Skip to content

join: add support for multibyte separators #6736

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
sylvestre merged 4 commits into from
Oct 6, 2024
Merged

join: add support for multibyte separators #6736

Changes from 1 commit
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
395c441
join: add test for multibyte separators
jtracey Jul 22, 2024
2e96f64
join: implement support for multibyte separators
jtracey Jul 22, 2024
c6cff91
join: use a trait instead of an enum for separator
jtracey Sep 25, 2024
879e1ea
join: test whitespace merging
jtracey Sep 27, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
Prev Previous commit
Next Next commit
join: implement support for multibyte separators
  • Loading branch information
@jtracey
jtracey committed Sep 25, 2024
commit 2e96f64a9d29e527f07b0b3e033f93afc88ca986
113 changes: 73 additions & 40 deletions src/uu/join/src/join.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -7,7 +7,7 @@

use clap::builder::ValueParser;
use clap::{crate_version, Arg, ArgAction, Command};
use memchr::{memchr3_iter, memchr_iter};
use memchr::{memchr3_iter, memchr_iter, memmem::Finder};
use std::cmp::Ordering;
use std::error::Error;
use std::ffi::OsString;
Expand All @@ -17,6 +17,7 @@ use std::io::{stdin, stdout, BufRead, BufReader, BufWriter, Split, Stdin, Write}
use std::num::IntErrorKind;
#[cfg(unix)]
use std::os::unix::ffi::OsStrExt;
use std::rc::Rc;
use uucore::display::Quotable;
use uucore::error::{set_exit_code, FromIo, UError, UResult, USimpleError};
use uucore::line_ending::LineEnding;
Expand Down Expand Up @@ -60,9 +61,10 @@ enum FileNum {
File2,
}

#[derive(Copy, Clone, PartialEq)]
#[derive(Clone)]
enum Sep {
Char(u8),
Byte(u8),
Char(Rc<Finder<'static>>),
Line,
Whitespaces,
}
Expand Down Expand Up @@ -113,13 +115,18 @@ impl Default for Settings {
/// Output representation.
struct Repr<'a> {
line_ending: LineEnding,
separator: u8,
separator: Vec<u8>,
format: &'a [Spec],
empty: &'a [u8],
}

impl<'a> Repr<'a> {
fn new(line_ending: LineEnding, separator: u8, format: &'a [Spec], empty: &'a [u8]) -> Self {
fn new(line_ending: LineEnding, separator: Sep, format: &'a [Spec], empty: &'a [u8]) -> Self {
let separator = match separator {
Sep::Byte(c) => vec![c],
Sep::Char(f) => f.needle().into(),
_ => vec![b' '],
};
Repr {
line_ending,
separator,
Expand Down Expand Up @@ -155,7 +162,7 @@ impl<'a> Repr<'a> {
) -> Result<(), std::io::Error> {
for i in 0..line.field_ranges.len() {
if i != index {
writer.write_all(&[self.separator])?;
writer.write_all(&self.separator)?;
writer.write_all(line.get_field(i).unwrap())?;
}
}
Expand All @@ -169,7 +176,7 @@ impl<'a> Repr<'a> {
{
for i in 0..self.format.len() {
if i > 0 {
writer.write_all(&[self.separator])?;
writer.write_all(&self.separator)?;
}

let field = match f(&self.format[i]) {
Expand Down Expand Up @@ -274,19 +281,30 @@ impl Line {
fn new(string: Vec<u8>, separator: Sep, len_guess: usize) -> Self {
let mut field_ranges = Vec::with_capacity(len_guess);
let mut last_end = 0;
if separator == Sep::Whitespaces {
// GNU join uses Bourne shell field splitters by default
for i in memchr3_iter(b' ', b'\t', b'\n', &string) {
if i > last_end {
match separator {
Sep::Whitespaces => {
// GNU join used Bourne shell field splitters by default
// FIXME: but now uses locale-dependent whitespace
for i in memchr3_iter(b' ', b'\t', b'\n', &string) {
if i > last_end {
field_ranges.push((last_end, i));
}
last_end = i + 1;
}
}
Sep::Byte(sep) => {
for i in memchr_iter(sep, &string) {
field_ranges.push((last_end, i));
last_end = i + 1;
}
last_end = i + 1;
}
} else if let Sep::Char(sep) = separator {
for i in memchr_iter(sep, &string) {
field_ranges.push((last_end, i));
last_end = i + 1;
Sep::Char(finder) => {
for i in finder.find_iter(&string) {
field_ranges.push((last_end, i));
last_end = i + finder.needle().len();
}
}
Sep::Line => (),
}
field_ranges.push((last_end, string.len()));

