Merge 634a0ac756 into ff38a011d5

Closes #9745
Authored by: bashonly

.github/workflows/build.yml

@@ -254,7 +254,7 @@ jobs:
           # We need to fuse our own universal2 wheels for curl_cffi
           python3 -m pip install -U --user delocate
           mkdir curl_cffi_whls curl_cffi_universal2
-          python3 devscripts/install_deps.py --print -o --include curl_cffi > requirements.txt
+          python3 devscripts/install_deps.py --print -o --include curl-cffi > requirements.txt
           for platform in "macosx_11_0_arm64" "macosx_11_0_x86_64"; do
             python3 -m pip download \
               --only-binary=:all: \
@@ -362,7 +362,7 @@ jobs:
       - name: Install Requirements
         run: | # Custom pyinstaller built with https://github.com/yt-dlp/pyinstaller-builds
           python devscripts/install_deps.py -o --include build
-          python devscripts/install_deps.py --include py2exe --include curl_cffi
+          python devscripts/install_deps.py --include py2exe --include curl-cffi
           python -m pip install -U "https://yt-dlp.github.io/Pyinstaller-Builds/x86_64/pyinstaller-5.8.0-py3-none-any.whl"
 
       - name: Prepare

README.md

@@ -202,7 +202,7 @@ While all the other dependencies are optional, `ffmpeg` and `ffprobe` are highly
 The following provide support for impersonating browser requests. This may be required for some sites that employ TLS fingerprinting. 
 
 * [**curl_cffi**](https://github.com/yifeikong/curl_cffi) (recommended) - Python binding for [curl-impersonate](https://github.com/lwthiker/curl-impersonate). Provides impersonation targets for Chrome, Edge and Safari. Licensed under [MIT](https://github.com/yifeikong/curl_cffi/blob/main/LICENSE)
-  * Can be installed with the `curl_cffi` group, e.g. `pip install yt-dlp[default,curl_cffi]`
+  * Can be installed with the `curl-cffi` group, e.g. `pip install yt-dlp[default,curl-cffi]`
   * Currently only included in `yt-dlp.exe` and `yt-dlp_macos` builds
 
 

pyproject.toml

@@ -53,7 +53,7 @@ dependencies = [
 
 [project.optional-dependencies]
 default = []
-curl_cffi = ["curl-cffi==0.5.10; implementation_name=='cpython'"]
+curl-cffi = ["curl-cffi==0.5.10; implementation_name=='cpython'"]
 secretstorage = [
     "cffi",
     "secretstorage",

test/test_InfoExtractor.py

@@ -1906,6 +1906,15 @@ jwplayer("mediaplayer").setup({"abouttext":"Visit Indie DB","aboutlink":"http:\/
             expected_status=TEAPOT_RESPONSE_STATUS)
         self.assertEqual(content, TEAPOT_RESPONSE_BODY)
 
+    def test_search_nextjs_data(self):
+        data = '<script id="__NEXT_DATA__" type="application/json">{"props":{}}</script>'
+        self.assertEqual(self.ie._search_nextjs_data(data, None), {'props': {}})
+        self.assertEqual(self.ie._search_nextjs_data('', None, fatal=False), {})
+        self.assertEqual(self.ie._search_nextjs_data('', None, default=None), None)
+        self.assertEqual(self.ie._search_nextjs_data('', None, default={}), {})
+        with self.assertRaises(DeprecationWarning):
+            self.assertEqual(self.ie._search_nextjs_data('', None, default='{}'), {})
+
 
 if __name__ == '__main__':
     unittest.main()

yt_dlp/extractor/asobistage.py

@@ -105,7 +105,7 @@ class AsobiStageIE(InfoExtractor):
         video_type = {'archive': 'archives', 'player': 'broadcasts'}[type_]
         webpage = self._download_webpage(url, video_id)
         event_data = traverse_obj(
-            self._search_nextjs_data(webpage, video_id, default='{}'),
+            self._search_nextjs_data(webpage, video_id, default={}),
             ('props', 'pageProps', 'eventCMSData', {
                 'title': ('event_name', {str}),
                 'thumbnail': ('event_thumbnail_image', {url_or_none}),

yt_dlp/extractor/common.py

@@ -1738,12 +1738,16 @@ class InfoExtractor:
         traverse_json_ld(json_ld)
         return filter_dict(info)
 
