Merge a376ab99f8 into 96da952504

Authored by: seproDev, pukkandan

devscripts/make_lazy_extractors.py

@@ -15,7 +15,7 @@ from devscripts.utils import get_filename_args, read_file, write_file
 NO_ATTR = object()
 STATIC_CLASS_PROPERTIES = [
     'IE_NAME', '_ENABLED', '_VALID_URL',  # Used for URL matching
-    '_WORKING', 'IE_DESC', '_NETRC_MACHINE', 'SEARCH_KEY',  # Used for --extractor-descriptions
+    '_REPORTED_BROKEN', 'IE_DESC', '_NETRC_MACHINE', 'SEARCH_KEY',  # Used for --extractor-descriptions
     'age_limit',  # Used for --age-limit (evaluated)
     '_RETURN_TYPE',  # Accessed in CLI only with instance (evaluated)
 ]

supportedsites.md

@@ -304,10 +304,10 @@
  - **CrowdBunker**
  - **CrowdBunkerChannel**
  - **Crtvg**
- - **crunchyroll**: [*crunchyroll*](## "netrc machine")
- - **crunchyroll:artist**: [*crunchyroll*](## "netrc machine")
- - **crunchyroll:music**: [*crunchyroll*](## "netrc machine")
- - **crunchyroll:playlist**: [*crunchyroll*](## "netrc machine")
+ - **crunchyroll**: [*crunchyroll*](## "netrc machine") ([**Currently broken**](https://github.com/yt-dlp/yt-dlp/issues/9453))
+ - **crunchyroll:artist**: [*crunchyroll*](## "netrc machine") ([**Currently broken**](https://github.com/yt-dlp/yt-dlp/issues/9453))
+ - **crunchyroll:music**: [*crunchyroll*](## "netrc machine") ([**Currently broken**](https://github.com/yt-dlp/yt-dlp/issues/9453))
+ - **crunchyroll:playlist**: [*crunchyroll*](## "netrc machine") ([**Currently broken**](https://github.com/yt-dlp/yt-dlp/issues/9453))
  - **CSpan**: C-SPAN
  - **CSpanCongress**
  - **CtsNews**: 華視新聞
@@ -553,7 +553,7 @@
  - **hgtv.com:show**
  - **HGTVDe**
  - **HGTVUsa**
- - **HiDive**: [*hidive*](## "netrc machine")
+ - **HiDive**: [*hidive*](## "netrc machine") ([**Currently broken**](https://github.com/yt-dlp/yt-dlp/issues/9385))
  - **HistoricFilms**
  - **history:player**
  - **history:topic**: History.com Topic

test/test_download.py

@@ -98,7 +98,7 @@ def generator(test_case, tname):
             self.skipTest(reason)
 
         if not ie.working():
-            print_skipping('IE marked as not _WORKING')
+            print_skipping('IE is _REPORTED_BROKEN')
 
         for tc in test_cases:
             if tc.get('expected_exception'):
@@ -117,7 +117,7 @@ def generator(test_case, tname):
 
         for other_ie in other_ies:
             if not other_ie.working():
-                print_skipping('test depends on %sIE, marked as not WORKING' % other_ie.ie_key())
+                print_skipping(f'test depends on {other_ie.ie_key()}IE, is _REPORTED_BROKEN')
 
         params = get_params(test_case.get('params', {}))
         params['outtmpl'] = tname + '_' + params['outtmpl']

test/test_subtitles.py

@@ -40,8 +40,8 @@ class BaseTestSubtitles(unittest.TestCase):
         self.ie = self.IE()
         self.DL.add_info_extractor(self.ie)
         if not self.IE.working():
-            print('Skipping: %s marked as not _WORKING' % self.IE.ie_key())
-            self.skipTest('IE marked as not _WORKING')
+            print(f'Skipping: {self.IE.ie_key()} is _REPORTED_BROKEN')
+            self.skipTest('IE is _REPORTED_BROKEN')
 
