mirror of
https://github.com/servo/servo.git
synced 2025-06-15 03:44:30 +00:00
362b64aa68
context, to place floats in layout 2020. The containing block for a float is not necessarily the same as the block formatting context the float is in per CSS 2.1 [1]: "For other elements, if the element’s position is relative or static, the containing block is formed by the content edge of the nearest block container ancestor box." This shows up in the simplest case: <html> <body> <div style="float: left">Hello</div> </body> </html> In this case, the `<html>` element is the block formatting context with inline size equal to the width of the window, but the `<body>` element with nonzero inline margins is the containing block for the float. The float placement must respect the content box of the `<body>` element (i.e. floats must not overlap the `<body>` element's margins), not that of the `<html>` element. Because a single block formatting context may contain floats with different containing blocks, the left and right "walls" of that containing block become properties of individual floats at the time of placement, not properties of the float context itself. Additionally, this commit generalizes the float placement logic a bit to allow the placement of arbitrary objects, not just floats. This is intended to support inline layout and block formatting context placement. This commit updates the `FloatContext` and associated tests only and doesn't actually wire the context up to the rest of layout, so floats in pages still aren't actually laid out. [1]: https://drafts.csswg.org/css2/#containing-block-details
834 lines
29 KiB
Rust
834 lines
29 KiB
Rust
/* This Source Code Form is subject to the terms of the Mozilla Public
|
|
* License, v. 2.0. If a copy of the MPL was not distributed with this
|
|
* file, You can obtain one at https://mozilla.org/MPL/2.0/. */
|
|
|
|
//! Property-based randomized testing for the core float layout algorithm.
|
|
|
|
#[macro_use]
|
|
extern crate lazy_static;
|
|
|
|
use euclid::num::Zero;
|
|
use layout::flow::float::{ClearSide, FloatBand, FloatBandNode, FloatBandTree, FloatContext};
|
|
use layout::flow::float::{FloatSide, PlacementInfo};
|
|
use layout::geom::flow_relative::{Rect, Vec2};
|
|
use quickcheck::{Arbitrary, Gen};
|
|
use std::f32;
|
|
use std::ops::Range;
|
|
use std::panic::{self, PanicInfo};
|
|
use std::sync::{Mutex, MutexGuard};
|
|
use std::thread;
|
|
use std::u32;
|
|
use style::values::computed::Length;
|
|
|
|
lazy_static! {
|
|
static ref PANIC_HOOK_MUTEX: Mutex<()> = Mutex::new(());
|
|
}
|
|
|
|
// Suppresses panic messages. Some tests need to fail and we don't want them to spam the console.
|
|
// Note that, because the panic hook is process-wide, tests that are expected to fail might
|
|
// suppress panic messages from other failing tests. To work around this, run failing tests one at
|
|
// a time or use only a single test thread.
|
|
struct PanicMsgSuppressor<'a> {
|
|
#[allow(dead_code)]
|
|
mutex_guard: MutexGuard<'a, ()>,
|
|
prev_hook: Option<Box<dyn Fn(&PanicInfo<'_>) + 'static + Sync + Send>>,
|
|
}
|
|
|
|
impl<'a> PanicMsgSuppressor<'a> {
|
|
fn new(mutex_guard: MutexGuard<'a, ()>) -> PanicMsgSuppressor<'a> {
|
|
let prev_hook = panic::take_hook();
|
|
panic::set_hook(Box::new(|_| ()));
|
|
PanicMsgSuppressor {
|
|
mutex_guard,
|
|
prev_hook: Some(prev_hook),
|
|
}
|
|
}
|
|
}
|
|
|
|
impl<'a> Drop for PanicMsgSuppressor<'a> {
|
|
fn drop(&mut self) {
|
|
panic::set_hook(self.prev_hook.take().unwrap())
|
|
}
|
|
}
|
|
|
|
// AA tree helpers
|
|
|
|
#[derive(Clone, Debug)]
|
|
struct FloatBandWrapper(FloatBand);
|
|
|
|
impl Arbitrary for FloatBandWrapper {
|
|
fn arbitrary<G>(generator: &mut G) -> FloatBandWrapper
|
|
where
|
|
G: Gen,
|
|
{
|
|
let top: u32 = Arbitrary::arbitrary(generator);
|
|
let left: Option<u32> = Arbitrary::arbitrary(generator);
|
|
let right: Option<u32> = Arbitrary::arbitrary(generator);
|
|
FloatBandWrapper(FloatBand {
|
|
top: Length::new(top as f32),
|
|
left: left.map(|value| Length::new(value as f32)),
|
|
right: right.map(|value| Length::new(value as f32)),
|
|
})
|
|
}
|
|
}
|
|
|
|
#[derive(Clone, Debug)]
|
|
struct FloatRangeInput {
|
|
start_index: u32,
|
|
band_count: u32,
|
|
side: FloatSide,
|
|
length: u32,
|
|
}
|
|
|
|
impl Arbitrary for FloatRangeInput {
|
|
fn arbitrary<G>(generator: &mut G) -> FloatRangeInput
