_ Git - youtube-dl/blob - 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
  56     def _parse_smil(self, video_id, smil_url):
  57         formats = []
  58         _SWITCH_XPATH = (
  59             './/{http://www.w3.org/2001/SMIL20/Language}body/'
  60             '{http://www.w3.org/2001/SMIL20/Language}switch')
  61         smil_doc = self._download_xml(
  62             smil_url, video_id,
  63             note='Downloading SMIL information',
  64             errnote='Unable to download SMIL information',
  65             fatal=False)
  66         if smil_doc is False:  # Download failed
  67             return formats
  68         title_node = find_xpath_attr(
  69             smil_doc, './/{http://www.w3.org/2001/SMIL20/Language}meta',
  70             'name', 'title')
  71         if title_node is None:
  72             self.report_warning('Cannot find SMIL id')
  73             switch_node = smil_doc.find(_SWITCH_XPATH)
  74         else:
  75             title_id = title_node.attrib['content']
  76             switch_node = find_xpath_attr(
  77                 smil_doc, _SWITCH_XPATH, 'id', title_id)
  78         if switch_node is None:
  79             raise ExtractorError('Cannot find switch node')
  80         video_nodes = switch_node.findall(
  81             '{http://www.w3.org/2001/SMIL20/Language}video')
  82
  83         for vn in video_nodes:
  84             tbr = int_or_none(vn.attrib.get('system-bitrate'))
  85             furl = (
  86                 'http://livestream-f.akamaihd.net/%s?v=3.0.3&fp=WIN%%2014,0,0,145' %
  87                 (vn.attrib['src']))
  88             if 'clipBegin' in vn.attrib:
  89                 furl += '&ssek=' + vn.attrib['clipBegin']
  90             formats.append({
  91                 'url': furl,
  92                 'format_id': 'smil_%d' % tbr,
  93                 'ext': 'flv',
  94                 'tbr': tbr,
  95                 'preference': -1000,
  96             })
  97         return formats
  98
  99     def _extract_video_info(self, video_data):
 100         video_id = compat_str(video_data['id'])
 101
 102         FORMAT_KEYS = (
 103             ('sd', 'progressive_url'),
 104             ('hd', 'progressive_url_hd'),
 105         )
 106         formats = [{
 107             'format_id': format_id,
 108             'url': video_data[key],
 109             'quality': i + 1,
 110         } for i, (format_id, key) in enumerate(FORMAT_KEYS)
 111             if video_data.get(key)]
 112
 113         smil_url = video_data.get('smil_url')
 114         if smil_url:
 115             formats.extend(self._parse_smil(video_id, smil_url))
 116         self._sort_formats(formats)
 117
 118         return {
 119             'id': video_id,
 120             'formats': formats,
 121             'title': video_data['caption'],
 122             'thumbnail': video_data.get('thumbnail_url'),
 123             'upload_date': video_data['updated_at'].replace('-', '')[:8],
 124             'like_count': video_data.get('likes', {}).get('total'),
 125             'view_count': video_data.get('views'),
 126         }
 127
 128     def _extract_event(self, info):
 129         event_id = compat_str(info['id'])
 130         account = compat_str(info['owner_account_id'])
 131         root_url = (
 132             'https://new.livestream.com/api/accounts/{account}/events/{event}/'
 133             'feed.json'.format(account=account, event=event_id))
 134
 135         def _extract_videos():
 136             last_video = None
 137             for i in itertools.count(1):
 138                 if last_video is None:
 139                     info_url = root_url
 140                 else:
 141                     info_url = '{root}?&id={id}&newer=-1&type=video'.format(
 142                         root=root_url, id=last_video)
 143                 videos_info = self._download_json(info_url, event_id, 'Downloading page {0}'.format(i))['data']
 144                 videos_info = [v['data'] for v in videos_info if v['type'] == 'video']
 145                 if not videos_info:
 146                     break
 147                 for v in videos_info:
 148                     yield self._extract_video_info(v)
 149                 last_video = videos_info[-1]['id']
 150         return self.playlist_result(_extract_videos(), event_id, info['full_name'])
 151
 152     def _real_extract(self, url):
 153         mobj = re.match(self._VALID_URL, url)
 154         video_id = mobj.group('id')
 155         event_name = mobj.group('event_name')
 156         webpage = self._download_webpage(url, video_id or event_name)
 157
 158         og_video = self._og_search_video_url(
 159             webpage, 'player url', fatal=False, default=None)
 160         if og_video is not None:
 161             query_str = compat_urllib_parse_urlparse(og_video).query
 162             query = compat_urlparse.parse_qs(query_str)
 163             if 'play_url' in query:
 164                 api_url = query['play_url'][0].replace('.smil', '')
