yt_dlp/extractor/twitter.py

   1 import re
   2
   3 from .common import InfoExtractor
   4 from ..compat import (
   5     compat_HTTPError,
   6     compat_parse_qs,
   7     compat_urllib_parse_unquote,
   8     compat_urllib_parse_urlparse,
   9 )
  10 from ..utils import (
  11     dict_get,
  12     ExtractorError,
  13     format_field,
  14     float_or_none,
  15     int_or_none,
  16     traverse_obj,
  17     try_get,
  18     strip_or_none,
  19     unified_timestamp,
  20     update_url_query,
  21     url_or_none,
  22     xpath_text,
  23 )
  24
  25 from .periscope import (
  26     PeriscopeBaseIE,
  27     PeriscopeIE,
  28 )
  29
  30
  31 class TwitterBaseIE(InfoExtractor):
  32     _API_BASE = 'https://api.twitter.com/1.1/'
  33     _BASE_REGEX = r'https?://(?:(?:www|m(?:obile)?)\.)?twitter\.com/'
  34     _GUEST_TOKEN = None
  35
  36     def _extract_variant_formats(self, variant, video_id):
  37         variant_url = variant.get('url')
  38         if not variant_url:
  39             return [], {}
  40         elif '.m3u8' in variant_url:
  41             return self._extract_m3u8_formats_and_subtitles(
  42                 variant_url, video_id, 'mp4', 'm3u8_native',
  43                 m3u8_id='hls', fatal=False)
  44         else:
  45             tbr = int_or_none(dict_get(variant, ('bitrate', 'bit_rate')), 1000) or None
  46             f = {
  47                 'url': variant_url,
  48                 'format_id': 'http' + ('-%d' % tbr if tbr else ''),
  49                 'tbr': tbr,
  50             }
  51             self._search_dimensions_in_video_url(f, variant_url)
  52             return [f], {}
  53
  54     def _extract_formats_from_vmap_url(self, vmap_url, video_id):
  55         vmap_url = url_or_none(vmap_url)
  56         if not vmap_url:
  57             return [], {}
  58         vmap_data = self._download_xml(vmap_url, video_id)
  59         formats = []
  60         subtitles = {}
  61         urls = []
  62         for video_variant in vmap_data.findall('.//{http://twitter.com/schema/videoVMapV2.xsd}videoVariant'):
  63             video_variant.attrib['url'] = compat_urllib_parse_unquote(
  64                 video_variant.attrib['url'])
  65             urls.append(video_variant.attrib['url'])
  66             fmts, subs = self._extract_variant_formats(
  67                 video_variant.attrib, video_id)
  68             formats.extend(fmts)
  69             subtitles = self._merge_subtitles(subtitles, subs)
  70         video_url = strip_or_none(xpath_text(vmap_data, './/MediaFile'))
  71         if video_url not in urls:
  72             fmts, subs = self._extract_variant_formats({'url': video_url}, video_id)
  73             formats.extend(fmts)
  74             subtitles = self._merge_subtitles(subtitles, subs)
  75         return formats, subtitles
  76
  77     @staticmethod
  78     def _search_dimensions_in_video_url(a_format, video_url):
  79         m = re.search(r'/(?P<width>\d+)x(?P<height>\d+)/', video_url)
  80         if m:
  81             a_format.update({
  82                 'width': int(m.group('width')),
  83                 'height': int(m.group('height')),
  84             })
  85
  86     def _call_api(self, path, video_id, query={}):
  87         headers = {
  88             'Authorization': 'Bearer AAAAAAAAAAAAAAAAAAAAAPYXBAAAAAAACLXUNDekMxqa8h%2F40K4moUkGsoc%3DTYfbDKbT3jJPCEVnMYqilB28NHfOPqkca3qaAxGfsyKCs0wRbw',
  89         }
  90         token = self._get_cookies(self._API_BASE).get('ct0')
  91         if token:
  92             headers['x-csrf-token'] = token.value
  93         if not self._GUEST_TOKEN:
  94             self._GUEST_TOKEN = self._download_json(
  95                 self._API_BASE + 'guest/activate.json', video_id,
  96                 'Downloading guest token', data=b'',
  97                 headers=headers)['guest_token']
  98         headers['x-guest-token'] = self._GUEST_TOKEN
  99         try:
