yt_dlp/extractor/instagram.py

   1 import hashlib
   2 import itertools
   3 import json
   4 import re
   5 import time
   6 import urllib.error
   7
   8 from .common import InfoExtractor
   9 from ..utils import (
  10     ExtractorError,
  11     decode_base_n,
  12     encode_base_n,
  13     float_or_none,
  14     format_field,
  15     get_element_by_attribute,
  16     int_or_none,
  17     lowercase_escape,
  18     str_or_none,
  19     str_to_int,
  20     traverse_obj,
  21     url_or_none,
  22     urlencode_postdata,
  23 )
  24
  25 _ENCODING_CHARS = 'ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789-_'
  26
  27
  28 def _pk_to_id(id):
  29     """Source: https://stackoverflow.com/questions/24437823/getting-instagram-post-url-from-media-id"""
  30     return encode_base_n(int(id.split('_')[0]), table=_ENCODING_CHARS)
  31
  32
  33 def _id_to_pk(shortcode):
  34     """Covert a shortcode to a numeric value"""
  35     return decode_base_n(shortcode[:11], table=_ENCODING_CHARS)
  36
  37
  38 class InstagramBaseIE(InfoExtractor):
  39     _NETRC_MACHINE = 'instagram'
  40     _IS_LOGGED_IN = False
  41
  42     _API_BASE_URL = 'https://i.instagram.com/api/v1'
  43     _LOGIN_URL = 'https://www.instagram.com/accounts/login'
  44     _API_HEADERS = {
  45         'X-IG-App-ID': '936619743392459',
  46         'X-ASBD-ID': '198387',
  47         'X-IG-WWW-Claim': '0',
  48         'Origin': 'https://www.instagram.com',
  49         'Accept': '*/*',
  50         'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/104.0.0.0 Safari/537.36',
  51     }
  52
  53     def _perform_login(self, username, password):
  54         if self._IS_LOGGED_IN:
  55             return
  56
  57         login_webpage = self._download_webpage(
  58             self._LOGIN_URL, None, note='Downloading login webpage', errnote='Failed to download login webpage')
  59
  60         shared_data = self._parse_json(self._search_regex(
  61             r'window\._sharedData\s*=\s*({.+?});', login_webpage, 'shared data', default='{}'), None)
  62
  63         login = self._download_json(
  64             f'{self._LOGIN_URL}/ajax/', None, note='Logging in', headers={
  65                 **self._API_HEADERS,
  66                 'X-Requested-With': 'XMLHttpRequest',
  67                 'X-CSRFToken': shared_data['config']['csrf_token'],
  68                 'X-Instagram-AJAX': shared_data['rollout_hash'],
  69                 'Referer': 'https://www.instagram.com/',
  70             }, data=urlencode_postdata({
  71                 'enc_password': f'#PWD_INSTAGRAM_BROWSER:0:{int(time.time())}:{password}',
  72                 'username': username,
  73                 'queryParams': '{}',
  74                 'optIntoOneTap': 'false',
  75                 'stopDeletionNonce': '',
  76                 'trustedDeviceRecords': '{}',
  77             }))
  78
  79         if not login.get('authenticated'):
  80             if login.get('message'):
  81                 raise ExtractorError(f'Unable to login: {login["message"]}')
  82             elif login.get('user'):
  83                 raise ExtractorError('Unable to login: Sorry, your password was incorrect. Please double-check your password.', expected=True)
  84             elif login.get('user') is False:
  85                 raise ExtractorError('Unable to login: The username you entered doesn\'t belong to an account. Please check your username and try again.', expected=True)
  86             raise ExtractorError('Unable to login')
  87         InstagramBaseIE._IS_LOGGED_IN = True
  88
  89     def _get_count(self, media, kind, *keys):
  90         return traverse_obj(
  91             media, (kind, 'count'), *((f'edge_media_{key}', 'count') for key in keys),
  92             expected_type=int_or_none)
  93
  94     def _get_dimension(self, name, media, webpage=None):
  95         return (
  96             traverse_obj(media, ('dimensions', name), expected_type=int_or_none)
  97             or int_or_none(self._html_search_meta(