|
|
where
|
|
G: Gen,
|
|
{
|
|
let start_index: u32 = Arbitrary::arbitrary(generator);
|
|
let band_count: u32 = Arbitrary::arbitrary(generator);
|
|
let is_left: bool = Arbitrary::arbitrary(generator);
|
|
let length: u32 = Arbitrary::arbitrary(generator);
|
|
FloatRangeInput {
|
|
start_index,
|
|
band_count,
|
|
side: if is_left {
|
|
FloatSide::Left
|
|
} else {
|
|
FloatSide::Right
|
|
},
|
|
length,
|
|
}
|
|
}
|
|
}
|
|
|
|
// AA tree predicates
|
|
|
|
fn check_node_ordering(node: &FloatBandNode) {
|
|
let mid = node.band.top;
|
|
if let Some(ref left) = node.left.0 {
|
|
assert!(left.band.top < mid);
|
|
}
|
|
if let Some(ref right) = node.right.0 {
|
|
assert!(right.band.top > mid);
|
|
}
|
|
if let Some(ref left) = node.left.0 {
|
|
check_node_ordering(left);
|
|
}
|
|
if let Some(ref right) = node.right.0 {
|
|
check_node_ordering(right);
|
|
}
|
|
}
|
|
|
|
// https://en.wikipedia.org/wiki/AA_tree#Balancing_rotations
|
|
fn check_node_balance(node: &FloatBandNode) {
|
|
// 1. The level of every leaf node is one.
|
|
if node.left.0.is_none() && node.right.0.is_none() {
|
|
assert_eq!(node.level, 1);
|
|
}
|
|
// 2. The level of every left child is exactly one less than that of its parent.
|
|
if let Some(ref left) = node.left.0 {
|
|
assert_eq!(left.level, node.level - 1);
|
|
}
|
|
// 3. The level of every right child is equal to or one less than that of its parent.
|
|
if let Some(ref right) = node.right.0 {
|
|
assert!(right.level == node.level || right.level == node.level - 1);
|
|
}
|
|
// 4. The level of every right grandchild is strictly less than that of its grandparent.
|
|
if let Some(ref right) = node.right.0 {
|
|
if let Some(ref right_right) = right.right.0 {
|
|
assert!(right_right.level < node.level);
|
|
}
|
|
}
|
|
// 5. Every node of level greater than one has two children.
|
|
if node.level > 1 {
|
|
assert!(node.left.0.is_some() && node.right.0.is_some());
|
|
}
|
|
}
|
|
|
|
fn check_tree_ordering(tree: FloatBandTree) {
|
|
if let Some(ref root) = tree.root.0 {
|
|
check_node_ordering(root);
|
|
}
|
|
}
|
|
|
|
fn check_tree_balance(tree: FloatBandTree) {
|
|
if let Some(ref root) = tree.root.0 {
|
|
check_node_balance(root);
|
|
}
|
|
}
|
|
|
|
fn check_tree_find(tree: &FloatBandTree, block_position: Length, sorted_bands: &[FloatBand]) {
|
|
let found_band = tree
|
|
.find(block_position)
|
|
.expect("Couldn't find the band in the tree!");
|
|
let reference_band_index = sorted_bands
|
|
.iter()
|
|
.position(|band| band.top > block_position)
|
|
.expect("Couldn't find the reference band!") -
|
|
1;
|
|
let reference_band = &sorted_bands[reference_band_index];
|
|
assert_eq!(found_band.top, reference_band.top);
|
|
assert_eq!(found_band.left, reference_band.left);
|
|
assert_eq!(found_band.right, reference_band.right);
|
|
}
|
|
|
|
fn check_tree_find_next(tree: &FloatBandTree, block_position: Length, sorted_bands: &[FloatBand]) {
|
|
let found_band = tree
|
|
.find_next(block_position)
|
|
.expect("Couldn't find the band in the tree!");
|
|
let reference_band_index = sorted_bands
|
|
.iter()
|
|
.position(|band| band.top > block_position)
|
|
.expect("Couldn't find the reference band!");
|
|
let reference_band = &sorted_bands[reference_band_index];
|
|
assert_eq!(found_band.top, reference_band.top);
|
|
assert_eq!(found_band.left, reference_band.left);
|
|
assert_eq!(found_band.right, reference_band.right);
|
|
}
|
|
|
|
fn check_node_range_setting(
|
|
node: &FloatBandNode,
|
|
block_range: &Range<Length>,
|
|
side: FloatSide,
|
|
value: Length,
|
|
) {
|
|
if node.band.top >= block_range.start && node.band.top < block_range.end {
|
|
match side {
|
|
FloatSide::Left => assert!(node.band.left.unwrap() >= value),
|
|
FloatSide::Right => assert!(node.band.right.unwrap() <= value),
|
|
}
|
|
}
|
|
|
|
if let Some(ref left) = node.left.0 {
|
|
check_node_range_setting(left, block_range, side, value)
|
|
}
|
|
if let Some(ref right) = node.right.0 {
|
|
check_node_range_setting(right, block_range, side, value)
|
|
}
|
|
}
|
|
|
|
fn check_tree_range_setting(
|
|
tree: &FloatBandTree,
|
|
block_range: &Range<Length>,
|
|
side: FloatSide,
|
|
value: Length,
|
|
) {
|
|
if let Some(ref root) = tree.root.0 {
|
|
check_node_range_setting(root, block_range, side, value)
|
|
}
|
|
}
|
|
|
|
// AA tree unit tests
|
|
|
|
// Tests that the tree is a properly-ordered binary tree.
