fix ox.web.youtube

ox/web/youtube.py

@@ -38,11 +38,27 @@ def video_url(youtubeId, format='mp4', timeout=cache_timeout):
         fmt=35
     elif format == 'webm':
         streams = videos(youtubeId, 'webm')
-        return streams[max(streams.keys())]['stream_url']
+        return streams[max(streams.keys())]['url']
 
     streams = videos(youtubeId)
     if str(fmt) in streams:
-        return streams[str(fmt)]['stream_url']
+        return streams[str(fmt)]['url']
+
+def get_video_info(id):
+    eurl = get_url(id)
+    data = read_url(eurl)
+    t = re.compile('\W[\'"]?t[\'"]?: ?[\'"](.+?)[\'"]').findall(data)
+    if t:
+        t = t[0]
+    else:
+        raise IOError
+    url = "http://www.youtube.com/get_video_info?&video_id=%s&el=$el&ps=default&eurl=%s&hl=en_US&t=%s" % (id, quote(eurl), quote(t))
+    data = read_url(url)
+    info = {}
+    for part in data.split('&'):
+        key, value = part.split('=')
+        info[key] = unquote_plus(value).replace('+', ' ')
+    return info
 
 def find(query, max_results=10, offset=1, orderBy='relevance'):
     query = quote(query)
@@ -121,21 +137,20 @@ def videos(id, format=''):
         'webm': 'video/webm',
         'mp4': 'video/mp4'
     }.get(format)
-    url = "http://www.youtube.com/watch?v=%s" % id
-    data = read_url(url)
-    match = re.compile('"url_encoded_fmt_stream_map": "(.*?)"').findall(data)
+    info = get_video_info(id)
+    stream_map = info['url_encoded_fmt_stream_map']
     streams = {}
-    if match:
-        for x in match[0].split(','):
-            stream = {}
-            for s in x.split('\\u0026'):
-                key, value = s.split('=')
-                value = unquote_plus(value)
-                stream[key] = value
-            if 'url' in stream and 'sig' in stream:
-                stream['stream_url'] = '%s&signature=%s' % (stream['url'], stream['sig'])
-            if not stream_type or stream['type'].startswith(stream_type):
-                streams[stream['itag']] = stream
+    for x in stream_map.split(','):
+        stream = {}
+        #for s in x.split('\\u0026'):
+        for s in x.split('&'):
+            key, value = s.split('=')
+            value = unquote_plus(value)
+            stream[key] = value
+        if 'url' in stream and 'sig' in stream:
+            stream['url'] = '%s&signature=%s' % (stream['url'], stream['sig'])
+        if not stream_type or stream['type'].startswith(stream_type):
+            streams[stream['itag']] = stream
     return streams
 
 def playlist(url):
@@ -173,7 +188,9 @@ def download_webm(id, filename):
             streams[stream['itag']] = stream
     if streams:
         s = max(streams.keys())
-        url = '%s&signature=%s' % (streams[s]['url'], streams[s]['sig'])
+        url = streams[s]['url']
+        if 'sig' in streams[s]:
+            url += 'signature=' + streams[s]['sig']
     else:
         return None
 
@@ -198,4 +215,3 @@ def get_config(id):
     if match:
         config = json.load(match[0])
     return config
-