yt_dlp/extractor/instagram.py

   1 # coding: utf-8
   2
   3 import itertools
   4 import hashlib
   5 import json
   6 import re
   7 import time
   8
   9 from .common import InfoExtractor
  10 from ..compat import (
  11     compat_HTTPError,
  12 )
  13 from ..utils import (
  14     ExtractorError,
  15     format_field,
  16     float_or_none,
  17     get_element_by_attribute,
  18     int_or_none,
  19     lowercase_escape,
  20     str_or_none,
  21     str_to_int,
  22     traverse_obj,
  23     url_or_none,
  24     urlencode_postdata,
  25 )
  26
  27
  28 class InstagramBaseIE(InfoExtractor):
  29     _NETRC_MACHINE = 'instagram'
  30     _IS_LOGGED_IN = False
  31
  32     def _login(self):
  33         username, password = self._get_login_info()
  34         if username is None or self._IS_LOGGED_IN:
  35             return
  36
  37         login_webpage = self._download_webpage(
  38             'https://www.instagram.com/accounts/login/', None,
  39             note='Downloading login webpage', errnote='Failed to download login webpage')
  40
  41         shared_data = self._parse_json(
  42             self._search_regex(
  43                 r'window\._sharedData\s*=\s*({.+?});',
  44                 login_webpage, 'shared data', default='{}'),
  45             None)
  46
  47         login = self._download_json('https://www.instagram.com/accounts/login/ajax/', None, note='Logging in', headers={
  48             'Accept': '*/*',
  49             'X-IG-App-ID': '936619743392459',
  50             'X-ASBD-ID': '198387',
  51             'X-IG-WWW-Claim': '0',
  52             'X-Requested-With': 'XMLHttpRequest',
  53             'X-CSRFToken': shared_data['config']['csrf_token'],
  54             'X-Instagram-AJAX': shared_data['rollout_hash'],
  55             'Referer': 'https://www.instagram.com/',
  56         }, data=urlencode_postdata({
  57             'enc_password': f'#PWD_INSTAGRAM_BROWSER:0:{int(time.time())}:{password}',
  58             'username': username,
  59             'queryParams': '{}',
  60             'optIntoOneTap': 'false',
  61             'stopDeletionNonce': '',
  62             'trustedDeviceRecords': '{}',
  63         }))
  64
  65         if not login.get('authenticated'):
  66             if login.get('message'):
  67                 raise ExtractorError(f'Unable to login: {login["message"]}')
  68             elif login.get('user'):
  69                 raise ExtractorError('Unable to login: Sorry, your password was incorrect. Please double-check your password.', expected=True)
  70             elif login.get('user') is False:
  71                 raise ExtractorError('Unable to login: The username you entered doesn\'t belong to an account. Please check your username and try again.', expected=True)
  72             raise ExtractorError('Unable to login')
  73         InstagramBaseIE._IS_LOGGED_IN = True
  74
  75     def _real_initialize(self):
  76         self._login()
  77
  78     def _get_count(self, media, kind, *keys):
  79         return traverse_obj(
  80             media, (kind, 'count'), *((f'edge_media_{key}', 'count') for key in keys),
  81             expected_type=int_or_none)
  82
  83     def _get_dimension(self, name, media, webpage=None):
  84         return (
  85             traverse_obj(media, ('dimensions', name), expected_type=int_or_none)
  86             or int_or_none(self._html_search_meta(
  87                 (f'og:video:{name}', f'video:{name}'), webpage or '', default=None)))
  88
  89     def _extract_nodes(self, nodes, is_direct=False):
  90         for idx, node in enumerate(nodes, start=1):
  91             if node.get('__typename') != 'GraphVideo' and node.get('is_video') is not True:
  92                 continue
  93
  94             video_id = node.get('shortcode')
  95
  96             if is_direct:
  97                 info = {
  98                     'id': video_id or node['id'],
  99                     'url': node.get('video_url'),
 100                     'width': self._get_dimension('width', node),
 101                     'height': self._get_dimension('height', node),
 102                     'http_headers': {
 103                         'Referer': 'https://www.instagram.com/',
 104                     }
 105                 }
 106             elif not video_id:
 107                 continue
