| 1 |
// Vendored from rustc-literal-escaper v0.0.5 |
| 2 |
// https://github.com/rust-lang/literal-escaper/tree/v0.0.5 |
| 3 |
|
| 4 |
//! Utilities for validating (raw) string, char, and byte literals and |
| 5 |
//! turning escape sequences into the values they represent. |
| 6 |
|
| 7 |
use crate::num::NonZeroChar; |
| 8 |
use std::ffi::CStr; |
| 9 |
use std::num::NonZeroU8; |
| 10 |
use std::ops::Range; |
| 11 |
use std::str::Chars; |
| 12 |
|
| 13 |
/// Errors and warnings that can occur during string, char, and byte unescaping. |
| 14 |
/// |
| 15 |
/// Mostly relating to malformed escape sequences, but also a few other problems. |
| 16 |
#[derive(Debug, PartialEq, Eq)] |
| 17 |
pub enum EscapeError { |
| 18 |
/// Expected 1 char, but 0 were found. |
| 19 |
ZeroChars, |
| 20 |
/// Expected 1 char, but more than 1 were found. |
| 21 |
MoreThanOneChar, |
| 22 |
|
| 23 |
/// Escaped '\' character without continuation. |
| 24 |
LoneSlash, |
| 25 |
/// Invalid escape character (e.g. '\z'). |
| 26 |
InvalidEscape, |
| 27 |
/// Raw '\r' encountered. |
| 28 |
BareCarriageReturn, |
| 29 |
/// Raw '\r' encountered in raw string. |
| 30 |
BareCarriageReturnInRawString, |
| 31 |
/// Unescaped character that was expected to be escaped (e.g. raw '\t'). |
| 32 |
EscapeOnlyChar, |
| 33 |
|
| 34 |
/// Numeric character escape is too short (e.g. '\x1'). |
| 35 |
TooShortHexEscape, |
| 36 |
/// Invalid character in numeric escape (e.g. '\xz') |
| 37 |
InvalidCharInHexEscape, |
| 38 |
/// Character code in numeric escape is non-ascii (e.g. '\xFF'). |
| 39 |
OutOfRangeHexEscape, |
| 40 |
|
| 41 |
/// '\u' not followed by '{'. |
| 42 |
NoBraceInUnicodeEscape, |
| 43 |
/// Non-hexadecimal value in '\u{..}'. |
| 44 |
InvalidCharInUnicodeEscape, |
| 45 |
/// '\u{}' |
| 46 |
EmptyUnicodeEscape, |
| 47 |
/// No closing brace in '\u{..}', e.g. '\u{12'. |
| 48 |
UnclosedUnicodeEscape, |
| 49 |
/// '\u{_12}' |
| 50 |
LeadingUnderscoreUnicodeEscape, |
| 51 |
/// More than 6 characters in '\u{..}', e.g. '\u{10FFFF_FF}' |
| 52 |
OverlongUnicodeEscape, |
| 53 |
/// Invalid in-bound unicode character code, e.g. '\u{DFFF}'. |
| 54 |
LoneSurrogateUnicodeEscape, |
| 55 |
/// Out of bounds unicode character code, e.g. '\u{FFFFFF}'. |
| 56 |
OutOfRangeUnicodeEscape, |
| 57 |
|
| 58 |
/// Unicode escape code in byte literal. |
| 59 |
UnicodeEscapeInByte, |
| 60 |
/// Non-ascii character in byte literal, byte string literal, or raw byte string literal. |
| 61 |
NonAsciiCharInByte, |
| 62 |
|
| 63 |
/// `\0` in a C string literal. |
| 64 |
NulInCStr, |
| 65 |
|
| 66 |
/// After a line ending with '\', the next line contains whitespace |
| 67 |
/// characters that are not skipped. |
| 68 |
UnskippedWhitespaceWarning, |
| 69 |
|
| 70 |
/// After a line ending with '\', multiple lines are skipped. |
| 71 |
MultipleSkippedLinesWarning, |
| 72 |
} |
| 73 |
|
| 74 |
impl EscapeError { |
| 75 |
/// Returns true for actual errors, as opposed to warnings. |
| 76 |
pub fn is_fatal(&self) -> bool { |
| 77 |
!matches!( |
| 78 |
self, |
| 79 |
EscapeError::UnskippedWhitespaceWarning | EscapeError::MultipleSkippedLinesWarning |
| 80 |
) |
| 81 |
} |
| 82 |
} |
| 83 |
|
| 84 |
