yt_dlp/extractor/patreon.py

   1 import itertools
   2 import urllib.parse
   3
   4 from .common import InfoExtractor
   5 from .sproutvideo import VidsIoIE
   6 from .vimeo import VimeoIE
   7 from ..networking.exceptions import HTTPError
   8 from ..utils import (
   9     KNOWN_EXTENSIONS,
  10     ExtractorError,
  11     clean_html,
  12     determine_ext,
  13     int_or_none,
  14     mimetype2ext,
  15     parse_iso8601,
  16     smuggle_url,
  17     str_or_none,
  18     traverse_obj,
  19     url_or_none,
  20     urljoin,
  21 )
  22
  23
  24 class PatreonBaseIE(InfoExtractor):
  25     USER_AGENT = 'Patreon/7.6.28 (Android; Android 11; Scale/2.10)'
  26
  27     def _call_api(self, ep, item_id, query=None, headers=None, fatal=True, note=None):
  28         if headers is None:
  29             headers = {}
  30         if 'User-Agent' not in headers:
  31             headers['User-Agent'] = self.USER_AGENT
  32         if query:
  33             query.update({'json-api-version': 1.0})
  34
  35         try:
  36             return self._download_json(
  37                 f'https://www.patreon.com/api/{ep}',
  38                 item_id, note=note if note else 'Downloading API JSON',
  39                 query=query, fatal=fatal, headers=headers)
  40         except ExtractorError as e:
  41             if not isinstance(e.cause, HTTPError) or mimetype2ext(e.cause.response.headers.get('Content-Type')) != 'json':
  42                 raise
  43             err_json = self._parse_json(self._webpage_read_content(e.cause.response, None, item_id), item_id, fatal=False)
  44             err_message = traverse_obj(err_json, ('errors', ..., 'detail'), get_all=False)
  45             if err_message:
  46                 raise ExtractorError(f'Patreon said: {err_message}', expected=True)
  47             raise
  48
  49
  50 class PatreonIE(PatreonBaseIE):
  51     _VALID_URL = r'https?://(?:www\.)?patreon\.com/(?:creation\?hid=|posts/(?:[\w-]+-)?)(?P<id>\d+)'
  52     _TESTS = [{
  53         'url': 'http://www.patreon.com/creation?hid=743933',
  54         'md5': 'e25505eec1053a6e6813b8ed369875cc',
  55         'info_dict': {
  56             'id': '743933',
  57             'ext': 'mp3',
  58             'title': 'Episode 166: David Smalley of Dogma Debate',
  59             'description': 'md5:34d207dd29aa90e24f1b3f58841b81c7',
  60             'uploader': 'Cognitive Dissonance Podcast',
  61             'thumbnail': 're:^https?://.*$',
  62             'timestamp': 1406473987,
  63             'upload_date': '20140727',
  64             'uploader_id': '87145',
  65             'like_count': int,
  66             'comment_count': int,
  67             'uploader_url': 'https://www.patreon.com/dissonancepod',
  68             'channel_id': '80642',
  69             'channel_url': 'https://www.patreon.com/dissonancepod',
  70             'channel_follower_count': int,
  71         },
  72     }, {
  73         'url': 'http://www.patreon.com/creation?hid=754133',
  74         'md5': '3eb09345bf44bf60451b8b0b81759d0a',
  75         'info_dict': {
  76             'id': '754133',
  77             'ext': 'mp3',
  78             'title': 'CD 167 Extra',
  79             'uploader': 'Cognitive Dissonance Podcast',
  80             'thumbnail': 're:^https?://.*$',
  81             'like_count': int,
  82             'comment_count': int,
  83             'uploader_url': 'https://www.patreon.com/dissonancepod',
  84         },
  85         'skip': 'Patron-only content',
  86     }, {
  87         'url': 'https://www.patreon.com/creation?hid=1682498',
  88         'info_dict': {
  89             'id': 'SU4fj_aEMVw',
  90             'ext': 'mp4',
  91             'title': 'I\'m on Patreon!',
  92             'uploader': 'TraciJHines',
  93             'thumbnail': 're:^https?://.*$',
  94             'upload_date': '20150211',
  95             'description': 'md5:8af6425f50bd46fbf29f3db0fc3a8364',
  96             'uploader_id': '@TraciHinesMusic',
  97             'categories': ['Entertainment'],
  98             'duration': 282,
  99             'view_count': int,
 100             'tags': 'count:39',
 101             'age_limit': 0,
 102             'channel': 'TraciJHines',
 103             'channel_url': 'https://www.youtube.com/channel/UCGLim4T2loE5rwCMdpCIPVg',
