mirrors/coreutils
1
0
Fork 0
mirror of https://github.com/uutils/coreutils.git synced 2025-07-07 21:45:01 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 5

fmt: fix the last two gnu issues

Browse source
This commit is contained in:
Sylvestre Ledru 2025-06-22 23:07:36 +02:00
parent 56ce0e28ad
commit 2df7f01c15
3 changed files with 95 additions and 10 deletions
Download patch file Download diff file Expand all files Collapse all files

10
src/uu/fmt/src/linebreak.rs
View file

@ -236,7 +236,11 @@ fn find_kp_breakpoints<'a, T: Iterator<Item = &'a WordInfo<'a>>>(
let mut next_active_breaks = vec![];
let stretch = args.opts.width - args.opts.goal;
let minlength = args.opts.goal.max(stretch + 1) - stretch;
let minlength = if args.opts.goal <= 10 {
1
} else {
args.opts.goal.max(stretch + 1) - stretch
};
let mut new_linebreaks = vec![];
let mut is_sentence_start = false;
let mut least_demerits = 0;
@ -384,11 +388,11 @@ fn build_best_path<'a>(paths: &[LineBreak<'a>], active: &[usize]) -> Vec<(&'a Wo
const BAD_INFTY: i64 = 10_000_000;
const BAD_INFTY_SQ: i64 = BAD_INFTY * BAD_INFTY;
// badness = BAD_MULT * abs(r) ^ 3
const BAD_MULT: f32 = 100.0;
const BAD_MULT: f32 = 200.0;
// DR_MULT is multiplier for delta-R between lines
const DR_MULT: f32 = 600.0;
// DL_MULT is penalty multiplier for short words at end of line
const DL_MULT: f32 = 300.0;
const DL_MULT: f32 = 10.0;
fn compute_demerits(delta_len: isize, stretch: usize, wlen: usize, prev_rat: f32) -> (i64, f32) {
// how much stretch are we using?

22
src/uu/fmt/src/parasplit.rs
View file

@ -26,6 +26,14 @@ fn char_width(c: char) -> usize {
}
}
// GNU fmt has a more restrictive definition of whitespace than Unicode.
// It only considers ASCII whitespace characters (space, tab, newline, etc.)
// and excludes many Unicode whitespace characters like non-breaking spaces.
fn is_fmt_whitespace(c: char) -> bool {
// Only ASCII whitespace characters are considered whitespace in GNU fmt
matches!(c, ' ' | '\t' | '\n' | '\r' | '\x0B' | '\x0C')
}
// lines with PSKIP, lacking PREFIX, or which are entirely blank are
// NoFormatLines; otherwise, they are FormatLines
#[derive(Debug)]
@ -109,7 +117,7 @@ impl FileLines<'_> {
for (i, char) in line.char_indices() {
if line[i..].starts_with(pfx) {
return (true, i);
} else if !char.is_whitespace() {
} else if !is_fmt_whitespace(char) {
break;
}
}
@ -128,7 +136,7 @@ impl FileLines<'_> {
prefix_len = indent_len;
}
if (os >= prefix_end) && !c.is_whitespace() {
if (os >= prefix_end) && !is_fmt_whitespace(c) {
// found first non-whitespace after prefix, this is indent_end
indent_end = os;
break;
@ -154,7 +162,7 @@ impl Iterator for FileLines<'_> {
// emit a blank line
// Err(true) indicates that this was a linebreak,
// which is important to know when detecting mail headers
if n.chars().all(char::is_whitespace) {
if n.chars().all(is_fmt_whitespace) {
return Some(Line::NoFormatLine(String::new(), true));
}
@ -174,7 +182,7 @@ impl Iterator for FileLines<'_> {
if pmatch
&& n[poffset + self.opts.prefix.as_ref().map_or(0, |s| s.len())..]
.chars()
.all(char::is_whitespace)
.all(is_fmt_whitespace)
{
return Some(Line::NoFormatLine(n, false));
}
@ -498,7 +506,7 @@ impl WordSplit<'_> {
let mut aftertab = 0;
let mut word_start = None;
for (os, c) in string.char_indices() {
if !c.is_whitespace() {
if !is_fmt_whitespace(c) {
word_start = Some(os);
break;
} else if c == '\t' {
@ -519,7 +527,7 @@ impl WordSplit<'_> {
impl WordSplit<'_> {
fn new<'b>(opts: &'b FmtOptions, string: &'b str) -> WordSplit<'b> {
// wordsplits *must* start at a non-whitespace character
let trim_string = string.trim_start();
let trim_string = string.trim_start_matches(is_fmt_whitespace);
WordSplit {
opts,
string: trim_string,
@ -571,7 +579,7 @@ impl<'a> Iterator for WordSplit<'a> {
// points to whitespace character OR end of string
let mut word_nchars = 0;
self.position = match self.string[word_start..].find(|x: char| {
if x.is_whitespace() {
if is_fmt_whitespace(x) {
true
} else {
word_nchars += char_width(x);