_ Git - youtube-dl/blob - youtube_dl/extractor/livestream.py

   1 from __future__ import unicode_literals
   2
   3 import re
   4 import itertools
   5
   6 from .common import InfoExtractor
   7 from ..compat import (
   8     compat_str,
   9     compat_urlparse,
  10 )
  11 from ..utils import (
  12     find_xpath_attr,
  13     xpath_attr,
  14     xpath_with_ns,
  15     xpath_text,
  16     orderedSet,
  17     int_or_none,
  18     float_or_none,
  19     parse_iso8601,
  20     determine_ext,
  21 )
  22
  23
  24 class LivestreamIE(InfoExtractor):
  25     IE_NAME = 'livestream'
  26     _VALID_URL = r'https?://(?:new\.)?livestream\.com/(?:accounts/(?P<account_id>\d+)|(?P<account_name>[^/]+))/(?:events/(?P<event_id>\d+)|(?P<event_name>[^/]+))(?:/videos/(?P<id>\d+))?'
  27     _TESTS = [{
  28         'url': 'http://new.livestream.com/CoheedandCambria/WebsterHall/videos/4719370',
  29         'md5': '53274c76ba7754fb0e8d072716f2292b',
  30         'info_dict': {
  31             'id': '4719370',
  32             'ext': 'mp4',
  33             'title': 'Live from Webster Hall NYC',
  34             'timestamp': 1350008072,
  35             'upload_date': '20121012',
  36             'duration': 5968.0,
  37             'like_count': int,
  38             'view_count': int,
  39             'thumbnail': 're:^http://.*\.jpg$'
  40         }
  41     }, {
  42         'url': 'http://new.livestream.com/tedx/cityenglish',
  43         'info_dict': {
  44             'title': 'TEDCity2.0 (English)',
  45             'id': '2245590',
  46         },
  47         'playlist_mincount': 4,
  48     }, {
  49         'url': 'http://new.livestream.com/chess24/tatasteelchess',
  50         'info_dict': {
  51             'title': 'Tata Steel Chess',
  52             'id': '3705884',
  53         },
  54         'playlist_mincount': 60,
  55     }, {
  56         'url': 'https://new.livestream.com/accounts/362/events/3557232/videos/67864563/player?autoPlay=false&height=360&mute=false&width=640',
  57         'only_matching': True,
  58     }, {
  59         'url': 'http://livestream.com/bsww/concacafbeachsoccercampeonato2015',
  60         'only_matching': True,
  61     }]
  62     _API_URL_TEMPLATE = 'http://livestream.com/api/accounts/%s/events/%s'
  63
  64     def _parse_smil_formats(self, smil, smil_url, video_id, namespace=None, f4m_params=None, transform_rtmp_url=None):
  65         base_ele = find_xpath_attr(smil, self._xpath_ns('.//meta', namespace), 'name', 'httpBase')
  66         base = base_ele.get('content') if base_ele else 'http://livestreamvod-f.akamaihd.net/'
  67
  68         formats = []
  69         video_nodes = smil.findall(self._xpath_ns('.//video', namespace))
  70
  71         for vn in video_nodes:
  72             tbr = int_or_none(vn.attrib.get('system-bitrate'), 1000)
  73             furl = (
  74                 '%s%s?v=3.0.3&fp=WIN%%2014,0,0,145' % (base, vn.attrib['src']))
  75             if 'clipBegin' in vn.attrib:
  76                 furl += '&ssek=' + vn.attrib['clipBegin']
  77             formats.append({
  78                 'url': furl,
  79                 'format_id': 'smil_%d' % tbr,
  80                 'ext': 'flv',
  81                 'tbr': tbr,
  82                 'preference': -1000,
  83             })
  84         return formats
  85
  86     def _extract_video_info(self, video_data):
  87         video_id = compat_str(video_data['id'])
  88
  89         FORMAT_KEYS = (
  90             ('sd', 'progressive_url'),
  91             ('hd', 'progressive_url_hd'),
  92         )
  93
  94         formats = []
  95         for format_id, key in FORMAT_KEYS:
  96             video_url = video_data.get(key)
  97             if video_url:
  98                 ext = determine_ext(video_url)
  99                 bitrate = int_or_none(self._search_regex(r'(\d+)\.%s' % ext, video_url, 'bitrate', default=None))
 100                 formats.append({
 101                     'url': video_url,
 102                     'format_id': format_id,
 103                     'tbr': bitrate,
 104                     'ext': ext,
 105                 })
 106
 107         smil_url = video_data.get('smil_url')
 108         if smil_url:
 109             smil_formats = self._extract_smil_formats(smil_url, video_id)
 110             if smil_formats:
 111                 formats.extend(smil_formats)
 112
 113         m3u8_url = video_data.get('m3u8_url')
 114         if m3u8_url:
 115             m3u8_formats = self._extract_m3u8_formats(
 116                 m3u8_url, video_id, 'mp4', 'm3u8_native', m3u8_id='hls', fatal=False)
 117             if m3u8_formats:
 118                 formats.extend(m3u8_formats)
 119
 120         f4m_url = video_data.get('f4m_url')
 121         if f4m_url:
