Merge 454bf0baf8 into 12d8ea8246

Closes #9554
Authored by: coletdjnz, bashonly

Co-authored-by: bashonly <88596187+bashonly@users.noreply.github.com>

Makefile

@@ -73,11 +73,11 @@ codetest:
 	flake8 .
 
 test:
-	$(PYTHON) -m pytest
+	$(PYTHON) -m pytest -Werror
 	$(MAKE) codetest
 
 offlinetest: codetest
-	$(PYTHON) -m pytest -k "not download"
+	$(PYTHON) -m pytest -Werror -m "not download"
 
 CODE_FOLDERS_CMD = find yt_dlp -type f -name '__init__.py' | sed 's,/__init__.py,,' | grep -v '/__' | sort
 CODE_FOLDERS != $(CODE_FOLDERS_CMD)

README.md

@@ -108,7 +108,6 @@ #### Alternatives
 [yt-dlp_x86.exe](https://github.com/yt-dlp/yt-dlp/releases/latest/download/yt-dlp_x86.exe)|Windows (Win7 SP1+) standalone x86 (32-bit) binary
 [yt-dlp_min.exe](https://github.com/yt-dlp/yt-dlp/releases/latest/download/yt-dlp_min.exe)|Windows (Win7 SP1+) standalone x64 binary built with `py2exe`<br/> ([Not recommended](#standalone-py2exe-builds-windows))
 [yt-dlp_linux](https://github.com/yt-dlp/yt-dlp/releases/latest/download/yt-dlp_linux)|Linux standalone x64 binary
-[yt-dlp_linux.zip](https://github.com/yt-dlp/yt-dlp/releases/latest/download/yt-dlp_linux.zip)|Unpackaged Linux executable (no auto-update)
 [yt-dlp_linux_armv7l](https://github.com/yt-dlp/yt-dlp/releases/latest/download/yt-dlp_linux_armv7l)|Linux standalone armv7l (32-bit) binary
 [yt-dlp_linux_aarch64](https://github.com/yt-dlp/yt-dlp/releases/latest/download/yt-dlp_linux_aarch64)|Linux standalone aarch64 (64-bit) binary
 [yt-dlp_win.zip](https://github.com/yt-dlp/yt-dlp/releases/latest/download/yt-dlp_win.zip)|Unpackaged Windows executable (no auto-update)
@@ -170,7 +169,7 @@ # To update to nightly from stable executable/binary:
 yt-dlp --update-to nightly
 
