kumi/yt-dlp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity

[ExtractAudio, cleanup] Refactor

Browse source
This commit is contained in:
pukkandan 2022-06-06 21:49:57 +05:30
parent b7c47b7438
commit 35faefee5d
No known key found for this signature in database
GPG key ID: 7EEE9E1E817D0A39
5 changed files with 57 additions and 92 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

13
README.md
View file

@ -871,23 +871,22 @@ You can also fork the project on github and run your fork's [build workflow](.gi
## Post-Processing Options:
-x, --extract-audio Convert video files to audio-only files
(requires ffmpeg and ffprobe)
--audio-format FORMAT Specify audio format to convert the audio to
when -x is used. Currently supported formats
are: best (default) or one of aac, flac,
mp3, m4a, opus, vorbis, wav, alac
--audio-format FORMAT Format to convert the audio to when -x is
used. (currently supported: best (default),
mp3, aac, m4a, opus, vorbis, flac, alac, wav)
--audio-quality QUALITY Specify ffmpeg audio quality to use when
converting the audio with -x. Insert a value
between 0 (best) and 10 (worst) for VBR or a
specific bitrate like 128K (default 5)
--remux-video FORMAT Remux the video into another container if
necessary (currently supported: mp4, mkv,
flv, webm, mov, avi, mka, ogg, aac, flac,
mp3, m4a, opus, vorbis, wav, alac). If
flv, webm, mov, avi, mka, ogg, mp3, aac,
m4a, opus, vorbis, flac, alac, wav). If
target container does not support the
video/audio codec, remuxing will fail. You
can specify multiple rules; Eg.
"aac>m4a/mov>mp4/mkv" will remux aac to m4a,
mov to mp4 and anything else to mkv.
mov to mp4 and anything else to mkv
--recode-video FORMAT Re-encode the video into another format if
necessary. The syntax and supported formats
are the same as --remux-video

12
yt_dlp/__init__.py
View file

@ -215,13 +215,9 @@ def validate_options(opts):
# Postprocessor formats
validate_in('audio format', opts.audioformat, ['best'] + list(FFmpegExtractAudioPP.SUPPORTED_EXTS))
validate_in('subtitle format', opts.convertsubtitles, FFmpegSubtitlesConvertorPP.SUPPORTED_EXTS)
for name, value, pp in (
('thumbnail format', opts.convertthumbnails, FFmpegThumbnailsConvertorPP),
('recode video format', opts.recodevideo, FFmpegVideoConvertorPP),
('remux video format', opts.remuxvideo, FFmpegVideoRemuxerPP),
):
if value is not None:
validate_regex(name, value.replace(' ', ''), pp.FORMAT_RE)
validate_regex('thumbnail format', opts.convertthumbnails, FFmpegThumbnailsConvertorPP.FORMAT_RE)
validate_regex('recode video format', opts.recodevideo, FFmpegVideoConvertorPP.FORMAT_RE)
validate_regex('remux video format', opts.remuxvideo, FFmpegVideoRemuxerPP.FORMAT_RE)
if opts.audioquality:
opts.audioquality = opts.audioquality.strip('k').strip('K')
# int_or_none prevents inf, nan
@ -653,7 +649,7 @@ def parse_options(argv=None):
final_ext = (
opts.recodevideo if opts.recodevideo in FFmpegVideoConvertorPP.SUPPORTED_EXTS
else opts.remuxvideo if opts.remuxvideo in FFmpegVideoRemuxerPP.SUPPORTED_EXTS
else opts.audioformat if (opts.extractaudio and opts.audioformat != 'best')
else opts.audioformat if (opts.extractaudio and opts.audioformat in FFmpegExtractAudioPP.SUPPORTED_EXTS)
else None)
return parser, opts, urls, {

