[extractor/iwara] Fix authentication (#7137)

Closes #7035, Closes #7207
Authored by: toomyzoom

yt_dlp/extractor/iwara.py

@@ -1,68 +1,83 @@
 import functools
 import urllib.parse
+import urllib.error
 import hashlib
 import json
+import time
 
 from .common import InfoExtractor
 from ..utils import (
     ExtractorError,
     OnDemandPagedList,
     int_or_none,
+    jwt_decode_hs256,
     mimetype2ext,
     qualities,
     traverse_obj,
+    try_call,
     unified_timestamp,
 )
 
 
-# https://github.com/yt-dlp/yt-dlp/issues/6671
 class IwaraBaseIE(InfoExtractor):
+    _NETRC_MACHINE = 'iwara'
     _USERTOKEN = None
     _MEDIATOKEN = None
-    _NETRC_MACHINE = 'iwara'
 
-    def _get_user_token(self, invalidate=False):
+    def _is_token_expired(self, token, token_type):
-        if not invalidate and self._USERTOKEN:
+        # User token TTL == ~3 weeks, Media token TTL == ~1 hour
-            return self._USERTOKEN
+        if (try_call(lambda: jwt_decode_hs256(token)['exp']) or 0) <= int(time.time() - 120):
+            self.to_screen(f'{token_type} token has expired')
+            return True
 
+    def _get_user_token(self):
         username, password = self._get_login_info()
-        IwaraBaseIE._USERTOKEN = username and self.cache.load(self._NETRC_MACHINE, username)
+        if not username or not password:
-        if not IwaraBaseIE._USERTOKEN or invalidate:
+            return
-            IwaraBaseIE._USERTOKEN = self._download_json(
+
+        user_token = IwaraBaseIE._USERTOKEN or self.cache.load(self._NETRC_MACHINE, username)
+        if not user_token or self._is_token_expired(user_token, 'User'):
+            response = self._download_json(
                 'https://api.iwara.tv/user/login', None, note='Logging in',
-                data=json.dumps({
+                headers={'Content-Type': 'application/json'}, data=json.dumps({
                     'email': username,
                     'password': password
-                }).encode('utf-8'),
+                }).encode(), expected_status=lambda x: True)
-                headers={
+            user_token = traverse_obj(response, ('token', {str}))
+            if not user_token:
+                error = traverse_obj(response, ('message', {str}))
+                if 'invalidLogin' in error:
+                    raise ExtractorError('Invalid login credentials', expected=True)
+                else:
+                    raise ExtractorError(f'Iwara API said: {error or "nothing"}')
+
+            self.cache.store(self._NETRC_MACHINE, username, user_token)
+
+        IwaraBaseIE._USERTOKEN = user_token
+
+    def _get_media_token(self):
+        self._get_user_token()
+        if not IwaraBaseIE._USERTOKEN:
+            return  # user has not passed credentials
+
+        if not IwaraBaseIE._MEDIATOKEN or self._is_token_expired(IwaraBaseIE._MEDIATOKEN, 'Media'):
+            IwaraBaseIE._MEDIATOKEN = self._download_json(
+                'https://api.iwara.tv/user/token', None, note='Fetching media token',
+                data=b'', headers={
+                    'Authorization': f'Bearer {IwaraBaseIE._USERTOKEN}',
                     'Content-Type': 'application/json'
-                })['token']
+                })['accessToken']
 
-            self.cache.store(self._NETRC_MACHINE, username, IwaraBaseIE._USERTOKEN)
+        return {'Authorization': f'Bearer {IwaraBaseIE._MEDIATOKEN}'}
 