 108             else:
 109                 info = {
 110                     '_type': 'url',
 111                     'ie_key': 'Instagram',
 112                     'id': video_id,
 113                     'url': f'https://instagram.com/p/{video_id}',
 114                 }
 115
 116             yield {
 117                 **info,
 118                 'title': node.get('title') or (f'Video {idx}' if is_direct else None),
 119                 'description': traverse_obj(
 120                     node, ('edge_media_to_caption', 'edges', 0, 'node', 'text'), expected_type=str),
 121                 'thumbnail': traverse_obj(
 122                     node, 'display_url', 'thumbnail_src', 'display_src', expected_type=url_or_none),
 123                 'duration': float_or_none(node.get('video_duration')),
 124                 'timestamp': int_or_none(node.get('taken_at_timestamp')),
 125                 'view_count': int_or_none(node.get('video_view_count')),
 126                 'comment_count': self._get_count(node, 'comments', 'preview_comment', 'to_comment', 'to_parent_comment'),
 127                 'like_count': self._get_count(node, 'likes', 'preview_like'),
 128             }
 129
 130     def _extract_product_media(self, product_media):
 131         media_id = product_media.get('code') or product_media.get('id')
 132         vcodec = product_media.get('video_codec')
 133         dash_manifest_raw = product_media.get('video_dash_manifest')
 134         videos_list = product_media.get('video_versions')
 135         if not (dash_manifest_raw or videos_list):
 136             return {}
 137
 138         formats = [{
 139             'format_id': format.get('id'),
 140             'url': format.get('url'),
 141             'width': format.get('width'),
 142             'height': format.get('height'),
 143             'vcodec': vcodec,
 144         } for format in videos_list or []]
 145         if dash_manifest_raw:
 146             formats.extend(self._parse_mpd_formats(self._parse_xml(dash_manifest_raw, media_id), mpd_id='dash'))
 147         self._sort_formats(formats)
 148
 149         thumbnails = [{
 150             'url': thumbnail.get('url'),
 151             'width': thumbnail.get('width'),
 152             'height': thumbnail.get('height')
 153         } for thumbnail in traverse_obj(product_media, ('image_versions2', 'candidates')) or []]
 154         return {
 155             'id': media_id,
 156             'duration': float_or_none(product_media.get('video_duration')),
 157             'formats': formats,
 158             'thumbnails': thumbnails
 159         }
 160
 161     def _extract_product(self, product_info):
 162         if isinstance(product_info, list):
 163             product_info = product_info[0]
 164
 165         user_info = product_info.get('user') or {}
 166         info_dict = {
 167             'id': product_info.get('code') or product_info.get('id'),
 168             'title': product_info.get('title') or f'Video by {user_info.get("username")}',
 169             'description': traverse_obj(product_info, ('caption', 'text'), expected_type=str_or_none),
 170             'timestamp': int_or_none(product_info.get('taken_at')),
 171             'channel': user_info.get('username'),
 172             'uploader': user_info.get('full_name'),
 173             'uploader_id': str_or_none(user_info.get('pk')),
 174             'view_count': int_or_none(product_info.get('view_count')),
 175             'like_count': int_or_none(product_info.get('like_count')),
 176             'comment_count': int_or_none(product_info.get('comment_count')),
 177             'http_headers': {
 178                 'Referer': 'https://www.instagram.com/',
 179             }
 180         }
 181         carousel_media = product_info.get('carousel_media')
 182         if carousel_media:
 183             return {
 184                 '_type': 'playlist',
 185                 **info_dict,
 186                 'title': f'Post by {user_info.get("username")}',
 187                 'entries': [{
 188                     **info_dict,
 189                     **self._extract_product_media(product_media),
 190                 } for product_media in carousel_media],
 191             }
 192
 193         return {
 194             **info_dict,
 195             **self._extract_product_media(product_info)
