_ 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     update_url_query,
  18     int_or_none,
  19     float_or_none,
  20     parse_iso8601,
  21     determine_ext,
  22 )
  23
  24
  25 class LivestreamIE(InfoExtractor):
  26     IE_NAME = 'livestream'
  27     _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+))?'
  28     _TESTS = [{
  29         'url': 'http://new.livestream.com/CoheedandCambria/WebsterHall/videos/4719370',
  30         'md5': '53274c76ba7754fb0e8d072716f2292b',
  31         'info_dict': {
  32             'id': '4719370',
  33             'ext': 'mp4',
  34             'title': 'Live from Webster Hall NYC',
  35             'timestamp': 1350008072,
  36             'upload_date': '20121012',
  37             'duration': 5968.0,
  38             'like_count': int,
  39             'view_count': int,
  40             'thumbnail': r're:^http://.*\.jpg$'
  41         }
  42     }, {
  43         'url': 'http://new.livestream.com/tedx/cityenglish',
  44         'info_dict': {
  45             'title': 'TEDCity2.0 (English)',
  46             'id': '2245590',
  47         },
  48         'playlist_mincount': 4,
  49     }, {
  50         'url': 'http://new.livestream.com/chess24/tatasteelchess',
  51         'info_dict': {
  52             'title': 'Tata Steel Chess',
  53             'id': '3705884',
  54         },
  55         'playlist_mincount': 60,
  56     }, {
  57         'url': 'https://new.livestream.com/accounts/362/events/3557232/videos/67864563/player?autoPlay=false&height=360&mute=false&width=640',
  58         'only_matching': True,
  59     }, {
  60         'url': 'http://livestream.com/bsww/concacafbeachsoccercampeonato2015',
  61         'only_matching': True,
  62     }]
  63     _API_URL_TEMPLATE = 'http://livestream.com/api/accounts/%s/events/%s'
  64
  65     def _parse_smil_formats(self, smil, smil_url, video_id, namespace=None, f4m_params=None, transform_rtmp_url=None):
  66         base_ele = find_xpath_attr(
  67             smil, self._xpath_ns('.//meta', namespace), 'name', 'httpBase')
  68         base = base_ele.get('content') if base_ele is not None else 'http://livestreamvod-f.akamaihd.net/'
  69
  70         formats = []
  71         video_nodes = smil.findall(self._xpath_ns('.//video', namespace))
  72
  73         for vn in video_nodes:
  74             tbr = int_or_none(vn.attrib.get('system-bitrate'), 1000)
  75             furl = (
  76                 update_url_query(compat_urlparse.urljoin(base, vn.attrib['src']), {
  77                     'v': '3.0.3',
  78                     'fp': 'WIN% 14,0,0,145',
  79                 }))
  80             if 'clipBegin' in vn.attrib:
  81                 furl += '&ssek=' + vn.attrib['clipBegin']
  82             formats.append({
  83                 'url': furl,
  84                 'format_id': 'smil_%d' % tbr,
  85                 'ext': 'flv',
  86                 'tbr': tbr,
  87                 'preference': -1000,
  88             })
  89         return formats
  90
  91     def _extract_video_info(self, video_data):
  92         video_id = compat_str(video_data['id'])
  93
  94         FORMAT_KEYS = (
  95             ('sd', 'progressive_url'),
  96             ('hd', 'progressive_url_hd'),
  97         )
  98
  99         formats = []
 100         for format_id, key in FORMAT_KEYS:
 101             video_url = video_data.get(key)
 102             if video_url:
 103                 ext = determine_ext(video_url)
 104                 if ext == 'm3u8':
 105                     continue
 106                 bitrate = int_or_none(self._search_regex(
 107                     r'(\d+)\.%s' % ext, video_url, 'bitrate', default=None))
 108                 formats.append({
 109                     'url': video_url,
 110                     'format_id': format_id,
 111                     'tbr': bitrate,
 112                     'ext': ext,
 113                 })
 114
 115         smil_url = video_data.get('smil_url')
 116         if smil_url:
 117             formats.extend(self._extract_smil_formats(smil_url, video_id, fatal=False))
 118
 119         m3u8_url = video_data.get('m3u8_url')
 120         if m3u8_url:
 121             formats.extend(self._extract_m3u8_formats(
 122                 m3u8_url, video_id, 'mp4', 'm3u8_native',
 123                 m3u8_id='hls', fatal=False))
 124
 125         f4m_url = video_data.get('f4m_url')
 126         if f4m_url:
 127             formats.extend(self._extract_f4m_formats(
 128                 f4m_url, video_id, f4m_id='hds', fatal=False))
 129         self._sort_formats(formats)
 130
 131         comments = [{
 132             'author_id': comment.get('author_id'),
