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

   1 from __future__ import unicode_literals
   2
   3 import json
   4 import re
   5
   6 from .subtitles import SubtitlesInfoExtractor
   7
   8 from ..utils import (
   9     compat_str,
  10 )
  11
  12
  13 class TEDIE(SubtitlesInfoExtractor):
  14     _VALID_URL = r'''(?x)http://(?P<type>www|embed)\.ted\.com/
  15         (
  16             (?P<type_playlist>playlists(?:/\d+)?) # We have a playlist
  17             |
  18             ((?P<type_talk>talks)) # We have a simple talk
  19         )
  20         (/lang/(.*?))? # The url may contain the language
  21         /(?P<name>\w+) # Here goes the name and then ".html"
  22         '''
  23     _TEST = {
  24         'url': 'http://www.ted.com/talks/dan_dennett_on_our_consciousness.html',
  25         'md5': '4ea1dada91e4174b53dac2bb8ace429d',
  26         'info_dict': {
  27             'id': '102',
  28             'ext': 'mp4',
  29             'title': 'The illusion of consciousness',
  30             'description': ('Philosopher Dan Dennett makes a compelling '
  31                 'argument that not only don\'t we understand our own '
  32                 'consciousness, but that half the time our brains are '
  33                 'actively fooling us.'),
  34             'uploader': 'Dan Dennett',
  35         }
  36     }
  37
  38     _FORMATS_PREFERENCE = {
  39         'low': 1,
  40         'medium': 2,
  41         'high': 3,
  42     }
  43
  44     def _extract_info(self, webpage):
  45         info_json = self._search_regex(r'q\("\w+.init",({.+})\)</script>',
  46             webpage, 'info json')
  47         return json.loads(info_json)
  48
  49     def _real_extract(self, url):
  50         m = re.match(self._VALID_URL, url, re.VERBOSE)
  51         if m.group('type') == 'embed': # if the _VALID_URL is an embed
  52             desktop_url = re.sub("embed", "www", url)
  53             return self.url_result(desktop_url, 'TED') # pass the desktop version to the extractor
  54         name = m.group('name')
  55         if m.group('type_talk'):
  56             return self._talk_info(url, name)
  57         else:
  58             return self._playlist_videos_info(url, name)
  59
  60     def _playlist_videos_info(self, url, name):
  61         '''Returns the videos of the playlist'''
  62
  63         webpage = self._download_webpage(url, name,
  64             'Downloading playlist webpage')
  65         info = self._extract_info(webpage)
  66         playlist_info = info['playlist']
  67
  68         playlist_entries = [
  69             self.url_result(u'http://www.ted.com/talks/' + talk['slug'], self.ie_key())
  70             for talk in info['talks']
  71         ]
  72         return self.playlist_result(
  73             playlist_entries,
  74             playlist_id=compat_str(playlist_info['id']),
  75             playlist_title=playlist_info['title'])
  76
  77     def _talk_info(self, url, video_name):
  78         webpage = self._download_webpage(url, video_name)
  79         self.report_extraction(video_name)
  80
  81         talk_info = self._extract_info(webpage)['talks'][0]
  82
  83         formats = [{
  84             'ext': 'mp4',
  85             'url': format_url,
  86             'format_id': format_id,
  87             'format': format_id,
  88             'preference': self._FORMATS_PREFERENCE.get(format_id, -1),
  89         } for (format_id, format_url) in talk_info['nativeDownloads'].items()]
  90         self._sort_formats(formats)
  91
  92         video_id = compat_str(talk_info['id'])
  93         # subtitles
  94         video_subtitles = self.extract_subtitles(video_id, talk_info)
  95         if self._downloader.params.get('listsubtitles', False):
  96             self._list_available_subtitles(video_id, talk_info)
  97             return
  98
  99         thumbnail = talk_info['thumb']
 100         if not thumbnail.startswith('http'):
 101             thumbnail = 'http://' + thumbnail
 102         return {
 103             'id': video_id,
 104             'title': talk_info['title'],
 105             'uploader': talk_info['speaker'],
 106             'thumbnail': thumbnail,
 107             'description': self._og_search_description(webpage),
 108             'subtitles': video_subtitles,
 109             'formats': formats,
 110         }
 111
 112     def _get_available_subtitles(self, video_id, talk_info):
 113         languages = [lang['languageCode'] for lang in talk_info.get('languages', [])]
 114         if languages:
 115             sub_lang_list = {}
 116             for l in languages:
 117                 url = 'http://www.ted.com/talks/subtitles/id/%s/lang/%s/format/srt' % (video_id, l)
 118                 sub_lang_list[l] = url
 119             return sub_lang_list
 120         else:
 121             self._downloader.report_warning(u'video doesn\'t have subtitles')
 122             return {}