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
 159     def _real_extract(self, url):
 160         video_id = self._match_id(url)
 161         post = self._call_api(
 162             f'posts/{video_id}', video_id, query={
 163                 'fields[media]': 'download_url,mimetype,size_bytes',
 164                 'fields[post]': 'comment_count,content,embed,image,like_count,post_file,published_at,title,current_user_can_view',
 165                 'fields[user]': 'full_name,url',
 166                 'fields[post_tag]': 'value',
 167                 'fields[campaign]': 'url,name,patron_count',
 168                 'json-api-use-default-includes': 'false',
 169                 'include': 'media,user,user_defined_tags,campaign',
 170             })
 171         attributes = post['data']['attributes']
 172         title = attributes['title'].strip()
 173         image = attributes.get('image') or {}
 174         info = {
 175             'id': video_id,
 176             'title': title,
 177             'description': clean_html(attributes.get('content')),
 178             'thumbnail': image.get('large_url') or image.get('url'),
 179             'timestamp': parse_iso8601(attributes.get('published_at')),
 180             'like_count': int_or_none(attributes.get('like_count')),
 181             'comment_count': int_or_none(attributes.get('comment_count')),
 182         }
 183         can_view_post = traverse_obj(attributes, 'current_user_can_view')
 184         if can_view_post and info['comment_count']:
 185             info['__post_extractor'] = self.extract_comments(video_id)
 186
 187         for i in post.get('included', []):
 188             i_type = i.get('type')
 189             if i_type == 'media':
 190                 media_attributes = i.get('attributes') or {}
 191                 download_url = media_attributes.get('download_url')
 192                 ext = mimetype2ext(media_attributes.get('mimetype'))
 193                 if download_url and ext in KNOWN_EXTENSIONS:
 194                     return {
 195                         **info,
 196                         'ext': ext,
 197                         'filesize': int_or_none(media_attributes.get('size_bytes')),
 198                         'url': download_url,
 199                     }
 200             elif i_type == 'user':
 201                 user_attributes = i.get('attributes')
 202                 if user_attributes:
 203                     info.update({
 204                         'uploader': user_attributes.get('full_name'),
 205                         'uploader_id': str_or_none(i.get('id')),
 206                         'uploader_url': user_attributes.get('url'),
 207                     })
 208
 209             elif i_type == 'post_tag':
 210                 info.setdefault('tags', []).append(traverse_obj(i, ('attributes', 'value')))
 211
 212             elif i_type == 'campaign':
 213                 info.update({
 214                     'channel': traverse_obj(i, ('attributes', 'title')),
 215                     'channel_id': str_or_none(i.get('id')),
 216                     'channel_url': traverse_obj(i, ('attributes', 'url')),
 217                     'channel_follower_count': int_or_none(traverse_obj(i, ('attributes', 'patron_count'))),
 218                 })
 219
 220         # handle Vimeo embeds
 221         if try_get(attributes, lambda x: x['embed']['provider']) == 'Vimeo':
 222             embed_html = try_get(attributes, lambda x: x['embed']['html'])
 223             v_url = url_or_none(compat_urllib_parse_unquote(
 224                 self._search_regex(r'(https(?:%3A%2F%2F|://)player\.vimeo\.com.+app_id(?:=|%3D)+\d+)', embed_html, 'vimeo url', fatal=False)))
 225             if v_url:
 226                 return {
 227                     **info,
 228                     '_type': 'url_transparent',
 229                     'url': VimeoIE._smuggle_referrer(v_url, 'https://patreon.com'),
 230                     'ie_key': 'Vimeo',
 231                 }
 232
 233         embed_url = try_get(attributes, lambda x: x['embed']['url'])
 234         if embed_url:
 235             return {
 236                 **info,
 237                 '_type': 'url',
 238                 'url': embed_url,
 239             }
 240
 241         post_file = traverse_obj(attributes, 'post_file')
 242         if post_file:
 243             name = post_file.get('name')
