mirror of
https://github.com/yt-dlp/yt-dlp.git
synced 2025-12-16 04:58:56 +00:00
Merge branch 'master' into yt-live-from-start-range
This commit is contained in:
bashonly
@@ -247,6 +247,8 @@ class InfoExtractor:
|
||||
(For internal use only)
|
||||
* http_chunk_size Chunk size for HTTP downloads
|
||||
* ffmpeg_args Extra arguments for ffmpeg downloader
|
||||
* is_dash_periods Whether the format is a result of merging
|
||||
multiple DASH periods.
|
||||
RTMP formats can also have the additional fields: page_url,
|
||||
app, play_path, tc_url, flash_version, rtmp_live, rtmp_conn,
|
||||
rtmp_protocol, rtmp_real_time
|
||||
@@ -278,7 +280,7 @@ class InfoExtractor:
|
||||
description: Full video description.
|
||||
uploader: Full name of the video uploader.
|
||||
license: License name the video is licensed under.
|
||||
creator: The creator of the video.
|
||||
creators: List of creators of the video.
|
||||
timestamp: UNIX timestamp of the moment the video was uploaded
|
||||
upload_date: Video upload date in UTC (YYYYMMDD).
|
||||
If not explicitly set, calculated from timestamp
|
||||
@@ -422,16 +424,16 @@ class InfoExtractor:
|
||||
track_number: Number of the track within an album or a disc, as an integer.
|
||||
track_id: Id of the track (useful in case of custom indexing, e.g. 6.iii),
|
||||
as a unicode string.
|
||||
artist: Artist(s) of the track.
|
||||
genre: Genre(s) of the track.
|
||||
artists: List of artists of the track.
|
||||
composers: List of composers of the piece.
|
||||
genres: List of genres of the track.
|
||||
album: Title of the album the track belongs to.
|
||||
album_type: Type of the album (e.g. "Demo", "Full-length", "Split", "Compilation", etc).
|
||||
album_artist: List of all artists appeared on the album (e.g.
|
||||
"Ash Borer / Fell Voices" or "Various Artists", useful for splits
|
||||
and compilations).
|
||||
album_artists: List of all artists appeared on the album.
|
||||
E.g. ["Ash Borer", "Fell Voices"] or ["Various Artists"].
|
||||
Useful for splits and compilations.
|
||||
disc_number: Number of the disc or other physical medium the track belongs to,
|
||||
as an integer.
|
||||
composer: Composer of the piece
|
||||
|
||||
The following fields should only be set for clips that should be cut from the original video:
|
||||
|
||||
@@ -442,6 +444,18 @@ class InfoExtractor:
|
||||
rows: Number of rows in each storyboard fragment, as an integer
|
||||
columns: Number of columns in each storyboard fragment, as an integer
|
||||
|
||||
The following fields are deprecated and should not be set by new code:
|
||||
composer: Use "composers" instead.
|
||||
Composer(s) of the piece, comma-separated.
|
||||
artist: Use "artists" instead.
|
||||
Artist(s) of the track, comma-separated.
|
||||
genre: Use "genres" instead.
|
||||
Genre(s) of the track, comma-separated.
|
||||
album_artist: Use "album_artists" instead.
|
||||
All artists appeared on the album, comma-separated.
|
||||
creator: Use "creators" instead.
|
||||
The creator of the video.
|
||||
|
||||
Unless mentioned otherwise, the fields should be Unicode strings.
|
||||
|
||||
Unless mentioned otherwise, None is equivalent to absence of information.
|
||||
@@ -2530,7 +2544,11 @@ class InfoExtractor:
|
||||
self._report_ignoring_subs('DASH')
|
||||
return fmts
|
||||
|
||||
def _extract_mpd_formats_and_subtitles(
|
||||
def _extract_mpd_formats_and_subtitles(self, *args, **kwargs):
|
||||
periods = self._extract_mpd_periods(*args, **kwargs)
|
||||
return self._merge_mpd_periods(periods)
|
||||
|
||||
def _extract_mpd_periods(
|
||||
self, mpd_url, video_id, mpd_id=None, note=None, errnote=None,
|
||||
fatal=True, data=None, headers={}, query={}):
|
||||
|
||||
@@ -2543,17 +2561,16 @@ class InfoExtractor:
|
||||
errnote='Failed to download MPD manifest' if errnote is None else errnote,
|
||||
fatal=fatal, data=data, headers=headers, query=query)
|
||||
if res is False:
|
||||
return [], {}
|
||||
return []
|
||||
mpd_doc, urlh = res
|
||||
if mpd_doc is None:
|
||||
return [], {}
|
||||
return []
