[utils] Fix inconsistent output of clean_html

`\s` in Python 2.x doesn't match unicode whitespace characters by
default

diff --git a/youtube_dl/utils.py b/youtube_dl/utils.py
index 91e235ff2f6166106c93e4b8de5bbeb18a6d8b7a..41bc205446a7594ece9f14dc43b0617808962053 100644 (file)
--- a/youtube_dl/utils.py
+++ b/youtube_dl/utils.py
@@ -421,8 +421,8 @@ def clean_html(html):
 
     # Newline vs <br />
     html = html.replace('\n', ' ')
-    html = re.sub(r'\s*<\s*br\s*/?\s*>\s*', '\n', html)
-    html = re.sub(r'<\s*/\s*p\s*>\s*<\s*p[^>]*>', '\n', html)
+    html = re.sub(r'(?u)\s*<\s*br\s*/?\s*>\s*', '\n', html)
+    html = re.sub(r'(?u)<\s*/\s*p\s*>\s*<\s*p[^>]*>', '\n', html)
     # Strip html tags
     html = re.sub('<.*?>', '', html)
     # Replace html entities