 244             ext = determine_ext(name)
 245             if ext in KNOWN_EXTENSIONS:
 246                 return {
 247                     **info,
 248                     'ext': ext,
 249                     'url': post_file['url'],
 250                 }
 251             elif name == 'video':
 252                 formats, subtitles = self._extract_m3u8_formats_and_subtitles(post_file['url'], video_id)
 253                 return {
 254                     **info,
 255                     'formats': formats,
 256                     'subtitles': subtitles,
 257                 }
 258
 259         if can_view_post is False:
 260             self.raise_no_formats('You do not have access to this post', video_id=video_id, expected=True)
 261         else:
 262             self.raise_no_formats('No supported media found in this post', video_id=video_id, expected=True)
 263         return info
 264
 265     def _get_comments(self, post_id):
 266         cursor = None
 267         count = 0
 268         params = {
 269             'page[count]': 50,
 270             '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',
 271             'fields[comment]': 'body,created,is_by_creator',
 272             'fields[user]': 'image_url,full_name,url',
 273             'filter[flair]': 'image_tiny_url,name',
 274             'sort': '-created',
 275             'json-api-version': 1.0,
 276             'json-api-use-default-includes': 'false',
 277         }
 278
 279         for page in itertools.count(1):
 280
 281             params.update({'page[cursor]': cursor} if cursor else {})
 282             response = self._call_api(
 283                 f'posts/{post_id}/comments', post_id, query=params, note='Downloading comments page %d' % page)
 284
 285             cursor = None
 286             for comment in traverse_obj(response, (('data', ('included', lambda _, v: v['type'] == 'comment')), ...), default=[]):
 287                 count += 1
 288                 comment_id = comment.get('id')
 289                 attributes = comment.get('attributes') or {}
 290                 if comment_id is None:
 291                     continue
 292                 author_id = traverse_obj(comment, ('relationships', 'commenter', 'data', 'id'))
 293                 author_info = traverse_obj(
 294                     response, ('included', lambda _, v: v['id'] == author_id and v['type'] == 'user', 'attributes'),
 295                     get_all=False, expected_type=dict, default={})
 296
 297                 yield {
 298                     'id': comment_id,
 299                     'text': attributes.get('body'),
 300                     'timestamp': parse_iso8601(attributes.get('created')),
 301                     'parent': traverse_obj(comment, ('relationships', 'parent', 'data', 'id'), default='root'),
 302                     'author_is_uploader': attributes.get('is_by_creator'),
 303                     'author_id': author_id,
 304                     'author': author_info.get('full_name'),
 305                     'author_thumbnail': author_info.get('image_url'),
 306                 }
 307
 308             if count < traverse_obj(response, ('meta', 'count')):
 309                 cursor = traverse_obj(response, ('data', -1, 'id'))
 310
 311             if cursor is None:
 312                 break
 313
 314
 315 class PatreonCampaignIE(PatreonBaseIE):
 316
 317     _VALID_URL = r'https?://(?:www\.)?patreon\.com/(?!rss)(?:(?:m/(?P<campaign_id>\d+))|(?P<vanity>[-\w]+))'
 318     _TESTS = [{
 319         'url': 'https://www.patreon.com/dissonancepod/',
 320         'info_dict': {
 321             'title': 'Cognitive Dissonance Podcast',
 322             'channel_url': 'https://www.patreon.com/dissonancepod',
 323             'id': '80642',
 324             'description': 'md5:eb2fa8b83da7ab887adeac34da6b7af7',
 325             'channel_id': '80642',
 326             'channel': 'Cognitive Dissonance Podcast',
 327             'age_limit': 0,
 328             'channel_follower_count': int,
 329             'uploader_id': '87145',
 330             'uploader_url': 'https://www.patreon.com/dissonancepod',
 331             'uploader': 'Cognitive Dissonance Podcast',
 332             'thumbnail': r're:^https?://.*$',
 333         },
 334         'playlist_mincount': 68,
 335     }, {
 336         'url': 'https://www.patreon.com/m/4767637/posts',