     def getInfoDict(self):
         info_dict = self.DL.extract_info(self.url, download=False)

yt_dlp/YoutubeDL.py

@@ -1582,8 +1582,12 @@ class YoutubeDL:
                 continue
 
             if not ie.working():
-                self.report_warning('The program functionality for this site has been marked as broken, '
-                                    'and will probably not work.')
+                self.report_warning(join_nonempty(
+                    f'[{ie.IE_NAME}] The program\'s functionality for this site has been marked as '
+                    f'{self._format_err("BROKEN", self.Styles.ERROR)}, and will probably not work.',
+                    format_field(ie._REPORTED_BROKEN, None, f'See  {self._format_err("%s", self.Styles.EMPHASIS)}'
+                                 '  for more information. Do NOT open a new issue for this.'),
+                    delim='\n         '))
 
             temp_id = ie.get_temp_id(url)
             if temp_id is not None and self.in_download_archive({'id': temp_id, 'ie_key': key}):
@@ -2136,6 +2140,11 @@ class YoutubeDL:
 
     def _check_formats(self, formats):
         for f in formats:
+            working = f.get('__working')
+            if working is not None:
+                if working:
+                    yield f
+                continue
             self.to_screen('[info] Testing format %s' % f['format_id'])
             path = self.get_output_path('temp')
             if not self._ensure_dir_exists(f'{path}/'):
@@ -2152,33 +2161,44 @@ class YoutubeDL:
                         os.remove(temp_file.name)
                     except OSError:
                         self.report_warning('Unable to delete temporary file "%s"' % temp_file.name)
+            f['__working'] = success
             if success:
                 yield f
             else:
                 self.to_screen('[info] Unable to download format %s. Skipping...' % f['format_id'])
 
+    def _select_formats(self, formats, selector):
+        return list(selector({
+            'formats': formats,
+            'has_merged_format': any('none' not in (f.get('acodec'), f.get('vcodec')) for f in formats),
+            'incomplete_formats': (all(f.get('vcodec') == 'none' for f in formats)  # No formats with video
+                                   or all(f.get('acodec') == 'none' for f in formats)),  # OR, No formats with audio
+        }))
+
     def _default_format_spec(self, info_dict, download=True):
+        download = download and not self.params.get('simulate')
+        prefer_best = download and (
+            self.params['outtmpl']['default'] == '-'
+            or info_dict.get('is_live') and not self.params.get('live_from_start'))
 
         def can_merge():
             merger = FFmpegMergerPP(self)
             return merger.available and merger.can_merge()
 
-        prefer_best = (
-            not self.params.get('simulate')
-            and download
-            and (
-                not can_merge()
-                or info_dict.get('is_live') and not self.params.get('live_from_start')
-                or self.params['outtmpl']['default'] == '-'))
-        compat = (
-            prefer_best
-            or self.params.get('allow_multiple_audio_streams', False)
-            or 'format-spec' in self.params['compat_opts'])
+        if not prefer_best and download and not can_merge():
+            prefer_best = True
+            formats = self._get_formats(info_dict)
+            evaluate_formats = lambda spec: self._select_formats(formats, self.build_format_selector(spec))
+            if evaluate_formats('b/bv+ba') != evaluate_formats('bv*+ba/b'):
+                self.report_warning('ffmpeg not found. The downloaded format may not be the best available. '
+                                    'Installing ffmpeg is strongly recommended: https://github.com/yt-dlp/yt-dlp#dependencies')
 
