1292 lines
44 KiB
Rust
1292 lines
44 KiB
Rust
// SPDX-License-Identifier: MIT OR Apache-2.0
|
|
|
|
#![allow(clippy::too_many_arguments)]
|
|
|
|
#[cfg(not(feature = "std"))]
|
|
use alloc::vec::Vec;
|
|
use core::cmp::{max, min};
|
|
use core::fmt;
|
|
use core::mem;
|
|
use core::ops::Range;
|
|
use unicode_script::{Script, UnicodeScript};
|
|
use unicode_segmentation::UnicodeSegmentation;
|
|
|
|
use crate::fallback::FontFallbackIter;
|
|
use crate::{
|
|
Align, AttrsList, Color, Font, FontSystem, LayoutGlyph, LayoutLine, ShapePlanCache, Wrap,
|
|
};
|
|
|
|
/// The shaping strategy of some text.
|
|
#[derive(Debug, Clone, Copy, PartialEq, Eq, Hash)]
|
|
pub enum Shaping {
|
|
/// Basic shaping with no font fallback.
|
|
///
|
|
/// This shaping strategy is very cheap, but it will not display complex
|
|
/// scripts properly nor try to find missing glyphs in your system fonts.
|
|
///
|
|
/// You should use this strategy when you have complete control of the text
|
|
/// and the font you are displaying in your application.
|
|
#[cfg(feature = "swash")]
|
|
Basic,
|
|
/// Advanced text shaping and font fallback.
|
|
///
|
|
/// You will need to enable this strategy if the text contains a complex
|
|
/// script, the font used needs it, and/or multiple fonts in your system
|
|
/// may be needed to display all of the glyphs.
|
|
Advanced,
|
|
}
|
|
|
|
impl Shaping {
|
|
fn run(
|
|
self,
|
|
scratch: &mut ShapeBuffer,
|
|
glyphs: &mut Vec<ShapeGlyph>,
|
|
font_system: &mut FontSystem,
|
|
line: &str,
|
|
attrs_list: &AttrsList,
|
|
start_run: usize,
|
|
end_run: usize,
|
|
span_rtl: bool,
|
|
) {
|
|
match self {
|
|
#[cfg(feature = "swash")]
|
|
Self::Basic => shape_skip(font_system, glyphs, line, attrs_list, start_run, end_run),
|
|
Self::Advanced => shape_run(
|
|
scratch,
|
|
glyphs,
|
|
font_system,
|
|
line,
|
|
attrs_list,
|
|
start_run,
|
|
end_run,
|
|
span_rtl,
|
|
),
|
|
}
|
|
}
|
|
}
|
|
|
|
/// A set of buffers containing allocations for shaped text.
|
|
#[derive(Default)]
|
|
pub struct ShapeBuffer {
|
|
/// Buffer for holding unicode text.
|
|
rustybuzz_buffer: Option<rustybuzz::UnicodeBuffer>,
|
|
|
|
/// Temporary buffers for scripts.
|
|
scripts: Vec<Script>,
|
|
|
|
/// Buffer for visual lines.
|
|
visual_lines: Vec<VisualLine>,
|
|
}
|
|
|
|
impl fmt::Debug for ShapeBuffer {
|
|
fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
|
|
f.pad("ShapeBuffer { .. }")
|
|
}
|
|
}
|
|
|
|
fn shape_fallback(
|
|
scratch: &mut ShapeBuffer,
|
|
glyphs: &mut Vec<ShapeGlyph>,
|
|
shape_plan_cache: &mut ShapePlanCache,
|
|
font: &Font,
|
|
line: &str,
|
|
attrs_list: &AttrsList,
|
|
start_run: usize,
|
|
end_run: usize,
|
|
span_rtl: bool,
|
|
) -> Vec<usize> {
|
|
let run = &line[start_run..end_run];
|
|
|
|
let font_scale = font.rustybuzz().units_per_em() as f32;
|
|
let ascent = font.rustybuzz().ascender() as f32 / font_scale;
|
|
let descent = -font.rustybuzz().descender() as f32 / font_scale;
|
|
|
|
let mut buffer = scratch.rustybuzz_buffer.take().unwrap_or_default();
|
|
buffer.set_direction(if span_rtl {
|
|
rustybuzz::Direction::RightToLeft
|
|
} else {
|
|
rustybuzz::Direction::LeftToRight
|
|
});
|
|
buffer.push_str(run);
|
|
buffer.guess_segment_properties();
|
|
|
|
let rtl = matches!(buffer.direction(), rustybuzz::Direction::RightToLeft);
|
|
assert_eq!(rtl, span_rtl);
|
|
|
|
let shape_plan = shape_plan_cache.get(font, &buffer);
|
|
let glyph_buffer = rustybuzz::shape_with_plan(font.rustybuzz(), shape_plan, buffer);
|
|
let glyph_infos = glyph_buffer.glyph_infos();
|
|
let glyph_positions = glyph_buffer.glyph_positions();
|
|
|
|
let mut missing = Vec::new();
|
|
glyphs.reserve(glyph_infos.len());
|
|
let glyph_start = glyphs.len();
|
|
for (info, pos) in glyph_infos.iter().zip(glyph_positions.iter()) {
|
|
let x_advance = pos.x_advance as f32 / font_scale;
|
|
let y_advance = pos.y_advance as f32 / font_scale;
|
|
let x_offset = pos.x_offset as f32 / font_scale;
|
|
let y_offset = pos.y_offset as f32 / font_scale;
|
|
|
|
let start_glyph = start_run + info.cluster as usize;
|
|
|
|
if info.glyph_id == 0 {
|
|
missing.push(start_glyph);
|
|
}
|
|
|
|
let attrs = attrs_list.get_span(start_glyph);
|
|
glyphs.push(ShapeGlyph {
|
|
start: start_glyph,
|
|
end: end_run, // Set later
|
|
x_advance,
|
|
y_advance,
|
|
x_offset,
|
|
y_offset,
|
|
ascent,
|
|
descent,
|
|
font_id: font.id(),
|
|
glyph_id: info.glyph_id.try_into().expect("failed to cast glyph ID"),
|
|
//TODO: color should not be related to shaping
|
|
color_opt: attrs.color_opt,
|
|
metadata: attrs.metadata,
|
|
});
|
|
}
|
|
|
|
// Adjust end of glyphs
|
|
if rtl {
|
|
for i in glyph_start + 1..glyphs.len() {
|
|
let next_start = glyphs[i - 1].start;
|
|
let next_end = glyphs[i - 1].end;
|
|
let prev = &mut glyphs[i];
|
|
if prev.start == next_start {
|
|
prev.end = next_end;
|
|
} else {
|
|
prev.end = next_start;
|
|
}
|
|
}
|
|
} else {
|
|
for i in (glyph_start + 1..glyphs.len()).rev() {
|
|
let next_start = glyphs[i].start;
|
|
let next_end = glyphs[i].end;
|
|
let prev = &mut glyphs[i - 1];
|
|
if prev.start == next_start {
|
|
prev.end = next_end;
|
|
} else {
|
|
prev.end = next_start;
|
|
}
|
|
}
|
|
}
|
|
|
|
// Restore the buffer to save an allocation.
