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

   1 from HTMLParser import HTMLParser
   2 import json
   3 import re
   4 from urlparse import urljoin
   5
   6 from .common import InfoExtractor
   7
   8
   9 class UstreamIE(InfoExtractor):
  10     _VALID_URL = r'https?://www\.ustream\.tv/recorded/(?P<videoID>\d+)'
  11     IE_NAME = u'ustream'
  12     _TEST = {
  13         u'url': u'http://www.ustream.tv/recorded/20274954',
  14         u'file': u'20274954.flv',
  15         u'md5': u'088f151799e8f572f84eb62f17d73e5c',
  16         u'info_dict': {
  17             u"uploader": u"Young Americans for Liberty",
  18             u"title": u"Young Americans for Liberty February 7, 2012 2:28 AM"
  19         }
  20     }
  21
  22     def _real_extract(self, url):
  23         m = re.match(self._VALID_URL, url)
  24         video_id = m.group('videoID')
  25
  26         video_url = u'http://tcdn.ustream.tv/video/%s' % video_id
  27         webpage = self._download_webpage(url, video_id)
  28
  29         self.report_extraction(video_id)
  30
  31         video_title = self._html_search_regex(r'data-title="(?P<title>.+)"',
  32             webpage, u'title')
  33
  34         uploader = self._html_search_regex(r'data-content-type="channel".*?>(?P<uploader>.*?)</a>',
  35             webpage, u'uploader', fatal=False, flags=re.DOTALL)
  36
  37         thumbnail = self._html_search_regex(r'<link rel="image_src" href="(?P<thumb>.*?)"',
  38             webpage, u'thumbnail', fatal=False)
  39
  40         info = {
  41                 'id': video_id,
  42                 'url': video_url,
  43                 'ext': 'flv',
  44                 'title': video_title,
  45                 'uploader': uploader,
  46                 'thumbnail': thumbnail,
  47                }
  48         return info
  49
  50 # More robust than regular expressions
  51
  52 class ChannelParser(HTMLParser):
  53     """
  54     <meta name="ustream:channel_id" content="1234">
  55     """
  56     channel_id = None
  57
  58     def handle_starttag(self, tag, attrs):
  59         if tag != 'meta':
  60             return
  61         values = dict(attrs)
  62         if values.get('name') != 'ustream:channel_id':
  63             return
  64         value = values.get('content', '')
  65         if value.isdigit():
  66             self.channel_id = value
  67
  68 class SocialstreamParser(HTMLParser):
  69     """
  70     <li class="content123 video" data-content-id="123" data-length="1452"
  71         data-href="/recorded/123" data-og-url="/recorded/123">
  72     """
  73     def __init__(self):
  74         HTMLParser.__init__(self)
  75         self.content_ids = []
  76
  77     def handle_starttag(self, tag, attrs):
  78         if tag != 'li':
  79             return
  80         for (attr, value) in attrs:
  81             if attr == 'data-content-id' and value.isdigit():
  82                 self.content_ids.append(value)
  83
  84 class UstreamChannelIE(InfoExtractor):
  85     _VALID_URL = r'https?://www\.ustream\.tv/channel/(?P<slug>.+)'
  86     IE_NAME = u'ustream:channel'
  87
  88     def _real_extract(self, url):
  89         m = re.match(self._VALID_URL, url)
  90         slug = m.group('slug')
  91         # Slugs can be non-ascii, but youtube-dl can't handle non-ascii command lines,
  92         # so if we got this far it's probably percent encoded and we needn't worry.
  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(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         # From YoutubeChannelIE
 110
 111         self._downloader.to_screen(u'[ustream] Channel %s: Found %i videos' % (channel_id, len(video_ids)))
 112
 113         urls = ['http://www.ustream.tv/recorded/' + vid for vid in video_ids]
 114         url_entries = [self.url_result(eurl, 'Ustream') for eurl in urls]
 115         return [self.playlist_result(url_entries, channel_id)]