yt_dlp/extractor/nebula.py

@@ -1,11 +1,10 @@
 import itertools
 import json
+import urllib.error
 
 from .common import InfoExtractor
-from ..networking.exceptions import HTTPError
 from ..utils import (
     ExtractorError,
-    int_or_none,
     make_archive_id,
     parse_iso8601,
     smuggle_url,
@@ -30,12 +29,15 @@ class NebulaBaseIE(InfoExtractor):
                 'https://nebula.tv/auth/login/', None,
                 'Logging in to Nebula', 'Login failed',
                 data=json.dumps({'email': username, 'password': password}).encode(),
-                headers={'content-type': 'application/json'})
+                headers={
+                    'content-type': 'application/json',
+                    'cookie': ''  # 'sessionid' cookie causes 403
+                })
         except ExtractorError as e:
-            if isinstance(e.cause, HTTPError) and e.cause.status == 400:
+            if isinstance(e.cause, urllib.error.HTTPError) and e.cause.status == 400:
                 raise ExtractorError('Login failed: Invalid username or password', expected=True)
             raise
-        self._api_token = traverse_obj(response, ('key', {str}))
+        self._api_token = response.get('key')
         if not self._api_token:
             raise ExtractorError('Login failed: No token')
 
@@ -45,10 +47,10 @@ class NebulaBaseIE(InfoExtractor):
         try:
             return self._download_json(*args, **kwargs)
         except ExtractorError as e:
-            if not isinstance(e.cause, HTTPError) or e.cause.status not in (401, 403):
+            if not isinstance(e.cause, urllib.error.HTTPError) or e.cause.status not in (401, 403):
                 raise
-            self.to_screen(
-                f'Reauthorizing with Nebula and retrying, because last API call resulted in error {e.cause.status}')
+            self.to_screen(f'Reautherizing with Nebula and retrying, '
+                           f'because last API call resulted in error {e.cause.status}')
             self._real_initialize()
             if self._token:
                 kwargs.setdefault('headers', {})['Authorization'] = f'Bearer {self._token}'
@@ -63,37 +65,38 @@ class NebulaBaseIE(InfoExtractor):
             headers={'Authorization': f'Token {self._api_token}'} if self._api_token else None,
             note='Authorizing to Nebula', data=b'')['token']
 
-    def _extract_formats(self, content_id, slug):
-        for retry in (False, True):
-            try:
-                fmts, subs = self._extract_m3u8_formats_and_subtitles(
-                    f'https://content.api.nebula.app/{content_id.split(":")[0]}s/{content_id}/manifest.m3u8',
-                    slug, 'mp4', query={
-                        'token': self._token,
-                        'app_version': '23.10.0',
-                        'platform': 'ios',
-                    })
-                return {'formats': fmts, 'subtitles': subs}
-            except ExtractorError as e:
-                if isinstance(e.cause, HTTPError) and e.cause.status == 401:
-                    self.raise_login_required()
-                if not retry and isinstance(e.cause, HTTPError) and e.cause.status == 403:
-                    self.to_screen('Reauthorizing with Nebula and retrying, because fetching video resulted in error')
-                    self._real_initialize()
-                    continue
+    def _extract_formats(self, content_id, slug, fatal=False):
+        api_path = 'lessons' if content_id.startswith('lesson:') else 'video_episodes'
+        try:
+            fmts, subs = self._extract_m3u8_formats_and_subtitles(
+                f'https://content.api.nebula.app/{api_path}/{content_id}/manifest.m3u8',
+                slug, 'mp4', query={
+                    'token': self._token,
+                    'app_version': '23.10.0',
+                    'platform': 'ios',
+                })
+        except ExtractorError as e:
+            if isinstance(e.cause, urllib.error.HTTPError) and e.cause.status == 401:
+                self.raise_login_required('This video is only available for users with an active subscription')
+            if fatal or not isinstance(e.cause, urllib.error.HTTPError) or e.cause.status != 403:
                 raise
+            self.to_screen('Reautherizing with Nebula and retrying, because fetching video resulted in error')
+            self._real_initialize()
+            return self._extract_formats(content_id, slug, fatal=True)
+
+        return {'formats': fmts, 'subtitles': subs}
 
     def _extract_video_metadata(self, episode):
         channel_url = traverse_obj(
             episode, (('channel_slug', 'class_slug'), {lambda x: urljoin('https://nebula.tv/', x)}), get_all=False)
         return {
-            'id': episode['id'].partition(':')[2],
+            'id': episode['id'].split(':', 1)[-1],
             **traverse_obj(episode, {
                 'display_id': 'slug',
                 'title': 'title',
                 'description': 'description',
                 'timestamp': ('published_at', {parse_iso8601}),
-                'duration': ('duration', {int_or_none}),
+                'duration': 'duration',
                 'channel_id': 'channel_slug',
                 'uploader_id': 'channel_slug',
                 'channel': 'channel_title',
@@ -101,7 +104,7 @@ class NebulaBaseIE(InfoExtractor):
                 'series': 'channel_title',
                 'creator': 'channel_title',
                 'thumbnail': ('images', 'thumbnail', 'src', {url_or_none}),
-                'episode_number': ('order', {int_or_none}),
+                'episode_number': 'order',
                 # Old code was wrongly setting extractor_key from NebulaSubscriptionsIE
                 '_old_archive_ids': ('zype_id', {lambda x: [
                     make_archive_id(NebulaIE, x), make_archive_id(NebulaSubscriptionsIE, x)] if x else None}),
@@ -236,7 +239,6 @@ class NebulaIE(NebulaBaseIE):
 
 
 class NebulaClassIE(NebulaBaseIE):
-    IE_NAME = 'nebula:class'
     _VALID_URL = rf'{_BASE_URL_RE}/(?P<id>[-\w]+)/(?P<ep>\d+)'
     _TESTS = [{
         'url': 'https://nebula.tv/copyright-for-fun-and-profit/14',
@@ -358,7 +360,7 @@ class NebulaChannelIE(NebulaBaseIE):
             if not next_url:
                 break
 
-    def _generate_class_entries(self, channel):
+    def _generate_playlist_entries_class(self, channel):
         for lesson in channel['lessons']:
             metadata = self._extract_video_metadata(lesson)
             yield self.url_result(smuggle_url(
@@ -372,7 +374,7 @@ class NebulaChannelIE(NebulaBaseIE):
             collection_slug, note='Retrieving channel')
 
         if channel.get('type') == 'class':
-            entries = self._generate_class_entries(channel)
+            entries = self._generate_playlist_entries_class(channel)
         else:
             entries = self._generate_playlist_entries(channel['id'], collection_slug)