_ 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     RegexNotFoundError,
  11 )
  12
  13
  14 class TEDIE(SubtitlesInfoExtractor):
  15     _VALID_URL = r'''(?x)http://www\.ted\.com/
  16         (
  17             (?P<type_playlist>playlists(?:/\d+)?) # We have a playlist
  18             |
  19             ((?P<type_talk>talks)) # We have a simple talk
  20         )
  21         (/lang/(.*?))? # The url may contain the language
  22         /(?P<name>\w+) # Here goes the name and then ".html"
  23         '''
  24     _TEST = {
  25         'url': 'http://www.ted.com/talks/dan_dennett_on_our_consciousness.html',
  26         'file': '102.mp4',
  27         'md5': '4ea1dada91e4174b53dac2bb8ace429d',
  28         'info_dict': {
  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         name = m.group('name')
  52         if m.group('type_talk'):
  53             return self._talk_info(url, name)
  54         else:
  55             return self._playlist_videos_info(url, name)
  56
  57     def _playlist_videos_info(self, url, name):
  58         '''Returns the videos of the playlist'''
  59
  60         webpage = self._download_webpage(url, name,
  61             'Downloading playlist webpage')
  62         info = self._extract_info(webpage)
  63         playlist_info = info['playlist']
  64
  65         playlist_entries = [
  66             self.url_result(u'http://www.ted.com/talks/' + talk['slug'], self.ie_key())
  67             for talk in info['talks']
  68         ]
  69         return self.playlist_result(
  70             playlist_entries,
  71             playlist_id=compat_str(playlist_info['id']),
  72             playlist_title=playlist_info['title'])
  73
  74     def _talk_info(self, url, video_name):
  75         webpage = self._download_webpage(url, video_name)
  76         self.report_extraction(video_name)
  77
  78         talk_info = self._extract_info(webpage)['talks'][0]
  79
  80         formats = [{
  81             'ext': 'mp4',
  82             'url': format_url,
  83             'format_id': format_id,
  84             'format': format_id,
  85             'preference': self._FORMATS_PREFERENCE.get(format_id, -1),
  86         } for (format_id, format_url) in talk_info['nativeDownloads'].items()]
  87         self._sort_formats(formats)
  88
  89         video_id = talk_info['id']
  90         # subtitles
  91         video_subtitles = self.extract_subtitles(video_id, talk_info)
  92         if self._downloader.params.get('listsubtitles', False):
  93             self._list_available_subtitles(video_id, talk_info)
  94             return
  95
  96         return {
  97             'id': video_id,
  98             'title': talk_info['title'],
  99             'uploader': talk_info['speaker'],
 100             'thumbnail': talk_info['thumb'],
 101             'description': self._og_search_description(webpage),
 102             'subtitles': video_subtitles,
 103             'formats': formats,
 104         }
 105
 106     def _get_available_subtitles(self, video_id, talk_info):
 107         languages = [lang['languageCode'] for lang in talk_info.get('languages', [])]
 108         if languages:
 109             sub_lang_list = {}
 110             for l in languages:
 111                 url = 'http://www.ted.com/talks/subtitles/id/%s/lang/%s/format/srt' % (video_id, l)
 112                 sub_lang_list[l] = url
 113             return sub_lang_list
 114         else:
 115             self._downloader.report_warning(u'video doesn\'t have subtitles')
 116             return {}