  98                 (f'og:video:{name}', f'video:{name}'), webpage or '', default=None)))
  99
 100     def _extract_nodes(self, nodes, is_direct=False):
 101         for idx, node in enumerate(nodes, start=1):
 102             if node.get('__typename') != 'GraphVideo' and node.get('is_video') is not True:
 103                 continue
 104
 105             video_id = node.get('shortcode')
 106
 107             if is_direct:
 108                 info = {
 109                     'id': video_id or node['id'],
 110                     'url': node.get('video_url'),
 111                     'width': self._get_dimension('width', node),
 112                     'height': self._get_dimension('height', node),
 113                     'http_headers': {
 114                         'Referer': 'https://www.instagram.com/',
 115                     }
 116                 }
 117             elif not video_id:
 118                 continue
 119             else:
 120                 info = {
 121                     '_type': 'url',
 122                     'ie_key': 'Instagram',
 123                     'id': video_id,
 124                     'url': f'https://instagram.com/p/{video_id}',
 125                 }
 126
 127             yield {
 128                 **info,
 129                 'title': node.get('title') or (f'Video {idx}' if is_direct else None),
 130                 'description': traverse_obj(
 131                     node, ('edge_media_to_caption', 'edges', 0, 'node', 'text'), expected_type=str),
 132                 'thumbnail': traverse_obj(
 133                     node, 'display_url', 'thumbnail_src', 'display_src', expected_type=url_or_none),
 134                 'duration': float_or_none(node.get('video_duration')),
 135                 'timestamp': int_or_none(node.get('taken_at_timestamp')),
 136                 'view_count': int_or_none(node.get('video_view_count')),
 137                 'comment_count': self._get_count(node, 'comments', 'preview_comment', 'to_comment', 'to_parent_comment'),
 138                 'like_count': self._get_count(node, 'likes', 'preview_like'),
 139             }
 140
 141     def _extract_product_media(self, product_media):
 142         media_id = product_media.get('code') or _pk_to_id(product_media.get('pk'))
 143         vcodec = product_media.get('video_codec')
 144         dash_manifest_raw = product_media.get('video_dash_manifest')
 145         videos_list = product_media.get('video_versions')
 146         if not (dash_manifest_raw or videos_list):
 147             return {}
 148
 149         formats = [{
 150             'format_id': format.get('id'),
 151             'url': format.get('url'),
 152             'width': format.get('width'),
 153             'height': format.get('height'),
 154             'vcodec': vcodec,
 155         } for format in videos_list or []]
 156         if dash_manifest_raw:
 157             formats.extend(self._parse_mpd_formats(self._parse_xml(dash_manifest_raw, media_id), mpd_id='dash'))
 158         self._sort_formats(formats)
 159
 160         thumbnails = [{
 161             'url': thumbnail.get('url'),
 162             'width': thumbnail.get('width'),
 163             'height': thumbnail.get('height')
 164         } for thumbnail in traverse_obj(product_media, ('image_versions2', 'candidates')) or []]
 165         return {
 166             'id': media_id,
 167             'duration': float_or_none(product_media.get('video_duration')),
 168             'formats': formats,
 169             'thumbnails': thumbnails
 170         }
 171
 172     def _extract_product(self, product_info):
 173         if isinstance(product_info, list):
 174             product_info = product_info[0]
 175
 176         comment_data = traverse_obj(product_info, ('edge_media_to_parent_comment', 'edges'))
 177         comments = [{
 178             'author': traverse_obj(comment_dict, ('node', 'owner', 'username')),
 179             'author_id': traverse_obj(comment_dict, ('node', 'owner', 'id')),
 180             'id': traverse_obj(comment_dict, ('node', 'id')),
