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

   1 import json
   2 import re
   3
   4 from .subtitles import SubtitlesInfoExtractor
   5
   6 class TEDIE(SubtitlesInfoExtractor):
   7     _VALID_URL=r'''http://www\.ted\.com/
   8                    (
   9                         ((?P<type_playlist>playlists)/(?P<playlist_id>\d+)) # We have a playlist
  10                         |
  11                         ((?P<type_talk>talks)) # We have a simple talk
  12                    )
  13                    (/lang/(.*?))? # The url may contain the language
  14                    /(?P<name>\w+) # Here goes the name and then ".html"
  15                    '''
  16     _TEST = {
  17         u'url': u'http://www.ted.com/talks/dan_dennett_on_our_consciousness.html',
  18         u'file': u'102.mp4',
  19         u'md5': u'2d76ee1576672e0bd8f187513267adf6',
  20         u'info_dict': {
  21             u"description": u"md5:c6fa72e6eedbd938c9caf6b2702f5922",
  22             u"title": u"Dan Dennett: The illusion of consciousness"
  23         }
  24     }
  25
  26     @classmethod
  27     def suitable(cls, url):
  28         """Receives a URL and returns True if suitable for this IE."""
  29         return re.match(cls._VALID_URL, url, re.VERBOSE) is not None
  30
  31     def _real_extract(self, url):
  32         m=re.match(self._VALID_URL, url, re.VERBOSE)
  33         if m.group('type_talk'):
  34             return [self._talk_info(url)]
  35         else :
  36             playlist_id=m.group('playlist_id')
  37             name=m.group('name')
  38             self.to_screen(u'Getting info of playlist %s: "%s"' % (playlist_id,name))
  39             return [self._playlist_videos_info(url,name,playlist_id)]
  40
  41     def _playlist_videos_info(self,url,name,playlist_id=0):
  42         '''Returns the videos of the playlist'''
  43         video_RE=r'''
  44                      <li\ id="talk_(\d+)"([.\s]*?)data-id="(?P<video_id>\d+)"
  45                      ([.\s]*?)data-playlist_item_id="(\d+)"
  46                      ([.\s]*?)data-mediaslug="(?P<mediaSlug>.+?)"
  47                      '''
  48         video_name_RE=r'<p\ class="talk-title"><a href="(?P<talk_url>/talks/(.+).html)">(?P<fullname>.+?)</a></p>'
  49         webpage=self._download_webpage(url, playlist_id, 'Downloading playlist webpage')
  50         m_videos=re.finditer(video_RE,webpage,re.VERBOSE)
  51         m_names=re.finditer(video_name_RE,webpage)
  52
  53         playlist_title = self._html_search_regex(r'div class="headline">\s*?<h1>\s*?<span>(.*?)</span>',
  54                                                  webpage, 'playlist title')
  55
  56         playlist_entries = []
  57         for m_video, m_name in zip(m_videos,m_names):
  58             talk_url='http://www.ted.com%s' % m_name.group('talk_url')
  59             playlist_entries.append(self.url_result(talk_url, 'TED'))
  60         return self.playlist_result(playlist_entries, playlist_id = playlist_id, playlist_title = playlist_title)
  61
  62     def _talk_info(self, url, video_id=0):
  63         """Return the video for the talk in the url"""
  64         m = re.match(self._VALID_URL, url,re.VERBOSE)
  65         video_name = m.group('name')
  66         webpage = self._download_webpage(url, video_id, 'Downloading \"%s\" page' % video_name)
  67         self.report_extraction(video_name)
  68         # If the url includes the language we get the title translated
  69         title = self._html_search_regex(r'<span .*?id="altHeadline".+?>(?P<title>.*)</span>',
  70                                         webpage, 'title')
  71         json_data = self._search_regex(r'<script.*?>var talkDetails = ({.*?})</script>',
  72                                     webpage, 'json data')
  73         info = json.loads(json_data)
  74         desc = self._html_search_regex(r'<div class="talk-intro">.*?<p.*?>(.*?)</p>',
  75                                        webpage, 'description', flags = re.DOTALL)
  76
  77         thumbnail = self._search_regex(r'</span>[\s.]*</div>[\s.]*<img src="(.*?)"',
  78                                        webpage, 'thumbnail')
  79         formats = [{
  80             'ext': 'mp4',
  81             'url': stream['file'],
  82             'format': stream['id']
  83             } for stream in info['htmlStreams']]
  84
  85         video_id = info['id']
  86
  87         # subtitles
  88         video_subtitles = self.extract_subtitles(video_id, webpage)
  89         if self._downloader.params.get('listsubtitles', False):
  90             self._list_available_subtitles(video_id, webpage)
  91             return
  92
  93         info = {
  94             'id': video_id,
  95             'title': title,
  96             'thumbnail': thumbnail,
  97             'description': desc,
  98             'subtitles': video_subtitles,
  99             'formats': formats,
 100         }
 101
 102         # TODO: Remove when #980 has been merged
 103         info.update(info['formats'][-1])
 104
 105         return info
 106
 107     def _get_available_subtitles(self, video_id, webpage):
 108         options = self._search_regex(r'(?:<select name="subtitles_language_select" id="subtitles_language_select">)(.*?)(?:</select>)', webpage, 'subtitles_language_select', flags=re.DOTALL)
 109         languages = re.findall(r'(?:<option value=")(\S+)"', options)
 110         if languages:
 111             sub_lang_list = {}
 112             for l in languages:
 113                 url = 'http://www.ted.com/talks/subtitles/id/%s/lang/%s/format/srt' % (video_id, l)
 114                 sub_lang_list[l] = url
 115             return sub_lang_list
 116         return {}