[extractor/instagram] Fix extraction (#4696)

Closes #4657, #4532, #4475
Authored by: bashonly, pritam20ps05

yt_dlp/extractor/instagram.py

@@ -39,25 +39,30 @@ class InstagramBaseIE(InfoExtractor):
     _NETRC_MACHINE = 'instagram'
     _IS_LOGGED_IN = False
 
+    _API_BASE_URL = 'https://i.instagram.com/api/v1'
+    _LOGIN_URL = 'https://www.instagram.com/accounts/login'
+    _API_HEADERS = {
+        'X-IG-App-ID': '936619743392459',
+        'X-ASBD-ID': '198387',
+        'X-IG-WWW-Claim': '0',
+        'Origin': 'https://www.instagram.com',
+        'Accept': '*/*',
+        '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',
+    }
+
     def _perform_login(self, username, password):
         if self._IS_LOGGED_IN:
             return
 
         login_webpage = self._download_webpage(
-            'https://www.instagram.com/accounts/login/', None,
+            self._LOGIN_URL, None, note='Downloading login webpage', errnote='Failed to download login webpage')
-            note='Downloading login webpage', errnote='Failed to download login webpage')
 
-        shared_data = self._parse_json(
+        shared_data = self._parse_json(self._search_regex(
-            self._search_regex(
+            r'window\._sharedData\s*=\s*({.+?});', login_webpage, 'shared data', default='{}'), None)
-                r'window\._sharedData\s*=\s*({.+?});',
-                login_webpage, 'shared data', default='{}'),
-            None)
 
-        login = self._download_json('https://www.instagram.com/accounts/login/ajax/', None, note='Logging in', headers={
+        login = self._download_json(
-            'Accept': '*/*',
+            f'{self._LOGIN_URL}/ajax/', None, note='Logging in', headers={
-            'X-IG-App-ID': '936619743392459',
+                **self._API_HEADERS,
-            'X-ASBD-ID': '198387',
-            'X-IG-WWW-Claim': '0',
                 'X-Requested-With': 'XMLHttpRequest',
                 'X-CSRFToken': shared_data['config']['csrf_token'],
                 'X-Instagram-AJAX': shared_data['rollout_hash'],
@@ -134,7 +139,7 @@ class InstagramBaseIE(InfoExtractor):
             }
 
     def _extract_product_media(self, product_media):
-        media_id = product_media.get('code') or product_media.get('id')
+        media_id = product_media.get('code') or _pk_to_id(product_media.get('pk'))
         vcodec = product_media.get('video_codec')
         dash_manifest_raw = product_media.get('video_dash_manifest')
         videos_list = product_media.get('video_versions')
@@ -179,7 +184,7 @@ class InstagramBaseIE(InfoExtractor):
 
         user_info = product_info.get('user') or {}
         info_dict = {
-            'id': product_info.get('code') or product_info.get('id'),
+            'id': product_info.get('code') or _pk_to_id(product_info.get('pk')),
             'title': product_info.get('title') or f'Video by {user_info.get("username")}',
             'description': traverse_obj(product_info, ('caption', 'text'), expected_type=str_or_none),
             'timestamp': int_or_none(product_info.get('taken_at')),
@@ -360,42 +365,67 @@ class InstagramIE(InstagramBaseIE):
 
     def _real_extract(self, url):
         video_id, url = self._match_valid_url(url).group('id', 'url')
-        general_info = self._download_json(
+        media, webpage = {}, ''
-            f'https://www.instagram.com/graphql/query/?query_hash=9f8827793ef34641b2fb195d4d41151c'
-            f'&variables=%7B"shortcode":"{video_id}",'
-            '"parent_comment_count":10,"has_threaded_comments":true}', video_id, fatal=False, errnote=False,
-            headers={
-                'Accept': '*',
-                'User-Agent': 'Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/103.0.0.0 Safari/537.36',
-                'Authority': 'www.instagram.com',
-                'Referer': 'https://www.instagram.com',
-                'x-ig-app-id': '936619743392459',
-            })
-        media = traverse_obj(general_info, ('data', 'shortcode_media')) or {}
-        if not media:
-            self.report_warning('General metadata extraction failed', video_id)
 
