Put esoteric line endings behind a feature flag

This commit is contained in:
Blaž Hrastnik 2022-03-16 18:02:37 +09:00
parent b4a282fd36
commit f67e1ee172
No known key found for this signature in database
GPG key ID: 1238B9C4AD889640
3 changed files with 58 additions and 8 deletions

View file

@ -17,7 +17,7 @@ unicode-lines = ["ropey/unicode_lines"]
helix-loader = { version = "0.6", path = "../helix-loader" } helix-loader = { version = "0.6", path = "../helix-loader" }
# ropey = "1.3" # ropey = "1.3"
ropey = { git = "https://github.com/cessen/ropey", default-features = false, features = ["cr_lines"] } ropey = { git = "https://github.com/cessen/ropey", default-features = false }
smallvec = "1.8" smallvec = "1.8"
smartstring = "1.0.0" smartstring = "1.0.0"
unicode-segmentation = "1.9" unicode-segmentation = "1.9"

View file

@ -10,12 +10,18 @@ pub const DEFAULT_LINE_ENDING: LineEnding = LineEnding::LF;
pub enum LineEnding { pub enum LineEnding {
Crlf, // CarriageReturn followed by LineFeed Crlf, // CarriageReturn followed by LineFeed
LF, // U+000A -- LineFeed LF, // U+000A -- LineFeed
VT, // U+000B -- VerticalTab #[cfg(feature = "unicode-lines")]
FF, // U+000C -- FormFeed VT, // U+000B -- VerticalTab
CR, // U+000D -- CarriageReturn #[cfg(feature = "unicode-lines")]
Nel, // U+0085 -- NextLine FF, // U+000C -- FormFeed
LS, // U+2028 -- Line Separator #[cfg(feature = "unicode-lines")]
PS, // U+2029 -- ParagraphSeparator CR, // U+000D -- CarriageReturn
#[cfg(feature = "unicode-lines")]
Nel, // U+0085 -- NextLine
#[cfg(feature = "unicode-lines")]
LS, // U+2028 -- Line Separator
#[cfg(feature = "unicode-lines")]
PS, // U+2029 -- ParagraphSeparator
} }
impl LineEnding { impl LineEnding {
@ -32,11 +38,17 @@ impl LineEnding {
match self { match self {
Self::Crlf => "\u{000D}\u{000A}", Self::Crlf => "\u{000D}\u{000A}",
Self::LF => "\u{000A}", Self::LF => "\u{000A}",
#[cfg(feature = "unicode-lines")]
Self::VT => "\u{000B}", Self::VT => "\u{000B}",
#[cfg(feature = "unicode-lines")]
Self::FF => "\u{000C}", Self::FF => "\u{000C}",
#[cfg(feature = "unicode-lines")]
Self::CR => "\u{000D}", Self::CR => "\u{000D}",
#[cfg(feature = "unicode-lines")]
Self::Nel => "\u{0085}", Self::Nel => "\u{0085}",
#[cfg(feature = "unicode-lines")]
Self::LS => "\u{2028}", Self::LS => "\u{2028}",
#[cfg(feature = "unicode-lines")]
Self::PS => "\u{2029}", Self::PS => "\u{2029}",
} }
} }
@ -45,11 +57,17 @@ impl LineEnding {
pub const fn from_char(ch: char) -> Option<LineEnding> { pub const fn from_char(ch: char) -> Option<LineEnding> {
match ch { match ch {
'\u{000A}' => Some(LineEnding::LF), '\u{000A}' => Some(LineEnding::LF),
#[cfg(feature = "unicode-lines")]
'\u{000B}' => Some(LineEnding::VT), '\u{000B}' => Some(LineEnding::VT),
#[cfg(feature = "unicode-lines")]
'\u{000C}' => Some(LineEnding::FF), '\u{000C}' => Some(LineEnding::FF),
#[cfg(feature = "unicode-lines")]
'\u{000D}' => Some(LineEnding::CR), '\u{000D}' => Some(LineEnding::CR),
#[cfg(feature = "unicode-lines")]
'\u{0085}' => Some(LineEnding::Nel), '\u{0085}' => Some(LineEnding::Nel),
#[cfg(feature = "unicode-lines")]
'\u{2028}' => Some(LineEnding::LS), '\u{2028}' => Some(LineEnding::LS),
#[cfg(feature = "unicode-lines")]
'\u{2029}' => Some(LineEnding::PS), '\u{2029}' => Some(LineEnding::PS),
// Not a line ending // Not a line ending
_ => None, _ => None,
@ -65,11 +83,17 @@ impl LineEnding {
match g { match g {
"\u{000D}\u{000A}" => Some(LineEnding::Crlf), "\u{000D}\u{000A}" => Some(LineEnding::Crlf),
"\u{000A}" => Some(LineEnding::LF), "\u{000A}" => Some(LineEnding::LF),
#[cfg(feature = "unicode-lines")]
"\u{000B}" => Some(LineEnding::VT), "\u{000B}" => Some(LineEnding::VT),
#[cfg(feature = "unicode-lines")]
"\u{000C}" => Some(LineEnding::FF), "\u{000C}" => Some(LineEnding::FF),
#[cfg(feature = "unicode-lines")]
"\u{000D}" => Some(LineEnding::CR), "\u{000D}" => Some(LineEnding::CR),
#[cfg(feature = "unicode-lines")]
"\u{0085}" => Some(LineEnding::Nel), "\u{0085}" => Some(LineEnding::Nel),
