Fix backtracking string regexes in JavascriptLexer und TypescriptLexer.

fixes #1637
pygments · Dec 17, 2020 · 7958cd6 · 7958cd6
1 parent a507918
commit 7958cd6
Show file tree

Hide file tree

Showing 3 changed files with 28 additions and 6 deletions.
diff --git a/CHANGES b/CHANGES
@@ -17,6 +17,7 @@ Version 2.7.4
 (not released yet)
 
 - Fixed infinite loop in SML lexer (#1625)
+- Fixed backtracking string regexes in JavaScript/TypeScript lexers (#1637)
 
 
 Version 2.7.3

diff --git a/pygments/lexers/javascript.py b/pygments/lexers/javascript.py
@@ -95,10 +95,20 @@ class JavascriptLexer(RegexLexer):
              r'Error|eval|isFinite|isNaN|isSafeInteger|parseFloat|parseInt|'
              r'document|this|window)\b', Name.Builtin),
             (JS_IDENT, Name.Other),
-            (r'"(\\\\|\\"|[^"])*"', String.Double),
-            (r"'(\\\\|\\'|[^'])*'", String.Single),
+            (r'"', String.Double, 'string-double'),
+            (r"'", String.Single, 'string-single'),
             (r'`', String.Backtick, 'interp'),
         ],
+        'string-double': [
+            (r'\\.', String.Double),
+            (r'[^\\"]+', String.Double),
+            (r'"', String.Double, '#pop'),
+        ],
+        'string-single': [
+            (r'\\.', String.Single),
+            (r"[^\\']+", String.Single),
+            (r"'", String.Single, '#pop'),
+        ],
         'interp': [
             (r'`', String.Backtick, '#pop'),
             (r'\\\\', String.Backtick),
@@ -112,7 +122,6 @@ class JavascriptLexer(RegexLexer):
             (r'\}', String.Interpol, '#pop'),
             include('root'),
         ],
-        # (\\\\|\\`|[^`])*`', String.Backtick),
     }
 
 
@@ -522,12 +531,22 @@ class TypeScriptLexer(RegexLexer):
             (r'[0-9][0-9]*\.[0-9]+([eE][0-9]+)?[fd]?', Number.Float),
             (r'0x[0-9a-fA-F]+', Number.Hex),
             (r'[0-9]+', Number.Integer),
-            (r'"(\\\\|\\"|[^"])*"', String.Double),
-            (r"'(\\\\|\\'|[^'])*'", String.Single),
+            (r'"', String.Double, 'string-double'),
+            (r"'", String.Single, 'string-single'),
             (r'`', String.Backtick, 'interp'),
             # Match stuff like: Decorators
             (r'@\w+', Keyword.Declaration),
         ],
+        'string-double': [
+            (r'\\.', String.Double),
+            (r'[^\\"]+', String.Double),
+            (r'"', String.Double, '#pop'),
+        ],
+        'string-single': [
+            (r'\\.', String.Single),
+            (r"[^\\']+", String.Single),
+            (r"'", String.Single, '#pop'),
+        ],
 
         # The 'interp*' rules match those in JavascriptLexer. Changes made
         # there should be reflected here as well.

diff --git a/tests/test_html_lexer.py b/tests/test_html_lexer.py
@@ -65,7 +65,9 @@ def test_long_unclosed_javascript_fragment(lexer_html):
     tokens_body = [
         (Token.Name.Other, 'alert'),
         (Token.Punctuation, '('),
-        (Token.Literal.String.Double, '"hi"'),
+        (Token.Literal.String.Double, '"'),
+        (Token.Literal.String.Double, 'hi'),
+        (Token.Literal.String.Double, '"'),
         (Token.Punctuation, ')'),
         (Token.Punctuation, ';'),
     ]