 # To install nightly with pip:
-python3 -m pip install -U --pre yt-dlp[default]
+python3 -m pip install -U --pre "yt-dlp[default]"
 ```
 
 ## DEPENDENCIES
@@ -202,7 +201,7 @@ #### Impersonation
 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
 
 
@@ -1760,7 +1759,7 @@ # EXTRACTOR ARGUMENTS
 #### youtube
 * `lang`: Prefer translated metadata (`title`, `description` etc) of this language code (case-sensitive). By default, the video primary language metadata is preferred, with a fallback to `en` translated. See [youtube.py](https://github.com/yt-dlp/yt-dlp/blob/c26f9b991a0681fd3ea548d535919cec1fbbd430/yt_dlp/extractor/youtube.py#L381-L390) for list of supported content language codes
 * `skip`: One or more of `hls`, `dash` or `translated_subs` to skip extraction of the m3u8 manifests, dash manifests and [auto-translated subtitles](https://github.com/yt-dlp/yt-dlp/issues/4090#issuecomment-1158102032) respectively
-* `player_client`: Clients to extract video data from. The main clients are `web`, `android` and `ios` with variants `_music`, `_embedded`, `_embedscreen`, `_creator` (e.g. `web_embedded`); and `mweb`, `mweb_embedscreen`, `mediaconnect` and `tv_embedded` (agegate bypass) with no variants. By default, `ios,android,web` is used, but `tv_embedded` and `creator` variants are added as required for age-gated videos. Similarly, the music variants are added for `music.youtube.com` urls. You can use `all` to use all the clients, and `default` for the default clients.
+* `player_client`: Clients to extract video data from. The main clients are `web`, `ios` and `android`, with variants `_music`, `_embedded`, `_embedscreen`, `_creator` (e.g. `web_embedded`); and `mweb`, `mweb_embedscreen` and `tv_embedded` (agegate bypass) with no variants. By default, `ios,web` is used, but `tv_embedded` and `creator` variants are added as required for age-gated videos. Similarly, the music variants are added for `music.youtube.com` urls. The `android` clients will always be given lowest priority since their formats are broken. You can use `all` to use all the clients, and `default` for the default clients.
 * `player_skip`: Skip some network requests that are generally needed for robust extraction. One or more of `configs` (skip client configs), `webpage` (skip initial webpage), `js` (skip js player). While these options can help reduce the number of requests needed or avoid some rate-limiting, they could cause some issues. See [#860](https://github.com/yt-dlp/yt-dlp/pull/860) for more details
 * `player_params`: YouTube player parameters to use for player requests. Will overwrite any default ones set by yt-dlp.
 * `comment_sort`: `top` or `new` (default) - choose comment sorting mode (on YouTube's side)

devscripts/run_tests.bat

@@ -1,4 +0,0 @@
-@echo off
-
->&2 echo run_tests.bat is deprecated. Please use `devscripts/run_tests.py` instead
-python %~dp0run_tests.py %~1

devscripts/run_tests.sh

@@ -1,4 +0,0 @@
-#!/usr/bin/env sh
-
->&2 echo 'run_tests.sh is deprecated. Please use `devscripts/run_tests.py` instead'
-python3 devscripts/run_tests.py "$1"

pyinst.py

@@ -1,17 +0,0 @@
-#!/usr/bin/env python3
-
-# Allow execution from anywhere
-import os
-import sys
-
-sys.path.insert(0, os.path.dirname(os.path.abspath(__file__)))
-
-import warnings
-
-from bundle.pyinstaller import main
-
-warnings.warn(DeprecationWarning('`pyinst.py` is deprecated and will be removed in a future version. '
-                                 'Use `bundle.pyinstaller` instead'))
-
-if __name__ == '__main__':
-    main()

setup.py

@@ -1,36 +0,0 @@
-#!/usr/bin/env python3
-
-# Allow execution from anywhere
-import os
-import sys
-
-sys.path.insert(0, os.path.dirname(os.path.abspath(__file__)))
-
-import warnings
-
-
-if sys.argv[1:2] == ['py2exe']:
-    warnings.warn(DeprecationWarning('`setup.py py2exe` is deprecated and will be removed in a future version. '
-                                     'Use `bundle.py2exe` instead'))
-
-    import bundle.py2exe
-
-    bundle.py2exe.main()
-
-elif 'build_lazy_extractors' in sys.argv:
-    warnings.warn(DeprecationWarning('`setup.py build_lazy_extractors` is deprecated and will be removed in a future version. '
-                                     'Use `devscripts.make_lazy_extractors` instead'))
-
-    import subprocess
-
-    os.chdir(sys.path[0])
-    print('running build_lazy_extractors')