|
|
#[test]
|
|
fn test_tree_ordering() {
|
|
let f: fn(Vec<FloatBandWrapper>) = check;
|
|
quickcheck::quickcheck(f);
|
|
fn check(bands: Vec<FloatBandWrapper>) {
|
|
let mut tree = FloatBandTree::new();
|
|
for FloatBandWrapper(band) in bands {
|
|
tree = tree.insert(band);
|
|
}
|
|
check_tree_ordering(tree);
|
|
}
|
|
}
|
|
|
|
// Tests that the tree is balanced (i.e. AA tree invariants are maintained).
|
|
#[test]
|
|
fn test_tree_balance() {
|
|
let f: fn(Vec<FloatBandWrapper>) = check;
|
|
quickcheck::quickcheck(f);
|
|
fn check(bands: Vec<FloatBandWrapper>) {
|
|
let mut tree = FloatBandTree::new();
|
|
for FloatBandWrapper(band) in bands {
|
|
tree = tree.insert(band);
|
|
}
|
|
check_tree_balance(tree);
|
|
}
|
|
}
|
|
|
|
// Tests that the `find()` method works.
|
|
#[test]
|
|
fn test_tree_find() {
|
|
let f: fn(Vec<FloatBandWrapper>, Vec<u32>) = check;
|
|
quickcheck::quickcheck(f);
|
|
fn check(bands: Vec<FloatBandWrapper>, lookups: Vec<u32>) {
|
|
let mut bands: Vec<FloatBand> = bands.into_iter().map(|band| band.0).collect();
|
|
bands.push(FloatBand {
|
|
top: Length::zero(),
|
|
left: None,
|
|
right: None,
|
|
});
|
|
bands.push(FloatBand {
|
|
top: Length::new(f32::INFINITY),
|
|
left: None,
|
|
right: None,
|
|
});
|
|
let mut tree = FloatBandTree::new();
|
|
for ref band in &bands {
|
|
tree = tree.insert((*band).clone());
|
|
}
|
|
bands.sort_by(|a, b| a.top.partial_cmp(&b.top).unwrap());
|
|
for lookup in lookups {
|
|
check_tree_find(&tree, Length::new(lookup as f32), &bands);
|
|
}
|
|
}
|
|
}
|
|
|
|
// Tests that the `find_next()` method works.
|
|
#[test]
|
|
fn test_tree_find_next() {
|
|
let f: fn(Vec<FloatBandWrapper>, Vec<u32>) = check;
|
|
quickcheck::quickcheck(f);
|
|
fn check(bands: Vec<FloatBandWrapper>, lookups: Vec<u32>) {
|
|
let mut bands: Vec<FloatBand> = bands.into_iter().map(|band| band.0).collect();
|
|
bands.push(FloatBand {
|
|
top: Length::zero(),
|
|
left: None,
|
|
right: None,
|
|
});
|
|
bands.push(FloatBand {
|
|
top: Length::new(f32::INFINITY),
|
|
left: None,
|
|
right: None,
|
|
});
|
|
bands.sort_by(|a, b| a.top.partial_cmp(&b.top).unwrap());
|
|
bands.dedup_by(|a, b| a.top == b.top);
|
|
let mut tree = FloatBandTree::new();
|
|
for ref band in &bands {
|
|
tree = tree.insert((*band).clone());
|
|
}
|
|
for lookup in lookups {
|
|
check_tree_find_next(&tree, Length::new(lookup as f32), &bands);
|
|
}
|
|
}
|
|
}
|
|
|
|
// Tests that `set_range()` works.