 165                 info = json.loads(self._download_webpage(
 166                     api_url, video_id, 'Downloading video info'))
 167                 return self._extract_video_info(info)
 168
 169         config_json = self._search_regex(
 170             r'window.config = ({.*?});', webpage, 'window config')
 171         info = json.loads(config_json)['event']
 172
 173         def is_relevant(vdata, vid):
 174             result = vdata['type'] == 'video'
 175             if video_id is not None:
 176                 result = result and compat_str(vdata['data']['id']) == vid
 177             return result
 178
 179         if video_id is None:
 180             # This is an event page:
 181             return self._extract_event(info)
 182         else:
 183             videos = [self._extract_video_info(video_data['data'])
 184                       for video_data in info['feed']['data']
 185                       if is_relevant(video_data, video_id)]
 186             if not videos:
 187                 raise ExtractorError('Cannot find video %s' % video_id)
 188             return videos[0]
 189
 190
 191 # The original version of Livestream uses a different system
 192 class LivestreamOriginalIE(InfoExtractor):
 193     IE_NAME = 'livestream:original'
 194     _VALID_URL = r'''(?x)https?://www\.livestream\.com/
 195         (?P<user>[^/]+)/(?P<type>video|folder)
 196         (?:\?.*?Id=|/)(?P<id>.*?)(&|$)
 197         '''
 198     _TESTS = [{
 199         'url': 'http://www.livestream.com/dealbook/video?clipId=pla_8aa4a3f1-ba15-46a4-893b-902210e138fb',
 200         'info_dict': {
 201             'id': 'pla_8aa4a3f1-ba15-46a4-893b-902210e138fb',
 202             'ext': 'flv',
 203             'title': 'Spark 1 (BitCoin) with Cameron Winklevoss & Tyler Winklevoss of Winklevoss Capital',
 204         },
 205         'params': {
 206             # rtmp
 207             'skip_download': True,
 208         },
 209     }, {
 210         'url': 'https://www.livestream.com/newplay/folder?dirId=a07bf706-d0e4-4e75-a747-b021d84f2fd3',
 211         'info_dict': {
 212             'id': 'a07bf706-d0e4-4e75-a747-b021d84f2fd3',
 213         },
 214         'playlist_mincount': 4,
 215     }]
 216
 217     def _extract_video(self, user, video_id):
 218         api_url = 'http://x{0}x.api.channel.livestream.com/2.0/clipdetails?extendedInfo=true&id={1}'.format(user, video_id)
 219
 220         info = self._download_xml(api_url, video_id)
 221         item = info.find('channel').find('item')
 222         ns = {'media': 'http://search.yahoo.com/mrss'}
 223         thumbnail_url = item.find(xpath_with_ns('media:thumbnail', ns)).attrib['url']
 224         # Remove the extension and number from the path (like 1.jpg)
 225         path = self._search_regex(r'(user-files/.+)_.*?\.jpg$', thumbnail_url, 'path')
 226
 227         return {
 228             'id': video_id,
 229             'title': item.find('title').text,
 230             'url': 'rtmp://extondemand.livestream.com/ondemand',
 231             'play_path': 'trans/dv15/mogulus-{0}'.format(path),
 232             'player_url': 'http://static.livestream.com/chromelessPlayer/v21/playerapi.swf?hash=5uetk&v=0803&classid=D27CDB6E-AE6D-11cf-96B8-444553540000&jsEnabled=false&wmode=opaque',
 233             'ext': 'flv',
 234             'thumbnail': thumbnail_url,
 235         }
 236
 237     def _extract_folder(self, url, folder_id):
 238         webpage = self._download_webpage(url, folder_id)
 239         paths = orderedSet(re.findall(
 240             r'''(?x)(?:
 241                 <li\s+class="folder">\s*<a\s+href="|
 242                 <a\s+href="(?=https?://livestre\.am/)
 243             )([^"]+)"''', webpage))
 244
 245         return {
 246             '_type': 'playlist',
 247             'id': folder_id,
 248             'entries': [{
 249                 '_type': 'url',
 250                 'url': compat_urlparse.urljoin(url, p),
 251             } for p in paths],
 252         }
 253
 254     def _real_extract(self, url):
 255         mobj = re.match(self._VALID_URL, url)
 256         id = mobj.group('id')
 257         user = mobj.group('user')
 258         url_type = mobj.group('type')
 259         if url_type == 'folder':
 260             return self._extract_folder(url, id)
 261         else:
 262             return self._extract_video(user, id)
 263
 264
 265 # The server doesn't support HEAD request, the generic extractor can't detect
 266 # the redirection
 267 class LivestreamShortenerIE(InfoExtractor):
 268     IE_NAME = 'livestream:shortener'
 269     IE_DESC = False  # Do not list
 270     _VALID_URL = r'https?://livestre\.am/(?P<id>.+)'
 271
 272     def _real_extract(self, url):
 273         mobj = re.match(self._VALID_URL, url)
 274         id = mobj.group('id')
 275         webpage = self._download_webpage(url, id)
 276
 277         return {
 278             '_type': 'url',
 279             'url': self._og_search_url(webpage),
 280         }