Completely change project name to yt-dlp (#85)

* All modules and binary names are changed
* All documentation references changed
* yt-dlp no longer loads youtube-dlc config files
* All URLs changed to point to organization account

Co-authored-by: Pccode66
Co-authored-by: pukkandan

yt_dlp/postprocessor/__init__.py

@@ -0,0 +1,48 @@
+from __future__ import unicode_literals
+
+from .embedthumbnail import EmbedThumbnailPP
+from .ffmpeg import (
+    FFmpegPostProcessor,
+    FFmpegEmbedSubtitlePP,
+    FFmpegExtractAudioPP,
+    FFmpegFixupStretchedPP,
+    FFmpegFixupM3u8PP,
+    FFmpegFixupM4aPP,
+    FFmpegMergerPP,
+    FFmpegMetadataPP,
+    FFmpegVideoConvertorPP,
+    FFmpegVideoRemuxerPP,
+    FFmpegSubtitlesConvertorPP,
+)
+from .xattrpp import XAttrMetadataPP
+from .execafterdownload import ExecAfterDownloadPP
+from .metadatafromfield import MetadataFromFieldPP
+from .metadatafromfield import MetadataFromTitlePP
+from .movefilesafterdownload import MoveFilesAfterDownloadPP
+from .sponskrub import SponSkrubPP
+
+
+def get_postprocessor(key):
+    return globals()[key + 'PP']
+
+
+__all__ = [
+    'EmbedThumbnailPP',
+    'ExecAfterDownloadPP',
+    'FFmpegEmbedSubtitlePP',
+    'FFmpegExtractAudioPP',
+    'FFmpegFixupM3u8PP',
+    'FFmpegFixupM4aPP',
+    'FFmpegFixupStretchedPP',
+    'FFmpegMergerPP',
+    'FFmpegMetadataPP',
+    'FFmpegPostProcessor',
+    'FFmpegSubtitlesConvertorPP',
+    'FFmpegVideoConvertorPP',
+    'FFmpegVideoRemuxerPP',
+    'MetadataFromFieldPP',
+    'MetadataFromTitlePP',
+    'MoveFilesAfterDownloadPP',
+    'SponSkrubPP',
+    'XAttrMetadataPP',
+]

yt_dlp/postprocessor/common.py

@@ -0,0 +1,101 @@
+from __future__ import unicode_literals
+
+import os
+
+from ..compat import compat_str
+from ..utils import (
+    cli_configuration_args,
+    encodeFilename,
+    PostProcessingError,
+)
+
+
+class PostProcessor(object):
+    """Post Processor class.
+
+    PostProcessor objects can be added to downloaders with their
+    add_post_processor() method. When the downloader has finished a
+    successful download, it will take its internal chain of PostProcessors
+    and start calling the run() method on each one of them, first with
+    an initial argument and then with the returned value of the previous
+    PostProcessor.
+
+    The chain will be stopped if one of them ever returns None or the end
+    of the chain is reached.
+
+    PostProcessor objects follow a "mutual registration" process similar
+    to InfoExtractor objects.
+
+    Optionally PostProcessor can use a list of additional command-line arguments
+    with self._configuration_args.
+    """
+
+    _downloader = None
+
+    def __init__(self, downloader=None):
+        self._downloader = downloader
+        self.PP_NAME = self.pp_key()
+
+    @classmethod
+    def pp_key(cls):
+        name = cls.__name__[:-2]
+        return compat_str(name[6:]) if name[:6].lower() == 'ffmpeg' else name
+
+    def to_screen(self, text, prefix=True, *args, **kwargs):
+        tag = '[%s] ' % self.PP_NAME if prefix else ''
+        if self._downloader:
+            return self._downloader.to_screen('%s%s' % (tag, text), *args, **kwargs)
+
+    def report_warning(self, text, *args, **kwargs):
+        if self._downloader:
+            return self._downloader.report_warning(text, *args, **kwargs)
+
+    def report_error(self, text, *args, **kwargs):
+        if self._downloader:
+            return self._downloader.report_error(text, *args, **kwargs)
+
+    def write_debug(self, text, prefix=True, *args, **kwargs):
+        tag = '[debug] ' if prefix else ''
+        if self.get_param('verbose', False) and self._downloader:
+            return self._downloader.to_screen('%s%s' % (tag, text), *args, **kwargs)
+
+    def get_param(self, name, default=None, *args, **kwargs):
+        if self._downloader:
+            return self._downloader.params.get(name, default, *args, **kwargs)
+        return default
+
+    def set_downloader(self, downloader):
+        """Sets the downloader for this PP."""
+        self._downloader = downloader
+
+    def run(self, information):
+        """Run the PostProcessor.
+
+        The "information" argument is a dictionary like the ones
+        composed by InfoExtractors. The only difference is that this
+        one has an extra field called "filepath" that points to the
+        downloaded file.
+
+        This method returns a tuple, the first element is a list of the files
+        that can be deleted, and the second of which is the updated
+        information.
+
+        In addition, this method may raise a PostProcessingError
+        exception if post processing fails.
+        """
+        return [], information  # by default, keep file and do nothing
+
+    def try_utime(self, path, atime, mtime, errnote='Cannot update utime of file'):
+        try:
+            os.utime(encodeFilename(path), (atime, mtime))
+        except Exception:
+            self.report_warning(errnote)
+
+    def _configuration_args(self, *args, **kwargs):
+        return cli_configuration_args(
+            self._downloader.params.get('postprocessor_args'),
+            self.pp_key().lower(), *args, **kwargs)
+
+
+class AudioConversionError(PostProcessingError):
+    pass