 181             'text': traverse_obj(comment_dict, ('node', 'text')),
 182             'timestamp': traverse_obj(comment_dict, ('node', 'created_at'), expected_type=int_or_none),
 183         } for comment_dict in comment_data] if comment_data else None
 184
 185         user_info = product_info.get('user') or {}
 186         info_dict = {
 187             'id': product_info.get('code') or _pk_to_id(product_info.get('pk')),
 188             'title': product_info.get('title') or f'Video by {user_info.get("username")}',
 189             'description': traverse_obj(product_info, ('caption', 'text'), expected_type=str_or_none),
 190             'timestamp': int_or_none(product_info.get('taken_at')),
 191             'channel': user_info.get('username'),
 192             'uploader': user_info.get('full_name'),
 193             'uploader_id': str_or_none(user_info.get('pk')),
 194             'view_count': int_or_none(product_info.get('view_count')),
 195             'like_count': int_or_none(product_info.get('like_count')),
 196             'comment_count': int_or_none(product_info.get('comment_count')),
 197             'comments': comments,
 198             'http_headers': {
 199                 'Referer': 'https://www.instagram.com/',
 200             }
 201         }
 202         carousel_media = product_info.get('carousel_media')
 203         if carousel_media:
 204             return {
 205                 '_type': 'playlist',
 206                 **info_dict,
 207                 'title': f'Post by {user_info.get("username")}',
 208                 'entries': [{
 209                     **info_dict,
 210                     **self._extract_product_media(product_media),
 211                 } for product_media in carousel_media],
 212             }
 213
 214         return {
 215             **info_dict,
 216             **self._extract_product_media(product_info)
 217         }
 218
 219
 220 class InstagramIOSIE(InfoExtractor):
 221     IE_DESC = 'IOS instagram:// URL'
 222     _VALID_URL = r'instagram://media\?id=(?P<id>[\d_]+)'
 223     _TESTS = [{
 224         'url': 'instagram://media?id=482584233761418119',
 225         'md5': '0d2da106a9d2631273e192b372806516',
 226         'info_dict': {
 227             'id': 'aye83DjauH',
 228             'ext': 'mp4',
 229             'title': 'Video by naomipq',
 230             'description': 'md5:1f17f0ab29bd6fe2bfad705f58de3cb8',
 231             'thumbnail': r're:^https?://.*\.jpg',
 232             'duration': 0,
 233             'timestamp': 1371748545,
 234             'upload_date': '20130620',
 235             'uploader_id': 'naomipq',
 236             'uploader': 'B E A U T Y  F O R  A S H E S',
 237             'like_count': int,
 238             'comment_count': int,
 239             'comments': list,
 240         },
 241         'add_ie': ['Instagram']
 242     }]
 243
 244     def _real_extract(self, url):
 245         video_id = _pk_to_id(self._match_id(url))
 246         return self.url_result(f'http://instagram.com/tv/{video_id}', InstagramIE, video_id)
 247
 248
 249 class InstagramIE(InstagramBaseIE):
 250     _VALID_URL = r'(?P<url>https?://(?:www\.)?instagram\.com(?:/[^/]+)?/(?:p|tv|reel)/(?P<id>[^/?#&]+))'
 251     _EMBED_REGEX = [r'<iframe[^>]+src=(["\'])(?P<url>(?:https?:)?//(?:www\.)?instagram\.com/p/[^/]+/embed.*?)\1']
 252     _TESTS = [{
 253         'url': 'https://instagram.com/p/aye83DjauH/?foo=bar#abc',
 254         'md5': '0d2da106a9d2631273e192b372806516',
 255         'info_dict': {
 256             'id': 'aye83DjauH',
 257             'ext': 'mp4',
 258             'title': 'Video by naomipq',
 259             'description': 'md5:1f17f0ab29bd6fe2bfad705f58de3cb8',
 260             'thumbnail': r're:^https?://.*\.jpg',
 261             'duration': 0,
 262             'timestamp': 1371748545,
 263             'upload_date': '20130620',
 264             'uploader_id': '2815873',
 265             'uploader': 'B E A U T Y  F O R  A S H E S',
 266             'channel': 'naomipq',
 267             'like_count': int,
 268             'comment_count': int,