 104             'live_status': 'not_live',
 105             'like_count': int,
 106             'channel_id': 'UCGLim4T2loE5rwCMdpCIPVg',
 107             'availability': 'public',
 108             'channel_follower_count': int,
 109             'playable_in_embed': True,
 110             'uploader_url': 'https://www.youtube.com/@TraciHinesMusic',
 111             'comment_count': int,
 112             'channel_is_verified': True,
 113             'chapters': 'count:4',
 114         },
 115         'params': {
 116             'noplaylist': True,
 117             'skip_download': True,
 118         },
 119     }, {
 120         'url': 'https://www.patreon.com/posts/episode-166-of-743933',
 121         'only_matching': True,
 122     }, {
 123         'url': 'https://www.patreon.com/posts/743933',
 124         'only_matching': True,
 125     }, {
 126         'url': 'https://www.patreon.com/posts/kitchen-as-seen-51706779',
 127         'md5': '96656690071f6d64895866008484251b',
 128         'info_dict': {
 129             'id': '555089736',
 130             'ext': 'mp4',
 131             'title': 'KITCHEN AS SEEN ON DEEZ NUTS EXTENDED!',
 132             'uploader': 'Cold Ones',
 133             'thumbnail': 're:^https?://.*$',
 134             'upload_date': '20210526',
 135             'description': 'md5:557a409bd79d3898689419094934ba79',
 136             'uploader_id': '14936315',
 137         },
 138         'skip': 'Patron-only content',
 139     }, {
 140         # m3u8 video (https://github.com/yt-dlp/yt-dlp/issues/2277)
 141         'url': 'https://www.patreon.com/posts/video-sketchbook-32452882',
 142         'info_dict': {
 143             'id': '32452882',
 144             'ext': 'mp4',
 145             'comment_count': int,
 146             'uploader_id': '4301314',
 147             'like_count': int,
 148             'timestamp': 1576696962,
 149             'upload_date': '20191218',
 150             'thumbnail': r're:^https?://.*$',
 151             'uploader_url': 'https://www.patreon.com/loish',
 152             'description': 'md5:e2693e97ee299c8ece47ffdb67e7d9d2',
 153             'title': 'VIDEO // sketchbook flipthrough',
 154             'uploader': 'Loish ',
 155             'tags': ['sketchbook', 'video'],
 156             'channel_id': '1641751',
 157             'channel_url': 'https://www.patreon.com/loish',
 158             'channel_follower_count': int,
 159         },
 160     }, {
 161         # bad videos under media (if media is included). Real one is under post_file
 162         'url': 'https://www.patreon.com/posts/premium-access-70282931',
 163         'info_dict': {
 164             'id': '70282931',
 165             'ext': 'mp4',
 166             'title': '[Premium Access + Uncut] The Office - 2x6 The Fight - Group Reaction',
 167             'channel_url': 'https://www.patreon.com/thenormies',
 168             'channel_id': '573397',
 169             'uploader_id': '2929435',
 170             'uploader': 'The Normies',
 171             'description': 'md5:79c9fd8778e2cef84049a94c058a5e23',
 172             'comment_count': int,
 173             'upload_date': '20220809',
 174             'thumbnail': r're:^https?://.*$',
 175             'channel_follower_count': int,
 176             'like_count': int,
 177             'timestamp': 1660052820,
 178             'tags': ['The Office', 'early access', 'uncut'],
 179             'uploader_url': 'https://www.patreon.com/thenormies',
 180         },
 181         'skip': 'Patron-only content',
 182     }, {
 183         # dead vimeo and embed URLs, need to extract post_file
 184         'url': 'https://www.patreon.com/posts/hunter-x-hunter-34007913',
 185         'info_dict': {
 186             'id': '34007913',
 187             'ext': 'mp4',
 188             'title': 'Hunter x Hunter | Kurapika DESTROYS Uvogin!!!',
 189             'like_count': int,
 190             'uploader': 'YaBoyRoshi',
 191             'timestamp': 1581636833,
 192             'channel_url': 'https://www.patreon.com/yaboyroshi',
 193             'thumbnail': r're:^https?://.*$',
 194             'tags': ['Hunter x Hunter'],
 195             'uploader_id': '14264111',
 196             'comment_count': int,
 197             'channel_follower_count': int,