yt_dlp/postprocessor/embedthumbnail.py

@@ -0,0 +1,196 @@
+# coding: utf-8
+from __future__ import unicode_literals
+
+
+import os
+import subprocess
+import struct
+import re
+import base64
+
+try:
+    import mutagen
+    has_mutagen = True
+except ImportError:
+    has_mutagen = False
+
+from .ffmpeg import FFmpegPostProcessor
+
+from ..utils import (
+    check_executable,
+    encodeArgument,
+    encodeFilename,
+    error_to_compat_str,
+    PostProcessingError,
+    prepend_extension,
+    process_communicate_or_kill,
+    replace_extension,
+    shell_quote,
+)
+
+
+class EmbedThumbnailPPError(PostProcessingError):
+    pass
+
+
+class EmbedThumbnailPP(FFmpegPostProcessor):
+
+    def __init__(self, downloader=None, already_have_thumbnail=False):
+        super(EmbedThumbnailPP, self).__init__(downloader)
+        self._already_have_thumbnail = already_have_thumbnail
+
+    def run(self, info):
+        filename = info['filepath']
+        temp_filename = prepend_extension(filename, 'temp')
+
+        if not info.get('thumbnails'):
+            self.to_screen('There aren\'t any thumbnails to embed')
+            return [], info
+
+        original_thumbnail = thumbnail_filename = info['thumbnails'][-1]['filename']
+
+        if not os.path.exists(encodeFilename(thumbnail_filename)):
+            self.report_warning('Skipping embedding the thumbnail because the file is missing.')
+            return [], info
+
+        def is_webp(path):
+            with open(encodeFilename(path), 'rb') as f:
+                b = f.read(12)
+            return b[0:4] == b'RIFF' and b[8:] == b'WEBP'
+
+        # Correct extension for WebP file with wrong extension (see #25687, #25717)
+        _, thumbnail_ext = os.path.splitext(thumbnail_filename)
+        if thumbnail_ext:
+            thumbnail_ext = thumbnail_ext[1:].lower()
+            if thumbnail_ext != 'webp' and is_webp(thumbnail_filename):
+                self.to_screen('Correcting extension to webp and escaping path for thumbnail "%s"' % thumbnail_filename)
+                thumbnail_webp_filename = replace_extension(thumbnail_filename, 'webp')
+                os.rename(encodeFilename(thumbnail_filename), encodeFilename(thumbnail_webp_filename))
+                original_thumbnail = thumbnail_filename = thumbnail_webp_filename
+                thumbnail_ext = 'webp'
+
+        # Convert unsupported thumbnail formats to JPEG (see #25687, #25717)
+        if thumbnail_ext not in ['jpg', 'png']:
+            # NB: % is supposed to be escaped with %% but this does not work
+            # for input files so working around with standard substitution
+            escaped_thumbnail_filename = thumbnail_filename.replace('%', '#')
+            os.rename(encodeFilename(thumbnail_filename), encodeFilename(escaped_thumbnail_filename))
+            escaped_thumbnail_jpg_filename = replace_extension(escaped_thumbnail_filename, 'jpg')
+            self.to_screen('Converting thumbnail "%s" to JPEG' % escaped_thumbnail_filename)
+            self.run_ffmpeg(escaped_thumbnail_filename, escaped_thumbnail_jpg_filename, ['-bsf:v', 'mjpeg2jpeg'])
+            thumbnail_jpg_filename = replace_extension(thumbnail_filename, 'jpg')
+            # Rename back to unescaped for further processing
+            os.rename(encodeFilename(escaped_thumbnail_filename), encodeFilename(thumbnail_filename))
+            os.rename(encodeFilename(escaped_thumbnail_jpg_filename), encodeFilename(thumbnail_jpg_filename))
+            thumbnail_filename = thumbnail_jpg_filename
+            thumbnail_ext = 'jpg'
+
+        success = True
+        if info['ext'] == 'mp3':
+            options = [
+                '-c', 'copy', '-map', '0:0', '-map', '1:0', '-id3v2_version', '3',
+                '-metadata:s:v', 'title="Album cover"', '-metadata:s:v', 'comment="Cover (front)"']
+
+            self.to_screen('Adding thumbnail to "%s"' % filename)
+            self.run_ffmpeg_multiple_files([filename, thumbnail_filename], temp_filename, options)
+
+        elif info['ext'] in ['mkv', 'mka']:
+            options = ['-c', 'copy', '-map', '0', '-dn']
+
+            mimetype = 'image/%s' % ('png' if thumbnail_ext == 'png' else 'jpeg')
+            old_stream, new_stream = self.get_stream_number(
+                filename, ('tags', 'mimetype'), mimetype)
+            if old_stream is not None:
+                options.extend(['-map', '-0:%d' % old_stream])
+                new_stream -= 1
+            options.extend([
+                '-attach', thumbnail_filename,
+                '-metadata:s:%d' % new_stream, 'mimetype=%s' % mimetype,
+                '-metadata:s:%d' % new_stream, 'filename=cover.%s' % thumbnail_ext])
+
+            self.to_screen('Adding thumbnail to "%s"' % filename)
+            self.run_ffmpeg(filename, temp_filename, options)
+
+        elif info['ext'] in ['m4a', 'mp4', 'mov']:
+            try:
+                options = ['-c', 'copy', '-map', '0', '-dn', '-map', '1']
+
+                old_stream, new_stream = self.get_stream_number(
+                    filename, ('disposition', 'attached_pic'), 1)
+                if old_stream is not None:
+                    options.extend(['-map', '-0:%d' % old_stream])
+                    new_stream -= 1
+                options.extend(['-disposition:%s' % new_stream, 'attached_pic'])
+
+                self.to_screen('Adding thumbnail to "%s"' % filename)
+                self.run_ffmpeg_multiple_files([filename, thumbnail_filename], temp_filename, options)
+
+            except PostProcessingError as err:
+                self.report_warning('unable to embed using ffprobe & ffmpeg; %s' % error_to_compat_str(err))
+                atomicparsley = next((
+                    x for x in ['AtomicParsley', 'atomicparsley']
+                    if check_executable(x, ['-v'])), None)
+                if atomicparsley is None:
+                    raise EmbedThumbnailPPError('AtomicParsley was not found. Please install.')
+
+                cmd = [encodeFilename(atomicparsley, True),
+                       encodeFilename(filename, True),
+                       encodeArgument('--artwork'),
+                       encodeFilename(thumbnail_filename, True),
+                       encodeArgument('-o'),
+                       encodeFilename(temp_filename, True)]
+                cmd += [encodeArgument(o) for o in self._configuration_args(exe='AtomicParsley')]
+
+                self.to_screen('Adding thumbnail to "%s"' % filename)
+                self.write_debug('AtomicParsley command line: %s' % shell_quote(cmd))
+                p = subprocess.Popen(cmd, stdout=subprocess.PIPE, stderr=subprocess.PIPE)
+                stdout, stderr = process_communicate_or_kill(p)
+                if p.returncode != 0:
+                    msg = stderr.decode('utf-8', 'replace').strip()
+                    raise EmbedThumbnailPPError(msg)
+                # for formats that don't support thumbnails (like 3gp) AtomicParsley
+                # won't create to the temporary file
+                if b'No changes' in stdout:
+                    self.report_warning('The file format doesn\'t support embedding a thumbnail')
+                    success = False
+
+        elif info['ext'] in ['ogg', 'opus']:
+            if not has_mutagen:
+                raise EmbedThumbnailPPError('module mutagen was not found. Please install using `python -m pip install mutagen`')
+            self.to_screen('Adding thumbnail to "%s"' % filename)
+
+            size_regex = r',\s*(?P<w>\d+)x(?P<h>\d+)\s*[,\[]'
+            size_result = self.run_ffmpeg(thumbnail_filename, thumbnail_filename, ['-hide_banner'])
+            mobj = re.search(size_regex, size_result)
+            width, height = int(mobj.group('w')), int(mobj.group('h'))
+            mimetype = ('image/%s' % ('png' if thumbnail_ext == 'png' else 'jpeg')).encode('ascii')
+
+            # https://xiph.org/flac/format.html#metadata_block_picture
+            data = bytearray()
+            data += struct.pack('>II', 3, len(mimetype))
+            data += mimetype
+            data += struct.pack('>IIIIII', 0, width, height, 8, 0, os.stat(thumbnail_filename).st_size)  # 32 if png else 24
+
+            fin = open(thumbnail_filename, "rb")
+            data += fin.read()
+            fin.close()
+
+            temp_filename = filename
+            f = mutagen.File(temp_filename)
+            f.tags['METADATA_BLOCK_PICTURE'] = base64.b64encode(data).decode('ascii')
+            f.save()
+
+        else:
+            raise EmbedThumbnailPPError('Supported filetypes for thumbnail embedding are: mp3, mkv/mka, ogg/opus, m4a/mp4/mov')
+
+        if success and temp_filename != filename:
+            os.remove(encodeFilename(filename))
+            os.rename(encodeFilename(temp_filename), encodeFilename(filename))
+
+        files_to_delete = [thumbnail_filename]
+        if self._already_have_thumbnail:
+            info['__files_to_move'][original_thumbnail] = replace_extension(
+                info['__thumbnail_filename'], os.path.splitext(original_thumbnail)[1][1:])
+            if original_thumbnail == thumbnail_filename:
+                files_to_delete = []
+        return files_to_delete, info

