_ 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     RegexNotFoundError,
  10 )
  11
  12
  13 class TEDIE(SubtitlesInfoExtractor):
  14     _VALID_URL=r'''(?x)http://www\.ted\.com/
  15                    (
  16                         ((?P<type_playlist>playlists)/(?P<playlist_id>\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         'file': '102.mp4',
  26         'md5': '4ea1dada91e4174b53dac2bb8ace429d',
  27         'info_dict': {
  28             'title': 'The illusion of consciousness',
  29             'description': 'Philosopher Dan Dennett makes a compelling argument that not only don\'t we understand our own consciousness, but that half the time our brains are actively fooling us.',
  30             'uploader': 'Dan Dennett',
  31         }
  32     }
  33
  34     _FORMATS_PREFERENCE = {
  35         'low': 1,
  36         'medium': 2,
  37         'high': 3,
  38     }
  39
  40     def _real_extract(self, url):
  41         m=re.match(self._VALID_URL, url, re.VERBOSE)
  42         if m.group('type_talk'):
  43             return self._talk_info(url)
  44         else :
  45             playlist_id=m.group('playlist_id')
  46             name=m.group('name')
  47             self.to_screen(u'Getting info of playlist %s: "%s"' % (playlist_id,name))
  48             return [self._playlist_videos_info(url,name,playlist_id)]
  49
  50
  51     def _playlist_videos_info(self, url, name, playlist_id):
  52         '''Returns the videos of the playlist'''
  53
  54         webpage = self._download_webpage(
  55             url, playlist_id, 'Downloading playlist webpage')
  56         matches = re.finditer(
  57             r'<p\s+class="talk-title[^"]*"><a\s+href="(?P<talk_url>/talks/[^"]+\.html)">[^<]*</a></p>',
  58             webpage)
  59
  60         playlist_title = self._html_search_regex(r'div class="headline">\s*?<h1>\s*?<span>(.*?)</span>',
  61                                                  webpage, 'playlist title')
  62
  63         playlist_entries = [
  64             self.url_result(u'http://www.ted.com' + m.group('talk_url'), 'TED')
  65             for m in matches
  66         ]
  67         return self.playlist_result(
  68             playlist_entries, playlist_id=playlist_id, playlist_title=playlist_title)
  69
  70     def _talk_info(self, url, video_id=0):
  71         """Return the video for the talk in the url"""
  72         m = re.match(self._VALID_URL, url)
  73         video_name = m.group('name')
  74         webpage = self._download_webpage(url, video_id, 'Downloading \"%s\" page' % video_name)
  75         self.report_extraction(video_name)
  76
  77         info_json = self._search_regex(r'"talkPage.init",({.+})\)</script>', webpage, 'info json')
  78         info = json.loads(info_json)
  79         talk_info = info['talks'][0]
  80
  81         formats = [{
  82             'ext': 'mp4',
  83             'url': format_url,
  84             'format_id': format_id,
  85             'format': format_id,
  86             'preference': self._FORMATS_PREFERENCE.get(format_id, -1),
  87         } for (format_id, format_url) in talk_info['nativeDownloads'].items()]
  88         self._sort_formats(formats)
  89
  90         video_id = talk_info['id']
  91         # subtitles
  92         video_subtitles = self.extract_subtitles(video_id, talk_info)
  93         if self._downloader.params.get('listsubtitles', False):
  94             self._list_available_subtitles(video_id, talk_info)
  95             return
  96
  97         return {
  98             'id': video_id,
  99             'title': talk_info['title'],
 100             'uploader': talk_info['speaker'],
 101             'thumbnail': talk_info['thumb'],
 102             'description': self._og_search_description(webpage),
 103             'subtitles': video_subtitles,
 104             'formats': formats,
 105         }
 106
 107     def _get_available_subtitles(self, video_id, talk_info):
 108         languages = [lang['languageCode'] for lang in talk_info.get('languages', [])]
 109         if languages:
 110             sub_lang_list = {}
 111             for l in languages:
 112                 url = 'http://www.ted.com/talks/subtitles/id/%s/lang/%s/format/srt' % (video_id, l)
 113                 sub_lang_list[l] = url
 114             return sub_lang_list
 115         else:
 116             self._downloader.report_warning(u'video doesn\'t have subtitles')
 117             return {}