yt_dlp/extractor/tumblr.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4
   5 from .common import InfoExtractor
   6 from ..utils import (
   7     ExtractorError,
   8     int_or_none,
   9     traverse_obj,
  10     urlencode_postdata
  11 )
  12
  13
  14 class TumblrIE(InfoExtractor):
  15     _VALID_URL = r'https?://(?P<blog_name>[^/?#&]+)\.tumblr\.com/(?:post|video)/(?P<id>[0-9]+)(?:$|[/?#])'
  16     _NETRC_MACHINE = 'tumblr'
  17     _LOGIN_URL = 'https://www.tumblr.com/login'
  18     _OAUTH_URL = 'https://www.tumblr.com/api/v2/oauth2/token'
  19     _TESTS = [{
  20         'url': 'http://tatianamaslanydaily.tumblr.com/post/54196191430/orphan-black-dvd-extra-behind-the-scenes',
  21         'md5': '479bb068e5b16462f5176a6828829767',
  22         'info_dict': {
  23             'id': '54196191430',
  24             'ext': 'mp4',
  25             'title': 'md5:dfac39636969fe6bf1caa2d50405f069',
  26             'description': 'md5:390ab77358960235b6937ab3b8528956',
  27             'uploader_id': 'tatianamaslanydaily',
  28             'uploader_url': 'https://tatianamaslanydaily.tumblr.com/',
  29             'thumbnail': r're:^https?://.*\.jpg',
  30             'duration': 127,
  31             'like_count': int,
  32             'repost_count': int,
  33             'age_limit': 0,
  34             'tags': ['Orphan Black', 'Tatiana Maslany', 'Interview', 'Video', 'OB S1 DVD Extras'],
  35         }
  36     }, {
  37         'note': 'multiple formats',
  38         'url': 'https://maskofthedragon.tumblr.com/post/626907179849564160/mona-talking-in-english',
  39         'md5': 'f43ff8a8861712b6cf0e0c2bd84cfc68',
  40         'info_dict': {
  41             'id': '626907179849564160',
  42             'ext': 'mp4',
  43             'title': 'Mona\xa0“talking” in\xa0“english”',
  44             'description': 'md5:082a3a621530cb786ad2b7592a6d9e2c',
  45             'uploader_id': 'maskofthedragon',
  46             'uploader_url': 'https://maskofthedragon.tumblr.com/',
  47             'thumbnail': r're:^https?://.*\.jpg',
  48             'duration': 7,
  49             'like_count': int,
  50             'repost_count': int,
  51             'age_limit': 0,
  52             'tags': 'count:19',
  53         },
  54         'params': {
  55             'format': 'hd',
  56         },
  57     }, {
  58         'note': 'non-iframe video (with related posts)',
  59         'url': 'https://shieldfoss.tumblr.com/post/675519763813908480',
  60         'md5': '12bdb75661ef443bffe5a4dac1dbf118',
  61         'info_dict': {
  62             'id': '675519763813908480',
  63             'ext': 'mp4',
  64             'title': 'Shieldfoss',
  65             'uploader_id': 'nerviovago',
  66             'uploader_url': 'https://nerviovago.tumblr.com/',
  67             'thumbnail': r're:^https?://.*\.jpg',
  68             'like_count': int,
  69             'repost_count': int,
  70             'age_limit': 0,
  71             'tags': [],
  72         }
  73     }, {
  74         'note': 'dashboard only (original post)',
  75         'url': 'https://jujanon.tumblr.com/post/159704441298/my-baby-eating',
  76         'md5': '029f7c91ab386701b211e3d494d2d95e',
  77         'info_dict': {
  78             'id': '159704441298',
  79             'ext': 'mp4',
  80             'title': 'md5:ba79365861101f4911452728d2950561',
  81             'description': 'md5:773738196cea76b6996ec71e285bdabc',
  82             'uploader_id': 'jujanon',
  83             'uploader_url': 'https://jujanon.tumblr.com/',
  84             'thumbnail': r're:^https?://.*\.jpg',
  85             'like_count': int,
  86             'repost_count': int,
  87             'age_limit': 0,
  88             'tags': ['crabs', 'my video', 'my pets'],
  89         }
  90     }, {
  91         'note': 'dashboard only (reblog)',
  92         'url': 'https://bartlebyshop.tumblr.com/post/180294460076/duality-of-bird',
  93         'md5': '04334e7cadb1af680d162912559f51a5',
  94         'info_dict': {
  95             'id': '180294460076',
  96             'ext': 'mp4',
  97             'title': 'duality of bird',
  98             'description': 'duality of bird',
  99             'uploader_id': 'todaysbird',
 100             'uploader_url': 'https://todaysbird.tumblr.com/',
 101             'thumbnail': r're:^https?://.*\.jpg',
 102             'like_count': int,
 103             'repost_count': int,
 104             'age_limit': 0,
 105             'tags': [],
 106         }
 107     }, {
 108         'note': 'dashboard only (external)',
