_ 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_urllib_parse_urlparse,
   9     compat_urlparse,
  10     xpath_with_ns,
  11     compat_str,
  12     orderedSet,
  13 )
  14
  15
  16 class LivestreamIE(InfoExtractor):
  17     IE_NAME = 'livestream'
  18     _VALID_URL = r'http://new\.livestream\.com/.*?/(?P<event_name>.*?)(/videos/(?P<id>\d+))?/?$'
  19     _TEST = {
  20         'url': 'http://new.livestream.com/CoheedandCambria/WebsterHall/videos/4719370',
  21         'md5': '53274c76ba7754fb0e8d072716f2292b',
  22         'info_dict': {
  23             'id': '4719370',
  24             'ext': 'mp4',
  25             'title': 'Live from Webster Hall NYC',
  26             'upload_date': '20121012',
  27         }
  28     }
  29
  30     def _extract_video_info(self, video_data):
  31         video_url = (
  32             video_data.get('progressive_url_hd') or
  33             video_data.get('progressive_url')
  34         )
  35         return {
  36             'id': compat_str(video_data['id']),
  37             'url': video_url,
  38             'title': video_data['caption'],
  39             'thumbnail': video_data['thumbnail_url'],
  40             'upload_date': video_data['updated_at'].replace('-', '')[:8],
  41         }
  42
  43     def _real_extract(self, url):
  44         mobj = re.match(self._VALID_URL, url)
  45         video_id = mobj.group('id')
  46         event_name = mobj.group('event_name')
  47         webpage = self._download_webpage(url, video_id or event_name)
  48
  49         if video_id is None:
  50             # This is an event page:
  51             config_json = self._search_regex(
  52                 r'window.config = ({.*?});', webpage, 'window config')
  53             info = json.loads(config_json)['event']
  54             videos = [self._extract_video_info(video_data['data'])
  55                 for video_data in info['feed']['data']
  56                 if video_data['type'] == 'video']
  57             return self.playlist_result(videos, info['id'], info['full_name'])
  58         else:
  59             og_video = self._og_search_video_url(webpage, 'player url')
  60             query_str = compat_urllib_parse_urlparse(og_video).query
  61             query = compat_urlparse.parse_qs(query_str)
  62             api_url = query['play_url'][0].replace('.smil', '')
  63             info = json.loads(self._download_webpage(
  64                 api_url, video_id, 'Downloading video info'))
  65             return self._extract_video_info(info)
  66
  67
  68 # The original version of Livestream uses a different system
  69 class LivestreamOriginalIE(InfoExtractor):
  70     IE_NAME = 'livestream:original'
  71     _VALID_URL = r'''(?x)https?://www\.livestream\.com/
  72         (?P<user>[^/]+)/(?P<type>video|folder)
  73         (?:\?.*?Id=|/)(?P<id>.*?)(&|$)
  74         '''
  75     _TEST = {
  76         'url': 'http://www.livestream.com/dealbook/video?clipId=pla_8aa4a3f1-ba15-46a4-893b-902210e138fb',
  77         'info_dict': {
  78             'id': 'pla_8aa4a3f1-ba15-46a4-893b-902210e138fb',
  79             'ext': 'flv',
  80             'title': 'Spark 1 (BitCoin) with Cameron Winklevoss & Tyler Winklevoss of Winklevoss Capital',
  81         },
  82         'params': {
  83             # rtmp
  84             'skip_download': True,
  85         },
  86     }
  87
  88     def _extract_video(self, user, video_id):
  89         api_url = 'http://x{0}x.api.channel.livestream.com/2.0/clipdetails?extendedInfo=true&id={1}'.format(user, video_id)
  90
  91         info = self._download_xml(api_url, video_id)
  92         item = info.find('channel').find('item')
  93         ns = {'media': 'http://search.yahoo.com/mrss'}
  94         thumbnail_url = item.find(xpath_with_ns('media:thumbnail', ns)).attrib['url']
  95         # Remove the extension and number from the path (like 1.jpg)
  96         path = self._search_regex(r'(user-files/.+)_.*?\.jpg$', thumbnail_url, 'path')
  97
  98         return {
  99             'id': video_id,
 100             'title': item.find('title').text,
 101             'url': 'rtmp://extondemand.livestream.com/ondemand',
 102             'play_path': 'mp4:trans/dv15/mogulus-{0}.mp4'.format(path),
 103             'ext': 'flv',
 104             'thumbnail': thumbnail_url,
 105         }
 106
 107     def _extract_folder(self, url, folder_id):
 108         webpage = self._download_webpage(url, folder_id)
 109         urls = orderedSet(re.findall(r'<a href="(https?://livestre\.am/.*?)"', webpage))
 110
 111         return {
 112             '_type': 'playlist',
 113             'id': folder_id,
 114             'entries': [{
 115                 '_type': 'url',
 116                 'url': video_url,
 117             } for video_url in urls],
 118         }
 119
 120     def _real_extract(self, url):
 121         mobj = re.match(self._VALID_URL, url)
 122         id = mobj.group('id')
 123         user = mobj.group('user')
 124         url_type = mobj.group('type')
 125         if url_type == 'folder':
 126             return self._extract_folder(url, id)
 127         else:
 128             return self._extract_video(user, id)
 129
 130
 131 # The server doesn't support HEAD request, the generic extractor can't detect
 132 # the redirection
 133 class LivestreamShortenerIE(InfoExtractor):
 134     IE_NAME = 'livestream:shortener'
 135     IE_DESC = False  # Do not list
 136     _VALID_URL = r'https?://livestre\.am/(?P<id>.+)'
 137
 138     def _real_extract(self, url):
 139         mobj = re.match(self._VALID_URL, url)
 140         id = mobj.group('id')
 141         webpage = self._download_webpage(url, id)
 142
 143         return {
 144             '_type': 'url',
 145             'url': self._og_search_url(webpage),
 146         }