summary refs log tree commit diff
path: root/compiler/rustc_parse/src
diff options
context:
space:
mode:
authorMatthias Krüger <matthias.krueger@famsik.de>2023-02-18 13:26:46 +0100
committerGitHub <noreply@github.com>2023-02-18 13:26:46 +0100
commit3035ccbcb972a7d762c8e3f64a6b200434677209 (patch)
tree14e538a50e8187007e1d40e81ad8b37c42b4ed48 /compiler/rustc_parse/src
parent3701bdc6333145410f009c83bd03f424eca05009 (diff)
parente3f9db5fc319c6d8eee5d47d216ea6a426070c41 (diff)
downloadrust-3035ccbcb972a7d762c8e3f64a6b200434677209.tar.gz
rust-3035ccbcb972a7d762c8e3f64a6b200434677209.zip
Rollup merge of #108031 - jieyouxu:issue-108019, r=estebank
Don't recover lifetimes/labels containing emojis as character literals

Fixes #108019.

Note that at the time of this commit, `unic-emoji-char` seems to have data tables only up to Unicode 5.0, but Unicode is already newer than this.

A newer emoji such as `🥺` will not be recognized as an emoji but older emojis such as `🐱` will.

This PR leaves a couple of FIXMEs where `unic_emoji_char::is_emoji` is used.
Diffstat (limited to 'compiler/rustc_parse/src')
-rw-r--r--compiler/rustc_parse/src/lexer/mod.rs9
1 files changed, 7 insertions, 2 deletions
diff --git a/compiler/rustc_parse/src/lexer/mod.rs b/compiler/rustc_parse/src/lexer/mod.rs
index bd998ed91d9..37449aaabed 100644
--- a/compiler/rustc_parse/src/lexer/mod.rs
+++ b/compiler/rustc_parse/src/lexer/mod.rs
@@ -200,16 +200,21 @@ impl<'a> StringReader<'a> {
                     };
                     token::Literal(token::Lit { kind, symbol, suffix })
                 }
-                rustc_lexer::TokenKind::Lifetime { starts_with_number } => {
+                rustc_lexer::TokenKind::Lifetime { starts_with_number, contains_emoji } => {
                     // Include the leading `'` in the real identifier, for macro
                     // expansion purposes. See #12512 for the gory details of why
                     // this is necessary.
                     let lifetime_name = self.str_from(start);
                     if starts_with_number {
                         let span = self.mk_sp(start, self.pos);
-                        let mut diag = self.sess.struct_err("lifetimes cannot start with a number");
+                        let mut diag = self.sess.struct_err("lifetimes or labels cannot start with a number");
                         diag.set_span(span);
                         diag.stash(span, StashKey::LifetimeIsChar);
+                    } else if contains_emoji {
+                        let span = self.mk_sp(start, self.pos);
+                        let mut diag = self.sess.struct_err("lifetimes or labels cannot contain emojis");
+                        diag.set_span(span);
+                        diag.stash(span, StashKey::LifetimeContainsEmoji);
                     }
                     let ident = Symbol::intern(lifetime_name);
                     token::Lifetime(ident)