With this change we no longer stretch "width: auto" for replaced
elements and also use "width calculation rules for block-level replaced
elements", like suggested by the spec.
If a calculation was simplified down to a single numeric node, then most
of the time we can instead return a regular StyleValue, for example
`calc(2px + 3px)` would be simplified down to a `5px` LengthStyleValue.
This means that parse_calculated_value() can't return a
CalculatedStyleValue directly, and its callers all have to handle
non-calculated values as well as calculated ones.
This simplification is reflected in the new test results. Serialization
is not yet correct in all cases but we're closer than we were. :^)
This allows us to disable test output, which performs expensive assert
tracking. This was making our imported tests run significantly slower
than tests run via `WPT.sh`.
Formatting the output ourselves also allows us to remove unnecessary
information from the test output.
This commit also rebaselines all existing imported WPT tests to follow
the new format.
This wins us 65 new WPT subtest passes! It also shows up that we're
doing the wrong thing in ShorthandStyleValue in places, notably with
the grid properties. However, having one place to fix instead of two
will make it easier to correct them. :^)
In order to be fully correct, we should use the algorithm here:
https://drafts.csswg.org/cssom/#serialize-a-css-value
However, it's quite hand-wavy. What we do have in the meantime is
`ShorthandStyleValue::to_string()`, where we special-case the
serialization rules for shorthands with a generic fallback that's
equivalent to what the previous `get_property_value()` code was doing.
This causes 36 new subtests to pass locally. :^)
Unfortunately at least one of these is flaky when it's able to load the
font file, apparently because we don't wait for the font and its
stylesheet to actually load before the tests run.