Raphaël G. Git Repositories - youtubedl/blob - youtube_dl/extractor/livestream.py

   1 from __future__ import unicode_literals
   2
   3 import re
   4 import json
   5
   6 from .common import InfoExtractor
   7 from ..compat import (
   8     compat_str,
   9     compat_urllib_parse_urlparse,
  10     compat_urlparse,
  11 )
  12 from ..utils import (
  13     ExtractorError,
  14     find_xpath_attr,
  15     int_or_none,
  16     orderedSet,
  17     xpath_with_ns,
  18 )
  19
  20
  21 class LivestreamIE(InfoExtractor):
  22     IE_NAME = 'livestream'
  23     _VALID_URL = r'https?://new\.livestream\.com/.*?/(?P<event_name>.*?)(/videos/(?P<id>[0-9]+)(?:/player)?)?/?(?:$|[?#])'
  24     _TESTS = [{
  25         'url': 'http://new.livestream.com/CoheedandCambria/WebsterHall/videos/4719370',
  26         'md5': '53274c76ba7754fb0e8d072716f2292b',
  27         'info_dict': {
  28             'id': '4719370',
  29             'ext': 'mp4',
  30             'title': 'Live from Webster Hall NYC',
  31             'upload_date': '20121012',
  32             'like_count': int,
  33             'view_count': int,
  34             'thumbnail': 're:^http://.*\.jpg$'
  35         }
  36     }, {
  37         'url': 'http://new.livestream.com/tedx/cityenglish',
  38         'info_dict': {
  39             'title': 'TEDCity2.0 (English)',
  40             'id': '2245590',
  41         },
  42         'playlist_mincount': 4,
  43     }, {
  44         'url': 'https://new.livestream.com/accounts/362/events/3557232/videos/67864563/player?autoPlay=false&height=360&mute=false&width=640',
  45         'only_matching': True,
  46     }]
  47
  48     def _parse_smil(self, video_id, smil_url):
  49         formats = []
  50         _SWITCH_XPATH = (
  51             './/{http://www.w3.org/2001/SMIL20/Language}body/'
  52             '{http://www.w3.org/2001/SMIL20/Language}switch')
  53         smil_doc = self._download_xml(
  54             smil_url, video_id,
  55             note='Downloading SMIL information',
  56             errnote='Unable to download SMIL information',
  57             fatal=False)
  58         if smil_doc is False:  # Download failed
  59             return formats
  60         title_node = find_xpath_attr(
  61             smil_doc, './/{http://www.w3.org/2001/SMIL20/Language}meta',
  62             'name', 'title')
  63         if title_node is None:
  64             self.report_warning('Cannot find SMIL id')
  65             switch_node = smil_doc.find(_SWITCH_XPATH)
  66         else:
  67             title_id = title_node.attrib['content']
  68             switch_node = find_xpath_attr(
  69                 smil_doc, _SWITCH_XPATH, 'id', title_id)
  70         if switch_node is None:
  71             raise ExtractorError('Cannot find switch node')
  72         video_nodes = switch_node.findall(
  73             '{http://www.w3.org/2001/SMIL20/Language}video')
  74
  75         for vn in video_nodes:
  76             tbr = int_or_none(vn.attrib.get('system-bitrate'))
  77             furl = (
  78                 'http://livestream-f.akamaihd.net/%s?v=3.0.3&fp=WIN%%2014,0,0,145' %
  79                 (vn.attrib['src']))
  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         formats = [{
  99             'format_id': format_id,
 100             'url': video_data[key],
 101             'quality': i + 1,
 102         } for i, (format_id, key) in enumerate(FORMAT_KEYS)
 103             if video_data.get(key)]
 104
 105         smil_url = video_data.get('smil_url')
 106         if smil_url:
 107             formats.extend(self._parse_smil(video_id, smil_url))
 108         self._sort_formats(formats)
 109
 110         return {
 111             'id': video_id,
 112             'formats': formats,
 113             'title': video_data['caption'],
 114             'thumbnail': video_data.get('thumbnail_url'),
 115             'upload_date': video_data['updated_at'].replace('-', '')[:8],
 116             'like_count': video_data.get('likes', {}).get('total'),
 117             'view_count': video_data.get('views'),
 118         }
 119
 120     def _real_extract(self, url):
 121         mobj = re.match(self._VALID_URL, url)
 122         video_id = mobj.group('id')
 123         event_name = mobj.group('event_name')
 124         webpage = self._download_webpage(url, video_id or event_name)
 125
 126         og_video = self._og_search_video_url(
 127             webpage, 'player url', fatal=False, default=None)
 128         if og_video is not None:
 129             query_str = compat_urllib_parse_urlparse(og_video).query
 130             query = compat_urlparse.parse_qs(query_str)
 131             if 'play_url' in query:
 132                 api_url = query['play_url'][0].replace('.smil', '')
 133                 info = json.loads(self._download_webpage(
 134                     api_url, video_id, 'Downloading video info'))