-    def _search_nextjs_data(self, webpage, video_id, *, transform_source=None, fatal=True, **kw):
-        return self._parse_json(
-            self._search_regex(
-                r'(?s)<script[^>]+id=[\'"]__NEXT_DATA__[\'"][^>]*>([^<]+)</script>',
-                webpage, 'next.js data', fatal=fatal, **kw),
-            video_id, transform_source=transform_source, fatal=fatal)
+    def _search_nextjs_data(self, webpage, video_id, *, fatal=True, default=NO_DEFAULT, **kw):
+        if default == '{}':
+            self._downloader.deprecation_warning('using `default=\'{}\'` is deprecated, use `default={}` instead')
+            default = {}
+        if default is not NO_DEFAULT:
+            fatal = False
+
+        return self._search_json(
+            r'<script[^>]+id=[\'"]__NEXT_DATA__[\'"][^>]*>', webpage, 'next.js data',
+            video_id, end_pattern='</script>', fatal=fatal, default=default, **kw)
 
     def _search_nuxt_data(self, webpage, video_id, context_name='__NUXT__', *, fatal=True, traverse=('data', 0)):
         """Parses Nuxt.js metadata. This works as long as the function __NUXT__ invokes is a pure function"""

yt_dlp/extractor/crunchyroll.py

@@ -24,11 +24,15 @@ class CrunchyrollBaseIE(InfoExtractor):
     _BASE_URL = 'https://www.crunchyroll.com'
     _API_BASE = 'https://api.crunchyroll.com'
     _NETRC_MACHINE = 'crunchyroll'
+    _REFRESH_TOKEN = None
     _AUTH_HEADERS = None
+    _AUTH_EXPIRY = None
     _API_ENDPOINT = None
-    _BASIC_AUTH = None
+    _BASIC_AUTH = 'Basic ' + base64.b64encode(':'.join((
+        't-kdgp2h8c3jub8fn0fq',
+        'yfLDfMfrYvKXh4JXS1LEI2cCqu1v5Wan',
+    )).encode()).decode()
     _IS_PREMIUM = None
-    _CLIENT_ID = ('cr_web', 'noaihdevm_6iyg0a8l0q')
     _LOCALE_LOOKUP = {
         'ar': 'ar-SA',
         'de': 'de-DE',
@@ -43,69 +47,74 @@ class CrunchyrollBaseIE(InfoExtractor):
         'hi': 'hi-IN',
     }
 
-    @property
-    def is_logged_in(self):
-        return bool(self._get_cookies(self._BASE_URL).get('etp_rt'))
+    def _set_auth_info(self, response):
+        CrunchyrollBaseIE._IS_PREMIUM = 'cr_premium' in traverse_obj(response, ('access_token', {jwt_decode_hs256}, 'benefits', ...))
+        CrunchyrollBaseIE._AUTH_HEADERS = {'Authorization': response['token_type'] + ' ' + response['access_token']}
+        CrunchyrollBaseIE._AUTH_EXPIRY = time_seconds(seconds=traverse_obj(response, ('expires_in', {float_or_none}), default=300) - 10)
+
+    def _request_token(self, headers, data, note='Requesting token', errnote='Failed to request token'):
+        try:  # TODO: Add impersonation support here
+            return self._download_json(
+                f'{self._BASE_URL}/auth/v1/token', None, note=note, errnote=errnote,
+                headers=headers, data=urlencode_postdata(data))
+        except ExtractorError as error:
+            if not isinstance(error.cause, HTTPError) or error.cause.status != 403:
+                raise
+            raise ExtractorError(
+                'Request blocked by Cloudflare; navigate to Crunchyroll in your browser, '
+                'then pass the fresh cookies (with --cookies-from-browser or --cookies) '
+                'and your browser\'s User-Agent (with --user-agent)', expected=True)
 
     def _perform_login(self, username, password):
-        if self.is_logged_in:
+        if not CrunchyrollBaseIE._REFRESH_TOKEN:
+            CrunchyrollBaseIE._REFRESH_TOKEN = self.cache.load(self._NETRC_MACHINE, username)
+        if CrunchyrollBaseIE._REFRESH_TOKEN:
             return
 
