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

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import re
   5
   6 from .common import InfoExtractor
   7 from ..compat import compat_urlparse
   8 from ..utils import (
   9     determine_ext,
  10     float_or_none,
  11     xpath_text,
  12     remove_end,
  13     int_or_none,
  14     ExtractorError,
  15 )
  16
  17 from .periscope import PeriscopeIE
  18
  19
  20 class TwitterBaseIE(InfoExtractor):
  21     def _get_vmap_video_url(self, vmap_url, video_id):
  22         vmap_data = self._download_xml(vmap_url, video_id)
  23         return xpath_text(vmap_data, './/MediaFile').strip()
  24
  25
  26 class TwitterCardIE(TwitterBaseIE):
  27     IE_NAME = 'twitter:card'
  28     _VALID_URL = r'https?://(?:www\.)?twitter\.com/i/(?:cards/tfw/v1|videos(?:/tweet)?)/(?P<id>\d+)'
  29     _TESTS = [
  30         {
  31             'url': 'https://twitter.com/i/cards/tfw/v1/560070183650213889',
  32             # MD5 checksums are different in different places
  33             'info_dict': {
  34                 'id': '560070183650213889',
  35                 'ext': 'mp4',
  36                 'title': 'Twitter Card',
  37                 'thumbnail': 're:^https?://.*\.jpg$',
  38                 'duration': 30.033,
  39             }
  40         },
  41         {
  42             'url': 'https://twitter.com/i/cards/tfw/v1/623160978427936768',
  43             'md5': '7ee2a553b63d1bccba97fbed97d9e1c8',
  44             'info_dict': {
  45                 'id': '623160978427936768',
  46                 'ext': 'mp4',
  47                 'title': 'Twitter Card',
  48                 'thumbnail': 're:^https?://.*\.jpg',
  49                 'duration': 80.155,
  50             },
  51         },
  52         {
  53             'url': 'https://twitter.com/i/cards/tfw/v1/654001591733886977',
  54             'md5': 'b6d9683dd3f48e340ded81c0e917ad46',
  55             'info_dict': {
  56                 'id': 'dq4Oj5quskI',
  57                 'ext': 'mp4',
  58                 'title': 'Ubuntu 11.10 Overview',
  59                 'description': 'md5:a831e97fa384863d6e26ce48d1c43376',
  60                 'upload_date': '20111013',
  61                 'uploader': 'OMG! Ubuntu!',
  62                 'uploader_id': 'omgubuntu',
  63             },
  64             'add_ie': ['Youtube'],
  65         },
  66         {
  67             'url': 'https://twitter.com/i/cards/tfw/v1/665289828897005568',
  68             'md5': 'ab2745d0b0ce53319a534fccaa986439',
  69             'info_dict': {
  70                 'id': 'iBb2x00UVlv',
  71                 'ext': 'mp4',
  72                 'upload_date': '20151113',
  73                 'uploader_id': '1189339351084113920',
  74                 'uploader': 'ArsenalTerje',
  75                 'title': 'Vine by ArsenalTerje',
  76             },
  77             'add_ie': ['Vine'],
  78         }, {
  79             'url': 'https://twitter.com/i/videos/tweet/705235433198714880',
  80             'md5': '3846d0a07109b5ab622425449b59049d',
  81             'info_dict': {
  82                 'id': '705235433198714880',
  83                 'ext': 'mp4',
  84                 'title': 'Twitter web player',
  85                 'thumbnail': 're:^https?://.*\.jpg',
  86             },
  87         }, {
  88             'url': 'https://twitter.com/i/videos/752274308186120192',
  89             'only_matching': True,
  90         },
  91     ]
  92
  93     def _real_extract(self, url):
  94         video_id = self._match_id(url)
  95
  96         config = None
  97         formats = []
  98         duration = None
  99
 100         webpage = self._download_webpage(url, video_id)
 101
 102         iframe_url = self._html_search_regex(
 103             r'<iframe[^>]+src="((?:https?:)?//(?:www.youtube.com/embed/[^"]+|(?:www\.)?vine\.co/v/\w+/card))"',
 104             webpage, 'video iframe', default=None)
 105         if iframe_url:
 106             return self.url_result(iframe_url)
 107
 108         config = self._parse_json(self._html_search_regex(
 109             r'data-(?:player-)?config="([^"]+)"', webpage,
 110             'data player config', default='{}'),
 111             video_id)
 112
 113         if config.get('source_type') == 'vine':
 114             return self.url_result(config['player_url'], 'Vine')
 115
 116         periscope_url = PeriscopeIE._extract_url(webpage)
 117         if periscope_url:
 118             return self.url_result(periscope_url, PeriscopeIE.ie_key())
 119