 196         }
 197
 198
 199 class InstagramIOSIE(InfoExtractor):
 200     IE_DESC = 'IOS instagram:// URL'
 201     _VALID_URL = r'instagram://media\?id=(?P<id>[\d_]+)'
 202     _TESTS = [{
 203         'url': 'instagram://media?id=482584233761418119',
 204         'md5': '0d2da106a9d2631273e192b372806516',
 205         'info_dict': {
 206             'id': 'aye83DjauH',
 207             'ext': 'mp4',
 208             'title': 'Video by naomipq',
 209             'description': 'md5:1f17f0ab29bd6fe2bfad705f58de3cb8',
 210             'thumbnail': r're:^https?://.*\.jpg',
 211             'duration': 0,
 212             'timestamp': 1371748545,
 213             'upload_date': '20130620',
 214             'uploader_id': 'naomipq',
 215             'uploader': 'B E A U T Y  F O R  A S H E S',
 216             'like_count': int,
 217             'comment_count': int,
 218             'comments': list,
 219         },
 220         'add_ie': ['Instagram']
 221     }]
 222
 223     def _get_id(self, id):
 224         """Source: https://stackoverflow.com/questions/24437823/getting-instagram-post-url-from-media-id"""
 225         chrs = 'ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789-_'
 226         media_id = int(id.split('_')[0])
 227         shortened_id = ''
 228         while media_id > 0:
 229             r = media_id % 64
 230             media_id = (media_id - r) // 64
 231             shortened_id = chrs[r] + shortened_id
 232         return shortened_id
 233
 234     def _real_extract(self, url):
 235         return {
 236             '_type': 'url_transparent',
 237             'url': f'http://instagram.com/tv/{self._get_id(self._match_id(url))}/',
 238             'ie_key': 'Instagram',
 239         }
 240
 241
 242 class InstagramIE(InstagramBaseIE):
 243     _VALID_URL = r'(?P<url>https?://(?:www\.)?instagram\.com(?:/[^/]+)?/(?:p|tv|reel)/(?P<id>[^/?#&]+))'
 244     _TESTS = [{
 245         'url': 'https://instagram.com/p/aye83DjauH/?foo=bar#abc',
 246         'md5': '0d2da106a9d2631273e192b372806516',
 247         'info_dict': {
 248             'id': 'aye83DjauH',
 249             'ext': 'mp4',
 250             'title': 'Video by naomipq',
 251             'description': 'md5:1f17f0ab29bd6fe2bfad705f58de3cb8',
 252             'thumbnail': r're:^https?://.*\.jpg',
 253             'duration': 0,
 254             'timestamp': 1371748545,
 255             'upload_date': '20130620',
 256             'uploader_id': '2815873',
 257             'uploader': 'B E A U T Y  F O R  A S H E S',
 258             'channel': 'naomipq',
 259             'like_count': int,
 260             'comment_count': int,
 261             'comments': list,
 262         },
 263     }, {
 264         # missing description
 265         'url': 'https://www.instagram.com/p/BA-pQFBG8HZ/?taken-by=britneyspears',
 266         'info_dict': {
 267             'id': 'BA-pQFBG8HZ',
 268             'ext': 'mp4',
 269             'title': 'Video by britneyspears',
 270             'thumbnail': r're:^https?://.*\.jpg',
 271             'duration': 0,
 272             'timestamp': 1453760977,
 273             'upload_date': '20160125',
 274             'uploader_id': '12246775',
 275             'uploader': 'Britney Spears',
 276             'channel': 'britneyspears',
 277             'like_count': int,
 278             'comment_count': int,
 279             'comments': list,
 280         },
 281         'params': {
 282             'skip_download': True,
 283         },
 284     }, {
 285         # multi video post
 286         'url': 'https://www.instagram.com/p/BQ0eAlwhDrw/',
 287         'playlist': [{
 288             'info_dict': {
 289                 'id': 'BQ0dSaohpPW',
 290                 'ext': 'mp4',
 291                 'title': 'Video 1',
 292             },
 293         }, {
 294             'info_dict': {
 295                 'id': 'BQ0dTpOhuHT',
 296                 'ext': 'mp4',
 297                 'title': 'Video 2',
 298             },
 299         }, {
 300             'info_dict': {
 301                 'id': 'BQ0dT7RBFeF',
 302                 'ext': 'mp4',
 303                 'title': 'Video 3',
 304             },
 305         }],
 306         'info_dict': {
 307             'id': 'BQ0eAlwhDrw',
 308             'title': 'Post by instagram',