-    subprocess.run([sys.executable, 'devscripts/make_lazy_extractors.py'])
-
-else:
-
-    print(
-        'ERROR: Building by calling `setup.py` is deprecated. '
-        'Use a build frontend like `build` instead. ',
-        'Refer to  https://build.pypa.io  for more info', file=sys.stderr)
-    sys.exit(1)

test/test_InfoExtractor.py

@@ -1912,7 +1912,7 @@ def test_search_nextjs_data(self):
         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):
+        with self.assertWarns(DeprecationWarning):
             self.assertEqual(self.ie._search_nextjs_data('', None, default='{}'), {})
 
 

yt_dlp/extractor/ceskatelevize.py

@@ -101,7 +101,7 @@ def _real_extract(self, url):
         site_name = self._og_search_property('site_name', webpage, fatal=False, default='Česká televize')
         playlist_title = self._og_search_title(webpage, default=None)
         if site_name and playlist_title:
-            playlist_title = re.split(r'\s*[—|]\s*%s' % (site_name, ), playlist_title, 1)[0]
+            playlist_title = re.split(r'\s*[—|]\s*%s' % (site_name, ), playlist_title, maxsplit=1)[0]
         playlist_description = self._og_search_description(webpage, default=None)
         if playlist_description:
             playlist_description = playlist_description.replace('\xa0', ' ')

yt_dlp/extractor/common.py

@@ -3531,7 +3531,7 @@ def _parse_jwplayer_formats(self, jwplayer_sources_data, video_id=None,
                     # See com/longtailvideo/jwplayer/media/RTMPMediaProvider.as
                     # of jwplayer.flash.swf
                     rtmp_url_parts = re.split(
-                        r'((?:mp4|mp3|flv):)', source_url, 1)
+                        r'((?:mp4|mp3|flv):)', source_url, maxsplit=1)
                     if len(rtmp_url_parts) == 3:
                         rtmp_url, prefix, play_path = rtmp_url_parts
                         a_format.update({

yt_dlp/extractor/thisvid.py

@@ -134,7 +134,7 @@ def _make_playlist_result(self, url):
         title = re.split(
             r'(?i)\s*\|\s*ThisVid\.com\s*$',
             self._og_search_title(webpage, default=None)
-            or self._html_search_regex(r'(?s)<title\b[^>]*>(.+?)</title', webpage, 'title', fatal=False) or '', 1)[0] or None
+            or self._html_search_regex(r'(?s)<title\b[^>]*>(.+?)</title', webpage, 'title', fatal=False) or '', maxsplit=1)[0] or None
 
         return self.playlist_from_matches(
             self._generate_playlist_entries(url, playlist_id, webpage),

yt_dlp/extractor/vk.py

@@ -467,13 +467,13 @@ def _real_extract(self, url):
                     'source_preference': 1,
                     'height': height,
                 })
-            elif format_id == 'hls':
+            elif format_id.startswith('hls') and format_id != 'hls_live_playback':
                 fmts, subs = self._extract_m3u8_formats_and_subtitles(
                     format_url, video_id, 'mp4', 'm3u8_native',
                     m3u8_id=format_id, fatal=False, live=is_live)
                 formats.extend(fmts)
                 self._merge_subtitles(subs, target=subtitles)
-            elif format_id.startswith('dash_'):
+            elif format_id.startswith('dash') and format_id not in ('dash_live_playback', 'dash_uni'):
                 fmts, subs = self._extract_mpd_formats_and_subtitles(
                     format_url, video_id, mpd_id=format_id, fatal=False)
                 formats.extend(fmts)

yt_dlp/extractor/youtube.py

@@ -2353,6 +2353,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
                 'format': '17',  # 3gp format available on android
                 'extractor_args': {'youtube': {'player_client': ['android']}},
             },
+            'skip': 'android client broken',
         },
         {
             # Skip download of additional client configs (remix client config in this case)
@@ -2730,7 +2731,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
                 'heatmap': 'count:100',
             },
             'params': {
-                'extractor_args': {'youtube': {'player_client': ['android'], 'player_skip': ['webpage']}},
+                'extractor_args': {'youtube': {'player_client': ['ios'], 'player_skip': ['webpage']}},
             },
         },
     ]
@@ -3662,8 +3663,6 @@ def _extract_player_response(self, client, video_id, master_ytcfg, player_ytcfg,
         yt_query = {
             'videoId': video_id,
         }
