Generic: use compat_urllib_parse_unquote to prevent utf8 mangling

of the entire page in python 2.

-requires- fixed compat_urllib_parse_unquote

example - the following will save with a mangled playlist title,
 instead of the kanji for 'tsunami'. This affects all utf8encoded
 urls as well

youtube-dl -f18 -o '%(playlist_title)s-%(title)s.%(ext)s' \
  https://gist.githubusercontent.com/atomicdryad/fcb97465e6060fc519e1/raw/61c14c1e3a4985471dcf56c281d24d7e781a4e0e/tsunami.html

youtube_dl/extractor/generic.py

@@ -1115,7 +1115,7 @@ class GenericIE(InfoExtractor):
         # Sometimes embedded video player is hidden behind percent encoding
         # (e.g. https://github.com/rg3/youtube-dl/issues/2448)
         # Unescaping the whole page allows to handle those cases in a generic way
-        webpage = compat_urllib_parse.unquote(webpage)
+        webpage = compat_urllib_parse_unquote(webpage)
 
         # it's tempting to parse this further, but you would
         # have to take into account all the variations like