From 80d7156427814042d6beb55714858d8c8b5b1796 Mon Sep 17 00:00:00 2001
From: Martin Paljak <martin@martinpaljak.net>
Date: Sat, 28 Dec 2013 20:16:50 +0000
Subject: [PATCH] Add support for unicode slugs

---
 pelican/utils.py | 32 +++++++++++++++++++-------------
 1 file changed, 19 insertions(+), 13 deletions(-)

diff --git a/pelican/utils.py b/pelican/utils.py
index 4b25ec7f..ebff1e5f 100644
--- a/pelican/utils.py
+++ b/pelican/utils.py
@@ -227,7 +227,7 @@ def pelican_open(filename):
     yield content
 
 
-def slugify(value, substitutions=()):
+def slugify(value, substitutions=(), use_unicode=True):
     """
     Normalizes string, converts to lowercase, removes non-alpha characters,
     and converts spaces to hyphens.
@@ -236,24 +236,30 @@ def slugify(value, substitutions=()):
     """
     # TODO Maybe steal again from current Django 1.5dev
     value = Markup(value).striptags()
+
     # value must be unicode per se
-    import unicodedata
-    from unidecode import unidecode
-    # unidecode returns str in Py2 and 3, so in Py2 we have to make
-    # it unicode again
-    value = unidecode(value)
-    if isinstance(value, six.binary_type):
-        value = value.decode('ascii')
+    if not use_unicode:
+        from unidecode import unidecode
+        # unidecode returns str in Py2 and 3, so in Py2 we have to make
+        # it unicode again
+        value = unidecode(value)
+        if isinstance(value, six.binary_type):
+            value = value.decode('ascii')
+
     # still unicode
-    value = unicodedata.normalize('NFKD', value).lower()
+    import unicodedata
+    value = unicodedata.normalize('NFKC' if use_unicode else 'NFKD', value).lower()
     for src, dst in substitutions:
         value = value.replace(src.lower(), dst.lower())
     value = re.sub('[^\w\s-]', '', value).strip()
     value = re.sub('[-\s]+', '-', value)
-    # we want only ASCII chars
-    value = value.encode('ascii', 'ignore')
-    # but Pelican should generally use only unicode
-    return value.decode('ascii')
+    if use_unicode:
+        return value
+    else:
+        # we want only ASCII chars
+        value = value.encode('ascii', 'ignore')
+        # but Pelican should generally use only unicode
+        return value.decode('ascii')
 
 
 def copy(path, source, destination, destination_path=None):