Raphaël G. Git Repositories - youtubedl/blob - youtube_dl/extractor/ted.py

   1 from __future__ import unicode_literals
   2
   3 import json
   4 import re
   5
   6 from .common import InfoExtractor
   7
   8 from ..compat import compat_str
   9 from ..utils import (
  10     int_or_none,
  11     try_get,
  12 )
  13
  14
  15 class TEDIE(InfoExtractor):
  16     IE_NAME = 'ted'
  17     _VALID_URL = r'''(?x)
  18         (?P<proto>https?://)
  19         (?P<type>www|embed(?:-ssl)?)(?P<urlmain>\.ted\.com/
  20         (
  21             (?P<type_playlist>playlists(?:/\d+)?) # We have a playlist
  22             |
  23             ((?P<type_talk>talks)) # We have a simple talk
  24             |
  25             (?P<type_watch>watch)/[^/]+/[^/]+
  26         )
  27         (/lang/(.*?))? # The url may contain the language
  28         /(?P<name>[\w-]+) # Here goes the name and then ".html"
  29         .*)$
  30         '''
  31     _TESTS = [{
  32         'url': 'http://www.ted.com/talks/dan_dennett_on_our_consciousness.html',
  33         'md5': '0de43ac406aa3e4ea74b66c9c7789b13',
  34         'info_dict': {
  35             'id': '102',
  36             'ext': 'mp4',
  37             'title': 'The illusion of consciousness',
  38             'description': ('Philosopher Dan Dennett makes a compelling '
  39                             'argument that not only don\'t we understand our own '
  40                             'consciousness, but that half the time our brains are '
  41                             'actively fooling us.'),
  42             'uploader': 'Dan Dennett',
  43             'width': 853,
  44             'duration': 1308,
  45         }
  46     }, {
  47         'url': 'http://www.ted.com/watch/ted-institute/ted-bcg/vishal-sikka-the-beauty-and-power-of-algorithms',
  48         'md5': 'b899ac15e345fb39534d913f7606082b',
  49         'info_dict': {
  50             'id': 'tSVI8ta_P4w',
  51             'ext': 'mp4',
  52             'title': 'Vishal Sikka: The beauty and power of algorithms',
  53             'thumbnail': r're:^https?://.+\.jpg',
  54             'description': 'md5:6261fdfe3e02f4f579cbbfc00aff73f4',
  55             'upload_date': '20140122',
  56             'uploader_id': 'TEDInstitute',
  57             'uploader': 'TED Institute',
  58         },
  59         'add_ie': ['Youtube'],
  60     }, {
  61         'url': 'http://www.ted.com/talks/gabby_giffords_and_mark_kelly_be_passionate_be_courageous_be_your_best',
  62         'md5': '71b3ab2f4233012dce09d515c9c39ce2',
  63         'info_dict': {
  64             'id': '1972',
  65             'ext': 'mp4',
  66             'title': 'Be passionate. Be courageous. Be your best.',
  67             'uploader': 'Gabby Giffords and Mark Kelly',
  68             'description': 'md5:5174aed4d0f16021b704120360f72b92',
  69             'duration': 1128,
  70         },
  71     }, {
  72         'url': 'http://www.ted.com/playlists/who_are_the_hackers',
  73         'info_dict': {
  74             'id': '10',
  75             'title': 'Who are the hackers?',
  76         },
  77         'playlist_mincount': 6,
  78     }, {
  79         # contains a youtube video
  80         'url': 'https://www.ted.com/talks/douglas_adams_parrots_the_universe_and_everything',
  81         'add_ie': ['Youtube'],
  82         'info_dict': {
  83             'id': '_ZG8HBuDjgc',
  84             'ext': 'webm',
  85             'title': 'Douglas Adams: Parrots the Universe and Everything',
  86             'description': 'md5:01ad1e199c49ac640cb1196c0e9016af',
  87             'uploader': 'University of California Television (UCTV)',
  88             'uploader_id': 'UCtelevision',
  89             'upload_date': '20080522',
  90         },
  91         'params': {
  92             'skip_download': True,
  93         },
  94     }, {
  95         # YouTube video
  96         'url': 'http://www.ted.com/talks/jeffrey_kluger_the_sibling_bond',
  97         'add_ie': ['Youtube'],
  98         'info_dict': {
  99             'id': 'aFBIPO-P7LM',
 100             'ext': 'mp4',
 101             'title': 'The hidden power of siblings: Jeff Kluger at TEDxAsheville',
 102             'description': 'md5:3d7a4f50d95ca5dd67104e2a20f43fe1',
 103             'uploader': 'TEDx Talks',
 104             'uploader_id': 'TEDxTalks',
 105             'upload_date': '20111216',
 106         },
 107         'params': {
 108             'skip_download': True,
 109         },
 110     }]
 111
 112     _NATIVE_FORMATS = {