/// Check a raw string literal for validity |
| 85 |
/// |
| 86 |
/// Takes the contents of a raw string literal (without quotes) |
| 87 |
/// and produces a sequence of characters or errors, |
| 88 |
/// which are returned by invoking `callback`. |
| 89 |
/// NOTE: Does no escaping, but produces errors for bare carriage return ('\r'). |
| 90 |
pub fn check_raw_str(src: &str, callback: impl FnMut(Range<usize>, Result<char, EscapeError>)) { |
| 91 |
str::check_raw(src, callback); |
| 92 |
} |
| 93 |
|
| 94 |
/// Check a raw byte string literal for validity |
| 95 |
/// |
| 96 |
/// Takes the contents of a raw byte string literal (without quotes) |
| 97 |
/// and produces a sequence of bytes or errors, |
| 98 |
/// which are returned by invoking `callback`. |
| 99 |
/// NOTE: Does no escaping, but produces errors for bare carriage return ('\r'). |
| 100 |
pub fn check_raw_byte_str(src: &str, callback: impl FnMut(Range<usize>, Result<u8, EscapeError>)) { |
| 101 |
<[u8]>::check_raw(src, callback); |
| 102 |
} |
| 103 |
|
| 104 |
/// Check a raw C string literal for validity |
| 105 |
/// |
| 106 |
/// Takes the contents of a raw C string literal (without quotes) |
| 107 |
/// and produces a sequence of characters or errors, |
| 108 |
/// which are returned by invoking `callback`. |
| 109 |
/// NOTE: Does no escaping, but produces errors for bare carriage return ('\r'). |
| 110 |
pub fn check_raw_c_str( |
| 111 |
src: &str, |
| 112 |
callback: impl FnMut(Range<usize>, Result<NonZeroChar, EscapeError>), |
| 113 |
) { |
| 114 |
CStr::check_raw(src, callback); |
| 115 |
} |
| 116 |
|
| 117 |
/// Trait for checking raw string literals for validity |
| 118 |
trait CheckRaw { |
| 119 |
/// Unit type of the implementing string type (`char` for string, `u8` for byte string) |
| 120 |
type RawUnit; |
| 121 |
|
| 122 |
/// Converts chars to the unit type of the literal type |
| 123 |
fn char2raw_unit(c: char) -> Result<Self::RawUnit, EscapeError>; |
| 124 |
|
| 125 |
/// Takes the contents of a raw literal (without quotes) |
| 126 |
/// and produces a sequence of `Result<Self::RawUnit, EscapeError>` |
| 127 |
/// which are returned via `callback`. |
| 128 |
/// |
| 129 |
/// NOTE: Does no escaping, but produces errors for bare carriage return ('\r'). |
| 130 |
fn check_raw( |
| 131 |
src: &str, |
| 132 |
mut callback: impl FnMut(Range<usize>, Result<Self::RawUnit, EscapeError>), |
| 133 |
) { |
| 134 |
let mut chars = src.chars(); |
| 135 |
while let Some(c) = chars.next() { |
| 136 |
let start = src.len() - chars.as_str().len() - c.len_utf8(); |
| 137 |
let res = match c { |
| 138 |
'\r' => Err(EscapeError::BareCarriageReturnInRawString), |
| 139 |
_ => Self::char2raw_unit(c), |
| 140 |
}; |
| 141 |
let end = src.len() - chars.as_str().len(); |
| 142 |
callback(start..end, res); |
| 143 |
} |
| 144 |
|
| 145 |
// Unfortunately, it is a bit unclear whether the following equivalent code is slower or faster: bug 141855 |
| 146 |
// src.char_indices().for_each(|(pos, c)| { |
| 147 |
// callback( |
| 148 |
// pos..pos + c.len_utf8(), |
| 149 |
// if c == '\r' { |
| 150 |
// Err(EscapeError::BareCarriageReturnInRawString) |
| 151 |
// } else { |
| 152 |