 198             'description': 'Kurapika is a walking cheat code!',
 199             'upload_date': '20200213',
 200             'channel_id': '2147162',
 201             'uploader_url': 'https://www.patreon.com/yaboyroshi',
 202         },
 203     }, {
 204         # NSFW vimeo embed URL
 205         'url': 'https://www.patreon.com/posts/4k-spiderman-4k-96414599',
 206         'info_dict': {
 207             'id': '902250943',
 208             'ext': 'mp4',
 209             'title': '❤️(4K) Spiderman Girl Yeonhwa’s Gift ❤️(4K) 스파이더맨걸 연화의 선물',
 210             'description': '❤️(4K) Spiderman Girl Yeonhwa’s Gift \n❤️(4K) 스파이더맨걸 연화의 선물',
 211             'uploader': 'Npickyeonhwa',
 212             'uploader_id': '90574422',
 213             'uploader_url': 'https://www.patreon.com/Yeonhwa726',
 214             'channel_id': '10237902',
 215             'channel_url': 'https://www.patreon.com/Yeonhwa726',
 216             'duration': 70,
 217             'timestamp': 1705150153,
 218             'upload_date': '20240113',
 219             'comment_count': int,
 220             'like_count': int,
 221             'thumbnail': r're:^https?://.+',
 222         },
 223         'params': {'skip_download': 'm3u8'},
 224     }, {
 225         # multiple attachments/embeds
 226         'url': 'https://www.patreon.com/posts/holy-wars-solos-100601977',
 227         'playlist_count': 3,
 228         'info_dict': {
 229             'id': '100601977',
 230             'title': '"Holy Wars" (Megadeth) Solos Transcription & Lesson/Analysis',
 231             'description': 'md5:d099ab976edfce6de2a65c2b169a88d3',
 232             'uploader': 'Bradley Hall',
 233             'uploader_id': '24401883',
 234             'uploader_url': 'https://www.patreon.com/bradleyhallguitar',
 235             'channel_id': '3193932',
 236             'channel_url': 'https://www.patreon.com/bradleyhallguitar',
 237             'channel_follower_count': int,
 238             'timestamp': 1710777855,
 239             'upload_date': '20240318',
 240             'like_count': int,
 241             'comment_count': int,
 242             'thumbnail': r're:^https?://.+',
 243         },
 244         'skip': 'Patron-only content',
 245     }]
 246     _RETURN_TYPE = 'video'
 247
 248     def _real_extract(self, url):
 249         video_id = self._match_id(url)
 250         post = self._call_api(
 251             f'posts/{video_id}', video_id, query={
 252                 'fields[media]': 'download_url,mimetype,size_bytes',
 253                 'fields[post]': 'comment_count,content,embed,image,like_count,post_file,published_at,title,current_user_can_view',
 254                 'fields[user]': 'full_name,url',
 255                 'fields[post_tag]': 'value',
 256                 'fields[campaign]': 'url,name,patron_count',
 257                 'json-api-use-default-includes': 'false',
 258                 'include': 'audio,user,user_defined_tags,campaign,attachments_media',
 259             })
 260         attributes = post['data']['attributes']
 261         info = traverse_obj(attributes, {
 262             'title': ('title', {str.strip}),
 263             'description': ('content', {clean_html}),
 264             'thumbnail': ('image', ('large_url', 'url'), {url_or_none}, any),
 265             'timestamp': ('published_at', {parse_iso8601}),
 266             'like_count': ('like_count', {int_or_none}),
 267             'comment_count': ('comment_count', {int_or_none}),
 268         })
 269
 270         entries = []
 271         idx = 0
 272         for include in traverse_obj(post, ('included', lambda _, v: v['type'])):
 273             include_type = include['type']
 274             if include_type == 'media':
 275                 media_attributes = traverse_obj(include, ('attributes', {dict})) or {}
 276                 download_url = url_or_none(media_attributes.get('download_url'))
 277                 ext = mimetype2ext(media_attributes.get('mimetype'))
 278
 279                 # if size_bytes is None, this media file is likely unavailable
 280                 # See: https://github.com/yt-dlp/yt-dlp/issues/4608
 281                 size_bytes = int_or_none(media_attributes.get('size_bytes'))
 282                 if download_url and ext in KNOWN_EXTENSIONS and size_bytes is not None:
