diff options
author | Valentin Popov <valentin@popov.link> | 2024-01-08 00:21:28 +0300 |
---|---|---|
committer | Valentin Popov <valentin@popov.link> | 2024-01-08 00:21:28 +0300 |
commit | 1b6a04ca5504955c571d1c97504fb45ea0befee4 (patch) | |
tree | 7579f518b23313e8a9748a88ab6173d5e030b227 /vendor/anstream/src/adapter | |
parent | 5ecd8cf2cba827454317368b68571df0d13d7842 (diff) | |
download | fparkan-1b6a04ca5504955c571d1c97504fb45ea0befee4.tar.xz fparkan-1b6a04ca5504955c571d1c97504fb45ea0befee4.zip |
Initial vendor packages
Signed-off-by: Valentin Popov <valentin@popov.link>
Diffstat (limited to 'vendor/anstream/src/adapter')
-rw-r--r-- | vendor/anstream/src/adapter/mod.rs | 15 | ||||
-rw-r--r-- | vendor/anstream/src/adapter/strip.rs | 513 | ||||
-rw-r--r-- | vendor/anstream/src/adapter/wincon.rs | 320 |
3 files changed, 848 insertions, 0 deletions
diff --git a/vendor/anstream/src/adapter/mod.rs b/vendor/anstream/src/adapter/mod.rs new file mode 100644 index 0000000..f266b68 --- /dev/null +++ b/vendor/anstream/src/adapter/mod.rs @@ -0,0 +1,15 @@ +//! Gracefully degrade styled output + +mod strip; +mod wincon; + +pub use strip::strip_bytes; +pub use strip::strip_str; +pub use strip::StripBytes; +pub use strip::StripBytesIter; +pub use strip::StripStr; +pub use strip::StripStrIter; +pub use strip::StrippedBytes; +pub use strip::StrippedStr; +pub use wincon::WinconBytes; +pub use wincon::WinconBytesIter; diff --git a/vendor/anstream/src/adapter/strip.rs b/vendor/anstream/src/adapter/strip.rs new file mode 100644 index 0000000..5078c51 --- /dev/null +++ b/vendor/anstream/src/adapter/strip.rs @@ -0,0 +1,513 @@ +use anstyle_parse::state::state_change; +use anstyle_parse::state::Action; +use anstyle_parse::state::State; + +/// Strip ANSI escapes from a `&str`, returning the printable content +/// +/// This can be used to take output from a program that includes escape sequences and write it +/// somewhere that does not easily support them, such as a log file. +/// +/// For non-contiguous data, see [`StripStr`]. +/// +/// # Example +/// +/// ```rust +/// use std::io::Write as _; +/// +/// let styled_text = "\x1b[32mfoo\x1b[m bar"; +/// let plain_str = anstream::adapter::strip_str(&styled_text).to_string(); +/// assert_eq!(plain_str, "foo bar"); +/// ``` +#[inline] +pub fn strip_str(data: &str) -> StrippedStr<'_> { + StrippedStr::new(data) +} + +/// See [`strip_str`] +#[derive(Default, Clone, Debug, PartialEq, Eq)] +pub struct StrippedStr<'s> { + bytes: &'s [u8], + state: State, +} + +impl<'s> StrippedStr<'s> { + #[inline] + fn new(data: &'s str) -> Self { + Self { + bytes: data.as_bytes(), + state: State::Ground, + } + } + + /// Create a [`String`] of the printable content + #[inline] + #[allow(clippy::inherent_to_string_shadow_display)] // Single-allocation implementation + pub fn to_string(&self) -> String { + use std::fmt::Write as _; + let mut stripped = String::with_capacity(self.bytes.len()); + let _ = write!(&mut stripped, "{}", self); + stripped + } +} + +impl<'s> std::fmt::Display for StrippedStr<'s> { + /// **Note:** this does *not* exhaust the [`Iterator`] + #[inline] + fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result { + let iter = Self { + bytes: self.bytes, + state: self.state, + }; + for printable in iter { + printable.fmt(f)?; + } + Ok(()) + } +} + +impl<'s> Iterator for StrippedStr<'s> { + type Item = &'s str; + + #[inline] + fn next(&mut self) -> Option<Self::Item> { + next_str(&mut self.bytes, &mut self.state) + } +} + +/// Incrementally strip non-contiguous data +#[derive(Default, Clone, Debug, PartialEq, Eq)] +pub struct StripStr { + state: State, +} + +impl StripStr { + /// Initial state + pub fn new() -> Self { + Default::default() + } + + /// Strip the next segment of data + pub fn strip_next<'s>(&'s mut self, data: &'s str) -> StripStrIter<'s> { + StripStrIter { + bytes: data.as_bytes(), + state: &mut self.state, + } + } +} + +/// See [`StripStr`] +#[derive(Debug, PartialEq, Eq)] +pub struct StripStrIter<'s> { + bytes: &'s [u8], + state: &'s mut State, +} + +impl<'s> Iterator for StripStrIter<'s> { + type Item = &'s str; + + #[inline] + fn next(&mut self) -> Option<Self::Item> { + next_str(&mut self.bytes, self.state) + } +} + +#[inline] +fn next_str<'s>(bytes: &mut &'s [u8], state: &mut State) -> Option<&'s str> { + let offset = bytes.iter().copied().position(|b| { + let (next_state, action) = state_change(*state, b); + if next_state != State::Anywhere { + *state = next_state; + } + is_printable_str(action, b) + }); + let (_, next) = bytes.split_at(offset.unwrap_or(bytes.len())); + *bytes = next; + *state = State::Ground; + + let offset = bytes.iter().copied().position(|b| { + let (_next_state, action) = state_change(State::Ground, b); + !is_printable_str(action, b) + }); + let (printable, next) = bytes.split_at(offset.unwrap_or(bytes.len())); + *bytes = next; + if printable.is_empty() { + None + } else { + let printable = unsafe { + from_utf8_unchecked( + printable, + "`bytes` was validated as UTF-8, the parser preserves UTF-8 continuations", + ) + }; + Some(printable) + } +} + +#[inline] +unsafe fn from_utf8_unchecked<'b>(bytes: &'b [u8], safety_justification: &'static str) -> &'b str { + if cfg!(debug_assertions) { + // Catch problems more quickly when testing + std::str::from_utf8(bytes).expect(safety_justification) + } else { + std::str::from_utf8_unchecked(bytes) + } +} + +#[inline] +fn is_printable_str(action: Action, byte: u8) -> bool { + // VT320 considered 0x7f to be `Print`able but we expect to be working in UTF-8 systems and not + // ISO Latin-1, making it DEL and non-printable + const DEL: u8 = 0x7f; + (action == Action::Print && byte != DEL) + || action == Action::BeginUtf8 + // since we know the input is valid UTF-8, the only thing we can do with + // continuations is to print them + || is_utf8_continuation(byte) + || (action == Action::Execute && byte.is_ascii_whitespace()) +} + +#[inline] +fn is_utf8_continuation(b: u8) -> bool { + matches!(b, 0x80..=0xbf) +} + +/// Strip ANSI escapes from bytes, returning the printable content +/// +/// This can be used to take output from a program that includes escape sequences and write it +/// somewhere that does not easily support them, such as a log file. +/// +/// # Example +/// +/// ```rust +/// use std::io::Write as _; +/// +/// let styled_text = "\x1b[32mfoo\x1b[m bar"; +/// let plain_str = anstream::adapter::strip_bytes(styled_text.as_bytes()).into_vec(); +/// assert_eq!(plain_str.as_slice(), &b"foo bar"[..]); +/// ``` +#[inline] +pub fn strip_bytes(data: &[u8]) -> StrippedBytes<'_> { + StrippedBytes::new(data) +} + +/// See [`strip_bytes`] +#[derive(Default, Clone, Debug, PartialEq, Eq)] +pub struct StrippedBytes<'s> { + bytes: &'s [u8], + state: State, + utf8parser: Utf8Parser, +} + +impl<'s> StrippedBytes<'s> { + /// See [`strip_bytes`] + #[inline] + pub fn new(bytes: &'s [u8]) -> Self { + Self { + bytes, + state: State::Ground, + utf8parser: Default::default(), + } + } + + /// Strip the next slice of bytes + /// + /// Used when the content is in several non-contiguous slices + /// + /// # Panic + /// + /// May panic if it is not exhausted / empty + #[inline] + pub fn extend(&mut self, bytes: &'s [u8]) { + debug_assert!