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 ..utils import (
   8     compat_str,
   9     compat_urllib_parse_urlparse,
  10     compat_urlparse,
  11     ExtractorError,
  12     find_xpath_attr,
  13     int_or_none,
  14     orderedSet,
  15     xpath_with_ns,
  16 )
  17
  18
  19 class LivestreamIE(InfoExtractor):
  20     IE_NAME = 'livestream'
  21     _VALID_URL = r'http://new\.livestream\.com/.*?/(?P<event_name>.*?)(/videos/(?P<id>\d+))?/?$'
  22     _TESTS = [{
  23         'url': 'http://new.livestream.com/CoheedandCambria/WebsterHall/videos/4719370',
  24         'md5': '53274c76ba7754fb0e8d072716f2292b',
  25         'info_dict': {
  26             'id': '4719370',
  27             'ext': 'mp4',
  28             'title': 'Live from Webster Hall NYC',
  29             'upload_date': '20121012',
  30             'like_count': int,
  31             'view_count': int,
  32             'thumbnail': 're:^http://.*\.jpg$'
  33         }
  34     }, {
  35         'url': 'http://new.livestream.com/tedx/cityenglish',
  36         'info_dict': {
  37             'title': 'TEDCity2.0 (English)',
  38         },
  39         'playlist_mincount': 4,
  40     }]
  41
  42     def _parse_smil(self, video_id, smil_url):
  43         formats = []
  44         _SWITCH_XPATH = (
  45             './/{http://www.w3.org/2001/SMIL20/Language}body/'
  46             '{http://www.w3.org/2001/SMIL20/Language}switch')
  47         smil_doc = self._download_xml(
  48             smil_url, video_id,
  49             note='Downloading SMIL information',
  50             errnote='Unable to download SMIL information',
  51             fatal=False)
  52         if smil_doc is False:  # Download failed
  53             return formats
  54         title_node = find_xpath_attr(
  55             smil_doc, './/{http://www.w3.org/2001/SMIL20/Language}meta',
  56             'name', 'title')
  57         if title_node is None:
  58             self.report_warning('Cannot find SMIL id')
  59             switch_node = smil_doc.find(_SWITCH_XPATH)
  60         else:
  61             title_id = title_node.attrib['content']
  62             switch_node = find_xpath_attr(
  63                 smil_doc, _SWITCH_XPATH, 'id', title_id)
  64         if switch_node is None:
  65             raise ExtractorError('Cannot find switch node')
  66         video_nodes = switch_node.findall(
  67             '{http://www.w3.org/2001/SMIL20/Language}video')
  68
  69         for vn in video_nodes:
  70             tbr = int_or_none(vn.attrib.get('system-bitrate'))
  71             furl = (
  72                 'http://livestream-f.akamaihd.net/%s?v=3.0.3&fp=WIN%%2014,0,0,145' %
  73                 (vn.attrib['src']))
  74             if 'clipBegin' in vn.attrib:
  75                 furl += '&ssek=' + vn.attrib['clipBegin']
  76             formats.append({
  77                 'url': furl,
  78                 'format_id': 'smil_%d' % tbr,
  79                 'ext': 'flv',
  80                 'tbr': tbr,
  81                 'preference': -1000,
  82             })
  83         return formats
  84
  85     def _extract_video_info(self, video_data):
  86         video_id = compat_str(video_data['id'])
  87
  88         FORMAT_KEYS = (
  89             ('sd', 'progressive_url'),
  90             ('hd', 'progressive_url_hd'),
  91         )
  92         formats = [{
  93             'format_id': format_id,
  94             'url': video_data[key],
  95             'quality': i + 1,
  96         } for i, (format_id, key) in enumerate(FORMAT_KEYS)
  97             if video_data.get(key)]
  98
  99         smil_url = video_data.get('smil_url')
 100         if smil_url:
 101             formats.extend(self._parse_smil(video_id, smil_url))
 102         self._sort_formats(formats)
 103
 104         return {
 105             'id': video_id,
 106             'formats': formats,
 107             'title': video_data['caption'],
 108             'thumbnail': video_data.get('thumbnail_url'),
 109             'upload_date': video_data['updated_at'].replace('-', '')[:8],
 110             'like_count': video_data.get('likes', {}).get('total'),
 111             'view_count': video_data.get('views'),
 112         }
 113
 114     def _real_extract(self, url):
 115         mobj = re.match(self._VALID_URL, url)
 116         video_id = mobj.group('id')
 117         event_name = mobj.group('event_name')
 118         webpage = self._download_webpage(url, video_id or event_name)
 119
 120         og_video = self._og_search_video_url(
 121             webpage, 'player url', fatal=False, default=None)
 122         if og_video is not None:
 123             query_str = compat_urllib_parse_urlparse(og_video).query
 124             query = compat_urlparse.parse_qs(query_str)
 125             if 'play_url' in query:
 126                 api_url = query['play_url'][0].replace('.smil', '')
 127                 info = json.loads(self._download_webpage(