|
|
scratch.rustybuzz_buffer = Some(glyph_buffer.clear());
|
|
|
|
missing
|
|
}
|
|
|
|
fn shape_run(
|
|
scratch: &mut ShapeBuffer,
|
|
glyphs: &mut Vec<ShapeGlyph>,
|
|
font_system: &mut FontSystem,
|
|
line: &str,
|
|
attrs_list: &AttrsList,
|
|
start_run: usize,
|
|
end_run: usize,
|
|
span_rtl: bool,
|
|
) {
|
|
// Re-use the previous script buffer if possible.
|
|
let mut scripts = {
|
|
let mut scripts = mem::take(&mut scratch.scripts);
|
|
scripts.clear();
|
|
scripts
|
|
};
|
|
for c in line[start_run..end_run].chars() {
|
|
match c.script() {
|
|
Script::Common | Script::Inherited | Script::Latin | Script::Unknown => (),
|
|
script => {
|
|
if !scripts.contains(&script) {
|
|
scripts.push(script);
|
|
}
|
|
}
|
|
}
|
|
}
|
|
|
|
log::trace!(" Run {:?}: '{}'", &scripts, &line[start_run..end_run],);
|
|
|
|
let attrs = attrs_list.get_span(start_run);
|
|
|
|
let fonts = font_system.get_font_matches(attrs);
|
|
|
|
let default_families = [&attrs.family];
|
|
let mut font_iter = FontFallbackIter::new(font_system, &fonts, &default_families, &scripts);
|
|
|
|
let font = font_iter.next().expect("no default font found");
|
|
|
|
let glyph_start = glyphs.len();
|
|
let mut missing = shape_fallback(
|
|
scratch,
|
|
glyphs,
|
|
font_iter.shape_plan_cache(),
|
|
&font,
|
|
line,
|
|
attrs_list,
|
|
start_run,
|
|
end_run,
|
|
span_rtl,
|
|
);
|
|
|
|
//TODO: improve performance!
|
|
while !missing.is_empty() {
|
|
let font = match font_iter.next() {
|
|
Some(some) => some,
|
|
None => break,
|
|
};
|
|
|
|
log::trace!(
|
|
"Evaluating fallback with font '{}'",
|
|
font_iter.face_name(font.id())
|
|
);
|
|
let mut fb_glyphs = Vec::new();
|
|
let fb_missing = shape_fallback(
|
|
scratch,
|
|
&mut fb_glyphs,
|
|
font_iter.shape_plan_cache(),
|
|
&font,
|
|
line,
|
|
attrs_list,
|
|
start_run,
|
|
end_run,
|
|
span_rtl,
|
|
);
|
|
|
|
// Insert all matching glyphs
|
|
let mut fb_i = 0;
|
|
while fb_i < fb_glyphs.len() {
|
|
let start = fb_glyphs[fb_i].start;
|
|
let end = fb_glyphs[fb_i].end;
|
|
|
|
// Skip clusters that are not missing, or where the fallback font is missing
|
|
if !missing.contains(&start) || fb_missing.contains(&start) {
|
|
fb_i += 1;
|
|
continue;
|
|
}
|
|
|
|
let mut missing_i = 0;
|
|
while missing_i < missing.len() {
|
|
if missing[missing_i] >= start && missing[missing_i] < end {
|
|
// println!("No longer missing {}", missing[missing_i]);
|
|
missing.remove(missing_i);
|
|
} else {
|
|
missing_i += 1;
|
|
}
|
|
}
|
|
|
|
// Find prior glyphs
|
|
let mut i = glyph_start;
|
|
while i < glyphs.len() {
|
|
if glyphs[i].start >= start && glyphs[i].end <= end {
|
|
break;
|
|
} else {
|
|
i += 1;
|
|
}
|
|
}
|
|
|
|
// Remove prior glyphs
|
|
while i < glyphs.len() {
|
|
if glyphs[i].start >= start && glyphs[i].end <= end {
|
|
let _glyph = glyphs.remove(i);
|
|
// log::trace!("Removed {},{} from {}", _glyph.start, _glyph.end, i);
|
|
} else {
|
|
break;
|
|
}
|
|
}
|
|
|
|
while fb_i < fb_glyphs.len() {
|
|
if fb_glyphs[fb_i].start >= start && fb_glyphs[fb_i].end <= end {
|
|
let fb_glyph = fb_glyphs.remove(fb_i);
|
|
// log::trace!("Insert {},{} from font {} at {}", fb_glyph.start, fb_glyph.end, font_i, i);
|
|
glyphs.insert(i, fb_glyph);
|
|
i += 1;
|
|
} else {
|
|
break;
|
|
}
|
|
}
|
|
}
|
|
}
|
|
|
|
// Debug missing font fallbacks
|
|
font_iter.check_missing(&line[start_run..end_run]);
|
|
|
|
/*
|
|
for glyph in glyphs.iter() {
|
|
log::trace!("'{}': {}, {}, {}, {}", &line[glyph.start..glyph.end], glyph.x_advance, glyph.y_advance, glyph.x_offset, glyph.y_offset);
|
|
}
|
|
*/
|
|
|
|
// Restore the scripts buffer.