 109         'url': 'https://afloweroutofstone.tumblr.com/post/675661759168823296/the-blues-remembers-everything-the-country-forgot',
 110         'info_dict': {
 111             'id': 'q67_fd7b8SU',
 112             'ext': 'mp4',
 113             'title': 'The Blues Remembers Everything the Country Forgot',
 114             'alt_title': 'The Blues Remembers Everything the Country Forgot',
 115             'description': 'md5:1a6b4097e451216835a24c1023707c79',
 116             'release_date': '20201224',
 117             'creator': 'md5:c2239ba15430e87c3b971ba450773272',
 118             'uploader': 'Moor Mother - Topic',
 119             'upload_date': '20201223',
 120             'uploader_id': 'UCxrMtFBRkFvQJ_vVM4il08w',
 121             'uploader_url': 'http://www.youtube.com/channel/UCxrMtFBRkFvQJ_vVM4il08w',
 122             'thumbnail': r're:^https?://i.ytimg.com/.*',
 123             'channel': 'Moor Mother - Topic',
 124             'channel_id': 'UCxrMtFBRkFvQJ_vVM4il08w',
 125             'channel_url': 'https://www.youtube.com/channel/UCxrMtFBRkFvQJ_vVM4il08w',
 126             'channel_follower_count': int,
 127             'duration': 181,
 128             'view_count': int,
 129             'like_count': int,
 130             'age_limit': 0,
 131             'categories': ['Music'],
 132             'tags': 'count:7',
 133             'live_status': 'not_live',
 134             'playable_in_embed': True,
 135             'availability': 'public',
 136             'track': 'The Blues Remembers Everything the Country Forgot',
 137             'artist': 'md5:c2239ba15430e87c3b971ba450773272',
 138             'album': 'Brass',
 139             'release_year': 2020,
 140         },
 141         'add_ie': ['Youtube'],
 142     }, {
 143         'url': 'http://naked-yogi.tumblr.com/post/118312946248/naked-smoking-stretching',
 144         'md5': 'de07e5211d60d4f3a2c3df757ea9f6ab',
 145         'info_dict': {
 146             'id': 'Wmur',
 147             'ext': 'mp4',
 148             'title': 'naked smoking & stretching',
 149             'upload_date': '20150506',
 150             'timestamp': 1430931613,
 151             'age_limit': 18,
 152             'uploader_id': '1638622',
 153             'uploader': 'naked-yogi',
 154         },
 155         # 'add_ie': ['Vidme'],
 156         'skip': 'dead embedded video host'
 157     }, {
 158         'url': 'https://prozdvoices.tumblr.com/post/673201091169681408/what-recording-voice-acting-sounds-like',
 159         'md5': 'a0063fc8110e6c9afe44065b4ea68177',
 160         'info_dict': {
 161             'id': 'eomhW5MLGWA',
 162             'ext': 'mp4',
 163             'title': 'what recording voice acting sounds like',
 164             'description': 'md5:1da3faa22d0e0b1d8b50216c284ee798',
 165             'uploader': 'ProZD',
 166             'upload_date': '20220112',
 167             'uploader_id': 'ProZD',
 168             'uploader_url': 'http://www.youtube.com/user/ProZD',
 169             'thumbnail': r're:^https?://i.ytimg.com/.*',
 170             'channel': 'ProZD',
 171             'channel_id': 'UC6MFZAOHXlKK1FI7V0XQVeA',
 172             'channel_url': 'https://www.youtube.com/channel/UC6MFZAOHXlKK1FI7V0XQVeA',
 173             'channel_follower_count': int,
 174             'duration': 20,
 175             'view_count': int,
 176             'like_count': int,
 177             'age_limit': 0,
 178             'categories': ['Film & Animation'],
 179             'tags': [],
 180             'live_status': 'not_live',
 181             'playable_in_embed': True,
 182             'availability': 'public',
 183         },
 184         'add_ie': ['Youtube'],
 185     }, {
 186         'url': 'https://dominustempori.tumblr.com/post/673572712813297664/youtubes-all-right-for-some-pretty-cool',
 187         'md5': '203e9eb8077e3f45bfaeb4c86c1467b8',
 188         'info_dict': {
 189             'id': '87816359',
 190             'ext': 'mov',
 191             'title': 'Harold Ramis',
 192             'description': 'md5:be8e68cbf56ce0785c77f0c6c6dfaf2c',
 193             'uploader': 'Resolution Productions Group',
 194             'uploader_id': 'resolutionproductions',
 195             'uploader_url': 'https://vimeo.com/resolutionproductions',
 196             'upload_date': '20140227',
 197             'thumbnail': r're:^https?://i.vimeocdn.com/video/.*',
 198             'timestamp': 1393523719,
 199             'duration': 291,
 200         },
 201         'add_ie': ['Vimeo'],
 202     }, {
 203         'url': 'http://sutiblr.tumblr.com/post/139638707273',
 204         'md5': '2dd184b3669e049ba40563a7d423f95c',
