fix(parser): improve lexing of jsx identifier to fix duplicated comments after jsx name (#2687)

crates/oxc_parser/examples/parser.rs

@@ -20,6 +20,15 @@ fn main() -> Result<(), String> {
     println!("AST:");
     println!("{}", serde_json::to_string_pretty(&ret.program).unwrap());
 
+    println!("Comments:");
+    let comments = ret
+        .trivias
+        .comments
+        .into_iter()
+        .map(|(start, end, _)| &source_text[start as usize..end as usize])
+        .collect::<Vec<_>>();
+    println!("{comments:?}");
+
     if ret.errors.is_empty() {
         println!("Parsed Successfully.");
     } else {

crates/oxc_parser/src/cursor.rs

@@ -221,9 +221,11 @@ impl<'a> ParserImpl<'a> {
         }
     }
 
-    /// Tell lexer to re-read a jsx identifier
+    /// Tell lexer to continue reading jsx identifier if the lexer character position is at `-` for `<component-name>`
-    pub(crate) fn re_lex_jsx_identifier(&mut self) {
+    pub(crate) fn continue_lex_jsx_identifier(&mut self) {
-        self.token = self.lexer.next_jsx_identifier(self.cur_token().start);
+        if let Some(token) = self.lexer.continue_lex_jsx_identifier() {
+            self.token = token;
+        }
     }
 
     pub(crate) fn re_lex_right_angle(&mut self) -> Kind {

crates/oxc_parser/src/jsx/mod.rs

@@ -362,8 +362,8 @@ impl<'a> ParserImpl<'a> {
         if !self.at(Kind::Ident) && !self.cur_kind().is_all_keyword() {
             return Err(self.unexpected());
         }
-        // we are at a valid normal Ident or Keyword, let's keep on lexing for `-`
+        // Currently at a valid normal Ident or Keyword, keep on lexing for `-` in `<component-name />`
-        self.re_lex_jsx_identifier();
+        self.continue_lex_jsx_identifier();
         self.bump_any();
         let span = self.end_span(span);
         let name = span.source_text(self.source_text);

crates/oxc_parser/src/lexer/jsx.rs

@@ -2,7 +2,7 @@ use super::{Kind, Lexer, Token};
 use crate::diagnostics;
 
 use memchr::{memchr, memchr2};
-use oxc_syntax::identifier::{is_identifier_part, is_identifier_start};
+use oxc_syntax::identifier::is_identifier_part;
 
 impl<'a> Lexer<'a> {
     /// `JSXDoubleStringCharacters` ::
@@ -47,13 +47,6 @@ impl<'a> Lexer<'a> {
         self.finish_next(kind)
     }
 
-    /// Expand the current token for `JSXIdentifier`
-    pub(crate) fn next_jsx_identifier(&mut self, start_offset: u32) -> Token {
-        let kind = self.read_jsx_identifier(start_offset);
-        self.lookahead.clear();
-        self.finish_next(kind)
-    }
-
     /// [`JSXChild`](https://facebook.github.io/jsx/#prod-JSXChild)
     /// `JSXChild` :
     /// `JSXText`
@@ -90,25 +83,28 @@ impl<'a> Lexer<'a> {
         }
     }
 
+    /// Expand the current `Ident` token for `JSXIdentifier`
+    ///
+    /// The current character is at `Ident`, continue reading for `JSXIdentifier` if it has a `-`
+    ///
     /// `JSXIdentifier` :
     ///   `IdentifierStart`
     ///   `JSXIdentifier` `IdentifierPart`
     ///   `JSXIdentifier` [no `WhiteSpace` or Comment here] -
-    fn read_jsx_identifier(&mut self, _start_offset: u32) -> Kind {
+    pub(crate) fn continue_lex_jsx_identifier(&mut self) -> Option<Token> {
+        if self.peek() != Some('-') {
+            return None;
+        }
+        self.consume_char();
         while let Some(c) = self.peek() {
-            if c == '-' || is_identifier_start(c) {
+            if c == '-' || is_identifier_part(c) {
                 self.consume_char();
-                while let Some(c) = self.peek() {
-                    if is_identifier_part(c) {
-                        self.consume_char();
-                    } else {
-                        break;
-                    }
-                }
             } else {
                 break;
             }
         }
-        Kind::Ident
+        // Clear the current lookahead `Minus` Token
+        self.lookahead.clear();
+        Some(self.finish_next(Kind::Ident))
     }
 }

tasks/coverage/parser_typescript.snap

@@ -16182,14 +16182,6 @@ Expect to Parse: "conformance/salsa/plainJSRedeclare3.ts"
   × Invalid characters after number
     ╭─[conformance/jsx/tsxAttributeInvalidNames.tsx:12:10]
  11 │ // Invalid names
- 12 │ <test1 32data={32} />;
-    ·          ────
- 13 │ <test2 -data={32} />;
-    ╰────
-
-  × Invalid characters after number
-    ╭─[conformance/jsx/tsxAttributeInvalidNames.tsx:12:10]
- 11 │ // Invalid names
  12 │ <test1 32data={32} />;
     ·          ────
  13 │ <test2 -data={32} />;