yt_dlp/postprocessor/execafterdownload.py

@@ -0,0 +1,36 @@
+from __future__ import unicode_literals
+
+import subprocess
+
+from .common import PostProcessor
+from ..compat import compat_shlex_quote
+from ..utils import (
+    encodeArgument,
+    PostProcessingError,
+)
+
+
+class ExecAfterDownloadPP(PostProcessor):
+
+    def __init__(self, downloader, exec_cmd):
+        super(ExecAfterDownloadPP, self).__init__(downloader)
+        self.exec_cmd = exec_cmd
+
+    @classmethod
+    def pp_key(cls):
+        return 'Exec'
+
+    def run(self, information):
+        cmd = self.exec_cmd
+        if '{}' not in cmd:
+            cmd += ' {}'
+
+        cmd = cmd.replace('{}', compat_shlex_quote(information['filepath']))
+
+        self.to_screen('Executing command: %s' % cmd)
+        retCode = subprocess.call(encodeArgument(cmd), shell=True)
+        if retCode != 0:
+            raise PostProcessingError(
+                'Command returned error code %d' % retCode)
+
+        return [], information

yt_dlp/postprocessor/ffmpeg.py

@@ -0,0 +1,760 @@
+from __future__ import unicode_literals
+
+import io
+import os
+import subprocess
+import time
+import re
+import json
+
+
+from .common import AudioConversionError, PostProcessor
+
+from ..utils import (
+    encodeArgument,
+    encodeFilename,
+    get_exe_version,
+    is_outdated_version,
+    PostProcessingError,
+    prepend_extension,
+    shell_quote,
+    subtitles_filename,
+    dfxp2srt,
+    ISO639Utils,
+    process_communicate_or_kill,
+    replace_extension,
+    traverse_dict,
+)
+
+
+EXT_TO_OUT_FORMATS = {
+    'aac': 'adts',
+    'flac': 'flac',
+    'm4a': 'ipod',
+    'mka': 'matroska',
+    'mkv': 'matroska',
+    'mpg': 'mpeg',
+    'ogv': 'ogg',
+    'ts': 'mpegts',
+    'wma': 'asf',
+    'wmv': 'asf',
+}
+ACODECS = {
+    'mp3': 'libmp3lame',
+    'aac': 'aac',
+    'flac': 'flac',
+    'm4a': 'aac',
+    'opus': 'libopus',
+    'vorbis': 'libvorbis',
+    'wav': None,
+}
+
+
+class FFmpegPostProcessorError(PostProcessingError):
+    pass
+
+
+class FFmpegPostProcessor(PostProcessor):
+    def __init__(self, downloader=None):
+        PostProcessor.__init__(self, downloader)
+        self._determine_executables()
+
+    def check_version(self):
+        if not self.available:
+            raise FFmpegPostProcessorError('ffmpeg not found. Please install')
+
+        required_version = '10-0' if self.basename == 'avconv' else '1.0'
+        if is_outdated_version(
+                self._versions[self.basename], required_version):
+            warning = 'Your copy of %s is outdated, update %s to version %s or newer if you encounter any errors.' % (
+                self.basename, self.basename, required_version)
+            self.report_warning(warning)
+
+    @staticmethod
+    def get_versions(downloader=None):
+        return FFmpegPostProcessor(downloader)._versions
+
+    def _determine_executables(self):
+        programs = ['avprobe', 'avconv', 'ffmpeg', 'ffprobe']
+        prefer_ffmpeg = True
+
+        def get_ffmpeg_version(path):
+            ver = get_exe_version(path, args=['-version'])
+            if ver:
+                regexs = [
+                    r'(?:\d+:)?([0-9.]+)-[0-9]+ubuntu[0-9.]+$',  # Ubuntu, see [1]
+                    r'n([0-9.]+)$',  # Arch Linux
+                    # 1. http://www.ducea.com/2006/06/17/ubuntu-package-version-naming-explanation/
+                ]
+                for regex in regexs:
+                    mobj = re.match(regex, ver)
+                    if mobj:
+                        ver = mobj.group(1)
+            return ver
+
+        self.basename = None
+        self.probe_basename = None
+
+        self._paths = None
+        self._versions = None
+        if self._downloader:
+            prefer_ffmpeg = self.get_param('prefer_ffmpeg', True)
+            location = self.get_param('ffmpeg_location')
+            if location is not None:
+                if not os.path.exists(location):
+                    self.report_warning(
+                        'ffmpeg-location %s does not exist! '
+                        'Continuing without ffmpeg.' % (location))
+                    self._versions = {}
+                    return
+                elif not os.path.isdir(location):
+                    basename = os.path.splitext(os.path.basename(location))[0]
+                    if basename not in programs:
+                        self.report_warning(
+                            'Cannot identify executable %s, its basename should be one of %s. '
+                            'Continuing without ffmpeg.' %
+                            (location, ', '.join(programs)))
+                        self._versions = {}
+                        return None
+                    location = os.path.dirname(os.path.abspath(location))
+                    if basename in ('ffmpeg', 'ffprobe'):
+                        prefer_ffmpeg = True
+
+                self._paths = dict(
+                    (p, os.path.join(location, p)) for p in programs)
+                self._versions = dict(
+                    (p, get_ffmpeg_version(self._paths[p])) for p in programs)
+        if self._versions is None:
+            self._versions = dict(
+                (p, get_ffmpeg_version(p)) for p in programs)
+            self._paths = dict((p, p) for p in programs)
+
+        if prefer_ffmpeg is False:
+            prefs = ('avconv', 'ffmpeg')
+        else:
+            prefs = ('ffmpeg', 'avconv')
+        for p in prefs:
+            if self._versions[p]:
+                self.basename = p
+                break
+
+        if prefer_ffmpeg is False:
+            prefs = ('avprobe', 'ffprobe')
+        else:
+            prefs = ('ffprobe', 'avprobe')
+        for p in prefs:
+            if self._versions[p]:
+                self.probe_basename = p
+                break
+
+    @property
+    def available(self):
+        return self.basename is not None
+
+    @property
+    def executable(self):
+        return self._paths[self.basename]
+
+    @property
+    def probe_available(self):
+        return self.probe_basename is not None
+
+    @property
+    def probe_executable(self):
+        return self._paths[self.probe_basename]
+
+    def get_audio_codec(self, path):
+        if not self.probe_available and not self.available:
+            raise PostProcessingError('ffprobe and ffmpeg not found. Please install')
+        try:
+            if self.probe_available:
+                cmd = [
+                    encodeFilename(self.probe_executable, True),
+                    encodeArgument('-show_streams')]
+            else:
+                cmd = [
+                    encodeFilename(self.executable, True),
+                    encodeArgument('-i')]
+            cmd.append(encodeFilename(self._ffmpeg_filename_argument(path), True))
+            self.write_debug('%s command line: %s' % (self.basename, shell_quote(cmd)))
+            handle = subprocess.Popen(
+                cmd, stderr=subprocess.PIPE,
+                stdout=subprocess.PIPE, stdin=subprocess.PIPE)
+            stdout_data, stderr_data = process_communicate_or_kill(handle)
+            expected_ret = 0 if self.probe_available else 1
+            if handle.wait() != expected_ret:
+                return None
+        except (IOError, OSError):
+            return None
+        output = (stdout_data if self.probe_available else stderr_data).decode('ascii', 'ignore')
+        if self.probe_available:
+            audio_codec = None
+            for line in output.split('\n'):
+                if line.startswith('codec_name='):
+                    audio_codec = line.split('=')[1].strip()
