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

   1 import json
   2 import re
   3
   4 from .common import InfoExtractor
   5
   6
   7 class TEDIE(InfoExtractor):
   8     _VALID_URL=r'''http://www\.ted\.com/
   9                    (
  10                         ((?P<type_playlist>playlists)/(?P<playlist_id>\d+)) # We have a playlist
  11                         |
  12                         ((?P<type_talk>talks)) # We have a simple talk
  13                    )
  14                    (/lang/(.*?))? # The url may contain the language
  15                    /(?P<name>\w+) # Here goes the name and then ".html"
  16                    '''
  17
  18     @classmethod
  19     def suitable(cls, url):
  20         """Receives a URL and returns True if suitable for this IE."""
  21         return re.match(cls._VALID_URL, url, re.VERBOSE) is not None
  22
  23     def _real_extract(self, url):
  24         m=re.match(self._VALID_URL, url, re.VERBOSE)
  25         if m.group('type_talk'):
  26             return [self._talk_info(url)]
  27         else :
  28             playlist_id=m.group('playlist_id')
  29             name=m.group('name')
  30             self.to_screen(u'Getting info of playlist %s: "%s"' % (playlist_id,name))
  31             return [self._playlist_videos_info(url,name,playlist_id)]
  32
  33     def _playlist_videos_info(self,url,name,playlist_id=0):
  34         '''Returns the videos of the playlist'''
  35         video_RE=r'''
  36                      <li\ id="talk_(\d+)"([.\s]*?)data-id="(?P<video_id>\d+)"
  37                      ([.\s]*?)data-playlist_item_id="(\d+)"
  38                      ([.\s]*?)data-mediaslug="(?P<mediaSlug>.+?)"
  39                      '''
  40         video_name_RE=r'<p\ class="talk-title"><a href="(?P<talk_url>/talks/(.+).html)">(?P<fullname>.+?)</a></p>'
  41         webpage=self._download_webpage(url, playlist_id, 'Downloading playlist webpage')
  42         m_videos=re.finditer(video_RE,webpage,re.VERBOSE)
  43         m_names=re.finditer(video_name_RE,webpage)
  44
  45         playlist_title = self._html_search_regex(r'div class="headline">\s*?<h1>\s*?<span>(.*?)</span>',
  46                                                  webpage, 'playlist title')
  47
  48         playlist_entries = []
  49         for m_video, m_name in zip(m_videos,m_names):
  50             talk_url='http://www.ted.com%s' % m_name.group('talk_url')
  51             playlist_entries.append(self.url_result(talk_url, 'TED'))
  52         return self.playlist_result(playlist_entries, playlist_id = playlist_id, playlist_title = playlist_title)
  53
  54     def _talk_info(self, url, video_id=0):
  55         """Return the video for the talk in the url"""
  56         m = re.match(self._VALID_URL, url,re.VERBOSE)
  57         video_name = m.group('name')
  58         webpage = self._download_webpage(url, video_id, 'Downloading \"%s\" page' % video_name)
  59         self.report_extraction(video_name)
  60         # If the url includes the language we get the title translated
  61         title = self._html_search_regex(r'<span id="altHeadline" >(?P<title>.*)</span>',
  62                                         webpage, 'title')
  63         json_data = self._search_regex(r'<script.*?>var talkDetails = ({.*?})</script>',
  64                                     webpage, 'json data')
  65         info = json.loads(json_data)
  66         desc = self._html_search_regex(r'<div class="talk-intro">.*?<p.*?>(.*?)</p>',
  67                                        webpage, 'description', flags = re.DOTALL)
  68
  69         thumbnail = self._search_regex(r'</span>[\s.]*</div>[\s.]*<img src="(.*?)"',
  70                                        webpage, 'thumbnail')
  71         info = {
  72                 'id': info['id'],
  73                 'url': info['htmlStreams'][-1]['file'],
  74                 'ext': 'mp4',
  75                 'title': title,
  76                 'thumbnail': thumbnail,
  77                 'description': desc,
  78                 }
  79         return info