 100             return self._download_json(
 101                 self._API_BASE + path, video_id, headers=headers, query=query)
 102         except ExtractorError as e:
 103             if isinstance(e.cause, compat_HTTPError) and e.cause.code == 403:
 104                 raise ExtractorError(self._parse_json(
 105                     e.cause.read().decode(),
 106                     video_id)['errors'][0]['message'], expected=True)
 107             raise
 108
 109
 110 class TwitterCardIE(InfoExtractor):
 111     IE_NAME = 'twitter:card'
 112     _VALID_URL = TwitterBaseIE._BASE_REGEX + r'i/(?:cards/tfw/v1|videos(?:/tweet)?)/(?P<id>\d+)'
 113     _TESTS = [
 114         {
 115             'url': 'https://twitter.com/i/cards/tfw/v1/560070183650213889',
 116             # MD5 checksums are different in different places
 117             'info_dict': {
 118                 'id': '560070183650213889',
 119                 'ext': 'mp4',
 120                 'title': "Twitter - You can now shoot, edit and share video on Twitter. Capture life's most moving moments from your perspective.",
 121                 'description': 'md5:18d3e24bb4f6e5007487dd546e53bd96',
 122                 'uploader': 'Twitter',
 123                 'uploader_id': 'Twitter',
 124                 'thumbnail': r're:^https?://.*\.jpg',
 125                 'duration': 30.033,
 126                 'timestamp': 1422366112,
 127                 'upload_date': '20150127',
 128             },
 129         },
 130         {
 131             'url': 'https://twitter.com/i/cards/tfw/v1/623160978427936768',
 132             'md5': '7137eca597f72b9abbe61e5ae0161399',
 133             'info_dict': {
 134                 'id': '623160978427936768',
 135                 'ext': 'mp4',
 136                 'title': "NASA - Fly over Pluto's icy Norgay Mountains and Sputnik Plain in this @NASANewHorizons #PlutoFlyby video.",
 137                 'description': "Fly over Pluto's icy Norgay Mountains and Sputnik Plain in this @NASANewHorizons #PlutoFlyby video. https://t.co/BJYgOjSeGA",
 138                 'uploader': 'NASA',
 139                 'uploader_id': 'NASA',
 140                 'timestamp': 1437408129,
 141                 'upload_date': '20150720',
 142             },
 143         },
 144         {
 145             'url': 'https://twitter.com/i/cards/tfw/v1/654001591733886977',
 146             'md5': 'b6d9683dd3f48e340ded81c0e917ad46',
 147             'info_dict': {
 148                 'id': 'dq4Oj5quskI',
 149                 'ext': 'mp4',
 150                 'title': 'Ubuntu 11.10 Overview',
 151                 'description': 'md5:a831e97fa384863d6e26ce48d1c43376',
 152                 'upload_date': '20111013',
 153                 'uploader': 'OMG! UBUNTU!',
 154                 'uploader_id': 'omgubuntu',
 155             },
 156             'add_ie': ['Youtube'],
 157         },
 158         {
 159             'url': 'https://twitter.com/i/cards/tfw/v1/665289828897005568',
 160             'md5': '6dabeaca9e68cbb71c99c322a4b42a11',
 161             'info_dict': {
 162                 'id': 'iBb2x00UVlv',
 163                 'ext': 'mp4',
 164                 'upload_date': '20151113',
 165                 'uploader_id': '1189339351084113920',
 166                 'uploader': 'ArsenalTerje',
 167                 'title': 'Vine by ArsenalTerje',
 168                 'timestamp': 1447451307,
 169             },
 170             'add_ie': ['Vine'],
 171         }, {
 172             'url': 'https://twitter.com/i/videos/tweet/705235433198714880',
 173             'md5': '884812a2adc8aaf6fe52b15ccbfa3b88',
 174             'info_dict': {
 175                 'id': '705235433198714880',
 176                 'ext': 'mp4',
 177                 'title': "Brent Yarina - Khalil Iverson's missed highlight dunk. And made highlight dunk. In one highlight.",
 178                 'description': "Khalil Iverson's missed highlight dunk. And made highlight dunk. In one highlight. https://t.co/OrxcJ28Bns",