( + self.is_empty(), + "current bytes must be processed to ensure we end at the right state" + ); + self.bytes = bytes; + } + + /// Report the bytes has been exhausted + #[inline] + pub fn is_empty(&self) -> bool { + self.bytes.is_empty() + } + + /// Create a [`Vec`] of the printable content + #[inline] + pub fn into_vec(self) -> Vec<u8> { + let mut stripped = Vec::with_capacity(self.bytes.len()); + for printable in self { + stripped.extend(printable); + } + stripped + } +} + +impl<'s> Iterator for StrippedBytes<'s> { + type Item = &'s [u8]; + + #[inline] + fn next(&mut self) -> Option<Self::Item> { + next_bytes(&mut self.bytes, &mut self.state, &mut self.utf8parser) + } +} + +/// Incrementally strip non-contiguous data +#[derive(Default, Clone, Debug, PartialEq, Eq)] +pub struct StripBytes { + state: State, + utf8parser: Utf8Parser, +} + +impl StripBytes { + /// Initial state + pub fn new() -> Self { + Default::default() + } + + /// Strip the next segment of data + pub fn strip_next<'s>(&'s mut self, bytes: &'s [u8]) -> StripBytesIter<'s> { + StripBytesIter { + bytes, + state: &mut self.state, + utf8parser: &mut self.utf8parser, + } + } +} + +/// See [`StripBytes`] +#[derive(Debug, PartialEq, Eq)] +pub struct StripBytesIter<'s> { + bytes: &'s [u8], + state: &'s mut State, + utf8parser: &'s mut Utf8Parser, +} + +impl<'s> Iterator for StripBytesIter<'s> { + type Item = &'s [u8]; + + #[inline] + fn next(&mut self) -> Option<Self::Item> { + next_bytes(&mut self.bytes, self.state, self.utf8parser) + } +} + +#[inline] +fn next_bytes<'s>( + bytes: &mut &'s [u8], + state: &mut State, + utf8parser: &mut Utf8Parser, +) -> Option<&'s [u8]> { + let offset = bytes.iter().copied().position(|b| { + if *state == State::Utf8 { + true + } else { + let (next_state, action) = state_change(*state, b); + if next_state != State::Anywhere { + *state = next_state; + } + is_printable_bytes(action, b) + } + }); + let (_, next) = bytes.split_at(offset.unwrap_or(bytes.len())); + *bytes = next; + + let offset = bytes.iter().copied().position(|b| { + if *state == State::Utf8 { + if utf8parser.add(b) { + *state = State::Ground; + } + false + } else { + let (next_state, action) = state_change(State::Ground, b); + if next_state != State::Anywhere { + *state = next_state; + } + if *state == State::Utf8 { + utf8parser.add(b); + false + } else { + !is_printable_bytes(action, b) + } + } + }); + let (printable, next) = bytes.split_at(offset.unwrap_or(bytes.len())); + *bytes = next; + if printable.is_empty() { + None + } else { + Some(printable) + } +} + +#[derive(Default, Clone, Debug, PartialEq, Eq)] +pub struct Utf8Parser { + utf8_parser: utf8parse::Parser, +} + +impl Utf8Parser { + fn add(&mut self, byte: u8) -> bool { + let mut b = false; + let mut receiver = VtUtf8Receiver(&mut b); + self.utf8_parser.advance(&mut receiver, byte); + b + } +} + +struct VtUtf8Receiver<'a>(&'a mut bool); + +impl<'a> utf8parse::Receiver for VtUtf8Receiver<'a> { + fn codepoint(&mut self, _: char) { + *self.0 = true; + } + + fn invalid_sequence(&mut self) { + *self.0 = true; + } +} + +#[inline] +fn is_printable_bytes(action: Action, byte: u8) -> bool { + // VT320 considered 0x7f to be `Print`able but we expect to be working in UTF-8 systems and not + // ISO Latin-1, making it DEL and non-printable + const DEL: u8 = 0x7f; + + // Continuations aren't included as they may also be control codes, requiring more context + (action == Action::Print && byte != DEL) + || action == Action::BeginUtf8 + || (action == Action::Execute && byte.is_ascii_whitespace()) +} + +#[cfg(test)] +mod test { + use super::*; + use proptest::prelude::*; + + /// Model