|
|
#[test]
|
|
fn test_tree_range_setting() {
|
|
let f: fn(Vec<FloatBandWrapper>, Vec<FloatRangeInput>) = check;
|
|
quickcheck::quickcheck(f);
|
|
fn check(bands: Vec<FloatBandWrapper>, ranges: Vec<FloatRangeInput>) {
|
|
let mut tree = FloatBandTree::new();
|
|
for FloatBandWrapper(ref band) in &bands {
|
|
tree = tree.insert((*band).clone());
|
|
}
|
|
|
|
let mut tops: Vec<Length> = bands.iter().map(|band| band.0.top).collect();
|
|
tops.push(Length::new(f32::INFINITY));
|
|
tops.sort_by(|a, b| a.px().partial_cmp(&b.px()).unwrap());
|
|
|
|
for range in ranges {
|
|
let start = range.start_index.min(tops.len() as u32 - 1);
|
|
let end = (range.start_index + range.length).min(tops.len() as u32 - 1);
|
|
let block_range = tops[start as usize]..tops[end as usize];
|
|
let length = Length::new(range.length as f32);
|
|
let new_tree = tree.set_range(&block_range, range.side, length);
|
|
check_tree_range_setting(&new_tree, &block_range, range.side, length);
|
|
}
|
|
}
|
|
}
|
|
|
|
// Float predicates
|
|
|
|
#[derive(Clone, Debug)]
|
|
struct FloatInput {
|
|
// Information needed to place the float.
|
|
info: PlacementInfo,
|
|
// The float may be placed no higher than this line. This simulates the effect of line boxes
|
|
// per CSS 2.1 § 9.5.1 rule 6.
|
|
ceiling: u32,
|
|
}
|
|
|
|
impl Arbitrary for FloatInput {
|
|
fn arbitrary<G>(generator: &mut G) -> FloatInput
|
|
where
|
|
G: Gen,
|
|
{
|
|
let width: u32 = Arbitrary::arbitrary(generator);
|
|
let height: u32 = Arbitrary::arbitrary(generator);
|
|
let is_left: bool = Arbitrary::arbitrary(generator);
|
|
let ceiling: u32 = Arbitrary::arbitrary(generator);
|
|
let left_wall: u32 = Arbitrary::arbitrary(generator);
|
|
let right_wall: u32 = Arbitrary::arbitrary(generator);
|
|
let clear: u8 = Arbitrary::arbitrary(generator);
|
|
FloatInput {
|
|
info: PlacementInfo {
|
|
size: Vec2 {
|
|
inline: Length::new(width as f32),
|
|
block: Length::new(height as f32),
|
|
},
|
|
side: if is_left {
|
|
FloatSide::Left
|
|
} else {
|
|
FloatSide::Right
|
|
},
|
|
clear: new_clear_side(clear),
|
|
left_wall: Length::new(left_wall as f32),
|
|
right_wall: Length::new(right_wall as f32),
|
|
},
|
|
ceiling,
|
|
}
|
|
}
|
|
|
|
fn shrink(&self) -> Box<dyn Iterator<Item = FloatInput>> {
|
|
let mut this = (*self).clone();
|
|
let mut shrunk = false;
|
|
if let Some(inline_size) = self.info.size.inline.px().shrink().next() {
|
|
this.info.size.inline = Length::new(inline_size);
|
|
shrunk = true;
|
|
}
|
|
if let Some(block_size) = self.info.size.block.px().shrink().next() {
|
|
this.info.size.block = Length::new(block_size);
|
|
shrunk = true;
|
|
}
|
|
if let Some(clear_side) = (self.info.clear as u8).shrink().next() {
|
|
this.info.clear = new_clear_side(clear_side);
|
|
shrunk = true;
|
|
}
|
|
if let Some(left_wall) = self.info.left_wall.px().shrink().next() {
|
|
this.info.left_wall = Length::new(left_wall);
|
|
shrunk = true;
|
|
}
|
|
if let Some(right_wall) = self.info.right_wall.px().shrink().next() {
|
|
this.info.right_wall = Length::new(right_wall);
|
|
shrunk = true;
|
|
}
|
|
if let Some(ceiling) = self.ceiling.shrink().next() {
|
|
this.ceiling = ceiling;
|
|
shrunk = true;
|
|
}
|
|
if shrunk {
|
|
quickcheck::single_shrinker(this)
|
|
} else {
|
|
quickcheck::empty_shrinker()
|
|
}
|
|
}
|
|
}
|
|
|
|
fn new_clear_side(value: u8) -> ClearSide {
|
|
match value & 3 {
|
|
0 => ClearSide::None,
|
|
1 => ClearSide::Left,
|
|
2 => ClearSide::Right,
|
|
_ => ClearSide::Both,
|
|
}
|
|
}
|
|
|
|
#[derive(Clone)]
|
|
struct FloatPlacement {
|
|
float_context: FloatContext,
|
|
placed_floats: Vec<PlacedFloat>,
|
|
}
|
|
|
|
// Information about the placement of a float.