// Self::char2raw_unit(c) |
| 153 |
// }, |
| 154 |
// ); |
| 155 |
// }); |
| 156 |
} |
| 157 |
} |
| 158 |
|
| 159 |
impl CheckRaw for str { |
| 160 |
type RawUnit = char; |
| 161 |
|
| 162 |
#[inline] |
| 163 |
fn char2raw_unit(c: char) -> Result<Self::RawUnit, EscapeError> { |
| 164 |
Ok(c) |
| 165 |
} |
| 166 |
} |
| 167 |
|
| 168 |
impl CheckRaw for [u8] { |
| 169 |
type RawUnit = u8; |
| 170 |
|
| 171 |
#[inline] |
| 172 |
fn char2raw_unit(c: char) -> Result<Self::RawUnit, EscapeError> { |
| 173 |
char2byte(c) |
| 174 |
} |
| 175 |
} |
| 176 |
|
| 177 |
/// Turn an ascii char into a byte |
| 178 |
#[inline] |
| 179 |
fn char2byte(c: char) -> Result<u8, EscapeError> { |
| 180 |
// do NOT do: c.try_into().ok_or(EscapeError::NonAsciiCharInByte) |
| 181 |
if c.is_ascii() { |
| 182 |
Ok(c as u8) |
| 183 |
} else { |
| 184 |
Err(EscapeError::NonAsciiCharInByte) |
| 185 |
} |
| 186 |
} |
| 187 |
|
| 188 |
impl CheckRaw for CStr { |
| 189 |
type RawUnit = NonZeroChar; |
| 190 |
|
| 191 |
#[inline] |
| 192 |
fn char2raw_unit(c: char) -> Result<Self::RawUnit, EscapeError> { |
| 193 |
NonZeroChar::new(c).ok_or(EscapeError::NulInCStr) |
| 194 |
} |
| 195 |
} |
| 196 |
|
| 197 |
/// Unescape a char literal |
| 198 |
/// |
| 199 |
/// Takes the contents of a char literal (without quotes), |
| 200 |
/// and returns an unescaped char or an error. |
| 201 |
#[inline] |
| 202 |
pub fn unescape_char(src: &str) -> Result<char, EscapeError> { |
| 203 |
str::unescape_single(&mut src.chars()) |
| 204 |
} |
| 205 |
|
| 206 |
/// Unescape a byte literal |
| 207 |
/// |
| 208 |
/// Takes the contents of a byte literal (without quotes), |
| 209 |
/// and returns an unescaped byte or an error. |
| 210 |
#[inline] |
| 211 |
pub fn unescape_byte(src: &str) -> Result<u8, EscapeError> { |
| 212 |
<[u8]>::unescape_single(&mut src.chars()) |
| 213 |
} |
| 214 |
|
| 215 |
/// Unescape a string literal |
| 216 |
/// |
| 217 |
/// Takes the contents of a string literal (without quotes) |
| 218 |
/// and produces a sequence of escaped characters or errors, |
| 219 |
/// which are returned by invoking `callback`. |
| 220 |
pub fn unescape_str(src: &str, callback: impl FnMut(Range<usize>, Result<char, EscapeError>)) { |
| 221 |
str::unescape(src, callback) |
| 222 |
} |
| 223 |
|
| 224 |
/// Unescape a byte string literal |
| 225 |
/// |
| 226 |
/// Takes the contents of a byte string literal (without quotes) |
| 227 |
/// and produces a sequence of escaped bytes or errors, |
| 228 |
/// which are returned by invoking `callback`. |
| 229 |
pub fn unescape_byte_str(src: &str, callback: impl FnMut(Range<usize>, Result<u8, EscapeError>)) { |
| 230 |
<[u8]>::unescape(src, callback) |
| 231 |
} |
| 232 |
|
| 233 |
/// Unescape a C string literal |
| 234 |
/// |
| 235 |
/// Takes the contents of a C string literal (without quotes) |
| 236 |
/// and produces a sequence of escaped MixedUnits or errors, |
| 237 |
/// which are returned by invoking `callback`. |
| 238 |
pub fn unescape_c_str( |
| 239 |
src: &str, |
| 240 |
callback: impl FnMut(Range<usize>, Result<MixedUnit, EscapeError>), |
| 241 |
) { |
| 242 |
CStr::unescape(src, callback) |
| 243 |
} |
| 244 |
|
| 245 |
