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