-        return self._USERTOKEN
+    def _perform_login(self, username, password):
-
+        self._get_media_token()
-    def _get_media_token(self, invalidate=False):
-        if not invalidate and self._MEDIATOKEN:
-            return self._MEDIATOKEN
-
-        IwaraBaseIE._MEDIATOKEN = self._download_json(
-            'https://api.iwara.tv/user/token', None, note='Fetching media token',
-            data=b'',  # Need to have some data here, even if it's empty
-            headers={
-                'Authorization': f'Bearer {self._get_user_token()}',
-                'Content-Type': 'application/json'
-            })['accessToken']
-
-        return self._MEDIATOKEN
 
 
 class IwaraIE(IwaraBaseIE):
     IE_NAME = 'iwara'
     _VALID_URL = r'https?://(?:www\.|ecchi\.)?iwara\.tv/videos?/(?P<id>[a-zA-Z0-9]+)'
     _TESTS = [{
-        # this video cannot be played because of migration
-        'only_matching': True,
         'url': 'https://www.iwara.tv/video/k2ayoueezfkx6gvq',
         'info_dict': {
             'id': 'k2ayoueezfkx6gvq',
@@ -79,25 +94,29 @@ class IwaraIE(IwaraBaseIE):
             'timestamp': 1677843869,
             'modified_timestamp': 1679056362,
         },
+        'skip': 'this video cannot be played because of migration',
     }, {
         'url': 'https://iwara.tv/video/1ywe1sbkqwumpdxz5/',
-        'md5': '20691ce1473ec2766c0788e14c60ce66',
+        'md5': '7645f966f069b8ec9210efd9130c9aad',
         'info_dict': {
             'id': '1ywe1sbkqwumpdxz5',
             'ext': 'mp4',
             'age_limit': 18,
-            'title': 'Aponia 阿波尼亚SEX  Party Tonight 手动脱衣 大奶 裸腿',
+            'title': 'Aponia アポニア SEX  Party Tonight 手の脱衣 巨乳 ',
-            'description': 'md5:0c4c310f2e0592d68b9f771d348329ca',
+            'description': 'md5:3f60016fff22060eef1ef26d430b1f67',
-            'uploader': '龙也zZZ',
+            'uploader': 'Lyu ya',
             'uploader_id': 'user792540',
             'tags': [
                 'uncategorized'
             ],
-            'like_count': 1809,
+            'like_count': int,
-            'view_count': 25156,
+            'view_count': int,
-            'comment_count': 1,
+            'comment_count': int,
             'timestamp': 1678732213,
-            'modified_timestamp': 1679110271,
+            'modified_timestamp': int,
+            'thumbnail': 'https://files.iwara.tv/image/thumbnail/581d12b5-46f4-4f15-beb2-cfe2cde5d13d/thumbnail-00.jpg',
+            'modified_date': '20230614',
+            'upload_date': '20230313',
         },
     }, {
         'url': 'https://iwara.tv/video/blggmfno8ghl725bg',
@@ -112,12 +131,15 @@ class IwaraIE(IwaraBaseIE):
             'tags': [
                 'pee'
             ],
-            'like_count': 192,
+            'like_count': int,
-            'view_count': 12119,
+            'view_count': int,
-            'comment_count': 0,
+            'comment_count': int,
             'timestamp': 1598880567,
-            'modified_timestamp': 1598908995,
+            'modified_timestamp': int,
-            'availability': 'needs_auth',
+            'upload_date': '20200831',
+            'modified_date': '20230605',
+            'thumbnail': 'https://files.iwara.tv/image/thumbnail/7693e881-d302-42a4-a780-f16d66b5dadd/thumbnail-00.jpg',
+            # 'availability': 'needs_auth',
         },
     }]
 
@@ -142,17 +164,16 @@ def _extract_formats(self, video_id, fileurl):
 
     def _real_extract(self, url):
         video_id = self._match_id(url)
-        username, password = self._get_login_info()
+        username, _ = self._get_login_info()
-        headers = {
+        video_data = self._download_json(
-            'Authorization': f'Bearer {self._get_media_token()}',
+            f'https://api.iwara.tv/video/{video_id}', video_id,
-        } if username and password else None
+            expected_status=lambda x: True, headers=self._get_media_token())
-        video_data = self._download_json(f'https://api.iwara.tv/video/{video_id}', video_id, expected_status=lambda x: True, headers=headers)
         errmsg = video_data.get('message')
         # at this point we can actually get uploaded user info, but do we need it?
         if errmsg == 'errors.privateVideo':
-            self.raise_login_required('Private video. Login if you have permissions to watch')
+            self.raise_login_required('Private video. Login if you have permissions to watch', method='password')
         elif errmsg == 'errors.notFound' and not username:
-            self.raise_login_required('Video may need login to view')
+            self.raise_login_required('Video may need login to view', method='password')
         elif errmsg:  # None if success
             raise ExtractorError(f'Iwara says: {errmsg}')
 