 309             'description': 'md5:0f9203fc6a2ce4d228da5754bcf54957',
 310         },
 311     }, {
 312         # IGTV
 313         'url': 'https://www.instagram.com/tv/BkfuX9UB-eK/',
 314         'info_dict': {
 315             'id': 'BkfuX9UB-eK',
 316             'ext': 'mp4',
 317             'title': 'Fingerboarding Tricks with @cass.fb',
 318             'thumbnail': r're:^https?://.*\.jpg',
 319             'duration': 53.83,
 320             'timestamp': 1530032919,
 321             'upload_date': '20180626',
 322             'uploader_id': '25025320',
 323             'uploader': 'Instagram',
 324             'channel': 'instagram',
 325             'like_count': int,
 326             'comment_count': int,
 327             'comments': list,
 328             'description': 'Meet Cass Hirst (@cass.fb), a fingerboarding pro who can perform tiny ollies and kickflips while blindfolded.',
 329         }
 330     }, {
 331         'url': 'https://instagram.com/p/-Cmh1cukG2/',
 332         'only_matching': True,
 333     }, {
 334         'url': 'http://instagram.com/p/9o6LshA7zy/embed/',
 335         'only_matching': True,
 336     }, {
 337         'url': 'https://www.instagram.com/tv/aye83DjauH/',
 338         'only_matching': True,
 339     }, {
 340         'url': 'https://www.instagram.com/reel/CDUMkliABpa/',
 341         'only_matching': True,
 342     }, {
 343         'url': 'https://www.instagram.com/marvelskies.fc/reel/CWqAgUZgCku/',
 344         'only_matching': True,
 345     }]
 346
 347     @staticmethod
 348     def _extract_embed_url(webpage):
 349         mobj = re.search(
 350             r'<iframe[^>]+src=(["\'])(?P<url>(?:https?:)?//(?:www\.)?instagram\.com/p/[^/]+/embed.*?)\1',
 351             webpage)
 352         if mobj:
 353             return mobj.group('url')
 354
 355         blockquote_el = get_element_by_attribute(
 356             'class', 'instagram-media', webpage)
 357         if blockquote_el is None:
 358             return
 359
 360         mobj = re.search(
 361             r'<a[^>]+href=([\'"])(?P<link>[^\'"]+)\1', blockquote_el)
 362         if mobj:
 363             return mobj.group('link')
 364
 365     def _real_extract(self, url):
 366         video_id, url = self._match_valid_url(url).group('id', 'url')
 367         webpage, urlh = self._download_webpage_handle(url, video_id)
 368         if 'www.instagram.com/accounts/login' in urlh.geturl():
 369             self.report_warning('Main webpage is locked behind the login page. '
 370                                 'Retrying with embed webpage (Note that some metadata might be missing)')
 371             webpage = self._download_webpage(
 372                 'https://www.instagram.com/p/%s/embed/' % video_id, video_id, note='Downloading embed webpage')
 373
 374         shared_data = self._parse_json(
 375             self._search_regex(
 376                 r'window\._sharedData\s*=\s*({.+?});',
 377                 webpage, 'shared data', default='{}'),
 378             video_id, fatal=False)
 379         media = traverse_obj(
 380             shared_data,
 381             ('entry_data', 'PostPage', 0, 'graphql', 'shortcode_media'),
 382             ('entry_data', 'PostPage', 0, 'media'),
 383             expected_type=dict)
 384
 385         # _sharedData.entry_data.PostPage is empty when authenticated (see
 386         # https://github.com/ytdl-org/youtube-dl/pull/22880)
 387         if not media:
 388             additional_data = self._parse_json(
 389                 self._search_regex(
 390                     r'window\.__additionalDataLoaded\s*\(\s*[^,]+,\s*({.+?})\s*\);',
 391                     webpage, 'additional data', default='{}'),
 392                 video_id, fatal=False)
 393             product_item = traverse_obj(additional_data, ('items', 0), expected_type=dict)
 394             if product_item:
 395                 return self._extract_product(product_item)
 396             media = traverse_obj(additional_data, ('graphql', 'shortcode_media'), 'shortcode_media', expected_type=dict) or {}
 397
 398         if not media and 'www.instagram.com/accounts/login' in urlh.geturl():
 399             self.raise_login_required('You need to log in to access this content')
 400
 401         username = traverse_obj(media, ('owner', 'username')) or self._search_regex(