 269             'comments': list,
 270         },
 271     }, {
 272         # missing description
 273         'url': 'https://www.instagram.com/p/BA-pQFBG8HZ/?taken-by=britneyspears',
 274         'info_dict': {
 275             'id': 'BA-pQFBG8HZ',
 276             'ext': 'mp4',
 277             'title': 'Video by britneyspears',
 278             'thumbnail': r're:^https?://.*\.jpg',
 279             'duration': 0,
 280             'timestamp': 1453760977,
 281             'upload_date': '20160125',
 282             'uploader_id': '12246775',
 283             'uploader': 'Britney Spears',
 284             'channel': 'britneyspears',
 285             'like_count': int,
 286             'comment_count': int,
 287             'comments': list,
 288         },
 289         'params': {
 290             'skip_download': True,
 291         },
 292     }, {
 293         # multi video post
 294         'url': 'https://www.instagram.com/p/BQ0eAlwhDrw/',
 295         'playlist': [{
 296             'info_dict': {
 297                 'id': 'BQ0dSaohpPW',
 298                 'ext': 'mp4',
 299                 'title': 'Video 1',
 300             },
 301         }, {
 302             'info_dict': {
 303                 'id': 'BQ0dTpOhuHT',
 304                 'ext': 'mp4',
 305                 'title': 'Video 2',
 306             },
 307         }, {
 308             'info_dict': {
 309                 'id': 'BQ0dT7RBFeF',
 310                 'ext': 'mp4',
 311                 'title': 'Video 3',
 312             },
 313         }],
 314         'info_dict': {
 315             'id': 'BQ0eAlwhDrw',
 316             'title': 'Post by instagram',
 317             'description': 'md5:0f9203fc6a2ce4d228da5754bcf54957',
 318         },
 319     }, {
 320         # IGTV
 321         'url': 'https://www.instagram.com/tv/BkfuX9UB-eK/',
 322         'info_dict': {
 323             'id': 'BkfuX9UB-eK',
 324             'ext': 'mp4',
 325             'title': 'Fingerboarding Tricks with @cass.fb',
 326             'thumbnail': r're:^https?://.*\.jpg',
 327             'duration': 53.83,
 328             'timestamp': 1530032919,
 329             'upload_date': '20180626',
 330             'uploader_id': '25025320',
 331             'uploader': 'Instagram',
 332             'channel': 'instagram',
 333             'like_count': int,
 334             'comment_count': int,
 335             'comments': list,
 336             'description': 'Meet Cass Hirst (@cass.fb), a fingerboarding pro who can perform tiny ollies and kickflips while blindfolded.',
 337         }
 338     }, {
 339         'url': 'https://instagram.com/p/-Cmh1cukG2/',
 340         'only_matching': True,
 341     }, {
 342         'url': 'http://instagram.com/p/9o6LshA7zy/embed/',
 343         'only_matching': True,
 344     }, {
 345         'url': 'https://www.instagram.com/tv/aye83DjauH/',
 346         'only_matching': True,
 347     }, {
 348         'url': 'https://www.instagram.com/reel/CDUMkliABpa/',
 349         'only_matching': True,
 350     }, {
 351         'url': 'https://www.instagram.com/marvelskies.fc/reel/CWqAgUZgCku/',
 352         'only_matching': True,
 353     }]
 354
 355     @classmethod
 356     def _extract_embed_urls(cls, url, webpage):
 357         res = tuple(super()._extract_embed_urls(url, webpage))
 358         if res:
 359             return res
 360
 361         mobj = re.search(r'<a[^>]+href=([\'"])(?P<link>[^\'"]+)\1',
 362                          get_element_by_attribute('class', 'instagram-media', webpage) or '')
 363         if mobj:
 364             return [mobj.group('link')]
 365
 366     def _real_extract(self, url):
 367         video_id, url = self._match_valid_url(url).group('id', 'url')
 368         media, webpage = {}, ''
 369
 370         api_check = self._download_json(
 371             f'{self._API_BASE_URL}/web/get_ruling_for_content/?content_type=MEDIA&target_id={_id_to_pk(video_id)}',
 372             video_id, headers=self._API_HEADERS, fatal=False, note='Setting up session', errnote=False) or {}