+                elif line.strip() == 'codec_type=audio' and audio_codec is not None:
+                    return audio_codec
+        else:
+            # Stream #FILE_INDEX:STREAM_INDEX[STREAM_ID](LANGUAGE): CODEC_TYPE: CODEC_NAME
+            mobj = re.search(
+                r'Stream\s*#\d+:\d+(?:\[0x[0-9a-f]+\])?(?:\([a-z]{3}\))?:\s*Audio:\s*([0-9a-z]+)',
+                output)
+            if mobj:
+                return mobj.group(1)
+        return None
+
+    def get_metadata_object(self, path, opts=[]):
+        if self.probe_basename != 'ffprobe':
+            if self.probe_available:
+                self.report_warning('Only ffprobe is supported for metadata extraction')
+            raise PostProcessingError('ffprobe not found. Please install.')
+        self.check_version()
+
+        cmd = [
+            encodeFilename(self.probe_executable, True),
+            encodeArgument('-hide_banner'),
+            encodeArgument('-show_format'),
+            encodeArgument('-show_streams'),
+            encodeArgument('-print_format'),
+            encodeArgument('json'),
+        ]
+
+        cmd += opts
+        cmd.append(encodeFilename(self._ffmpeg_filename_argument(path), True))
+        self.write_debug('ffprobe command line: %s' % shell_quote(cmd))
+        p = subprocess.Popen(cmd, stdout=subprocess.PIPE, stderr=subprocess.PIPE, stdin=subprocess.PIPE)
+        stdout, stderr = p.communicate()
+        return json.loads(stdout.decode('utf-8', 'replace'))
+
+    def get_stream_number(self, path, keys, value):
+        streams = self.get_metadata_object(path)['streams']
+        num = next(
+            (i for i, stream in enumerate(streams) if traverse_dict(stream, keys, casesense=False) == value),
+            None)
+        return num, len(streams)
+
+    def run_ffmpeg_multiple_files(self, input_paths, out_path, opts):
+        self.check_version()
+
+        oldest_mtime = min(
+            os.stat(encodeFilename(path)).st_mtime for path in input_paths)
+
+        cmd = [encodeFilename(self.executable, True), encodeArgument('-y')]
+        # avconv does not have repeat option
+        if self.basename == 'ffmpeg':
+            cmd += [encodeArgument('-loglevel'), encodeArgument('repeat+info')]
+
+        def make_args(file, pre=[], post=[], *args, **kwargs):
+            args = pre + self._configuration_args(*args, **kwargs) + post
+            return (
+                [encodeArgument(o) for o in args]
+                + [encodeFilename(self._ffmpeg_filename_argument(file), True)])
+
+        for i, path in enumerate(input_paths):
+            cmd += make_args(path, post=['-i'], exe='%s_i%d' % (self.basename, i + 1), use_default_arg=False)
+        cmd += make_args(out_path, pre=opts, exe=self.basename)
+
+        self.write_debug('ffmpeg command line: %s' % shell_quote(cmd))
+        p = subprocess.Popen(cmd, stdout=subprocess.PIPE, stderr=subprocess.PIPE, stdin=subprocess.PIPE)
+        stdout, stderr = process_communicate_or_kill(p)
+        if p.returncode != 0:
+            stderr = stderr.decode('utf-8', 'replace').strip()
+            if self.get_param('verbose', False):
+                self.report_error(stderr)
+            raise FFmpegPostProcessorError(stderr.split('\n')[-1])
+        self.try_utime(out_path, oldest_mtime, oldest_mtime)
+        return stderr.decode('utf-8', 'replace')
+
+    def run_ffmpeg(self, path, out_path, opts):
+        return self.run_ffmpeg_multiple_files([path], out_path, opts)
+
+    def _ffmpeg_filename_argument(self, fn):
+        # Always use 'file:' because the filename may contain ':' (ffmpeg
+        # interprets that as a protocol) or can start with '-' (-- is broken in
+        # ffmpeg, see https://ffmpeg.org/trac/ffmpeg/ticket/2127 for details)
+        # Also leave '-' intact in order not to break streaming to stdout.
+        if fn.startswith(('http://', 'https://')):
+            return fn
+        return 'file:' + fn if fn != '-' else fn
+
+
+class FFmpegExtractAudioPP(FFmpegPostProcessor):
+    COMMON_AUDIO_EXTENSIONS = ('wav', 'flac', 'm4a', 'aiff', 'mp3', 'ogg', 'mka', 'opus', 'wma')
+
+    def __init__(self, downloader=None, preferredcodec=None, preferredquality=None, nopostoverwrites=False):
+        FFmpegPostProcessor.__init__(self, downloader)
+        if preferredcodec is None:
+            preferredcodec = 'best'
+        self._preferredcodec = preferredcodec
+        self._preferredquality = preferredquality
+        self._nopostoverwrites = nopostoverwrites
+
+    def run_ffmpeg(self, path, out_path, codec, more_opts):
+        if codec is None:
+            acodec_opts = []
+        else:
+            acodec_opts = ['-acodec', codec]
+        opts = ['-vn'] + acodec_opts + more_opts
+        try:
+            FFmpegPostProcessor.run_ffmpeg(self, path, out_path, opts)
+        except FFmpegPostProcessorError as err:
+            raise AudioConversionError(err.msg)
+
+    def run(self, information):
+        path = information['filepath']
+        orig_ext = information['ext']
+
+        if self._preferredcodec == 'best' and orig_ext in self.COMMON_AUDIO_EXTENSIONS:
+            self.to_screen('Skipping audio extraction since 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'
+            else:
+                # MP3 otherwise.
+                acodec = 'libmp3lame'
+                extension = 'mp3'
+                more_opts = []
+                if self._preferredquality is not None:
+                    if int(self._preferredquality) < 10:
+                        more_opts += ['-q:a', self._preferredquality]
+                    else:
+                        more_opts += ['-b:a', self._preferredquality + 'k']
+        else:
+            # We convert the audio (lossy if codec is lossy)
+            acodec = ACODECS[self._preferredcodec]
+            extension = self._preferredcodec
+            more_opts = []
+            if self._preferredquality is not None:
+                # The opus codec doesn't support the -aq option
+                if int(self._preferredquality) < 10 and extension != 'opus':
+                    more_opts += ['-q:a', self._preferredquality]
+                else:
+                    more_opts += ['-b:a', self._preferredquality + 'k']
+            if self._preferredcodec == 'aac':
+                more_opts += ['-f', 'adts']
+            if self._preferredcodec == 'm4a':
+                more_opts += ['-bsf:a', 'aac_adtstoasc']
+            if self._preferredcodec == 'vorbis':
+                extension = 'ogg'
+            if self._preferredcodec == 'wav':
+                extension = 'wav'
+                more_opts += ['-f', 'wav']
+
+        prefix, sep, ext = path.rpartition('.')  # not os.path.splitext, since the latter does not work on unicode in all setups
+        new_path = prefix + sep + extension
+
+        information['filepath'] = new_path
+        information['ext'] = extension
+
+        # If we download foo.mp3 and convert it to... foo.mp3, then don't delete foo.mp3, silly.
+        if (new_path == path
+                or (self._nopostoverwrites and os.path.exists(encodeFilename(new_path)))):
+            self.to_screen('Post-process file %s exists, skipping' % new_path)
+            return [], information
+
+        try:
+            self.to_screen('Destination: ' + new_path)
+            self.run_ffmpeg(path, new_path, acodec, more_opts)
+        except AudioConversionError as e:
+            raise PostProcessingError(
+                'audio conversion failed: ' + e.msg)
+        except Exception:
+            raise PostProcessingError('error running ' + self.basename)
+
+        # 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')
+
+        return [path], information
+
+
