perf(lexer): add #[cold] to unicode path

crates/oxc_parser/src/lexer/source.rs

@@ -375,7 +375,12 @@ impl<'a> Source<'a> {
             unsafe { self.ptr = self.ptr.add(1) };
             return Some(byte as char);
         }
+        self.next_unicode_char(byte)
+    }
 
+    #[expect(clippy::unnecessary_wraps)]
+    #[cold] // Unicode is rare.
+    fn next_unicode_char(&mut self, byte: u8) -> Option<char> {
         // Multi-byte Unicode character.
         // Check invariant that `ptr` is on a UTF-8 character boundary.
         debug_assert!(!is_utf8_cont_byte(byte));
@@ -404,6 +409,12 @@ impl<'a> Source<'a> {
             return Some([byte1 as char, byte2 as char]);
         }
 
+        // Handle Unicode characters
+        self.next_2_unicode_chars(byte1)
+    }
+
+    #[cold] // Unicode is rare.
+    fn next_2_unicode_chars(&mut self, byte1: u8) -> Option<[char; 2]> {
         // Multi-byte Unicode character.
         // Check invariant that `ptr` is on a UTF-8 character boundary.
         debug_assert!(!is_utf8_cont_byte(byte1));
@@ -505,7 +516,12 @@ impl<'a> Source<'a> {
         if byte.is_ascii() {
             return Some(byte as char);
         }
+        self.peek_unicode_char(byte)
+    }
 
+    #[expect(clippy::unnecessary_wraps)]
+    #[cold] // Unicode is rare.
+    fn peek_unicode_char(&self, byte: u8) -> Option<char> {
         // Multi-byte Unicode character.
         // Check invariant that `ptr` is on a UTF-8 character boundary.
         debug_assert!(!is_utf8_cont_byte(byte));

crates/oxc_parser/src/lexer/string.rs

@@ -237,9 +237,13 @@ impl<'a> Lexer<'a> {
     /// This reduces the overall memory consumption while keeping the `Token` size small
     /// Strings without escaped values can be retrieved as is from the token span
     pub(super) fn save_string(&mut self, has_escape: bool, s: &'a str) {
-        if !has_escape {
-            return;
+        if has_escape {
+            self.save_escaped_string(s);
         }
+    }
+
+    #[cold]
+    fn save_escaped_string(&mut self, s: &'a str) {
         self.escaped_strings.insert(self.token.start(), s);
         self.token.set_escaped(true);
     }

crates/oxc_parser/src/lexer/unicode.rs

@@ -37,25 +37,34 @@ impl<'a> Lexer<'a> {
                 self.identifier_tail_after_unicode(start_pos);
                 Kind::Ident
             }
-            c if is_irregular_whitespace(c) => {
-                self.consume_char();
-                self.trivia_builder.add_irregular_whitespace(self.token.start(), self.offset());
-                Kind::Skip
-            }
-            c if is_irregular_line_terminator(c) => {
-                self.consume_char();
-                self.token.set_is_on_new_line(true);
-                self.trivia_builder.add_irregular_whitespace(self.token.start(), self.offset());
-                Kind::Skip
-            }
-            _ => {
-                self.consume_char();
-                self.error(diagnostics::invalid_character(c, self.unterminated_range()));
-                Kind::Undetermined
-            }
+            c if is_irregular_whitespace(c) => self.handle_irregular_whitespace(c),
+            c if is_irregular_line_terminator(c) => self.handle_irregular_line_terminator(c),
+            _ => self.handle_invalid_unicode_char(c),
         }
     }
 
+    #[cold]
+    fn handle_irregular_whitespace(&mut self, _c: char) -> Kind {
+        self.consume_char();
+        self.trivia_builder.add_irregular_whitespace(self.token.start(), self.offset());
+        Kind::Skip
+    }
+
+    #[cold]
+    fn handle_irregular_line_terminator(&mut self, _c: char) -> Kind {
+        self.consume_char();
+        self.token.set_is_on_new_line(true);
+        self.trivia_builder.add_irregular_whitespace(self.token.start(), self.offset());
+        Kind::Skip
+    }
+
+    #[cold]
+    fn handle_invalid_unicode_char(&mut self, c: char) -> Kind {
+        self.consume_char();
+        self.error(diagnostics::invalid_character(c, self.unterminated_range()));
+        Kind::Undetermined
+    }
+
     /// Identifier `UnicodeEscapeSequence`
     ///   \u `Hex4Digits`
     ///   \u{ `CodePoint` }