In order to support size keywords in block layout, we may need to call
`inline_content_sizes()` in order to compute the min/max-content sizes.
But this required a mutable reference in order the update the cache,
and in various places we already had mutable references.
So this switches the cache into a RwLock to avoid needing mutable refs.
Note OnceCell wouldn't work because it's not thread-safe.
Signed-off-by: Oriol Brufau <obrufau@igalia.com>
This is the second flexbox caching change. It seeks to detect when a
relayout can be avoided in the case of a stretching flex item. This
heuristic can be combined, because currently we still do relayout
sometimes when we do not need to.
For instance currently we always relayout when a flex child is itself a
column flex. This only needs to happen when the grandchildren themselves
grow or shrink. That optimization is perhaps a lower priority as
`flex-grow: 0 / flex-shrink: 1` is the default behavior for flex.
Since this change means we more consistenly zero out the percentage part
of `calc` expressions when they have circular dependencies, this causes one
test to start failing (`/css/css-values/calc-min-height-block-1.html`).
This is related to w3c/csswg-drafts#10969, which is pending on further
discussion in the working group.
Signed-off-by: Martin Robinson <mrobinson@igalia.com>
Co-authored-by: Oriol Brufau <obrufau@igalia.com>
This is the first part of caching intermediary layout during flexbox
layout. A later change will try to reuse these layouts, when possible,
for actual item layout and re-layout due to stretching.
Signed-off-by: Martin Robinson <mrobinson@igalia.com>
Co-authored-by: Oriol Brufau <obrufau@igalia.com>
The result of `inline_content_sizes()` may depend on the block size of
the containing block, so we were always recomputing in case we got
a different block size.
However, if no content has a vertical percentage or stretches vertically,
then we don't need to recompute: the result will be the same anyways.
Signed-off-by: Oriol Brufau <obrufau@igalia.com>
Co-authored-by: Martin Robinson <mrobinson@igalia.com>
This adds supports for right-to-left text assigning bidi levels to all
line items when necessary. This includes support for the `dir` attribute
as well as corresponding CSS properties like `unicode-bidi`. It only
implements right-to-left rendering for inline layout at the moment and
doesn't include support for `dir=auto`. Because of missing features,
this causes quite a few tests to start failing, as references become
incorrect due to right-to-left rendering being active in some cases,
but not others (before it didn't exist at all).
Analysis of most of the new failures:
```
- /css/css-flexbox/gap-001-rtl.html
/css/css-flexbox/gap-004-rtl.html
- Require implementing BiDi in Flexbox, because the start and
end inline margins are opposite the order of items.
- /css/CSS2/bidi-text/direction-applies-to-*.xht
/css/CSS2/bidi-text/direction-applies-to-002.xht
/css/CSS2/bidi-text/direction-applies-to-003.xht
/css/CSS2/bidi-text/direction-applies-to-004.xht
- Broken due to a bug in tables, not allocating the
right amount of width for a column.
- /css/css-lists/inline-list.html
- This fails because we wrongly insert a soft wrap opportunity between the
start of an inline box and its first content.
- /css/css-text/bidi/bidi-lines-001.html
/css/css-text/bidi/bidi-lines-002.html
/css/CSS2/text/bidi-flag-emoji.html
- We do not fully support unicode-bidi: plaintext
- /css/css-text/text-align/text-align-end-010.html
/css/css-text/text-align/text-align-justify-006.html
/css/css-text/text-align/text-align-start-010.html
/html/dom/elements/global-attributes/*
- We do not support dir=auto yet.
- /css/css-text/white-space/tab-bidi-001.html
- Servo doesn't support tab stops
- /css/CSS2/positioning/abspos-block-level-001.html
/css/css-text/word-break/word-break-normal-ar-000.html
- Do not yet support RTL layout in block
- /css/css-text/white-space/pre-wrap-018.html
- Even in RTL contexts, spaces at the end of the line must hang and
not be reordered
- /css/css-text/white-space/trailing-space-and-text-alignment-rtl-002.html
- We are letting spaces hang with white-space: pre, but they shouldn't
hang.
```
Signed-off-by: Martin Robinson <mrobinson@igalia.com>
Co-authored-by: Rakhi Sharma <atbrakhi@igalia.com>
This change adds an expensive layout for the determination of minimum
automatic size and flex basis in process of flexbox layout. Currently,
the layout is not cached, so may be performed up to 2 more times than
necessary.
Signed-off-by: Martin Robinson <mrobinson@igalia.com>
Co-authored-by: Oriol Brufau <obrufau@igalia.com>
This only paints text in input fields. Selection and cursor are still
not painted.