Expand Down Expand Up @@ -445,7 +463,7 @@ impl<'a> State<'a> {
}

fn reset_read_line(&mut self, input: &Input) -> Result<(), std::io::Error> {
let line = self.read_line(input.separator)?;
let line = self.read_line(input.separator.clone())?;
self.reset(line);
Ok(())
}
Expand Down Expand Up @@ -507,7 +525,7 @@ impl<'a> State<'a> {

/// Get the next line with the order check.
fn next_line(&mut self, input: &Input) -> Result<Option<Line>, JoinError> {
if let Some(line) = self.read_line(input.separator)? {
if let Some(line) = self.read_line(input.separator.clone())? {
if input.check_order == CheckOrder::Disabled {
return Ok(Some(line));
}
Expand Down Expand Up @@ -574,25 +592,43 @@ impl<'a> State<'a> {
}

fn parse_separator(value_os: &OsString) -> UResult<Sep> {
// Five possible separator values:
// No argument supplied, separate on whitespace; handled implicitly as the default elsewhere
// An empty string arg, whole line sparation
// On unix-likes only, a single arbitrary byte
// The two-character "\0" string, interpreted as a single 0 byte
// A single scalar valid in the locale encoding (currently only UTF-8)

if value_os.is_empty() {
return Ok(Sep::Line);
}

#[cfg(unix)]
let value = value_os.as_bytes();
#[cfg(not(unix))]
let value = match value_os.to_str() {
Some(value) => value.as_bytes(),
None => {
return Err(USimpleError::new(
1,
"unprintable field separators are only supported on unix-like platforms",
));
{
let value = value_os.as_bytes();
if value.len() == 1 {
return Ok(Sep::Byte(value[0]));
}
}

let Some(value) = value_os.to_str() else {
#[cfg(unix)]
return Err(USimpleError::new(1, "non-UTF-8 multi-byte tab"));
#[cfg(not(unix))]
return Err(USimpleError::new(
1,
"unprintable field separators are only supported on unix-like platforms",
));
};
match value.len() {
0 => Ok(Sep::Line),
1 => Ok(Sep::Char(value[0])),
2 if value[0] == b'\\' && value[1] == b'0' => Ok(Sep::Char(0)),

let mut chars = value.chars();
let c = chars.next().expect("valid string with at least one byte");
match chars.next() {
None => Ok(Sep::Char(Finder::new(value).into_owned().into())),
Some('0') if c == '\\' => Ok(Sep::Byte(0)),
_ => Err(USimpleError::new(
1,
format!("multi-character tab {}", value_os.to_string_lossy()),
format!("multi-character tab {}", value),
)),
}
}
Expand Down Expand Up @@ -838,15 +874,15 @@ fn exec(file1: &str, file2: &str, settings: Settings) -> UResult<()> {
)?;

let input = Input::new(
settings.separator,
settings.separator.clone(),
settings.ignore_case,
settings.check_order,
);

let format = if settings.autoformat {
let mut format = vec![Spec::Key];
let mut initialize = |state: &mut State| {
let max_fields = state.initialize(settings.separator, settings.autoformat);
let max_fields = state.initialize(settings.separator.clone(), settings.autoformat);
for i in 0..max_fields {
if i != state.key {
format.push(Spec::Field(state.file_num, i));
Expand All @@ -857,17 +893,14 @@ fn exec(file1: &str, file2: &str, settings: Settings) -> UResult<()> {
initialize(&mut state2);
format
} else {
state1.initialize(settings.separator, settings.autoformat);
state2.initialize(settings.separator, settings.autoformat);
state1.initialize(settings.separator.clone(), settings.autoformat);
state2.initialize(settings.separator.clone(), settings.autoformat);
settings.format
};

let repr = Repr::new(
settings.line_ending,
match settings.separator {
Sep::Char(sep) => sep,
_ => b' ',
},
settings.separator,
&format,
&settings.empty,
);
Expand Down