/// Enum representing either a char or a byte |
| 246 |
/// |
| 247 |
/// Used for mixed utf8 string literals, i.e. those that allow both unicode |
| 248 |
/// chars and high bytes. |
| 249 |
#[derive(Copy, Clone, Debug, PartialEq, Eq)] |
| 250 |
pub enum MixedUnit { |
| 251 |
/// Used for ASCII chars (written directly or via `\x00`..`\x7f` escapes) |
| 252 |
/// and Unicode chars (written directly or via `\u` escapes). |
| 253 |
/// |
| 254 |
/// For example, if '¥' appears in a string it is represented here as |
| 255 |
/// `MixedUnit::Char('¥')`, and it will be appended to the relevant byte |
| 256 |
/// string as the two-byte UTF-8 sequence `[0xc2, 0xa5]` |
| 257 |
Char(NonZeroChar), |
| 258 |
|
| 259 |
/// Used for high bytes (`\x80`..`\xff`). |
| 260 |
/// |
| 261 |
/// For example, if `\xa5` appears in a string it is represented here as |
| 262 |
/// `MixedUnit::HighByte(0xa5)`, and it will be appended to the relevant |
| 263 |
/// byte string as the single byte `0xa5`. |
| 264 |
HighByte(NonZeroU8), |
| 265 |
} |
| 266 |
|
| 267 |
impl From<NonZeroChar> for MixedUnit { |
| 268 |
#[inline] |
| 269 |
fn from(c: NonZeroChar) -> Self { |
| 270 |
MixedUnit::Char(c) |
| 271 |
} |
| 272 |
} |
| 273 |
|
| 274 |
impl From<NonZeroU8> for MixedUnit { |
| 275 |
#[inline] |
| 276 |
fn from(byte: NonZeroU8) -> Self { |
| 277 |
if byte.get().is_ascii() { |
| 278 |
MixedUnit::Char(NonZeroChar::new(byte.get() as char).unwrap()) |
| 279 |
} else { |
| 280 |
MixedUnit::HighByte(byte) |
| 281 |
} |
| 282 |
} |
| 283 |
} |
| 284 |
|
| 285 |
impl TryFrom<char> for MixedUnit { |
| 286 |
type Error = EscapeError; |
| 287 |
|
| 288 |
#[inline] |
| 289 |
fn try_from(c: char) -> Result<Self, EscapeError> { |
| 290 |
NonZeroChar::new(c) |
| 291 |
.map(MixedUnit::Char) |
| 292 |
.ok_or(EscapeError::NulInCStr) |
| 293 |
} |
| 294 |
} |
| 295 |
|
| 296 |
impl TryFrom<u8> for MixedUnit { |
| 297 |
type Error = EscapeError; |
| 298 |
|
| 299 |
#[inline] |
| 300 |
fn try_from(byte: u8) -> Result<Self, EscapeError> { |
| 301 |
NonZeroU8::new(byte) |
| 302 |
.map(From::from) |
| 303 |
.ok_or(EscapeError::NulInCStr) |
| 304 |
} |
| 305 |
} |
| 306 |
|
| 307 |
/// Trait for unescaping escape sequences in strings |
| 308 |
trait Unescape { |
| 309 |
/// Unit type of the implementing string type (`char` for string, `u8` for byte string) |
| 310 |
type Unit; |
| 311 |
|
| 312 |
/// Result of unescaping the zero char ('\0') |
| 313 |
const ZERO_RESULT: Result<Self::Unit, EscapeError>; |
| 314 |
|
| 315 |
/// Converts non-zero bytes to the unit type |
| 316 |
fn nonzero_byte2unit(b: NonZeroU8) -> Self::Unit; |
| 317 |
|
| 318 |
/// Converts chars to the unit type |
| 319 |
fn char2unit(c: char) -> Result<Self::Unit, EscapeError>; |
| 320 |
|
| 321 |
/// Converts the byte of a hex escape to the unit type |
| 322 |
fn hex2unit(b: u8) -> Result<Self::Unit, EscapeError>; |
| 323 |
|
| 324 |
/// Converts the result of a unicode escape to the unit type |
| 325 |
fn unicode2unit(r: Result<char, EscapeError>) -> Result<Self::Unit, EscapeError>; |
| 326 |
|
| 327 |
/// Unescape a single unit (single quote syntax) |
| 328 |
fn unescape_single(chars: &mut Chars<'_>) -> Result<Self::Unit, EscapeError> { |
