_ 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(
  66             smil, self._xpath_ns('.//meta', namespace), 'name', 'httpBase')
  67         base = base_ele.get('content') if base_ele else 'http://livestreamvod-f.akamaihd.net/'
  68
  69         formats = []
  70         video_nodes = smil.findall(self._xpath_ns('.//video', namespace))
  71
  72         for vn in video_nodes:
  73             tbr = int_or_none(vn.attrib.get('system-bitrate'), 1000)
  74             furl = (
  75                 '%s%s?v=3.0.3&fp=WIN%%2014,0,0,145' % (base, vn.attrib['src']))
  76             if 'clipBegin' in vn.attrib:
  77                 furl += '&ssek=' + vn.attrib['clipBegin']
  78             formats.append({
  79                 'url': furl,
  80                 'format_id': 'smil_%d' % tbr,
  81                 'ext': 'flv',
  82                 'tbr': tbr,
  83                 'preference': -1000,
  84             })
  85         return formats
  86
  87     def _extract_video_info(self, video_data):
  88         video_id = compat_str(video_data['id'])
  89
  90         FORMAT_KEYS = (
  91             ('sd', 'progressive_url'),
  92             ('hd', 'progressive_url_hd'),
  93         )
  94
  95         formats = []
  96         for format_id, key in FORMAT_KEYS:
  97             video_url = video_data.get(key)
  98             if video_url:
  99                 ext = determine_ext(video_url)
 100                 if ext == 'm3u8':
 101                     continue
 102                 bitrate = int_or_none(self._search_regex(
 103                     r'(\d+)\.%s' % ext, video_url, 'bitrate', default=None))
 104                 formats.append({
 105                     'url': video_url,
 106                     'format_id': format_id,
 107                     'tbr': bitrate,
 108                     'ext': ext,
 109                 })
 110
 111         smil_url = video_data.get('smil_url')
 112         if smil_url:
 113             smil_formats = self._extract_smil_formats(smil_url, video_id)
 114             if smil_formats:
 115                 formats.extend(smil_formats)
 116
 117         m3u8_url = video_data.get('m3u8_url')
 118         if m3u8_url:
 119             m3u8_formats = self._extract_m3u8_formats(
 120                 m3u8_url, video_id, 'mp4', 'm3u8_native', m3u8_id='hls', fatal=False)
 121             if m3u8_formats:
 122                 formats.extend(m3u8_formats)
 123
 124         f4m_url = video_data.get('f4m_url')
 125         if f4m_url:
 126             f4m_formats = self._extract_f4m_formats(
 127                 f4m_url, video_id, f4m_id='hds', fatal=False)
 128             if f4m_formats:
 129                 formats.extend(f4m_formats)
 130         self._sort_formats(formats)
 131
 132         comments = [{
 133             'author_id': comment.get('author_id'),
 134             'author': comment.get('author', {}).get('full_name'),
 135             'id': comment.get('id'),
 136             'text': comment['text'],
 137             'timestamp': parse_iso8601(comment.get('created_at')),
 138         } for comment in video_data.get('comments', {}).get('data', [])]
 139
 140         return {
 141             'id': video_id,
 142             'formats': formats,
 143             'title': video_data['caption'],
 144             'description': video_data.get('description'),
 145             'thumbnail': video_data.get('thumbnail_url'),
 146             'duration': float_or_none(video_data.get('duration'), 1000),
 147             'timestamp': parse_iso8601(video_data.get('publish_at')),
 148             'like_count': video_data.get('likes', {}).get('total'),
 149             'comment_count': video_data.get('comments', {}).get('total'),
 150             'view_count': video_data.get('views'),
 151             'comments': comments,
 152         }
 153
 154     def _extract_stream_info(self, stream_info):
 155         broadcast_id = stream_info['broadcast_id']
 156         is_live = stream_info.get('is_live')
 157
 158         formats = []
 159         smil_url = stream_info.get('play_url')
 160         if smil_url:
 161             smil_formats = self._extract_smil_formats(smil_url, broadcast_id)
 162             if smil_formats:
 163                 formats.extend(smil_formats)
 164
 165         entry_protocol = 'm3u8' if is_live else 'm3u8_native'
 166         m3u8_url = stream_info.get('m3u8_url')
 167         if m3u8_url:
 168             m3u8_formats = self._extract_m3u8_formats(
