[textutils] use NFKD decomposition in unormalize()

The normal form KD (NFKD) will apply the compatibility decomposition, i.e.
replace all compatibility characters with their equivalents.

1 files changed, 9 insertions, 1 deletions

diff --git a/textutils.py b/textutils.py
index db69d3b..cf4deb2 100644
--- a/textutils.py
+++ b/textutils.py
@@ -73,6 +73,14 @@ MANUAL_UNICODE_MAP = {
 
 def unormalize(ustring, ignorenonascii=False):
     """replace diacritical characters with their corresponding ascii characters
+    
+    Convert the unicode string to its long normalized form (unicode character
+    will be transform into several characters) and keep the first one only.
+    The normal form KD (NFKD) will apply the compatibility decomposition, i.e.
+    replace all compatibility characters with their equivalents.
+    
+    :see: Another project about ASCII transliterations of Unicode text
+          http://pypi.python.org/pypi/Unidecode
     """
     res = []
     for letter in ustring[:]:
@@ -83,7 +91,7 @@ def unormalize(ustring, ignorenonascii=False):
                 if ignorenonascii:
                     continue
                 raise ValueError("can't deal with non-ascii based characters")
-            replacement = _uninormalize('NFD', letter)[0]
+            replacement = _uninormalize('NFKD', letter)[0]
         res.append(replacement)
     return u''.join(res)