| 329 |
let res = match chars.next().ok_or(EscapeError::ZeroChars)? { |
| 330 |
'\\' => Self::unescape_1(chars), |
| 331 |
'\n' | '\t' | '\'' => Err(EscapeError::EscapeOnlyChar), |
| 332 |
'\r' => Err(EscapeError::BareCarriageReturn), |
| 333 |
c => Self::char2unit(c), |
| 334 |
}?; |
| 335 |
if chars.next().is_some() { |
| 336 |
return Err(EscapeError::MoreThanOneChar); |
| 337 |
} |
| 338 |
Ok(res) |
| 339 |
} |
| 340 |
|
| 341 |
/// Unescape the first unit of a string (double quoted syntax) |
| 342 |
fn unescape_1(chars: &mut Chars<'_>) -> Result<Self::Unit, EscapeError> { |
| 343 |
// Previous character was '\\', unescape what follows. |
| 344 |
let c = chars.next().ok_or(EscapeError::LoneSlash)?; |
| 345 |
if c == '0' { |
| 346 |
Self::ZERO_RESULT |
| 347 |
} else { |
| 348 |
simple_escape(c) |
| 349 |
.map(|b| Self::nonzero_byte2unit(b)) |
| 350 |
.or_else(|c| match c { |
| 351 |
'x' => Self::hex2unit(hex_escape(chars)?), |
| 352 |
'u' => Self::unicode2unit({ |
| 353 |
let value = unicode_escape(chars)?; |
| 354 |
if value > char::MAX as u32 { |
| 355 |
Err(EscapeError::OutOfRangeUnicodeEscape) |
| 356 |
} else { |
| 357 |
char::from_u32(value).ok_or(EscapeError::LoneSurrogateUnicodeEscape) |
| 358 |
} |
| 359 |
}), |
| 360 |
_ => Err(EscapeError::InvalidEscape), |
| 361 |
}) |
| 362 |
} |
| 363 |
} |
| 364 |
|
| 365 |
/// Unescape a string literal |
| 366 |
/// |
| 367 |
/// Takes the contents of a raw string literal (without quotes) |
| 368 |
/// and produces a sequence of `Result<Self::Unit, EscapeError>` |
| 369 |
/// which are returned via `callback`. |
| 370 |
fn unescape( |
| 371 |
src: &str, |
| 372 |
mut callback: impl FnMut(Range<usize>, Result<Self::Unit, EscapeError>), |
| 373 |
) { |
| 374 |
let mut chars = src.chars(); |
| 375 |
while let Some(c) = chars.next() { |
| 376 |
let start = src.len() - chars.as_str().len() - c.len_utf8(); |
| 377 |
let res = match c { |
| 378 |
'\\' => { |
| 379 |
if let Some(b'\n') = chars.as_str().as_bytes().first() { |
| 380 |
let _ = chars.next(); |
| 381 |
// skip whitespace for backslash newline, see [Rust language reference] |
| 382 |
// (https://doc.rust-lang.org/reference/tokens.html#string-literals). |
| 383 |
let callback_err = |range, err| callback(range, Err(err)); |
| 384 |
skip_ascii_whitespace(&mut chars, start, callback_err); |
| 385 |
continue; |
| 386 |
} else { |
| 387 |
Self::unescape_1(&mut chars) |
| 388 |
} |
| 389 |
} |
| 390 |
'"' => Err(EscapeError::EscapeOnlyChar), |
| 391 |
'\r' => Err(EscapeError::BareCarriageReturn), |
| 392 |
c => Self::char2unit(c), |
| 393 |
}; |
| 394 |
let end = src.len() - chars.as_str().len(); |
| 395 |
callback(start..end, res); |
| 396 |
} |
| 397 |
} |
| 398 |
} |
| 399 |
|
| 400 |
/// Interpret a non-nul ASCII escape |
| 401 |
/// |
| 402 |
/// Parses the character of an ASCII escape (except nul) without the leading backslash. |
| 403 |
#[inline] // single use in Unescape::unescape_1 |
| 404 |
fn simple_escape(c: char) -> Result<NonZeroU8, char> { |
| 405 |
// Previous character was '\\', unescape what follows. |
| 406 |
Ok(NonZeroU8::new(match c { |
| 407 |
'"' => b'"', |
| 408 |
'n' => b'\n', |
| 409 |