 120         def _search_dimensions_in_video_url(a_format, video_url):
 121             m = re.search(r'/(?P<width>\d+)x(?P<height>\d+)/', video_url)
 122             if m:
 123                 a_format.update({
 124                     'width': int(m.group('width')),
 125                     'height': int(m.group('height')),
 126                 })
 127
 128         video_url = config.get('video_url') or config.get('playlist', [{}])[0].get('source')
 129
 130         if video_url:
 131             if determine_ext(video_url) == 'm3u8':
 132                 formats.extend(self._extract_m3u8_formats(video_url, video_id, ext='mp4', m3u8_id='hls'))
 133             else:
 134                 f = {
 135                     'url': video_url,
 136                 }
 137
 138                 _search_dimensions_in_video_url(f, video_url)
 139
 140                 formats.append(f)
 141
 142         vmap_url = config.get('vmapUrl') or config.get('vmap_url')
 143         if vmap_url:
 144             formats.append({
 145                 'url': self._get_vmap_video_url(vmap_url, video_id),
 146             })
 147
 148         media_info = None
 149
 150         for entity in config.get('status', {}).get('entities', []):
 151             if 'mediaInfo' in entity:
 152                 media_info = entity['mediaInfo']
 153
 154         if media_info:
 155             for media_variant in media_info['variants']:
 156                 media_url = media_variant['url']
 157                 if media_url.endswith('.m3u8'):
 158                     formats.extend(self._extract_m3u8_formats(media_url, video_id, ext='mp4', m3u8_id='hls'))
 159                 elif media_url.endswith('.mpd'):
 160                     formats.extend(self._extract_mpd_formats(media_url, video_id, mpd_id='dash'))
 161                 else:
 162                     vbr = int_or_none(media_variant.get('bitRate'), scale=1000)
 163                     a_format = {
 164                         'url': media_url,
 165                         'format_id': 'http-%d' % vbr if vbr else 'http',
 166                         'vbr': vbr,
 167                     }
 168                     # Reported bitRate may be zero
 169                     if not a_format['vbr']:
 170                         del a_format['vbr']
 171
 172                     _search_dimensions_in_video_url(a_format, media_url)
 173
 174                     formats.append(a_format)
 175
 176             duration = float_or_none(media_info.get('duration', {}).get('nanos'), scale=1e9)
 177
 178         self._sort_formats(formats)
 179
 180         title = self._search_regex(r'<title>([^<]+)</title>', webpage, 'title')
 181         thumbnail = config.get('posterImageUrl') or config.get('image_src')
 182         duration = float_or_none(config.get('duration')) or duration
 183
 184         return {
 185             'id': video_id,
 186             'title': title,
 187             'thumbnail': thumbnail,
 188             'duration': duration,
 189             'formats': formats,
 190         }
 191
 192
 193 class TwitterIE(InfoExtractor):
 194     IE_NAME = 'twitter'
 195     _VALID_URL = r'https?://(?:www\.|m\.|mobile\.)?twitter\.com/(?P<user_id>[^/]+)/status/(?P<id>\d+)'
 196     _TEMPLATE_URL = 'https://twitter.com/%s/status/%s'
 197
 198     _TESTS = [{
 199         'url': 'https://twitter.com/freethenipple/status/643211948184596480',
 200         'info_dict': {
 201             'id': '643211948184596480',
 202             'ext': 'mp4',
 203             'title': 'FREE THE NIPPLE - FTN supporters on Hollywood Blvd today!',
 204             'thumbnail': 're:^https?://.*\.jpg',
 205             'description': 'FREE THE NIPPLE on Twitter: "FTN supporters on Hollywood Blvd today! http://t.co/c7jHH749xJ"',
 206             'uploader': 'FREE THE NIPPLE',
 207             'uploader_id': 'freethenipple',
 208         },
 209         'params': {
 210             'skip_download': True,  # requires ffmpeg
 211         },
 212     }, {
 213         'url': 'https://twitter.com/giphz/status/657991469417025536/photo/1',
 214         'md5': 'f36dcd5fb92bf7057f155e7d927eeb42',
 215         'info_dict': {
 216             'id': '657991469417025536',
 217             'ext': 'mp4',
 218             'title': 'Gifs - tu vai cai tu vai cai tu nao eh capaz disso tu vai cai',
 219             'description': 'Gifs on Twitter: "tu vai cai tu vai cai tu nao eh capaz disso tu vai cai https://t.co/tM46VHFlO5"',
 220             'thumbnail': 're:^https?://.*\.png',
 221             'uploader': 'Gifs',
