youtube_dl/extractor/livestream.py

   1 import re
   2 import json
   3 import xml.etree.ElementTree
   4
   5 from .common import InfoExtractor
   6 from ..utils import (
   7     compat_urllib_parse_urlparse,
   8     compat_urlparse,
   9     get_meta_content,
  10     xpath_with_ns,
  11     ExtractorError,
  12 )
  13
  14
  15 class LivestreamIE(InfoExtractor):
  16     IE_NAME = u'livestream'
  17     _VALID_URL = r'http://new.livestream.com/.*?/(?P<event_name>.*?)(/videos/(?P<id>\d+))?/?$'
  18     _TEST = {
  19         u'url': u'http://new.livestream.com/CoheedandCambria/WebsterHall/videos/4719370',
  20         u'file': u'4719370.mp4',
  21         u'md5': u'0d2186e3187d185a04b3cdd02b828836',
  22         u'info_dict': {
  23             u'title': u'Live from Webster Hall NYC',
  24             u'upload_date': u'20121012',
  25         }
  26     }
  27
  28     def _extract_video_info(self, video_data):
  29         video_url = video_data.get('progressive_url_hd') or video_data.get('progressive_url')
  30         return {'id': video_data['id'],
  31                 'url': video_url,
  32                 'ext': 'mp4',
  33                 'title': video_data['caption'],
  34                 'thumbnail': video_data['thumbnail_url'],
  35                 'upload_date': video_data['updated_at'].replace('-','')[:8],
  36                 }
  37
  38     def _real_extract(self, url):
  39         mobj = re.match(self._VALID_URL, url)
  40         video_id = mobj.group('id')
  41         event_name = mobj.group('event_name')
  42         webpage = self._download_webpage(url, video_id or event_name)
  43
  44         if video_id is None:
  45             # This is an event page:
  46             config_json = self._search_regex(r'window.config = ({.*?});',
  47                 webpage, u'window config')
  48             info = json.loads(config_json)['event']
  49             videos = [self._extract_video_info(video_data['data'])
  50                 for video_data in info['feed']['data'] if video_data['type'] == u'video']
  51             return self.playlist_result(videos, info['id'], info['full_name'])
  52         else:
  53             og_video = self._og_search_video_url(webpage, name=u'player url')
  54             query_str = compat_urllib_parse_urlparse(og_video).query
  55             query = compat_urlparse.parse_qs(query_str)
  56             api_url = query['play_url'][0].replace('.smil', '')
  57             info = json.loads(self._download_webpage(api_url, video_id,
  58                                                      u'Downloading video info'))
  59             return self._extract_video_info(info)
  60
  61
  62 # The original version of Livestream uses a different system
  63 class LivestreamOriginalIE(InfoExtractor):
  64     IE_NAME = u'livestream:original'
  65     _VALID_URL = r'https?://www\.livestream\.com/(?P<user>[^/]+)/video\?.*?clipId=(?P<id>.*?)(&|$)'
  66     _TEST = {
  67         u'url': u'http://www.livestream.com/dealbook/video?clipId=pla_8aa4a3f1-ba15-46a4-893b-902210e138fb',
  68         u'info_dict': {
  69             u'id': u'pla_8aa4a3f1-ba15-46a4-893b-902210e138fb',
  70             u'ext': u'flv',
  71             u'title': u'Spark 1 (BitCoin) with Cameron Winklevoss & Tyler Winklevoss of Winklevoss Capital',
  72         },
  73         u'params': {
  74             # rtmp
  75             u'skip_download': True,
  76         },
  77     }
  78
  79     def _real_extract(self, url):
  80         mobj = re.match(self._VALID_URL, url)
  81         video_id = mobj.group('id')
  82         user = mobj.group('user')
  83         api_url = 'http://x{0}x.api.channel.livestream.com/2.0/clipdetails?extendedInfo=true&id={1}'.format(user, video_id)
  84
  85         api_response = self._download_webpage(api_url, video_id)
  86         info = xml.etree.ElementTree.fromstring(api_response.encode('utf-8'))
  87         item = info.find('channel').find('item')
  88         ns = {'media': 'http://search.yahoo.com/mrss'}
  89         thumbnail_url = item.find(xpath_with_ns('media:thumbnail', ns)).attrib['url']
  90         # Remove the extension and number from the path (like 1.jpg)
  91         path = self._search_regex(r'(user-files/.+)_.*?\.jpg$', thumbnail_url, u'path')
  92
  93         return {
  94             'id': video_id,
  95             'title': item.find('title').text,
  96             'url': 'rtmp://extondemand.livestream.com/ondemand',
  97             'play_path': 'mp4:trans/dv15/mogulus-{0}.mp4'.format(path),
  98             'ext': 'flv',
  99             'thumbnail': thumbnail_url,
 100         }