<!-- Reviewable:start -->
---
This change is [<img src="https://reviewable.io/review_button.svg" height="34" align="absmiddle" alt="Reviewable"/>](https://reviewable.io/reviews/servo/servo/16623)
<!-- Reviewable:end -->
Three keyword types are created through a macro but have some of their
impl's handled elsewhere. Since all impl's are the same, this commit
moves them into the macro to have them auto generated, for more concise
code.
1. We add a new arm to TransitionProperty, TransitionProperty::Unsupported,
which contains an Atom, so it's better to remove the Copy trait from
TransitionProperty.
2. TransitionProperty::Unsupported(Atom) represents any non-animatable, custom,
or unrecognized property, and we use Atom to store the ident string for
serialization.