|
|
#[derive(Clone)]
|
|
struct PlacedFloat {
|
|
origin: Vec2<Length>,
|
|
info: PlacementInfo,
|
|
ceiling: Length,
|
|
}
|
|
|
|
impl Drop for FloatPlacement {
|
|
fn drop(&mut self) {
|
|
if !thread::panicking() {
|
|
return;
|
|
}
|
|
|
|
// Dump the float context for debugging.
|
|
eprintln!("Failing float placement:");
|
|
for placed_float in &self.placed_floats {
|
|
eprintln!(
|
|
" * {:?} @ {:?}, {:?}",
|
|
placed_float.info, placed_float.origin, placed_float.ceiling
|
|
);
|
|
}
|
|
eprintln!("Bands:\n{:?}\n", self.float_context.bands);
|
|
}
|
|
}
|
|
|
|
impl PlacedFloat {
|
|
fn rect(&self) -> Rect<Length> {
|
|
Rect {
|
|
start_corner: self.origin.clone(),
|
|
size: self.info.size.clone(),
|
|
}
|
|
}
|
|
}
|
|
|
|
impl FloatPlacement {
|
|
fn place(floats: Vec<FloatInput>) -> FloatPlacement {
|
|
let mut float_context = FloatContext::new();
|
|
let mut placed_floats = vec![];
|
|
for float in floats {
|
|
let ceiling = Length::new(float.ceiling as f32);
|
|
float_context.lower_ceiling(ceiling);
|
|
placed_floats.push(PlacedFloat {
|
|
origin: float_context.add_float(&float.info),
|
|
info: float.info,
|
|
ceiling,
|
|
})
|
|
}
|
|
FloatPlacement {
|
|
float_context,
|
|
placed_floats,
|
|
}
|
|
}
|
|
}
|
|
|
|
// From CSS 2.1 § 9.5.1 [1].
|
|
//
|
|
// [1]: https://www.w3.org/TR/CSS2/visuren.html#float-position
|
|
|
|
// 1. The left outer edge of a left-floating box may not be to the left of the left edge of its
|
|
// containing block. An analogous rule holds for right-floating elements.
|
|
fn check_floats_rule_1(placement: &FloatPlacement) {
|
|
for placed_float in &placement.placed_floats {
|
|
match placed_float.info.side {
|
|
FloatSide::Left => assert!(placed_float.origin.inline >= placed_float.info.left_wall),
|
|
FloatSide::Right => {
|
|
assert!(placed_float.rect().max_inline_position() <= placed_float.info.right_wall)
|
|
},
|
|
}
|
|
}
|
|
}
|
|
|
|
// 2. If the current box is left-floating, and there are any left-floating boxes generated by
|
|
// elements earlier in the source document, then for each such earlier box, either the left
|
|
// outer edge of the current box must be to the right of the right outer edge of the earlier
|
|
// box, or its top must be lower than the bottom of the earlier box. Analogous rules hold for
|
|
// right-floating boxes.
|
|
fn check_floats_rule_2(placement: &FloatPlacement) {
|
|
for (this_float_index, this_float) in placement.placed_floats.iter().enumerate() {
|
|
for prev_float in &placement.placed_floats[0..this_float_index] {
|
|
match (this_float.info.side, prev_float.info.side) {
|
|
(FloatSide::Left, FloatSide::Left) => {
|
|
assert!(
|
|
this_float.origin.inline >= prev_float.rect().max_inline_position() ||
|
|
this_float.origin.block >= prev_float.rect().max_block_position()
|
|
);
|
|
},
|
|
(FloatSide::Right, FloatSide::Right) => {
|
|
assert!(
|
|
this_float.rect().max_inline_position() <= prev_float.origin.inline ||
|
|
this_float.origin.block >= prev_float.rect().max_block_position()
|
|
);
|
|
},
|
|
(FloatSide::Left, FloatSide::Right) | (FloatSide::Right, FloatSide::Left) => {},
|
|
}
|
|
}
|
|
}
|
|
}
|
|
|
|
// 3. The right outer edge of a left-floating box may not be to the right of the left outer edge of
|
|
// any right-floating box that is next to it. Analogous rules hold for right-floating elements.
