yt_dlp/extractor/imgur.py

   1 import functools
   2 import re
   3
   4 from .common import InfoExtractor
   5 from ..utils import (
   6     ExtractorError,
   7     determine_ext,
   8     float_or_none,
   9     int_or_none,
  10     js_to_json,
  11     mimetype2ext,
  12     parse_iso8601,
  13     str_or_none,
  14     strip_or_none,
  15     traverse_obj,
  16     url_or_none,
  17 )
  18
  19
  20 class ImgurBaseIE(InfoExtractor):
  21     _CLIENT_ID = '546c25a59c58ad7'
  22
  23     @classmethod
  24     def _imgur_result(cls, item_id):
  25         return cls.url_result(f'https://imgur.com/{item_id}', ImgurIE, item_id)
  26
  27     def _call_api(self, endpoint, video_id, **kwargs):
  28         return self._download_json(
  29             f'https://api.imgur.com/post/v1/{endpoint}/{video_id}?client_id={self._CLIENT_ID}&include=media,account',
  30             video_id, **kwargs)
  31
  32     @staticmethod
  33     def get_description(s):
  34         if 'Discover the magic of the internet at Imgur' in s:
  35             return None
  36         return s or None
  37
  38
  39 class ImgurIE(ImgurBaseIE):
  40     _VALID_URL = r'https?://(?:i\.)?imgur\.com/(?!(?:a|gallery|t|topic|r)/)(?P<id>[a-zA-Z0-9]+)'
  41
  42     _TESTS = [{
  43         'url': 'https://imgur.com/A61SaA1',
  44         'info_dict': {
  45             'id': 'A61SaA1',
  46             'ext': 'mp4',
  47             'title': 'MRW gifv is up and running without any bugs',
  48             'timestamp': 1416446068,
  49             'upload_date': '20141120',
  50             'dislike_count': int,
  51             'comment_count': int,
  52             'release_timestamp': 1416446068,
  53             'release_date': '20141120',
  54             'like_count': int,
  55             'thumbnail': 'https://i.imgur.com/A61SaA1h.jpg',
  56         },
  57     }, {
  58         'url': 'https://i.imgur.com/A61SaA1.gifv',
  59         'only_matching': True,
  60     }, {
  61         'url': 'https://i.imgur.com/crGpqCV.mp4',
  62         'only_matching': True,
  63     }, {
  64         'url': 'https://i.imgur.com/jxBXAMC.gifv',
  65         'info_dict': {
  66             'id': 'jxBXAMC',
  67             'ext': 'mp4',
  68             'title': 'Fahaka puffer feeding',
  69             'timestamp': 1533835503,
  70             'upload_date': '20180809',
  71             'release_date': '20180809',
  72             'like_count': int,
  73             'duration': 30.0,
  74             'comment_count': int,
  75             'release_timestamp': 1533835503,
  76             'thumbnail': 'https://i.imgur.com/jxBXAMCh.jpg',
  77             'dislike_count': int,
  78         },
  79     }, {
  80         # needs Accept header, ref: https://github.com/yt-dlp/yt-dlp/issues/9458
  81         'url': 'https://imgur.com/zV03bd5',
  82         'md5': '59df97884e8ba76143ff6b640a0e2904',
  83         'info_dict': {
  84             'id': 'zV03bd5',
  85             'ext': 'mp4',
  86             'title': 'Ive - Liz',
  87             'timestamp': 1710491255,
  88             'upload_date': '20240315',
  89             'like_count': int,
  90             'dislike_count': int,
  91             'duration': 56.92,
  92             'comment_count': int,
  93             'release_timestamp': 1710491255,
  94             'release_date': '20240315',
  95         },
  96     }]
  97
  98     def _real_extract(self, url):
  99         video_id = self._match_id(url)
 100         data = self._call_api('media', video_id)
 101         if not traverse_obj(data, ('media', 0, (
 102                 ('type', {lambda t: t == 'video' or None}),
 103                 ('metadata', 'is_animated'))), get_all=False):
 104             raise ExtractorError(f'{video_id} is not a video or animated image', expected=True)
 105         webpage = self._download_webpage(
 106             f'https://i.imgur.com/{video_id}.gifv', video_id, fatal=False) or ''
 107         formats = []
 108
 109         media_fmt = traverse_obj(data, ('media', 0, {