 179                 'uploader': 'Brent Yarina',
 180                 'uploader_id': 'BTNBrentYarina',
 181                 'timestamp': 1456976204,
 182                 'upload_date': '20160303',
 183             },
 184             'skip': 'This content is no longer available.',
 185         }, {
 186             'url': 'https://twitter.com/i/videos/752274308186120192',
 187             'only_matching': True,
 188         },
 189     ]
 190
 191     def _real_extract(self, url):
 192         status_id = self._match_id(url)
 193         return self.url_result(
 194             'https://twitter.com/statuses/' + status_id,
 195             TwitterIE.ie_key(), status_id)
 196
 197
 198 class TwitterIE(TwitterBaseIE):
 199     IE_NAME = 'twitter'
 200     _VALID_URL = TwitterBaseIE._BASE_REGEX + r'(?:(?:i/web|[^/]+)/status|statuses)/(?P<id>\d+)'
 201
 202     _TESTS = [{
 203         'url': 'https://twitter.com/freethenipple/status/643211948184596480',
 204         'info_dict': {
 205             'id': '643211948184596480',
 206             'ext': 'mp4',
 207             'title': 'FREE THE NIPPLE - FTN supporters on Hollywood Blvd today!',
 208             'thumbnail': r're:^https?://.*\.jpg',
 209             'description': 'FTN supporters on Hollywood Blvd today! http://t.co/c7jHH749xJ',
 210             'uploader': 'FREE THE NIPPLE',
 211             'uploader_id': 'freethenipple',
 212             'duration': 12.922,
 213             'timestamp': 1442188653,
 214             'upload_date': '20150913',
 215             'age_limit': 18,
 216         },
 217     }, {
 218         'url': 'https://twitter.com/giphz/status/657991469417025536/photo/1',
 219         'md5': 'f36dcd5fb92bf7057f155e7d927eeb42',
 220         'info_dict': {
 221             'id': '657991469417025536',
 222             'ext': 'mp4',
 223             'title': 'Gifs - tu vai cai tu vai cai tu nao eh capaz disso tu vai cai',
 224             'description': 'Gifs on Twitter: "tu vai cai tu vai cai tu nao eh capaz disso tu vai cai https://t.co/tM46VHFlO5"',
 225             'thumbnail': r're:^https?://.*\.png',
 226             'uploader': 'Gifs',
 227             'uploader_id': 'giphz',
 228         },
 229         'expected_warnings': ['height', 'width'],
 230         'skip': 'Account suspended',
 231     }, {
 232         'url': 'https://twitter.com/starwars/status/665052190608723968',
 233         'info_dict': {
 234             'id': '665052190608723968',
 235             'ext': 'mp4',
 236             'title': 'Star Wars - A new beginning is coming December 18. Watch the official 60 second #TV spot for #StarWars: #TheForceAwakens.',
 237             'description': 'A new beginning is coming December 18. Watch the official 60 second #TV spot for #StarWars: #TheForceAwakens. https://t.co/OkSqT2fjWJ',
 238             'uploader_id': 'starwars',
 239             'uploader': 'Star Wars',
 240             'timestamp': 1447395772,
 241             'upload_date': '20151113',
 242         },
 243     }, {
 244         'url': 'https://twitter.com/BTNBrentYarina/status/705235433198714880',
 245         'info_dict': {
 246             'id': '705235433198714880',
 247             'ext': 'mp4',
 248             'title': "Brent Yarina - Khalil Iverson's missed highlight dunk. And made highlight dunk. In one highlight.",
 249             'description': "Khalil Iverson's missed highlight dunk. And made highlight dunk. In one highlight. https://t.co/OrxcJ28Bns",
 250             'uploader_id': 'BTNBrentYarina',
 251             'uploader': 'Brent Yarina',
 252             'timestamp': 1456976204,
 253             'upload_date': '20160303',
 254         },
 255         'params': {
 256             # The same video as https://twitter.com/i/videos/tweet/705235433198714880
 257             # Test case of TwitterCardIE
 258             'skip_download': True,
 259         },
 260     }, {
 261         'url': 'https://twitter.com/jaydingeer/status/700207533655363584',