 222             'uploader_id': 'giphz',
 223         },
 224         'expected_warnings': ['height', 'width'],
 225         'skip': 'Account suspended',
 226     }, {
 227         'url': 'https://twitter.com/starwars/status/665052190608723968',
 228         'md5': '39b7199856dee6cd4432e72c74bc69d4',
 229         'info_dict': {
 230             'id': '665052190608723968',
 231             'ext': 'mp4',
 232             'title': 'Star Wars - A new beginning is coming December 18. Watch the official 60 second #TV spot for #StarWars: #TheForceAwakens.',
 233             'description': 'Star Wars on Twitter: "A new beginning is coming December 18. Watch the official 60 second #TV spot for #StarWars: #TheForceAwakens."',
 234             'uploader_id': 'starwars',
 235             'uploader': 'Star Wars',
 236         },
 237     }, {
 238         'url': 'https://twitter.com/BTNBrentYarina/status/705235433198714880',
 239         'info_dict': {
 240             'id': '705235433198714880',
 241             'ext': 'mp4',
 242             'title': 'Brent Yarina - Khalil Iverson\'s missed highlight dunk. And made highlight dunk. In one highlight.',
 243             'description': 'Brent Yarina on Twitter: "Khalil Iverson\'s missed highlight dunk. And made highlight dunk. In one highlight."',
 244             'uploader_id': 'BTNBrentYarina',
 245             'uploader': 'Brent Yarina',
 246         },
 247         'params': {
 248             # The same video as https://twitter.com/i/videos/tweet/705235433198714880
 249             # Test case of TwitterCardIE
 250             'skip_download': True,
 251         },
 252     }, {
 253         'url': 'https://twitter.com/jaydingeer/status/700207533655363584',
 254         'md5': '',
 255         'info_dict': {
 256             'id': '700207533655363584',
 257             'ext': 'mp4',
 258             'title': 'JG - BEAT PROD: @suhmeduh #Damndaniel',
 259             'description': 'JG on Twitter: "BEAT PROD: @suhmeduh  https://t.co/HBrQ4AfpvZ #Damndaniel https://t.co/byBooq2ejZ"',
 260             'thumbnail': 're:^https?://.*\.jpg',
 261             'uploader': 'JG',
 262             'uploader_id': 'jaydingeer',
 263         },
 264         'params': {
 265             'skip_download': True,  # requires ffmpeg
 266         },
 267     }, {
 268         'url': 'https://twitter.com/Filmdrunk/status/713801302971588609',
 269         'md5': '89a15ed345d13b86e9a5a5e051fa308a',
 270         'info_dict': {
 271             'id': 'MIOxnrUteUd',
 272             'ext': 'mp4',
 273             'title': 'Dr.Pepperの飲み方 #japanese #バカ #ドクペ #電動ガン',
 274             'uploader': 'TAKUMA',
 275             'uploader_id': '1004126642786242560',
 276             'upload_date': '20140615',
 277         },
 278         'add_ie': ['Vine'],
 279     }, {
 280         'url': 'https://twitter.com/captainamerica/status/719944021058060289',
 281         'info_dict': {
 282             'id': '719944021058060289',
 283             'ext': 'mp4',
 284             'title': 'Captain America - @King0fNerd Are you sure you made the right choice? Find out in theaters.',
 285             'description': 'Captain America on Twitter: "@King0fNerd Are you sure you made the right choice? Find out in theaters. https://t.co/GpgYi9xMJI"',
 286             'uploader_id': 'captainamerica',
 287             'uploader': 'Captain America',
 288         },
 289         'params': {
 290             'skip_download': True,  # requires ffmpeg
 291         },
 292     }, {
 293         'url': 'https://twitter.com/OPP_HSD/status/779210622571536384',
 294         'info_dict': {
 295             'id': '1zqKVVlkqLaKB',
 296             'ext': 'mp4',
 297             'title': 'Sgt Kerry Schmidt - Ontario Provincial Police - Road rage, mischief, assault, rollover and fire in one occurrence',
 298             'upload_date': '20160923',
 299             'uploader_id': 'OPP_HSD',
 300             'uploader': 'Sgt Kerry Schmidt - Ontario Provincial Police',
 301             'timestamp': 1474613214,
 302         },
 303         'add_ie': ['Periscope'],
 304     }]
 305
 306     def _real_extract(self, url):
 307         mobj = re.match(self._VALID_URL, url)
 308         user_id = mobj.group('user_id')
 309         twid = mobj.group('id')
 310
 311         webpage, urlh = self._download_webpage_handle(
 312             self._TEMPLATE_URL % (user_id, twid), twid)
 313
 314         if 'twitter.com/account/suspended' in urlh.geturl():
 315             raise ExtractorError('Account suspended by Twitter.', expected=True)
 316
