1 from __future__ import unicode_literals
6 from .subtitles import SubtitlesInfoExtractor
13 class TEDIE(SubtitlesInfoExtractor):
16 (?P<type>www|embed)(?P<urlmain>\.ted\.com/
18 (?P<type_playlist>playlists(?:/\d+)?) # We have a playlist
20 ((?P<type_talk>talks)) # We have a simple talk
22 (?P<type_watch>watch)/[^/]+/[^/]+
24 (/lang/(.*?))? # The url may contain the language
25 /(?P<name>[\w-]+) # Here goes the name and then ".html"
29 'url': 'http://www.ted.com/talks/dan_dennett_on_our_consciousness.html',
30 'md5': '4ea1dada91e4174b53dac2bb8ace429d',
34 'title': 'The illusion of consciousness',
35 'description': ('Philosopher Dan Dennett makes a compelling '
36 'argument that not only don\'t we understand our own '
37 'consciousness, but that half the time our brains are '
38 'actively fooling us.'),
39 'uploader': 'Dan Dennett',
42 'url': 'http://www.ted.com/watch/ted-institute/ted-bcg/vishal-sikka-the-beauty-and-power-of-algorithms',
43 'md5': '226f4fb9c62380d11b7995efa4c87994',
45 'id': 'vishal-sikka-the-beauty-and-power-of-algorithms',
47 'title': 'Vishal Sikka: The beauty and power of algorithms',
48 'thumbnail': 're:^https?://.+\.jpg',
49 'description': 'Adaptive, intelligent, and consistent, algorithms are emerging as the ultimate app for everything from matching consumers to products to assessing medical diagnoses. Vishal Sikka shares his appreciation for the algorithm, charting both its inherent beauty and its growing power.',
53 _FORMATS_PREFERENCE = {
59 def _extract_info(self, webpage):
60 info_json = self._search_regex(r'q\("\w+.init",({.+})\)</script>',
62 return json.loads(info_json)
64 def _real_extract(self, url):
65 m = re.match(self._VALID_URL, url, re.VERBOSE)
66 if m.group('type') == 'embed':
67 desktop_url = m.group('proto') + 'www' + m.group('urlmain')
68 return self.url_result(desktop_url, 'TED')
69 name = m.group('name')
70 if m.group('type_talk'):
71 return self._talk_info(url, name)
72 elif m.group('type_watch'):
73 return self._watch_info(url, name)
75 return self._playlist_videos_info(url, name)
77 def _playlist_videos_info(self, url, name):
78 '''Returns the videos of the playlist'''
80 webpage = self._download_webpage(url, name,
81 'Downloading playlist webpage')
82 info = self._extract_info(webpage)
83 playlist_info = info['playlist']
86 self.url_result(u'http://www.ted.com/talks/' + talk['slug'], self.ie_key())
87 for talk in info['talks']
89 return self.playlist_result(
91 playlist_id=compat_str(playlist_info['id']),
92 playlist_title=playlist_info['title'])
94 def _talk_info(self, url, video_name):
95 webpage = self._download_webpage(url, video_name)
96 self.report_extraction(video_name)
98 talk_info = self._extract_info(webpage)['talks'][0]
103 'format_id': format_id,
105 'preference': self._FORMATS_PREFERENCE.get(format_id, -1),
106 } for (format_id, format_url) in talk_info['nativeDownloads'].items()]
107 self._sort_formats(formats)
109 video_id = compat_str(talk_info['id'])
111 video_subtitles = self.extract_subtitles(video_id, talk_info)
112 if self._downloader.params.get('listsubtitles', False):
113 self._list_available_subtitles(video_id, talk_info)
116 thumbnail = talk_info['thumb']
117 if not thumbnail.startswith('http'):
118 thumbnail = 'http://' + thumbnail
121 'title': talk_info['title'],
122 'uploader': talk_info['speaker'],
123 'thumbnail': thumbnail,
124 'description': self._og_search_description(webpage),
125 'subtitles': video_subtitles,
129 def _get_available_subtitles(self, video_id, talk_info):
130 languages = [lang['languageCode'] for lang in talk_info.get('languages', [])]
134 url = 'http://www.ted.com/talks/subtitles/id/%s/lang/%s/format/srt' % (video_id, l)
135 sub_lang_list[l] = url
138 self._downloader.report_warning(u'video doesn\'t have subtitles')
141 def _watch_info(self, url, name):
142 webpage = self._download_webpage(url, name)
144 config_json = self._html_search_regex(
145 r"data-config='([^']+)", webpage, 'config')
146 config = json.loads(config_json)
147 video_url = config['video']['url']
148 thumbnail = config.get('image', {}).get('url')
150 title = self._html_search_regex(
151 r"(?s)<h1(?:\s+class='[^']+')?>(.+?)</h1>", webpage, 'title')
152 description = self._html_search_regex(
153 r'(?s)<h4 class="[^"]+" id="h3--about-this-talk">.*?</h4>(.*?)</div>',
154 webpage, 'description', fatal=False)
160 'thumbnail': thumbnail,
161 'description': description,