based off full parser + fn parser_strip(bytes: &[u8]) -> String { + #[derive(Default)] + struct Strip(String); + impl Strip { + fn with_capacity(capacity: usize) -> Self { + Self(String::with_capacity(capacity)) + } + } + impl anstyle_parse::Perform for Strip { + fn print(&mut self, c: char) { + self.0.push(c); + } + + fn execute(&mut self, byte: u8) { + if byte.is_ascii_whitespace() { + self.0.push(byte as char); + } + } + } + + let mut stripped = Strip::with_capacity(bytes.len()); + let mut parser = anstyle_parse::Parser::<anstyle_parse::DefaultCharAccumulator>::new(); + for byte in bytes { + parser.advance(&mut stripped, *byte); + } + stripped.0 + } + + /// Model verifying incremental parsing + fn strip_char(mut s: &str) -> String { + let mut result = String::new(); + let mut state = StripStr::new(); + while !s.is_empty() { + let mut indices = s.char_indices(); + indices.next(); // current + let offset = indices.next().map(|(i, _)| i).unwrap_or_else(|| s.len()); + let (current, remainder) = s.split_at(offset); + for printable in state.strip_next(current) { + result.push_str(printable); + } + s = remainder; + } + result + } + + /// Model verifying incremental parsing + fn strip_byte(s: &[u8]) -> Vec<u8> { + let mut result = Vec::new(); + let mut state = StripBytes::default(); + for start in 0..s.len() { + let current = &s[start..=start]; + for printable in state.strip_next(current) { + result.extend(printable); + } + } + result + } + + #[test] + fn test_strip_bytes_multibyte() { + let bytes = [240, 145, 141, 139]; + let expected = parser_strip(&bytes); + let actual = String::from_utf8(strip_bytes(&bytes).into_vec()).unwrap(); + assert_eq!(expected, actual); + } + + #[test] + fn test_strip_byte_multibyte() { + let bytes = [240, 145, 141, 139]; + let expected = parser_strip(&bytes); + let actual = String::from_utf8(strip_byte(&bytes).to_vec()).unwrap(); + assert_eq!(expected, actual); + } + + #[test] + fn test_strip_str_del() { + let input = std::str::from_utf8(&[0x7f]).unwrap(); + let expected = ""; + let actual = strip_str(input).to_string(); + assert_eq!(expected, actual); + } + + #[test] + fn test_strip_byte_del() { + let bytes = [0x7f]; + let expected = ""; + let actual = String::from_utf8(strip_byte(&bytes).to_vec()).unwrap(); + assert_eq!(expected, actual); + } + + proptest! { + #[test] + #[cfg_attr(miri, ignore)] // See https://github.com/AltSysrq/proptest/issues/253 + fn strip_str_no_escapes(s in "\\PC*") { + let expected = parser_strip(s.as_bytes()); + let actual = strip_str(&s).to_string(); + assert_eq!(expected, actual); + } + + #[test] + #[cfg_attr(miri, ignore)] // See https://github.com/AltSysrq/proptest/issues/253 + fn strip_char_no_escapes(s in "\\PC*") { + let expected = parser_strip(s.as_bytes()); + let actual = strip_char(&s); + assert_eq!(expected, actual); + } + + #[test] + #[cfg_attr(miri, ignore)] // See https://github.com/AltSysrq/proptest/issues/253 + fn strip_bytes_no_escapes(s in "\\PC*") { + dbg!(&s); + dbg!(s.as_bytes()); + let expected = parser_strip(s.as_bytes()); + let actual = String::from_utf8(strip_bytes(s.as_bytes()).into_vec()).unwrap(); + assert_eq!(expected, actual); + } + + #[test] + #[cfg_attr(miri, ignore)] // See https://github.com/AltSysrq/proptest/issues/253 + fn strip_byte_no_escapes(s in "\\PC*") { + dbg!(&s); + dbg!(s.as_bytes()); + let expected = parser_strip(s.as_bytes()); + let actual = String::from_utf8(strip_byte(s.as_bytes()).to_vec()).unwrap(); + assert_eq!