_ Git - youtube-dl/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         },
  41         'playlist_mincount': 4,
  42     }, {
  43         'url': 'https://new.livestream.com/accounts/362/events/3557232/videos/67864563/player?autoPlay=false&height=360&mute=false&width=640',
  44         'only_matching': True,
  45     }]
  46
  47     def _parse_smil(self, video_id, smil_url):
  48         formats = []
  49         _SWITCH_XPATH = (
  50             './/{http://www.w3.org/2001/SMIL20/Language}body/'
  51             '{http://www.w3.org/2001/SMIL20/Language}switch')
  52         smil_doc = self._download_xml(
  53             smil_url, video_id,
  54             note='Downloading SMIL information',
  55             errnote='Unable to download SMIL information',
  56             fatal=False)
  57         if smil_doc is False:  # Download failed
  58             return formats
  59         title_node = find_xpath_attr(
  60             smil_doc, './/{http://www.w3.org/2001/SMIL20/Language}meta',
  61             'name', 'title')
  62         if title_node is None:
  63             self.report_warning('Cannot find SMIL id')
  64             switch_node = smil_doc.find(_SWITCH_XPATH)
  65         else:
  66             title_id = title_node.attrib['content']
  67             switch_node = find_xpath_attr(
  68                 smil_doc, _SWITCH_XPATH, 'id', title_id)
  69         if switch_node is None:
  70             raise ExtractorError('Cannot find switch node')
  71         video_nodes = switch_node.findall(
  72             '{http://www.w3.org/2001/SMIL20/Language}video')
  73
  74         for vn in video_nodes:
  75             tbr = int_or_none(vn.attrib.get('system-bitrate'))
  76             furl = (
  77                 'http://livestream-f.akamaihd.net/%s?v=3.0.3&fp=WIN%%2014,0,0,145' %
  78                 (vn.attrib['src']))
  79             if 'clipBegin' in vn.attrib:
  80                 furl += '&ssek=' + vn.attrib['clipBegin']
  81             formats.append({
  82                 'url': furl,
  83                 'format_id': 'smil_%d' % tbr,
  84                 'ext': 'flv',
  85                 'tbr': tbr,
  86                 'preference': -1000,
  87             })
  88         return formats
  89
  90     def _extract_video_info(self, video_data):
  91         video_id = compat_str(video_data['id'])
  92
  93         FORMAT_KEYS = (
  94             ('sd', 'progressive_url'),
  95             ('hd', 'progressive_url_hd'),
  96         )
  97         formats = [{
  98             'format_id': format_id,
  99             'url': video_data[key],
 100             'quality': i + 1,
 101         } for i, (format_id, key) in enumerate(FORMAT_KEYS)
 102             if video_data.get(key)]
 103
 104         smil_url = video_data.get('smil_url')
 105         if smil_url:
 106             formats.extend(self._parse_smil(video_id, smil_url))
 107         self._sort_formats(formats)
 108
 109         return {
 110             'id': video_id,
 111             'formats': formats,
 112             'title': video_data['caption'],
 113             'thumbnail': video_data.get('thumbnail_url'),
 114             'upload_date': video_data['updated_at'].replace('-', '')[:8],
 115             'like_count': video_data.get('likes', {}).get('total'),
 116             'view_count': video_data.get('views'),
 117         }
 118
 119     def _real_extract(self, url):
 120         mobj = re.match(self._VALID_URL, url)
 121         video_id = mobj.group('id')
 122         event_name = mobj.group('event_name')
 123         webpage = self._download_webpage(url, video_id or event_name)
 124
 125         og_video = self._og_search_video_url(
 126             webpage, 'player url', fatal=False, default=None)
 127         if og_video is not None:
 128             query_str = compat_urllib_parse_urlparse(og_video).query
 129             query = compat_urlparse.parse_qs(query_str)
 130             if 'play_url' in query:
 131                 api_url = query['play_url'][0].replace('.smil', '')
 132                 info = json.loads(self._download_webpage(
 133                     api_url, video_id, 'Downloading video info'))
 134                 return self._extract_video_info(info)
 135
 136         config_json = self._search_regex(
 137             r'window.config = ({.*?});', webpage, 'window config')
 138         info = json.loads(config_json)['event']
 139
 140         def is_relevant(vdata, vid):
 141             result = vdata['type'] == 'video'