16
yt_dlp/options.py
View file

@ -1423,20 +1423,22 @@ def create_parser():
postproc.add_option(
'--audio-format', metavar='FORMAT', dest='audioformat', default='best',
help=(
'Specify audio format to convert the audio to when -x is used. Currently supported formats are: '
'best (default) or one of %s' % ', '.join(FFmpegExtractAudioPP.SUPPORTED_EXTS)))
'Format to convert the audio to when -x is used. '
f'(currently supported: best (default), {", ".join(FFmpegExtractAudioPP.SUPPORTED_EXTS)})'))
postproc.add_option(
'--audio-quality', metavar='QUALITY',
dest='audioquality', default='5',
help='Specify ffmpeg audio quality to use when converting the audio with -x. Insert a value between 0 (best) and 10 (worst) for VBR or a specific bitrate like 128K (default %default)')
help=(
'Specify ffmpeg audio quality to use when converting the audio with -x. '
'Insert a value between 0 (best) and 10 (worst) for VBR or a specific bitrate like 128K (default %default)'))
postproc.add_option(
'--remux-video',
metavar='FORMAT', dest='remuxvideo', default=None,
help=(
'Remux the video into another container if necessary (currently supported: %s). '
'If target container does not support the video/audio codec, remuxing will fail. '
'You can specify multiple rules; Eg. "aac>m4a/mov>mp4/mkv" will remux aac to m4a, mov to mp4 '
'and anything else to mkv.' % ', '.join(FFmpegVideoRemuxerPP.SUPPORTED_EXTS)))
'Remux the video into another container if necessary '
f'(currently supported: {", ".join(FFmpegVideoRemuxerPP.SUPPORTED_EXTS)}). '
'If target container does not support the video/audio codec, remuxing will fail. You can specify multiple rules; '
'Eg. "aac>m4a/mov>mp4/mkv" will remux aac to m4a, mov to mp4 and anything else to mkv'))
postproc.add_option(
'--recode-video',
metavar='FORMAT', dest='recodevideo', default=None,

2
yt_dlp/postprocessor/common.py
View file

@ -216,5 +216,5 @@ class PostProcessor(metaclass=PostProcessorMetaClass):
raise PostProcessingError(f'Unable to communicate with {self.PP_NAME} API: {e}')
class AudioConversionError(PostProcessingError):
class AudioConversionError(PostProcessingError): # Deprecated
pass