 283                     idx += 1
 284                     entries.append({
 285                         'id': f'{video_id}-{idx}',
 286                         'ext': ext,
 287                         'filesize': size_bytes,
 288                         'url': download_url,
 289                     })
 290
 291             elif include_type == 'user':
 292                 info.update(traverse_obj(include, {
 293                     'uploader': ('attributes', 'full_name', {str}),
 294                     'uploader_id': ('id', {str_or_none}),
 295                     'uploader_url': ('attributes', 'url', {url_or_none}),
 296                 }))
 297
 298             elif include_type == 'post_tag':
 299                 if post_tag := traverse_obj(include, ('attributes', 'value', {str})):
 300                     info.setdefault('tags', []).append(post_tag)
 301
 302             elif include_type == 'campaign':
 303                 info.update(traverse_obj(include, {
 304                     'channel': ('attributes', 'title', {str}),
 305                     'channel_id': ('id', {str_or_none}),
 306                     'channel_url': ('attributes', 'url', {url_or_none}),
 307                     'channel_follower_count': ('attributes', 'patron_count', {int_or_none}),
 308                 }))
 309
 310         # all-lowercase 'referer' so we can smuggle it to Generic, SproutVideo, Vimeo
 311         headers = {'referer': 'https://patreon.com/'}
 312
 313         # handle Vimeo embeds
 314         if traverse_obj(attributes, ('embed', 'provider')) == 'Vimeo':
 315             v_url = urllib.parse.unquote(self._html_search_regex(
 316                 r'(https(?:%3A%2F%2F|://)player\.vimeo\.com.+app_id(?:=|%3D)+\d+)',
 317                 traverse_obj(attributes, ('embed', 'html', {str})), 'vimeo url', fatal=False) or '')
 318             if url_or_none(v_url) and self._request_webpage(
 319                     v_url, video_id, 'Checking Vimeo embed URL', headers=headers,
 320                     fatal=False, errnote=False, expected_status=429):  # 429 is TLS fingerprint rejection
 321                 entries.append(self.url_result(
 322                     VimeoIE._smuggle_referrer(v_url, 'https://patreon.com/'),
 323                     VimeoIE, url_transparent=True))
 324
 325         embed_url = traverse_obj(attributes, ('embed', 'url', {url_or_none}))
 326         if embed_url and (urlh := self._request_webpage(
 327                 embed_url, video_id, 'Checking embed URL', headers=headers,
 328                 fatal=False, errnote=False, expected_status=403)):
 329             # Password-protected vids.io embeds return 403 errors w/o --video-password or session cookie
 330             if urlh.status != 403 or VidsIoIE.suitable(embed_url):
 331                 entries.append(self.url_result(smuggle_url(embed_url, headers)))
 332
 333         post_file = traverse_obj(attributes, ('post_file', {dict}))
 334         if post_file:
 335             name = post_file.get('name')
 336             ext = determine_ext(name)
 337             if ext in KNOWN_EXTENSIONS:
 338                 entries.append({
 339                     'id': video_id,
 340                     'ext': ext,
 341                     'url': post_file['url'],
 342                 })
 343             elif name == 'video' or determine_ext(post_file.get('url')) == 'm3u8':
 344                 formats, subtitles = self._extract_m3u8_formats_and_subtitles(post_file['url'], video_id)
 345                 entries.append({
 346                     'id': video_id,
 347                     'formats': formats,
 348                     'subtitles': subtitles,
 349                 })
 350
 351         can_view_post = traverse_obj(attributes, 'current_user_can_view')
 352         comments = None
 353         if can_view_post and info.get('comment_count'):
 354             comments = self.extract_comments(video_id)
 355
 356         if not entries and can_view_post is False:
 357             self.raise_no_formats('You do not have access to this post', video_id=video_id, expected=True)
 358         elif not entries:
 359             self.raise_no_formats('No supported media found in this post', video_id=video_id, expected=True)
 360         elif len(entries) == 1:
 361             info.update(entries[0])
 362         else:
 363             for entry in entries:
 364                 entry.update(info)
 365             return self.playlist_result(entries, video_id, **info, __post_extractor=comments)
 366
 367         info['id'] = video_id
 368         info['__post_extractor'] = comments
 369         return info
 370
 371     def _get_comments(self, post_id):
 372         cursor = None
 373         count = 0
 374         params = {
 375             'page[count]': 50,
 376             'include': 'parent.commenter.campaign,parent.post.user,parent.post.campaign.creator,parent.replies.parent,parent.replies.commenter.campaign,parent.replies.post.user,parent.replies.post.campaign.creator,commenter.campaign,post.user,post.campaign.creator,replies.parent,replies.commenter.campaign,replies.post.user,replies.post.campaign.creator,on_behalf_of_campaign',
 377             'fields[comment]': 'body,created,is_by_creator',
 378             'fields[user]': 'image_url,full_name,url',
 379             'filter[flair]': 'image_tiny_url,name',
 380             'sort': '-created',
 381             'json-api-version': 1.0,
 382             'json-api-use-default-includes': 'false',
 383         }
 384
 385         for page in itertools.count(1):
 386
 387             params.update({'page[cursor]': cursor} if cursor else {})
 388             response = self._call_api(
 389                 f'posts/{post_id}/comments', post_id, query=params, note=f'Downloading comments page {page}')
 390
 391             cursor = None
 392             for comment in traverse_obj(response, (('data', ('included', lambda _, v: v['type'] == 'comment')), ...)):
 393                 count += 1
 394                 comment_id = comment.get('id')
 395                 attributes = comment.get('attributes') or {}
 396                 if comment_id is None:
 397                     continue
 398                 author_id = traverse_obj(comment, ('relationships', 'commenter', 'data', 'id'))
 399                 author_info = traverse_obj(
 400                     response, ('included', lambda _, v: v['id'] == author_id and v['type'] == 'user', 'attributes'),
 401                     get_all=False, expected_type=dict, default={})
 402
 403                 yield {
 404                     'id': comment_id,
 405                     'text': attributes.get('body'),
 406                     'timestamp': parse_iso8601(attributes.get('created')),
 407                     'parent': traverse_obj(comment, ('relationships', 'parent', 'data', 'id'), default='root'),
 408                     'author_is_uploader': attributes.get('is_by_creator'),
 409                     'author_id': author_id,
 410                     'author': author_info.get('full_name'),
 411                     'author_thumbnail': author_info.get('image_url'),
 412                 }
 413
 414             if count < traverse_obj(response, ('meta', 'count')):
 415                 cursor = traverse_obj(response, ('data', -1, 'id'))
 416
 417             if cursor is None:
 418                 break
 419
 420
 421 class PatreonCampaignIE(PatreonBaseIE):
 422
 423     _VALID_URL = r'https?://(?:www\.)?patreon\.com/(?!rss)(?:(?:m/(?P<campaign_id>\d+))|(?P<vanity>[-\w]+))'
 424     _TESTS = [{
 425         'url': 'https://www.patreon.com/dissonancepod/',
 426         'info_dict': {
 427             'title': 'Cognitive Dissonance Podcast',
 428             'channel_url': 'https://www.patreon.com/dissonancepod',
 429             'id': '80642',
 430             'description': 'md5:eb2fa8b83da7ab887adeac34da6b7af7',
 431             'channel_id': '80642',
 432             'channel': 'Cognitive Dissonance Podcast',
 433             'age_limit': 0,
 434             'channel_follower_count': int,
 435             'uploader_id': '87145',
 436             'uploader_url': 'https://www.patreon.com/dissonancepod',
 437             'uploader': 'Cognitive Dissonance Podcast',
 438             'thumbnail': r're:^https?://.*$',
 439         },
 440         'playlist_mincount': 68,
 441     }, {
 442         'url': 'https://www.patreon.com/m/4767637/posts',
 443         'info_dict': {
 444             'title': 'Not Just Bikes',
 445             'channel_follower_count': int,
 446             'id': '4767637',
 447             'channel_id': '4767637',