 262         'info_dict': {
 263             'id': '700207533655363584',
 264             'ext': 'mp4',
 265             'title': 'simon vertugo - BEAT PROD: @suhmeduh #Damndaniel',
 266             'description': 'BEAT PROD: @suhmeduh  https://t.co/HBrQ4AfpvZ #Damndaniel https://t.co/byBooq2ejZ',
 267             'thumbnail': r're:^https?://.*\.jpg',
 268             'uploader': 'simon vertugo',
 269             'uploader_id': 'simonvertugo',
 270             'duration': 30.0,
 271             'timestamp': 1455777459,
 272             'upload_date': '20160218',
 273         },
 274     }, {
 275         'url': 'https://twitter.com/Filmdrunk/status/713801302971588609',
 276         'md5': '89a15ed345d13b86e9a5a5e051fa308a',
 277         'info_dict': {
 278             'id': 'MIOxnrUteUd',
 279             'ext': 'mp4',
 280             'title': 'Dr.Pepperの飲み方 #japanese #バカ #ドクペ #電動ガン',
 281             'uploader': 'TAKUMA',
 282             'uploader_id': '1004126642786242560',
 283             'timestamp': 1402826626,
 284             'upload_date': '20140615',
 285         },
 286         'add_ie': ['Vine'],
 287     }, {
 288         'url': 'https://twitter.com/captainamerica/status/719944021058060289',
 289         'info_dict': {
 290             'id': '719944021058060289',
 291             'ext': 'mp4',
 292             'title': 'Captain America - @King0fNerd Are you sure you made the right choice? Find out in theaters.',
 293             'description': '@King0fNerd Are you sure you made the right choice? Find out in theaters. https://t.co/GpgYi9xMJI',
 294             'uploader_id': 'CaptainAmerica',
 295             'uploader': 'Captain America',
 296             'duration': 3.17,
 297             'timestamp': 1460483005,
 298             'upload_date': '20160412',
 299         },
 300     }, {
 301         'url': 'https://twitter.com/OPP_HSD/status/779210622571536384',
 302         'info_dict': {
 303             'id': '1zqKVVlkqLaKB',
 304             'ext': 'mp4',
 305             'title': 'Sgt Kerry Schmidt - Ontario Provincial Police - Road rage, mischief, assault, rollover and fire in one occurrence',
 306             'upload_date': '20160923',
 307             'uploader_id': '1PmKqpJdOJQoY',
 308             'uploader': 'Sgt Kerry Schmidt - Ontario Provincial Police',
 309             'timestamp': 1474613214,
 310         },
 311         'add_ie': ['Periscope'],
 312     }, {
 313         # has mp4 formats via mobile API
 314         'url': 'https://twitter.com/news_al3alm/status/852138619213144067',
 315         'info_dict': {
 316             'id': '852138619213144067',
 317             'ext': 'mp4',
 318             'title': 'عالم الأخبار - كلمة تاريخية بجلسة الجناسي التاريخية.. النائب خالد مؤنس العتيبي للمعارضين : اتقوا الله .. الظلم ظلمات يوم القيامة',
 319             'description': 'كلمة تاريخية بجلسة الجناسي التاريخية.. النائب خالد مؤنس العتيبي للمعارضين : اتقوا الله .. الظلم ظلمات يوم القيامة   https://t.co/xg6OhpyKfN',
 320             'uploader': 'عالم الأخبار',
 321             'uploader_id': 'news_al3alm',
 322             'duration': 277.4,
 323             'timestamp': 1492000653,
 324             'upload_date': '20170412',
 325         },
 326         'skip': 'Account suspended',
 327     }, {
 328         'url': 'https://twitter.com/i/web/status/910031516746514432',
 329         'info_dict': {
 330             'id': '910031516746514432',
 331             'ext': 'mp4',
 332             'title': 'Préfet de Guadeloupe - [Direct] #Maria Le centre se trouve actuellement au sud de Basse-Terre. Restez confinés. Réfugiez-vous dans la pièce la + sûre.',
 333             'thumbnail': r're:^https?://.*\.jpg',
 334             'description': '[Direct] #Maria Le centre se trouve actuellement au sud de Basse-Terre. Restez confinés. Réfugiez-vous dans la pièce la + sûre. https://t.co/mwx01Rs4lo',
 335             'uploader': 'Préfet de Guadeloupe',
 336             'uploader_id': 'Prefet971',