|
|
fn check_floats_rule_3(placement: &FloatPlacement) {
|
|
for (this_float_index, this_float) in placement.placed_floats.iter().enumerate() {
|
|
for other_float in &placement.placed_floats[0..this_float_index] {
|
|
// This logic to check intersection is complicated by the fact that we need to treat
|
|
// zero-height floats later in the document as "next to" floats earlier in the
|
|
// document. Otherwise we might end up with a situation like:
|
|
//
|
|
// <div id="a" style="float: left; width: 32px; height: 32px"></div>
|
|
// <div id="b" style="float: right; width: 0px; height: 0px"></div>
|
|
//
|
|
// Where the top of `b` should probably be 32px per Rule 3, but unless this distinction
|
|
// is made the top of `b` could legally be 0px.
|
|
if this_float.origin.block >= other_float.rect().max_block_position() ||
|
|
(this_float.info.size.block == Length::zero() &&
|
|
this_float.rect().max_block_position() < other_float.origin.block) ||
|
|
(this_float.info.size.block > Length::zero() &&
|
|
this_float.rect().max_block_position() <= other_float.origin.block)
|
|
{
|
|
continue;
|
|
}
|
|
|
|
match (this_float.info.side, other_float.info.side) {
|
|
(FloatSide::Left, FloatSide::Right) => {
|
|
assert!(this_float.rect().max_inline_position() <= other_float.origin.inline);
|
|
},
|
|
(FloatSide::Right, FloatSide::Left) => {
|
|
assert!(this_float.origin.inline >= other_float.rect().max_inline_position());
|
|
},
|
|
(FloatSide::Left, FloatSide::Left) | (FloatSide::Right, FloatSide::Right) => {},
|
|
}
|
|
}
|
|
}
|
|
}
|
|
|
|
// 4. A floating box's outer top may not be higher than the top of its containing block. When the
|
|
// float occurs between two collapsing margins, the float is positioned as if it had an
|
|
// otherwise empty anonymous block parent taking part in the flow. The position of such a parent
|
|
// is defined by the rules in the section on margin collapsing.
|
|
fn check_floats_rule_4(placement: &FloatPlacement) {
|
|
for placed_float in &placement.placed_floats {
|
|
assert!(placed_float.origin.block >= Length::zero());
|
|
}
|
|
}
|
|
|
|
// 5. The outer top of a floating box may not be higher than the outer top of any block or floated
|
|
// box generated by an element earlier in the source document.
|
|
fn check_floats_rule_5(placement: &FloatPlacement) {
|
|
let mut block_position = Length::zero();
|
|
for placed_float in &placement.placed_floats {
|
|
assert!(placed_float.origin.block >= block_position);
|
|
block_position = placed_float.origin.block;
|
|
}
|
|
}
|
|
|
|
// 6. The outer top of an element's floating box may not be higher than the top of any line-box
|
|
// containing a box generated by an element earlier in the source document.
|
|
fn check_floats_rule_6(placement: &FloatPlacement) {
|
|
for placed_float in &placement.placed_floats {
|
|
assert!(placed_float.origin.block >= placed_float.ceiling);
|
|
}
|
|
}
|
|
|
|
// 7. A left-floating box that has another left-floating box to its left may not have its right
|
|
// outer edge to the right of its containing block's right edge. (Loosely: a left float may not
|
|
// stick out at the right edge, unless it is already as far to the left as possible.) An
|
|
// analogous rule holds for right-floating elements.
|
|
fn check_floats_rule_7(placement: &FloatPlacement) {
|
|
for (placed_float_index, placed_float) in placement.placed_floats.iter().enumerate() {
|
|
// Only consider floats that stick out.
|
|
match placed_float.info.side {
|
|
FloatSide::Left => {
|
|
if placed_float.rect().max_inline_position() <= placed_float.info.right_wall {
|
|
continue;
|
|
}
|
|
},
|
|
FloatSide::Right => {
|
|
if placed_float.origin.inline >= placed_float.info.left_wall {
|
|
continue;
|
|
}
|
|
},
|
|
}
|
|
|
|
// Make sure there are no previous floats to the left or right.
|
|
for prev_float in &placement.placed_floats[0..placed_float_index] {
|
|
assert!(
|
|
prev_float.info.side != placed_float.info.side ||
|
|
prev_float.rect().max_block_position() <= placed_float.origin.block ||
|
|
prev_float.origin.block >= placed_float.rect().max_block_position()
|
|
);
|
|
}
|
|
}
|
|
}
|
|
|
|
// 8. A floating box must be placed as high as possible.