 133             'author': comment.get('author', {}).get('full_name'),
 134             'id': comment.get('id'),
 135             'text': comment['text'],
 136             'timestamp': parse_iso8601(comment.get('created_at')),
 137         } for comment in video_data.get('comments', {}).get('data', [])]
 138
 139         return {
 140             'id': video_id,
 141             'formats': formats,
 142             'title': video_data['caption'],
 143             'description': video_data.get('description'),
 144             'thumbnail': video_data.get('thumbnail_url'),
 145             'duration': float_or_none(video_data.get('duration'), 1000),
 146             'timestamp': parse_iso8601(video_data.get('publish_at')),
 147             'like_count': video_data.get('likes', {}).get('total'),
 148             'comment_count': video_data.get('comments', {}).get('total'),
 149             'view_count': video_data.get('views'),
 150             'comments': comments,
 151         }
 152
 153     def _extract_stream_info(self, stream_info):
 154         broadcast_id = compat_str(stream_info['broadcast_id'])
 155         is_live = stream_info.get('is_live')
 156
 157         formats = []
 158         smil_url = stream_info.get('play_url')
 159         if smil_url:
 160             formats.extend(self._extract_smil_formats(smil_url, broadcast_id))
 161
 162         m3u8_url = stream_info.get('m3u8_url')
 163         if m3u8_url:
 164             formats.extend(self._extract_m3u8_formats(
 165                 m3u8_url, broadcast_id, 'mp4', 'm3u8_native',
 166                 m3u8_id='hls', fatal=False))
 167
 168         rtsp_url = stream_info.get('rtsp_url')
 169         if rtsp_url:
 170             formats.append({
 171                 'url': rtsp_url,
 172                 'format_id': 'rtsp',
 173             })
 174         self._sort_formats(formats)
 175
 176         return {
 177             'id': broadcast_id,
 178             'formats': formats,
 179             'title': self._live_title(stream_info['stream_title']) if is_live else stream_info['stream_title'],
 180             'thumbnail': stream_info.get('thumbnail_url'),
 181             'is_live': is_live,
 182         }
 183
 184     def _extract_event(self, event_data):
 185         event_id = compat_str(event_data['id'])
 186         account_id = compat_str(event_data['owner_account_id'])
 187         feed_root_url = self._API_URL_TEMPLATE % (account_id, event_id) + '/feed.json'
 188
 189         stream_info = event_data.get('stream_info')
 190         if stream_info:
 191             return self._extract_stream_info(stream_info)
 192
 193         last_video = None
 194         entries = []
 195         for i in itertools.count(1):
 196             if last_video is None:
 197                 info_url = feed_root_url
 198             else:
 199                 info_url = '{root}?&id={id}&newer=-1&type=video'.format(
 200                     root=feed_root_url, id=last_video)
 201             videos_info = self._download_json(
 202                 info_url, event_id, 'Downloading page {0}'.format(i))['data']
 203             videos_info = [v['data'] for v in videos_info if v['type'] == 'video']
 204             if not videos_info:
 205                 break
 206             for v in videos_info:
 207                 v_id = compat_str(v['id'])
 208                 entries.append(self.url_result(
 209                     'http://livestream.com/accounts/%s/events/%s/videos/%s' % (account_id, event_id, v_id),
 210                     'Livestream', v_id, v.get('caption')))
 211             last_video = videos_info[-1]['id']
 212         return self.playlist_result(entries, event_id, event_data['full_name'])
 213
 214     def _real_extract(self, url):
 215         mobj = re.match(self._VALID_URL, url)
 216         video_id = mobj.group('id')
 217         event = mobj.group('event_id') or mobj.group('event_name')
 218         account = mobj.group('account_id') or mobj.group('account_name')
 219         api_url = self._API_URL_TEMPLATE % (account, event)
 220         if video_id:
 221             video_data = self._download_json(
 222                 api_url + '/videos/%s' % video_id, video_id)
 223             return self._extract_video_info(video_data)
 224         else:
 225             event_data = self._download_json(api_url, video_id)
 226             return self._extract_event(event_data)
 227
 228
 229 # The original version of Livestream uses a different system
 230 class LivestreamOriginalIE(InfoExtractor):
 231     IE_NAME = 'livestream:original'
 232     _VALID_URL = r'''(?x)https?://original\.livestream\.com/
 233         (?P<user>[^/\?#]+)(?:/(?P<type>video|folder)
 234         (?:(?:\?.*?Id=|/)(?P<id>.*?)(&|$))?)?
