yt_dlp/extractor/tumblr.py

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