|
||||
|
||||
# We could have been redirected to a new url when we retrieved our mpd file.
|
||||
mpd_url = urlh.url
|
||||
mpd_base_url = base_url(mpd_url)
|
||||
|
||||
return self._parse_mpd_formats_and_subtitles(
|
||||
mpd_doc, mpd_id, mpd_base_url, mpd_url)
|
||||
return self._parse_mpd_periods(mpd_doc, mpd_id, mpd_base_url, mpd_url)
|
||||
|
||||
def _parse_mpd_formats(self, *args, **kwargs):
|
||||
fmts, subs = self._parse_mpd_formats_and_subtitles(*args, **kwargs)
|
||||
@@ -2561,8 +2578,39 @@ class InfoExtractor:
|
||||
self._report_ignoring_subs('DASH')
|
||||
return fmts
|
||||
|
||||
def _parse_mpd_formats_and_subtitles(
|
||||
self, mpd_doc, mpd_id=None, mpd_base_url='', mpd_url=None):
|
||||
def _parse_mpd_formats_and_subtitles(self, *args, **kwargs):
|
||||
periods = self._parse_mpd_periods(*args, **kwargs)
|
||||
return self._merge_mpd_periods(periods)
|
||||
|
||||
def _merge_mpd_periods(self, periods):
|
||||
"""
|
||||
Combine all formats and subtitles from an MPD manifest into a single list,
|
||||
by concatenate streams with similar formats.
|
||||
"""
|
||||
formats, subtitles = {}, {}
|
||||
for period in periods:
|
||||
for f in period['formats']:
|
||||
assert 'is_dash_periods' not in f, 'format already processed'
|
||||
f['is_dash_periods'] = True
|
||||
format_key = tuple(v for k, v in f.items() if k not in (
|
||||
('format_id', 'fragments', 'manifest_stream_number')))
|
||||
if format_key not in formats:
|
||||
formats[format_key] = f
|
||||
elif 'fragments' in f:
|
||||
formats[format_key].setdefault('fragments', []).extend(f['fragments'])
|
||||
|
||||
if subtitles and period['subtitles']:
|
||||
self.report_warning(bug_reports_message(
|
||||
'Found subtitles in multiple periods in the DASH manifest; '
|
||||
'if part of the subtitles are missing,'
|
||||
), only_once=True)
|
||||
|
||||
for sub_lang, sub_info in period['subtitles'].items():
|
||||
subtitles.setdefault(sub_lang, []).extend(sub_info)
|
||||
|
||||
return list(formats.values()), subtitles
|
||||
|
||||
def _parse_mpd_periods(self, mpd_doc, mpd_id=None, mpd_base_url='', mpd_url=None):
|
||||
"""
|
||||
Parse formats from MPD manifest.
|
||||
References:
|
||||
@@ -2643,14 +2691,17 @@ class InfoExtractor:
|
||||
mpd_duration = parse_duration(mpd_doc.get('mediaPresentationDuration'))
|
||||
availability_start_time = unified_timestamp(
|
||||
mpd_doc.get('availabilityStartTime'), with_milliseconds=True) or 0
|
||||
formats, subtitles = [], {}
|
||||
stream_numbers = collections.defaultdict(int)
|
||||
for period in mpd_doc.findall(_add_ns('Period')):
|
||||
for period_idx, period in enumerate(mpd_doc.findall(_add_ns('Period'))):
|
||||
# segmentIngestTime is completely out of spec, but YT Livestream do this
|
||||
segment_ingest_time = period.get('{http://youtube.com/yt/2012/10/10}segmentIngestTime')
|
||||
if segment_ingest_time:
|
||||
availability_start_time = unified_timestamp(segment_ingest_time, with_milliseconds=True)
|
||||
|
||||
period_entry = {
|
||||
'id': period.get('id', f'period-{period_idx}'),
|
||||
'formats': [],
|
||||
'subtitles': collections.defaultdict(list),
|
||||
}
|
||||
period_duration = parse_duration(period.get('duration')) or mpd_duration
|
||||
period_ms_info = extract_multisegment_info(period, {
|
||||
'start_number': 1,
|
||||
@@ -2908,11 +2959,10 @@ class InfoExtractor:
|
||||
if content_type in ('video', 'audio', 'image/jpeg'):
|
||||
f['manifest_stream_number'] = stream_numbers[f['url']]
|
||||
stream_numbers[f['url']] += 1
|
||||
formats.append(f)
|
||||
period_entry['formats'].append(f)
|
||||
elif content_type == 'text':
|
||||
subtitles.setdefault(lang or 'und', []).append(f)
|
||||
|
||||
return formats, subtitles
|
||||
period_entry['subtitles'][lang or 'und'].append(f)
|
||||
yield period_entry
|
||||
|
||||
def _extract_ism_formats(self, *args, **kwargs):
|
||||
fmts, subs = self._extract_ism_formats_and_subtitles(*args, **kwargs)
|
||||
|
||||
Reference in New Issue
Block a user