 128                     api_url, video_id, 'Downloading video info'))
 129                 return self._extract_video_info(info)
 130
 131         config_json = self._search_regex(
 132             r'window.config = ({.*?});', webpage, 'window config')
 133         info = json.loads(config_json)['event']
 134
 135         def is_relevant(vdata, vid):
 136             result = vdata['type'] == 'video'
 137             if video_id is not None:
 138                 result = result and compat_str(vdata['data']['id']) == vid
 139             return result
 140
 141         videos = [self._extract_video_info(video_data['data'])
 142                   for video_data in info['feed']['data']
 143                   if is_relevant(video_data, video_id)]
 144         if video_id is None:
 145             # This is an event page:
 146             return self.playlist_result(videos, info['id'], info['full_name'])
 147         else:
 148             if not videos:
 149                 raise ExtractorError('Cannot find video %s' % video_id)
 150             return videos[0]
 151
 152
 153 # The original version of Livestream uses a different system
 154 class LivestreamOriginalIE(InfoExtractor):
 155     IE_NAME = 'livestream:original'
 156     _VALID_URL = r'''(?x)https?://www\.livestream\.com/
 157         (?P<user>[^/]+)/(?P<type>video|folder)
 158         (?:\?.*?Id=|/)(?P<id>.*?)(&|$)
 159         '''
 160     _TESTS = [{
 161         'url': 'http://www.livestream.com/dealbook/video?clipId=pla_8aa4a3f1-ba15-46a4-893b-902210e138fb',
 162         'info_dict': {
 163             'id': 'pla_8aa4a3f1-ba15-46a4-893b-902210e138fb',
 164             'ext': 'flv',
 165             'title': 'Spark 1 (BitCoin) with Cameron Winklevoss & Tyler Winklevoss of Winklevoss Capital',
 166         },
 167         'params': {
 168             # rtmp
 169             'skip_download': True,
 170         },
 171     }, {
 172         'url': 'https://www.livestream.com/newplay/folder?dirId=a07bf706-d0e4-4e75-a747-b021d84f2fd3',
 173         'info_dict': {
 174             'id': 'a07bf706-d0e4-4e75-a747-b021d84f2fd3',
 175         },
 176         'playlist_mincount': 4,
 177     }]
 178
 179     def _extract_video(self, user, video_id):
 180         api_url = 'http://x{0}x.api.channel.livestream.com/2.0/clipdetails?extendedInfo=true&id={1}'.format(user, video_id)
 181
 182         info = self._download_xml(api_url, video_id)
 183         item = info.find('channel').find('item')
 184         ns = {'media': 'http://search.yahoo.com/mrss'}
 185         thumbnail_url = item.find(xpath_with_ns('media:thumbnail', ns)).attrib['url']
 186         # Remove the extension and number from the path (like 1.jpg)
 187         path = self._search_regex(r'(user-files/.+)_.*?\.jpg$', thumbnail_url, 'path')
 188
 189         return {
 190             'id': video_id,
 191             'title': item.find('title').text,
 192             'url': 'rtmp://extondemand.livestream.com/ondemand',
 193             'play_path': 'trans/dv15/mogulus-{0}'.format(path),
 194             'player_url': 'http://static.livestream.com/chromelessPlayer/v21/playerapi.swf?hash=5uetk&v=0803&classid=D27CDB6E-AE6D-11cf-96B8-444553540000&jsEnabled=false&wmode=opaque',
 195             'ext': 'flv',
 196             'thumbnail': thumbnail_url,
 197         }
 198
 199     def _extract_folder(self, url, folder_id):
 200         webpage = self._download_webpage(url, folder_id)
 201         paths = orderedSet(re.findall(
 202             r'''(?x)(?:
 203                 <li\s+class="folder">\s*<a\s+href="|
 204                 <a\s+href="(?=https?://livestre\.am/)
 205             )([^"]+)"''', webpage))
 206
 207         return {
 208             '_type': 'playlist',
 209             'id': folder_id,
 210             'entries': [{
 211                 '_type': 'url',
 212                 'url': compat_urlparse.urljoin(url, p),
 213             } for p in paths],
 214         }
 215
 216     def _real_extract(self, url):
 217         mobj = re.match(self._VALID_URL, url)
 218         id = mobj.group('id')
 219         user = mobj.group('user')
 220         url_type = mobj.group('type')
 221         if url_type == 'folder':
 222             return self._extract_folder(url, id)
 223         else:
 224             return self._extract_video(user, id)
 225
 226
 227 # The server doesn't support HEAD request, the generic extractor can't detect
 228 # the redirection
 229 class LivestreamShortenerIE(InfoExtractor):
 230     IE_NAME = 'livestream:shortener'
 231     IE_DESC = False  # Do not list
 232     _VALID_URL = r'https?://livestre\.am/(?P<id>.+)'
 233
 234     def _real_extract(self, url):
 235         mobj = re.match(self._VALID_URL, url)
 236         id = mobj.group('id')
 237         webpage = self._download_webpage(url, id)
 238
 239         return {
 240             '_type': 'url',
 241             'url': self._og_search_url(webpage),
 242         }