Issues #814253, #9179: Group references and conditional group references now

work in lookbehind assertions in regular expressions.

Lib/re.py

@@ -351,10 +351,11 @@ class Scanner:
         s = sre_parse.Pattern()
         s.flags = flags
         for phrase, action in lexicon:
+            gid = s.opengroup()
             p.append(sre_parse.SubPattern(s, [
-                (SUBPATTERN, (len(p)+1, sre_parse.parse(phrase, flags))),
+                (SUBPATTERN, (gid, sre_parse.parse(phrase, flags))),
                 ]))
-        s.groups = len(p)+1
+            s.closegroup(gid, p[-1])
         p = sre_parse.SubPattern(s, [(BRANCH, (None, p))])
         self.scanner = sre_compile.compile(p)
     def scan(self, string):

Lib/sre_parse.py

@@ -69,12 +69,14 @@ class Pattern:
     # master pattern object.  keeps track of global attributes
     def __init__(self):
         self.flags = 0
-        self.open = []
-        self.groups = 1
         self.groupdict = {}
+        self.subpatterns = [None]  # group 0
+    @property
+    def groups(self):
+        return len(self.subpatterns)
     def opengroup(self, name=None):
         gid = self.groups
-        self.groups = gid + 1
+        self.subpatterns.append(None)
         if self.groups > MAXGROUPS:
             raise error("groups number is too large")
         if name is not None:
@@ -83,12 +85,11 @@ class Pattern:
                 raise error("redefinition of group name %s as group %d; "
                             "was group %d" % (repr(name), gid,  ogid))
             self.groupdict[name] = gid
-        self.open.append(gid)
         return gid
-    def closegroup(self, gid):
+    def closegroup(self, gid, p):
-        self.open.remove(gid)
+        self.subpatterns[gid] = p
     def checkgroup(self, gid):
-        return gid < self.groups and gid not in self.open
+        return gid < self.groups and self.subpatterns[gid] is not None
 
 class SubPattern:
     # a subpattern, in intermediate form
@@ -184,7 +185,21 @@ class SubPattern:
             elif op in _UNITCODES:
                 lo = lo + 1
                 hi = hi + 1
-            elif op == SUCCESS:
+            elif op is GROUPREF:
+                i, j = self.pattern.subpatterns[av].getwidth()
+                lo = lo + i
+                hi = hi + j
+            elif op is GROUPREF_EXISTS:
+                i, j = av[1].getwidth()
+                if av[2] is not None:
+                    l, h = av[2].getwidth()
+                    i = min(i, l)
+                    j = max(j, h)
+                else:
+                    i = 0
+                lo = lo + i
+                hi = hi + j
+            elif op is SUCCESS:
                 break
         self.width = min(lo, MAXREPEAT - 1), min(hi, MAXREPEAT)
         return self.width
@@ -705,7 +720,7 @@ def _parse(source, state):
                 if not sourcematch(")"):
                     raise error("unbalanced parenthesis")
                 if group is not None:
-                    state.closegroup(group)
+                    state.closegroup(group, p)
                 subpatternappend((SUBPATTERN, (group, p)))
             else:
                 while True:

Lib/test/test_re.py

@@ -575,7 +575,7 @@ class ReTests(unittest.TestCase):
         self.assertEqual(re.match("a.*b", "a\n\nb", re.DOTALL).group(0),
                          "a\n\nb")
 
-    def test_non_consuming(self):
+    def test_lookahead(self):
         self.assertEqual(re.match("(a(?=\s[^a]))", "a b").group(1), "a")
         self.assertEqual(re.match("(a(?=\s[^a]*))", "a b").group(1), "a")
         self.assertEqual(re.match("(a(?=\s[abc]))", "a b").group(1), "a")
@@ -589,6 +589,42 @@ class ReTests(unittest.TestCase):
         self.assertEqual(re.match(r"(a)(?!\s\1)", "a b").group(1), "a")
         self.assertEqual(re.match(r"(a)(?!\s(abc|a))", "a b").group(1), "a")
 
+        # Group reference.
+        self.assertTrue(re.match(r'(a)b(?=\1)a', 'aba'))
+        self.assertIsNone(re.match(r'(a)b(?=\1)c', 'abac'))
+        # Conditional group reference.
+        self.assertTrue(re.match('(?:(a)|(x))b(?=(?(2)x|c))c', 'abc'))
+        self.assertIsNone(re.match('(?:(a)|(x))b(?=(?(2)c|x))c', 'abc'))
+        self.assertTrue(re.match('(?:(a)|(x))b(?=(?(2)x|c))c', 'abc'))
+        self.assertIsNone(re.match('(?:(a)|(x))b(?=(?(1)b|x))c', 'abc'))
+        self.assertTrue(re.match('(?:(a)|(x))b(?=(?(1)c|x))c', 'abc'))
+        # Group used before defined.
+        self.assertTrue(re.match('(a)b(?=(?(2)x|c))(c)', 'abc'))
+        self.assertIsNone(re.match('(a)b(?=(?(2)b|x))(c)', 'abc'))
+        self.assertTrue(re.match('(a)b(?=(?(1)c|x))(c)', 'abc'))
+
+    def test_lookbehind(self):
+        self.assertTrue(re.match('ab(?<=b)c', 'abc'))
+        self.assertIsNone(re.match('ab(?<=c)c', 'abc'))
+        self.assertIsNone(re.match('ab(?<!b)c', 'abc'))
+        self.assertTrue(re.match('ab(?<!c)c', 'abc'))
+        # Group reference.
+        self.assertTrue(re.match(r'(a)a(?<=\1)c', 'aac'))
+        self.assertIsNone(re.match(r'(a)b(?<=\1)a', 'abaa'))
+        self.assertIsNone(re.match(r'(a)a(?<!\1)c', 'aac'))
+        self.assertTrue(re.match(r'(a)b(?<!\1)a', 'abaa'))
+        # Conditional group reference.
+        self.assertIsNone(re.match('(?:(a)|(x))b(?<=(?(2)x|c))c', 'abc'))
+        self.assertIsNone(re.match('(?:(a)|(x))b(?<=(?(2)b|x))c', 'abc'))
+        self.assertTrue(re.match('(?:(a)|(x))b(?<=(?(2)x|b))c', 'abc'))
+        self.assertIsNone(re.match('(?:(a)|(x))b(?<=(?(1)c|x))c', 'abc'))
+        self.assertTrue(re.match('(?:(a)|(x))b(?<=(?(1)b|x))c', 'abc'))
+        # Group used before defined.
+        self.assertIsNone(re.match('(a)b(?<=(?(2)x|c))(c)', 'abc'))
+        self.assertIsNone(re.match('(a)b(?<=(?(2)b|x))(c)', 'abc'))
+        self.assertIsNone(re.match('(a)b(?<=(?(1)c|x))(c)', 'abc'))
+        self.assertTrue(re.match('(a)b(?<=(?(1)b|x))(c)', 'abc'))
+
     def test_ignore_case(self):
         self.assertEqual(re.match("abc", "ABC", re.I).group(0), "ABC")
         self.assertEqual(re.match(b"abc", b"ABC", re.I).group(0), b"ABC")

Misc/NEWS

@@ -183,6 +183,9 @@ Core and Builtins
 Library
 -------
 
+- Issues #814253, #9179: Group references and conditional group references now
+  work in lookbehind assertions in regular expressions.
+
 - Issue #22406: Fixed the uu_codec codec incorrectly ported to 3.x.
   Based on patch by Martin Panter.