-        upsell_response = self._download_json(
-            f'{self._API_BASE}/get_upsell_data.0.json', None, 'Getting session id',
-            query={
-                'sess_id': 1,
-                'device_id': 'whatvalueshouldbeforweb',
-                'device_type': 'com.crunchyroll.static',
-                'access_token': 'giKq5eY27ny3cqz',
-                'referer': f'{self._BASE_URL}/welcome/login'
-            })
-        if upsell_response['code'] != 'ok':
-            raise ExtractorError('Could not get session id')
-        session_id = upsell_response['data']['session_id']
-
-        login_response = self._download_json(
-            f'{self._API_BASE}/login.1.json', None, 'Logging in',
-            data=urlencode_postdata({
-                'account': username,
-                'password': password,
-                'session_id': session_id
-            }))
-        if login_response['code'] != 'ok':
-            raise ExtractorError('Login failed. Server message: %s' % login_response['message'], expected=True)
-        if not self.is_logged_in:
-            raise ExtractorError('Login succeeded but did not set etp_rt cookie')
-
-    def _update_auth(self):
-        if CrunchyrollBaseIE._AUTH_HEADERS and CrunchyrollBaseIE._AUTH_REFRESH > time_seconds():
-            return
-
-        if not CrunchyrollBaseIE._BASIC_AUTH:
-            cx_api_param = self._CLIENT_ID[self.is_logged_in]
-            self.write_debug(f'Using cxApiParam={cx_api_param}')
-            CrunchyrollBaseIE._BASIC_AUTH = 'Basic ' + base64.b64encode(f'{cx_api_param}:'.encode()).decode()
-
-        auth_headers = {'Authorization': CrunchyrollBaseIE._BASIC_AUTH}
-        if self.is_logged_in:
-            grant_type = 'etp_rt_cookie'
-        else:
-            grant_type = 'client_id'
-            auth_headers['ETP-Anonymous-ID'] = uuid.uuid4()
         try:
-            auth_response = self._download_json(
-                f'{self._BASE_URL}/auth/v1/token', None, note=f'Authenticating with grant_type={grant_type}',
-                headers=auth_headers, data=f'grant_type={grant_type}'.encode())
+            login_response = self._request_token(
+                headers={'Authorization': self._BASIC_AUTH}, data={
+                    'username': username,
+                    'password': password,
+                    'grant_type': 'password',
+                    'scope': 'offline_access',
+                }, note='Logging in', errnote='Failed to log in')
         except ExtractorError as error:
-            if isinstance(error.cause, HTTPError) and error.cause.status == 403:
-                raise ExtractorError(
-                    'Request blocked by Cloudflare; navigate to Crunchyroll in your browser, '
-                    'then pass the fresh cookies (with --cookies-from-browser or --cookies) '
-                    'and your browser\'s User-Agent (with --user-agent)', expected=True)
+            if isinstance(error.cause, HTTPError) and error.cause.status == 401:
+                raise ExtractorError('Invalid username and/or password', expected=True)
             raise
 