 337         'info_dict': {
 338             'title': 'Not Just Bikes',
 339             'channel_follower_count': int,
 340             'id': '4767637',
 341             'channel_id': '4767637',
 342             'channel_url': 'https://www.patreon.com/notjustbikes',
 343             'description': 'md5:595c6e7dca76ae615b1d38c298a287a1',
 344             'age_limit': 0,
 345             'channel': 'Not Just Bikes',
 346             'uploader_url': 'https://www.patreon.com/notjustbikes',
 347             'uploader': 'Not Just Bikes',
 348             'uploader_id': '37306634',
 349             'thumbnail': r're:^https?://.*$',
 350         },
 351         'playlist_mincount': 71
 352     }, {
 353         'url': 'https://www.patreon.com/dissonancepod/posts',
 354         'only_matching': True
 355     }, {
 356         'url': 'https://www.patreon.com/m/5932659',
 357         'only_matching': True
 358     }]
 359
 360     @classmethod
 361     def suitable(cls, url):
 362         return False if PatreonIE.suitable(url) else super(PatreonCampaignIE, cls).suitable(url)
 363
 364     def _entries(self, campaign_id):
 365         cursor = None
 366         params = {
 367             'fields[post]': 'patreon_url,url',
 368             'filter[campaign_id]': campaign_id,
 369             'filter[is_draft]': 'false',
 370             'sort': '-published_at',
 371             'json-api-use-default-includes': 'false',
 372         }
 373
 374         for page in itertools.count(1):
 375
 376             params.update({'page[cursor]': cursor} if cursor else {})
 377             posts_json = self._call_api('posts', campaign_id, query=params, note='Downloading posts page %d' % page)
 378
 379             cursor = traverse_obj(posts_json, ('meta', 'pagination', 'cursors', 'next'))
 380             for post in posts_json.get('data') or []:
 381                 yield self.url_result(url_or_none(traverse_obj(post, ('attributes', 'patreon_url'))), 'Patreon')
 382
 383             if cursor is None:
 384                 break
 385
 386     def _real_extract(self, url):
 387
 388         campaign_id, vanity = self._match_valid_url(url).group('campaign_id', 'vanity')
 389         if campaign_id is None:
 390             webpage = self._download_webpage(url, vanity, headers={'User-Agent': self.USER_AGENT})
 391             campaign_id = self._search_regex(r'https://www.patreon.com/api/campaigns/(\d+)/?', webpage, 'Campaign ID')
 392
 393         params = {
 394             'json-api-use-default-includes': 'false',
 395             'fields[user]': 'full_name,url',
 396             'fields[campaign]': 'name,summary,url,patron_count,creation_count,is_nsfw,avatar_photo_url',
 397             'include': 'creator'
 398         }
 399
 400         campaign_response = self._call_api(
 401             f'campaigns/{campaign_id}', campaign_id,
 402             note='Downloading campaign info', fatal=False,
 403             query=params) or {}
 404
 405         campaign_info = campaign_response.get('data') or {}
 406         channel_name = traverse_obj(campaign_info, ('attributes', 'name'))
 407         user_info = traverse_obj(
 408             campaign_response, ('included', lambda _, v: v['type'] == 'user'),
 409             default={}, expected_type=dict, get_all=False)
 410
 411         return {
 412             '_type': 'playlist',
 413             'id': campaign_id,
 414             'title': channel_name,
 415             'entries': self._entries(campaign_id),
 416             'description': clean_html(traverse_obj(campaign_info, ('attributes', 'summary'))),
 417             'channel_url': traverse_obj(campaign_info, ('attributes', 'url')),
 418             'channel_follower_count': int_or_none(traverse_obj(campaign_info, ('attributes', 'patron_count'))),
 419             'channel_id': campaign_id,
 420             'channel': channel_name,
 421             'uploader_url': traverse_obj(user_info, ('attributes', 'url')),
 422             'uploader_id': str_or_none(user_info.get('id')),
 423             'uploader': traverse_obj(user_info, ('attributes', 'full_name')),
 424             'playlist_count': traverse_obj(campaign_info, ('attributes', 'creation_count')),
 425             'age_limit': 18 if traverse_obj(campaign_info, ('attributes', 'is_nsfw')) else 0,
 426             'thumbnail': url_or_none(traverse_obj(campaign_info, ('attributes', 'avatar_photo_url'))),
 427         }