|
|
scratch.scripts = scripts;
|
|
}
|
|
|
|
#[cfg(feature = "swash")]
|
|
fn shape_skip(
|
|
font_system: &mut FontSystem,
|
|
glyphs: &mut Vec<ShapeGlyph>,
|
|
line: &str,
|
|
attrs_list: &AttrsList,
|
|
start_run: usize,
|
|
end_run: usize,
|
|
) {
|
|
let attrs = attrs_list.get_span(start_run);
|
|
let fonts = font_system.get_font_matches(attrs);
|
|
|
|
let default_families = [&attrs.family];
|
|
let mut font_iter = FontFallbackIter::new(font_system, &fonts, &default_families, &[]);
|
|
|
|
let font = font_iter.next().expect("no default font found");
|
|
let font_id = font.id();
|
|
let font = font.as_swash();
|
|
|
|
let charmap = font.charmap();
|
|
let metrics = font.metrics(&[]);
|
|
let glyph_metrics = font.glyph_metrics(&[]).scale(1.0);
|
|
|
|
let ascent = metrics.ascent / f32::from(metrics.units_per_em);
|
|
let descent = metrics.descent / f32::from(metrics.units_per_em);
|
|
|
|
glyphs.extend(
|
|
line[start_run..end_run]
|
|
.chars()
|
|
.enumerate()
|
|
.map(|(i, codepoint)| {
|
|
let glyph_id = charmap.map(codepoint);
|
|
let x_advance = glyph_metrics.advance_width(glyph_id);
|
|
let attrs = attrs_list.get_span(i);
|
|
|
|
ShapeGlyph {
|
|
start: i,
|
|
end: i + 1,
|
|
x_advance,
|
|
y_advance: 0.0,
|
|
x_offset: 0.0,
|
|
y_offset: 0.0,
|
|
ascent,
|
|
descent,
|
|
font_id,
|
|
glyph_id,
|
|
color_opt: attrs.color_opt,
|
|
metadata: attrs.metadata,
|
|
}
|
|
}),
|
|
);
|
|
}
|
|
|
|
/// A shaped glyph
|
|
#[derive(Debug)]
|
|
pub struct ShapeGlyph {
|
|
pub start: usize,
|
|
pub end: usize,
|
|
pub x_advance: f32,
|
|
pub y_advance: f32,
|
|
pub x_offset: f32,
|
|
pub y_offset: f32,
|
|
pub ascent: f32,
|
|
pub descent: f32,
|
|
pub font_id: fontdb::ID,
|
|
pub glyph_id: u16,
|
|
pub color_opt: Option<Color>,
|
|
pub metadata: usize,
|
|
}
|
|
|
|
impl ShapeGlyph {
|
|
fn layout(
|
|
&self,
|
|
font_size: f32,
|
|
x: f32,
|
|
y: f32,
|
|
w: f32,
|
|
level: unicode_bidi::Level,
|
|
) -> LayoutGlyph {
|
|
LayoutGlyph {
|
|
start: self.start,
|
|
end: self.end,
|
|
font_size,
|
|
font_id: self.font_id,
|
|
glyph_id: self.glyph_id,
|
|
x,
|
|
y,
|
|
w,
|
|
level,
|
|
x_offset: self.x_offset,
|
|
y_offset: self.y_offset,
|
|
color_opt: self.color_opt,
|
|
metadata: self.metadata,
|
|
}
|
|
}
|
|
}
|
|
|
|
/// A shaped word (for word wrapping)
|
|
#[derive(Debug)]
|
|
pub struct ShapeWord {
|
|
pub blank: bool,
|
|
pub glyphs: Vec<ShapeGlyph>,
|
|
pub x_advance: f32,
|
|
pub y_advance: f32,
|
|
}
|
|
|
|
impl ShapeWord {
|
|
pub fn new(
|
|
font_system: &mut FontSystem,
|
|
line: &str,
|
|
attrs_list: &AttrsList,
|
|
word_range: Range<usize>,
|
|
level: unicode_bidi::Level,
|
|
blank: bool,
|
|
shaping: Shaping,
|
|
) -> Self {
|
|
Self::new_in_buffer(
|
|
&mut ShapeBuffer::default(),
|
|
font_system,
|
|
line,
|
|
attrs_list,
|
|
word_range,
|
|
level,
|
|
blank,
|
|
shaping,
|
|
)
|
|
}
|
|
|
|
/// Shape a word into a set of glyphs, using a scratch buffer.
