yt_dlp/extractor/patreon.py

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