 113         'low': {'width': 320, 'height': 180},
 114         'medium': {'width': 512, 'height': 288},
 115         'high': {'width': 854, 'height': 480},
 116     }
 117
 118     def _extract_info(self, webpage):
 119         info_json = self._search_regex(
 120             r'(?s)q\(\s*"\w+.init"\s*,\s*({.+})\)\s*</script>',
 121             webpage, 'info json')
 122         return json.loads(info_json)
 123
 124     def _real_extract(self, url):
 125         m = re.match(self._VALID_URL, url, re.VERBOSE)
 126         if m.group('type').startswith('embed'):
 127             desktop_url = m.group('proto') + 'www' + m.group('urlmain')
 128             return self.url_result(desktop_url, 'TED')
 129         name = m.group('name')
 130         if m.group('type_talk'):
 131             return self._talk_info(url, name)
 132         elif m.group('type_watch'):
 133             return self._watch_info(url, name)
 134         else:
 135             return self._playlist_videos_info(url, name)
 136
 137     def _playlist_videos_info(self, url, name):
 138         '''Returns the videos of the playlist'''
 139
 140         webpage = self._download_webpage(url, name,
 141                                          'Downloading playlist webpage')
 142         info = self._extract_info(webpage)
 143
 144         playlist_info = try_get(
 145             info, lambda x: x['__INITIAL_DATA__']['playlist'],
 146             dict) or info['playlist']
 147
 148         playlist_entries = [
 149             self.url_result('http://www.ted.com/talks/' + talk['slug'], self.ie_key())
 150             for talk in try_get(
 151                 info, lambda x: x['__INITIAL_DATA__']['talks'],
 152                 dict) or info['talks']
 153         ]
 154         return self.playlist_result(
 155             playlist_entries,
 156             playlist_id=compat_str(playlist_info['id']),
 157             playlist_title=playlist_info['title'])
 158
 159     def _talk_info(self, url, video_name):
 160         webpage = self._download_webpage(url, video_name)
 161
 162         info = self._extract_info(webpage)
 163
 164         talk_info = try_get(
 165             info, lambda x: x['__INITIAL_DATA__']['talks'][0],
 166             dict) or info['talks'][0]
 167
 168         title = talk_info['title'].strip()
 169
 170         external = talk_info.get('external')
 171         if external:
 172             service = external['service']
 173             self.to_screen('Found video from %s' % service)
 174             ext_url = None
 175             if service.lower() == 'youtube':
 176                 ext_url = external.get('code')
 177             return {
 178                 '_type': 'url',
 179                 'url': ext_url or external['uri'],
 180             }
 181
 182         native_downloads = try_get(
 183             talk_info, lambda x: x['downloads']['nativeDownloads'],
 184             dict) or talk_info['nativeDownloads']
 185
 186         formats = [{
 187             'url': format_url,
 188             'format_id': format_id,
 189             'format': format_id,
 190         } for (format_id, format_url) in native_downloads.items() if format_url is not None]
 191         if formats:
 192             for f in formats:
 193                 finfo = self._NATIVE_FORMATS.get(f['format_id'])
 194                 if finfo:
 195                     f.update(finfo)
 196
 197         player_talk = talk_info['player_talks'][0]
 198
 199         resources_ = player_talk.get('resources') or talk_info.get('resources')
 200
 201         http_url = None
 202         for format_id, resources in resources_.items():
 203             if format_id == 'h264':
 204                 for resource in resources:
 205                     h264_url = resource.get('file')
 206                     if not h264_url:
 207                         continue
 208                     bitrate = int_or_none(resource.get('bitrate'))
 209                     formats.append({
 210                         'url': h264_url,
 211                         'format_id': '%s-%sk' % (format_id, bitrate),
 212                         'tbr': bitrate,
 213                     })
 214                     if re.search(r'\d+k', h264_url):
 215                         http_url = h264_url
 216             elif format_id == 'rtmp':
 217                 streamer = talk_info.get('streamer')
 218                 if not streamer:
 219                     continue
 220                 for resource in resources:
