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>
* use app unit in replaced elements
* more app unit usage
* Avoid unnecessary into()
* Run ./mach fmt
* use scaleby
* update
---------
Co-authored-by: Oriol Brufau <obrufau@igalia.com>
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. -->
Before counting whitepsace-only `GlyphStore`s where counted as a single
justification opportunity when trimming whitespace from the front and
back of lines. This isn't correct, instead count the actual number of
word seperators of the trimmed `GlyphStore`s.
These two counts can be different in the case where whitespace collapse
isn't happening yet (flexbox). In addition, using word seperators means
the code is making less assumptions about the contents of the line and
is more robust.
This fixes some crashes in flexbox tests on debug builds.
Co-authored-by: Rakhi Sharma <atbrakhi@igalia.com>
Instead of a tricky stack of enum iterators expose a `foreach()`
function on InlineFormattingContext, which takes a `FnMut`. This
prevents callers wanting to iterate from keeping a stack of iterators
and will potentially allow a future version of this function to avoid
borrowing the ArcRefCell<...> of inline boxes for every iteration
(presumably using something like OwnedRef).
Convert `inline_content_sizes` to use this new `foreach()` function and
move the `Computation` out of the function body to
`ContentSizesComputation`. This reduces the stack depth of inline size
computation, because `foreach()` is iterative and not recursive.
This is a preliminary change to removing the second round of text shaping
during layout, because shaping will use this new iterator.
This is just a bit of code movement that trims down the size of the
`inline.rs` file in order to make it a bit more manageable. It leads the
way to more refactoring and cleanup in the future.
* layout: Implement support for `line-height` and `vertical-align`
This is an initial implementation of proper `line-height` and
`vertical-align` support. While this change includes the bulk of the
work there are still many missing pieces for full support. In particular
some big missing things are:
- Flex containers do not properly compute their baselines. The idea is
to tackle this in a followup change. This causes various flex tests
to start failing because everything used to be top aligned.
- The implementation of the line-height quirks (only active in quirks
mode) are incomplete. While the quirk works in many cases, there are
still some cases where it is handled incorrectly. This requires more
redesign and refinement, better suited for a followup.
- Most of the features are CSS 3 such as precision control of the
baseline and first and last baselines are not implemented. This
change gets us close to CSS 2.x support.
While there are many new test passes with this change some tests are
starting to fail. An accounting of new failures:
Tests failing also in Layout 2013:
- /css/css2/positioning/toogle-abspos-on-relpos-inline-child.html (only passes in Chrome)
- /css/CSS2/fonts/font-applies-to-001.xht (potentially an issue with font size)
Invalid tests:
- /css/CSS2/visudet/inline-block-baseline-003.xht
- /css/CSS2/visudet/inline-block-baseline-004.xht
- These are are failing in all browsers. See https://bugs.chromium.org/p/chromium/issues/detail?id=1222151.
Missing table support:
- /_mozilla/mozilla/table_valign_middle.html
Missing `font-size-adjust` support :
- /css/css-fonts/font-size-adjust-zero-2.html (also failing in 2013)
Incomplete form field support :
- /html/rendering/widgets/the-select-element/option-add-label-quirks.html (label isn't rendered so button isn't the right size in quirks mode due to line height quirk)
Need support for calculating flexbox baseline:
- /css/css-flexbox/fieldset-baseline-alignment.html
- /css/css-flexbox/flex-inline.html
- /css/css-flexbox/flexbox-baseline-multi-line-horiz-001.html
- /css/css-flexbox/flexbox-baseline-single-item-001a.html
- /css/css-flexbox/flexbox-baseline-single-item-001b.html
Failing because we don't create anonymous inline boxes for text children of blocks:
- /css/CSS2/linebox/anonymous-inline-inherit-001.html
Passes locally (potentially related to fonts):
- /css/CSS2/css1/c414-flt-fit-004.xht
- /css/css-transforms/transform-input-017.html
- /html/obsolete/requirements-for-implementations/the-marquee-element-0/marquee-min-intrinsic-size.html
- /css/css-fonts/first-available-font-005.html
- /css/css-fonts/first-available-font-006.html
* Some cleanups after live review with @mukilan
Also update results.