 402             r'"owner"\s*:\s*{\s*"username"\s*:\s*"(.+?)"', webpage, 'username', fatal=False)
 403
 404         description = (
 405             traverse_obj(media, ('edge_media_to_caption', 'edges', 0, 'node', 'text'), expected_type=str)
 406             or media.get('caption'))
 407         if not description:
 408             description = self._search_regex(
 409                 r'"caption"\s*:\s*"(.+?)"', webpage, 'description', default=None)
 410             if description is not None:
 411                 description = lowercase_escape(description)
 412
 413         video_url = media.get('video_url')
 414         if not video_url:
 415             nodes = traverse_obj(media, ('edge_sidecar_to_children', 'edges', ..., 'node'), expected_type=dict) or []
 416             if nodes:
 417                 return self.playlist_result(
 418                     self._extract_nodes(nodes, True), video_id,
 419                     format_field(username, template='Post by %s'), description)
 420
 421             video_url = self._og_search_video_url(webpage, secure=False)
 422
 423         formats = [{
 424             'url': video_url,
 425             'width': self._get_dimension('width', media, webpage),
 426             'height': self._get_dimension('height', media, webpage),
 427         }]
 428         dash = traverse_obj(media, ('dash_info', 'video_dash_manifest'))
 429         if dash:
 430             formats.extend(self._parse_mpd_formats(self._parse_xml(dash, video_id), mpd_id='dash'))
 431         self._sort_formats(formats)
 432
 433         comment_data = traverse_obj(media, ('edge_media_to_parent_comment', 'edges'))
 434         comments = [{
 435             'author': traverse_obj(comment_dict, ('node', 'owner', 'username')),
 436             'author_id': traverse_obj(comment_dict, ('node', 'owner', 'id')),
 437             'id': traverse_obj(comment_dict, ('node', 'id')),
 438             'text': traverse_obj(comment_dict, ('node', 'text')),
 439             'timestamp': traverse_obj(comment_dict, ('node', 'created_at'), expected_type=int_or_none),
 440         } for comment_dict in comment_data] if comment_data else None
 441
 442         display_resources = (
 443             media.get('display_resources')
 444             or [{'src': media.get(key)} for key in ('display_src', 'display_url')]
 445             or [{'src': self._og_search_thumbnail(webpage)}])
 446         thumbnails = [{
 447             'url': thumbnail['src'],
 448             'width': thumbnail.get('config_width'),
 449             'height': thumbnail.get('config_height'),
 450         } for thumbnail in display_resources if thumbnail.get('src')]
 451
 452         return {
 453             'id': video_id,
 454             'formats': formats,
 455             'title': media.get('title') or 'Video by %s' % username,
 456             'description': description,
 457             'duration': float_or_none(media.get('video_duration')),
 458             'timestamp': traverse_obj(media, 'taken_at_timestamp', 'date', expected_type=int_or_none),
 459             'uploader_id': traverse_obj(media, ('owner', 'id')),
 460             'uploader': traverse_obj(media, ('owner', 'full_name')),
 461             'channel': username,
 462             'like_count': self._get_count(media, 'likes', 'preview_like') or str_to_int(self._search_regex(
 463                 r'data-log-event="likeCountClick"[^>]*>[^\d]*([\d,\.]+)', webpage, 'like count', fatal=False)),
 464             'comment_count': self._get_count(media, 'comments', 'preview_comment', 'to_comment', 'to_parent_comment'),
 465             'comments': comments,
 466             'thumbnails': thumbnails,
 467             'http_headers': {
 468                 'Referer': 'https://www.instagram.com/',
 469             }
 470         }
 471
 472
 473 class InstagramPlaylistBaseIE(InstagramBaseIE):
 474     _gis_tmpl = None  # used to cache GIS request type
 475
 476     def _parse_graphql(self, webpage, item_id):
 477         # Reads a webpage and returns its GraphQL data.
 478         return self._parse_json(
 479             self._search_regex(
 480                 r'sharedData\s*=\s*({.+?})\s*;\s*[<\n]', webpage, 'data'),
 481             item_id)
 482
 483     def _extract_graphql(self, data, url):
 484         # Parses GraphQL queries containing videos and generates a playlist.