 169                 m3u8_url, broadcast_id, 'mp4', entry_protocol, m3u8_id='hls', fatal=False)
 170             if m3u8_formats:
 171                 formats.extend(m3u8_formats)
 172
 173         rtsp_url = stream_info.get('rtsp_url')
 174         if rtsp_url:
 175             formats.append({
 176                 'url': rtsp_url,
 177                 'format_id': 'rtsp',
 178             })
 179         self._sort_formats(formats)
 180
 181         return {
 182             'id': broadcast_id,
 183             'formats': formats,
 184             'title': self._live_title(stream_info['stream_title']) if is_live else stream_info['stream_title'],
 185             'thumbnail': stream_info.get('thumbnail_url'),
 186             'is_live': is_live,
 187         }
 188
 189     def _extract_event(self, event_data):
 190         event_id = compat_str(event_data['id'])
 191         account_id = compat_str(event_data['owner_account_id'])
 192         feed_root_url = self._API_URL_TEMPLATE % (account_id, event_id) + '/feed.json'
 193
 194         stream_info = event_data.get('stream_info')
 195         if stream_info:
 196             return self._extract_stream_info(stream_info)
 197
 198         last_video = None
 199         entries = []
 200         for i in itertools.count(1):
 201             if last_video is None:
 202                 info_url = feed_root_url
 203             else:
 204                 info_url = '{root}?&id={id}&newer=-1&type=video'.format(
 205                     root=feed_root_url, id=last_video)
 206             videos_info = self._download_json(
 207                 info_url, event_id, 'Downloading page {0}'.format(i))['data']
 208             videos_info = [v['data'] for v in videos_info if v['type'] == 'video']
 209             if not videos_info:
 210                 break
 211             for v in videos_info:
 212                 entries.append(self.url_result(
 213                     'http://livestream.com/accounts/%s/events/%s/videos/%s' % (account_id, event_id, v['id']),
 214                     'Livestream', v['id'], v['caption']))
 215             last_video = videos_info[-1]['id']
 216         return self.playlist_result(entries, event_id, event_data['full_name'])
 217
 218     def _real_extract(self, url):
 219         mobj = re.match(self._VALID_URL, url)
 220         video_id = mobj.group('id')
 221         event = mobj.group('event_id') or mobj.group('event_name')
 222         account = mobj.group('account_id') or mobj.group('account_name')
 223         api_url = self._API_URL_TEMPLATE % (account, event)
 224         if video_id:
 225             video_data = self._download_json(
 226                 api_url + '/videos/%s' % video_id, video_id)
 227             return self._extract_video_info(video_data)
 228         else:
 229             event_data = self._download_json(api_url, video_id)
 230             return self._extract_event(event_data)
 231
 232
 233 # The original version of Livestream uses a different system
 234 class LivestreamOriginalIE(InfoExtractor):
 235     IE_NAME = 'livestream:original'
 236     _VALID_URL = r'''(?x)https?://original\.livestream\.com/
 237         (?P<user>[^/\?#]+)(?:/(?P<type>video|folder)
 238         (?:(?:\?.*?Id=|/)(?P<id>.*?)(&|$))?)?
 239         '''
 240     _TESTS = [{
 241         'url': 'http://original.livestream.com/dealbook/video?clipId=pla_8aa4a3f1-ba15-46a4-893b-902210e138fb',
 242         'info_dict': {
 243             'id': 'pla_8aa4a3f1-ba15-46a4-893b-902210e138fb',
 244             'ext': 'mp4',
 245             'title': 'Spark 1 (BitCoin) with Cameron Winklevoss & Tyler Winklevoss of Winklevoss Capital',
 246             'duration': 771.301,
 247             'view_count': int,
 248         },
 249     }, {
 250         'url': 'https://original.livestream.com/newplay/folder?dirId=a07bf706-d0e4-4e75-a747-b021d84f2fd3',
 251         'info_dict': {
 252             'id': 'a07bf706-d0e4-4e75-a747-b021d84f2fd3',
 253         },
 254         'playlist_mincount': 4,
 255     }, {
 256         # live stream
 257         'url': 'http://original.livestream.com/znsbahamas',
 258         'only_matching': True,
 259     }]
 260
 261     def _extract_video_info(self, user, video_id):
 262         api_url = 'http://x%sx.api.channel.livestream.com/2.0/clipdetails?extendedInfo=true&id=%s' % (user, video_id)
 263         info = self._download_xml(api_url, video_id)
 264
 265         item = info.find('channel').find('item')
 266         title = xpath_text(item, 'title')
