[noz] Fix extraction on python 2.6 by means of using compat_xpath

youtube_dl/extractor/noz.py

@@ -2,7 +2,10 @@
 from __future__ import unicode_literals
 
 from .common import InfoExtractor
-from ..compat import compat_urllib_parse_unquote
+from ..compat import (
+    compat_urllib_parse_unquote,
+    compat_xpath,
+)
 from ..utils import (
     int_or_none,
     find_xpath_attr,
@@ -47,7 +50,7 @@ class NozIE(InfoExtractor):
         duration = int_or_none(xpath_text(
             doc, './/article/movie/file/duration'))
         formats = []
-        for qnode in doc.findall('.//article/movie/file/qualities/qual'):
+        for qnode in doc.findall(compat_xpath('.//article/movie/file/qualities/qual')):
             http_url_ele = find_xpath_attr(
                 qnode, './html_urls/video_url', 'format', 'video/mp4')
             http_url = http_url_ele.text if http_url_ele is not None else None