-        return (
-            'best/bestvideo+bestaudio' if prefer_best
-            else 'bestvideo*+bestaudio/best' if not compat
-            else 'bestvideo+bestaudio/best')
+        compat = (self.params.get('allow_multiple_audio_streams')
+                  or 'format-spec' in self.params['compat_opts'])
+
+        return ('best/bestvideo+bestaudio' if prefer_best
+                else 'bestvideo+bestaudio/best' if compat
+                else 'bestvideo*+bestaudio/best')
 
     def build_format_selector(self, format_spec):
         def syntax_error(note, start):
@@ -2928,12 +2948,7 @@ class YoutubeDL:
                 self.write_debug(f'Default format spec: {req_format}')
                 format_selector = self.build_format_selector(req_format)
 
-            formats_to_download = list(format_selector({
-                'formats': formats,
-                'has_merged_format': any('none' not in (f.get('acodec'), f.get('vcodec')) for f in formats),
-                'incomplete_formats': (all(f.get('vcodec') == 'none' for f in formats)  # No formats with video
-                                       or all(f.get('acodec') == 'none' for f in formats)),  # OR, No formats with audio
-            }))
+            formats_to_download = self._select_formats(formats, format_selector)
             if interactive_format_selection and not formats_to_download:
                 self.report_error('Requested format is not available', tb=False, is_error=False)
                 continue

yt_dlp/extractor/common.py

@@ -546,8 +546,9 @@ class InfoExtractor:
     The _ENABLED attribute should be set to False for IEs that
     are disabled by default and must be explicitly enabled.
 
-    The _WORKING attribute should be set to False for broken IEs
+    For broken extractors, the _REPORTED_BROKEN attribute can be set to the issue URL
     in order to warn the users and skip the tests.
+    [Deprecated] If there is no open issue, set _WORKING = False instead.
     """
 
     _ready = False
@@ -613,10 +614,13 @@ class InfoExtractor:
         except (IndexError, AttributeError):
             return None
 
+    @classproperty(cache=True)
+    def _REPORTED_BROKEN(cls):
+        return not cls._WORKING and ''
+
     @classmethod
     def working(cls):
-        """Getter method for _WORKING."""
-        return cls._WORKING
+        return cls._REPORTED_BROKEN is False
 
     @classmethod
     def supports_login(cls):
@@ -3674,7 +3678,12 @@ class InfoExtractor:
                 _COUNTS = ('', '5', '10', 'all')
                 desc += f' (e.g. "{cls.SEARCH_KEY}{random.choice(_COUNTS)}:{random.choice(search_examples)}")'
         if not cls.working():
-            desc += ' (**Currently broken**)' if markdown else ' (Currently broken)'
+            msg = 'Currently broken'
+            if markdown:
+                msg = f'**{msg}**'
+                if cls._REPORTED_BROKEN:
+                    msg = f'[{msg}]({cls._REPORTED_BROKEN})'
+            desc += f' ({msg})'
 
         # Escape emojis. Ref: https://github.com/github/markup/issues/1153
         name = (' - **%s**' % re.sub(r':(\w+:)', ':\u200B\\g<1>', cls.IE_NAME)) if markdown else cls.IE_NAME

yt_dlp/extractor/crunchyroll.py

@@ -21,6 +21,7 @@ from ..utils import (
 
 
 class CrunchyrollBaseIE(InfoExtractor):
+    _REPORTED_BROKEN = 'https://github.com/yt-dlp/yt-dlp/issues/9453'
     _BASE_URL = 'https://www.crunchyroll.com'
     _API_BASE = 'https://api.crunchyroll.com'
     _NETRC_MACHINE = 'crunchyroll'

yt_dlp/extractor/hidive.py

@@ -9,6 +9,7 @@ from ..utils import (
 
 
 class HiDiveIE(InfoExtractor):
+    _REPORTED_BROKEN = 'https://github.com/yt-dlp/yt-dlp/issues/9385'
     _VALID_URL = r'https?://(?:www\.)?hidive\.com/stream/(?P<id>(?P<title>[^/]+)/(?P<key>[^/?#&]+))'
     # Using X-Forwarded-For results in 403 HTTP error for HLS fragments,
     # so disabling geo bypass completely