diff options
author | Daniel Baumann <daniel.baumann@progress-linux.org> | 2024-05-18 02:49:42 +0000 |
---|---|---|
committer | Daniel Baumann <daniel.baumann@progress-linux.org> | 2024-05-18 02:49:42 +0000 |
commit | 837b550238aa671a591ccf282dddeab29cadb206 (patch) | |
tree | 914b6b8862bace72bd3245ca184d374b08d8a672 /vendor/proc-macro2/src/parse.rs | |
parent | Adding debian version 1.70.0+dfsg2-1. (diff) | |
download | rustc-837b550238aa671a591ccf282dddeab29cadb206.tar.xz rustc-837b550238aa671a591ccf282dddeab29cadb206.zip |
Merging upstream version 1.71.1+dfsg1.
Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>
Diffstat (limited to 'vendor/proc-macro2/src/parse.rs')
-rw-r--r-- | vendor/proc-macro2/src/parse.rs | 276 |
1 files changed, 194 insertions, 82 deletions
diff --git a/vendor/proc-macro2/src/parse.rs b/vendor/proc-macro2/src/parse.rs index 82291da1b..21e324136 100644 --- a/vendor/proc-macro2/src/parse.rs +++ b/vendor/proc-macro2/src/parse.rs @@ -27,7 +27,18 @@ impl<'a> Cursor<'a> { self.rest.starts_with(s) } - fn is_empty(&self) -> bool { + pub fn starts_with_char(&self, ch: char) -> bool { + self.rest.starts_with(ch) + } + + pub fn starts_with_fn<Pattern>(&self, f: Pattern) -> bool + where + Pattern: FnMut(char) -> bool, + { + self.rest.starts_with(f) + } + + pub fn is_empty(&self) -> bool { self.rest.is_empty() } @@ -97,7 +108,7 @@ fn skip_whitespace(input: Cursor) -> Cursor { s = s.advance(1); continue; } - b if b <= 0x7f => {} + b if b.is_ascii() => {} _ => { let ch = s.chars().next().unwrap(); if is_whitespace(ch) { @@ -262,9 +273,11 @@ fn leaf_token(input: Cursor) -> PResult<TokenTree> { } fn ident(input: Cursor) -> PResult<crate::Ident> { - if ["r\"", "r#\"", "r##", "b\"", "b\'", "br\"", "br#"] - .iter() - .any(|prefix| input.starts_with(prefix)) + if [ + "r\"", "r#\"", "r##", "b\"", "b\'", "br\"", "br#", "c\"", "cr\"", "cr#", + ] + .iter() + .any(|prefix| input.starts_with(prefix)) { Err(Reject) } else { @@ -322,6 +335,8 @@ fn literal_nocapture(input: Cursor) -> Result<Cursor, Reject> { Ok(ok) } else if let Ok(ok) = byte_string(input) { Ok(ok) + } else if let Ok(ok) = c_string(input) { + Ok(ok) } else if let Ok(ok) = byte(input) { Ok(ok) } else if let Ok(ok) = character(input) { @@ -352,8 +367,8 @@ fn string(input: Cursor) -> Result<Cursor, Reject> { } } -fn cooked_string(input: Cursor) -> Result<Cursor, Reject> { - let mut chars = input.char_indices().peekable(); +fn cooked_string(mut input: Cursor) -> Result<Cursor, Reject> { + let mut chars = input.char_indices(); while let Some((i, ch)) = chars.next() { match ch { @@ -367,31 +382,17 @@ fn cooked_string(input: Cursor) -> Result<Cursor, Reject> { }, '\\' => match chars.next() { Some((_, 'x')) => { - if !backslash_x_char(&mut chars) { - break; - } + backslash_x_char(&mut chars)?; } Some((_, 'n')) | Some((_, 'r')) | Some((_, 't')) | Some((_, '\\')) | Some((_, '\'')) | Some((_, '"')) | Some((_, '0')) => {} Some((_, 'u')) => { - if !backslash_u(&mut chars) { - break; - } + backslash_u(&mut chars)?; } - Some((_, ch @ '\n')) | Some((_, ch @ '\r')) => { - let mut last = ch; - loop { - if last == '\r' && chars.next().map_or(true, |(_, ch)| ch != '\n') { - return Err(Reject); - } - match chars.peek() { - Some((_, ch)) if ch.is_whitespace() => { - last = *ch; - chars.next(); - } - _ => break, - } - } + Some((newline, ch @ '\n')) | Some((newline, ch @ '\r')) => { + input = input.advance(newline + 1); + trailing_backslash(&mut input, ch as u8)?; + chars = input.char_indices(); } _ => break, }, @@ -401,11 +402,30 @@ fn cooked_string(input: Cursor) -> Result<Cursor, Reject> { Err(Reject) } +fn raw_string(input: Cursor) -> Result<Cursor, Reject> { + let (input, delimiter) = delimiter_of_raw_string(input)?; + let mut bytes = input.bytes().enumerate(); + while let Some((i, byte)) = bytes.next() { + match byte { + b'"' if input.rest[i + 1..].starts_with(delimiter) => { + let rest = input.advance(i + 1 + delimiter.len()); + return Ok(literal_suffix(rest)); + } + b'\r' => match bytes.next() { + Some((_, b'\n')) => {} + _ => break, + }, + _ => {} + } + } + Err(Reject) +} + fn byte_string(input: Cursor) -> Result<Cursor, Reject> { if let Ok(input) = input.parse("b\"") { cooked_byte_string(input) } else if let Ok(input) = input.parse("br") { - raw_string(input) + raw_byte_string(input) } else { Err(Reject) } @@ -425,68 +445,127 @@ fn cooked_byte_string(mut input: Cursor) -> Result<Cursor, Reject> { }, b'\\' => match bytes.next() { Some((_, b'x')) => { - if !backslash_x_byte(&mut bytes) { - break; - } + backslash_x_byte(&mut bytes)?; } Some((_, b'n')) | Some((_, b'r')) | Some((_, b't')) | Some((_, b'\\')) | Some((_, b'0')) | Some((_, b'\'')) | Some((_, b'"')) => {} Some((newline, b @ b'\n')) | Some((newline, b @ b'\r')) => { - let mut last = b as char; - let rest = input.advance(newline + 1); - let mut chars = rest.char_indices(); - loop { - if last == '\r' && chars.next().map_or(true, |(_, ch)| ch != '\n') { - return Err(Reject); - } - match chars.next() { - Some((_, ch)) if ch.is_whitespace() => last = ch, - Some((offset, _)) => { - input = rest.advance(offset); - bytes = input.bytes().enumerate(); - break; - } - None => return Err(Reject), - } - } + input = input.advance(newline + 1); + trailing_backslash(&mut input, b)?; + bytes = input.bytes().enumerate(); } _ => break, }, - b if b < 0x80 => {} + b if b.is_ascii() => {} _ => break, } } Err(Reject) } -fn raw_string(input: Cursor) -> Result<Cursor, Reject> { - let mut chars = input.char_indices(); - let mut n = 0; - for (i, ch) in &mut chars { - match ch { - '"' => { - n = i; - break; +fn delimiter_of_raw_string(input: Cursor) -> PResult<&str> { + for (i, byte) in input.bytes().enumerate() { + match byte { + b'"' => { + if i > 255 { + // https://github.com/rust-lang/rust/pull/95251 + return Err(Reject); + } + return Ok((input.advance(i + 1), &input.rest[..i])); } - '#' => {} - _ => return Err(Reject), + b'#' => {} + _ => break, } } - if n > 255 { - // https://github.com/rust-lang/rust/pull/95251 - return Err(Reject); + Err(Reject) +} + +fn raw_byte_string(input: Cursor) -> Result<Cursor, Reject> { + let (input, delimiter) = delimiter_of_raw_string(input)?; + let mut bytes = input.bytes().enumerate(); + while let Some((i, byte)) = bytes.next() { + match byte { + b'"' if input.rest[i + 1..].starts_with(delimiter) => { + let rest = input.advance(i + 1 + delimiter.len()); + return Ok(literal_suffix(rest)); + } + b'\r' => match bytes.next() { + Some((_, b'\n')) => {} + _ => break, + }, + other => { + if !other.is_ascii() { + break; + } + } + } + } + Err(Reject) +} + +fn c_string(input: Cursor) -> Result<Cursor, Reject> { + if let Ok(input) = input.parse("c\"") { + cooked_c_string(input) + } else if let Ok(input) = input.parse("cr") { + raw_c_string(input) + } else { + Err(Reject) + } +} + +fn raw_c_string(input: Cursor) -> Result<Cursor, Reject> { + let (input, delimiter) = delimiter_of_raw_string(input)?; + let mut bytes = input.bytes().enumerate(); + while let Some((i, byte)) = bytes.next() { + match byte { + b'"' if input.rest[i + 1..].starts_with(delimiter) => { + let rest = input.advance(i + 1 + delimiter.len()); + return Ok(literal_suffix(rest)); + } + b'\r' => match bytes.next() { + Some((_, b'\n')) => {} + _ => break, + }, + b'\0' => break, + _ => {} + } } + Err(Reject) +} + +fn cooked_c_string(mut input: Cursor) -> Result<Cursor, Reject> { + let mut chars = input.char_indices(); + while let Some((i, ch)) = chars.next() { match ch { - '"' if input.rest[i + 1..].starts_with(&input.rest[..n]) => { - let rest = input.advance(i + 1 + n); - return Ok(literal_suffix(rest)); + '"' => { + let input = input.advance(i + 1); + return Ok(literal_suffix(input)); } '\r' => match chars.next() { Some((_, '\n')) => {} _ => break, }, - _ => {} + '\\' => match chars.next() { + Some((_, 'x')) => { + backslash_x_nonzero(&mut chars)?; + } + Some((_, 'n')) | Some((_, 'r')) | Some((_, 't')) | Some((_, '\\')) + | Some((_, '\'')) | Some((_, '"')) => {} + Some((_, 'u')) => { + if backslash_u(&mut chars)? == '\0' { + break; + } + } + Some((newline, ch @ '\n')) | Some((newline, ch @ '\r')) => { + input = input.advance(newline + 1); + trailing_backslash(&mut input, ch as u8)?; + chars = input.char_indices(); + } + _ => break, + }, + '\0' => break, + _ch => {} } } Err(Reject) @@ -497,7 +576,7 @@ fn byte(input: Cursor) -> Result<Cursor, Reject> { let mut bytes = input.bytes().enumerate(); let ok = match bytes.next().map(|(_, b)| b) { Some(b'\\') => match bytes.next().map(|(_, b)| b) { - Some(b'x') => backslash_x_byte(&mut bytes), + Some(b'x') => backslash_x_byte(&mut bytes).is_ok(), Some(b'n') | Some(b'r') | Some(b't') | Some(b'\\') | Some(b'0') | Some(b'\'') | Some(b'"') => true, _ => false, @@ -520,8 +599,8 @@ fn character(input: Cursor) -> Result<Cursor, Reject> { let mut chars = input.char_indices(); let ok = match chars.next().map(|(_, ch)| ch) { Some('\\') => match chars.next().map(|(_, ch)| ch) { - Some('x') => backslash_x_char(&mut chars), - Some('u') => backslash_u(&mut chars), + Some('x') => backslash_x_char(&mut chars).is_ok(), + Some('u') => backslash_u(&mut chars).is_ok(), Some('n') | Some('r') | Some('t') | Some('\\') | Some('0') | Some('\'') | Some('"') => { true } @@ -542,32 +621,45 @@ macro_rules! next_ch { match $chars.next() { Some((_, ch)) => match ch { $pat $(| $rest)* => ch, - _ => return false, + _ => return Err(Reject), }, - None => return false, + None => return Err(Reject), } }; } -fn backslash_x_char<I>(chars: &mut I) -> bool +fn backslash_x_char<I>(chars: &mut I) -> Result<(), Reject> where I: Iterator<Item = (usize, char)>, { next_ch!(chars @ '0'..='7'); next_ch!(chars @ '0'..='9' | 'a'..='f' | 'A'..='F'); - true + Ok(()) } -fn backslash_x_byte<I>(chars: &mut I) -> bool +fn backslash_x_byte<I>(chars: &mut I) -> Result<(), Reject> where I: Iterator<Item = (usize, u8)>, { next_ch!(chars @ b'0'..=b'9' | b'a'..=b'f' | b'A'..=b'F'); next_ch!(chars @ b'0'..=b'9' | b'a'..=b'f' | b'A'..=b'F'); - true + Ok(()) } -fn backslash_u<I>(chars: &mut I) -> bool +fn backslash_x_nonzero<I>(chars: &mut I) -> Result<(), Reject> +where + I: Iterator<Item = (usize, char)>, +{ + let first = next_ch!(chars @ '0'..='9' | 'a'..='f' | 'A'..='F'); + let second = next_ch!(chars @ '0'..='9' | 'a'..='f' | 'A'..='F'); + if first == '0' && second == '0' { + Err(Reject) + } else { + Ok(()) + } +} + +fn backslash_u<I>(chars: &mut I) -> Result<char, Reject> where I: Iterator<Item = (usize, char)>, { @@ -580,17 +672,37 @@ where 'a'..='f' => 10 + ch as u8 - b'a', 'A'..='F' => 10 + ch as u8 - b'A', '_' if len > 0 => continue, - '}' if len > 0 => return char::from_u32(value).is_some(), - _ => return false, + '}' if len > 0 => return char::from_u32(value).ok_or(Reject), + _ => break, }; if len == 6 { - return false; + break; } value *= 0x10; value += u32::from(digit); len += 1; } - false + Err(Reject) +} + +fn trailing_backslash(input: &mut Cursor, mut last: u8) -> Result<(), Reject> { + let mut whitespace = input.bytes().enumerate(); + loop { + if last == b'\r' && whitespace.next().map_or(true, |(_, b)| b != b'\n') { + return Err(Reject); + } + match whitespace.next() { + Some((_, b @ b' ')) | Some((_, b @ b'\t')) | Some((_, b @ b'\n')) + | Some((_, b @ b'\r')) => { + last = b; + } + Some((offset, _)) => { + *input = input.advance(offset); + return Ok(()); + } + None => return Err(Reject), + } + } } fn float(input: Cursor) -> Result<Cursor, Reject> { @@ -756,7 +868,7 @@ fn digits(mut input: Cursor) -> Result<Cursor, Reject> { fn punct(input: Cursor) -> PResult<Punct> { let (rest, ch) = punct_char(input)?; if ch == '\'' { - if ident_any(rest)?.0.starts_with("'") { + if ident_any(rest)?.0.starts_with_char('\'') { Err(Reject) } else { Ok((rest, Punct::new('\'', Spacing::Joint))) @@ -848,7 +960,7 @@ fn doc_comment_contents(input: Cursor) -> PResult<(&str, bool)> { Ok((input, (&s[3..s.len() - 2], true))) } else if input.starts_with("///") { let input = input.advance(3); - if input.starts_with("/") { + if input.starts_with_char('/') { return Err(Reject); } let (input, s) = take_until_newline_or_eof(input); |