Skip to content

Commit

Permalink
Tweak Mode.
Browse files Browse the repository at this point in the history
- Add `use Mode::*` to avoid all the qualifiers.
- Reorder the variants. The existing order makes no particular sense,
  which has bugged me for some time. I've chosen an order that makes
  sense to me.
  • Loading branch information
nnethercote committed Dec 8, 2023
1 parent f9f5555 commit 1d031fa
Showing 1 changed file with 38 additions and 34 deletions.
72 changes: 38 additions & 34 deletions compiler/rustc_lexer/src/unescape.rs
Original file line number Diff line number Diff line change
Expand Up @@ -4,6 +4,8 @@
use std::ops::Range;
use std::str::Chars;

use Mode::*;

#[cfg(test)]
mod tests;

Expand Down Expand Up @@ -84,15 +86,14 @@ where
F: FnMut(Range<usize>, Result<char, EscapeError>),
{
match mode {
Mode::Char | Mode::Byte => {
Char | Byte => {
let mut chars = src.chars();
let res = unescape_char_or_byte(&mut chars, mode);
callback(0..(src.len() - chars.as_str().len()), res);
}
Mode::Str | Mode::ByteStr => unescape_str_common(src, mode, callback),

Mode::RawStr | Mode::RawByteStr => unescape_raw_str_or_raw_byte_str(src, mode, callback),
Mode::CStr | Mode::RawCStr => unreachable!(),
Str | ByteStr => unescape_str_common(src, mode, callback),
RawStr | RawByteStr => unescape_raw_str_or_raw_byte_str(src, mode, callback),
CStr | RawCStr => unreachable!(),
}
}

Expand All @@ -118,84 +119,87 @@ pub fn unescape_c_string<F>(src: &str, mode: Mode, callback: &mut F)
where
F: FnMut(Range<usize>, Result<CStrUnit, EscapeError>),
{
if mode == Mode::RawCStr {
unescape_raw_str_or_raw_byte_str(src, mode, &mut |r, result| {
callback(r, result.map(CStrUnit::Char))
});
} else {
unescape_str_common(src, mode, callback);
match mode {
CStr => {
unescape_str_common(src, mode, callback);
}
RawCStr => {
unescape_raw_str_or_raw_byte_str(src, mode, &mut |r, result| {
callback(r, result.map(CStrUnit::Char))
});
}
Char | Byte | Str | RawStr | ByteStr | RawByteStr => unreachable!(),
}
}

/// Takes a contents of a char literal (without quotes), and returns an
/// unescaped char or an error.
pub fn unescape_char(src: &str) -> Result<char, EscapeError> {
unescape_char_or_byte(&mut src.chars(), Mode::Char)
unescape_char_or_byte(&mut src.chars(), Char)
}

/// Takes a contents of a byte literal (without quotes), and returns an
/// unescaped byte or an error.
pub fn unescape_byte(src: &str) -> Result<u8, EscapeError> {
unescape_char_or_byte(&mut src.chars(), Mode::Byte).map(byte_from_char)
unescape_char_or_byte(&mut src.chars(), Byte).map(byte_from_char)
}

/// What kind of literal do we parse.
#[derive(Debug, Clone, Copy, PartialEq)]
pub enum Mode {
Char,
Str,

Byte,
ByteStr,

Str,
RawStr,

ByteStr,
RawByteStr,

CStr,
RawCStr,
}

impl Mode {
pub fn in_double_quotes(self) -> bool {
match self {
Mode::Str
| Mode::ByteStr
| Mode::RawStr
| Mode::RawByteStr
| Mode::CStr
| Mode::RawCStr => true,
Mode::Char | Mode::Byte => false,
Str | RawStr | ByteStr | RawByteStr | CStr | RawCStr => true,
Char | Byte => false,
}
}

/// Non-byte literals should have `\xXX` escapes that are within the ASCII range.
fn ascii_escapes_should_be_ascii(self) -> bool {
match self {
Mode::Char | Mode::Str => true,
Mode::Byte | Mode::ByteStr | Mode::CStr => false,
Mode::RawStr | Mode::RawByteStr | Mode::RawCStr => unreachable!(),
Char | Str => true,
Byte | ByteStr | CStr => false,
RawStr | RawByteStr | RawCStr => unreachable!(),
}
}

/// Whether characters within the literal must be within the ASCII range
/// Whether characters within the literal must be within the ASCII range.
#[inline]
fn chars_should_be_ascii(self) -> bool {
match self {
Mode::Byte | Mode::ByteStr | Mode::RawByteStr => true,
Mode::Char | Mode::Str | Mode::RawStr | Mode::CStr | Mode::RawCStr => false,
Byte | ByteStr | RawByteStr => true,
Char | Str | RawStr | CStr | RawCStr => false,
}
}

/// Byte literals do not allow unicode escape.
fn is_unicode_escape_disallowed(self) -> bool {
match self {
Mode::Byte | Mode::ByteStr | Mode::RawByteStr => true,
Mode::Char | Mode::Str | Mode::RawStr | Mode::CStr | Mode::RawCStr => false,
Byte | ByteStr | RawByteStr => true,
Char | Str | RawStr | CStr | RawCStr => false,
}
}

pub fn prefix_noraw(self) -> &'static str {
match self {
Mode::Byte | Mode::ByteStr | Mode::RawByteStr => "b",
Mode::CStr | Mode::RawCStr => "c",
Mode::Char | Mode::Str | Mode::RawStr => "",
Char | Str | RawStr => "",
Byte | ByteStr | RawByteStr => "b",
CStr | RawCStr => "c",
}
}
}
Expand Down Expand Up @@ -410,7 +414,7 @@ where
#[inline]
pub fn byte_from_char(c: char) -> u8 {
let res = c as u32;
debug_assert!(res <= u8::MAX as u32, "guaranteed because of Mode::ByteStr");
debug_assert!(res <= u8::MAX as u32, "guaranteed because of ByteStr");
res as u8
}

Expand Down

0 comments on commit 1d031fa

Please sign in to comment.