[extractor/youtube] Consider language in format de-duplication

This commit is contained in:
pukkandan 2022-11-15 05:23:32 +05:30
parent d7b460d0e5
commit a4894d3e25
No known key found for this signature in database
GPG key ID: 7EEE9E1E817D0A39

View file

@ -1,5 +1,6 @@
import base64 import base64
import calendar import calendar
import collections
import copy import copy
import datetime import datetime
import enum import enum
@ -2480,6 +2481,34 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
'note': '6 channel audio', 'note': '6 channel audio',
'url': 'https://www.youtube.com/watch?v=zgdo7-RRjgo', 'url': 'https://www.youtube.com/watch?v=zgdo7-RRjgo',
'only_matching': True, 'only_matching': True,
}, {
'note': 'Multiple HLS formats with same itag',
'url': 'https://www.youtube.com/watch?v=kX3nB4PpJko',
'info_dict': {
'id': 'kX3nB4PpJko',
'ext': 'mp4',
'categories': ['Entertainment'],
'description': 'md5:e8031ff6e426cdb6a77670c9b81f6fa6',
'uploader_url': 'http://www.youtube.com/user/MrBeast6000',
'live_status': 'not_live',
'duration': 937,
'channel_follower_count': int,
'thumbnail': 'https://i.ytimg.com/vi_webp/kX3nB4PpJko/maxresdefault.webp',
'title': 'Last To Take Hand Off Jet, Keeps It!',
'channel': 'MrBeast',
'playable_in_embed': True,
'view_count': int,
'upload_date': '20221112',
'uploader': 'MrBeast',
'uploader_id': 'MrBeast6000',
'channel_url': 'https://www.youtube.com/channel/UCX6OQ3DkcsbYNE6H8uQQuVA',
'age_limit': 0,
'availability': 'public',
'channel_id': 'UCX6OQ3DkcsbYNE6H8uQQuVA',
'like_count': int,
'tags': [],
},
'params': {'extractor_args': {'youtube': {'player_client': ['ios']}}, 'format': '233-1'},
} }
] ]
@ -3472,7 +3501,7 @@ def _needs_live_processing(self, live_status, duration):
return live_status return live_status
def _extract_formats_and_subtitles(self, streaming_data, video_id, player_url, live_status, duration): def _extract_formats_and_subtitles(self, streaming_data, video_id, player_url, live_status, duration):
itags, stream_ids = {}, [] itags, stream_ids = collections.defaultdict(set), []
itag_qualities, res_qualities = {}, {0: None} itag_qualities, res_qualities = {}, {0: None}
q = qualities([ q = qualities([
# Normally tiny is the smallest video-only formats. But # Normally tiny is the smallest video-only formats. But
@ -3554,10 +3583,6 @@ def _extract_formats_and_subtitles(self, streaming_data, video_id, player_url, l
video_id=video_id, only_once=True) video_id=video_id, only_once=True)
throttled = True throttled = True
if itag:
itags[itag] = 'https'
stream_ids.append(stream_id)
tbr = float_or_none(fmt.get('averageBitrate') or fmt.get('bitrate'), 1000) tbr = float_or_none(fmt.get('averageBitrate') or fmt.get('bitrate'), 1000)
language_preference = ( language_preference = (
10 if audio_track.get('audioIsDefault') and 10 10 if audio_track.get('audioIsDefault') and 10
@ -3616,6 +3641,10 @@ def _extract_formats_and_subtitles(self, streaming_data, video_id, player_url, l
} }
if dct.get('ext'): if dct.get('ext'):
dct['container'] = dct['ext'] + '_dash' dct['container'] = dct['ext'] + '_dash'
if itag:
itags[itag].add(('https', dct.get('language')))
stream_ids.append(stream_id)
yield dct yield dct
needs_live_processing = self._needs_live_processing(live_status, duration) needs_live_processing = self._needs_live_processing(live_status, duration)
@ -3636,13 +3665,15 @@ def _extract_formats_and_subtitles(self, streaming_data, video_id, player_url, l
skip_manifests.add('dash') skip_manifests.add('dash')
def process_manifest_format(f, proto, itag): def process_manifest_format(f, proto, itag):
if itag in itags: key = (proto, f.get('language'))
if itags[itag] == proto or f'{itag}-{proto}' in itags: if key in itags[itag]:
return False return False
itag = f'{itag}-{proto}' itags[itag].add(key)
if itag:
if any(p != proto for p, _ in itags[itag]):
f['format_id'] = f'{itag}-{proto}'
elif itag:
f['format_id'] = itag f['format_id'] = itag
itags[itag] = proto
f['quality'] = q(itag_qualities.get(try_get(f, lambda f: f['format_id'].split('-')[0]), -1)) f['quality'] = q(itag_qualities.get(try_get(f, lambda f: f['format_id'].split('-')[0]), -1))
if f['quality'] == -1 and f.get('height'): if f['quality'] == -1 and f.get('height'):