|
|
#[allow(clippy::too_many_arguments)]
|
|
pub fn new_in_buffer(
|
|
scratch: &mut ShapeBuffer,
|
|
font_system: &mut FontSystem,
|
|
line: &str,
|
|
attrs_list: &AttrsList,
|
|
word_range: Range<usize>,
|
|
level: unicode_bidi::Level,
|
|
blank: bool,
|
|
shaping: Shaping,
|
|
) -> Self {
|
|
let word = &line[word_range.clone()];
|
|
|
|
log::trace!(
|
|
" Word{}: '{}'",
|
|
if blank { " BLANK" } else { "" },
|
|
word
|
|
);
|
|
|
|
let mut glyphs = Vec::new();
|
|
let span_rtl = level.is_rtl();
|
|
|
|
let mut start_run = word_range.start;
|
|
let mut attrs = attrs_list.defaults();
|
|
for (egc_i, _egc) in word.grapheme_indices(true) {
|
|
let start_egc = word_range.start + egc_i;
|
|
let attrs_egc = attrs_list.get_span(start_egc);
|
|
if !attrs.compatible(&attrs_egc) {
|
|
shaping.run(
|
|
scratch,
|
|
&mut glyphs,
|
|
font_system,
|
|
line,
|
|
attrs_list,
|
|
start_run,
|
|
start_egc,
|
|
span_rtl,
|
|
);
|
|
|
|
start_run = start_egc;
|
|
attrs = attrs_egc;
|
|
}
|
|
}
|
|
if start_run < word_range.end {
|
|
shaping.run(
|
|
scratch,
|
|
&mut glyphs,
|
|
font_system,
|
|
line,
|
|
attrs_list,
|
|
start_run,
|
|
word_range.end,
|
|
span_rtl,
|
|
);
|
|
}
|
|
|
|
let mut x_advance = 0.0;
|
|
let mut y_advance = 0.0;
|
|
for glyph in &glyphs {
|
|
x_advance += glyph.x_advance;
|
|
y_advance += glyph.y_advance;
|
|
}
|
|
|
|
Self {
|
|
blank,
|
|
glyphs,
|
|
x_advance,
|
|
y_advance,
|
|
}
|
|
}
|
|
}
|
|
|
|
/// A shaped span (for bidirectional processing)
|
|
#[derive(Debug)]
|
|
pub struct ShapeSpan {
|
|
pub level: unicode_bidi::Level,
|
|
pub words: Vec<ShapeWord>,
|
|
}
|
|
|
|
impl ShapeSpan {
|
|
pub fn new(
|
|
font_system: &mut FontSystem,
|
|
line: &str,
|
|
attrs_list: &AttrsList,
|
|
span_range: Range<usize>,
|
|
line_rtl: bool,
|
|
level: unicode_bidi::Level,
|
|
shaping: Shaping,
|
|
) -> Self {
|
|
Self::new_in_buffer(
|
|
&mut ShapeBuffer::default(),
|
|
font_system,
|
|
line,
|
|
attrs_list,
|
|
span_range,
|
|
line_rtl,
|
|
level,
|
|
shaping,
|
|
)
|
|
}
|
|
|
|
/// Shape a span into a set of words, using a scratch buffer.
|
|
pub fn new_in_buffer(
|
|
scratch: &mut ShapeBuffer,
|
|
font_system: &mut FontSystem,
|
|
line: &str,
|
|
attrs_list: &AttrsList,
|
|
span_range: Range<usize>,
|
|
line_rtl: bool,
|
|
level: unicode_bidi::Level,
|
|
shaping: Shaping,
|
|
) -> Self {
|
|
let span = &line[span_range.start..span_range.end];
|
|
|
|
log::trace!(
|
|
" Span {}: '{}'",
|
|
if level.is_rtl() { "RTL" } else { "LTR" },
|
|
span
|
|
);
|
|
|
|
let mut words = Vec::new();
|
|
|
|
let mut start_word = 0;
|
|
for (end_lb, _) in unicode_linebreak::linebreaks(span) {
|
|
let mut start_lb = end_lb;
|
|
for (i, c) in span[start_word..end_lb].char_indices().rev() {
|
|
// TODO: Not all whitespace characters are linebreakable, e.g. 00A0 (No-break
|
|
// space)
|
|
// https://www.unicode.org/reports/tr14/#GL
|
|
// https://www.unicode.org/Public/UCD/latest/ucd/PropList.txt
|
|
if c.is_whitespace() {
|
|
start_lb = start_word + i;
|
|
} else {
|
|
break;
|
|
}
|
|
}
|
|
if start_word < start_lb {
|
|
words.push(ShapeWord::new_in_buffer(
|
|
scratch,
|
|
font_system,
|
|
line,
|
|
attrs_list,
|
|
(span_range.start + start_word)..(span_range.start + start_lb),
|
|
level,
|
|
false,
|
|
shaping,
|
|
));
|
|
}
|
|
if start_lb < end_lb {
|
|
for (i, c) in span[start_lb..end_lb].char_indices() {
|
|
// assert!(c.is_whitespace());
|
|
words.push(ShapeWord::new_in_buffer(
|
|
scratch,
|
|
font_system,
|
|
line,
|
|
attrs_list,
|
|
(span_range.start + start_lb + i)
|
|
..(span_range.start + start_lb + i + c.len_utf8()),
|
|
level,
|
|
true,
|
|
shaping,
|
|
));
|
|
}
|
|
}
|
|
start_word = end_lb;
|
|
}
|
|
|
|
// Reverse glyphs in RTL lines
|
|
if line_rtl {
|
|
for word in &mut words {
|
|
word.glyphs.reverse();
|
|
}
|
|
}
|
|
|
|
// Reverse words in spans that do not match line direction
|
|
if line_rtl != level.is_rtl() {
|
|
words.reverse();
|
|
}
|
|
|
|
ShapeSpan { level, words }
|
|
}
|
|
}
|
|
|
|
/// A shaped line (or paragraph)
|
|
#[derive(Debug)]
|
|
pub struct ShapeLine {
|
|
pub rtl: bool,
|
|
pub spans: Vec<ShapeSpan>,
|
|
}
|
|
|
|
// Visual Line Ranges: (span_index, (first_word_index, first_glyph_index), (last_word_index, last_glyph_index))
|
|
type VlRange = (usize, (usize, usize), (usize, usize));
|
|
|
|
#[derive(Default)]
|
|
struct VisualLine {
|
|
ranges: Vec<VlRange>,
|
|
spaces: u32,
|
|
w: f32,
|
|
}
|
|
|
|
impl ShapeLine {
|
|
/// # Panics
|
|
///
|
|
/// Will panic if `line` contains more than one paragraph.