 317         username = remove_end(self._og_search_title(webpage), ' on Twitter')
 318
 319         title = description = self._og_search_description(webpage).strip('').replace('\n', ' ').strip('“”')
 320
 321         # strip  'https -_t.co_BJYgOjSeGA' junk from filenames
 322         title = re.sub(r'\s+(https?://[^ ]+)', '', title)
 323
 324         info = {
 325             'uploader_id': user_id,
 326             'uploader': username,
 327             'webpage_url': url,
 328             'description': '%s on Twitter: "%s"' % (username, description),
 329             'title': username + ' - ' + title,
 330         }
 331
 332         mobj = re.search(r'''(?x)
 333             <video[^>]+class="animated-gif"(?P<more_info>[^>]+)>\s*
 334                 <source[^>]+video-src="(?P<url>[^"]+)"
 335         ''', webpage)
 336
 337         if mobj:
 338             more_info = mobj.group('more_info')
 339             height = int_or_none(self._search_regex(
 340                 r'data-height="(\d+)"', more_info, 'height', fatal=False))
 341             width = int_or_none(self._search_regex(
 342                 r'data-width="(\d+)"', more_info, 'width', fatal=False))
 343             thumbnail = self._search_regex(
 344                 r'poster="([^"]+)"', more_info, 'poster', fatal=False)
 345             info.update({
 346                 'id': twid,
 347                 'url': mobj.group('url'),
 348                 'height': height,
 349                 'width': width,
 350                 'thumbnail': thumbnail,
 351             })
 352             return info
 353
 354         twitter_card_url = None
 355         if 'class="PlayableMedia' in webpage:
 356             twitter_card_url = '%s//twitter.com/i/videos/tweet/%s' % (self.http_scheme(), twid)
 357         else:
 358             twitter_card_iframe_url = self._search_regex(
 359                 r'data-full-card-iframe-url=([\'"])(?P<url>(?:(?!\1).)+)\1',
 360                 webpage, 'Twitter card iframe URL', default=None, group='url')
 361             if twitter_card_iframe_url:
 362                 twitter_card_url = compat_urlparse.urljoin(url, twitter_card_iframe_url)
 363
 364         if twitter_card_url:
 365             info.update({
 366                 '_type': 'url_transparent',
 367                 'ie_key': 'TwitterCard',
 368                 'url': twitter_card_url,
 369             })
 370             return info
 371
 372         raise ExtractorError('There\'s no video in this tweet.')
 373
 374
 375 class TwitterAmplifyIE(TwitterBaseIE):
 376     IE_NAME = 'twitter:amplify'
 377     _VALID_URL = r'https?://amp\.twimg\.com/v/(?P<id>[0-9a-f\-]{36})'
 378
 379     _TEST = {
 380         'url': 'https://amp.twimg.com/v/0ba0c3c7-0af3-4c0a-bed5-7efd1ffa2951',
 381         'md5': '7df102d0b9fd7066b86f3159f8e81bf6',
 382         'info_dict': {
 383             'id': '0ba0c3c7-0af3-4c0a-bed5-7efd1ffa2951',
 384             'ext': 'mp4',
 385             'title': 'Twitter Video',
 386             'thumbnail': 're:^https?://.*',
 387         },
 388     }
 389
 390     def _real_extract(self, url):
 391         video_id = self._match_id(url)
 392         webpage = self._download_webpage(url, video_id)
 393
 394         vmap_url = self._html_search_meta(
 395             'twitter:amplify:vmap', webpage, 'vmap url')
 396         video_url = self._get_vmap_video_url(vmap_url, video_id)
 397
 398         thumbnails = []
 399         thumbnail = self._html_search_meta(
 400             'twitter:image:src', webpage, 'thumbnail', fatal=False)
 401
 402         def _find_dimension(target):
 403             w = int_or_none(self._html_search_meta(
 404                 'twitter:%s:width' % target, webpage, fatal=False))
 405             h = int_or_none(self._html_search_meta(
 406                 'twitter:%s:height' % target, webpage, fatal=False))
 407             return w, h
 408
 409         if thumbnail:
 410             thumbnail_w, thumbnail_h = _find_dimension('image')
 411             thumbnails.append({
 412                 'url': thumbnail,
 413                 'width': thumbnail_w,
 414                 'height': thumbnail_h,
 415             })
 416
 417         video_w, video_h = _find_dimension('player')
 418         formats = [{
 419             'url': video_url,
 420             'width': video_w,
 421             'height': video_h,
 422         }]
 423
 424         return {
 425             'id': video_id,
 426             'title': 'Twitter Video',
 427             'formats': formats,
 428             'thumbnails': thumbnails,
 429         }