 337             'duration': 47.48,
 338             'timestamp': 1505803395,
 339             'upload_date': '20170919',
 340         },
 341         'params': {
 342             'skip_download': True,  # requires ffmpeg
 343         },
 344     }, {
 345         # card via api.twitter.com/1.1/videos/tweet/config
 346         'url': 'https://twitter.com/LisPower1/status/1001551623938805763',
 347         'info_dict': {
 348             'id': '1001551623938805763',
 349             'ext': 'mp4',
 350             'title': 're:.*?Shep is on a roll today.*?',
 351             'thumbnail': r're:^https?://.*\.jpg',
 352             'description': 'md5:37b9f2ff31720cef23b2bd42ee8a0f09',
 353             'uploader': 'Lis Power',
 354             'uploader_id': 'LisPower1',
 355             'duration': 111.278,
 356             'timestamp': 1527623489,
 357             'upload_date': '20180529',
 358         },
 359         'params': {
 360             'skip_download': True,  # requires ffmpeg
 361         },
 362     }, {
 363         'url': 'https://twitter.com/foobar/status/1087791357756956680',
 364         'info_dict': {
 365             'id': '1087791357756956680',
 366             'ext': 'mp4',
 367             'title': 'Twitter - A new is coming.  Some of you got an opt-in to try it now. Check out the emoji button, quick keyboard shortcuts, upgraded trends, advanced search, and more. Let us know your thoughts!',
 368             'thumbnail': r're:^https?://.*\.jpg',
 369             'description': 'md5:6dfd341a3310fb97d80d2bf7145df976',
 370             'uploader': 'Twitter',
 371             'uploader_id': 'Twitter',
 372             'duration': 61.567,
 373             'timestamp': 1548184644,
 374             'upload_date': '20190122',
 375         },
 376     }, {
 377         # not available in Periscope
 378         'url': 'https://twitter.com/ViviEducation/status/1136534865145286656',
 379         'info_dict': {
 380             'id': '1vOGwqejwoWxB',
 381             'ext': 'mp4',
 382             'title': 'Vivi - Vivi founder @lior_rauchy announcing our new student feedback tool live at @EduTECH_AU #EduTECH2019',
 383             'uploader': 'Vivi',
 384             'uploader_id': '1eVjYOLGkGrQL',
 385         },
 386         'add_ie': ['TwitterBroadcast'],
 387     }, {
 388         # unified card
 389         'url': 'https://twitter.com/BrooklynNets/status/1349794411333394432?s=20',
 390         'info_dict': {
 391             'id': '1349794411333394432',
 392             'ext': 'mp4',
 393             'title': 'md5:d1c4941658e4caaa6cb579260d85dcba',
 394             'thumbnail': r're:^https?://.*\.jpg',
 395             'description': 'md5:71ead15ec44cee55071547d6447c6a3e',
 396             'uploader': 'Brooklyn Nets',
 397             'uploader_id': 'BrooklynNets',
 398             'duration': 324.484,
 399             'timestamp': 1610651040,
 400             'upload_date': '20210114',
 401         },
 402         'params': {
 403             'skip_download': True,
 404         },
 405     }, {
 406         # Twitch Clip Embed
 407         'url': 'https://twitter.com/GunB1g/status/1163218564784017422',
 408         'only_matching': True,
 409     }, {
 410         # promo_video_website card
 411         'url': 'https://twitter.com/GunB1g/status/1163218564784017422',
 412         'only_matching': True,
 413     }, {
 414         # promo_video_convo card
 415         'url': 'https://twitter.com/poco_dandy/status/1047395834013384704',
 416         'only_matching': True,
 417     }, {
 418         # appplayer card
 419         'url': 'https://twitter.com/poco_dandy/status/1150646424461176832',
 420         'only_matching': True,
 421     }, {
 422         # video_direct_message card
 423         'url': 'https://twitter.com/qarev001/status/1348948114569269251',
 424         'only_matching': True,
 425     }, {
 426         # poll2choice_video card
 427         'url': 'https://twitter.com/CAF_Online/status/1349365911120195585',