|
|
pub fn new(
|
|
font_system: &mut FontSystem,
|
|
line: &str,
|
|
attrs_list: &AttrsList,
|
|
shaping: Shaping,
|
|
) -> Self {
|
|
Self::new_in_buffer(
|
|
&mut ShapeBuffer::default(),
|
|
font_system,
|
|
line,
|
|
attrs_list,
|
|
shaping,
|
|
)
|
|
}
|
|
|
|
/// Shape a line into a set of spans, using a scratch buffer.
|
|
///
|
|
/// # Panics
|
|
///
|
|
/// Will panic if `line` contains more than one paragraph.
|
|
pub fn new_in_buffer(
|
|
scratch: &mut ShapeBuffer,
|
|
font_system: &mut FontSystem,
|
|
line: &str,
|
|
attrs_list: &AttrsList,
|
|
shaping: Shaping,
|
|
) -> Self {
|
|
let mut spans = Vec::new();
|
|
|
|
let bidi = unicode_bidi::BidiInfo::new(line, None);
|
|
let rtl = if bidi.paragraphs.is_empty() {
|
|
false
|
|
} else {
|
|
assert_eq!(bidi.paragraphs.len(), 1);
|
|
let para_info = &bidi.paragraphs[0];
|
|
let line_rtl = para_info.level.is_rtl();
|
|
|
|
log::trace!("Line {}: '{}'", if line_rtl { "RTL" } else { "LTR" }, line);
|
|
|
|
let line_range = para_info.range.clone();
|
|
let levels = Self::adjust_levels(&unicode_bidi::Paragraph::new(&bidi, para_info));
|
|
|
|
// Find consecutive level runs. We use this to create Spans.
|
|
// Each span is a set of characters with equal levels.
|
|
let mut start = line_range.start;
|
|
let mut run_level = levels[start];
|
|
spans.reserve(line_range.end - start + 1);
|
|
|
|
for (i, &new_level) in levels
|
|
.iter()
|
|
.enumerate()
|
|
.take(line_range.end)
|
|
.skip(start + 1)
|
|
{
|
|
if new_level != run_level {
|
|
// End of the previous run, start of a new one.
|
|
spans.push(ShapeSpan::new_in_buffer(
|
|
scratch,
|
|
font_system,
|
|
line,
|
|
attrs_list,
|
|
start..i,
|
|
line_rtl,
|
|
run_level,
|
|
shaping,
|
|
));
|
|
start = i;
|
|
run_level = new_level;
|
|
}
|
|
}
|
|
spans.push(ShapeSpan::new_in_buffer(
|
|
scratch,
|
|
font_system,
|
|
line,
|
|
attrs_list,
|
|
start..line_range.end,
|
|
line_rtl,
|
|
run_level,
|
|
shaping,
|
|
));
|
|
line_rtl
|
|
};
|
|
|
|
Self { rtl, spans }
|
|
}
|
|
|
|
// A modified version of first part of unicode_bidi::bidi_info::visual_run
|
|
fn adjust_levels(para: &unicode_bidi::Paragraph) -> Vec<unicode_bidi::Level> {
|
|
use unicode_bidi::BidiClass::*;
|
|
let text = para.info.text;
|
|
let levels = ¶.info.levels;
|
|
let original_classes = ¶.info.original_classes;
|
|
|
|
let mut levels = levels.clone();
|
|
let line_classes = &original_classes[..];
|
|
let line_levels = &mut levels[..];
|
|
|
|
// Reset some whitespace chars to paragraph level.
|
|
// <http://www.unicode.org/reports/tr9/#L1>
|
|
let mut reset_from: Option<usize> = Some(0);
|
|
let mut reset_to: Option<usize> = None;
|
|
for (i, c) in text.char_indices() {
|
|
match line_classes[i] {
|
|
// Ignored by X9
|
|
RLE | LRE | RLO | LRO | PDF | BN => {}
|
|
// Segment separator, Paragraph separator
|
|
B | S => {
|
|
assert_eq!(reset_to, None);
|
|
reset_to = Some(i + c.len_utf8());
|
|
if reset_from.is_none() {
|
|
reset_from = Some(i);
|
|
}
|
|
}
|
|
// Whitespace, isolate formatting
|
|
WS | FSI | LRI | RLI | PDI => {
|
|
if reset_from.is_none() {
|
|
reset_from = Some(i);
|
|
}
|
|
}
|
|
_ => {
|
|
reset_from = None;
|
|
}
|
|
}
|
|
if let (Some(from), Some(to)) = (reset_from, reset_to) {
|
|
for level in &mut line_levels[from..to] {
|
|
*level = para.para.level;
|
|
}
|
|
reset_from = None;
|
|
reset_to = None;
|
|
}
|
|
}
|
|
if let Some(from) = reset_from {
|
|
for level in &mut line_levels[from..] {
|
|
*level = para.para.level;
|
|
}
|
|
}
|
|
levels
|
|
}
|
|
|
|
// A modified version of second part of unicode_bidi::bidi_info::visual run
|
|
fn reorder(&self, line_range: &[VlRange]) -> Vec<Range<usize>> {
|
|
let line: Vec<unicode_bidi::Level> = line_range
|
|
.iter()
|
|
.map(|(span_index, _, _)| self.spans[*span_index].level)
|
|
.collect();
|
|
// Find consecutive level runs.
