[extractor/youtube] Construct fragment list lazily

Ref: yt-dlp/yt-dlp/commit/e389d17
See: yt-dlp/yt-dlp#6517
This commit is contained in:
pukkandan 2023-03-12 14:46:09 +05:30 committed by dirkf
parent f35b757c82
commit 88f28f620b

View file

@ -31,6 +31,7 @@ from ..utils import (
get_element_by_attribute, get_element_by_attribute,
int_or_none, int_or_none,
js_to_json, js_to_json,
LazyList,
merge_dicts, merge_dicts,
mimetype2ext, mimetype2ext,
parse_codecs, parse_codecs,
@ -1986,9 +1987,19 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
itags = [] itags = []
itag_qualities = {} itag_qualities = {}
q = qualities(['tiny', 'small', 'medium', 'large', 'hd720', 'hd1080', 'hd1440', 'hd2160', 'hd2880', 'highres']) q = qualities(['tiny', 'small', 'medium', 'large', 'hd720', 'hd1080', 'hd1440', 'hd2160', 'hd2880', 'highres'])
CHUNK_SIZE = 10 << 20
streaming_data = player_response.get('streamingData') or {} streaming_data = player_response.get('streamingData') or {}
streaming_formats = streaming_data.get('formats') or [] streaming_formats = streaming_data.get('formats') or []
streaming_formats.extend(streaming_data.get('adaptiveFormats') or []) streaming_formats.extend(streaming_data.get('adaptiveFormats') or [])
def build_fragments(f):
return LazyList({
'url': update_url_query(f['url'], {
'range': '{0}-{1}'.format(range_start, min(range_start + CHUNK_SIZE - 1, f['filesize']))
})
} for range_start in range(0, f['filesize'], CHUNK_SIZE))
for fmt in streaming_formats: for fmt in streaming_formats:
if fmt.get('targetDurationSec') or fmt.get('drmFamilies'): if fmt.get('targetDurationSec') or fmt.get('drmFamilies'):
continue continue
@ -2048,15 +2059,10 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
if no_video: if no_video:
dct['abr'] = tbr dct['abr'] = tbr
if no_audio or no_video: if no_audio or no_video:
CHUNK_SIZE = 10 << 20
# avoid Youtube throttling # avoid Youtube throttling
dct.update({ dct.update({
'protocol': 'http_dash_segments', 'protocol': 'http_dash_segments',
'fragments': [{ 'fragments': build_fragments(dct),
'url': update_url_query(dct['url'], {
'range': '{0}-{1}'.format(range_start, min(range_start + CHUNK_SIZE - 1, dct['filesize']))
})
} for range_start in range(0, dct['filesize'], CHUNK_SIZE)]
} if dct['filesize'] else { } if dct['filesize'] else {
'downloader_options': {'http_chunk_size': CHUNK_SIZE} # No longer useful? 'downloader_options': {'http_chunk_size': CHUNK_SIZE} # No longer useful?
}) })