 428         'only_matching': True,
 429     }, {
 430         # poll3choice_video card
 431         'url': 'https://twitter.com/SamsungMobileSA/status/1348609186725289984',
 432         'only_matching': True,
 433     }, {
 434         # poll4choice_video card
 435         'url': 'https://twitter.com/SouthamptonFC/status/1347577658079641604',
 436         'only_matching': True,
 437     }]
 438
 439     def _real_extract(self, url):
 440         twid = self._match_id(url)
 441         status = self._call_api(
 442             'statuses/show/%s.json' % twid, twid, {
 443                 'cards_platform': 'Web-12',
 444                 'include_cards': 1,
 445                 'include_reply_count': 1,
 446                 'include_user_entities': 0,
 447                 'tweet_mode': 'extended',
 448             })
 449
 450         title = description = status['full_text'].replace('\n', ' ')
 451         # strip  'https -_t.co_BJYgOjSeGA' junk from filenames
 452         title = re.sub(r'\s+(https?://[^ ]+)', '', title)
 453         user = status.get('user') or {}
 454         uploader = user.get('name')
 455         if uploader:
 456             title = '%s - %s' % (uploader, title)
 457         uploader_id = user.get('screen_name')
 458
 459         tags = []
 460         for hashtag in (try_get(status, lambda x: x['entities']['hashtags'], list) or []):
 461             hashtag_text = hashtag.get('text')
 462             if not hashtag_text:
 463                 continue
 464             tags.append(hashtag_text)
 465
 466         info = {
 467             'id': twid,
 468             'title': title,
 469             'description': description,
 470             'uploader': uploader,
 471             'timestamp': unified_timestamp(status.get('created_at')),
 472             'uploader_id': uploader_id,
 473             'uploader_url': format_field(uploader_id, None, 'https://twitter.com/%s'),
 474             'like_count': int_or_none(status.get('favorite_count')),
 475             'repost_count': int_or_none(status.get('retweet_count')),
 476             'comment_count': int_or_none(status.get('reply_count')),
 477             'age_limit': 18 if status.get('possibly_sensitive') else 0,
 478             'tags': tags,
 479         }
 480
 481         def extract_from_video_info(media):
 482             video_info = media.get('video_info') or {}
 483
 484             formats = []
 485             subtitles = {}
 486             for variant in video_info.get('variants', []):
 487                 fmts, subs = self._extract_variant_formats(variant, twid)
 488                 subtitles = self._merge_subtitles(subtitles, subs)
 489                 formats.extend(fmts)
 490             self._sort_formats(formats, ('res', 'br', 'size', 'proto'))  # The codec of http formats are unknown
 491
 492             thumbnails = []
 493             media_url = media.get('media_url_https') or media.get('media_url')
 494             if media_url:
 495                 def add_thumbnail(name, size):
 496                     thumbnails.append({
 497                         'id': name,
 498                         'url': update_url_query(media_url, {'name': name}),
 499                         'width': int_or_none(size.get('w') or size.get('width')),
 500                         'height': int_or_none(size.get('h') or size.get('height')),
 501                     })
 502                 for name, size in media.get('sizes', {}).items():
 503                     add_thumbnail(name, size)
 504                 add_thumbnail('orig', media.get('original_info') or {})
 505
 506             info.update({
 507                 'formats': formats,
 508                 'subtitles': subtitles,
 509                 'thumbnails': thumbnails,
 510                 'duration': float_or_none(video_info.get('duration_millis'), 1000),
 511             })
 512
 513         media = traverse_obj(status, ((None, 'quoted_status'), 'extended_entities', 'media', 0), get_all=False)