#[cfg(feature = "unicode-lines")]
"\u{2028}" => Some(LineEnding::LS), "\u{2028}" => Some(LineEnding::LS),
#[cfg(feature = "unicode-lines")]
"\u{2029}" => Some(LineEnding::PS), "\u{2029}" => Some(LineEnding::PS),
// Not a line ending // Not a line ending
_ => None, _ => None,
@ -101,7 +125,9 @@ pub fn auto_detect_line_ending(doc: &Rope) -> Option<LineEnding> {
// are being matched, as they might be special-use only // are being matched, as they might be special-use only
for line in doc.lines().take(100) { for line in doc.lines().take(100) {
match get_line_ending(&line) { match get_line_ending(&line) {
None | Some(LineEnding::VT) | Some(LineEnding::FF) | Some(LineEnding::PS) => {} None => {}
#[cfg(feature = "unicode-lines")]
Some(LineEnding::VT) | Some(LineEnding::FF) | Some(LineEnding::PS) => {}
ending => return ending, ending => return ending,
} }
} }
@ -128,6 +154,19 @@ pub fn get_line_ending(line: &RopeSlice) -> Option<LineEnding> {
LineEnding::from_str(g2).or_else(|| LineEnding::from_str(g1)) LineEnding::from_str(g2).or_else(|| LineEnding::from_str(g1))
} }
#[cfg(not(feature = "unicode-lines"))]
/// Returns the passed line's line ending, if any.
pub fn get_line_ending_of_str(line: &str) -> Option<LineEnding> {
if line.ends_with("\u{000D}\u{000A}") {
Some(LineEnding::Crlf)
} else if line.ends_with('\u{000A}') {
Some(LineEnding::LF)
} else {
None
}
}
#[cfg(feature = "unicode-lines")]
/// Returns the passed line's line ending, if any. /// Returns the passed line's line ending, if any.
pub fn get_line_ending_of_str(line: &str) -> Option<LineEnding> { pub fn get_line_ending_of_str(line: &str) -> Option<LineEnding> {
if line.ends_with("\u{000D}\u{000A}") { if line.ends_with("\u{000D}\u{000A}") {
@ -211,6 +250,7 @@ mod line_ending_tests {
#[test] #[test]
fn str_to_line_ending() { fn str_to_line_ending() {
#[cfg(feature = "unicode-lines")]
assert_eq!(LineEnding::from_str("\r"), Some(LineEnding::CR)); assert_eq!(LineEnding::from_str("\r"), Some(LineEnding::CR));
assert_eq!(LineEnding::from_str("\n"), Some(LineEnding::LF)); assert_eq!(LineEnding::from_str("\n"), Some(LineEnding::LF));
assert_eq!(LineEnding::from_str("\r\n"), Some(LineEnding::Crlf)); assert_eq!(LineEnding::from_str("\r\n"), Some(LineEnding::Crlf));
@ -220,6 +260,7 @@ mod line_ending_tests {
#[test] #[test]
fn rope_slice_to_line_ending() { fn rope_slice_to_line_ending() {
let r = Rope::from_str("hello\r\n"); let r = Rope::from_str("hello\r\n");
#[cfg(feature = "unicode-lines")]
assert_eq!( assert_eq!(
LineEnding::from_rope_slice(&r.slice(5..6)), LineEnding::from_rope_slice(&r.slice(5..6)),
Some(LineEnding::CR) Some(LineEnding::CR)
@ -238,6 +279,7 @@ mod line_ending_tests {
#[test] #[test]
fn get_line_ending_rope_slice() { fn get_line_ending_rope_slice() {
let r = Rope::from_str("Hello\rworld\nhow\r\nare you?"); let r = Rope::from_str("Hello\rworld\nhow\r\nare you?");
#[cfg(feature = "unicode-lines")]
assert_eq!(get_line_ending(&r.slice(..6)), Some(LineEnding::CR)); assert_eq!(get_line_ending(&r.slice(..6)), Some(LineEnding::CR));
assert_eq!(get_line_ending(&r.slice(..12)), Some(LineEnding::LF)); assert_eq!(get_line_ending(&r.slice(..12)), Some(LineEnding::LF));
assert_eq!(get_line_ending(&r.slice(..17)), Some(LineEnding::Crlf)); assert_eq!(get_line_ending(&r.slice(..17)), Some(LineEnding::Crlf));
@ -247,6 +289,7 @@ mod line_ending_tests {
#[test] #[test]
fn get_line_ending_str() { fn get_line_ending_str() {
let text = "Hello\rworld\nhow\r\nare you?"; let text = "Hello\rworld\nhow\r\nare you?";
#[cfg(feature = "unicode-lines")]
assert_eq!(get_line_ending_of_str(&text[..6]), Some(LineEnding::CR)); assert_eq!(get_line_ending_of_str(&text[..6]), Some(LineEnding::CR));
assert_eq!(get_line_ending_of_str(&text[..12]), Some(LineEnding::LF)); assert_eq!(get_line_ending_of_str(&text[..12]), Some(LineEnding::LF));
assert_eq!(get_line_ending_of_str(&text[..17]), Some(LineEnding::Crlf)); assert_eq!(get_line_ending_of_str(&text[..17]), Some(LineEnding::Crlf));