-        info = self._download_json(
+        api_check = self._download_json(
-            f'https://i.instagram.com/api/v1/media/{_id_to_pk(video_id)}/info/', video_id,
+            f'{self._API_BASE_URL}/web/get_ruling_for_content/?content_type=MEDIA&target_id={_id_to_pk(video_id)}',
-            fatal=False, note='Downloading video info', errnote=False, headers={
+            video_id, headers=self._API_HEADERS, fatal=False, note='Setting up session', errnote=False) or {}
-                'Accept': '*',
+        csrf_token = self._get_cookies('https://www.instagram.com').get('csrftoken')
-                'User-Agent': 'Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/103.0.0.0 Safari/537.36',
+
-                'Authority': 'www.instagram.com',
+        if not csrf_token:
-                'Referer': 'https://www.instagram.com',
+            self.report_warning('No csrf token set by Instagram API', video_id)
-                'x-ig-app-id': '936619743392459',
+        elif api_check.get('status') != 'ok':
-            })
+            self.report_warning('Instagram API is not granting access', video_id)
-        if info:
+        else:
-            media.update(info['items'][0])
+            if self._get_cookies(url).get('sessionid'):
+                media = traverse_obj(self._download_json(
+                    f'{self._API_BASE_URL}/media/{_id_to_pk(video_id)}/info/', video_id,
+                    fatal=False, note='Downloading video info', headers={
+                        **self._API_HEADERS,
+                        'X-CSRFToken': csrf_token.value,
+                    }), ('items', 0))
+                if media:
                     return self._extract_product(media)
 
-        webpage = self._download_webpage(
+            variables = {
-            f'https://www.instagram.com/p/{video_id}/embed/', video_id,
+                'shortcode': video_id,
-            note='Downloading embed webpage', fatal=False)
+                'child_comment_count': 3,
-        if not webpage:
+                'fetch_comment_count': 40,
-            self.raise_login_required('Requested content was not found, the content might be private')
+                'parent_comment_count': 24,
+                'has_threaded_comments': True,
+            }
+            general_info = self._download_json(
+                'https://www.instagram.com/graphql/query/', video_id, fatal=False,
+                headers={
+                    **self._API_HEADERS,
+                    'X-CSRFToken': csrf_token.value,
+                    'X-Requested-With': 'XMLHttpRequest',
+                    'Referer': url,
+                }, query={
+                    'query_hash': '9f8827793ef34641b2fb195d4d41151c',
+                    'variables': json.dumps(variables, separators=(',', ':')),
+                })
+            media = traverse_obj(general_info, ('data', 'shortcode_media'))
 
+        if not media:
+            self.report_warning('General metadata extraction failed (some metadata might be missing).', video_id)
+            webpage, urlh = self._download_webpage_handle(url, video_id)
+            shared_data = self._search_json(
+                r'window\._sharedData\s*=', webpage, 'shared data', video_id, fatal=False)
+
+            if self._LOGIN_URL not in urlh.geturl():
+                media.update(traverse_obj(
+                    shared_data, ('entry_data', 'PostPage', 0, 'graphql', 'shortcode_media'),
+                    ('entry_data', 'PostPage', 0, 'media'), expected_type=dict) or {})
+            else:
+                self.report_warning('Main webpage is locked behind the login page. Retrying with embed webpage')
+                webpage = self._download_webpage(
+                    f'{url}/embed/', video_id, note='Downloading embed webpage', fatal=False)
                 additional_data = self._search_json(
                     r'window\.__additionalDataLoaded\s*\(\s*[^,]+,\s*', webpage, 'additional data', video_id, fatal=False)
+                if not additional_data:
+                    self.raise_login_required('Requested content was not found, the content might be private')
+
                 product_item = traverse_obj(additional_data, ('items', 0), expected_type=dict)
                 if product_item:
                     media.update(product_item)
@@ -649,12 +679,8 @@ class InstagramStoryIE(InstagramBaseIE):
 
         story_info_url = user_id if username != 'highlights' else f'highlight:{story_id}'
         videos = traverse_obj(self._download_json(
-            f'https://i.instagram.com/api/v1/feed/reels_media/?reel_ids={story_info_url}',
+            f'{self._API_BASE_URL}/feed/reels_media/?reel_ids={story_info_url}',
-            story_id, errnote=False, fatal=False, headers={
+            story_id, errnote=False, fatal=False, headers=self._API_HEADERS), 'reels')
-                'X-IG-App-ID': 936619743392459,
-                'X-ASBD-ID': 198387,
-                'X-IG-WWW-Claim': 0,
-            }), 'reels')
         if not videos:
             self.raise_login_required('You need to log in to access this content')