 514         if media and media.get('type') != 'photo':
 515             extract_from_video_info(media)
 516         else:
 517             card = status.get('card')
 518             if card:
 519                 binding_values = card['binding_values']
 520
 521                 def get_binding_value(k):
 522                     o = binding_values.get(k) or {}
 523                     return try_get(o, lambda x: x[x['type'].lower() + '_value'])
 524
 525                 card_name = card['name'].split(':')[-1]
 526                 if card_name == 'player':
 527                     info.update({
 528                         '_type': 'url',
 529                         'url': get_binding_value('player_url'),
 530                     })
 531                 elif card_name == 'periscope_broadcast':
 532                     info.update({
 533                         '_type': 'url',
 534                         'url': get_binding_value('url') or get_binding_value('player_url'),
 535                         'ie_key': PeriscopeIE.ie_key(),
 536                     })
 537                 elif card_name == 'broadcast':
 538                     info.update({
 539                         '_type': 'url',
 540                         'url': get_binding_value('broadcast_url'),
 541                         'ie_key': TwitterBroadcastIE.ie_key(),
 542                     })
 543                 elif card_name == 'summary':
 544                     info.update({
 545                         '_type': 'url',
 546                         'url': get_binding_value('card_url'),
 547                     })
 548                 elif card_name == 'unified_card':
 549                     media_entities = self._parse_json(get_binding_value('unified_card'), twid)['media_entities']
 550                     extract_from_video_info(next(iter(media_entities.values())))
 551                 # amplify, promo_video_website, promo_video_convo, appplayer,
 552                 # video_direct_message, poll2choice_video, poll3choice_video,
 553                 # poll4choice_video, ...
 554                 else:
 555                     is_amplify = card_name == 'amplify'
 556                     vmap_url = get_binding_value('amplify_url_vmap') if is_amplify else get_binding_value('player_stream_url')
 557                     content_id = get_binding_value('%s_content_id' % (card_name if is_amplify else 'player'))
 558                     formats, subtitles = self._extract_formats_from_vmap_url(vmap_url, content_id or twid)
 559                     self._sort_formats(formats)
 560
 561                     thumbnails = []
 562                     for suffix in ('_small', '', '_large', '_x_large', '_original'):
 563                         image = get_binding_value('player_image' + suffix) or {}
 564                         image_url = image.get('url')
 565                         if not image_url or '/player-placeholder' in image_url:
 566                             continue
 567                         thumbnails.append({
 568                             'id': suffix[1:] if suffix else 'medium',
 569                             'url': image_url,
 570                             'width': int_or_none(image.get('width')),
 571                             'height': int_or_none(image.get('height')),
 572                         })
 573
 574                     info.update({
 575                         'formats': formats,
 576                         'subtitles': subtitles,
 577                         'thumbnails': thumbnails,
 578                         'duration': int_or_none(get_binding_value(
 579                             'content_duration_seconds')),
 580                     })
 581             else:
 582                 expanded_url = try_get(status, lambda x: x['entities']['urls'][0]['expanded_url'])
 583                 if not expanded_url:
 584                     raise ExtractorError("There's no video in this tweet.")