+class FFmpegVideoRemuxerPP(FFmpegPostProcessor):
+    def __init__(self, downloader=None, preferedformat=None):
+        super(FFmpegVideoRemuxerPP, self).__init__(downloader)
+        self._preferedformats = preferedformat.lower().split('/')
+
+    def run(self, information):
+        path = information['filepath']
+        sourceext, targetext = information['ext'].lower(), None
+        for pair in self._preferedformats:
+            kv = pair.split('>')
+            if len(kv) == 1 or kv[0].strip() == sourceext:
+                targetext = kv[-1].strip()
+                break
+
+        _skip_msg = (
+            'could not find a mapping for %s' if not targetext
+            else 'already is in target format %s' if sourceext == targetext
+            else None)
+        if _skip_msg:
+            self.to_screen('Not remuxing media file %s; %s' % (path, _skip_msg % sourceext))
+            return [], information
+
+        options = ['-c', 'copy', '-map', '0', '-dn']
+        if targetext in ['mp4', 'm4a', 'mov']:
+            options.extend(['-movflags', '+faststart'])
+        prefix, sep, oldext = path.rpartition('.')
+        outpath = prefix + sep + targetext
+        self.to_screen('Remuxing video from %s to %s; Destination: %s' % (sourceext, targetext, outpath))
+        self.run_ffmpeg(path, outpath, options)
+        information['filepath'] = outpath
+        information['format'] = targetext
+        information['ext'] = targetext
+        return [path], information
+
+
+class FFmpegVideoConvertorPP(FFmpegPostProcessor):
+    def __init__(self, downloader=None, preferedformat=None):
+        super(FFmpegVideoConvertorPP, self).__init__(downloader)
+        self._preferedformat = preferedformat
+
+    def run(self, information):
+        path = information['filepath']
+        if information['ext'] == self._preferedformat:
+            self.to_screen('Not converting video file %s - already is in target format %s' % (path, self._preferedformat))
+            return [], information
+        options = []
+        if self._preferedformat == 'avi':
+            options.extend(['-c:v', 'libxvid', '-vtag', 'XVID'])
+        prefix, sep, ext = path.rpartition('.')
+        outpath = prefix + sep + self._preferedformat
+        self.to_screen('Converting video from %s to %s, Destination: ' % (information['ext'], self._preferedformat) + outpath)
+        self.run_ffmpeg(path, outpath, options)
+        information['filepath'] = outpath
+        information['format'] = self._preferedformat
+        information['ext'] = self._preferedformat
+        return [path], information
+
+
+class FFmpegEmbedSubtitlePP(FFmpegPostProcessor):
+    def __init__(self, downloader=None, already_have_subtitle=False):
+        super(FFmpegEmbedSubtitlePP, self).__init__(downloader)
+        self._already_have_subtitle = already_have_subtitle
+
+    def run(self, information):
+        if information['ext'] not in ('mp4', 'webm', 'mkv'):
+            self.to_screen('Subtitles can only be embedded in mp4, webm or mkv files')
+            return [], information
+        subtitles = information.get('requested_subtitles')
+        if not subtitles:
+            self.to_screen('There aren\'t any subtitles to embed')
+            return [], information
+
+        filename = information['filepath']
+
+        ext = information['ext']
+        sub_langs = []
+        sub_filenames = []
+        webm_vtt_warn = False
+        mp4_ass_warn = False
+
+        for lang, sub_info in subtitles.items():
+            sub_ext = sub_info['ext']
+            if sub_ext == 'json':
+                self.report_warning('JSON subtitles cannot be embedded')
+            elif ext != 'webm' or ext == 'webm' and sub_ext == 'vtt':
+                sub_langs.append(lang)
+                sub_filenames.append(subtitles_filename(filename, lang, sub_ext, ext))
+            else:
+                if not webm_vtt_warn and ext == 'webm' and sub_ext != 'vtt':
+                    webm_vtt_warn = True
+                    self.report_warning('Only WebVTT subtitles can be embedded in webm files')
+            if not mp4_ass_warn and ext == 'mp4' and sub_ext == 'ass':
+                mp4_ass_warn = True
+                self.report_warning('ASS subtitles cannot be properly embedded in mp4 files; expect issues')
+
+        if not sub_langs:
+            return [], information
+
+        input_files = [filename] + sub_filenames
+
+        opts = [
+            '-c', 'copy', '-map', '0', '-dn',
+            # Don't copy the existing subtitles, we may be running the
+            # postprocessor a second time
+            '-map', '-0:s',
+            # Don't copy Apple TV chapters track, bin_data (see #19042, #19024,
+            # https://trac.ffmpeg.org/ticket/6016)
+            '-map', '-0:d',
+        ]
+        if information['ext'] == 'mp4':
+            opts += ['-c:s', 'mov_text']
+        for (i, lang) in enumerate(sub_langs):
+            opts.extend(['-map', '%d:0' % (i + 1)])
+            lang_code = ISO639Utils.short2long(lang) or lang
+            opts.extend(['-metadata:s:s:%d' % i, 'language=%s' % lang_code])
+
+        temp_filename = prepend_extension(filename, 'temp')
+        self.to_screen('Embedding subtitles in "%s"' % filename)
+        self.run_ffmpeg_multiple_files(input_files, temp_filename, opts)
+        os.remove(encodeFilename(filename))
+        os.rename(encodeFilename(temp_filename), encodeFilename(filename))
+
+        files_to_delete = [] if self._already_have_subtitle else sub_filenames
+        return files_to_delete, information
+
+
+class FFmpegMetadataPP(FFmpegPostProcessor):
+    def run(self, info):
+        metadata = {}
+
+        def add(meta_list, info_list=None):
+            if not info_list:
+                info_list = meta_list
+            if not isinstance(meta_list, (list, tuple)):
+                meta_list = (meta_list,)
+            if not isinstance(info_list, (list, tuple)):
+                info_list = (info_list,)
+            for info_f in info_list:
+                if info.get(info_f) is not None:
+                    for meta_f in meta_list:
+                        metadata[meta_f] = info[info_f]
+                    break
+
+        # See [1-4] for some info on media metadata/metadata supported
+        # by ffmpeg.
+        # 1. https://kdenlive.org/en/project/adding-meta-data-to-mp4-video/
+        # 2. https://wiki.multimedia.cx/index.php/FFmpeg_Metadata
+        # 3. https://kodi.wiki/view/Video_file_tagging
+
+        add('title', ('track', 'title'))
+        add('date', 'upload_date')
+        add(('description', 'comment'), 'description')
+        add('purl', 'webpage_url')
+        add('track', 'track_number')
+        add('artist', ('artist', 'creator', 'uploader', 'uploader_id'))
+        add('genre')
+        add('album')
+        add('album_artist')
+        add('disc', 'disc_number')
+        add('show', 'series')
+        add('season_number')
+        add('episode_id', ('episode', 'episode_id'))
+        add('episode_sort', 'episode_number')
+
+        if not metadata:
+            self.to_screen('There isn\'t any metadata to add')
+            return [], info
+
+        filename = info['filepath']
+        temp_filename = prepend_extension(filename, 'temp')
+        in_filenames = [filename]
+        options = ['-map', '0', '-dn']
+
+        if info['ext'] == 'm4a':
+            options.extend(['-vn', '-acodec', 'copy'])
+        else:
+            options.extend(['-c', 'copy'])
+
+        for (name, value) in metadata.items():
+            options.extend(['-metadata', '%s=%s' % (name, value)])
+
+        chapters = info.get('chapters', [])
+        if chapters:
+            metadata_filename = replace_extension(filename, 'meta')
+            with io.open(metadata_filename, 'wt', encoding='utf-8') as f:
+                def ffmpeg_escape(text):
+                    return re.sub(r'(=|;|#|\\|\n)', r'\\\1', text)
+
+                metadata_file_content = ';FFMETADATA1\n'
+                for chapter in chapters:
+                    metadata_file_content += '[CHAPTER]\nTIMEBASE=1/1000\n'
+                    metadata_file_content += 'START=%d\n' % (chapter['start_time'] * 1000)
+                    metadata_file_content += 'END=%d\n' % (chapter['end_time'] * 1000)
+                    chapter_title = chapter.get('title')
+                    if chapter_title:
+                        metadata_file_content += 'title=%s\n' % ffmpeg_escape(chapter_title)
+                f.write(metadata_file_content)
+                in_filenames.append(metadata_filename)
+                options.extend(['-map_metadata', '1'])
+
+        if '__infojson_filename' in info and info['ext'] in ('mkv', 'mka'):
+            old_stream, new_stream = self.get_stream_number(
+                filename, ('tags', 'mimetype'), 'application/json')
+            if old_stream is not None:
+                options.extend(['-map', '-0:%d' % old_stream])
+                new_stream -= 1
+
+            options.extend([
+                '-attach', info['__infojson_filename'],
+                '-metadata:s:%d' % new_stream, 'mimetype=application/json'
+            ])
+
+        self.to_screen('Adding metadata to \'%s\'' % filename)
+        self.run_ffmpeg_multiple_files(in_filenames, temp_filename, options)
+        if chapters:
+            os.remove(metadata_filename)
+        os.remove(encodeFilename(filename))
+        os.rename(encodeFilename(temp_filename), encodeFilename(filename))
+        return [], info
+
+
+class FFmpegMergerPP(FFmpegPostProcessor):
+    def run(self, info):
+        filename = info['filepath']
+        temp_filename = prepend_extension(filename, 'temp')
+        args = ['-c', 'copy']
+        for (i, fmt) in enumerate(info['requested_formats']):
+            if fmt.get('acodec') != 'none':
+                args.extend(['-map', '%u:a:0' % (i)])
+            if fmt.get('vcodec') != 'none':
+                args.extend(['-map', '%u:v:0' % (i)])
+        self.to_screen('Merging formats into "%s"' % filename)
+        self.run_ffmpeg_multiple_files(info['__files_to_merge'], temp_filename, args)
+        os.rename(encodeFilename(temp_filename), encodeFilename(filename))
+        return info['__files_to_merge'], info
+
+    def can_merge(self):
+        # TODO: figure out merge-capable ffmpeg version
+        if self.basename != 'avconv':
+            return True
+
+        required_version = '10-0'
+        if is_outdated_version(
+                self._versions[self.basename], required_version):
+            warning = ('Your copy of %s is outdated and unable to properly mux separate video and audio files, '
+                       'yt-dlp will download single file media. '
+                       'Update %s to version %s or newer to fix this.') % (
+                           self.basename, self.basename, required_version)
+            self.report_warning(warning)
+            return False
+        return True
+
+
+class FFmpegFixupStretchedPP(FFmpegPostProcessor):
+    def run(self, info):
+        stretched_ratio = info.get('stretched_ratio')
+        if stretched_ratio is None or stretched_ratio == 1:
+            return [], info
+
+        filename = info['filepath']
+        temp_filename = prepend_extension(filename, 'temp')
+
+        options = ['-c', 'copy', '-map', '0', '-dn', '-aspect', '%f' % stretched_ratio]
+        self.to_screen('Fixing aspect ratio in "%s"' % filename)
+        self.run_ffmpeg(filename, temp_filename, options)
+
+        os.remove(encodeFilename(filename))
+        os.rename(encodeFilename(temp_filename), encodeFilename(filename))
+
+        return [], info
+
+
+class FFmpegFixupM4aPP(FFmpegPostProcessor):
+    def run(self, info):
+        if info.get('container') != 'm4a_dash':
+            return [], info
+
+        filename = info['filepath']
+        temp_filename = prepend_extension(filename, 'temp')
+
+        options = ['-c', 'copy', '-map', '0', '-dn', '-f', 'mp4']
+        self.to_screen('Correcting container in "%s"' % filename)
+        self.run_ffmpeg(filename, temp_filename, options)
+
+        os.remove(encodeFilename(filename))
+        os.rename(encodeFilename(temp_filename), encodeFilename(filename))
+
+        return [], info
+
+
+class FFmpegFixupM3u8PP(FFmpegPostProcessor):
+    def run(self, info):
+        filename = info['filepath']
+        if self.get_audio_codec(filename) == 'aac':
+            temp_filename = prepend_extension(filename, 'temp')
+
+            options = ['-c', 'copy', '-map', '0', '-dn', '-f', 'mp4', '-bsf:a', 'aac_adtstoasc']
+            self.to_screen('Fixing malformed AAC bitstream in "%s"' % filename)
+            self.run_ffmpeg(filename, temp_filename, options)
+
+            os.remove(encodeFilename(filename))
+            os.rename(encodeFilename(temp_filename), encodeFilename(filename))
+        return [], info
+
+
+class FFmpegSubtitlesConvertorPP(FFmpegPostProcessor):
+    def __init__(self, downloader=None, format=None):
+        super(FFmpegSubtitlesConvertorPP, self).__init__(downloader)
+        self.format = format
+
+    def run(self, info):
+        subs = info.get('requested_subtitles')
+        filename = info['filepath']
+        new_ext = self.format
+        new_format = new_ext
+        if new_format == 'vtt':
+            new_format = 'webvtt'
+        if subs is None:
+            self.to_screen('There aren\'t any subtitles to convert')
+            return [], info
+        self.to_screen('Converting subtitles')
+        sub_filenames = []
+        for lang, sub in subs.items():
+            ext = sub['ext']
+            if ext == new_ext:
+                self.to_screen('Subtitle file for %s is already in the requested format' % new_ext)
+                continue
+            elif ext == 'json':
+                self.to_screen(
+                    'You have requested to convert json subtitles into another format, '
+                    'which is currently not possible')
+                continue
+            old_file = subtitles_filename(filename, lang, ext, info.get('ext'))
+            sub_filenames.append(old_file)
+            new_file = subtitles_filename(filename, lang, new_ext, info.get('ext'))
+
+            if ext in ('dfxp', 'ttml', 'tt'):
+                self.report_warning(
+                    'You have requested to convert dfxp (TTML) subtitles into another format, '
+                    'which results in style information loss')
+
+                dfxp_file = old_file
+                srt_file = subtitles_filename(filename, lang, 'srt', info.get('ext'))
+
+                with open(dfxp_file, 'rb') as f:
+                    srt_data = dfxp2srt(f.read())
+
+                with io.open(srt_file, 'wt', encoding='utf-8') as f:
+                    f.write(srt_data)
+                old_file = srt_file
+
+                subs[lang] = {
+                    'ext': 'srt',
+                    'data': srt_data
+                }
+
+                if new_ext == 'srt':
+                    continue
+                else:
+                    sub_filenames.append(srt_file)
+
+            self.run_ffmpeg(old_file, new_file, ['-f', new_format])
+
+            with io.open(new_file, 'rt', encoding='utf-8') as f:
+                subs[lang] = {
+                    'ext': new_ext,
+                    'data': f.read(),
+                }
+
+        return sub_filenames, info