 235         '''
 236     _TESTS = [{
 237         'url': 'http://original.livestream.com/dealbook/video?clipId=pla_8aa4a3f1-ba15-46a4-893b-902210e138fb',
 238         'info_dict': {
 239             'id': 'pla_8aa4a3f1-ba15-46a4-893b-902210e138fb',
 240             'ext': 'mp4',
 241             'title': 'Spark 1 (BitCoin) with Cameron Winklevoss & Tyler Winklevoss of Winklevoss Capital',
 242             'duration': 771.301,
 243             'view_count': int,
 244         },
 245     }, {
 246         'url': 'https://original.livestream.com/newplay/folder?dirId=a07bf706-d0e4-4e75-a747-b021d84f2fd3',
 247         'info_dict': {
 248             'id': 'a07bf706-d0e4-4e75-a747-b021d84f2fd3',
 249         },
 250         'playlist_mincount': 4,
 251     }, {
 252         # live stream
 253         'url': 'http://original.livestream.com/znsbahamas',
 254         'only_matching': True,
 255     }]
 256
 257     def _extract_video_info(self, user, video_id):
 258         api_url = 'http://x%sx.api.channel.livestream.com/2.0/clipdetails?extendedInfo=true&id=%s' % (user, video_id)
 259         info = self._download_xml(api_url, video_id)
 260
 261         item = info.find('channel').find('item')
 262         title = xpath_text(item, 'title')
 263         media_ns = {'media': 'http://search.yahoo.com/mrss'}
 264         thumbnail_url = xpath_attr(
 265             item, xpath_with_ns('media:thumbnail', media_ns), 'url')
 266         duration = float_or_none(xpath_attr(
 267             item, xpath_with_ns('media:content', media_ns), 'duration'))
 268         ls_ns = {'ls': 'http://api.channel.livestream.com/2.0'}
 269         view_count = int_or_none(xpath_text(
 270             item, xpath_with_ns('ls:viewsCount', ls_ns)))
 271
 272         return {
 273             'id': video_id,
 274             'title': title,
 275             'thumbnail': thumbnail_url,
 276             'duration': duration,
 277             'view_count': view_count,
 278         }
 279
 280     def _extract_video_formats(self, video_data, video_id):
 281         formats = []
 282
 283         progressive_url = video_data.get('progressiveUrl')
 284         if progressive_url:
 285             formats.append({
 286                 'url': progressive_url,
 287                 'format_id': 'http',
 288             })
 289
 290         m3u8_url = video_data.get('httpUrl')
 291         if m3u8_url:
 292             formats.extend(self._extract_m3u8_formats(
 293                 m3u8_url, video_id, 'mp4', 'm3u8_native',
 294                 m3u8_id='hls', fatal=False))
 295
 296         rtsp_url = video_data.get('rtspUrl')
 297         if rtsp_url:
 298             formats.append({
 299                 'url': rtsp_url,
 300                 'format_id': 'rtsp',
 301             })
 302
 303         self._sort_formats(formats)
 304         return formats
 305
 306     def _extract_folder(self, url, folder_id):
 307         webpage = self._download_webpage(url, folder_id)
 308         paths = orderedSet(re.findall(
 309             r'''(?x)(?:
 310                 <li\s+class="folder">\s*<a\s+href="|
 311                 <a\s+href="(?=https?://livestre\.am/)
 312             )([^"]+)"''', webpage))
 313
 314         entries = [{
 315             '_type': 'url',
 316             'url': compat_urlparse.urljoin(url, p),
 317         } for p in paths]
 318
 319         return self.playlist_result(entries, folder_id)
 320
 321     def _real_extract(self, url):
 322         mobj = re.match(self._VALID_URL, url)
 323         user = mobj.group('user')
 324         url_type = mobj.group('type')
 325         content_id = mobj.group('id')
 326         if url_type == 'folder':
 327             return self._extract_folder(url, content_id)
 328         else:
 329             # this url is used on mobile devices
 330             stream_url = 'http://x%sx.api.channel.livestream.com/3.0/getstream.json' % user
 331             info = {}
 332             if content_id:
 333                 stream_url += '?id=%s' % content_id
 334                 info = self._extract_video_info(user, content_id)
 335             else:
 336                 content_id = user
 337                 webpage = self._download_webpage(url, content_id)
 338                 info = {
 339                     'title': self._og_search_title(webpage),
 340                     'description': self._og_search_description(webpage),
 341                     'thumbnail': self._search_regex(r'channelLogo\.src\s*=\s*"([^"]+)"', webpage, 'thumbnail', None),
 342                 }
 343             video_data = self._download_json(stream_url, content_id)
 344             is_live = video_data.get('isLive')
 345             info.update({
 346                 'id': content_id,
 347                 'title': self._live_title(info['title']) if is_live else info['title'],
 348                 'formats': self._extract_video_formats(video_data, content_id),
 349                 'is_live': is_live,
 350             })
 351             return info
 352
 353
 354 # The server doesn't support HEAD request, the generic extractor can't detect
 355 # the redirection
 356 class LivestreamShortenerIE(InfoExtractor):
 357     IE_NAME = 'livestream:shortener'
 358     IE_DESC = False  # Do not list
 359     _VALID_URL = r'https?://livestre\.am/(?P<id>.+)'
 360
 361     def _real_extract(self, url):
 362         mobj = re.match(self._VALID_URL, url)
 363         id = mobj.group('id')
 364         webpage = self._download_webpage(url, id)
 365
 366         return self.url_result(self._og_search_url(webpage))