youtube_dl/extractor/livestream.py

   1 from __future__ import unicode_literals
   2
   3 import re
   4 import json
   5 import itertools
   6
   7 from .common import InfoExtractor
   8 from ..compat import (
   9     compat_str,
  10     compat_urllib_parse_urlparse,
  11     compat_urlparse,
  12 )
  13 from ..utils import (
  14     ExtractorError,
  15     find_xpath_attr,
  16     int_or_none,
  17     orderedSet,
  18     xpath_with_ns,
  19 )
  20
  21
  22 class LivestreamIE(InfoExtractor):
  23     IE_NAME = 'livestream'
  24     _VALID_URL = r'https?://(?:new\.)?livestream\.com/.*?/(?P<event_name>.*?)(/videos/(?P<id>[0-9]+)(?:/player)?)?/?(?:$|[?#])'
  25     _TESTS = [{
  26         'url': 'http://new.livestream.com/CoheedandCambria/WebsterHall/videos/4719370',
  27         'md5': '53274c76ba7754fb0e8d072716f2292b',
  28         'info_dict': {
  29             'id': '4719370',
  30             'ext': 'mp4',
  31             'title': 'Live from Webster Hall NYC',
  32             'upload_date': '20121012',
  33             'like_count': int,
  34             'view_count': int,
  35             'thumbnail': 're:^http://.*\.jpg$'
  36         }
  37     }, {
  38         'url': 'http://new.livestream.com/tedx/cityenglish',
  39         'info_dict': {
  40             'title': 'TEDCity2.0 (English)',
  41             'id': '2245590',
  42         },
  43         'playlist_mincount': 4,
  44     }, {
  45         'url': 'http://new.livestream.com/chess24/tatasteelchess',
  46         'info_dict': {
  47             'title': 'Tata Steel Chess',
  48             'id': '3705884',
  49         },
  50         'playlist_mincount': 60,
  51     }, {
  52         'url': 'https://new.livestream.com/accounts/362/events/3557232/videos/67864563/player?autoPlay=false&height=360&mute=false&width=640',
  53         'only_matching': True,
  54     }, {
  55         'url': 'http://livestream.com/bsww/concacafbeachsoccercampeonato2015',
  56         'only_matching': True,
  57     }]
  58
  59     def _parse_smil(self, video_id, smil_url):
  60         formats = []
  61         _SWITCH_XPATH = (
  62             './/{http://www.w3.org/2001/SMIL20/Language}body/'
  63             '{http://www.w3.org/2001/SMIL20/Language}switch')
  64         smil_doc = self._download_xml(
  65             smil_url, video_id,
  66             note='Downloading SMIL information',
  67             errnote='Unable to download SMIL information',
  68             fatal=False)
  69         if smil_doc is False:  # Download failed
  70             return formats
  71         title_node = find_xpath_attr(
  72             smil_doc, './/{http://www.w3.org/2001/SMIL20/Language}meta',
  73             'name', 'title')
  74         if title_node is None:
  75             self.report_warning('Cannot find SMIL id')
  76             switch_node = smil_doc.find(_SWITCH_XPATH)
  77         else:
  78             title_id = title_node.attrib['content']
  79             switch_node = find_xpath_attr(
  80                 smil_doc, _SWITCH_XPATH, 'id', title_id)
  81         if switch_node is None:
  82             raise ExtractorError('Cannot find switch node')
  83         video_nodes = switch_node.findall(
  84             '{http://www.w3.org/2001/SMIL20/Language}video')
  85
  86         for vn in video_nodes:
  87             tbr = int_or_none(vn.attrib.get('system-bitrate'))
  88             furl = (
  89                 'http://livestream-f.akamaihd.net/%s?v=3.0.3&fp=WIN%%2014,0,0,145' %
  90                 (vn.attrib['src']))
  91             if 'clipBegin' in vn.attrib:
  92                 furl += '&ssek=' + vn.attrib['clipBegin']
  93             formats.append({
  94                 'url': furl,
  95                 'format_id': 'smil_%d' % tbr,
  96                 'ext': 'flv',
  97                 'tbr': tbr,
  98                 'preference': -1000,