yt_dlp/postprocessor/metadatafromfield.py

@@ -0,0 +1,71 @@
+from __future__ import unicode_literals
+
+import re
+
+from .common import PostProcessor
+from ..compat import compat_str
+from ..utils import str_or_none
+
+
+class MetadataFromFieldPP(PostProcessor):
+    regex = r'(?P<field>\w+):(?P<format>.+)$'
+
+    def __init__(self, downloader, formats):
+        PostProcessor.__init__(self, downloader)
+        assert isinstance(formats, (list, tuple))
+        self._data = []
+        for f in formats:
+            assert isinstance(f, compat_str)
+            match = re.match(self.regex, f)
+            assert match is not None
+            self._data.append({
+                'field': match.group('field'),
+                'format': match.group('format'),
+                'regex': self.format_to_regex(match.group('format'))})
+
+    def format_to_regex(self, fmt):
+        r"""
+        Converts a string like
+           '%(title)s - %(artist)s'
+        to a regex like
+           '(?P<title>.+)\ \-\ (?P<artist>.+)'
+        """
+        if not re.search(r'%\(\w+\)s', fmt):
+            return fmt
+        lastpos = 0
+        regex = ''
+        # replace %(..)s with regex group and escape other string parts
+        for match in re.finditer(r'%\((\w+)\)s', fmt):
+            regex += re.escape(fmt[lastpos:match.start()])
+            regex += r'(?P<' + match.group(1) + r'>[^\r\n]+)'
+            lastpos = match.end()
+        if lastpos < len(fmt):
+            regex += re.escape(fmt[lastpos:])
+        return regex
+
+    def run(self, info):
+        for dictn in self._data:
+            field, regex = dictn['field'], dictn['regex']
+            if field not in info:
+                self.report_warning('Video doesnot have a %s' % field)
+                continue
+            data_to_parse = str_or_none(info[field])
+            if data_to_parse is None:
+                self.report_warning('Field %s cannot be parsed' % field)
+                continue
+            self.write_debug('Searching for r"%s" in %s' % (regex, field))
+            match = re.search(regex, data_to_parse)
+            if match is None:
+                self.report_warning('Could not interpret video %s as "%s"' % (field, dictn['format']))
+                continue
+            for attribute, value in match.groupdict().items():
+                info[attribute] = value
+                self.to_screen('parsed %s from %s: %s' % (attribute, field, value if value is not None else 'NA'))
+        return [], info
+
+
+class MetadataFromTitlePP(MetadataFromFieldPP):  # for backward compatibility
+    def __init__(self, downloader, titleformat):
+        super(MetadataFromTitlePP, self).__init__(downloader, ['title:%s' % titleformat])
+        self._titleformat = titleformat
+        self._titleregex = self._data[0]['regex']

yt_dlp/postprocessor/movefilesafterdownload.py

@@ -0,0 +1,54 @@
+from __future__ import unicode_literals
+import os
+import shutil
+
+from .common import PostProcessor
+from ..utils import (
+    decodeFilename,
+    encodeFilename,
+    make_dir,
+    PostProcessingError,
+)
+
+
+class MoveFilesAfterDownloadPP(PostProcessor):
+
+    def __init__(self, downloader, files_to_move):
+        PostProcessor.__init__(self, downloader)
+        self.files_to_move = files_to_move
+
+    @classmethod
+    def pp_key(cls):
+        return 'MoveFiles'
+
+    def run(self, info):
+        dl_path, dl_name = os.path.split(encodeFilename(info['filepath']))
+        finaldir = info.get('__finaldir', dl_path)
+        finalpath = os.path.join(finaldir, dl_name)
+        self.files_to_move.update(info['__files_to_move'])
+        self.files_to_move[info['filepath']] = decodeFilename(finalpath)
+
+        make_newfilename = lambda old: decodeFilename(os.path.join(finaldir, os.path.basename(encodeFilename(old))))
+        for oldfile, newfile in self.files_to_move.items():
+            if not newfile:
+                newfile = make_newfilename(oldfile)
+            if os.path.abspath(encodeFilename(oldfile)) == os.path.abspath(encodeFilename(newfile)):
+                continue
+            if not os.path.exists(encodeFilename(oldfile)):
+                self.report_warning('File "%s" cannot be found' % oldfile)
+                continue
+            if os.path.exists(encodeFilename(newfile)):
+                if self.get_param('overwrites', True):
+                    self.report_warning('Replacing existing file "%s"' % newfile)
+                    os.remove(encodeFilename(newfile))
+                else:
+                    self.report_warning(
+                        'Cannot move file "%s" out of temporary directory since "%s" already exists. '
+                        % (oldfile, newfile))
+                    continue
+            make_dir(newfile, PostProcessingError)
+            self.to_screen('Moving file "%s" to "%s"' % (oldfile, newfile))
+            shutil.move(oldfile, newfile)  # os.rename cannot move between volumes
+
+        info['filepath'] = finalpath
+        return [], info

