Patch # 1140 (my code, approved by Effbot).

Make sure the type of the return value of re.sub(x, y, z) is the type
of y+x (i.e. unicode if either is unicode, str if they are both str)
even if there are no substitutions or if x==z (which triggered various
special cases in join_list()).

Could be backported to 2.5; no need to port to 3.0.

1 files changed, 25 insertions, 0 deletions

diff --git a/Lib/test/test_re.py b/Lib/test/test_re.py
index cfb949c..aa403ba 100644
--- a/Lib/test/test_re.py
+++ b/Lib/test/test_re.py
@@ -83,6 +83,31 @@ class ReTests(unittest.TestCase):
         self.assertEqual(re.sub('\r\n', '\n', 'abc\r\ndef\r\n'),
                          'abc\ndef\n')
 
+    def test_bug_1140(self):
+        # re.sub(x, y, u'') should return u'', not '', and
+        # re.sub(x, y, '') should return '', not u''.
+        # Also:
+        # re.sub(x, y, unicode(x)) should return unicode(y), and
+        # re.sub(x, y, str(x)) should return
+        #     str(y) if isinstance(y, str) else unicode(y).
+        for x in 'x', u'x':
+            for y in 'y', u'y':
+                z = re.sub(x, y, u'')
+                self.assertEqual(z, u'')
+                self.assertEqual(type(z), unicode)
+                #
+                z = re.sub(x, y, '')
+                self.assertEqual(z, '')
+                self.assertEqual(type(z), str)
+                #
+                z = re.sub(x, y, unicode(x))
+                self.assertEqual(z, y)
+                self.assertEqual(type(z), unicode)
+                #
+                z = re.sub(x, y, str(x))
+                self.assertEqual(z, y)
+                self.assertEqual(type(z), type(y))
+
     def test_sub_template_numeric_escape(self):
         # bug 776311 and friends
         self.assertEqual(re.sub('x', r'\0', 'x'), '\0')