 221                     formats.append({
 222                         'format_id': '%s-%s' % (format_id, resource.get('name')),
 223                         'url': streamer,
 224                         'play_path': resource['file'],
 225                         'ext': 'flv',
 226                         'width': int_or_none(resource.get('width')),
 227                         'height': int_or_none(resource.get('height')),
 228                         'tbr': int_or_none(resource.get('bitrate')),
 229                     })
 230             elif format_id == 'hls':
 231                 formats.extend(self._extract_m3u8_formats(
 232                     resources.get('stream'), video_name, 'mp4', m3u8_id=format_id, fatal=False))
 233
 234         m3u8_formats = list(filter(
 235             lambda f: f.get('protocol') == 'm3u8' and f.get('vcodec') != 'none',
 236             formats))
 237         if http_url:
 238             for m3u8_format in m3u8_formats:
 239                 bitrate = self._search_regex(r'(\d+k)', m3u8_format['url'], 'bitrate', default=None)
 240                 if not bitrate:
 241                     continue
 242                 f = m3u8_format.copy()
 243                 f.update({
 244                     'url': re.sub(r'\d+k', bitrate, http_url),
 245                     'format_id': m3u8_format['format_id'].replace('hls', 'http'),
 246                     'protocol': 'http',
 247                 })
 248                 formats.append(f)
 249
 250         audio_download = talk_info.get('audioDownload')
 251         if audio_download:
 252             formats.append({
 253                 'url': audio_download,
 254                 'format_id': 'audio',
 255                 'vcodec': 'none',
 256             })
 257
 258         self._sort_formats(formats)
 259
 260         video_id = compat_str(talk_info['id'])
 261
 262         return {
 263             'id': video_id,
 264             'title': title,
 265             'uploader': player_talk.get('speaker') or talk_info.get('speaker'),
 266             'thumbnail': player_talk.get('thumb') or talk_info.get('thumb'),
 267             'description': self._og_search_description(webpage),
 268             'subtitles': self._get_subtitles(video_id, talk_info),
 269             'formats': formats,
 270             'duration': talk_info.get('duration'),
 271         }
 272
 273     def _get_subtitles(self, video_id, talk_info):
 274         sub_lang_list = {}
 275         for language in try_get(
 276                 talk_info,
 277                 (lambda x: x['downloads']['languages'],
 278                  lambda x: x['languages']), list):
 279             lang_code = language.get('languageCode') or language.get('ianaCode')
 280             if not lang_code:
 281                 continue
 282             sub_lang_list[lang_code] = [
 283                 {
 284                     'url': 'http://www.ted.com/talks/subtitles/id/%s/lang/%s/format/%s' % (video_id, lang_code, ext),
 285                     'ext': ext,
 286                 }
 287                 for ext in ['ted', 'srt']
 288             ]
 289         return sub_lang_list
 290
 291     def _watch_info(self, url, name):
 292         webpage = self._download_webpage(url, name)
 293
 294         config_json = self._html_search_regex(
 295             r'"pages\.jwplayer"\s*,\s*({.+?})\s*\)\s*</script>',
 296             webpage, 'config', default=None)
 297         if not config_json:
 298             embed_url = self._search_regex(
 299                 r"<iframe[^>]+class='pages-video-embed__video__object'[^>]+src='([^']+)'", webpage, 'embed url')
 300             return self.url_result(self._proto_relative_url(embed_url))
 301         config = json.loads(config_json)['config']
 302         video_url = config['video']['url']
 303         thumbnail = config.get('image', {}).get('url')
 304
 305         title = self._html_search_regex(
 306             r"(?s)<h1(?:\s+class='[^']+')?>(.+?)</h1>", webpage, 'title')
 307         description = self._html_search_regex(
 308             [
 309                 r'(?s)<h4 class="[^"]+" id="h3--about-this-talk">.*?</h4>(.*?)</div>',
 310                 r'(?s)<p><strong>About this talk:</strong>\s+(.*?)</p>',
 311             ],
 312             webpage, 'description', fatal=False)
 313
 314         return {
 315             'id': name,
 316             'url': video_url,
 317             'title': title,
 318             'thumbnail': thumbnail,
 319             'description': description,
 320         }