 110             'url': ('url', {url_or_none}),
 111             'ext': ('ext', {str}),
 112             'width': ('width', {int_or_none}),
 113             'height': ('height', {int_or_none}),
 114             'filesize': ('size', {int_or_none}),
 115             'acodec': ('metadata', 'has_sound', {lambda b: None if b else 'none'}),
 116         }))
 117         media_url = media_fmt.get('url')
 118         if media_url:
 119             if not media_fmt.get('ext'):
 120                 media_fmt['ext'] = mimetype2ext(traverse_obj(
 121                     data, ('media', 0, 'mime_type'))) or determine_ext(media_url)
 122             if traverse_obj(data, ('media', 0, 'type')) == 'image':
 123                 media_fmt['acodec'] = 'none'
 124                 media_fmt.setdefault('preference', -10)
 125             formats.append(media_fmt)
 126
 127         video_elements = self._search_regex(
 128             r'(?s)<div class="video-elements">(.*?)</div>',
 129             webpage, 'video elements', default=None)
 130
 131         if video_elements:
 132             def og_get_size(media_type):
 133                 return {
 134                     p: int_or_none(self._og_search_property(f'{media_type}:{p}', webpage, default=None))
 135                     for p in ('width', 'height')
 136                 }
 137
 138             size = og_get_size('video')
 139             if not any(size.values()):
 140                 size = og_get_size('image')
 141
 142             formats = traverse_obj(
 143                 re.finditer(r'<source\s+src="(?P<src>[^"]+)"\s+type="(?P<type>[^"]+)"', video_elements),
 144                 (..., {
 145                     'format_id': ('type', {lambda s: s.partition('/')[2]}),
 146                     'url': ('src', {self._proto_relative_url}),
 147                     'ext': ('type', {mimetype2ext}),
 148                 }))
 149             for f in formats:
 150                 f.update(size)
 151
 152             # We can get the original gif format from the webpage as well
 153             gif_json = traverse_obj(self._search_json(
 154                 r'var\s+videoItem\s*=', webpage, 'GIF info', video_id,
 155                 transform_source=js_to_json, fatal=False), {
 156                     'url': ('gifUrl', {self._proto_relative_url}),
 157                     'filesize': ('size', {int_or_none}),
 158             })
 159             if gif_json:
 160                 gif_json.update(size)
 161                 gif_json.update({
 162                     'format_id': 'gif',
 163                     'preference': -10,  # gifs < videos
 164                     'ext': 'gif',
 165                     'acodec': 'none',
 166                     'vcodec': 'gif',
 167                     'container': 'gif',
 168                 })
 169                 formats.append(gif_json)
 170
 171         search = functools.partial(self._html_search_meta, html=webpage, default=None)
 172
 173         twitter_fmt = {
 174             'format_id': 'twitter',
 175             'url': url_or_none(search('twitter:player:stream')),
 176             'ext': mimetype2ext(search('twitter:player:stream:content_type')),
 177             'width': int_or_none(search('twitter:width')),
 178             'height': int_or_none(search('twitter:height')),
 179         }
 180         if twitter_fmt['url']:
 181             formats.append(twitter_fmt)
 182
 183         if not formats:
 184             self.raise_no_formats(
 185                 f'No sources found for video {video_id}. Maybe a plain image?', expected=True)
 186         self._remove_duplicate_formats(formats)
 187
 188         return {
 189             'title': self._og_search_title(webpage, default=None),
 190             'description': self.get_description(self._og_search_description(webpage, default='')),
 191             **traverse_obj(data, {
 192                 'uploader_id': ('account_id', {lambda a: str(a) if int_or_none(a) else None}),
 193                 'uploader': ('account', 'username', {lambda x: strip_or_none(x) or None}),