-        if _split_innertube_client(client)[0] in ('android', 'android_embedscreen'):
-            yt_query['params'] = 'CgIIAQ=='
 
         pp_arg = self._configuration_arg('player_params', [None], casesense=True)[0]
         if pp_arg:
@@ -3679,19 +3678,24 @@ def _extract_player_response(self, client, video_id, master_ytcfg, player_ytcfg,
 
     def _get_requested_clients(self, url, smuggled_data):
         requested_clients = []
-        default = ['ios', 'android', 'web']
+        android_clients = []
+        default = ['ios', 'web']
         allowed_clients = sorted(
             (client for client in INNERTUBE_CLIENTS.keys() if client[:1] != '_'),
             key=lambda client: INNERTUBE_CLIENTS[client]['priority'], reverse=True)
         for client in self._configuration_arg('player_client'):
-            if client in allowed_clients:
-                requested_clients.append(client)
-            elif client == 'default':
+            if client == 'default':
                 requested_clients.extend(default)
             elif client == 'all':
                 requested_clients.extend(allowed_clients)
-            else:
+            elif client not in allowed_clients:
                 self.report_warning(f'Skipping unsupported client {client}')
+            elif client.startswith('android'):
+                android_clients.append(client)
+            else:
+                requested_clients.append(client)
+        # Force deprioritization of broken Android clients for format de-duplication
+        requested_clients.extend(android_clients)
         if not requested_clients:
             requested_clients = default
 
@@ -3910,6 +3914,14 @@ def build_fragments(f):
                     f'{video_id}: Some formats are possibly damaged. They will be deprioritized', only_once=True)
 
             client_name = fmt.get(STREAMING_DATA_CLIENT_NAME)
+            # Android client formats are broken due to integrity check enforcement
+            # Ref: https://github.com/yt-dlp/yt-dlp/issues/9554
+            is_broken = client_name and client_name.startswith(short_client_name('android'))
+            if is_broken:
+                self.report_warning(
+                    f'{video_id}: Android client formats are broken and may yield HTTP Error 403. '
+                    'They will be deprioritized', only_once=True)
+
             name = fmt.get('qualityLabel') or quality.replace('audio_quality_', '') or ''
             fps = int_or_none(fmt.get('fps')) or 0
             dct = {
@@ -3922,7 +3934,7 @@ def build_fragments(f):
                     name, fmt.get('isDrc') and 'DRC',
                     try_get(fmt, lambda x: x['projectionType'].replace('RECTANGULAR', '').lower()),
                     try_get(fmt, lambda x: x['spatialAudioType'].replace('SPATIAL_AUDIO_TYPE_', '').lower()),
-                    throttled and 'THROTTLED', is_damaged and 'DAMAGED',
+                    throttled and 'THROTTLED', is_damaged and 'DAMAGED', is_broken and 'BROKEN',
                     (self.get_param('verbose') or all_formats) and client_name,
                     delim=', '),
                 # Format 22 is likely to be damaged. See https://github.com/yt-dlp/yt-dlp/issues/3372
@@ -3940,8 +3952,8 @@ def build_fragments(f):
                 'language': join_nonempty(audio_track.get('id', '').split('.')[0],
                                           'desc' if language_preference < -1 else '') or None,
                 'language_preference': language_preference,
-                # Strictly de-prioritize damaged and 3gp formats
-                'preference': -10 if is_damaged else -2 if itag == '17' else None,
+                # Strictly de-prioritize broken, damaged and 3gp formats
+                'preference': -20 if is_broken else -10 if is_damaged else -2 if itag == '17' else None,
             }
             mime_mobj = re.match(
                 r'((?:[^/]+)/(?:[^;]+))(?:;\s*codecs="([^"]+)")?', fmt.get('mimeType') or '')

yt_dlp/utils/_utils.py

@@ -2522,7 +2522,7 @@ def fixup(url):
             return False
         # "#" cannot be stripped out since it is part of the URI
         # However, it can be safely stripped out if following a whitespace
-        return re.split(r'\s#', url, 1)[0].rstrip()
+        return re.split(r'\s#', url, maxsplit=1)[0].rstrip()
 
     with contextlib.closing(batch_fd) as fd:
         return [url for url in map(fixup, fd) if url]