|
|
fn check_floats_rule_8(floats_and_perturbations: Vec<(FloatInput, u32)>) {
|
|
let floats = floats_and_perturbations
|
|
.iter()
|
|
.map(|&(ref float, _)| (*float).clone())
|
|
.collect();
|
|
let placement = FloatPlacement::place(floats);
|
|
|
|
for (float_index, &(_, perturbation)) in floats_and_perturbations.iter().enumerate() {
|
|
if perturbation == 0 {
|
|
continue;
|
|
}
|
|
|
|
let mut placement = placement.clone();
|
|
placement.placed_floats[float_index].origin.block =
|
|
placement.placed_floats[float_index].origin.block - Length::new(perturbation as f32);
|
|
|
|
let result = {
|
|
let mutex_guard = PANIC_HOOK_MUTEX.lock().unwrap();
|
|
let _suppressor = PanicMsgSuppressor::new(mutex_guard);
|
|
panic::catch_unwind(|| check_basic_float_rules(&placement))
|
|
};
|
|
assert!(result.is_err());
|
|
}
|
|
}
|
|
|
|
// 9. A left-floating box must be put as far to the left as possible, a right-floating box as far
|
|
// to the right as possible. A higher position is preferred over one that is further to the
|
|
// left/right.
|
|
fn check_floats_rule_9(floats_and_perturbations: Vec<(FloatInput, u32)>) {
|
|
let floats = floats_and_perturbations
|
|
.iter()
|
|
.map(|&(ref float, _)| (*float).clone())
|
|
.collect();
|
|
let placement = FloatPlacement::place(floats);
|
|
|
|
for (float_index, &(_, perturbation)) in floats_and_perturbations.iter().enumerate() {
|
|
if perturbation == 0 {
|
|
continue;
|
|
}
|
|
|
|
let mut placement = placement.clone();
|
|
{
|
|
let mut placed_float = &mut placement.placed_floats[float_index];
|
|
let perturbation = Length::new(perturbation as f32);
|
|
match placed_float.info.side {
|
|
FloatSide::Left => {
|
|
placed_float.origin.inline = placed_float.origin.inline - perturbation
|
|
},
|
|
FloatSide::Right => {
|
|
placed_float.origin.inline = placed_float.origin.inline + perturbation
|
|
},
|
|
}
|
|
}
|
|
|
|
let result = {
|
|
let mutex_guard = PANIC_HOOK_MUTEX.lock().unwrap();
|
|
let _suppressor = PanicMsgSuppressor::new(mutex_guard);
|
|
panic::catch_unwind(|| check_basic_float_rules(&placement))
|
|
};
|
|
assert!(result.is_err());
|
|
}
|
|
}
|
|
|
|
// From CSS 2.1 § 9.5.2 (https://www.w3.org/TR/CSS2/visuren.html#propdef-clear):
|
|
//
|
|
// 10. The top outer edge of the float must be below the bottom outer edge of all earlier
|
|
// left-floating boxes (in the case of 'clear: left'), or all earlier right-floating boxes (in
|
|
// the case of 'clear: right'), or both ('clear: both').
|
|
fn check_floats_rule_10(placement: &FloatPlacement) {
|
|
let mut block_position = Length::zero();
|
|
for placed_float in &placement.placed_floats {
|
|
assert!(placed_float.origin.block >= block_position);
|
|
block_position = placed_float.origin.block;
|
|
}
|
|
|
|
for (this_float_index, this_float) in placement.placed_floats.iter().enumerate() {
|
|
if this_float.info.clear == ClearSide::None {
|
|
continue;
|
|
}
|
|
|
|
for other_float in &placement.placed_floats[0..this_float_index] {
|
|
// This logic to check intersection is complicated by the fact that we need to treat
|
|
// zero-height floats later in the document as "next to" floats earlier in the
|
|
// document. Otherwise we might end up with a situation like:
|
|
//
|
|
// <div id="a" style="float: left; width: 32px; height: 32px"></div>
|
|
// <div id="b" style="float: right; width: 0px; height: 0px"></div>
|
|
//
|
|
// Where the top of `b` should probably be 32px per Rule 3, but unless this distinction
|
|
// is made the top of `b` could legally be 0px.
