fixes #1625: infinite loop in SML lexer

Reason was a lookahead-only pattern which was included in the state
where the lookahead was transitioning to.

2 files changed, 14 insertions, 6 deletions

diff --git a/CHANGES b/CHANGES
index 5e0c8683..fbfeeb1c 100644
--- a/CHANGES
+++ b/CHANGES
@@ -11,6 +11,14 @@ Version 2.8.0
 -------------
 (not released yet)
 
+
+Version 2.7.4
+-------------
+(not released yet)
+
+- Fixed infinite loop in SML lexer (#1625)
+
+
 Version 2.7.3
 -------------
 (released December 6, 2020)
diff --git a/pygments/lexers/ml.py b/pygments/lexers/ml.py
index 8ca8ce3e..f2ac367c 100644
--- a/pygments/lexers/ml.py
+++ b/pygments/lexers/ml.py
@@ -142,7 +142,7 @@ class SMLLexer(RegexLexer):
             (r'#\s+(%s)' % symbolicid_re, Name.Label),
             # Some reserved words trigger a special, local lexer state change
             (r'\b(datatype|abstype)\b(?!\')', Keyword.Reserved, 'dname'),
-            (r'(?=\b(exception)\b(?!\'))', Text, ('ename')),
+            (r'\b(exception)\b(?!\')', Keyword.Reserved, 'ename'),
             (r'\b(functor|include|open|signature|structure)\b(?!\')',
              Keyword.Reserved, 'sname'),
             (r'\b(type|eqtype)\b(?!\')', Keyword.Reserved, 'tname'),
@@ -315,15 +315,14 @@ class SMLLexer(RegexLexer):
         'ename': [
             include('whitespace'),
 
-            (r'(exception|and)\b(\s+)(%s)' % alphanumid_re,
+            (r'(and\b)(\s+)(%s)' % alphanumid_re,
              bygroups(Keyword.Reserved, Text, Name.Class)),
-            (r'(exception|and)\b(\s*)(%s)' % symbolicid_re,
+            (r'(and\b)(\s*)(%s)' % symbolicid_re,
              bygroups(Keyword.Reserved, Text, Name.Class)),
             (r'\b(of)\b(?!\')', Keyword.Reserved),
+            (r'(%s)|(%s)' % (alphanumid_re, symbolicid_re), Name.Class),
 
-            include('breakout'),
-            include('core'),
-            (r'\S+', Error),
+            default('#pop'),
         ],
 
         'datcon': [
@@ -445,6 +444,7 @@ class OcamlLexer(RegexLexer):
         ],
     }
 
+
 class OpaLexer(RegexLexer):
     """
     Lexer for the Opa language (http://opalang.org).