unit tests to unic nfc and nfd handling

jussimalinen · jussimalinen · commit feefd4070849 · 2010-05-05T11:39:32.000Z
diff --git a/src/robot/utils/unic.py b/src/robot/utils/unic.py
@@ -19,12 +19,12 @@
 def unic(item, *args):
     # Based on a recipe from http://code.activestate.com/recipes/466341
     try:
-        return unicodedata.normalize("NFC",unicode(item, *args))
+        return unicodedata.normalize('NFC', unicode(item, *args))
     except UnicodeError:
         try:
             ascii_text = str(item).encode('string_escape')
         except UnicodeError:
-            return "<unrepresentable object '%s'>" % item.__class__.__name__
+            return u"<unrepresentable object '%s'>" % item.__class__.__name__
         else:
             return unicode(ascii_text)
 
diff --git a/utest/utils/test_unic.py b/utest/utils/test_unic.py
@@ -1,4 +1,5 @@
 import unittest
+import unicodedata
 from robot.utils import unic, is_jython
 from robot.utils.asserts import assert_equals, assert_true
 if is_jython:
@@ -8,7 +9,7 @@
 
 
 if is_jython:
-    class TestUnic(unittest.TestCase):
+    class TestJavaUnic(unittest.TestCase):
 
         def test_with_java_object(self):
             data = u'This is unicode \xe4\xf6'
@@ -29,6 +30,13 @@ def test_with_iterator(self):
 
 class TestUnic(unittest.TestCase):
 
+    def test_unicode_nfc_and_nfd_decomposition_equality(self):
+        text = u'Hyv\xe4'
+        assert_equals(unic(unicodedata.normalize('NFC', text)), text)
+        # In Mac filesystem umlaut characters are presented in NFD-format.
+        # This is to check that unic normalizes all strings to NFC 
+        assert_equals(unic(unicodedata.normalize('NFD', text)), text)
+
     def test_object_containing_unicode_repr(self):
         assert_equals(unic(UnicodeRepr()), u'Hyv\xe4')
 
