From 533a1a627e54e9bfb5acc420ce6a6e0c4882fa9c Mon Sep 17 00:00:00 2001
From: Will Thompson <will@willthompson.co.uk>
Date: Tue, 24 Nov 2015 18:05:27 +0000
Subject: [PATCH] ox.html.sanitize_fragment: documentation, tests

---
 ox/html.py | 15 +++++++++++++++
 1 file changed, 15 insertions(+)

diff --git a/ox/html.py b/ox/html.py
index d69c2df..202a036 100644
--- a/ox/html.py
+++ b/ox/html.py
@@ -393,6 +393,21 @@ def split_tags(string):
     return [item for sublist in zip(strings, tags) for item in sublist][:-1]
 
 def sanitize_fragment(html):
+    '''
+    Ensures that tags are closed (or not, as appropriate), attributes
+    are quoted, etc. Does not strip potentially-malicious HTML: use
+    sanitize_html() for that.
+
+    >>> sanitize_fragment(u'<span lang="en">')
+    u'<span lang="en"></span>'
+    >>> sanitize_fragment(u'<span lang=en></span>')
+    u'<span lang="en"></span>'
+    >>> sanitize_fragment(u'<br><br/></br>')
+    u'<br><br>'
+    >>> sanitize_fragment(u'<a href="javascript:alert()">foo</a>')
+    u'<a href="javascript:alert()">foo</a>'
+    '''
+
     '''
     #html5lib reorders arguments, so not usable
     import html5lib