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