  99             })
 100         return formats
 101
 102     def _extract_video_info(self, video_data):
 103         video_id = compat_str(video_data['id'])
 104
 105         FORMAT_KEYS = (
 106             ('sd', 'progressive_url'),
 107             ('hd', 'progressive_url_hd'),
 108         )
 109         formats = [{
 110             'format_id': format_id,
 111             'url': video_data[key],
 112             'quality': i + 1,
 113         } for i, (format_id, key) in enumerate(FORMAT_KEYS)
 114             if video_data.get(key)]
 115
 116         smil_url = video_data.get('smil_url')
 117         if smil_url:
 118             formats.extend(self._parse_smil(video_id, smil_url))
 119         self._sort_formats(formats)
 120
 121         return {
 122             'id': video_id,
 123             'formats': formats,
 124             'title': video_data['caption'],
 125             'thumbnail': video_data.get('thumbnail_url'),
 126             'upload_date': video_data['updated_at'].replace('-', '')[:8],
 127             'like_count': video_data.get('likes', {}).get('total'),
 128             'view_count': video_data.get('views'),
 129         }
 130
 131     def _extract_event(self, info):
 132         event_id = compat_str(info['id'])
 133         account = compat_str(info['owner_account_id'])
 134         root_url = (
 135             'https://new.livestream.com/api/accounts/{account}/events/{event}/'
 136             'feed.json'.format(account=account, event=event_id))
 137
 138         def _extract_videos():
 139             last_video = None
 140             for i in itertools.count(1):
 141                 if last_video is None:
 142                     info_url = root_url
 143                 else:
 144                     info_url = '{root}?&id={id}&newer=-1&type=video'.format(
 145                         root=root_url, id=last_video)
 146                 videos_info = self._download_json(info_url, event_id, 'Downloading page {0}'.format(i))['data']
 147                 videos_info = [v['data'] for v in videos_info if v['type'] == 'video']
 148                 if not videos_info:
 149                     break
 150                 for v in videos_info:
 151                     yield self._extract_video_info(v)
 152                 last_video = videos_info[-1]['id']
 153         return self.playlist_result(_extract_videos(), event_id, info['full_name'])
 154
 155     def _real_extract(self, url):
 156         mobj = re.match(self._VALID_URL, url)
 157         video_id = mobj.group('id')
 158         event_name = mobj.group('event_name')
 159         webpage = self._download_webpage(url, video_id or event_name)
 160
 161         og_video = self._og_search_video_url(
 162             webpage, 'player url', fatal=False, default=None)
 163         if og_video is not None:
 164             query_str = compat_urllib_parse_urlparse(og_video).query
 165             query = compat_urlparse.parse_qs(query_str)
 166             if 'play_url' in query:
 167                 api_url = query['play_url'][0].replace('.smil', '')
 168                 info = json.loads(self._download_webpage(
 169                     api_url, video_id, 'Downloading video info'))
 170                 return self._extract_video_info(info)
 171
 172         config_json = self._search_regex(
 173             r'window.config = ({.*?});', webpage, 'window config')
 174         info = json.loads(config_json)['event']
 175
 176         def is_relevant(vdata, vid):
 177             result = vdata['type'] == 'video'
 178             if video_id is not None:
 179                 result = result and compat_str(vdata['data']['id']) == vid
 180             return result
 181
 182         if video_id is None:
 183             # This is an event page:
 184             return self._extract_event(info)
 185         else:
