bjoernager/rust
bjoernager/rust
1
Fork 0
Code Activity

Revert "Don't recover lifetimes/labels containing emojis as character literals"

Browse source 
Reverts PR #108031
Fixes (doesnt close until beta backported) #109746

This reverts commit e3f9db5fc3.
This reverts commit 98b82aedba.
This reverts commit 380fa26413.
This commit is contained in:
Michael Goulet 2023-04-10 06:52:18 +00:00
parent 3c2e2dd5c5
commit a047064d6b
8 changed files with 17 additions and 178 deletions
Download patch file Download diff file Expand all files Collapse all files

43
compiler/rustc_lexer/src/lib.rs
View file

@ -95,7 +95,7 @@ pub enum TokenKind {
Literal { kind: LiteralKind, suffix_start: u32 },
/// "'a"
Lifetime { starts_with_number: bool, contains_emoji: bool },
Lifetime { starts_with_number: bool },
// One-char tokens:
/// ";"
@ -632,13 +632,7 @@ impl Cursor<'_> {
// If the first symbol is valid for identifier, it can be a lifetime.
// Also check if it's a number for a better error reporting (so '0 will
// be reported as invalid lifetime and not as unterminated char literal).
// We also have to account for potential `'🐱` emojis to avoid reporting
// it as an unterminated char literal.
is_id_start(self.first())
|| self.first().is_digit(10)
// FIXME(#108019): `unic-emoji-char` seems to have data tables only up to Unicode
// 5.0, but Unicode is already newer than this.
|| unic_emoji_char::is_emoji(self.first())
is_id_start(self.first()) || self.first().is_digit(10)
};
if !can_be_a_lifetime {
@ -651,33 +645,16 @@ impl Cursor<'_> {
return Literal { kind, suffix_start };
}
// Either a lifetime or a character literal.
// Either a lifetime or a character literal with
// length greater than 1.
let starts_with_number = self.first().is_digit(10);
let mut contains_emoji = false;
// FIXME(#108019): `unic-emoji-char` seems to have data tables only up to Unicode
// 5.0, but Unicode is already newer than this.
if unic_emoji_char::is_emoji(self.first()) {
contains_emoji = true;
} else {
// Skip the literal contents.
// First symbol can be a number (which isn't a valid identifier start),
// so skip it without any checks.
self.bump();
}
self.eat_while(|c| {
if is_id_continue(c) {
true
// FIXME(#108019): `unic-emoji-char` seems to have data tables only up to Unicode
// 5.0, but Unicode is already newer than this.
} else if unic_emoji_char::is_emoji(c) {
contains_emoji = true;
true
} else {
false
}
});
// Skip the literal contents.
// First symbol can be a number (which isn't a valid identifier start),
// so skip it without any checks.
self.bump();
self.eat_while(is_id_continue);
// Check if after skipping literal contents we've met a closing
// single quote (which means that user attempted to create a
@ -687,7 +664,7 @@ impl Cursor<'_> {
let kind = Char { terminated: true };
Literal { kind, suffix_start: self.pos_within_token() }
} else {
Lifetime { starts_with_number, contains_emoji }
Lifetime { starts_with_number }
}
}

2
compiler/rustc_lexer/src/tests.rs
View file

@ -235,7 +235,7 @@ fn lifetime() {
check_lexing(
"'abc",
expect![[r#"
Token { kind: Lifetime { starts_with_number: false, contains_emoji: false }, len: 4 }
Token { kind: Lifetime { starts_with_number: false }, len: 4 }
"#]],
);
}