 194                 'uploader_url': ('account', 'avatar_url', {url_or_none}),
 195                 'like_count': ('upvote_count', {int_or_none}),
 196                 'dislike_count': ('downvote_count', {int_or_none}),
 197                 'comment_count': ('comment_count', {int_or_none}),
 198                 'age_limit': ('is_mature', {lambda x: 18 if x else None}),
 199                 'timestamp': (('updated_at', 'created_at'), {parse_iso8601}),
 200                 'release_timestamp': ('created_at', {parse_iso8601}),
 201             }, get_all=False),
 202             **traverse_obj(data, ('media', 0, 'metadata', {
 203                 'title': ('title', {lambda x: strip_or_none(x) or None}),
 204                 'description': ('description', {self.get_description}),
 205                 'duration': ('duration', {float_or_none}),
 206                 'timestamp': (('updated_at', 'created_at'), {parse_iso8601}),
 207                 'release_timestamp': ('created_at', {parse_iso8601}),
 208             }), get_all=False),
 209             'id': video_id,
 210             'formats': formats,
 211             'thumbnail': url_or_none(search('thumbnailUrl')),
 212             'http_headers': {'Accept': '*/*'},
 213         }
 214
 215
 216 class ImgurGalleryBaseIE(ImgurBaseIE):
 217     _GALLERY = True
 218
 219     def _real_extract(self, url):
 220         gallery_id = self._match_id(url)
 221
 222         data = self._call_api('albums', gallery_id, fatal=False, expected_status=404)
 223
 224         info = traverse_obj(data, {
 225             'title': ('title', {lambda x: strip_or_none(x) or None}),
 226             'description': ('description', {self.get_description}),
 227         })
 228
 229         if traverse_obj(data, 'is_album'):
 230
 231             def yield_media_ids():
 232                 for m_id in traverse_obj(data, (
 233                         'media', lambda _, v: v.get('type') == 'video' or v['metadata']['is_animated'],
 234                         'id', {lambda x: str_or_none(x) or None})):
 235                     yield m_id
 236
 237             # if a gallery with exactly one video, apply album metadata to video
 238             media_id = (
 239                 self._GALLERY
 240                 and traverse_obj(data, ('image_count', {lambda c: c == 1}))
 241                 and next(yield_media_ids(), None))
 242
 243             if not media_id:
 244                 result = self.playlist_result(
 245                     map(self._imgur_result, yield_media_ids()), gallery_id)
 246                 result.update(info)
 247                 return result
 248             gallery_id = media_id
 249
 250         result = self._imgur_result(gallery_id)
 251         info['_type'] = 'url_transparent'
 252         result.update(info)
 253         return result
 254
 255
 256 class ImgurGalleryIE(ImgurGalleryBaseIE):
 257     IE_NAME = 'imgur:gallery'
 258     _VALID_URL = r'https?://(?:i\.)?imgur\.com/(?:gallery|(?:t(?:opic)?|r)/[^/?#]+)/(?P<id>[a-zA-Z0-9]+)'
 259
 260     _TESTS = [{
 261         'url': 'http://imgur.com/gallery/Q95ko',
 262         'info_dict': {
 263             'id': 'Q95ko',
 264             'title': 'Adding faces make every GIF better',
 265         },
 266         'playlist_count': 25,
 267         'skip': 'Zoinks! You\'ve taken a wrong turn.',
 268     }, {
 269         # TODO: static images - replace with animated/video gallery
 270         'url': 'http://imgur.com/topic/Aww/ll5Vk',
 271         'only_matching': True,
 272     }, {
 273         'url': 'https://imgur.com/gallery/YcAQlkx',
 274         'add_ies': ['Imgur'],
 275         'info_dict': {
 276             'id': 'YcAQlkx',
 277             'ext': 'mp4',
 278             'title': 'Classic Steve Carell gif...cracks me up everytime....damn the repost downvotes....',
 279             'timestamp': 1358554297,
 280             'upload_date': '20130119',
 281             'uploader_id': '1648642',