(expected, actual); + } + } +} diff --git a/vendor/anstream/src/adapter/wincon.rs b/vendor/anstream/src/adapter/wincon.rs new file mode 100644 index 0000000..de3da65 --- /dev/null +++ b/vendor/anstream/src/adapter/wincon.rs @@ -0,0 +1,320 @@ +/// Incrementally convert to wincon calls for non-contiguous data +#[derive(Default, Clone, Debug, PartialEq, Eq)] +pub struct WinconBytes { + parser: anstyle_parse::Parser, + capture: WinconCapture, +} + +impl WinconBytes { + /// Initial state + pub fn new() -> Self { + Default::default() + } + + /// Strip the next segment of data + pub fn extract_next<'s>(&'s mut self, bytes: &'s [u8]) -> WinconBytesIter<'s> { + self.capture.reset(); + self.capture.printable.reserve(bytes.len()); + WinconBytesIter { + bytes, + parser: &mut self.parser, + capture: &mut self.capture, + } + } +} + +/// See [`WinconBytes`] +#[derive(Debug, PartialEq, Eq)] +pub struct WinconBytesIter<'s> { + bytes: &'s [u8], + parser: &'s mut anstyle_parse::Parser, + capture: &'s mut WinconCapture, +} + +impl<'s> Iterator for WinconBytesIter<'s> { + type Item = (anstyle::Style, String); + + #[inline] + fn next(&mut self) -> Option<Self::Item> { + next_bytes(&mut self.bytes, self.parser, self.capture) + } +} + +#[inline] +fn next_bytes( + bytes: &mut &[u8], + parser: &mut anstyle_parse::Parser, + capture: &mut WinconCapture, +) -> Option<(anstyle::Style, String)> { + capture.reset(); + while capture.ready.is_none() { + let byte = if let Some((byte, remainder)) = (*bytes).split_first() { + *bytes = remainder; + *byte + } else { + break; + }; + parser.advance(capture, byte); + } + if capture.printable.is_empty() { + return None; + } + + let style = capture.ready.unwrap_or(capture.style); + Some((style, std::mem::take(&mut capture.printable))) +} + +#[derive(Default, Clone, Debug, PartialEq, Eq)] +struct WinconCapture { + style: anstyle::Style, + printable: String, + ready: Option<anstyle::Style>, +} + +impl WinconCapture { + fn reset(&mut self) { + self.ready = None; + } +} + +impl anstyle_parse::Perform for WinconCapture { + /// Draw a character to the screen and update states. + fn print(&mut self, c: char) { + self.printable.push(c); + } + + /// Execute a C0 or C1 control function. + fn execute(&mut self, byte: u8) { + if byte.is_ascii_whitespace() { + self.printable.push(byte as char); + } + } + + fn csi_dispatch( + &mut self, + params: &anstyle_parse::Params, + _intermediates: &[u8], + ignore: bool, + action: u8, + ) { + if ignore { + return; + } + if action != b'm' { + return; + } + + let mut style = self.style; + // param/value differences are dependent on the escape code + let mut state = State::Normal; + let mut r = None; + let mut g = None; + let mut is_bg = false; + for param in params { + for value in param { + match (state, *value) { + (State::Normal, 0) => { + style = anstyle::Style::default(); + break; + } + (State::Normal, 1) => { + style = style.bold(); + break; + } + (State::Normal, 4) => { + style = style.underline(); + break; + } + (State::Normal, 30..=37) => { + let color = to_ansi_color(value - 30).unwrap(); + style = style.fg_color(Some(color.into())); + break; + } + (State::Normal, 38) => { + is_bg = false; + state = State::PrepareCustomColor; + } + (State::Normal, 39) => { + style = style.fg_color(None); + break; + } + (State::Normal, 40..=47) => { + let color = to_ansi_color(value - 40).unwrap(); + style = style.bg_color(Some(color.into())); + break; + } + (State::Normal, 48) => { + is_bg = true; + state = State::PrepareCustomColor; + } + (State::Normal, 49) => { + style = style.bg_color(None); + break; + } + (State::Normal, 90..=97) => { + let color = to_ansi_color(value - 90).unwrap().bright(true); + style = style.fg_color(Some(color.into())); + break; + } + (State::Normal, 100..