Raphaël G. Git Repositories - youtubedl/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     _TEST = {
  18         u'url': u'http://www.ted.com/talks/dan_dennett_on_our_consciousness.html',
  19         u'file': u'102.mp4',
  20         u'md5': u'2d76ee1576672e0bd8f187513267adf6',
  21         u'info_dict': {
  22             u"description": u"md5:c6fa72e6eedbd938c9caf6b2702f5922",
  23             u"title": u"Dan Dennett: The illusion of consciousness"
  24         }
  25     }
  26
  27     @classmethod
  28     def suitable(cls, url):
  29         """Receives a URL and returns True if suitable for this IE."""
  30         return re.match(cls._VALID_URL, url, re.VERBOSE) is not None
  31
  32     def _real_extract(self, url):
  33         m=re.match(self._VALID_URL, url, re.VERBOSE)
  34         if m.group('type_talk'):
  35             return [self._talk_info(url)]
  36         else :
  37             playlist_id=m.group('playlist_id')
  38             name=m.group('name')
  39             self.to_screen(u'Getting info of playlist %s: "%s"' % (playlist_id,name))
  40             return [self._playlist_videos_info(url,name,playlist_id)]
  41
  42     def _playlist_videos_info(self,url,name,playlist_id=0):
  43         '''Returns the videos of the playlist'''
  44         video_RE=r'''
  45                      <li\ id="talk_(\d+)"([.\s]*?)data-id="(?P<video_id>\d+)"
  46                      ([.\s]*?)data-playlist_item_id="(\d+)"
  47                      ([.\s]*?)data-mediaslug="(?P<mediaSlug>.+?)"
  48                      '''
  49         video_name_RE=r'<p\ class="talk-title"><a href="(?P<talk_url>/talks/(.+).html)">(?P<fullname>.+?)</a></p>'
  50         webpage=self._download_webpage(url, playlist_id, 'Downloading playlist webpage')
  51         m_videos=re.finditer(video_RE,webpage,re.VERBOSE)
  52         m_names=re.finditer(video_name_RE,webpage)
  53
  54         playlist_title = self._html_search_regex(r'div class="headline">\s*?<h1>\s*?<span>(.*?)</span>',
  55                                                  webpage, 'playlist title')
  56
  57         playlist_entries = []
  58         for m_video, m_name in zip(m_videos,m_names):
  59             talk_url='http://www.ted.com%s' % m_name.group('talk_url')
  60             playlist_entries.append(self.url_result(talk_url, 'TED'))
  61         return self.playlist_result(playlist_entries, playlist_id = playlist_id, playlist_title = playlist_title)
  62
  63     def _talk_info(self, url, video_id=0):
  64         """Return the video for the talk in the url"""
  65         m = re.match(self._VALID_URL, url,re.VERBOSE)
  66         video_name = m.group('name')
  67         webpage = self._download_webpage(url, video_id, 'Downloading \"%s\" page' % video_name)
  68         self.report_extraction(video_name)
  69         # If the url includes the language we get the title translated
  70         title = self._html_search_regex(r'<span .*?id="altHeadline".+?>(?P<title>.*)</span>',
  71                                         webpage, 'title')
  72         json_data = self._search_regex(r'<script.*?>var talkDetails = ({.*?})</script>',
  73                                     webpage, 'json data')
  74         info = json.loads(json_data)
  75         desc = self._html_search_regex(r'<div class="talk-intro">.*?<p.*?>(.*?)</p>',
  76                                        webpage, 'description', flags = re.DOTALL)
  77
  78         thumbnail = self._search_regex(r'</span>[\s.]*</div>[\s.]*<img src="(.*?)"',
  79                                        webpage, 'thumbnail')
  80         formats = [{
  81             'ext': 'mp4',
  82             'url': stream['file'],
  83             'format': stream['id']
  84             } for stream in info['htmlStreams']]
  85         info = {
  86             'id': info['id'],
  87             'title': title,
  88             'thumbnail': thumbnail,
  89             'description': desc,
  90             'formats': formats,
  91         }
  92
  93         # TODO: Remove when #980 has been merged
  94         info.update(info['formats'][-1])
  95
  96         return info