 373         csrf_token = self._get_cookies('https://www.instagram.com').get('csrftoken')
 374
 375         if not csrf_token:
 376             self.report_warning('No csrf token set by Instagram API', video_id)
 377         elif api_check.get('status') != 'ok':
 378             self.report_warning('Instagram API is not granting access', video_id)
 379         else:
 380             if self._get_cookies(url).get('sessionid'):
 381                 media.update(traverse_obj(self._download_json(
 382                     f'{self._API_BASE_URL}/media/{_id_to_pk(video_id)}/info/', video_id,
 383                     fatal=False, note='Downloading video info', headers={
 384                         **self._API_HEADERS,
 385                         'X-CSRFToken': csrf_token.value,
 386                     }), ('items', 0)) or {})
 387                 if media:
 388                     return self._extract_product(media)
 389
 390             variables = {
 391                 'shortcode': video_id,
 392                 'child_comment_count': 3,
 393                 'fetch_comment_count': 40,
 394                 'parent_comment_count': 24,
 395                 'has_threaded_comments': True,
 396             }
 397             general_info = self._download_json(
 398                 'https://www.instagram.com/graphql/query/', video_id, fatal=False,
 399                 headers={
 400                     **self._API_HEADERS,
 401                     'X-CSRFToken': csrf_token.value,
 402                     'X-Requested-With': 'XMLHttpRequest',
 403                     'Referer': url,
 404                 }, query={
 405                     'query_hash': '9f8827793ef34641b2fb195d4d41151c',
 406                     'variables': json.dumps(variables, separators=(',', ':')),
 407                 })
 408             media.update(traverse_obj(general_info, ('data', 'shortcode_media')) or {})
 409
 410         if not media:
 411             self.report_warning('General metadata extraction failed (some metadata might be missing).', video_id)
 412             webpage, urlh = self._download_webpage_handle(url, video_id)
 413             shared_data = self._search_json(
 414                 r'window\._sharedData\s*=', webpage, 'shared data', video_id, fatal=False) or {}
 415
 416             if shared_data and self._LOGIN_URL not in urlh.geturl():
 417                 media.update(traverse_obj(
 418                     shared_data, ('entry_data', 'PostPage', 0, 'graphql', 'shortcode_media'),
 419                     ('entry_data', 'PostPage', 0, 'media'), expected_type=dict) or {})
 420             else:
 421                 self.report_warning('Main webpage is locked behind the login page. Retrying with embed webpage')
 422                 webpage = self._download_webpage(
 423                     f'{url}/embed/', video_id, note='Downloading embed webpage', fatal=False)
 424                 additional_data = self._search_json(
 425                     r'window\.__additionalDataLoaded\s*\(\s*[^,]+,\s*', webpage, 'additional data', video_id, fatal=False)
 426                 if not additional_data:
 427                     self.raise_login_required('Requested content is not available, rate-limit reached or login required')
 428
 429                 product_item = traverse_obj(additional_data, ('items', 0), expected_type=dict)
 430                 if product_item:
 431                     media.update(product_item)
 432                     return self._extract_product(media)
 433
 434                 media.update(traverse_obj(
 435                     additional_data, ('graphql', 'shortcode_media'), 'shortcode_media', expected_type=dict) or {})
 436
 437         username = traverse_obj(media, ('owner', 'username')) or self._search_regex(
 438             r'"owner"\s*:\s*{\s*"username"\s*:\s*"(.+?)"', webpage, 'username', fatal=False)
 439
 440         description = (
 441             traverse_obj(media, ('edge_media_to_caption', 'edges', 0, 'node', 'text'), expected_type=str)
 442             or media.get('caption'))
