caisin/typst
1
0
Fork 0
mirror of https://github.com/typst/typst synced 2025-05-19 11:35:27 +08:00
Code Issues Packages Projects Releases Wiki Activity

Don't hyphenate on no-break characters (#2396)

Browse Source
This commit is contained in:
Eric Biedert 2023-10-22 13:36:33 +02:00 committed by GitHub
parent aaac1dbd68
commit 3ed6462ee0
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 18 additions and 8 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

26
crates/typst-library/src/layout/par.rs
View File

@ -1109,7 +1109,7 @@ static SEGMENTER: Lazy<LineSegmenter> = Lazy::new(|| {
LineSegmenter::try_new_lstm_with_buffer_provider(&provider).unwrap() LineSegmenter::try_new_lstm_with_buffer_provider(&provider).unwrap()
}); });
/// The Unicode line break properties for each code point. /// The line break segmenter for Chinese/Japanese text.
static CJ_SEGMENTER: Lazy<LineSegmenter> = Lazy::new(|| { static CJ_SEGMENTER: Lazy<LineSegmenter> = Lazy::new(|| {
let provider = BlobDataProvider::try_new_from_static_blob(ICU_DATA).unwrap(); let provider = BlobDataProvider::try_new_from_static_blob(ICU_DATA).unwrap();
let cj_blob = BlobDataProvider::try_new_from_static_blob(CJ_LINEBREAK_DATA).unwrap(); let cj_blob = BlobDataProvider::try_new_from_static_blob(CJ_LINEBREAK_DATA).unwrap();
@ -1117,7 +1117,7 @@ static CJ_SEGMENTER: Lazy<LineSegmenter> = Lazy::new(|| {
LineSegmenter::try_new_lstm_with_buffer_provider(&cj_provider).unwrap() LineSegmenter::try_new_lstm_with_buffer_provider(&cj_provider).unwrap()
}); });
/// The line break segmenter for Chinese/Jpanese text. /// The Unicode line break properties for each code point.
static LINEBREAK_DATA: Lazy<CodePointMapData<LineBreak>> = Lazy::new(|| { static LINEBREAK_DATA: Lazy<CodePointMapData<LineBreak>> = Lazy::new(|| {
let provider = BlobDataProvider::try_new_from_static_blob(ICU_DATA).unwrap(); let provider = BlobDataProvider::try_new_from_static_blob(ICU_DATA).unwrap();
let deser_provider = provider.as_deserializing(); let deser_provider = provider.as_deserializing();
@ -1170,6 +1170,8 @@ impl Iterator for Breakpoints<'_> {
type Item = (usize, bool, bool); type Item = (usize, bool, bool);
fn next(&mut self) -> Option<Self::Item> { fn next(&mut self) -> Option<Self::Item> {
let lb = LINEBREAK_DATA.as_borrowed();
// If we're currently in a hyphenated "word", process the next syllable. // If we're currently in a hyphenated "word", process the next syllable.
if let Some(syllable) = self.syllables.as_mut().and_then(Iterator::next) { if let Some(syllable) = self.syllables.as_mut().and_then(Iterator::next) {
self.offset += syllable.len(); self.offset += syllable.len();
@ -1177,18 +1179,26 @@ impl Iterator for Breakpoints<'_> {
self.offset = self.end; self.offset = self.end;
} }
// Filter out hyphenation opportunities where hyphenation was
// actually disabled.
let hyphen = self.offset < self.end; let hyphen = self.offset < self.end;
if hyphen && !self.hyphenate(self.offset) { if hyphen {
return self.next(); // Filter out hyphenation opportunities where hyphenation was
// actually disabled.
if !self.hyphenate(self.offset) {
return self.next();
}
// Filter out forbidden hyphenation opportunities.
if matches!(
syllable.chars().last().map(|c| lb.get(c)),
Some(LineBreak::Glue | LineBreak::WordJoiner | LineBreak::ZWJ)
) {
return self.next();
}
} }
return Some((self.offset, self.mandatory && !hyphen, hyphen)); return Some((self.offset, self.mandatory && !hyphen, hyphen));
} }
let lb = LINEBREAK_DATA.as_borrowed();
loop { loop {
// Get the next "word". // Get the next "word".
self.end = self.linebreaks.next()?; self.end = self.linebreaks.next()?;