This brings the version of WebRender used in Servo up-to-date with Gecko
upstream. The big change here is that HiDPI is no longer handled via
WebRender. Instead this happens via a scale applied to the root layer in
the compositor. In addition to this change, various changes are made to
Servo to adapt to the new WebRender API.
Co-authored-by: Mukilan Thiyagarajan <mukilan@igalia.com>
Put table cell content fragments into a hieararchy of fragments that
include their table row and table row group fragments. This ensures that
things like relative positioning and transforms set on rows and row
groups properly affect cells and cell content.
Co-authored-by: Oriol Brufau <obrufau@igalia.com>
At the root of an inline formatting context, we used its vertical-align
in order to compute the strut. That was wrong, since vertical-align
on a block container shouldn't affect the contents, it should only
affect the alignment of the block container (if it's inline-level)
within the parent IFC.
This was only working well if the block container was block-level, since
effective_vertical_align_for_inline_layout returned `baseline` for
block-level boxes.
Instead of the outer display type, this patch changes the logic to check
whether we are at the root of the IFC.
The specification gives instructions for how these values should be
propagated. The other big changs here is that they aren't applied to the
`<body>`.
Co-authored-by: Oriol Brufau <obrufau@igalia.com>
Instead of letting Stylo filter `@font-face` rules, handle this
filtering in Servo. It doesn't make sense that Stylo knows about what
fonts Servo supports. This also cleans up a bit the way that this is
handled, giving an entire stylesheet of rules to the font cache to
process instead of letting each layout thread walk the rules. This
brings more of the font-related code into the FontCacheThread itself.
This is the first step toward adding WOFF2 support and fixing various
web font related bugs.
This also ignores a clippy warning for a new function (and a similar
existing one), until this code can be refactored to use temporary Rust
strutures to carry display list building state.
There are a few new test failures here:
- FAIL [expected PASS] /css/css-images/image-set/image-set-conic-gradient-rendering.html
- FAIL [expected PASS] /css/css-images/image-set/image-set-repeating-conic-gradient-rendering.html
These fail because Servo does not yet support `image-set()`.
- FAIL [expected PASS] /css/filter-effects/filter-function/filter-function-conic-gradient.html
- FAIL [expected PASS] /css/filter-effects/filter-function/filter-function-repeating-conic-gradient.html
These fail because Servo does not support the very early filter effects
specification.
- FAIL [expected PASS] /html/canvas/element/manual/fill-and-stroke-styles/conic-gradient-rotation.html
- FAIL [expected PASS] /html/canvas/element/manual/fill-and-stroke-styles/conic-gradient.html
These fail because this change only adds support for CSS conical
gradients. Another set of changes will be necessary to support this for
Canvas.
Anonymous tables should not use legacy pseudos, as the legacy layout
engine had them inherit lots of random properites that lead to bad
layout in the new layout engine.
Co-authored-by: Oriol Brufau <obrufau@igalia.com>
* Add basic support for enumerating computed styles
The old code was so broken, it was enumerating the style attribute
instead of the computed styles.
So this patch implements a basic enumeration, but not including custom
properties whose computed value is not the guaranteed-invalid value.
* Feedback
* Enable `css-text` explicitly in include.ini
The WPT test runner has some strange logic for determining
the key's value for a node like `[css]` or `[css-text]`.
In this logic, if the node doesn't have an explicit value for
a key (here `skip`), then the implicit root node's setting
(i.e key/value pair at the top of the file that is not nested
under a heading) is used as the default fallback value [1]. Only
when the implicit root node doesn't have an explicit value set
does the logic starts looking at the current node's parent [2].
In our case, in `include.ini` the default value for `skip`
is `true` as that is the first line in the file.
Since `[css-text]` doesn't have `skip` set explicitly, the
default value of `true` is used even though the parent's
value is `false`.
[1]: 2bafcf9f18/tests/wpt/tests/tools/wptrunner/wptrunner/wptmanifest/backends/conditional.py (L265)
[2]: 2bafcf9f18/tests/wpt/tests/tools/wptrunner/wptrunner/manifestinclude.py (L59)
Signed-off-by: Mukilan Thiyagarajan <mukilan@igalia.com>
* Update text expectations
---------
Signed-off-by: Mukilan Thiyagarajan <mukilan@igalia.com>
Co-authored-by: Oriol Brufau <obrufau@igalia.com>
The min-content size of a table track was >= the max-content size. So
- For the min-content size of a column, now we just use min-inline-size,
ignoring inline-size and max-inline-size. This matches Gecko, Blink
and WebKit.
- For the max-content size of a column, we keep matching Gecko.
Note that Blink and WebKit are different, they ignore max-inline-size.
- For both the min-content and max-content sizes of a row, now we just
use block-size. This matches Gecko, Blink and WebKit.
Also, if the computed value contains percentages, now we treat it as
the initial value, instead of resolving percentages against zero.
This matches Gecko and Blink, but not WebKit for rows.
* Treat indefinite percentages as auto offsets in relative positioning
Instead of just resolving the percentages against zero.
The spec is not clear (https://github.com/w3c/csswg-drafts/issues/9353),
but this way we match Gecko, Blink and WebKit.