 443         if not description:
 444             description = self._search_regex(
 445                 r'"caption"\s*:\s*"(.+?)"', webpage, 'description', default=None)
 446             if description is not None:
 447                 description = lowercase_escape(description)
 448
 449         video_url = media.get('video_url')
 450         if not video_url:
 451             nodes = traverse_obj(media, ('edge_sidecar_to_children', 'edges', ..., 'node'), expected_type=dict) or []
 452             if nodes:
 453                 return self.playlist_result(
 454                     self._extract_nodes(nodes, True), video_id,
 455                     format_field(username, None, 'Post by %s'), description)
 456
 457             video_url = self._og_search_video_url(webpage, secure=False)
 458
 459         formats = [{
 460             'url': video_url,
 461             'width': self._get_dimension('width', media, webpage),
 462             'height': self._get_dimension('height', media, webpage),
 463         }]
 464         dash = traverse_obj(media, ('dash_info', 'video_dash_manifest'))
 465         if dash:
 466             formats.extend(self._parse_mpd_formats(self._parse_xml(dash, video_id), mpd_id='dash'))
 467         self._sort_formats(formats)
 468
 469         comment_data = traverse_obj(media, ('edge_media_to_parent_comment', 'edges'))
 470         comments = [{
 471             'author': traverse_obj(comment_dict, ('node', 'owner', 'username')),
 472             'author_id': traverse_obj(comment_dict, ('node', 'owner', 'id')),
 473             'id': traverse_obj(comment_dict, ('node', 'id')),
 474             'text': traverse_obj(comment_dict, ('node', 'text')),
 475             'timestamp': traverse_obj(comment_dict, ('node', 'created_at'), expected_type=int_or_none),
 476         } for comment_dict in comment_data] if comment_data else None
 477
 478         display_resources = (
 479             media.get('display_resources')
 480             or [{'src': media.get(key)} for key in ('display_src', 'display_url')]
 481             or [{'src': self._og_search_thumbnail(webpage)}])
 482         thumbnails = [{
 483             'url': thumbnail['src'],
 484             'width': thumbnail.get('config_width'),
 485             'height': thumbnail.get('config_height'),
 486         } for thumbnail in display_resources if thumbnail.get('src')]
 487
 488         return {
 489             'id': video_id,
 490             'formats': formats,
 491             'title': media.get('title') or 'Video by %s' % username,
 492             'description': description,
 493             'duration': float_or_none(media.get('video_duration')),
 494             'timestamp': traverse_obj(media, 'taken_at_timestamp', 'date', expected_type=int_or_none),
 495             'uploader_id': traverse_obj(media, ('owner', 'id')),
 496             'uploader': traverse_obj(media, ('owner', 'full_name')),
 497             'channel': username,
 498             'like_count': self._get_count(media, 'likes', 'preview_like') or str_to_int(self._search_regex(
 499                 r'data-log-event="likeCountClick"[^>]*>[^\d]*([\d,\.]+)', webpage, 'like count', fatal=False)),
 500             'comment_count': self._get_count(media, 'comments', 'preview_comment', 'to_comment', 'to_parent_comment'),
 501             'comments': comments,
 502             'thumbnails': thumbnails,
 503             'http_headers': {
 504                 'Referer': 'https://www.instagram.com/',
 505             }
 506         }
 507
 508
 509 class InstagramPlaylistBaseIE(InstagramBaseIE):
 510     _gis_tmpl = None  # used to cache GIS request type
 511
 512     def _parse_graphql(self, webpage, item_id):
 513         # Reads a webpage and returns its GraphQL data.
 514         return self._parse_json(
 515             self._search_regex(
 516                 r'sharedData\s*=\s*({.+?})\s*;\s*[<\n]', webpage, 'data'),
 517             item_id)
 518
 519     def _extract_graphql(self, data, url):
 520         # Parses GraphQL queries containing videos and generates a playlist.