View file

@ -288,11 +288,15 @@ fn set_line_ending(
cx.editor.set_status(match line_ending { cx.editor.set_status(match line_ending {
Crlf => "crlf", Crlf => "crlf",
LF => "line feed", LF => "line feed",
#[cfg(feature = "unicode-lines")]
FF => "form feed", FF => "form feed",
#[cfg(feature = "unicode-lines")]
CR => "carriage return", CR => "carriage return",
#[cfg(feature = "unicode-lines")]
Nel => "next line", Nel => "next line",
// These should never be a document's default line ending. // These should never be a document's default line ending.
#[cfg(feature = "unicode-lines")]
VT | LS | PS => "error", VT | LS | PS => "error",
}); });
@ -307,10 +311,13 @@ fn set_line_ending(
// Attempt to parse argument as a line ending. // Attempt to parse argument as a line ending.
let line_ending = match arg { let line_ending = match arg {
// We check for CR first because it shares a common prefix with CRLF. // We check for CR first because it shares a common prefix with CRLF.
#[cfg(feature = "unicode-lines")]
arg if arg.starts_with("cr") => CR, arg if arg.starts_with("cr") => CR,
arg if arg.starts_with("crlf") => Crlf, arg if arg.starts_with("crlf") => Crlf,
arg if arg.starts_with("lf") => LF, arg if arg.starts_with("lf") => LF,
#[cfg(feature = "unicode-lines")]
arg if arg.starts_with("ff") => FF, arg if arg.starts_with("ff") => FF,
#[cfg(feature = "unicode-lines")]
arg if arg.starts_with("nel") => Nel, arg if arg.starts_with("nel") => Nel,
_ => bail!("invalid line ending"), _ => bail!("invalid line ending"),
}; };