@@ -181,15 +202,6 @@ def _real_extract(self, url):
             'formats': list(self._extract_formats(video_id, video_data.get('fileUrl'))),
         }
 
-    def _perform_login(self, username, password):
-        if self.cache.load(self._NETRC_MACHINE, username) and self._get_media_token():
-            self.write_debug('Skipping logging in')
-            return
-
-        IwaraBaseIE._USERTOKEN = self._get_user_token(True)
-        self._get_media_token(True)
-        self.cache.store(self._NETRC_MACHINE, username, IwaraBaseIE._USERTOKEN)
-
 
 class IwaraUserIE(IwaraBaseIE):
     _VALID_URL = r'https?://(?:www\.)?iwara\.tv/profile/(?P<id>[^/?#&]+)'
@@ -200,12 +212,14 @@ class IwaraUserIE(IwaraBaseIE):
         'url': 'https://iwara.tv/profile/user792540/videos',
         'info_dict': {
             'id': 'user792540',
+            'title': 'Lyu ya',
         },
-        'playlist_mincount': 80,
+        'playlist_mincount': 70,
     }, {
         'url': 'https://iwara.tv/profile/theblackbirdcalls/videos',
         'info_dict': {
             'id': 'theblackbirdcalls',
+            'title': 'TheBlackbirdCalls',
         },
         'playlist_mincount': 723,
     }, {
@@ -214,6 +228,13 @@ class IwaraUserIE(IwaraBaseIE):
     }, {
         'url': 'https://iwara.tv/profile/theblackbirdcalls',
         'only_matching': True,
+    }, {
+        'url': 'https://www.iwara.tv/profile/lumymmd',
+        'info_dict': {
+            'id': 'lumymmd',
+            'title': 'Lumy MMD',
+        },
+        'playlist_mincount': 1,
     }]
 
     def _entries(self, playlist_id, user_id, page):
@@ -225,7 +246,7 @@ def _entries(self, playlist_id, user_id, page):
                 'sort': 'date',
                 'user': user_id,
                 'limit': self._PER_PAGE,
-            })
+            }, headers=self._get_media_token())
         for x in traverse_obj(videos, ('results', ..., 'id')):
             yield self.url_result(f'https://iwara.tv/video/{x}')
 
@@ -244,7 +265,6 @@ def _real_extract(self, url):
 
 
 class IwaraPlaylistIE(IwaraBaseIE):
-    # the ID is an UUID but I don't think it's necessary to write concrete regex
     _VALID_URL = r'https?://(?:www\.)?iwara\.tv/playlist/(?P<id>[0-9a-f-]+)'
     IE_NAME = 'iwara:playlist'
     _PER_PAGE = 32
@@ -260,7 +280,8 @@ class IwaraPlaylistIE(IwaraBaseIE):
     def _entries(self, playlist_id, first_page, page):
         videos = self._download_json(
             'https://api.iwara.tv/videos', playlist_id, f'Downloading page {page}',
-            query={'page': page, 'limit': self._PER_PAGE}) if page else first_page
+            query={'page': page, 'limit': self._PER_PAGE},
+            headers=self._get_media_token()) if page else first_page
         for x in traverse_obj(videos, ('results', ..., 'id')):
             yield self.url_result(f'https://iwara.tv/video/{x}')
 
@@ -268,7 +289,7 @@ def _real_extract(self, url):
         playlist_id = self._match_id(url)
         page_0 = self._download_json(
             f'https://api.iwara.tv/playlist/{playlist_id}?page=0&limit={self._PER_PAGE}', playlist_id,
-            note='Requesting playlist info')
+            note='Requesting playlist info', headers=self._get_media_token())
 
         return self.playlist_result(
             OnDemandPagedList(