* Update test expectations
* Fix size of tables in flow layout
The contents of a table can make it bigger than what we would expect
from its 'width', 'min-width', 'height' and ' min-height' properties.
Also, 'width: auto' doesn't stretch it to fill the containing block.
We had to refactor the resolution of margins to happen after layout,
otherwise 'auto' margins wouldn't align correctly.
Co-authored-by: Martin Robinson <mrobinson@igalia.com>
* Feedback
* Consistently use `containing_block_for_table` in table layout
* Update test result
---------
Co-authored-by: Martin Robinson <mrobinson@igalia.com>
This change adds a version of row height distribution that follows the
distribtuion algorithm used for tables in Blink's LayoutNG. This is just
an intermediate step toward implementing a distribution algorithm for
both rows and columns more similar to Layout NG.
The CSS Table 3 specification is often wrong with regard to web
compatability, which is why we have abandoned it in favor of the Layout
NG algorithm for row height distribution. this work.
Co-authored-by: Oriol Brufau <obrufau@igalia.com>
Synthetic small caps is supported by the font subsystem, but this is
disabled in Layout 2020. We can turn this on to bring support to parity
with the old layout system.
In addition to turning on synthetic small-caps this change also improves
the way that they work. Before, synthetic small caps meant that every
character was a small version of capitalized character. After this
change, capital letters are larger than small caps versions of small
letters -- matching other browsers and the common expectation of how
small caps works.
Absolutes need to be placed at their hypothetical position as if the
position value was static. This position differs based on the value they
had before blockification. The code for placing absolutes was taking
into account the original display for the inline value, but not for the
block value. A static `display: block` box would placed at a new block
position past the end of the linebox.
This adds basic support for `text-transform` in a way that is more
complete than legacy layout. There are still many missing elements of
proper `text-transform` support such as:
1. Support for `full-width` and `full-size-kana`
2. Support for grapheme based uppercasing, lowercasing, and
capitalization. These are all done per-code point right now.
3. Support for the language-specific `SpecialCasing.txt` cases for case
mapping such as the ones for Irish and Turkish.
Co-authored-by: Rakhi Sharma <atbrakhi@igalia.com>
This deviates from css2, but it's mandated by css-align, and matches
what other browsers do when no margin is 'auto'.
When some margin is 'auto', this should keep the proper round-tripping
behavior that Gecko and WebKit lack, and Blink recently adopted.
This adds support for table rows, columns, rowgroups and colgroups.
There are few additions here:
1. The createion of fragments, which allows script queries and hit
testing to work properly. These fragments are empty as all cells are
still direct descendants of the table fragment.
2. Properly handling size information from tracks and track groups as
well as frustrating rules about reordering rowgroups.
3. Painting a background seemlessly across track groups and groups. This
is a thing that isn't done in legacy layout (nor WebKit)!
Co-authored-by: Oriol Brufau <obrufau@igalia.com>
* Update fontsan to v0.5.0
Updates the C font-sanitizer library `ots` and dependencies
to the latest version.
Signed-off-by: Jonathan Schwender <schwenderjonathan@gmail.com>
* Update test expectations
---------
Signed-off-by: Jonathan Schwender <schwenderjonathan@gmail.com>
Co-authored-by: Martin Robinson <mrobinson@igalia.com>
In layout2020, 'list-style-position' is disabled behind a pref, so the
list_style_position field is an Option.
The serialization of the 'list-style' shorthand wasn't correctly
handling the case of it being None.
This adds an initial implementation of font fallback, on part with the
one used in legacy layout. There are still issues. For instance, font
matching is done per unicode character rather than based on graphemes or
the shape first approach of Chrome. The idea is that these changes can
be made later.
* Respond to shorthand property requests with real values
Signed-off-by: Sebastian C <sebsebmc@gmail.com>
* Cleanup formatting and old comments
Signed-off-by: Sebastian C <sebsebmc@gmail.com>
* Update WPT expectations
Signed-off-by: Sebastian C <sebsebmc@gmail.com>
* Refactor out helper fn
Signed-off-by: Sebastian C <sebsebmc@gmail.com>
* Cleanup
Signed-off-by: Sebastian C <sebsebmc@gmail.com>
---------
Signed-off-by: Sebastian C <sebsebmc@gmail.com>
This implements a very naive row height allocation approach. It has just
enough to implement `vertical-align` in table cells. Rowspanned cells
get enough space for their content, with the extra space necessary being
allocated to the last row. There's still a lot missing here, including
proper distribution of row height to rowspanned cells.
Co-authored-by: Oriol Brufau <obrufau@igalia.com>
For example,
style.cssText = "margin: 1px 2px 3px 4px";
[...style];
will now be
["margin-top", "margin-right", "margin-bottom", "margin-left"]
instead of
["1px", "2px", "3px", "4px"]
Instead of using the border widths from the style, use the ones recorded
by the `BoxFragment`. This is necessary because inline layout can
override these border widths during fragmentation. For instance, when a
box is split across two lines only one fragment should have an inline
start border.
This change starts collecting the starting baseline set for fragments,
which is necessary for some layout modes (flex and tables, namely) as
well as being important for the implementation of `align-items`. In
addition, it converts baseline measurement to use `Au` everywhere.
Co-authored-by: Oriol Brufau <obrufau@igalia.com>