 142             if video_id is not None:
 143                 result = result and compat_str(vdata['data']['id']) == vid
 144             return result
 145
 146         videos = [self._extract_video_info(video_data['data'])
 147                   for video_data in info['feed']['data']
 148                   if is_relevant(video_data, video_id)]
 149         if video_id is None:
 150             # This is an event page:
 151             return self.playlist_result(videos, info['id'], info['full_name'])
 152         else:
 153             if not videos:
 154                 raise ExtractorError('Cannot find video %s' % video_id)
 155             return videos[0]
 156
 157
 158 # The original version of Livestream uses a different system
 159 class LivestreamOriginalIE(InfoExtractor):
 160     IE_NAME = 'livestream:original'
 161     _VALID_URL = r'''(?x)https?://www\.livestream\.com/
 162         (?P<user>[^/]+)/(?P<type>video|folder)
 163         (?:\?.*?Id=|/)(?P<id>.*?)(&|$)
 164         '''
 165     _TESTS = [{
 166         'url': 'http://www.livestream.com/dealbook/video?clipId=pla_8aa4a3f1-ba15-46a4-893b-902210e138fb',
 167         'info_dict': {
 168             'id': 'pla_8aa4a3f1-ba15-46a4-893b-902210e138fb',
 169             'ext': 'flv',
 170             'title': 'Spark 1 (BitCoin) with Cameron Winklevoss & Tyler Winklevoss of Winklevoss Capital',
 171         },
 172         'params': {
 173             # rtmp
 174             'skip_download': True,
 175         },
 176     }, {
 177         'url': 'https://www.livestream.com/newplay/folder?dirId=a07bf706-d0e4-4e75-a747-b021d84f2fd3',
 178         'info_dict': {
 179             'id': 'a07bf706-d0e4-4e75-a747-b021d84f2fd3',
 180         },
 181         'playlist_mincount': 4,
 182     }]
 183
 184     def _extract_video(self, user, video_id):
 185         api_url = 'http://x{0}x.api.channel.livestream.com/2.0/clipdetails?extendedInfo=true&id={1}'.format(user, video_id)
 186
 187         info = self._download_xml(api_url, video_id)
 188         item = info.find('channel').find('item')
 189         ns = {'media': 'http://search.yahoo.com/mrss'}
 190         thumbnail_url = item.find(xpath_with_ns('media:thumbnail', ns)).attrib['url']
 191         # Remove the extension and number from the path (like 1.jpg)
 192         path = self._search_regex(r'(user-files/.+)_.*?\.jpg$', thumbnail_url, 'path')
 193
 194         return {
 195             'id': video_id,
 196             'title': item.find('title').text,
 197             'url': 'rtmp://extondemand.livestream.com/ondemand',
 198             'play_path': 'trans/dv15/mogulus-{0}'.format(path),
 199             'player_url': 'http://static.livestream.com/chromelessPlayer/v21/playerapi.swf?hash=5uetk&v=0803&classid=D27CDB6E-AE6D-11cf-96B8-444553540000&jsEnabled=false&wmode=opaque',
 200             'ext': 'flv',
 201             'thumbnail': thumbnail_url,
 202         }
 203
 204     def _extract_folder(self, url, folder_id):
 205         webpage = self._download_webpage(url, folder_id)
 206         paths = orderedSet(re.findall(
 207             r'''(?x)(?:
 208                 <li\s+class="folder">\s*<a\s+href="|
 209                 <a\s+href="(?=https?://livestre\.am/)
 210             )([^"]+)"''', webpage))
 211
 212         return {
 213             '_type': 'playlist',
 214             'id': folder_id,
 215             'entries': [{
 216                 '_type': 'url',
 217                 'url': compat_urlparse.urljoin(url, p),
 218             } for p in paths],
 219         }
 220
 221     def _real_extract(self, url):
 222         mobj = re.match(self._VALID_URL, url)
 223         id = mobj.group('id')
 224         user = mobj.group('user')
 225         url_type = mobj.group('type')
 226         if url_type == 'folder':
 227             return self._extract_folder(url, id)
 228         else:
 229             return self._extract_video(user, id)
 230
 231
 232 # The server doesn't support HEAD request, the generic extractor can't detect
 233 # the redirection
 234 class LivestreamShortenerIE(InfoExtractor):
 235     IE_NAME = 'livestream:shortener'
 236     IE_DESC = False  # Do not list
 237     _VALID_URL = r'https?://livestre\.am/(?P<id>.+)'
 238
 239     def _real_extract(self, url):
 240         mobj = re.match(self._VALID_URL, url)
 241         id = mobj.group('id')
 242         webpage = self._download_webpage(url, id)
 243
 244         return {
 245             '_type': 'url',
 246             'url': self._og_search_url(webpage),
 247         }