-        CrunchyrollBaseIE._IS_PREMIUM = 'cr_premium' in traverse_obj(auth_response, ('access_token', {jwt_decode_hs256}, 'benefits', ...))
-        CrunchyrollBaseIE._AUTH_HEADERS = {'Authorization': auth_response['token_type'] + ' ' + auth_response['access_token']}
-        CrunchyrollBaseIE._AUTH_REFRESH = time_seconds(seconds=traverse_obj(auth_response, ('expires_in', {float_or_none}), default=300) - 10)
+        CrunchyrollBaseIE._REFRESH_TOKEN = login_response['refresh_token']
+        self.cache.store(self._NETRC_MACHINE, username, CrunchyrollBaseIE._REFRESH_TOKEN)
+        self._set_auth_info(login_response)
+
+    def _update_auth(self):
+        if CrunchyrollBaseIE._AUTH_HEADERS and CrunchyrollBaseIE._AUTH_EXPIRY > time_seconds():
+            return
+
+        auth_headers = {'Authorization': self._BASIC_AUTH}
+        if CrunchyrollBaseIE._REFRESH_TOKEN:
+            data = {
+                'refresh_token': CrunchyrollBaseIE._REFRESH_TOKEN,
+                'grant_type': 'refresh_token',
+                'scope': 'offline_access',
+            }
+        else:
+            data = {'grant_type': 'client_id'}
+            auth_headers['ETP-Anonymous-ID'] = uuid.uuid4()
+        try:
+            auth_response = self._request_token(auth_headers, data)
+        except ExtractorError as error:
+            username, password = self._get_login_info()
+            if not username or not isinstance(error.cause, HTTPError) or error.cause.status != 400:
+                raise
+            self.to_screen('Refresh token has expired. Re-logging in')
+            CrunchyrollBaseIE._REFRESH_TOKEN = None
+            self.cache.store(self._NETRC_MACHINE, username, None)
+            self._perform_login(username, password)
+            return
+
+        self._set_auth_info(auth_response)
 
     def _locale_from_language(self, language):
         config_locale = self._configuration_arg('metadata', ie_key=CrunchyrollBetaIE, casesense=True)
@@ -168,7 +177,8 @@ class CrunchyrollBaseIE(InfoExtractor):
         self._update_auth()
         stream_response = self._download_json(
             f'https://cr-play-service.prd.crunchyrollsvc.com/v1/{identifier}/console/switch/play',
-            display_id, note='Downloading stream info', headers=CrunchyrollBaseIE._AUTH_HEADERS)
+            display_id, note='Downloading stream info', errnote='Failed to download stream info',
+            headers=CrunchyrollBaseIE._AUTH_HEADERS)
 
         available_formats = {'': ('', '', stream_response['url'])}
         for hardsub_lang, stream in traverse_obj(stream_response, ('hardSubs', {dict.items}, lambda _, v: v[1]['url'])):
@@ -383,9 +393,9 @@ class CrunchyrollBetaIE(CrunchyrollCmsBaseIE):
 
         if not self._IS_PREMIUM and traverse_obj(response, (f'{object_type}_metadata', 'is_premium_only')):
             message = f'This {object_type} is for premium members only'
-            if self.is_logged_in:
+            if CrunchyrollBaseIE._REFRESH_TOKEN:
                 raise ExtractorError(message, expected=True)
-            self.raise_login_required(message)
+            self.raise_login_required(message, method='password')
 
         result['formats'], result['subtitles'] = self._extract_stream(internal_id)
 
@@ -575,9 +585,9 @@ class CrunchyrollMusicIE(CrunchyrollBaseIE):
 
         if not self._IS_PREMIUM and response.get('isPremiumOnly'):
             message = f'This {response.get("type") or "media"} is for premium members only'
-            if self.is_logged_in:
+            if CrunchyrollBaseIE._REFRESH_TOKEN:
                 raise ExtractorError(message, expected=True)
-            self.raise_login_required(message)
+            self.raise_login_required(message, method='password')
 
         result = self._transform_music_response(response)
         result['formats'], _ = self._extract_stream(f'music/{internal_id}', internal_id)

yt_dlp/extractor/facebook.py

@@ -560,7 +560,7 @@ class FacebookIE(InfoExtractor):
                     js_data, lambda x: x['jsmods']['instances'], list) or [])
 
         def extract_dash_manifest(video, formats):
-            dash_manifest = video.get('dash_manifest')
+            dash_manifest = traverse_obj(video, 'dash_manifest', 'playlist', expected_type=str)
             if dash_manifest:
                 formats.extend(self._parse_mpd_formats(
                     compat_etree_fromstring(urllib.parse.unquote_plus(dash_manifest)),