 186             videos = [self._extract_video_info(video_data['data'])
 187                       for video_data in info['feed']['data']
 188                       if is_relevant(video_data, video_id)]
 189             if not videos:
 190                 raise ExtractorError('Cannot find video %s' % video_id)
 191             return videos[0]
 192
 193
 194 # The original version of Livestream uses a different system
 195 class LivestreamOriginalIE(InfoExtractor):
 196     IE_NAME = 'livestream:original'
 197     _VALID_URL = r'''(?x)https?://original\.livestream\.com/
 198         (?P<user>[^/]+)/(?P<type>video|folder)
 199         (?:\?.*?Id=|/)(?P<id>.*?)(&|$)
 200         '''
 201     _TESTS = [{
 202         'url': 'http://original.livestream.com/dealbook/video?clipId=pla_8aa4a3f1-ba15-46a4-893b-902210e138fb',
 203         'info_dict': {
 204             'id': 'pla_8aa4a3f1-ba15-46a4-893b-902210e138fb',
 205             'ext': 'mp4',
 206             'title': 'Spark 1 (BitCoin) with Cameron Winklevoss & Tyler Winklevoss of Winklevoss Capital',
 207         },
 208     }, {
 209         'url': 'https://original.livestream.com/newplay/folder?dirId=a07bf706-d0e4-4e75-a747-b021d84f2fd3',
 210         'info_dict': {
 211             'id': 'a07bf706-d0e4-4e75-a747-b021d84f2fd3',
 212         },
 213         'playlist_mincount': 4,
 214     }]
 215
 216     def _extract_video(self, user, video_id):
 217         api_url = 'http://x{0}x.api.channel.livestream.com/2.0/clipdetails?extendedInfo=true&id={1}'.format(user, video_id)
 218
 219         info = self._download_xml(api_url, video_id)
 220         # this url is used on mobile devices
 221         stream_url = 'http://x{0}x.api.channel.livestream.com/3.0/getstream.json?id={1}'.format(user, video_id)
 222         stream_info = self._download_json(stream_url, video_id)
 223         item = info.find('channel').find('item')
 224         ns = {'media': 'http://search.yahoo.com/mrss'}
 225         thumbnail_url = item.find(xpath_with_ns('media:thumbnail', ns)).attrib['url']
 226
 227         return {
 228             'id': video_id,
 229             'title': item.find('title').text,
 230             'url': stream_info['progressiveUrl'],
 231             'thumbnail': thumbnail_url,
 232         }
 233
 234     def _extract_folder(self, url, folder_id):
 235         webpage = self._download_webpage(url, folder_id)
 236         paths = orderedSet(re.findall(
 237             r'''(?x)(?:
 238                 <li\s+class="folder">\s*<a\s+href="|
 239                 <a\s+href="(?=https?://livestre\.am/)
 240             )([^"]+)"''', webpage))
 241
 242         return {
 243             '_type': 'playlist',
 244             'id': folder_id,
 245             'entries': [{
 246                 '_type': 'url',
 247                 'url': compat_urlparse.urljoin(url, p),
 248             } for p in paths],
 249         }
 250
 251     def _real_extract(self, url):
 252         mobj = re.match(self._VALID_URL, url)
 253         id = mobj.group('id')
 254         user = mobj.group('user')
 255         url_type = mobj.group('type')
 256         if url_type == 'folder':
 257             return self._extract_folder(url, id)
 258         else:
 259             return self._extract_video(user, id)
 260
 261
 262 # The server doesn't support HEAD request, the generic extractor can't detect
 263 # the redirection
 264 class LivestreamShortenerIE(InfoExtractor):
 265     IE_NAME = 'livestream:shortener'
 266     IE_DESC = False  # Do not list
 267     _VALID_URL = r'https?://livestre\.am/(?P<id>.+)'
 268
 269     def _real_extract(self, url):
 270         mobj = re.match(self._VALID_URL, url)
 271         id = mobj.group('id')
 272         webpage = self._download_webpage(url, id)
 273
 274         return {
 275             '_type': 'url',
 276             'url': self._og_search_url(webpage),
 277         }