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