yt_dlp/extractor/patreon.py

@@ -1,8 +1,8 @@
 import itertools
+import urllib.parse
 
 from .common import InfoExtractor
 from .vimeo import VimeoIE
-from ..compat import compat_urllib_parse_unquote
 from ..networking.exceptions import HTTPError
 from ..utils import (
     KNOWN_EXTENSIONS,
@@ -14,7 +14,6 @@ from ..utils import (
     parse_iso8601,
     str_or_none,
     traverse_obj,
-    try_get,
     url_or_none,
     urljoin,
 )
@@ -199,6 +198,27 @@ class PatreonIE(PatreonBaseIE):
             'channel_id': '2147162',
             'uploader_url': 'https://www.patreon.com/yaboyroshi',
         },
+    }, {
+        # NSFW vimeo embed URL
+        'url': 'https://www.patreon.com/posts/4k-spiderman-4k-96414599',
+        'info_dict': {
+            'id': '902250943',
+            'ext': 'mp4',
+            'title': '❤️(4K) Spiderman Girl Yeonhwa’s Gift ❤️(4K) 스파이더맨걸 연화의 선물',
+            'description': '❤️(4K) Spiderman Girl Yeonhwa’s Gift \n❤️(4K) 스파이더맨걸 연화의 선물',
+            'uploader': 'Npickyeonhwa',
+            'uploader_id': '90574422',
+            'uploader_url': 'https://www.patreon.com/Yeonhwa726',
+            'channel_id': '10237902',
+            'channel_url': 'https://www.patreon.com/Yeonhwa726',
+            'duration': 70,
+            'timestamp': 1705150153,
+            'upload_date': '20240113',
+            'comment_count': int,
+            'like_count': int,
+            'thumbnail': r're:^https?://.+',
+        },
+        'params': {'skip_download': 'm3u8'},
     }]
 
     def _real_extract(self, url):
@@ -268,16 +288,19 @@ class PatreonIE(PatreonBaseIE):
                 })
 
         # handle Vimeo embeds
-        if try_get(attributes, lambda x: x['embed']['provider']) == 'Vimeo':
-            embed_html = try_get(attributes, lambda x: x['embed']['html'])
-            v_url = url_or_none(compat_urllib_parse_unquote(
-                self._search_regex(r'(https(?:%3A%2F%2F|://)player\.vimeo\.com.+app_id(?:=|%3D)+\d+)', embed_html, 'vimeo url', fatal=False)))
-            if v_url:
-                v_url = VimeoIE._smuggle_referrer(v_url, 'https://patreon.com')
-                if self._request_webpage(v_url, video_id, 'Checking Vimeo embed URL', fatal=False, errnote=False):
-                    return self.url_result(v_url, VimeoIE, url_transparent=True, **info)
+        if traverse_obj(attributes, ('embed', 'provider')) == 'Vimeo':
+            v_url = urllib.parse.unquote(self._html_search_regex(
+                r'(https(?:%3A%2F%2F|://)player\.vimeo\.com.+app_id(?:=|%3D)+\d+)',
+                traverse_obj(attributes, ('embed', 'html', {str})), 'vimeo url', fatal=False) or '')
+            if url_or_none(v_url) and self._request_webpage(
+                    v_url, video_id, 'Checking Vimeo embed URL',
+                    headers={'Referer': 'https://patreon.com/'},
+                    fatal=False, errnote=False):
+                return self.url_result(
+                    VimeoIE._smuggle_referrer(v_url, 'https://patreon.com/'),
+                    VimeoIE, url_transparent=True, **info)
 
-        embed_url = try_get(attributes, lambda x: x['embed']['url'])
+        embed_url = traverse_obj(attributes, ('embed', 'url', {url_or_none}))
         if embed_url and self._request_webpage(embed_url, video_id, 'Checking embed URL', fatal=False, errnote=False):
             return self.url_result(embed_url, **info)
 

yt_dlp/extractor/stacommu.py

@@ -174,7 +174,7 @@ class TheaterComplexTownBaseIE(StacommuBaseIE):
 