|
|
let mut runs = Vec::new();
|
|
let mut start = 0;
|
|
let mut run_level = line[start];
|
|
let mut min_level = run_level;
|
|
let mut max_level = run_level;
|
|
|
|
for (i, &new_level) in line.iter().enumerate().skip(start + 1) {
|
|
if new_level != run_level {
|
|
// End of the previous run, start of a new one.
|
|
runs.push(start..i);
|
|
start = i;
|
|
run_level = new_level;
|
|
min_level = min(run_level, min_level);
|
|
max_level = max(run_level, max_level);
|
|
}
|
|
}
|
|
runs.push(start..line.len());
|
|
|
|
let run_count = runs.len();
|
|
|
|
// Re-order the odd runs.
|
|
// <http://www.unicode.org/reports/tr9/#L2>
|
|
|
|
// Stop at the lowest *odd* level.
|
|
min_level = min_level.new_lowest_ge_rtl().expect("Level error");
|
|
|
|
while max_level >= min_level {
|
|
// Look for the start of a sequence of consecutive runs of max_level or higher.
|
|
let mut seq_start = 0;
|
|
while seq_start < run_count {
|
|
if line[runs[seq_start].start] < max_level {
|
|
seq_start += 1;
|
|
continue;
|
|
}
|
|
|
|
// Found the start of a sequence. Now find the end.
|
|
let mut seq_end = seq_start + 1;
|
|
while seq_end < run_count {
|
|
if line[runs[seq_end].start] < max_level {
|
|
break;
|
|
}
|
|
seq_end += 1;
|
|
}
|
|
|
|
// Reverse the runs within this sequence.
|
|
runs[seq_start..seq_end].reverse();
|
|
|
|
seq_start = seq_end;
|
|
}
|
|
max_level
|
|
.lower(1)
|
|
.expect("Lowering embedding level below zero");
|
|
}
|
|
|
|
runs
|
|
}
|
|
|
|
pub fn layout(
|
|
&self,
|
|
font_size: f32,
|
|
line_width: f32,
|
|
wrap: Wrap,
|
|
align: Option<Align>,
|
|
) -> Vec<LayoutLine> {
|
|
let mut lines = Vec::with_capacity(1);
|
|
self.layout_to_buffer(
|
|
&mut ShapeBuffer::default(),
|
|
font_size,
|
|
line_width,
|
|
wrap,
|
|
align,
|
|
&mut lines,
|
|
);
|
|
lines
|
|
}
|
|
|
|
pub fn layout_to_buffer(
|
|
&self,
|
|
scratch: &mut ShapeBuffer,
|
|
font_size: f32,
|
|
line_width: f32,
|
|
wrap: Wrap,
|
|
align: Option<Align>,
|
|
layout_lines: &mut Vec<LayoutLine>,
|
|
) {
|
|
// For each visual line a list of (span index, and range of words in that span)
|
|
// Note that a BiDi visual line could have multiple spans or parts of them
|
|
// let mut vl_range_of_spans = Vec::with_capacity(1);
|
|
let mut visual_lines: Vec<VisualLine> = {
|
|
let mut visual_lines = mem::take(&mut scratch.visual_lines);
|
|
visual_lines.clear();
|
|
visual_lines
|
|
};
|
|
|
|
fn add_to_visual_line(
|
|
vl: &mut VisualLine,
|
|
span_index: usize,
|
|
start: (usize, usize),
|
|
end: (usize, usize),
|
|
width: f32,
|
|
number_of_blanks: u32,
|
|
) {
|
|
if end == start {
|
|
return;
|
|
}
|
|
|
|
vl.ranges.push((span_index, start, end));
|
|
vl.w += width;
|
|
vl.spaces += number_of_blanks;
|
|
}
|
|
|
|
// This would keep the maximum number of spans that would fit on a visual line
|
|
// If one span is too large, this variable will hold the range of words inside that span
|
|
// that fits on a line.
|
|
// let mut current_visual_line: Vec<VlRange> = Vec::with_capacity(1);
|
|
let mut current_visual_line = VisualLine::default();
|
|
|
|
if wrap == Wrap::None {
|
|
for (span_index, span) in self.spans.iter().enumerate() {
|
|
let mut word_range_width = 0.;
|
|
let mut number_of_blanks: u32 = 0;
|
|
for word in span.words.iter() {
|
|
let word_width = font_size * word.x_advance;
|
|
word_range_width += word_width;
|
|
if word.blank {
|
|
number_of_blanks += 1;
|
|
}
|
|
}
|
|
add_to_visual_line(
|
|
&mut current_visual_line,
|
|
span_index,
|
|
(0, 0),
|
|
(span.words.len(), 0),
|
|
word_range_width,
|
|
number_of_blanks,
|
|
);
|
|
}
|
|
} else {
|
|
for (span_index, span) in self.spans.iter().enumerate() {
|
|
let mut word_range_width = 0.;
|
|
let mut width_before_last_blank = 0.;
|
|
let mut number_of_blanks: u32 = 0;
|
|
|
|
// Create the word ranges that fits in a visual line
|
|
if self.rtl != span.level.is_rtl() {
|
|
// incongruent directions
|
|
let mut fitting_start = (span.words.len(), 0);
|
|
for (i, word) in span.words.iter().enumerate().rev() {
|
|
let word_width = font_size * word.x_advance;
|
|
|
|
// Addition in the same order used to compute the final width, so that
|
|
// relayouts with that width as the `line_width` will produce the same
|
|
// wrapping results.