This change adds support for `text-align-last` as well as ensuring that
it also applies to lines before forced line breaks. Two tests start to
fail because they rely on right-to-left text to pass:
- /css/css-text/text-align/text-align-last-010.html.ini
- /css/css-text/text-align/text-align-last-011.html.ini
This data structure has all of the metrics needed to render a font and
is in `Au`. We'll need more of these metrics for implementing
`vertical-align` and its use doesn't increase the size of the Fragment
tree (as the BoxFragment is still larger). In addition, this will be
helpful when switching layout to `Au`.
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>
The changes in #30740, fixed an issue where certain characters should
prevent line break opportunity after atomics. This change extends that
to also apply to before atomics, which is what the specification says
should happen.
We previously sent a " " to the linebreaker in order to ensure that the
next text had a soft wrap opportunity at the start. Calling `next(" ")`
without waiting until the returned index was 1, violated some
invariants of linebreaker ultimately causing a panic.
Instead of using the linebreaker for this, simply keep a flag in the
IFC layout state, which avoids the problem entirely.
Fixes#30703.
Earlier versions of inline layout in the new layout system did not
properly support line breaking when unbreakable segments spanned
multiple inline boxes. This change updates inline layout to add support
for that. Now items are added to an unbreakable segment before being
committed to a line.
Co-authored-by: Mukilan Thiyagarajan <mukilan@igalia.com>
Co-authored-by: Oriol Brufau <obrufau@igalia.com>
Flattening the LineItem tree into a token stream will allow for handling
the case where an unbreakable line segment spans multiple inline boxes
which might have different hierarchies. This change also fixes the
handling of the second anonymous fragment of a block-in-inline-split.
Co-authored-by: Oriol Brufau <obrufau@igalia.com>
Co-authored-by: Mukilan Thiyagarajan <mukilan@igalia.com>
After placing a float, FloatBox's layout_into_line_items() was calling
place_line_among_floats() with ifc.current_line.inline_position as the
width of needed by the contents of the line.
The problem is that this amount includes the trailing whitespace advance
and thus it could seem that the in-flow contents wouldn't fit next to
the float.
That's not the case, since collapsible whitespace at the end of the line
is removed, and preserved whitespace hangs.
So this patch subtracts ifc.current_line.trailing_whitespace_advance
when calling place_line_among_floats(), like it was already happening
when computing the available_inline_size.
Fixes#30561
Improve the calculation of the block size of line boxes and all their
component elements. Even empty spans can increase the size of the line
based on their font-size. Elements that have a line-height should
increase the block size of the line, but that setting should not effect
their own size.
In addition to the new passes there are some new failures
Failing because a progression exposes the real issue these tests are
testing:
- css/css-color/t32-opacity-offscreen-multiple-boxes-1-c.xht
- css/css-color/t32-opacity-offscreen-multiple-boxes-2-c.xht
Likely failing because of vertical-align and another sizing issue:
- css/css-transforms/perspective-untransformable-no-stacking-context.html
Failing because a progression reveals another failure:
- html/rendering/non-replaced-elements/hidden-elements.html
Co-authored-by: Mukilan Thiyagarajan <mukilan@igalia.com>
* Take forced line breaks into account for intrinsic size
Fixes#30350.
* Don't linebreak on collapsible whitespace
This whitespace can hang off the end of the line, because it will be
trimmed LineItem layout.
* Update float placement after line breakage
Also don't ever line break for collapsible whitespace.
* Fix a few more test cases and clean up
* Renaming according to review comments
This makes the names of flow relative geometry consistent with what is
used in the style crate and removes them from a module. With this change
it's more obvious what makes these types different from the ones in
`euclid`.
This makes the nesting architecture of inline layout a lot simpler. Now
PartialInlineBoxFragment and InlineNestingLevelState are replaced by
two structs:
- InlineContainerState
- InlineBoxContainerState
InlineContainerState holds state for the root of the inline formatting
context, while InlineBoxContainerState holds state for inline boxes.
InlineBoxContainerState has an InlineContainerState as the first
element, thus "extends" it.
Now the inline iterators are stack variables directly in the `layout()`
method. They are no longer stored in the state. This avoids the weird
nesting of state and instead relies on a normal vector to hold the
stack of state.
