_ Git - youtube-dl/blob - youtube_dl/extractor/ustream.py

   1 import json
   2 import re
   3
   4 from .common import InfoExtractor
   5 from ..utils import (
   6     compat_urlparse,
   7     compat_html_parser,
   8 )
   9
  10
  11 class UstreamIE(InfoExtractor):
  12     _VALID_URL = r'https?://www\.ustream\.tv/recorded/(?P<videoID>\d+)'
  13     IE_NAME = u'ustream'
  14     _TEST = {
  15         u'url': u'http://www.ustream.tv/recorded/20274954',
  16         u'file': u'20274954.flv',
  17         u'md5': u'088f151799e8f572f84eb62f17d73e5c',
  18         u'info_dict': {
  19             u"uploader": u"Young Americans for Liberty",
  20             u"title": u"Young Americans for Liberty February 7, 2012 2:28 AM"
  21         }
  22     }
  23
  24     def _real_extract(self, url):
  25         m = re.match(self._VALID_URL, url)
  26         video_id = m.group('videoID')
  27
  28         video_url = u'http://tcdn.ustream.tv/video/%s' % video_id
  29         webpage = self._download_webpage(url, video_id)
  30
  31         self.report_extraction(video_id)
  32
  33         video_title = self._html_search_regex(r'data-title="(?P<title>.+)"',
  34             webpage, u'title')
  35
  36         uploader = self._html_search_regex(r'data-content-type="channel".*?>(?P<uploader>.*?)</a>',
  37             webpage, u'uploader', fatal=False, flags=re.DOTALL)
  38
  39         thumbnail = self._html_search_regex(r'<link rel="image_src" href="(?P<thumb>.*?)"',
  40             webpage, u'thumbnail', fatal=False)
  41
  42         info = {
  43                 'id': video_id,
  44                 'url': video_url,
  45                 'ext': 'flv',
  46                 'title': video_title,
  47                 'uploader': uploader,
  48                 'thumbnail': thumbnail,
  49                }
  50         return info
  51
  52 # More robust than regular expressions
  53
  54 class ChannelParser(compat_html_parser.HTMLParser):
  55     """
  56     <meta name="ustream:channel_id" content="1234">
  57     """
  58     channel_id = None
  59
  60     def handle_starttag(self, tag, attrs):
  61         if tag != 'meta':
  62             return
  63         values = dict(attrs)
  64         if values.get('name') != 'ustream:channel_id':
  65             return
  66         value = values.get('content', '')
  67         if value.isdigit():
  68             self.channel_id = value
  69
  70 class SocialstreamParser(compat_html_parser.HTMLParser):
  71     """
  72     <li class="content123 video" data-content-id="123" data-length="1452"
  73         data-href="/recorded/123" data-og-url="/recorded/123">
  74     """
  75     def __init__(self):
  76         compat_html_parser.HTMLParser.__init__(self)
  77         self.content_ids = []
  78
  79     def handle_starttag(self, tag, attrs):
  80         if tag != 'li':
  81             return
  82         for (attr, value) in attrs:
  83             if attr == 'data-content-id' and value.isdigit():
  84                 self.content_ids.append(value)
  85
  86 class UstreamChannelIE(InfoExtractor):
  87     _VALID_URL = r'https?://www\.ustream\.tv/channel/(?P<slug>.+)'
  88     IE_NAME = u'ustream:channel'
  89
  90     def _real_extract(self, url):
  91         m = re.match(self._VALID_URL, url)
  92         slug = m.group('slug')
  93
  94         p = ChannelParser()
  95         p.feed(self._download_webpage(url, slug))
  96         p.close()
  97         channel_id = p.channel_id
  98
  99         p = SocialstreamParser()
 100         BASE = 'http://www.ustream.tv'
 101         next_url = '/ajax/socialstream/videos/%s/1.json' % channel_id
 102         while next_url:
 103             reply = json.loads(self._download_webpage(compat_urlparse.urljoin(BASE, next_url), channel_id))
 104             p.feed(reply['data'])
 105             next_url = reply['nextUrl']
 106         p.close()
 107         video_ids = p.content_ids
 108
 109         urls = ['http://www.ustream.tv/recorded/' + vid for vid in video_ids]
 110         url_entries = [self.url_result(eurl, 'Ustream') for eurl in urls]
 111         return self.playlist_result(url_entries, channel_id)