|
|
if current_visual_line.w + (word_range_width + word_width)
|
|
<= line_width
|
|
// Include one blank word over the width limit since it won't be
|
|
// counted in the final width
|
|
|| (word.blank
|
|
&& (current_visual_line.w + word_range_width) <= line_width)
|
|
{
|
|
// fits
|
|
if word.blank {
|
|
number_of_blanks += 1;
|
|
width_before_last_blank = word_range_width;
|
|
}
|
|
word_range_width += word_width;
|
|
continue;
|
|
} else if wrap == Wrap::Glyph {
|
|
for (glyph_i, glyph) in word.glyphs.iter().enumerate().rev() {
|
|
let glyph_width = font_size * glyph.x_advance;
|
|
if current_visual_line.w + (word_range_width + glyph_width)
|
|
<= line_width
|
|
{
|
|
word_range_width += glyph_width;
|
|
continue;
|
|
} else {
|
|
add_to_visual_line(
|
|
&mut current_visual_line,
|
|
span_index,
|
|
(i, glyph_i + 1),
|
|
fitting_start,
|
|
word_range_width,
|
|
number_of_blanks,
|
|
);
|
|
visual_lines.push(current_visual_line);
|
|
current_visual_line = VisualLine::default();
|
|
|
|
number_of_blanks = 0;
|
|
word_range_width = glyph_width;
|
|
fitting_start = (i, glyph_i + 1);
|
|
}
|
|
}
|
|
} else {
|
|
// Wrap::Word
|
|
|
|
// TODO: What if the previous span ended with whitespace and the next
|
|
// span wraps a new line? Is that possible?
|
|
//
|
|
// TODO: This comment it outdated, the current word can be a
|
|
// whitespace.
|
|
//
|
|
// Current word causing a wrap is not whitespace, so we ignore the
|
|
// previous word if it's a whitespace
|
|
let trailing_blank = span
|
|
.words
|
|
.get(i + 1)
|
|
.map_or(false, |previous_word| previous_word.blank);
|
|
if trailing_blank {
|
|
number_of_blanks = number_of_blanks.saturating_sub(1);
|
|
add_to_visual_line(
|
|
&mut current_visual_line,
|
|
span_index,
|
|
(i + 2, 0),
|
|
fitting_start,
|
|
width_before_last_blank,
|
|
number_of_blanks,
|
|
);
|
|
} else {
|
|
add_to_visual_line(
|
|
&mut current_visual_line,
|
|
span_index,
|
|
(i + 1, 0),
|
|
fitting_start,
|
|
word_range_width,
|
|
number_of_blanks,
|
|
);
|
|
}
|
|
visual_lines.push(current_visual_line);
|
|
current_visual_line = VisualLine::default();
|
|
|
|
number_of_blanks = 0;
|
|
if word.blank {
|
|
word_range_width = 0.;
|
|
fitting_start = (i, 0);
|
|
} else {
|
|
word_range_width = word_width;
|
|
fitting_start = (i + 1, 0);
|
|
}
|
|
}
|
|
}
|
|
add_to_visual_line(
|
|
&mut current_visual_line,
|
|
span_index,
|
|
(0, 0),
|
|
fitting_start,
|
|
word_range_width,
|
|
number_of_blanks,
|
|
);
|
|
} else {
|
|
// congruent direction
|
|
let mut fitting_start = (0, 0);
|
|
for (i, word) in span.words.iter().enumerate() {
|
|
let word_width = font_size * word.x_advance;
|
|
if current_visual_line.w + (word_range_width + word_width)
|
|
<= line_width
|
|
// Include one blank word over the width limit since it won't be
|
|
// counted in the final width.
|
|
|| (word.blank
|
|
&& (current_visual_line.w + word_range_width) <= line_width)
|
|
{
|
|
// fits
|
|
if word.blank {
|
|
number_of_blanks += 1;
|
|
width_before_last_blank = word_range_width;
|
|
}
|
|
word_range_width += word_width;
|
|
continue;
|
|
} else if wrap == Wrap::Glyph {
|
|
for (glyph_i, glyph) in word.glyphs.iter().enumerate() {
|
|
let glyph_width = font_size * glyph.x_advance;
|
|
if current_visual_line.w + (word_range_width + glyph_width)
|
|
<= line_width
|
|
{
|
|
word_range_width += glyph_width;
|
|
continue;
|
|
} else {
|
|
add_to_visual_line(
|
|
&mut current_visual_line,
|
|
span_index,
|
|
fitting_start,
|
|
(i, glyph_i),
|
|
word_range_width,
|
|
number_of_blanks,
|
|
);
|
|
visual_lines.push(current_visual_line);
|
|
current_visual_line = VisualLine::default();
|
|
|
|
number_of_blanks = 0;
|
|
word_range_width = glyph_width;
|
|
fitting_start = (i, glyph_i);
|
|
}
|
|
}
|
|
} else {
|
|
// Wrap::Word
|
|
|
|
// Current word causing a wrap is not whitespace, so we ignore the
|
|
// previous word if it's a whitespace
|
|
let trailing_blank = i > 0 && span.words[i - 1].blank;
|
|
if trailing_blank {
|
|
number_of_blanks = number_of_blanks.saturating_sub(1);
|
|
add_to_visual_line(
|
|
&mut current_visual_line,
|
|
span_index,
|
|
fitting_start,
|
|
(i - 1, 0),
|
|
width_before_last_blank,
|
|
number_of_blanks,
|
|
);
|
|
} else {
|
|
add_to_visual_line(
|
|
&mut current_visual_line,
|
|
span_index,
|
|
fitting_start,
|
|
(i, 0),
|
|
word_range_width,
|
|
number_of_blanks,
|
|
);
|
|
}
|
|
visual_lines.push(current_visual_line);
|
|
current_visual_line = VisualLine::default();
|
|
number_of_blanks = 0;
|
|
|
|
if word.blank {
|
|
word_range_width = 0.;
|
|
fitting_start = (i + 1, 0);
|
|
} else {
|
|
word_range_width = word_width;
|
|
fitting_start = (i, 0);