 205         'info_dict': {
 206             'id': 'ir7qBEIKqvq',
 207             'ext': 'mp4',
 208             'title': 'Vine by sutiblr',
 209             'alt_title': 'Vine by sutiblr',
 210             'uploader': 'sutiblr',
 211             'uploader_id': '1198993975374495744',
 212             'upload_date': '20160220',
 213             'like_count': int,
 214             'comment_count': int,
 215             'repost_count': int,
 216             'thumbnail': r're:^https?://.*\.jpg',
 217             'timestamp': 1455940159,
 218             'view_count': int,
 219         },
 220         'add_ie': ['Vine'],
 221     }, {
 222         'url': 'https://silami.tumblr.com/post/84250043974/my-bad-river-flows-in-you-impression-on-maschine',
 223         'md5': '3c92d7c3d867f14ccbeefa2119022277',
 224         'info_dict': {
 225             'id': 'nYtvtTPuTl',
 226             'ext': 'mp4',
 227             'title': 'Video by silbulterman',
 228             'description': '#maschine',
 229             'uploader_id': '242859024',
 230             'thumbnail': r're:^https?://.*\.jpg',
 231             'timestamp': 1398801174,
 232             'like_count': int,
 233             'uploader': 'Sil',
 234             'channel': 'silbulterman',
 235             'comment_count': int,
 236             'upload_date': '20140429',
 237         },
 238         'add_ie': ['Instagram'],
 239     }]
 240
 241     _providers = {
 242         'instagram': 'Instagram',
 243         'vimeo': 'Vimeo',
 244         'vine': 'Vine',
 245         'youtube': 'Youtube',
 246     }
 247
 248     _ACCESS_TOKEN = None
 249
 250     def _initialize_pre_login(self):
 251         login_page = self._download_webpage(
 252             self._LOGIN_URL, None, 'Downloading login page', fatal=False)
 253         if login_page:
 254             self._ACCESS_TOKEN = self._search_regex(
 255                 r'"API_TOKEN":\s*"(\w+)"', login_page, 'API access token', fatal=False)
 256         if not self._ACCESS_TOKEN:
 257             self.report_warning('Failed to get access token; metadata will be missing and some videos may not work')
 258
 259     def _perform_login(self, username, password):
 260         if not self._ACCESS_TOKEN:
 261             return
 262
 263         self._download_json(
 264             self._OAUTH_URL, None, 'Logging in',
 265             data=urlencode_postdata({
 266                 'password': password,
 267                 'grant_type': 'password',
 268                 'username': username,
 269             }), headers={
 270                 'Content-Type': 'application/x-www-form-urlencoded',
 271                 'Authorization': f'Bearer {self._ACCESS_TOKEN}',
 272             },
 273             errnote='Login failed', fatal=False)
 274
 275     def _real_extract(self, url):
 276         blog, video_id = self._match_valid_url(url).groups()
 277
 278         url = f'http://{blog}.tumblr.com/post/{video_id}/'
 279         webpage, urlh = self._download_webpage_handle(url, video_id)
 280
 281         redirect_url = urlh.geturl()
 282
 283         api_only = bool(self._search_regex(
 284             r'(tumblr.com|^)/(safe-mode|login_required|blog/view)',
 285             redirect_url, 'redirect', default=None))
 286
 287         if api_only and not self._ACCESS_TOKEN:
 288             raise ExtractorError('Cannot get data for dashboard-only post without access token')
 289
 290         post_json = {}
 291         if self._ACCESS_TOKEN:
 292             post_json = traverse_obj(
 293                 self._download_json(
 294                     f'https://www.tumblr.com/api/v2/blog/{blog}/posts/{video_id}/permalink',
 295                     video_id, headers={'Authorization': f'Bearer {self._ACCESS_TOKEN}'}, fatal=False),
 296                 ('response', 'timeline', 'elements', 0)) or {}
 297         content_json = traverse_obj(post_json, ('trail', 0, 'content'), ('content')) or []
 298         video_json = next(
 299             (item for item in content_json if item.get('type') == 'video'), {})
 300         media_json = video_json.get('media') or {}
 301         if api_only and not media_json.get('url') and not video_json.get('url'):
 302             raise ExtractorError('Failed to find video data for dashboard-only post')
 303
 304         if not media_json.get('url') and video_json.get('url'):
 305             # external video host
 306             return self.url_result(
 307                 video_json['url'],
 308                 self._providers.get(video_json.get('provider'), 'Generic'))
 309