'r' => b'\r', |
| 410 |
't' => b'\t', |
| 411 |
'\\' => b'\\', |
| 412 |
'\'' => b'\'', |
| 413 |
_ => Err(c)?, |
| 414 |
}) |
| 415 |
.unwrap()) |
| 416 |
} |
| 417 |
|
| 418 |
/// Interpret a hexadecimal escape |
| 419 |
/// |
| 420 |
/// Parses the two hexadecimal characters of a hexadecimal escape without the leading r"\x". |
| 421 |
#[inline] // single use in Unescape::unescape_1 |
| 422 |
fn hex_escape(chars: &mut impl Iterator<Item = char>) -> Result<u8, EscapeError> { |
| 423 |
let hi = chars.next().ok_or(EscapeError::TooShortHexEscape)?; |
| 424 |
let hi = hi.to_digit(16).ok_or(EscapeError::InvalidCharInHexEscape)?; |
| 425 |
|
| 426 |
let lo = chars.next().ok_or(EscapeError::TooShortHexEscape)?; |
| 427 |
let lo = lo.to_digit(16).ok_or(EscapeError::InvalidCharInHexEscape)?; |
| 428 |
|
| 429 |
Ok((hi * 16 + lo) as u8) |
| 430 |
} |
| 431 |
|
| 432 |
/// Interpret a unicode escape |
| 433 |
/// |
| 434 |
/// Parse the braces with hexadecimal characters (and underscores) part of a unicode escape. |
| 435 |
/// This r"{...}" normally comes after r"\u" and cannot start with an underscore. |
| 436 |
#[inline] // single use in Unescape::unescape_1 |
| 437 |
fn unicode_escape(chars: &mut impl Iterator<Item = char>) -> Result<u32, EscapeError> { |
| 438 |
if chars.next() != Some('{') { |
| 439 |
return Err(EscapeError::NoBraceInUnicodeEscape); |
| 440 |
} |
| 441 |
|
| 442 |
// First character must be a hexadecimal digit. |
| 443 |
let mut value: u32 = match chars.next().ok_or(EscapeError::UnclosedUnicodeEscape)? { |
| 444 |
'_' => return Err(EscapeError::LeadingUnderscoreUnicodeEscape), |
| 445 |
'}' => return Err(EscapeError::EmptyUnicodeEscape), |
| 446 |
c => c |
| 447 |
.to_digit(16) |
| 448 |
.ok_or(EscapeError::InvalidCharInUnicodeEscape)?, |
| 449 |
}; |
| 450 |
|
| 451 |
// First character is valid, now parse the rest of the number |
| 452 |
// and closing brace. |
| 453 |
let mut n_digits = 1; |
| 454 |
loop { |
| 455 |
match chars.next() { |
| 456 |
None => return Err(EscapeError::UnclosedUnicodeEscape), |
| 457 |
Some('_') => continue, |
| 458 |
Some('}') => { |
| 459 |
// Incorrect syntax has higher priority for error reporting |
| 460 |
// than unallowed value for a literal. |
| 461 |
return if n_digits > 6 { |
| 462 |
Err(EscapeError::OverlongUnicodeEscape) |
| 463 |
} else { |
| 464 |
Ok(value) |
| 465 |
}; |
| 466 |
} |
| 467 |
Some(c) => { |
| 468 |
let digit: u32 = c |
| 469 |
.to_digit(16) |
| 470 |
.ok_or(EscapeError::InvalidCharInUnicodeEscape)?; |
| 471 |
n_digits += 1; |
| 472 |
if n_digits > 6 { |
| 473 |
// Stop updating value since we're sure that it's incorrect already. |
| 474 |
continue; |
| 475 |
} |
| 476 |
value = value * 16 + digit; |
| 477 |
} |
| 478 |
}; |
| 479 |
} |
| 480 |
} |
| 481 |
|
| 482 |
/// Interpret a string continuation escape (https://doc.rust-lang.org/reference/expressions/literal-expr.html#string-continuation-escapes) |
| 483 |
/// |
| 484 |
/// Skip ASCII whitespace, except for the formfeed character |
| 485 |
/// (see [this issue](https://github.com/rust-lang/rust/issues/136600)). |
| 486 |
/// Warns on unescaped newline and following non-ASCII whitespace. |
| 487 |