 485         uploader_id = self._match_id(url)
 486         csrf_token = data['config']['csrf_token']
 487         rhx_gis = data.get('rhx_gis') or '3c7ca9dcefcf966d11dacf1f151335e8'
 488
 489         cursor = ''
 490         for page_num in itertools.count(1):
 491             variables = {
 492                 'first': 12,
 493                 'after': cursor,
 494             }
 495             variables.update(self._query_vars_for(data))
 496             variables = json.dumps(variables)
 497
 498             if self._gis_tmpl:
 499                 gis_tmpls = [self._gis_tmpl]
 500             else:
 501                 gis_tmpls = [
 502                     '%s' % rhx_gis,
 503                     '',
 504                     '%s:%s' % (rhx_gis, csrf_token),
 505                     '%s:%s:%s' % (rhx_gis, csrf_token, self.get_param('http_headers')['User-Agent']),
 506                 ]
 507
 508             # try all of the ways to generate a GIS query, and not only use the
 509             # first one that works, but cache it for future requests
 510             for gis_tmpl in gis_tmpls:
 511                 try:
 512                     json_data = self._download_json(
 513                         'https://www.instagram.com/graphql/query/', uploader_id,
 514                         'Downloading JSON page %d' % page_num, headers={
 515                             'X-Requested-With': 'XMLHttpRequest',
 516                             'X-Instagram-GIS': hashlib.md5(
 517                                 ('%s:%s' % (gis_tmpl, variables)).encode('utf-8')).hexdigest(),
 518                         }, query={
 519                             'query_hash': self._QUERY_HASH,
 520                             'variables': variables,
 521                         })
 522                     media = self._parse_timeline_from(json_data)
 523                     self._gis_tmpl = gis_tmpl
 524                     break
 525                 except ExtractorError as e:
 526                     # if it's an error caused by a bad query, and there are
 527                     # more GIS templates to try, ignore it and keep trying
 528                     if isinstance(e.cause, compat_HTTPError) and e.cause.code == 403:
 529                         if gis_tmpl != gis_tmpls[-1]:
 530                             continue
 531                     raise
 532
 533             nodes = traverse_obj(media, ('edges', ..., 'node'), expected_type=dict) or []
 534             if not nodes:
 535                 break
 536             yield from self._extract_nodes(nodes)
 537
 538             has_next_page = traverse_obj(media, ('page_info', 'has_next_page'))
 539             cursor = traverse_obj(media, ('page_info', 'end_cursor'), expected_type=str)
 540             if not has_next_page or not cursor:
 541                 break
 542
 543     def _real_extract(self, url):
 544         user_or_tag = self._match_id(url)
 545         webpage = self._download_webpage(url, user_or_tag)
 546         data = self._parse_graphql(webpage, user_or_tag)
 547
 548         self._set_cookie('instagram.com', 'ig_pr', '1')
 549
 550         return self.playlist_result(
 551             self._extract_graphql(data, url), user_or_tag, user_or_tag)
 552
 553
 554 class InstagramUserIE(InstagramPlaylistBaseIE):
 555     _VALID_URL = r'https?://(?:www\.)?instagram\.com/(?P<id>[^/]{2,})/?(?:$|[?#])'
 556     IE_DESC = 'Instagram user profile'
 557     IE_NAME = 'instagram:user'
 558     _TESTS = [{
 559         'url': 'https://instagram.com/porsche',
 560         'info_dict': {
 561             'id': 'porsche',
 562             'title': 'porsche',
 563         },
 564         'playlist_count': 5,
 565         'params': {
 566             'extract_flat': True,
 567             'skip_download': True,
 568             'playlistend': 5,
 569         }
 570     }]
 571
 572     _QUERY_HASH = '42323d64886122307be10013ad2dcc44',
 573
 574     @staticmethod
 575     def _parse_timeline_from(data):
 576         # extracts the media timeline data from a GraphQL result
 577         return data['data']['user']['edge_owner_to_timeline_media']
 578
 579     @staticmethod
 580     def _query_vars_for(data):
 581         # returns a dictionary of variables to add to the timeline query based
