filesystem: fixing utf-8 issues switching to jinja

[treecutter.git] / treecutter / sitemap.py

diff --git a/treecutter/sitemap.py b/treecutter/sitemap.py
index 24b53fe5b4ab3bd84d65caca16cd1e1e07db54a6..0ecefe4278efae22c581d702bb7c513630e43700 100644 (file)
--- a/treecutter/sitemap.py
+++ b/treecutter/sitemap.py
@@ -1,5 +1,6 @@
 #!/usr/bin/python
 import os
+import codecs
 import re
 import shutil
 import gettext
@@ -30,17 +31,18 @@ class Sitemap():
     # representing the text on the site.
     # A link can have several pages in different languages.
     def add_link(self, link):
-        tokens = filter(None,re.split(r'(^/[\w-]*/|[\w-]*/)',link))
+        tokens = filter(None,re.split(r'(^/[\w:-]*$|^/[\w:-]*/|[\w:-]*/)',link,flags=re.UNICODE))
         self._tree.add(tokens,Link(link))
 
     def write_map(self):
-        f = open(self._file,'w')
-        f.write('\n'.join(link.link() for link in self._tree))
+        f = codecs.open(self._file,'w','utf-8')
+        s = '\n'.join(link.link() for link in self._tree)
+        f.write(s)
         f.close()
 
     def read_map(self):
         try:
-            f = open(self._file)
+            f = codecs.open(self._file, 'r', 'utf-8')
             sml = f.read().split()
             f.close()
             for line in sml:
@@ -113,10 +115,10 @@ class Sitemap():
             ln = self._isocode.xpath(isoxml)[0].get('name')
             if lang != 'en':
                 ln = self._tranlang[lang].gettext(ln)
-            p = link.link()
-            if p[-1] == '/':
-                p = p +'index'
-            p = p+'.'+l
+            p = unicode(link.link())
+            if p[-1] == u'/':
+                p = p +u'index'
+            p = p+u'.'+l
             li = html.li(html.a(ln.decode('utf-8'),
                                 href=self._subdir+p,hreflang=l))
             menu.append(li)