#[inline] // single use in Unescape::unescape |
| 488 |
fn skip_ascii_whitespace( |
| 489 |
chars: &mut Chars<'_>, |
| 490 |
start: usize, |
| 491 |
mut callback: impl FnMut(Range<usize>, EscapeError), |
| 492 |
) { |
| 493 |
let rest = chars.as_str(); |
| 494 |
let first_non_space = rest |
| 495 |
.bytes() |
| 496 |
.position(|b| b != b' ' && b != b'\t' && b != b'\n' && b != b'\r') |
| 497 |
.unwrap_or(rest.len()); |
| 498 |
let (space, rest) = rest.split_at(first_non_space); |
| 499 |
// backslash newline adds 2 bytes |
| 500 |
let end = start + 2 + first_non_space; |
| 501 |
if space.contains('\n') { |
| 502 |
callback(start..end, EscapeError::MultipleSkippedLinesWarning); |
| 503 |
} |
| 504 |
*chars = rest.chars(); |
| 505 |
if let Some(c) = chars.clone().next() { |
| 506 |
if c.is_whitespace() { |
| 507 |
// for error reporting, include the character that was not skipped in the span |
| 508 |
callback( |
| 509 |
start..end + c.len_utf8(), |
| 510 |
EscapeError::UnskippedWhitespaceWarning, |
| 511 |
); |
| 512 |
} |
| 513 |
} |
| 514 |
} |
| 515 |
|
| 516 |
impl Unescape for str { |
| 517 |
type Unit = char; |
| 518 |
|
| 519 |
const ZERO_RESULT: Result<Self::Unit, EscapeError> = Ok('\0'); |
| 520 |
|
| 521 |
#[inline] |
| 522 |
fn nonzero_byte2unit(b: NonZeroU8) -> Self::Unit { |
| 523 |
b.get().into() |
| 524 |
} |
| 525 |
|
| 526 |
#[inline] |
| 527 |
fn char2unit(c: char) -> Result<Self::Unit, EscapeError> { |
| 528 |
Ok(c) |
| 529 |
} |
| 530 |
|
| 531 |
#[inline] |
| 532 |
fn hex2unit(b: u8) -> Result<Self::Unit, EscapeError> { |
| 533 |
if b.is_ascii() { |
| 534 |
Ok(b as char) |
| 535 |
} else { |
| 536 |
Err(EscapeError::OutOfRangeHexEscape) |
| 537 |
} |
| 538 |
} |
| 539 |
|
| 540 |
#[inline] |
| 541 |
fn unicode2unit(r: Result<char, EscapeError>) -> Result<Self::Unit, EscapeError> { |
| 542 |
r |
| 543 |
} |
| 544 |
} |
| 545 |
|
| 546 |
impl Unescape for [u8] { |
| 547 |
type Unit = u8; |
| 548 |
|
| 549 |
const ZERO_RESULT: Result<Self::Unit, EscapeError> = Ok(b'\0'); |
| 550 |
|
| 551 |
#[inline] |
| 552 |
fn nonzero_byte2unit(b: NonZeroU8) -> Self::Unit { |
| 553 |
b.get() |
| 554 |
} |
| 555 |
|
| 556 |
#[inline] |
| 557 |
fn char2unit(c: char) -> Result<Self::Unit, EscapeError> { |
| 558 |
char2byte(c) |
| 559 |
} |
| 560 |
|
| 561 |
#[inline] |
| 562 |
fn hex2unit(b: u8) -> Result<Self::Unit, EscapeError> { |
| 563 |
Ok(b) |
| 564 |
} |
| 565 |
|
| 566 |
#[inline] |
| 567 |
fn unicode2unit(_r: Result<char, EscapeError>) -> Result<Self::Unit, EscapeError> { |
| 568 |
Err(EscapeError::UnicodeEscapeInByte) |
| 569 |
} |
| 570 |
} |
| 571 |
|
| 572 |
impl Unescape for CStr { |
| 573 |
type Unit = MixedUnit; |
| 574 |
|
| 575 |
const ZERO_RESULT: Result<Self::Unit, EscapeError> = Err(EscapeError::NulInCStr); |
| 576 |
|
| 577 |
#[inline] |
| 578 |
fn nonzero_byte2unit(b: NonZeroU8) -> Self::Unit { |
| 579 |
b.into() |
| 580 |
} |
| 581 |
|
| 582 |
#[inline] |
| 583 |
fn char2unit(c: char) -> Result<Self::Unit, EscapeError> { |
| 584 |
c.try_into() |
| 585 |
} |
| 586 |
|
| 587 |
#[inline] |
| 588 |
fn hex2unit(byte: u8) -> Result<Self::Unit, EscapeError> { |
| 589 |
byte.try_into() |
| 590 |
} |
| 591 |
|
| 592 |
#[inline] |
| 593 |
