_ 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&seek=%s' %
  77                     (vn.attrib['src'], vn.attrib['clipBegin']))
  78                 formats.append({
  79                     'url': furl,
  80                     'format_id': 'smil_%d' % tbr,
  81                     'ext': 'flv',
  82                     'tbr': tbr,
  83                     'preference': -1000,
  84                 })
  85         self._sort_formats(formats)
  86
  87         return {
  88             'id': video_id,
  89             'formats': formats,
  90             'title': video_data['caption'],
  91             'thumbnail': video_data.get('thumbnail_url'),
  92             'upload_date': video_data['updated_at'].replace('-', '')[:8],
  93             'like_count': video_data.get('likes', {}).get('total'),
  94             'view_count': video_data.get('views'),
  95         }
  96
  97     def _real_extract(self, url):
  98         mobj = re.match(self._VALID_URL, url)
  99         video_id = mobj.group('id')
 100         event_name = mobj.group('event_name')
 101         webpage = self._download_webpage(url, video_id or event_name)
 102
 103         if video_id is None:
 104             # This is an event page:
 105             config_json = self._search_regex(
 106                 r'window.config = ({.*?});', webpage, 'window config')
 107             info = json.loads(config_json)['event']
 108             videos = [self._extract_video_info(video_data['data'])
 109                 for video_data in info['feed']['data']
 110                 if video_data['type'] == 'video']
 111             return self.playlist_result(videos, info['id'], info['full_name'])
 112         else:
 113             og_video = self._og_search_video_url(webpage, 'player url')
 114             query_str = compat_urllib_parse_urlparse(og_video).query
 115             query = compat_urlparse.parse_qs(query_str)
 116             api_url = query['play_url'][0].replace('.smil', '')
 117             info = json.loads(self._download_webpage(
 118                 api_url, video_id, 'Downloading video info'))
 119             return self._extract_video_info(info)
 120
 121
 122 # The original version of Livestream uses a different system
 123 class LivestreamOriginalIE(InfoExtractor):
 124     IE_NAME = 'livestream:original'
 125     _VALID_URL = r'''(?x)https?://www\.livestream\.com/
 126         (?P<user>[^/]+)/(?P<type>video|folder)
 127         (?:\?.*?Id=|/)(?P<id>.*?)(&|$)
 128         '''
 129     _TEST = {
 130         'url': 'http://www.livestream.com/dealbook/video?clipId=pla_8aa4a3f1-ba15-46a4-893b-902210e138fb',
 131         'info_dict': {
 132             'id': 'pla_8aa4a3f1-ba15-46a4-893b-902210e138fb',
 133             'ext': 'flv',
 134             'title': 'Spark 1 (BitCoin) with Cameron Winklevoss & Tyler Winklevoss of Winklevoss Capital',
 135         },
 136         'params': {
 137             # rtmp
 138             'skip_download': True,
 139         },
 140     }
 141
 142     def _extract_video(self, user, video_id):
 143         api_url = 'http://x{0}x.api.channel.livestream.com/2.0/clipdetails?extendedInfo=true&id={1}'.format(user, video_id)
 144
 145         info = self._download_xml(api_url, video_id)
 146         item = info.find('channel').find('item')
 147         ns = {'media': 'http://search.yahoo.com/mrss'}
 148         thumbnail_url = item.find(xpath_with_ns('media:thumbnail', ns)).attrib['url']
 149         # Remove the extension and number from the path (like 1.jpg)
 150         path = self._search_regex(r'(user-files/.+)_.*?\.jpg$', thumbnail_url, 'path')
 151
 152         return {
 153             'id': video_id,
 154             'title': item.find('title').text,
 155             'url': 'rtmp://extondemand.livestream.com/ondemand',
 156             'play_path': 'mp4:trans/dv15/mogulus-{0}.mp4'.format(path),
 157             'ext': 'flv',
 158             'thumbnail': thumbnail_url,
 159         }
 160
 161     def _extract_folder(self, url, folder_id):
 162         webpage = self._download_webpage(url, folder_id)
 163         urls = orderedSet(re.findall(r'<a href="(https?://livestre\.am/.*?)"', webpage))
 164
 165         return {
 166             '_type': 'playlist',
 167             'id': folder_id,
 168             'entries': [{
 169                 '_type': 'url',
 170                 'url': video_url,
 171             } for video_url in urls],
 172         }
 173
 174     def _real_extract(self, url):
 175         mobj = re.match(self._VALID_URL, url)
 176         id = mobj.group('id')
 177         user = mobj.group('user')
 178         url_type = mobj.group('type')
 179         if url_type == 'folder':
 180             return self._extract_folder(url, id)
 181         else:
 182             return self._extract_video(user, id)
 183
 184
 185 # The server doesn't support HEAD request, the generic extractor can't detect
 186 # the redirection
 187 class LivestreamShortenerIE(InfoExtractor):
 188     IE_NAME = 'livestream:shortener'
 189     IE_DESC = False  # Do not list
 190     _VALID_URL = r'https?://livestre\.am/(?P<id>.+)'
 191
 192     def _real_extract(self, url):
 193         mobj = re.match(self._VALID_URL, url)
 194         id = mobj.group('id')
 195         webpage = self._download_webpage(url, id)
 196
 197         return {
 198             '_type': 'url',
 199             'url': self._og_search_url(webpage),
 200         }