In addition to adding this feature, the change also updates the
user-agent.css with the latest from the HTML specification. Extra
padding and extraneous settings (such as a bogus line-height and
min-height) are also removed from servo.css. This leads to some new
passes.
There are some new passes, this introduces failures as inserting text
reveals issues that were hidden before. Notably:
- failures in `/html/editing/editing-0/spelling-and-grammar-checking/`:
We do not support spell-checking.
- Most of the rest of the new failures are missing features of input
boxes that are also missing in legacy layout.
The main change here is that collapsed and `text-transform`'d text is
computed as it's processed by DOM traversal. This single transformed
text is stored in the root of the `InlineFormattingContext`.
This will eventually allow performing linebreaking and shaping of the
entire inline formatting context at once. Allowing for intelligent
processing of linebreaking and also shaping across elements. This
matches more closely what LayoutNG does.
This shouldn't have any (or negligable) behavioral changes, but will
allow us to prevent linebreaking inside of clusters in a followup
change.
Co-authored-by: Rakhi Sharma <atbrakhi@igalia.com>
This doesn't really have observable behavior right now, as much as I
tried to trigger some kind of bug. On the other hand, it's just wrong
and is very obvious when you dump the Fragment tree. If you create a
`display: table-cell` that is a child of the `<body>` all parts of the
anonymous table are flagged as if they are the `<body>` element.
This moves white space collapse to right before breaking and shaping
happens, which is more similar to what happens in legacy layout. This is
the first step toward making this procedure more efficient (avoiding
string copies) and also implementing support for `text-transform`.
Co-authored-by: Rakhi Sharma <atbrakhi@igalia.com>
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.
Shape text during InlineFormattingContext construction rather than doing
it twice during fragment tree construction. This is a step on the way
toward proper font fallback.
This also moves all `TextRun` related code into `text_run.rs` to try to
trim down the size of `inline.rs`.
<!-- 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 do not require tests because this should only have
performance impacts.
<!-- 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. -->
This is the first part of constructing the box tree for table layout. No
layout is actually done and the construction of tables is now hidden
behind a flag (in order to not regress WPT). Notably, this does not
handle anonymous table part construction, when the DOM does not reflect
a fully-formed table. That's part two.
Progress toward #27459.
Co-authored-by: Oriol Brufau <obrufau@igalia.com>
Co-authored-by: Manish Goregaokar <manishsmail@gmail.com>
In legacy layout, anonymous text wrappers were inheriting the `overflow`
and `text-overflow` properties. This results in the creation of extra
clipping for these anonymous wrappers which could clip away floats. We
will likely implement `text-overflow` differently in non-legacy layout.
This change marks all legacy layout pseudo elements as "legacy" and also
adds a new pseudo element for non-legacy layout that does not inherit
`overflow`.
Fixes#30562.
Co-authored-by: Oriol Brufau <obrufau@igalia.com>
* Add initial support for css-text-3 whitespace handling
This adds initial support for whitespace handling from the CSS
specification for Layout 2020. In general, the basics are covered. Since
test output is very sensitive to whitespace handling, this change
incorporates several fixes:
1. Whitespace is collapsed according to the Phase 1 rules of the
specification, though language-specific unbreaking rules are not
handled properly yet.
2. Whitespace is mostly trimmed and positioned according to the Phase 2
rules, but full support for removing whitespace at the end of lines
is pending on a temporary data structure to hold lines under
construction.
3. Completely empty box fragments left over immediately after line
breaks are now trimmed from the fragment tree.
4. This change tries to detect when an inline formatting context
collapses through.
Fixes#29994.
Co-authored-by: Mukilan Thiyagarajan <me@mukilan.in>
Co-authored-by: Martin Robinson <mrobinson@igalia.com>
Signed-off-by: Martin Robinson <mrobinson@igalia.com>
* Update test results
---------
Signed-off-by: Martin Robinson <mrobinson@igalia.com>
Co-authored-by: Mukilan Thiyagarajan <me@mukilan.in>
Co-authored-by: Martin Robinson <mrobinson@igalia.com>
During layout it is often useful, for various specification reasons, to
know if an element is the `<body>` element of an `<html>` element root. There
are a couple places where a brittle heuristic is used to detect `<body>`
elements. This information is going to be even more important to
properly handle `<html>` elements that inherit their overflow property from
their `<body>` children.
Implementing this properly requires updating the DOM wrapper interface.
This check does reach up to the parent of thread-safe nodes, but this is
essentially the same kind of operation that `parent_style()` does, so is
ostensibly safe.
This change should not change any behavior and is just a preparation
step for properly handle `<body>` overflow.
