Raphaël G. Git Repositories - youtubedl/blob - youtube_dl/extractor/livestream.py

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