Co-authored-by: Mukilan Thiyagarajan <mukilan@igalia.com>
This implements the rest of the bulk of float support. Now inline
element flow around floats and floats can be pushed down by inline
elements before them.
Co-authored-by: Oriol Brufau <obrufau@igalia.com>
When layout was split into two phases, floats were laid out as direct
children of the inline formatting context. This meant that they were
positioned properly, but not properly made children of their inline
ancestors' stacking contexts. This change maintains the proper
positioning of floats, but positions them relatively to their inline
ancestors.
The big change here is that `text-align` needs to be taken into account
before actually laying out LineItems. This has the added benefit of
setting inline layout for the implementation of `text-align: full`. Now
all line items are laid out at the real final position and we can adjust
the `start_corner` property of float `BoxFragments` when their ancestors
are laid out.
In the first phase, we gather LineItems and then when we have enough to
form a line we turn them into Fragments. This will make it possible to
more simply implement `vertical-align` and `text-align: justify` because
we need to measure the different aspects of the candidate line and then
produce a Fragments.
This is a general refactor of the way that inline layout works, so comes
with some progressions. In addition there are some new failures.
New failures:
Some tests are now failing because only the test or reference is getting
proper line height when it wasn't before. These should be fixed in a
followup change that properly calculate line-height in more cases:
- /_mozilla/css/list_style_position_a.html
- /css/CSS2/floats/float-no-content-beside-001.html
- /css/css-content/pseudo-element-inline-box.html
- /css/css-flexbox/flexbox_flex-none-wrappable-content.html
Some tests are now failing because floats are now placed properly, but
are no longer in their inline box stacking contexts. These will be fixed
by a followup change which properly parents them:
- /css/filter-effects/filtered-inline-applies-to-float.html.ini
- /css/css-color/inline-opacity-float-child.html.ini
One test is failing due to floating point precision errors:
- /css/CSS2/floats-clear/floats-141.xht.ini
Signed-off-by: Martin Robinson <mrobinson@igalia.com>
Co-authored-by: Mukilan Thiyagarajan <mukilan@igalia.com>
When descending and we have the option, don't create new
PositioningContexts just to update the static position of laid out
abspos descendants. Instead, use the new PositioningContextLength to
only update the newly added hoisted abspos boxes.
* 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>
Remove rayon_croissant and refactor the way that information about
floats in flows bubbles up. This simplifies the code a good deal and
lets us take advantage of some more optimized functions provided by
rayon. This removes 2 crates from the dependency tree.
In addition, this allows avoiding passing `contains_floats` up from
every box tree construction function. This makes things simpler, but
also opens up the possibility of passing more of these flags up in the
future (such as `contains_counters`).
layout_2020: Do linebreak for atomic inline-level elements
<!-- 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#29591 (GitHub issue number if applicable)
<!-- Either: -->
- [x] There are tests for these changes OR
<!-- 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. -->
Clearance was implemented as a Length, where zero meant no clearance.
However, having a clearance of 0px should be different than having
no clearance, since the former can still prevent margin collapse.
This patch keeps the existing behavior, so it won't be possible to get
a clearance of Some(Length::zero()), but it prepares the terrain for
a follow-up to fix calculate_clearance to return the proper thing.
Margins should be able to collapse through floats when collapsing with
parent blocks (the containing block). To properly place floats in this
situation, we need to look at these subsequent floats to find out how
much of the margin will collapse with the parent.
This initial implementation is very basic and the second step would be
to cache this in order to avoid having to constantly recalculate it.
Fixes#29915.
Co-authored-by: Oriol Brufau <obrufau@igalia.com>
Absolutes with static insets need to be laid out at their ancestor
containing blocks, but their position is dependent on their parent's
layout. The static layout position is passed up the tree during hoisting
and ancestors each add their own offset to the position until it is
relative to the containing block that contains the absolute.
This is currently done with a closure and a fairly tricky "tree rank"
numbering system that needs to be threaded through the entire layout.
This change replaces that system.
Every time a child is laid out we create a positioning context to hold
any absolute children (this can be optimized away at a later time). At
each of these moments, we call a method to aggregate offsets to the
static insets of hoisted absolutes. This makes the logic easier to
follow and will also allow implementing this behavior for inline-blocks,
which was impossible with the old system.