 448             'channel_url': 'https://www.patreon.com/notjustbikes',
 449             'description': 'md5:595c6e7dca76ae615b1d38c298a287a1',
 450             'age_limit': 0,
 451             'channel': 'Not Just Bikes',
 452             'uploader_url': 'https://www.patreon.com/notjustbikes',
 453             'uploader': 'Not Just Bikes',
 454             'uploader_id': '37306634',
 455             'thumbnail': r're:^https?://.*$',
 456         },
 457         'playlist_mincount': 71,
 458     }, {
 459         'url': 'https://www.patreon.com/dissonancepod/posts',
 460         'only_matching': True,
 461     }, {
 462         'url': 'https://www.patreon.com/m/5932659',
 463         'only_matching': True,
 464     }]
 465
 466     @classmethod
 467     def suitable(cls, url):
 468         return False if PatreonIE.suitable(url) else super().suitable(url)
 469
 470     def _entries(self, campaign_id):
 471         cursor = None
 472         params = {
 473             'fields[post]': 'patreon_url,url',
 474             'filter[campaign_id]': campaign_id,
 475             'filter[is_draft]': 'false',
 476             'sort': '-published_at',
 477             'json-api-use-default-includes': 'false',
 478         }
 479
 480         for page in itertools.count(1):
 481
 482             params.update({'page[cursor]': cursor} if cursor else {})
 483             posts_json = self._call_api('posts', campaign_id, query=params, note=f'Downloading posts page {page}')
 484
 485             cursor = traverse_obj(posts_json, ('meta', 'pagination', 'cursors', 'next'))
 486             for post_url in traverse_obj(posts_json, ('data', ..., 'attributes', 'patreon_url')):
 487                 yield self.url_result(urljoin('https://www.patreon.com/', post_url), PatreonIE)
 488
 489             if cursor is None:
 490                 break
 491
 492     def _real_extract(self, url):
 493
 494         campaign_id, vanity = self._match_valid_url(url).group('campaign_id', 'vanity')
 495         if campaign_id is None:
 496             webpage = self._download_webpage(url, vanity, headers={'User-Agent': self.USER_AGENT})
 497             campaign_id = self._search_nextjs_data(
 498                 webpage, vanity)['props']['pageProps']['bootstrapEnvelope']['pageBootstrap']['campaign']['data']['id']
 499
 500         params = {
 501             'json-api-use-default-includes': 'false',
 502             'fields[user]': 'full_name,url',
 503             'fields[campaign]': 'name,summary,url,patron_count,creation_count,is_nsfw,avatar_photo_url',
 504             'include': 'creator',
 505         }
 506
 507         campaign_response = self._call_api(
 508             f'campaigns/{campaign_id}', campaign_id,
 509             note='Downloading campaign info', fatal=False,
 510             query=params) or {}
 511
 512         campaign_info = campaign_response.get('data') or {}
 513         channel_name = traverse_obj(campaign_info, ('attributes', 'name'))
 514         user_info = traverse_obj(
 515             campaign_response, ('included', lambda _, v: v['type'] == 'user'),
 516             default={}, expected_type=dict, get_all=False)
 517
 518         return {
 519             '_type': 'playlist',
 520             'id': campaign_id,
 521             'title': channel_name,
 522             'entries': self._entries(campaign_id),
 523             'description': clean_html(traverse_obj(campaign_info, ('attributes', 'summary'))),
 524             'channel_url': traverse_obj(campaign_info, ('attributes', 'url')),
 525             'channel_follower_count': int_or_none(traverse_obj(campaign_info, ('attributes', 'patron_count'))),
 526             'channel_id': campaign_id,
 527             'channel': channel_name,
 528             'uploader_url': traverse_obj(user_info, ('attributes', 'url')),
 529             'uploader_id': str_or_none(user_info.get('id')),
 530             'uploader': traverse_obj(user_info, ('attributes', 'full_name')),
 531             'playlist_count': traverse_obj(campaign_info, ('attributes', 'creation_count')),
 532             'age_limit': 18 if traverse_obj(campaign_info, ('attributes', 'is_nsfw')) else 0,
 533             'thumbnail': url_or_none(traverse_obj(campaign_info, ('attributes', 'avatar_photo_url'))),
 534         }