 521         uploader_id = self._match_id(url)
 522         csrf_token = data['config']['csrf_token']
 523         rhx_gis = data.get('rhx_gis') or '3c7ca9dcefcf966d11dacf1f151335e8'
 524
 525         cursor = ''
 526         for page_num in itertools.count(1):
 527             variables = {
 528                 'first': 12,
 529                 'after': cursor,
 530             }
 531             variables.update(self._query_vars_for(data))
 532             variables = json.dumps(variables)
 533
 534             if self._gis_tmpl:
 535                 gis_tmpls = [self._gis_tmpl]
 536             else:
 537                 gis_tmpls = [
 538                     '%s' % rhx_gis,
 539                     '',
 540                     '%s:%s' % (rhx_gis, csrf_token),
 541                     '%s:%s:%s' % (rhx_gis, csrf_token, self.get_param('http_headers')['User-Agent']),
 542                 ]
 543
 544             # try all of the ways to generate a GIS query, and not only use the
 545             # first one that works, but cache it for future requests
 546             for gis_tmpl in gis_tmpls:
 547                 try:
 548                     json_data = self._download_json(
 549                         'https://www.instagram.com/graphql/query/', uploader_id,
 550                         'Downloading JSON page %d' % page_num, headers={
 551                             'X-Requested-With': 'XMLHttpRequest',
 552                             'X-Instagram-GIS': hashlib.md5(
 553                                 ('%s:%s' % (gis_tmpl, variables)).encode('utf-8')).hexdigest(),
 554                         }, query={
 555                             'query_hash': self._QUERY_HASH,
 556                             'variables': variables,
 557                         })
 558                     media = self._parse_timeline_from(json_data)
 559                     self._gis_tmpl = gis_tmpl
 560                     break
 561                 except ExtractorError as e:
 562                     # if it's an error caused by a bad query, and there are
 563                     # more GIS templates to try, ignore it and keep trying
 564                     if isinstance(e.cause, urllib.error.HTTPError) and e.cause.code == 403:
 565                         if gis_tmpl != gis_tmpls[-1]:
 566                             continue
 567                     raise
 568
 569             nodes = traverse_obj(media, ('edges', ..., 'node'), expected_type=dict) or []
 570             if not nodes:
 571                 break
 572             yield from self._extract_nodes(nodes)
 573
 574             has_next_page = traverse_obj(media, ('page_info', 'has_next_page'))
 575             cursor = traverse_obj(media, ('page_info', 'end_cursor'), expected_type=str)
 576             if not has_next_page or not cursor:
 577                 break
 578
 579     def _real_extract(self, url):
 580         user_or_tag = self._match_id(url)
 581         webpage = self._download_webpage(url, user_or_tag)
 582         data = self._parse_graphql(webpage, user_or_tag)
 583
 584         self._set_cookie('instagram.com', 'ig_pr', '1')
 585
 586         return self.playlist_result(
 587             self._extract_graphql(data, url), user_or_tag, user_or_tag)
 588
 589
 590 class InstagramUserIE(InstagramPlaylistBaseIE):
 591     _VALID_URL = r'https?://(?:www\.)?instagram\.com/(?P<id>[^/]{2,})/?(?:$|[?#])'
 592     IE_DESC = 'Instagram user profile'
 593     IE_NAME = 'instagram:user'
 594     _TESTS = [{
 595         'url': 'https://instagram.com/porsche',
 596         'info_dict': {
 597             'id': 'porsche',
 598             'title': 'porsche',
 599         },
 600         'playlist_count': 5,
 601         'params': {
 602             'extract_flat': True,
 603             'skip_download': True,
 604             'playlistend': 5,
 605         }
 606     }]
 607
 608     _QUERY_HASH = '42323d64886122307be10013ad2dcc44',
 609
 610     @staticmethod
 611     def _parse_timeline_from(data):