106
yt_dlp/postprocessor/ffmpeg.py
View file

@ -6,7 +6,7 @@ import re
import subprocess
import time
from .common import AudioConversionError, PostProcessor
from .common import PostProcessor
from ..compat import functools, imghdr
from ..utils import (
ISO639Utils,
@ -45,19 +45,20 @@ EXT_TO_OUT_FORMATS = {
'vtt': 'webvtt',
}
ACODECS = {
'mp3': 'libmp3lame',
'aac': 'aac',
'flac': 'flac',
'm4a': 'aac',
'opus': 'libopus',
'vorbis': 'libvorbis',
'wav': None,
'alac': None,
# name: (ext, encoder, opts)
'mp3': ('mp3', 'libmp3lame', ()),
'aac': ('m4a', 'aac', ('-f', 'adts')),
'm4a': ('m4a', 'aac', ('-bsf:a', 'aac_adtstoasc')),
'opus': ('opus', 'libopus', ()),
'vorbis': ('ogg', 'libvorbis', ()),
'flac': ('flac', 'flac', ()),
'alac': ('m4a', None, ('-acodec', 'alac')),
'wav': ('wav', None, ('-f', 'wav')),
}
def create_mapping_re(supported):
return re.compile(r'{0}(?:/{0})*$'.format(r'(?:\w+>)?(?:%s)' % '|'.join(supported)))
return re.compile(r'{0}(?:/{0})*$'.format(r'(?:\s*\w+\s*>)?\s*(?:%s)\s*' % '|'.join(supported)))
def resolve_mapping(source, mapping):
@ -424,7 +425,7 @@ class FFmpegPostProcessor(PostProcessor):
class FFmpegExtractAudioPP(FFmpegPostProcessor):
COMMON_AUDIO_EXTS = ('wav', 'flac', 'm4a', 'aiff', 'mp3', 'ogg', 'mka', 'opus', 'wma')
SUPPORTED_EXTS = ('aac', 'flac', 'mp3', 'm4a', 'opus', 'vorbis', 'wav', 'alac')
SUPPORTED_EXTS = tuple(ACODECS.keys())
def __init__(self, downloader=None, preferredcodec=None, preferredquality=None, nopostoverwrites=False):
FFmpegPostProcessor.__init__(self, downloader)
@ -463,71 +464,45 @@ class FFmpegExtractAudioPP(FFmpegPostProcessor):
try:
FFmpegPostProcessor.run_ffmpeg(self, path, out_path, opts)
except FFmpegPostProcessorError as err:
raise AudioConversionError(err.msg)
raise PostProcessingError(f'audio conversion failed: {err.msg}')
@PostProcessor._restrict_to(images=False)
def run(self, information):
orig_path = path = information['filepath']
orig_ext = information['ext']
if self._preferredcodec == 'best' and orig_ext in self.COMMON_AUDIO_EXTS:
self.to_screen('Skipping audio extraction since the file is already in a common audio format')
target_format = self._preferredcodec
if target_format == 'best' and information['ext'] in self.COMMON_AUDIO_EXTS:
self.to_screen(f'Not converting audio {orig_path}; the file is already in a common audio format')
return [], information
filecodec = self.get_audio_codec(path)
if filecodec is None:
raise PostProcessingError('WARNING: unable to obtain file audio codec with ffprobe')
more_opts = []
if self._preferredcodec == 'best' or self._preferredcodec == filecodec or (self._preferredcodec == 'm4a' and filecodec == 'aac'):
if filecodec == 'aac' and self._preferredcodec in ['m4a', 'best']:
# Lossless, but in another container
acodec = 'copy'
extension = 'm4a'
more_opts = ['-bsf:a', 'aac_adtstoasc']
elif filecodec in ['aac', 'flac', 'mp3', 'vorbis', 'opus']:
# Lossless if possible
acodec = 'copy'
extension = filecodec
if filecodec == 'aac':
more_opts = ['-f', 'adts']
if filecodec == 'vorbis':
extension = 'ogg'
elif filecodec == 'alac':
acodec = None
extension = 'm4a'
more_opts += ['-acodec', 'alac']
else:
# MP3 otherwise.
acodec = 'libmp3lame'
extension = 'mp3'
more_opts = self._quality_args(acodec)
if filecodec == 'aac' and target_format in ('m4a', 'best'):
# Lossless, but in another container
extension, _, more_opts, acodec = *ACODECS['m4a'], 'copy'
elif target_format == 'best' or target_format == filecodec:
# Lossless if possible
try:
extension, _, more_opts, acodec = *ACODECS[filecodec], 'copy'
except KeyError:
extension, acodec, more_opts = ACODECS['mp3']
else:
# We convert the audio (lossy if codec is lossy)
acodec = ACODECS[self._preferredcodec]
extension, acodec, more_opts = ACODECS[target_format]
if acodec == 'aac' and self._features.get('fdk'):
acodec = 'libfdk_aac'
extension = self._preferredcodec
more_opts = self._quality_args(acodec)
if self._preferredcodec == 'aac':
more_opts += ['-f', 'adts']
elif self._preferredcodec == 'm4a':
more_opts += ['-bsf:a', 'aac_adtstoasc']
elif self._preferredcodec == 'vorbis':
extension = 'ogg'
elif self._preferredcodec == 'wav':
extension = 'wav'
more_opts += ['-f', 'wav']
elif self._preferredcodec == 'alac':
extension = 'm4a'
more_opts += ['-acodec', 'alac']
acodec, more_opts = 'libfdk_aac', []
prefix, sep, ext = path.rpartition('.') # not os.path.splitext, since the latter does not work on unicode in all setups
temp_path = new_path = prefix + sep + extension
more_opts = list(more_opts)
if acodec != 'copy':
more_opts = self._quality_args(acodec)
# not os.path.splitext, since the latter does not work on unicode in all setups
temp_path = new_path = f'{path.rpartition(".")[0]}.{extension}'
if new_path == path:
if acodec == 'copy':
self.to_screen(f'File is already in target format {self._preferredcodec}, skipping')
self.to_screen(f'Not converting audio {orig_path}; file is already in target format {target_format}')
return [], information
orig_path = prepend_extension(path, 'orig')
temp_path = prepend_extension(path, 'temp')
@ -536,14 +511,8 @@ class FFmpegExtractAudioPP(FFmpegPostProcessor):
self.to_screen('Post-process file %s exists, skipping' % new_path)
return [], information
try:
self.to_screen(f'Destination: {new_path}')
self.run_ffmpeg(path, temp_path, acodec, more_opts)
except AudioConversionError as e:
raise PostProcessingError(
'audio conversion failed: ' + e.msg)
except Exception:
raise PostProcessingError('error running ' + self.basename)
self.to_screen(f'Destination: {new_path}')
self.run_ffmpeg(path, temp_path, acodec, more_opts)
os.replace(path, orig_path)
os.replace(temp_path, new_path)
@ -553,8 +522,7 @@ class FFmpegExtractAudioPP(FFmpegPostProcessor):
# Try to update the date time for extracted audio file.
if information.get('filetime') is not None:
self.try_utime(
new_path, time.time(), information['filetime'],
errnote='Cannot update utime of audio file')
new_path, time.time(), information['filetime'], errnote='Cannot update utime of audio file')
return [orig_path], information