fn unicode2unit(r: Result<char, EscapeError>) -> Result<Self::Unit, EscapeError> { |
| 594 |
Self::char2unit(r?) |
| 595 |
} |
| 596 |
} |
| 597 |
|
| 598 |
/// Enum of the different kinds of literal |
| 599 |
#[derive(Debug, Clone, Copy, PartialEq)] |
| 600 |
pub enum Mode { |
| 601 |
/// `'a'` |
| 602 |
Char, |
| 603 |
|
| 604 |
/// `b'a'` |
| 605 |
Byte, |
| 606 |
|
| 607 |
/// `"hello"` |
| 608 |
Str, |
| 609 |
/// `r"hello"` |
| 610 |
RawStr, |
| 611 |
|
| 612 |
/// `b"hello"` |
| 613 |
ByteStr, |
| 614 |
/// `br"hello"` |
| 615 |
RawByteStr, |
| 616 |
|
| 617 |
/// `c"hello"` |
| 618 |
CStr, |
| 619 |
/// `cr"hello"` |
| 620 |
RawCStr, |
| 621 |
} |
| 622 |
|
| 623 |
impl Mode { |
| 624 |
pub fn in_double_quotes(self) -> bool { |
| 625 |
match self { |
| 626 |
Mode::Str |
| 627 |
| Mode::RawStr |
| 628 |
| Mode::ByteStr |
| 629 |
| Mode::RawByteStr |
| 630 |
| Mode::CStr |
| 631 |
| Mode::RawCStr => true, |
| 632 |
Mode::Char | Mode::Byte => false, |
| 633 |
} |
| 634 |
} |
| 635 |
|
| 636 |
pub fn prefix_noraw(self) -> &'static str { |
| 637 |
match self { |
| 638 |
Mode::Char | Mode::Str | Mode::RawStr => "", |
| 639 |
Mode::Byte | Mode::ByteStr | Mode::RawByteStr => "b", |
| 640 |
Mode::CStr | Mode::RawCStr => "c", |
| 641 |
} |
| 642 |
} |
| 643 |
} |
| 644 |
|
| 645 |
/// Check a literal only for errors |
| 646 |
/// |
| 647 |
/// Takes the contents of a literal (without quotes) |
| 648 |
/// and produces a sequence of only errors, |
| 649 |
/// which are returned by invoking `error_callback`. |
| 650 |
/// |
| 651 |
/// NB Does not produce any output other than errors |
| 652 |
pub fn check_for_errors( |
| 653 |
src: &str, |
| 654 |
mode: Mode, |
| 655 |
mut error_callback: impl FnMut(Range<usize>, EscapeError), |
| 656 |
) { |
| 657 |
match mode { |
| 658 |
Mode::Char => { |
| 659 |
let mut chars = src.chars(); |
| 660 |
if let Err(e) = str::unescape_single(&mut chars) { |
| 661 |
error_callback(0..(src.len() - chars.as_str().len()), e); |
| 662 |
} |
| 663 |
} |
| 664 |
Mode::Byte => { |
| 665 |
let mut chars = src.chars(); |
| 666 |
if let Err(e) = <[u8]>::unescape_single(&mut chars) { |
| 667 |
error_callback(0..(src.len() - chars.as_str().len()), e); |
| 668 |
} |
| 669 |
} |
| 670 |
Mode::Str => unescape_str(src, |range, res| { |
| 671 |
if let Err(e) = res { |
| 672 |
error_callback(range, e); |
| 673 |
} |
| 674 |
}), |
| 675 |
Mode::ByteStr => unescape_byte_str(src, |range, res| { |
| 676 |
if let Err(e) = res { |
| 677 |
error_callback(range, e); |
| 678 |
} |
| 679 |
}), |
| 680 |
Mode::CStr => unescape_c_str(src, |range, res| { |
| 681 |
if let Err(e) = res { |
| 682 |
error_callback(range, e); |
| 683 |
} |
| 684 |
}), |
| 685 |
Mode::RawStr => check_raw_str(src, |range, res| { |
| 686 |
if let Err(e) = res { |
| 687 |
error_callback(range, e); |
| 688 |
} |
| 689 |
}), |
| 690 |
Mode::RawByteStr => check_raw_byte_str(src, |range, res| { |
| 691 |
if let Err(e) = res { |
| 692 |
error_callback(range, e); |
| 693 |
} |
| 694 |
}), |
| 695 |
Mode::RawCStr => check_raw_c_str(src, |range, res| { |
| 696 |
if let Err(e) = res { |
| 697 |
error_callback(range, e); |
| 698 |
} |
| 699 |
}), |
| 700 |
} |
| 701 |
} |