 582         # on the GraphQL of the original page
 583         return {
 584             'id': data['entry_data']['ProfilePage'][0]['graphql']['user']['id']
 585         }
 586
 587
 588 class InstagramTagIE(InstagramPlaylistBaseIE):
 589     _VALID_URL = r'https?://(?:www\.)?instagram\.com/explore/tags/(?P<id>[^/]+)'
 590     IE_DESC = 'Instagram hashtag search URLs'
 591     IE_NAME = 'instagram:tag'
 592     _TESTS = [{
 593         'url': 'https://instagram.com/explore/tags/lolcats',
 594         'info_dict': {
 595             'id': 'lolcats',
 596             'title': 'lolcats',
 597         },
 598         'playlist_count': 50,
 599         'params': {
 600             'extract_flat': True,
 601             'skip_download': True,
 602             'playlistend': 50,
 603         }
 604     }]
 605
 606     _QUERY_HASH = 'f92f56d47dc7a55b606908374b43a314',
 607
 608     @staticmethod
 609     def _parse_timeline_from(data):
 610         # extracts the media timeline data from a GraphQL result
 611         return data['data']['hashtag']['edge_hashtag_to_media']
 612
 613     @staticmethod
 614     def _query_vars_for(data):
 615         # returns a dictionary of variables to add to the timeline query based
 616         # on the GraphQL of the original page
 617         return {
 618             'tag_name':
 619                 data['entry_data']['TagPage'][0]['graphql']['hashtag']['name']
 620         }
 621
 622
 623 class InstagramStoryIE(InstagramBaseIE):
 624     _VALID_URL = r'https?://(?:www\.)?instagram\.com/stories/(?P<user>[^/]+)/(?P<id>\d+)'
 625     IE_NAME = 'instagram:story'
 626
 627     _TESTS = [{
 628         'url': 'https://www.instagram.com/stories/highlights/18090946048123978/',
 629         'info_dict': {
 630             'id': '18090946048123978',
 631             'title': 'Rare',
 632         },
 633         'playlist_mincount': 50
 634     }]
 635
 636     def _real_extract(self, url):
 637         username, story_id = self._match_valid_url(url).groups()
 638
 639         story_info_url = f'{username}/{story_id}/?__a=1' if username == 'highlights' else f'{username}/?__a=1'
 640         story_info = self._download_json(f'https://www.instagram.com/stories/{story_info_url}', story_id, headers={
 641             'X-IG-App-ID': 936619743392459,
 642             'X-ASBD-ID': 198387,
 643             'X-IG-WWW-Claim': 0,
 644             'X-Requested-With': 'XMLHttpRequest',
 645             'Referer': url,
 646         })
 647         user_id = story_info['user']['id']
 648         highlight_title = traverse_obj(story_info, ('highlight', 'title'))
 649
 650         story_info_url = user_id if username != 'highlights' else f'highlight:{story_id}'
 651         videos = self._download_json(f'https://i.instagram.com/api/v1/feed/reels_media/?reel_ids={story_info_url}', story_id, headers={
 652             'X-IG-App-ID': 936619743392459,
 653             'X-ASBD-ID': 198387,
 654             'X-IG-WWW-Claim': 0,
 655         })['reels']
 656
 657         full_name = traverse_obj(videos, ('user', 'full_name'))
 658
 659         user_info = {}
 660         if not (username and username != 'highlights' and full_name):
 661             user_info = self._download_json(
 662                 f'https://i.instagram.com/api/v1/users/{user_id}/info/', story_id, headers={
 663                     'User-Agent': 'Mozilla/5.0 (Linux; Android 11; SM-A505F Build/RP1A.200720.012; wv) AppleWebKit/537.36 (KHTML, like Gecko) Version/4.0 Chrome/96.0.4664.45 Mobile Safari/537.36 Instagram 214.1.0.29.120 Android (30/11; 450dpi; 1080x2122; samsung; SM-A505F; a50; exynos9610; en_US; 333717274)',
 664                 }, note='Downloading user info')
 665
 666         username = traverse_obj(user_info, ('user', 'username')) or username
 667         full_name = traverse_obj(user_info, ('user', 'full_name')) or full_name
 668
 669         highlights = traverse_obj(videos, (f'highlight:{story_id}', 'items'), (str(user_id), 'items'))
 670         return self.playlist_result([{
 671             **self._extract_product(highlight),
 672             'title': f'Story by {username}',
 673             'uploader': full_name,
 674             'uploader_id': user_id,
 675         } for highlight in highlights], playlist_id=story_id, playlist_title=highlight_title)