=107) => { + let color = to_ansi_color(value - 100).unwrap().bright(true); + style = style.bg_color(Some(color.into())); + break; + } + (State::PrepareCustomColor, 5) => { + state = State::Ansi256; + } + (State::PrepareCustomColor, 2) => { + state = State::Rgb; + r = None; + g = None; + } + (State::Ansi256, n) => { + let color = anstyle::Ansi256Color(n as u8); + if is_bg { + style = style.bg_color(Some(color.into())); + } else { + style = style.fg_color(Some(color.into())); + } + break; + } + (State::Rgb, b) => match (r, g) { + (None, _) => { + r = Some(b); + } + (Some(_), None) => { + g = Some(b); + } + (Some(r), Some(g)) => { + let color = anstyle::RgbColor(r as u8, g as u8, b as u8); + if is_bg { + style = style.bg_color(Some(color.into())); + } else { + style = style.fg_color(Some(color.into())); + } + break; + } + }, + _ => { + break; + } + } + } + } + + if style != self.style && !self.printable.is_empty() { + self.ready = Some(self.style); + } + self.style = style; + } +} + +#[derive(Copy, Clone, PartialEq, Eq, Debug)] +enum State { + Normal, + PrepareCustomColor, + Ansi256, + Rgb, +} + +fn to_ansi_color(digit: u16) -> Option<anstyle::AnsiColor> { + match digit { + 0 => Some(anstyle::AnsiColor::Black), + 1 => Some(anstyle::AnsiColor::Red), + 2 => Some(anstyle::AnsiColor::Green), + 3 => Some(anstyle::AnsiColor::Yellow), + 4 => Some(anstyle::AnsiColor::Blue), + 5 => Some(anstyle::AnsiColor::Magenta), + 6 => Some(anstyle::AnsiColor::Cyan), + 7 => Some(anstyle::AnsiColor::White), + _ => None, + } +} + +#[cfg(test)] +mod test { + use super::*; + use owo_colors::OwoColorize as _; + use proptest::prelude::*; + + #[track_caller] + fn verify(input: &str, expected: Vec<(anstyle::Style, &str)>) { + let expected = expected + .into_iter() + .map(|(style, value)| (style, value.to_owned())) + .collect::<Vec<_>>(); + let mut state = WinconBytes::new(); + let actual = state.extract_next(input.as_bytes()).collect::<Vec<_>>(); + assert_eq!(expected, actual, "{input:?}"); + } + + #[test] + fn start() { + let input = format!("{} world!", "Hello".green().on_red()); + let expected = vec![ + ( + anstyle::AnsiColor::Green.on(anstyle::AnsiColor::Red), + "Hello", + ), + (anstyle::Style::default(), " world!"), + ]; + verify(&input, expected); + } + + #[test] + fn middle() { + let input = format!("Hello {}!", "world".green().on_red()); + let expected = vec![ + (anstyle::Style::default(), "Hello "), + ( + anstyle::AnsiColor::Green.on(anstyle::AnsiColor::Red), + "world", + ), + (anstyle::Style::default(), "!"), + ]; + verify(&input, expected); + } + + #[test] + fn end() { + let input = format!("Hello {}", "world!".green().on_red()); + let expected = vec![ + (anstyle::Style::default(), "Hello "), + ( + anstyle::AnsiColor::Green.on(anstyle::AnsiColor::Red), + "world!", + ), + ]; + verify(&input, expected); + } + + #[test] + fn ansi256_colors() { + // termcolor only supports "brights" via these + let input = format!( + "Hello {}!", + "world".color(owo_colors::XtermColors::UserBrightYellow) + ); + let expected = vec![ + (anstyle::Style::default(), "Hello "), + (anstyle::Ansi256Color(11).on_default(), "world"), + (anstyle::Style::default(), "!"), + ]; + verify(&input, expected); + } + + proptest! { + #[test] + #[cfg_attr(miri, ignore)] // See https://github.com/AltSysrq/proptest/issues/253 + fn wincon_no_escapes(s in "\\PC*") { + let expected = if s.is_empty() { + vec![] + } else { + vec![(anstyle::Style::default(), s.clone())] + }; + let mut state = WinconBytes::new(); + let actual = state.extract_next(s.as_bytes()).collect::<Vec<_>>(); + assert_eq!(expected, actual); + } + } +} |