Make _str in tests correctly process unicode escapes.

opottone · Feb 21, 2015 · 174820b · scoder · Feb 21, 2015 · opottone
1 parent b1c7416
commit 174820b
Show file tree

Hide file tree

Showing 3 changed files with 12 additions and 2 deletions.
diff --git a/src/lxml/tests/common_imports.py b/src/lxml/tests/common_imports.py
@@ -142,9 +142,14 @@ def make_doctest(filename):
                 doctests, {}, os.path.basename(filename), filename, 0))
 else:
     # Python 2
+    unichr_escape = re.compile(r'\\u[0-9a-fA-F]{4}|\\U[0-9a-fA-F]{8}')
+
     from __builtin__ import unicode
     def _str(s, encoding="UTF-8"):
-        return unicode(s, encoding=encoding)
+        s = unicode(s, encoding=encoding)
+        return unichr_escape.sub(lambda x:
+                                     x.group(0).decode('unicode-escape'),
+                                 s)
     def _bytes(s, encoding="UTF-8"):
         return s
     from io import BytesIO

diff --git a/src/lxml/tests/test_etree.py b/src/lxml/tests/test_etree.py
@@ -3818,7 +3818,6 @@ def _writeElement(self, element, encoding='us-ascii', compression=0):
             data = zlib.decompress(data)
         return canonicalize(data)
 
-
 class _XIncludeTestCase(HelperTestCase):
     def test_xinclude_text(self):
         filename = fileInTestDir('test_broken.xml')

diff --git a/src/lxml/tests/test_unicode.py b/src/lxml/tests/test_unicode.py
@@ -28,6 +28,12 @@
 
 
 class UnicodeTestCase(HelperTestCase):
+    def test__str(self):
+        # test the testing framework, namely _str from common_imports
+        self.assertEqual(_str('\x10'), _str('\u0010'))
+        self.assertEqual(_str('\x10'), _str('\U00000010'))
+        self.assertEqual(_str('\u1234'), _str('\U00001234'))
+
     def test_unicode_xml(self):
         tree = etree.XML('<p>%s</p>' % uni)
         self.assertEqual(uni, tree.text)