 
 class TheaterComplexTownVODIE(TheaterComplexTownBaseIE):
-    _VALID_URL = r'https?://(?:www\.)?theater-complex\.town/(?:en/)?videos/episodes/(?P<id>\w+)'
+    _VALID_URL = r'https?://(?:www\.)?theater-complex\.town/(?:(?:en|ja)/)?videos/episodes/(?P<id>\w+)'
     IE_NAME = 'theatercomplextown:vod'
     _TESTS = [{
         'url': 'https://www.theater-complex.town/videos/episodes/hoxqidYNoAn7bP92DN6p78',
@@ -195,6 +195,9 @@ class TheaterComplexTownVODIE(TheaterComplexTownBaseIE):
     }, {
         'url': 'https://www.theater-complex.town/en/videos/episodes/6QT7XYwM9dJz5Gf9VB6K5y',
         'only_matching': True,
+    }, {
+        'url': 'https://www.theater-complex.town/ja/videos/episodes/hoxqidYNoAn7bP92DN6p78',
+        'only_matching': True,
     }]
 
     _API_PATH = 'videoEpisodes'
@@ -204,7 +207,7 @@ class TheaterComplexTownVODIE(TheaterComplexTownBaseIE):
 
 
 class TheaterComplexTownPPVIE(TheaterComplexTownBaseIE):
-    _VALID_URL = r'https?://(?:www\.)?theater-complex\.town/(?:en/)?ppv/(?P<id>\w+)'
+    _VALID_URL = r'https?://(?:www\.)?theater-complex\.town/(?:(?:en|ja)/)?ppv/(?P<id>\w+)'
     IE_NAME = 'theatercomplextown:ppv'
     _TESTS = [{
         'url': 'https://www.theater-complex.town/ppv/wytW3X7khrjJBUpKuV3jen',
@@ -223,6 +226,9 @@ class TheaterComplexTownPPVIE(TheaterComplexTownBaseIE):
     }, {
         'url': 'https://www.theater-complex.town/en/ppv/wytW3X7khrjJBUpKuV3jen',
         'only_matching': True,
+    }, {
+        'url': 'https://www.theater-complex.town/ja/ppv/qwUVmLmGEiZ3ZW6it9uGys',
+        'only_matching': True,
     }]
 
     _API_PATH = 'events'

yt_dlp/extractor/stv.py

@@ -41,7 +41,7 @@ class STVPlayerIE(InfoExtractor):
         ptype, video_id = self._match_valid_url(url).groups()
 
         webpage = self._download_webpage(url, video_id, fatal=False) or ''
-        props = self._search_nextjs_data(webpage, video_id, default='{}').get('props') or {}
+        props = self._search_nextjs_data(webpage, video_id, default={}).get('props') or {}
         player_api_cache = try_get(
             props, lambda x: x['initialReduxState']['playerApiCache']) or {}
 

yt_dlp/extractor/tiktok.py

@@ -776,7 +776,7 @@ class TikTokIE(TikTokBaseIE):
             status = traverse_obj(sigi_data, ('VideoPage', 'statusCode', {int})) or 0
             video_data = traverse_obj(sigi_data, ('ItemModule', video_id, {dict}))
 
-        elif next_data := self._search_nextjs_data(webpage, video_id, default='{}'):
+        elif next_data := self._search_nextjs_data(webpage, video_id, default={}):
             self.write_debug('Found next.js data')
             status = traverse_obj(next_data, ('props', 'pageProps', 'statusCode', {int})) or 0
             video_data = traverse_obj(next_data, ('props', 'pageProps', 'itemInfo', 'itemStruct', {dict}))

yt_dlp/extractor/wrestleuniverse.py

@@ -147,7 +147,7 @@ class WrestleUniverseBaseIE(InfoExtractor):
         metadata = self._call_api(video_id, msg='metadata', query={'al': lang or 'ja'}, auth=False, fatal=False)
         if not metadata:
             webpage = self._download_webpage(url, video_id)
-            nextjs_data = self._search_nextjs_data(webpage, video_id)
+            nextjs_data = self._search_nextjs_data(webpage, video_id, fatal=False)
             metadata = traverse_obj(nextjs_data, (
                 'props', 'pageProps', *variadic(props_keys, (str, bytes, dict, set)), {dict})) or {}
         return metadata