 282             'uploader': 'wittyusernamehere',
 283             'release_timestamp': 1358554297,
 284             'thumbnail': 'https://i.imgur.com/YcAQlkxh.jpg',
 285             'release_date': '20130119',
 286             'uploader_url': 'https://i.imgur.com/u3R4I2S_d.png?maxwidth=290&fidelity=grand',
 287             'comment_count': int,
 288             'dislike_count': int,
 289             'like_count': int,
 290         },
 291     }, {
 292         # TODO: static image - replace with animated/video gallery
 293         'url': 'http://imgur.com/topic/Funny/N8rOudd',
 294         'only_matching': True,
 295     }, {
 296         'url': 'http://imgur.com/r/aww/VQcQPhM',
 297         'add_ies': ['Imgur'],
 298         'info_dict': {
 299             'id': 'VQcQPhM',
 300             'ext': 'mp4',
 301             'title': 'The boss is here',
 302             'timestamp': 1476494751,
 303             'upload_date': '20161015',
 304             'uploader_id': '19138530',
 305             'uploader': 'thematrixcam',
 306             'comment_count': int,
 307             'dislike_count': int,
 308             'uploader_url': 'https://i.imgur.com/qCjr5Pi_d.png?maxwidth=290&fidelity=grand',
 309             'release_timestamp': 1476494751,
 310             'like_count': int,
 311             'release_date': '20161015',
 312             'thumbnail': 'https://i.imgur.com/VQcQPhMh.jpg',
 313         },
 314     },
 315         # from https://github.com/ytdl-org/youtube-dl/pull/16674
 316         {
 317         'url': 'https://imgur.com/t/unmuted/6lAn9VQ',
 318         'info_dict': {
 319             'id': '6lAn9VQ',
 320             'title': 'Penguins !',
 321         },
 322         'playlist_count': 3,
 323     }, {
 324         'url': 'https://imgur.com/t/unmuted/kx2uD3C',
 325         'add_ies': ['Imgur'],
 326         'info_dict': {
 327             'id': 'ZVMv45i',
 328             'ext': 'mp4',
 329             'title': 'Intruder',
 330             'timestamp': 1528129683,
 331             'upload_date': '20180604',
 332             'release_timestamp': 1528129683,
 333             'release_date': '20180604',
 334             'like_count': int,
 335             'dislike_count': int,
 336             'comment_count': int,
 337             'duration': 30.03,
 338             'thumbnail': 'https://i.imgur.com/ZVMv45ih.jpg',
 339         },
 340     }, {
 341         'url': 'https://imgur.com/t/unmuted/wXSK0YH',
 342         'add_ies': ['Imgur'],
 343         'info_dict': {
 344             'id': 'JCAP4io',
 345             'ext': 'mp4',
 346             'title': 're:I got the blues$',
 347             'description': 'Luka’s vocal stylings.\n\nFP edit: don’t encourage me. I’ll never stop posting Luka and friends.',
 348             'timestamp': 1527809525,
 349             'upload_date': '20180531',
 350             'like_count': int,
 351             'dislike_count': int,
 352             'duration': 30.03,
 353             'comment_count': int,
 354             'release_timestamp': 1527809525,
 355             'thumbnail': 'https://i.imgur.com/JCAP4ioh.jpg',
 356             'release_date': '20180531',
 357         },
 358     }]
 359
 360
 361 class ImgurAlbumIE(ImgurGalleryBaseIE):
 362     IE_NAME = 'imgur:album'
 363     _VALID_URL = r'https?://(?:i\.)?imgur\.com/a/(?P<id>[a-zA-Z0-9]+)'
 364     _GALLERY = False
 365     _TESTS = [{
 366         # TODO: only static images - replace with animated/video gallery
 367         'url': 'http://imgur.com/a/j6Orj',
 368         'only_matching': True,
 369     },
 370         # from https://github.com/ytdl-org/youtube-dl/pull/21693
 371         {
 372         'url': 'https://imgur.com/a/iX265HX',
 373         'info_dict': {
 374             'id': 'iX265HX',
 375             'title': 'enen-no-shouboutai'
 376         },
 377         'playlist_count': 2,
 378     }, {
 379         'url': 'https://imgur.com/a/8pih2Ed',
 380         'info_dict': {
 381             'id': '8pih2Ed'
 382         },
 383         'playlist_mincount': 1,
 384     }]