yt_dlp/postprocessor/sponskrub.py

@@ -0,0 +1,93 @@
+from __future__ import unicode_literals
+import os
+import subprocess
+
+from .common import PostProcessor
+from ..compat import compat_shlex_split
+from ..utils import (
+    check_executable,
+    encodeArgument,
+    encodeFilename,
+    shell_quote,
+    str_or_none,
+    PostProcessingError,
+    prepend_extension,
+    process_communicate_or_kill,
+)
+
+
+class SponSkrubPP(PostProcessor):
+    _temp_ext = 'spons'
+    _exe_name = 'sponskrub'
+
+    def __init__(self, downloader, path='', args=None, ignoreerror=False, cut=False, force=False):
+        PostProcessor.__init__(self, downloader)
+        self.force = force
+        self.cutout = cut
+        self.args = str_or_none(args) or ''  # For backward compatibility
+        self.path = self.get_exe(path)
+
+        if not ignoreerror and self.path is None:
+            if path:
+                raise PostProcessingError('sponskrub not found in "%s"' % path)
+            else:
+                raise PostProcessingError('sponskrub not found. Please install or provide the path using --sponskrub-path.')
+
+    def get_exe(self, path=''):
+        if not path or not check_executable(path, ['-h']):
+            path = os.path.join(path, self._exe_name)
+            if not check_executable(path, ['-h']):
+                return None
+        return path
+
+    def run(self, information):
+        if self.path is None:
+            return [], information
+
+        filename = information['filepath']
+        if not os.path.exists(encodeFilename(filename)):  # no download
+            return [], information
+
+        if information['extractor_key'].lower() != 'youtube':
+            self.to_screen('Skipping sponskrub since it is not a YouTube video')
+            return [], information
+        if self.cutout and not self.force and not information.get('__real_download', False):
+            self.report_warning(
+                'Skipping sponskrub since the video was already downloaded. '
+                'Use --sponskrub-force to run sponskrub anyway')
+            return [], information
+
+        self.to_screen('Trying to %s sponsor sections' % ('remove' if self.cutout else 'mark'))
+        if self.cutout:
+            self.report_warning('Cutting out sponsor segments will cause the subtitles to go out of sync.')
+            if not information.get('__real_download', False):
+                self.report_warning('If sponskrub is run multiple times, unintended parts of the video could be cut out.')
+
+        temp_filename = prepend_extension(filename, self._temp_ext)
+        if os.path.exists(encodeFilename(temp_filename)):
+            os.remove(encodeFilename(temp_filename))
+
+        cmd = [self.path]
+        if not self.cutout:
+            cmd += ['-chapter']
+        cmd += compat_shlex_split(self.args)  # For backward compatibility
+        cmd += self._configuration_args(exe=self._exe_name, use_default_arg='no_compat')
+        cmd += ['--', information['id'], filename, temp_filename]
+        cmd = [encodeArgument(i) for i in cmd]
+
+        self.write_debug('sponskrub command line: %s' % shell_quote(cmd))
+        pipe = None if self.get_param('verbose') else subprocess.PIPE
+        p = subprocess.Popen(cmd, stdout=pipe)
+        stdout = process_communicate_or_kill(p)[0]
+
+        if p.returncode == 0:
+            os.remove(encodeFilename(filename))
+            os.rename(encodeFilename(temp_filename), encodeFilename(filename))
+            self.to_screen('Sponsor sections have been %s' % ('removed' if self.cutout else 'marked'))
+        elif p.returncode == 3:
+            self.to_screen('No segments in the SponsorBlock database')
+        else:
+            msg = stdout.decode('utf-8', 'replace').strip() if stdout else ''
+            msg = msg.split('\n')[0 if msg.lower().startswith('unrecognised') else -1]
+            raise PostProcessingError(msg if msg else 'sponskrub failed with error code %s' % p.returncode)
+        return [], information

yt_dlp/postprocessor/xattrpp.py

@@ -0,0 +1,78 @@
+from __future__ import unicode_literals
+
+from .common import PostProcessor
+from ..compat import compat_os_name
+from ..utils import (
+    hyphenate_date,
+    write_xattr,
+    XAttrMetadataError,
+    XAttrUnavailableError,
+)
+
+
+class XAttrMetadataPP(PostProcessor):
+    #
+    # More info about extended attributes for media:
+    #   http://freedesktop.org/wiki/CommonExtendedAttributes/
+    #   http://www.freedesktop.org/wiki/PhreedomDraft/
+    #   http://dublincore.org/documents/usageguide/elements.shtml
+    #
+    # TODO:
+    #  * capture youtube keywords and put them in 'user.dublincore.subject' (comma-separated)
+    #  * figure out which xattrs can be used for 'duration', 'thumbnail', 'resolution'
+    #
+
+    def run(self, info):
+        """ Set extended attributes on downloaded file (if xattr support is found). """
+
+        # Write the metadata to the file's xattrs
+        self.to_screen('Writing metadata to file\'s xattrs')
+
+        filename = info['filepath']
+
+        try:
+            xattr_mapping = {
+                'user.xdg.referrer.url': 'webpage_url',
+                # 'user.xdg.comment':            'description',
+                'user.dublincore.title': 'title',
+                'user.dublincore.date': 'upload_date',
+                'user.dublincore.description': 'description',
+                'user.dublincore.contributor': 'uploader',
+                'user.dublincore.format': 'format',
+            }
+
+            num_written = 0
+            for xattrname, infoname in xattr_mapping.items():
+
+                value = info.get(infoname)
+
+                if value:
+                    if infoname == 'upload_date':
+                        value = hyphenate_date(value)
+
+                    byte_value = value.encode('utf-8')
+                    write_xattr(filename, xattrname, byte_value)
+                    num_written += 1
+
+            return [], info
+
+        except XAttrUnavailableError as e:
+            self.report_error(str(e))
+            return [], info
+
+        except XAttrMetadataError as e:
+            if e.reason == 'NO_SPACE':
+                self.report_warning(
+                    'There\'s no disk space left, disk quota exceeded or filesystem xattr limit exceeded. '
+                    + (('Some ' if num_written else '') + 'extended attributes are not written.').capitalize())
+            elif e.reason == 'VALUE_TOO_LONG':
+                self.report_warning(
+                    'Unable to write extended attributes due to too long values.')
+            else:
+                msg = 'This filesystem doesn\'t support extended attributes. '
+                if compat_os_name == 'nt':
+                    msg += 'You need to use NTFS.'
+                else:
+                    msg += '(You may have to enable them in your /etc/fstab)'
+                self.report_error(msg)
+            return [], info