 135                 return self._extract_video_info(info)
 136
 137         config_json = self._search_regex(
 138             r'window.config = ({.*?});', webpage, 'window config')
 139         info = json.loads(config_json)['event']
 140
 141         def is_relevant(vdata, vid):
 142             result = vdata['type'] == 'video'
 143             if video_id is not None:
 144                 result = result and compat_str(vdata['data']['id']) == vid
 145             return result
 146
 147         videos = [self._extract_video_info(video_data['data'])
 148                   for video_data in info['feed']['data']
 149                   if is_relevant(video_data, video_id)]
 150         if video_id is None:
 151             # This is an event page:
 152             return self.playlist_result(
 153                 videos, '%s' % info['id'], info['full_name'])
 154         else:
 155             if not videos:
 156                 raise ExtractorError('Cannot find video %s' % video_id)
 157             return videos[0]
 158
 159
 160 # The original version of Livestream uses a different system
 161 class LivestreamOriginalIE(InfoExtractor):
 162     IE_NAME = 'livestream:original'
 163     _VALID_URL = r'''(?x)https?://www\.livestream\.com/
 164         (?P<user>[^/]+)/(?P<type>video|folder)
 165         (?:\?.*?Id=|/)(?P<id>.*?)(&|$)
 166         '''
 167     _TESTS = [{
 168         'url': 'http://www.livestream.com/dealbook/video?clipId=pla_8aa4a3f1-ba15-46a4-893b-902210e138fb',
 169         'info_dict': {
 170             'id': 'pla_8aa4a3f1-ba15-46a4-893b-902210e138fb',
 171             'ext': 'flv',
 172             'title': 'Spark 1 (BitCoin) with Cameron Winklevoss & Tyler Winklevoss of Winklevoss Capital',
 173         },
 174         'params': {
 175             # rtmp
 176             'skip_download': True,
 177         },
 178     }, {
 179         'url': 'https://www.livestream.com/newplay/folder?dirId=a07bf706-d0e4-4e75-a747-b021d84f2fd3',
 180         'info_dict': {
 181             'id': 'a07bf706-d0e4-4e75-a747-b021d84f2fd3',
 182         },
 183         'playlist_mincount': 4,
 184     }]
 185
 186     def _extract_video(self, user, video_id):
 187         api_url = 'http://x{0}x.api.channel.livestream.com/2.0/clipdetails?extendedInfo=true&id={1}'.format(user, video_id)
 188
 189         info = self._download_xml(api_url, video_id)
 190         item = info.find('channel').find('item')
 191         ns = {'media': 'http://search.yahoo.com/mrss'}
 192         thumbnail_url = item.find(xpath_with_ns('media:thumbnail', ns)).attrib['url']
 193         # Remove the extension and number from the path (like 1.jpg)
 194         path = self._search_regex(r'(user-files/.+)_.*?\.jpg$', thumbnail_url, 'path')
 195
 196         return {
 197             'id': video_id,
 198             'title': item.find('title').text,
 199             'url': 'rtmp://extondemand.livestream.com/ondemand',
 200             'play_path': 'trans/dv15/mogulus-{0}'.format(path),
 201             'player_url': 'http://static.livestream.com/chromelessPlayer/v21/playerapi.swf?hash=5uetk&v=0803&classid=D27CDB6E-AE6D-11cf-96B8-444553540000&jsEnabled=false&wmode=opaque',
 202             'ext': 'flv',
 203             'thumbnail': thumbnail_url,
 204         }
 205
 206     def _extract_folder(self, url, folder_id):
 207         webpage = self._download_webpage(url, folder_id)
 208         paths = orderedSet(re.findall(
 209             r'''(?x)(?:
 210                 <li\s+class="folder">\s*<a\s+href="|
 211                 <a\s+href="(?=https?://livestre\.am/)
 212             )([^"]+)"''', webpage))
 213
 214         return {
 215             '_type': 'playlist',
 216             'id': folder_id,
 217             'entries': [{
 218                 '_type': 'url',
 219                 'url': compat_urlparse.urljoin(url, p),
 220             } for p in paths],
 221         }
 222
 223     def _real_extract(self, url):
 224         mobj = re.match(self._VALID_URL, url)
 225         id = mobj.group('id')
 226         user = mobj.group('user')
 227         url_type = mobj.group('type')
 228         if url_type == 'folder':
 229             return self._extract_folder(url, id)
 230         else:
 231             return self._extract_video(user, id)
 232
 233
 234 # The server doesn't support HEAD request, the generic extractor can't detect
 235 # the redirection
 236 class LivestreamShortenerIE(InfoExtractor):
 237     IE_NAME = 'livestream:shortener'
 238     IE_DESC = False  # Do not list
 239     _VALID_URL = r'https?://livestre\.am/(?P<id>.+)'
 240
 241     def _real_extract(self, url):
 242         mobj = re.match(self._VALID_URL, url)
 243         id = mobj.group('id')
 244         webpage = self._download_webpage(url, id)
 245
 246         return {
 247             '_type': 'url',
 248             'url': self._og_search_url(webpage),
 249         }