[youtube] Fix error reporting of "Incomplete data"

[yt-dlp.git] / yt_dlp / extractor / youtube.py

diff --git a/yt_dlp/extractor/youtube.py b/yt_dlp/extractor/youtube.py
index 1b4e47b5f92735112a093e8742f6d5d4a5c54f92..12634483e643bac8b4677b124729af7581c9d586 100644 (file)
--- a/yt_dlp/extractor/youtube.py
+++ b/yt_dlp/extractor/youtube.py
@@ -109,7 +109,8 @@
         'INNERTUBE_CONTEXT': {
             'client': {
                 'clientName': 'ANDROID',
-                'clientVersion': '17.28.34',
+                'clientVersion': '17.29.34',
+                'androidSdkVersion': 30
             }
         },
         'INNERTUBE_CONTEXT_CLIENT_NAME': 3,
@@ -120,7 +121,8 @@
         'INNERTUBE_CONTEXT': {
             'client': {
                 'clientName': 'ANDROID_EMBEDDED_PLAYER',
-                'clientVersion': '17.28.34',
+                'clientVersion': '17.29.34',
+                'androidSdkVersion': 30
             },
         },
         'INNERTUBE_CONTEXT_CLIENT_NAME': 55,
@@ -132,6 +134,7 @@
             'client': {
                 'clientName': 'ANDROID_MUSIC',
                 'clientVersion': '5.16.51',
+                'androidSdkVersion': 30
             }
         },
         'INNERTUBE_CONTEXT_CLIENT_NAME': 21,
@@ -143,6 +146,7 @@
             'client': {
                 'clientName': 'ANDROID_CREATOR',
                 'clientVersion': '22.28.100',
+                'androidSdkVersion': 30
             },
         },
         'INNERTUBE_CONTEXT_CLIENT_NAME': 14,
@@ -805,7 +809,7 @@ def _extract_response(self, item_id, query, note='Downloading API JSON', headers
             # Youtube sometimes sends incomplete data
             # See: https://github.com/ytdl-org/youtube-dl/issues/28194
             if not traverse_obj(response, *variadic(check_get_keys)):
-                retry.error = ExtractorError('Incomplete data received')
+                retry.error = ExtractorError('Incomplete data received', expected=True)
                 continue
 
             return response
@@ -2254,6 +2258,10 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
                 'tags': [],
                 'uploader_url': 'http://www.youtube.com/user/nao20010128nao',
             }
+        }, {
+            'note': '6 channel audio',
+            'url': 'https://www.youtube.com/watch?v=zgdo7-RRjgo',
+            'only_matching': True,
         }
     ]
 
@@ -2645,7 +2653,8 @@ def _extract_n_function(self, video_id, player_url):
         if self.get_param('youtube_print_sig_code'):
             self.to_screen(f'Extracted nsig function from {player_id}:\n{func_code[1]}\n')
 
-        return lambda s: jsi.extract_function_from_code(*func_code)([s])
+        func = jsi.extract_function_from_code(*func_code)
+        return lambda s: func([s])
 
     def _extract_signature_timestamp(self, video_id, player_url, ytcfg=None, fatal=False):
         """
@@ -3138,7 +3147,7 @@ def append_client(*client_names):
                 pr_video_id = traverse_obj(pr, ('videoDetails', 'videoId'))
                 if pr_video_id and pr_video_id != video_id:
                     self.report_warning(
-                        f'{client} client returned a player response for "{pr_video_id}" instead of "{video_id}"' + bug_reports_message())
+                        f'Skipping player response from {client} client (got player response for video "{pr_video_id}" instead of "{video_id}")' + bug_reports_message())
                 else:
                     prs.append(pr)
 
@@ -3159,7 +3168,7 @@ def append_client(*client_names):
 
     def _extract_formats_and_subtitles(self, streaming_data, video_id, player_url, is_live, duration):
         itags, stream_ids = {}, []
-        itag_qualities, res_qualities = {}, {}
+        itag_qualities, res_qualities = {}, {0: -1}
         q = qualities([
             # Normally tiny is the smallest video-only formats. But
             # audio-only formats with unknown quality may get tagged as tiny
@@ -3238,9 +3247,9 @@ def _extract_formats_and_subtitles(self, streaming_data, video_id, player_url, i
                 else -10 if 'descriptive' in (audio_track.get('displayName') or '').lower() and -10
                 else -1)
             # Some formats may have much smaller duration than others (possibly damaged during encoding)
-            # Eg: 2-nOtRESiUc Ref: https://github.com/yt-dlp/yt-dlp/issues/2823
+            # E.g. 2-nOtRESiUc Ref: https://github.com/yt-dlp/yt-dlp/issues/2823
             # Make sure to avoid false positives with small duration differences.
-            # Eg: __2ABJjxzNo, ySuUZEjARPY
+            # E.g. __2ABJjxzNo, ySuUZEjARPY
             is_damaged = try_get(fmt, lambda x: float(x['approxDurationMs']) / duration < 500)
             if is_damaged:
                 self.report_warning(
@@ -3253,10 +3262,13 @@ def _extract_formats_and_subtitles(self, streaming_data, video_id, player_url, i
                     '%s%s' % (audio_track.get('displayName') or '',
                               ' (default)' if language_preference > 0 else ''),
                     fmt.get('qualityLabel') or quality.replace('audio_quality_', ''),
+                    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', delim=', '),
                 # Format 22 is likely to be damaged. See https://github.com/yt-dlp/yt-dlp/issues/3372
                 'source_preference': -10 if throttled else -5 if itag == '22' else -1,
                 'fps': int_or_none(fmt.get('fps')) or None,
+                'audio_channels': fmt.get('audioChannels'),
                 'height': height,
                 'quality': q(quality),
                 'has_drm': bool(fmt.get('drmFamilies')),
@@ -3308,10 +3320,9 @@ def process_manifest_format(f, proto, itag):
                 f['format_id'] = itag
                 itags[itag] = proto
 
-            f['quality'] = next((
-                q(qdict[val])
-                for val, qdict in ((f.get('format_id', '').split('-')[0], itag_qualities), (f.get('height'), res_qualities))
-                if val in qdict), -1)
+            f['quality'] = itag_qualities.get(try_get(f, lambda f: f['format_id'].split('-')[0]), -1)
+            if f['quality'] == -1 and f.get('height'):
+                f['quality'] = q(res_qualities[min(res_qualities, key=lambda x: abs(x - f['height']))])
             return True
 
         subtitles = {}
@@ -3577,7 +3588,8 @@ def feed_entry(name):
         formats.extend(self._extract_storyboard(player_responses, duration))
 
         # source_preference is lower for throttled/potentially damaged formats
-        self._sort_formats(formats, ('quality', 'res', 'fps', 'hdr:12', 'source', 'codec:vp9.2', 'lang', 'proto'))
+        self._sort_formats(formats, (
+            'quality', 'res', 'fps', 'hdr:12', 'source', 'vcodec:vp9.2', 'channels', 'acodec', 'lang', 'proto'))
 
         info = {
             'id': video_id,
@@ -5821,7 +5833,7 @@ def _real_extract(self, url):
 
 
 class YoutubeMusicSearchURLIE(YoutubeTabBaseInfoExtractor):
-    IE_DESC = 'YouTube music search URLs with selectable sections (Eg: #songs)'
+    IE_DESC = 'YouTube music search URLs with selectable sections, e.g. #songs'
     IE_NAME = 'youtube:music:search_url'
     _VALID_URL = r'https?://music\.youtube\.com/search\?([^#]+&)?(?:search_query|q)=(?:[^&]+)(?:[&#]|$)'
     _TESTS = [{