 122             f4m_formats = self._extract_f4m_formats(f4m_url, video_id, f4m_id='hds', fatal=False)
 123             if f4m_formats:
 124                 formats.extend(f4m_formats)
 125         self._sort_formats(formats)
 126
 127         comments = [{
 128             'author_id': comment.get('author_id'),
 129             'author': comment.get('author', {}).get('full_name'),
 130             'id': comment.get('id'),
 131             'text': comment['text'],
 132             'timestamp': parse_iso8601(comment.get('created_at')),
 133         } for comment in video_data.get('comments', {}).get('data', [])]
 134
 135         return {
 136             'id': video_id,
 137             'formats': formats,
 138             'title': video_data['caption'],
 139             'description': video_data.get('description'),
 140             'thumbnail': video_data.get('thumbnail_url'),
 141             'duration': float_or_none(video_data.get('duration'), 1000),
 142             'timestamp': parse_iso8601(video_data.get('publish_at')),
 143             'like_count': video_data.get('likes', {}).get('total'),
 144             'comment_count': video_data.get('comments', {}).get('total'),
 145             'view_count': video_data.get('views'),
 146             'comments': comments,
 147         }
 148
 149     def _extract_stream_info(self, stream_info):
 150         broadcast_id = stream_info['broadcast_id']
 151         is_live = stream_info.get('is_live')
 152
 153         formats = []
 154         smil_url = stream_info.get('play_url')
 155         if smil_url:
 156             smil_formats = self._extract_smil_formats(smil_url, broadcast_id)
 157             if smil_formats:
 158                 formats.extend(smil_formats)
 159
 160         m3u8_url = stream_info.get('m3u8_url')
 161         if m3u8_url:
 162             m3u8_formats = self._extract_m3u8_formats(
 163                 m3u8_url, broadcast_id, 'mp4', 'm3u8_native', m3u8_id='hls', fatal=False)
 164             if m3u8_formats:
 165                 formats.extend(m3u8_formats)
 166
 167         rtsp_url = stream_info.get('rtsp_url')
 168         if rtsp_url:
 169             formats.append({
 170                 'url': rtsp_url,
 171                 'format_id': 'rtsp',
 172             })
 173         self._sort_formats(formats)
 174
 175         return {
 176             'id': broadcast_id,
 177             'formats': formats,
 178             'title': self._live_title(stream_info['stream_title']) if is_live else stream_info['stream_title'],
 179             'thumbnail': stream_info.get('thumbnail_url'),
 180             'is_live': is_live,
 181         }
 182
 183     def _extract_event(self, event_data):
 184         event_id = compat_str(event_data['id'])
 185         account_id = compat_str(event_data['owner_account_id'])
 186         feed_root_url = self._API_URL_TEMPLATE % (account_id, event_id) + '/feed.json'
 187
 188         stream_info = event_data.get('stream_info')
 189         if stream_info:
 190             return self._extract_stream_info(stream_info)
 191
 192         last_video = None
 193         entries = []
 194         for i in itertools.count(1):
 195             if last_video is None:
 196                 info_url = feed_root_url
 197             else:
 198                 info_url = '{root}?&id={id}&newer=-1&type=video'.format(
 199                     root=feed_root_url, id=last_video)
 200             videos_info = self._download_json(info_url, event_id, 'Downloading page {0}'.format(i))['data']
 201             videos_info = [v['data'] for v in videos_info if v['type'] == 'video']
 202             if not videos_info:
 203                 break
 204             for v in videos_info:
 205                 entries.append(self.url_result(
 206                     'http://livestream.com/accounts/%s/events/%s/videos/%s' % (account_id, event_id, v['id']),
 207                     'Livestream', v['id'], v['caption']))
 208             last_video = videos_info[-1]['id']
 209         return self.playlist_result(entries, event_id, event_data['full_name'])
 210
 211     def _real_extract(self, url):
 212         mobj = re.match(self._VALID_URL, url)
 213         video_id = mobj.group('id')
 214         event = mobj.group('event_id') or mobj.group('event_name')
 215         account = mobj.group('account_id') or mobj.group('account_name')
 216         api_url = self._API_URL_TEMPLATE % (account, event)
 217         if video_id:
 218             video_data = self._download_json(api_url + '/videos/%s' % video_id, video_id)
 219             return self._extract_video_info(video_data)
 220         else:
 221             event_data = self._download_json(api_url, video_id)
 222             return self._extract_event(event_data)