 267         media_ns = {'media': 'http://search.yahoo.com/mrss'}
 268         thumbnail_url = xpath_attr(
 269             item, xpath_with_ns('media:thumbnail', media_ns), 'url')
 270         duration = float_or_none(xpath_attr(
 271             item, xpath_with_ns('media:content', media_ns), 'duration'))
 272         ls_ns = {'ls': 'http://api.channel.livestream.com/2.0'}
 273         view_count = int_or_none(xpath_text(
 274             item, xpath_with_ns('ls:viewsCount', ls_ns)))
 275
 276         return {
 277             'id': video_id,
 278             'title': title,
 279             'thumbnail': thumbnail_url,
 280             'duration': duration,
 281             'view_count': view_count,
 282         }
 283
 284     def _extract_video_formats(self, video_data, video_id, entry_protocol):
 285         formats = []
 286
 287         progressive_url = video_data.get('progressiveUrl')
 288         if progressive_url:
 289             formats.append({
 290                 'url': progressive_url,
 291                 'format_id': 'http',
 292             })
 293
 294         m3u8_url = video_data.get('httpUrl')
 295         if m3u8_url:
 296             m3u8_formats = self._extract_m3u8_formats(
 297                 m3u8_url, video_id, 'mp4', entry_protocol, m3u8_id='hls', fatal=False)
 298             if m3u8_formats:
 299                 formats.extend(m3u8_formats)
 300
 301         rtsp_url = video_data.get('rtspUrl')
 302         if rtsp_url:
 303             formats.append({
 304                 'url': rtsp_url,
 305                 'format_id': 'rtsp',
 306             })
 307
 308         self._sort_formats(formats)
 309         return formats
 310
 311     def _extract_folder(self, url, folder_id):
 312         webpage = self._download_webpage(url, folder_id)
 313         paths = orderedSet(re.findall(
 314             r'''(?x)(?:
 315                 <li\s+class="folder">\s*<a\s+href="|
 316                 <a\s+href="(?=https?://livestre\.am/)
 317             )([^"]+)"''', webpage))
 318
 319         entries = [{
 320             '_type': 'url',
 321             'url': compat_urlparse.urljoin(url, p),
 322         } for p in paths]
 323
 324         return self.playlist_result(entries, folder_id)
 325
 326     def _real_extract(self, url):
 327         mobj = re.match(self._VALID_URL, url)
 328         user = mobj.group('user')
 329         url_type = mobj.group('type')
 330         content_id = mobj.group('id')
 331         if url_type == 'folder':
 332             return self._extract_folder(url, content_id)
 333         else:
 334             # this url is used on mobile devices
 335             stream_url = 'http://x%sx.api.channel.livestream.com/3.0/getstream.json' % user
 336             info = {}
 337             if content_id:
 338                 stream_url += '?id=%s' % content_id
 339                 info = self._extract_video_info(user, content_id)
 340             else:
 341                 content_id = user
 342                 webpage = self._download_webpage(url, content_id)
 343                 info = {
 344                     'title': self._og_search_title(webpage),
 345                     'description': self._og_search_description(webpage),
 346                     'thumbnail': self._search_regex(r'channelLogo.src\s*=\s*"([^"]+)"', webpage, 'thumbnail', None),
 347                 }
 348             video_data = self._download_json(stream_url, content_id)
 349             is_live = video_data.get('isLive')
 350             entry_protocol = 'm3u8' if is_live else 'm3u8_native'
 351             info.update({
 352                 'id': content_id,
 353                 'title': self._live_title(info['title']) if is_live else info['title'],
 354                 'formats': self._extract_video_formats(video_data, content_id, entry_protocol),
 355                 'is_live': is_live,
 356             })
 357             return info
 358
 359
 360 # The server doesn't support HEAD request, the generic extractor can't detect
 361 # the redirection
 362 class LivestreamShortenerIE(InfoExtractor):
 363     IE_NAME = 'livestream:shortener'
 364     IE_DESC = False  # Do not list
 365     _VALID_URL = r'https?://livestre\.am/(?P<id>.+)'
 366
 367     def _real_extract(self, url):
 368         mobj = re.match(self._VALID_URL, url)
 369         id = mobj.group('id')
 370         webpage = self._download_webpage(url, id)
 371
 372         return {
 373             '_type': 'url',
 374             'url': self._og_search_url(webpage),
 375         }