 585                 info.update({
 586                     '_type': 'url',
 587                     'url': expanded_url,
 588                 })
 589         return info
 590
 591
 592 class TwitterAmplifyIE(TwitterBaseIE):
 593     IE_NAME = 'twitter:amplify'
 594     _VALID_URL = r'https?://amp\.twimg\.com/v/(?P<id>[0-9a-f\-]{36})'
 595
 596     _TEST = {
 597         'url': 'https://amp.twimg.com/v/0ba0c3c7-0af3-4c0a-bed5-7efd1ffa2951',
 598         'md5': '7df102d0b9fd7066b86f3159f8e81bf6',
 599         'info_dict': {
 600             'id': '0ba0c3c7-0af3-4c0a-bed5-7efd1ffa2951',
 601             'ext': 'mp4',
 602             'title': 'Twitter Video',
 603             'thumbnail': 're:^https?://.*',
 604         },
 605     }
 606
 607     def _real_extract(self, url):
 608         video_id = self._match_id(url)
 609         webpage = self._download_webpage(url, video_id)
 610
 611         vmap_url = self._html_search_meta(
 612             'twitter:amplify:vmap', webpage, 'vmap url')
 613         formats = self._extract_formats_from_vmap_url(vmap_url, video_id)
 614
 615         thumbnails = []
 616         thumbnail = self._html_search_meta(
 617             'twitter:image:src', webpage, 'thumbnail', fatal=False)
 618
 619         def _find_dimension(target):
 620             w = int_or_none(self._html_search_meta(
 621                 'twitter:%s:width' % target, webpage, fatal=False))
 622             h = int_or_none(self._html_search_meta(
 623                 'twitter:%s:height' % target, webpage, fatal=False))
 624             return w, h
 625
 626         if thumbnail:
 627             thumbnail_w, thumbnail_h = _find_dimension('image')
 628             thumbnails.append({
 629                 'url': thumbnail,
 630                 'width': thumbnail_w,
 631                 'height': thumbnail_h,
 632             })
 633
 634         video_w, video_h = _find_dimension('player')
 635         formats[0].update({
 636             'width': video_w,
 637             'height': video_h,
 638         })
 639
 640         return {
 641             'id': video_id,
 642             'title': 'Twitter Video',
 643             'formats': formats,
 644             'thumbnails': thumbnails,
 645         }
 646
 647
 648 class TwitterBroadcastIE(TwitterBaseIE, PeriscopeBaseIE):
 649     IE_NAME = 'twitter:broadcast'
 650     _VALID_URL = TwitterBaseIE._BASE_REGEX + r'i/broadcasts/(?P<id>[0-9a-zA-Z]{13})'
 651
 652     _TEST = {
 653         # untitled Periscope video
 654         'url': 'https://twitter.com/i/broadcasts/1yNGaQLWpejGj',
 655         'info_dict': {
 656             'id': '1yNGaQLWpejGj',
 657             'ext': 'mp4',
 658             'title': 'Andrea May Sahouri - Periscope Broadcast',
 659             'uploader': 'Andrea May Sahouri',
 660             'uploader_id': '1PXEdBZWpGwKe',
 661         },
 662     }
 663
 664     def _real_extract(self, url):
 665         broadcast_id = self._match_id(url)
 666         broadcast = self._call_api(
 667             'broadcasts/show.json', broadcast_id,
 668             {'ids': broadcast_id})['broadcasts'][broadcast_id]
 669         info = self._parse_broadcast_data(broadcast, broadcast_id)
 670         media_key = broadcast['media_key']
 671         source = self._call_api(
 672             'live_video_stream/status/' + media_key, media_key)['source']
 673         m3u8_url = source.get('noRedirectPlaybackUrl') or source['location']
 674         if '/live_video_stream/geoblocked/' in m3u8_url:
 675             self.raise_geo_restricted()
 676         m3u8_id = compat_parse_qs(compat_urllib_parse_urlparse(
 677             m3u8_url).query).get('type', [None])[0]
 678         state, width, height = self._extract_common_format_info(broadcast)
 679         info['formats'] = self._extract_pscp_m3u8_formats(
 680             m3u8_url, broadcast_id, m3u8_id, state, width, height)
 681         return info
 682
 683
 684 class TwitterShortenerIE(TwitterBaseIE):
 685     IE_NAME = 'twitter:shortener'
 686     _VALID_URL = r'https?://t.co/(?P<id>[^?]+)|tco:(?P<eid>[^?]+)'
 687     _BASE_URL = 'https://t.co/'
 688
 689     def _real_extract(self, url):
 690         mobj = self._match_valid_url(url)
 691         eid, id = mobj.group('eid', 'id')
 692         if eid:
 693             id = eid
 694             url = self._BASE_URL + id
 695         new_url = self._request_webpage(url, id, headers={'User-Agent': 'curl'}).geturl()
 696         __UNSAFE_LINK = "https://twitter.com/safety/unsafe_link_warning?unsafe_link="
 697         if new_url.startswith(__UNSAFE_LINK):
 698             new_url = new_url.replace(__UNSAFE_LINK, "")
 699         return self.url_result(new_url)