 310         video_url = self._og_search_video_url(webpage, default=None)
 311         duration = None
 312         formats = []
 313
 314         # iframes can supply duration and sometimes additional formats, so check for one
 315         iframe_url = self._search_regex(
 316             fr'src=\'(https?://www\.tumblr\.com/video/{blog}/{video_id}/[^\']+)\'',
 317             webpage, 'iframe url', default=None)
 318         if iframe_url:
 319             iframe = self._download_webpage(
 320                 iframe_url, video_id, 'Downloading iframe page',
 321                 headers={'Referer': redirect_url})
 322
 323             options = self._parse_json(
 324                 self._search_regex(
 325                     r'data-crt-options=(["\'])(?P<options>.+?)\1', iframe,
 326                     'hd video url', default='', group='options'),
 327                 video_id, fatal=False)
 328             if options:
 329                 duration = int_or_none(options.get('duration'))
 330
 331                 hd_url = options.get('hdUrl')
 332                 if hd_url:
 333                     # there are multiple formats; extract them
 334                     # ignore other sources of width/height data as they may be wrong
 335                     sources = []
 336                     sd_url = self._search_regex(
 337                         r'<source[^>]+src=(["\'])(?P<url>.+?)\1', iframe,
 338                         'sd video url', default=None, group='url')
 339                     if sd_url:
 340                         sources.append((sd_url, 'sd'))
 341                     sources.append((hd_url, 'hd'))
 342
 343                     formats = [{
 344                         'url': video_url,
 345                         'format_id': format_id,
 346                         'height': int_or_none(self._search_regex(
 347                             r'_(\d+)\.\w+$', video_url, 'height', default=None)),
 348                         'quality': quality,
 349                     } for quality, (video_url, format_id) in enumerate(sources)]
 350
 351         if not media_json.get('url') and not video_url and not iframe_url:
 352             # external video host (but we weren't able to figure it out from the api)
 353             iframe_url = self._search_regex(
 354                 r'src=["\'](https?://safe\.txmblr\.com/svc/embed/inline/[^"\']+)["\']',
 355                 webpage, 'embed iframe url', default=None)
 356             return self.url_result(iframe_url or redirect_url, 'Generic')
 357
 358         formats = formats or [{
 359             'url': media_json.get('url') or video_url,
 360             'width': int_or_none(
 361                 media_json.get('width') or self._og_search_property('video:width', webpage, default=None)),
 362             'height': int_or_none(
 363                 media_json.get('height') or self._og_search_property('video:height', webpage, default=None)),
 364         }]
 365         self._sort_formats(formats)
 366
 367         # the url we're extracting from might be an original post or it might be a reblog.
 368         # if it's a reblog, og:description will be the reblogger's comment, not the uploader's.
 369         # content_json is always the op, so if it exists but has no text, there's no description
 370         if content_json:
 371             description = '\n\n'.join((
 372                 item.get('text') for item in content_json if item.get('type') == 'text')) or None
 373         else:
 374             description = self._og_search_description(webpage, default=None)
 375         uploader_id = traverse_obj(post_json, 'reblogged_root_name', 'blog_name')
 376
 377         return {
 378             'id': video_id,
 379             'title': post_json.get('summary') or (blog if api_only else self._html_search_regex(
 380                 r'(?s)<title>(?P<title>.*?)(?: \| Tumblr)?</title>', webpage, 'title')),
 381             'description': description,
 382             'thumbnail': (traverse_obj(video_json, ('poster', 0, 'url'))
 383                           or self._og_search_thumbnail(webpage, default=None)),
 384             'uploader_id': uploader_id,
 385             'uploader_url': f'https://{uploader_id}.tumblr.com/' if uploader_id else None,
 386             'duration': duration,
 387             'like_count': post_json.get('like_count'),
 388             'repost_count': post_json.get('reblog_count'),
 389             'age_limit': {True: 18, False: 0}.get(post_json.get('is_nsfw')),
 390             'tags': post_json.get('tags'),
 391             'formats': formats,
 392         }