 612         # extracts the media timeline data from a GraphQL result
 613         return data['data']['user']['edge_owner_to_timeline_media']
 614
 615     @staticmethod
 616     def _query_vars_for(data):
 617         # returns a dictionary of variables to add to the timeline query based
 618         # on the GraphQL of the original page
 619         return {
 620             'id': data['entry_data']['ProfilePage'][0]['graphql']['user']['id']
 621         }
 622
 623
 624 class InstagramTagIE(InstagramPlaylistBaseIE):
 625     _VALID_URL = r'https?://(?:www\.)?instagram\.com/explore/tags/(?P<id>[^/]+)'
 626     IE_DESC = 'Instagram hashtag search URLs'
 627     IE_NAME = 'instagram:tag'
 628     _TESTS = [{
 629         'url': 'https://instagram.com/explore/tags/lolcats',
 630         'info_dict': {
 631             'id': 'lolcats',
 632             'title': 'lolcats',
 633         },
 634         'playlist_count': 50,
 635         'params': {
 636             'extract_flat': True,
 637             'skip_download': True,
 638             'playlistend': 50,
 639         }
 640     }]
 641
 642     _QUERY_HASH = 'f92f56d47dc7a55b606908374b43a314',
 643
 644     @staticmethod
 645     def _parse_timeline_from(data):
 646         # extracts the media timeline data from a GraphQL result
 647         return data['data']['hashtag']['edge_hashtag_to_media']
 648
 649     @staticmethod
 650     def _query_vars_for(data):
 651         # returns a dictionary of variables to add to the timeline query based
 652         # on the GraphQL of the original page
 653         return {
 654             'tag_name':
 655                 data['entry_data']['TagPage'][0]['graphql']['hashtag']['name']
 656         }
 657
 658
 659 class InstagramStoryIE(InstagramBaseIE):
 660     _VALID_URL = r'https?://(?:www\.)?instagram\.com/stories/(?P<user>[^/]+)/(?P<id>\d+)'
 661     IE_NAME = 'instagram:story'
 662
 663     _TESTS = [{
 664         'url': 'https://www.instagram.com/stories/highlights/18090946048123978/',
 665         'info_dict': {
 666             'id': '18090946048123978',
 667             'title': 'Rare',
 668         },
 669         'playlist_mincount': 50
 670     }]
 671
 672     def _real_extract(self, url):
 673         username, story_id = self._match_valid_url(url).groups()
 674         story_info = self._download_webpage(url, story_id)
 675         user_info = self._search_json(r'"user":', story_info, 'user info', story_id, fatal=False)
 676         if not user_info:
 677             self.raise_login_required('This content is unreachable')
 678         user_id = user_info.get('id')
 679
 680         story_info_url = user_id if username != 'highlights' else f'highlight:{story_id}'
 681         videos = traverse_obj(self._download_json(
 682             f'{self._API_BASE_URL}/feed/reels_media/?reel_ids={story_info_url}',
 683             story_id, errnote=False, fatal=False, headers=self._API_HEADERS), 'reels')
 684         if not videos:
 685             self.raise_login_required('You need to log in to access this content')
 686
 687         full_name = traverse_obj(videos, (f'highlight:{story_id}', 'user', 'full_name'), (str(user_id), 'user', 'full_name'))
 688         story_title = traverse_obj(videos, (f'highlight:{story_id}', 'title'))
 689         if not story_title:
 690             story_title = f'Story by {username}'
 691
 692         highlights = traverse_obj(videos, (f'highlight:{story_id}', 'items'), (str(user_id), 'items'))
 693         info_data = []
 694         for highlight in highlights:
 695             highlight_data = self._extract_product(highlight)
 696             if highlight_data.get('formats'):
 697                 info_data.append({
 698                     **highlight_data,
 699                     'uploader': full_name,
 700                     'uploader_id': user_id,
 701                 })
 702         return self.playlist_result(info_data, playlist_id=story_id, playlist_title=story_title)