|
|
if this_float.origin.block >= other_float.rect().max_block_position() ||
|
|
(this_float.info.size.block == Length::zero() &&
|
|
this_float.rect().max_block_position() < other_float.origin.block) ||
|
|
(this_float.info.size.block > Length::zero() &&
|
|
this_float.rect().max_block_position() <= other_float.origin.block)
|
|
{
|
|
continue;
|
|
}
|
|
|
|
match this_float.info.clear {
|
|
ClearSide::Left => assert_ne!(other_float.info.side, FloatSide::Left),
|
|
ClearSide::Right => assert_ne!(other_float.info.side, FloatSide::Right),
|
|
ClearSide::Both => assert!(false),
|
|
ClearSide::None => unreachable!(),
|
|
}
|
|
}
|
|
}
|
|
}
|
|
|
|
// Checks that rule 1-7 and rule 10 hold (i.e. all rules that don't specify that floats are placed
|
|
// "as far as possible" in some direction).
|
|
fn check_basic_float_rules(placement: &FloatPlacement) {
|
|
check_floats_rule_1(placement);
|
|
check_floats_rule_2(placement);
|
|
check_floats_rule_3(placement);
|
|
check_floats_rule_4(placement);
|
|
check_floats_rule_5(placement);
|
|
check_floats_rule_6(placement);
|
|
check_floats_rule_7(placement);
|
|
check_floats_rule_10(placement);
|
|
}
|
|
|
|
// Float unit tests
|
|
|
|
#[test]
|
|
fn test_floats_rule_1() {
|
|
let f: fn(Vec<FloatInput>) = check;
|
|
quickcheck::quickcheck(f);
|
|
fn check(floats: Vec<FloatInput>) {
|
|
check_floats_rule_1(&FloatPlacement::place(floats));
|
|
}
|
|
}
|
|
|
|
#[test]
|
|
fn test_floats_rule_2() {
|
|
let f: fn(Vec<FloatInput>) = check;
|
|
quickcheck::quickcheck(f);
|
|
fn check(floats: Vec<FloatInput>) {
|
|
check_floats_rule_2(&FloatPlacement::place(floats));
|
|
}
|
|
}
|
|
|
|
#[test]
|
|
fn test_floats_rule_3() {
|
|
let f: fn(Vec<FloatInput>) = check;
|
|
quickcheck::quickcheck(f);
|
|
fn check(floats: Vec<FloatInput>) {
|
|
check_floats_rule_3(&FloatPlacement::place(floats));
|
|
}
|
|
}
|
|
|
|
#[test]
|
|
fn test_floats_rule_4() {
|
|
let f: fn(Vec<FloatInput>) = check;
|
|
quickcheck::quickcheck(f);
|
|
fn check(floats: Vec<FloatInput>) {
|
|
check_floats_rule_4(&FloatPlacement::place(floats));
|
|
}
|
|
}
|
|
|
|
#[test]
|
|
fn test_floats_rule_5() {
|
|
let f: fn(Vec<FloatInput>) = check;
|
|
quickcheck::quickcheck(f);
|
|
fn check(floats: Vec<FloatInput>) {
|
|
check_floats_rule_5(&FloatPlacement::place(floats));
|
|
}
|
|
}
|
|
|
|
#[test]
|
|
fn test_floats_rule_6() {
|
|
let f: fn(Vec<FloatInput>) = check;
|
|
quickcheck::quickcheck(f);
|
|
fn check(floats: Vec<FloatInput>) {
|
|
check_floats_rule_6(&FloatPlacement::place(floats));
|
|
}
|
|
}
|
|
|
|
#[test]
|
|
fn test_floats_rule_7() {
|
|
let f: fn(Vec<FloatInput>) = check;
|
|
quickcheck::quickcheck(f);
|
|
fn check(floats: Vec<FloatInput>) {
|
|
check_floats_rule_7(&FloatPlacement::place(floats));
|
|
}
|
|
}
|
|
|
|
#[test]
|
|
fn test_floats_rule_8() {
|
|
let f: fn(Vec<(FloatInput, u32)>) = check;
|
|
quickcheck::quickcheck(f);
|
|
fn check(floats: Vec<(FloatInput, u32)>) {
|
|
check_floats_rule_8(floats);
|
|
}
|
|
}
|
|
|
|
#[test]
|
|
fn test_floats_rule_9() {
|
|
let f: fn(Vec<(FloatInput, u32)>) = check;
|
|
quickcheck::quickcheck(f);
|
|
fn check(floats: Vec<(FloatInput, u32)>) {
|
|
check_floats_rule_9(floats);
|
|
}
|
|
}
|
|
|
|
#[test]
|
|
fn test_floats_rule_10() {
|
|
let f: fn(Vec<FloatInput>) = check;
|
|
quickcheck::quickcheck(f);
|
|
fn check(floats: Vec<FloatInput>) {
|
|
check_floats_rule_10(&FloatPlacement::place(floats));
|
|
}
|
|
}
|