diff options
author | Daniel Baumann <daniel.baumann@progress-linux.org> | 2024-04-17 12:02:58 +0000 |
---|---|---|
committer | Daniel Baumann <daniel.baumann@progress-linux.org> | 2024-04-17 12:02:58 +0000 |
commit | 698f8c2f01ea549d77d7dc3338a12e04c11057b9 (patch) | |
tree | 173a775858bd501c378080a10dca74132f05bc50 /compiler/rustc_ast/src/util/unicode.rs | |
parent | Initial commit. (diff) | |
download | rustc-698f8c2f01ea549d77d7dc3338a12e04c11057b9.tar.xz rustc-698f8c2f01ea549d77d7dc3338a12e04c11057b9.zip |
Adding upstream version 1.64.0+dfsg1.upstream/1.64.0+dfsg1
Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>
Diffstat (limited to '')
-rw-r--r-- | compiler/rustc_ast/src/util/unicode.rs | 35 |
1 files changed, 35 insertions, 0 deletions
diff --git a/compiler/rustc_ast/src/util/unicode.rs b/compiler/rustc_ast/src/util/unicode.rs new file mode 100644 index 000000000..f009f7b30 --- /dev/null +++ b/compiler/rustc_ast/src/util/unicode.rs @@ -0,0 +1,35 @@ +pub const TEXT_FLOW_CONTROL_CHARS: &[char] = &[ + '\u{202A}', '\u{202B}', '\u{202D}', '\u{202E}', '\u{2066}', '\u{2067}', '\u{2068}', '\u{202C}', + '\u{2069}', +]; + +#[inline] +pub fn contains_text_flow_control_chars(s: &str) -> bool { + // Char - UTF-8 + // U+202A - E2 80 AA + // U+202B - E2 80 AB + // U+202C - E2 80 AC + // U+202D - E2 80 AD + // U+202E - E2 80 AE + // U+2066 - E2 81 A6 + // U+2067 - E2 81 A7 + // U+2068 - E2 81 A8 + // U+2069 - E2 81 A9 + let mut bytes = s.as_bytes(); + loop { + match core::slice::memchr::memchr(0xE2, &bytes) { + Some(idx) => { + // bytes are valid UTF-8 -> E2 must be followed by two bytes + let ch = &bytes[idx..idx + 3]; + match ch { + [_, 0x80, 0xAA..=0xAE] | [_, 0x81, 0xA6..=0xA9] => break true, + _ => {} + } + bytes = &bytes[idx + 3..]; + } + None => { + break false; + } + } + } +} |