proper handling of 'rel' attribute for links (Bug 971754)

2 files changed, 17 insertions, 1 deletions

diff --git a/src/lxml/html/clean.py b/src/lxml/html/clean.py
index dabc1257..e94eec26 100644
--- a/src/lxml/html/clean.py
+++ b/src/lxml/html/clean.py
@@ -399,7 +399,10 @@ class Cleaner(object):
         if self.add_nofollow:
             for el in _find_external_links(doc):
                 if not self.allow_follow(el):
-                    el.set('rel', 'nofollow')
+                    rel = 'nofollow'
+                    if el.get('rel'):
+                        rel = el.get('rel') + ' ' + rel
+                    el.set('rel', rel)
 
     def allow_follow(self, anchor):
         """
diff --git a/src/lxml/html/tests/test_clean.txt b/src/lxml/html/tests/test_clean.txt
index d87a6619..21bd12c6 100644
--- a/src/lxml/html/tests/test_clean.txt
+++ b/src/lxml/html/tests/test_clean.txt
@@ -27,6 +27,8 @@
 ...       Password: <input type="password" name="password">
 ...     </form>
 ...     <a href="evil-site">spam spam SPAM!</a>
+...     <a href="http://example.com" rel="author">Author</a>
+...     <a href="http://example.com">Text</a>
 ...     <img src="evil!">
 ...   </body>
 ... </html>'''
@@ -57,6 +59,8 @@
       Password: <input type="password" name="password">
     </form>
     <a href="evil-site">spam spam SPAM!</a>
+    <a href="http://example.com" rel="author">Author</a>
+    <a href="http://example.com">Text</a>
     <img src="evil!">
   </body>
 </html>
@@ -87,6 +91,8 @@
       Password: <input type="password" name="password">
     </form>
     <a href="evil-site">spam spam SPAM!</a>
+    <a href="http://example.com" rel="author">Author</a>
+    <a href="http://example.com">Text</a>
     <img src="evil!">
   </body>
 </html>
@@ -105,6 +111,8 @@
     of EVIL!
     Password:
     <a href="evil-site">spam spam SPAM!</a>
+    <a href="http://example.com" rel="author">Author</a>
+    <a href="http://example.com">Text</a>
     <img src="evil!">
   </body>
 </html>
@@ -122,6 +130,8 @@
     of EVIL!
     Password:
     <a href="evil-site" rel="nofollow">spam spam SPAM!</a>
+    <a href="http://example.com" rel="author nofollow">Author</a>
+    <a href="http://example.com" rel="nofollow">Text</a>
     <img src="evil!">
   </body>
 </html>
@@ -143,6 +153,9 @@
     of EVIL!
     Password:
     <a href="evil-site">spam spam SPAM!</a>
+    <a href="http://example.com" rel="author">Author</a>
+    <a href="http://example.com">Text</a>
     <img src="evil!">
   </body>
 </html>
+