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