 223
 224
 225 # The original version of Livestream uses a different system
 226 class LivestreamOriginalIE(InfoExtractor):
 227     IE_NAME = 'livestream:original'
 228     _VALID_URL = r'''(?x)https?://original\.livestream\.com/
 229         (?P<user>[^/]+)/(?P<type>video|folder)
 230         (?:\?.*?Id=|/)(?P<id>.*?)(&|$)
 231         '''
 232     _TESTS = [{
 233         'url': 'http://original.livestream.com/dealbook/video?clipId=pla_8aa4a3f1-ba15-46a4-893b-902210e138fb',
 234         'info_dict': {
 235             'id': 'pla_8aa4a3f1-ba15-46a4-893b-902210e138fb',
 236             'ext': 'mp4',
 237             'title': 'Spark 1 (BitCoin) with Cameron Winklevoss & Tyler Winklevoss of Winklevoss Capital',
 238             'duration': 771.301,
 239             'view_count': int,
 240         },
 241     }, {
 242         'url': 'https://original.livestream.com/newplay/folder?dirId=a07bf706-d0e4-4e75-a747-b021d84f2fd3',
 243         'info_dict': {
 244             'id': 'a07bf706-d0e4-4e75-a747-b021d84f2fd3',
 245         },
 246         'playlist_mincount': 4,
 247     }]
 248
 249     def _extract_video(self, user, video_id):
 250         api_url = 'http://x{0}x.api.channel.livestream.com/2.0/clipdetails?extendedInfo=true&id={1}'.format(user, video_id)
 251
 252         info = self._download_xml(api_url, video_id)
 253         # this url is used on mobile devices
 254         stream_url = 'http://x{0}x.api.channel.livestream.com/3.0/getstream.json?id={1}'.format(user, video_id)
 255         stream_info = self._download_json(stream_url, video_id)
 256         is_live = stream_info.get('isLive')
 257         item = info.find('channel').find('item')
 258         media_ns = {'media': 'http://search.yahoo.com/mrss'}
 259         thumbnail_url = xpath_attr(item, xpath_with_ns('media:thumbnail', media_ns), 'url')
 260         duration = float_or_none(xpath_attr(item, xpath_with_ns('media:content', media_ns), 'duration'))
 261         ls_ns = {'ls': 'http://api.channel.livestream.com/2.0'}
 262         view_count = int_or_none(xpath_text(item, xpath_with_ns('ls:viewsCount', ls_ns)))
 263
 264         formats = [{
 265             'url': stream_info['progressiveUrl'],
 266             'format_id': 'http',
 267         }]
 268
 269         m3u8_url = stream_info.get('httpUrl')
 270         if m3u8_url:
 271             m3u8_formats = self._extract_m3u8_formats(
 272                 m3u8_url, video_id, 'mp4', 'm3u8_native', m3u8_id='hls', fatal=False)
 273             if m3u8_formats:
 274                 formats.extend(m3u8_formats)
 275
 276         rtsp_url = stream_info.get('rtspUrl')
 277         if rtsp_url:
 278             formats.append({
 279                 'url': rtsp_url,
 280                 'format_id': 'rtsp',
 281             })
 282         self._sort_formats(formats)
 283
 284         return {
 285             'id': video_id,
 286             'title': self._live_title(xpath_text(item, 'title')) if is_live else xpath_text(item, 'title'),
 287             'formats': formats,
 288             'thumbnail': thumbnail_url,
 289             'duration': duration,
 290             'view_count': view_count,
 291             'is_live': is_live,
 292         }
 293
 294     def _extract_folder(self, url, folder_id):
 295         webpage = self._download_webpage(url, folder_id)
 296         paths = orderedSet(re.findall(
 297             r'''(?x)(?:
 298                 <li\s+class="folder">\s*<a\s+href="|
 299                 <a\s+href="(?=https?://livestre\.am/)
 300             )([^"]+)"''', webpage))
 301
 302         entries = [{
 303             '_type': 'url',
 304             'url': compat_urlparse.urljoin(url, p),
 305         } for p in paths]
 306
 307         return self.playlist_result(entries, folder_id)
 308
 309     def _real_extract(self, url):
 310         mobj = re.match(self._VALID_URL, url)
 311         id = mobj.group('id')
 312         user = mobj.group('user')
 313         url_type = mobj.group('type')
 314         if url_type == 'folder':
 315             return self._extract_folder(url, id)
 316         else:
 317             return self._extract_video(user, id)
 318
 319
 320 # The server doesn't support HEAD request, the generic extractor can't detect
 321 # the redirection
 322 class LivestreamShortenerIE(InfoExtractor):
 323     IE_NAME = 'livestream:shortener'
 324     IE_DESC = False  # Do not list
 325     _VALID_URL = r'https?://livestre\.am/(?P<id>.+)'
 326
 327     def _real_extract(self, url):
 328         mobj = re.match(self._VALID_URL, url)
 329         id = mobj.group('id')
 330         webpage = self._download_webpage(url, id)
 331
 332         return {
 333             '_type': 'url',
 334             'url': self._og_search_url(webpage),
 335         }