|
|
}
|
|
}
|
|
}
|
|
add_to_visual_line(
|
|
&mut current_visual_line,
|
|
span_index,
|
|
fitting_start,
|
|
(span.words.len(), 0),
|
|
word_range_width,
|
|
number_of_blanks,
|
|
);
|
|
}
|
|
}
|
|
}
|
|
|
|
if !current_visual_line.ranges.is_empty() {
|
|
visual_lines.push(current_visual_line);
|
|
}
|
|
|
|
// Create the LayoutLines using the ranges inside visual lines
|
|
let align = align.unwrap_or({
|
|
if self.rtl {
|
|
Align::Right
|
|
} else {
|
|
Align::Left
|
|
}
|
|
});
|
|
|
|
let start_x = if self.rtl { line_width } else { 0.0 };
|
|
|
|
let number_of_visual_lines = visual_lines.len();
|
|
for (index, visual_line) in visual_lines.iter().enumerate() {
|
|
if visual_line.ranges.is_empty() {
|
|
continue;
|
|
}
|
|
let new_order = self.reorder(&visual_line.ranges);
|
|
let mut glyphs = Vec::with_capacity(1);
|
|
let mut x = start_x;
|
|
let mut y = 0.;
|
|
let mut max_ascent: f32 = 0.;
|
|
let mut max_descent: f32 = 0.;
|
|
let alignment_correction = match (align, self.rtl) {
|
|
(Align::Left, true) => line_width - visual_line.w,
|
|
(Align::Left, false) => 0.,
|
|
(Align::Right, true) => 0.,
|
|
(Align::Right, false) => line_width - visual_line.w,
|
|
(Align::Center, _) => (line_width - visual_line.w) / 2.0,
|
|
(Align::End, _) => line_width - visual_line.w,
|
|
(Align::Justified, _) => 0.,
|
|
};
|
|
|
|
if self.rtl {
|
|
x -= alignment_correction;
|
|
} else {
|
|
x += alignment_correction;
|
|
}
|
|
|
|
// TODO: Only certain `is_whitespace` chars are typically expanded but this is what is
|
|
// currently used to compute `visual_line.spaces`.
|
|
//
|
|
// https://www.unicode.org/reports/tr14/#Introduction
|
|
// > When expanding or compressing interword space according to common
|
|
// > typographical practice, only the spaces marked by U+0020 SPACE and U+00A0
|
|
// > NO-BREAK SPACE are subject to compression, and only spaces marked by U+0020
|
|
// > SPACE, U+00A0 NO-BREAK SPACE, and occasionally spaces marked by U+2009 THIN
|
|
// > SPACE are subject to expansion. All other space characters normally have
|
|
// > fixed width.
|
|
//
|
|
// (also some spaces aren't followed by potential linebreaks but they could
|
|
// still be expanded)
|
|
|
|
// Amount of extra width added to each blank space within a line.
|
|
let justification_expansion = if matches!(align, Align::Justified)
|
|
&& visual_line.spaces > 0
|
|
// Don't justify the last line in a paragraph.
|
|
&& index != number_of_visual_lines - 1
|
|
{
|
|
(line_width - visual_line.w) / visual_line.spaces as f32
|
|
} else {
|
|
0.
|
|
};
|
|
|
|
let mut process_range = |range: Range<usize>| {
|
|
for &(span_index, (starting_word, starting_glyph), (ending_word, ending_glyph)) in
|
|
visual_line.ranges[range.clone()].iter()
|
|
{
|
|
let span = &self.spans[span_index];
|
|
// If ending_glyph is not 0 we need to include glyphs from the ending_word
|
|
for i in starting_word..ending_word + usize::from(ending_glyph != 0) {
|
|
let word = &span.words[i];
|
|
let included_glyphs = match (i == starting_word, i == ending_word) {
|
|
(false, false) => &word.glyphs[..],
|
|
(true, false) => &word.glyphs[starting_glyph..],
|
|
(false, true) => &word.glyphs[..ending_glyph],
|
|
(true, true) => &word.glyphs[starting_glyph..ending_glyph],
|
|
};
|
|
for glyph in included_glyphs {
|
|
let x_advance = font_size * glyph.x_advance
|
|
+ if word.blank {
|
|
justification_expansion
|
|
} else {
|
|
0.0
|
|
};
|
|
if self.rtl {
|
|
x -= x_advance;
|
|
}
|
|
let y_advance = font_size * glyph.y_advance;
|
|
glyphs.push(glyph.layout(font_size, x, y, x_advance, span.level));
|
|
if !self.rtl {
|
|
x += x_advance;
|
|
}
|
|
y += y_advance;
|
|
max_ascent = max_ascent.max(glyph.ascent);
|
|
max_descent = max_descent.max(glyph.descent);
|
|
}
|
|
}
|
|
}
|
|
};
|
|
|
|
if self.rtl {
|
|
for range in new_order.into_iter().rev() {
|
|
process_range(range);
|
|
}
|
|
} else {
|
|
/* LTR */
|
|
for range in new_order {
|
|
process_range(range);
|
|
}
|
|
}
|
|
|
|
layout_lines.push(LayoutLine {
|
|
w: if align != Align::Justified {
|
|
visual_line.w
|
|
} else if self.rtl {
|
|
start_x - x
|
|
} else {
|
|
x
|
|
},
|
|
max_ascent: max_ascent * font_size,
|
|
max_descent: max_descent * font_size,
|
|
glyphs,
|
|
});
|
|
}
|
|
|
|
// This is used to create a visual line for empty lines (e.g. lines with only a <CR>)
|
|
if layout_lines.is_empty() {
|
|
layout_lines.push(LayoutLine {
|
|
w: 0.0,
|
|
max_ascent: 0.0,
|
|
max_descent: 0.0,
|
|
glyphs: Default::default(),
|
|
});
|
|
}
|
|
|
|
// Restore the buffer to the scratch set to prevent reallocations.
|
|
scratch.visual_lines = visual_lines;
|
|
}
|
|
}
|