update ox.web.youtube

ox/web/youtube.py

@@ -5,19 +5,26 @@ import urllib2
 import cookielib
 import re
 from xml.dom.minidom import parseString
+import json
 
 import feedparser
 import ox
 from ox.cache import read_url, cache_timeout
 
 
+def get_id(url):
+    match = re.compile('v=(.+?)($|&)').findall(url)
+    if match:
+        return match[0][0]
+
+def get_url(id):
+    return 'http://www.youtube.com/watch?v=%s' % id
+
 def video_url(youtubeId, format='mp4', timeout=cache_timeout):
     """
         youtubeId - if of video
         format - video format, options: webm, 1080p, 720p, mp4, high
     """
-    def get_url(stream):
-        return '%s&signature=%s' % (stream['url'], stream['sig'])
     fmt = None
     if format == '4k':
         fmt=38
@@ -31,11 +38,11 @@ def video_url(youtubeId, format='mp4', timeout=cache_timeout):
         fmt=35
     elif format == 'webm':
         streams = videos(youtubeId, 'webm')
-        return get_url(streams[max(streams.keys())])
+        return streams[max(streams.keys())]['stream_url']
 
     streams = videos(youtubeId)
     if str(fmt) in streams:
-        return get_url(streams[str(fmt)])
+        return streams[str(fmt)]['stream_url']
 
 def find(query, max_results=10, offset=1, orderBy='relevance'):
     query = quote(query)
@@ -54,10 +61,15 @@ def find(query, max_results=10, offset=1, orderBy='relevance'):
 
 def info(id):
     info = {}
+    if id.startswith('http'):
+        id = get_id(id)
+        if not id:
+            return info
     url = "http://gdata.youtube.com/feeds/api/videos/%s?v=2" % id
     data = read_url(url)
     xml = parseString(data)
-    info['url'] = 'http://www.youtube.com/watch?v=%s' % id
+    info['id'] = id
+    info['url'] = get_url(id)
     info['title'] = xml.getElementsByTagName('title')[0].firstChild.data
     info['description'] = xml.getElementsByTagName('media:description')[0].firstChild.data
     info['date'] = xml.getElementsByTagName('published')[0].firstChild.data.split('T')[0]
@@ -70,14 +82,13 @@ def info(id):
     k = xml.getElementsByTagName('media:keywords')[0].firstChild
     if k:
         info['keywords'] = k.data.split(', ')
-    url = "http://www.youtube.com/watch?v=%s" % id
-    data = read_url(url)
+    data = read_url(info['url'])
     match = re.compile('<h4>License:</h4>(.*?)</p>', re.DOTALL).findall(data)
     if match:
         info['license'] = match[0].strip()
         info['license'] = re.sub('<.+?>', '', info['license']).strip()
 
-    url = "http://www.youtube.com/api/timedtext?hl=en&type=list&tlangs=1&v=%s&asrs=1"%id
+    url = "http://www.youtube.com/api/timedtext?hl=en&type=list&tlangs=1&v=%s&asrs=1" % id
     data = read_url(url)
     xml = parseString(data)
     languages = [t.getAttribute('lang_code') for t in xml.getElementsByTagName('track')]
@@ -113,12 +124,15 @@ def videos(id, format=''):
     data = read_url(url)
     match = re.compile('"url_encoded_fmt_stream_map": "(.*?)"').findall(data)
     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
     return streams
@@ -172,3 +186,15 @@ def download_webm(id, filename):
     f.close()
     u.close()
     return filename
+
+def get_config(id):
+    if id.startswith('http'):
+        url = id
+    else:
+        url = get_url(id)
+    data = read_url(url)
+    match = re.compile('ytplayer.config = (.*?);<').findall(data)
+    if match:
+        config = json.load(match[0])
+    return config
+