-try:
- import contextvars # noqa: F401
-except Exception:
- raise Exception(
- f'You are using an unsupported version of Python. Only Python versions 3.7 and above are supported by yt-dlp') # noqa: F541
+import sys
+
+if sys.version_info < (3, 8):
+ raise ImportError(
+ f'You are using an unsupported version of Python. Only Python versions 3.8 and above are supported by yt-dlp') # noqa: F541
-__license__ = 'Public Domain'
+__license__ = 'The Unlicense'
import collections
import getpass
import optparse
import os
import re
-import sys
+import traceback
-from .compat import compat_shlex_quote
+from .compat import compat_os_name, compat_shlex_quote
from .cookies import SUPPORTED_BROWSERS, SUPPORTED_KEYRINGS
-from .downloader import FileDownloader
from .downloader.external import get_external_downloader
from .extractor import list_extractor_classes
from .extractor.adobepass import MSO_INFO
-from .extractor.common import InfoExtractor
+from .networking.impersonate import ImpersonateTarget
from .options import parseOpts
from .postprocessor import (
FFmpegExtractAudioPP,
DateRange,
DownloadCancelled,
DownloadError,
+ FormatSorter,
GeoUtils,
PlaylistEntries,
SameFileError,
float_or_none,
format_field,
int_or_none,
+ join_nonempty,
match_filter_func,
+ parse_bytes,
parse_duration,
preferredencoding,
read_batch_urls,
read_stdin,
render_table,
setproctitle,
- std_headers,
traverse_obj,
variadic,
write_string,
)
+from .utils.networking import std_headers
from .YoutubeDL import YoutubeDL
_IN_CLI = False
def get_urls(urls, batchfile, verbose):
- # Batch file verification
+ """
+ @param verbose -1: quiet, 0: normal, 1: verbose
+ """
batch_urls = []
if batchfile is not None:
try:
batch_urls = read_batch_urls(
- read_stdin('URLs') if batchfile == '-'
+ read_stdin(None if verbose == -1 else 'URLs') if batchfile == '-'
else open(expand_path(batchfile), encoding='utf-8', errors='ignore'))
- if verbose:
+ if verbose == 1:
write_string('[debug] Batch file urls: ' + repr(batch_urls) + '\n')
except OSError:
_exit(f'ERROR: batch file {batchfile} could not be read')
def print_extractor_information(opts, urls):
- # Importing GenericIE is currently slow since it imports other extractors
- # TODO: Move this back to module level after generalization of embed detection
- from .extractor.generic import GenericIE
-
out = ''
if opts.list_extractors:
+ # Importing GenericIE is currently slow since it imports YoutubeIE
+ from .extractor.generic import GenericIE
+
urls = dict.fromkeys(urls, False)
for ie in list_extractor_classes(opts.age_limit):
out += ie.IE_NAME + (' (CURRENTLY BROKEN)' if not ie.working() else '') + '\n'
else:
opts.embed_infojson = False
if 'format-sort' in opts.compat_opts:
- opts.format_sort.extend(InfoExtractor.FormatSort.ytdl_default)
+ opts.format_sort.extend(FormatSorter.ytdl_default)
_video_multistreams_set = set_default_compat('multistreams', 'allow_multiple_video_streams', False, remove_compat=False)
_audio_multistreams_set = set_default_compat('multistreams', 'allow_multiple_audio_streams', False, remove_compat=False)
if _video_multistreams_set is False and _audio_multistreams_set is False:
raise ValueError(f'{max_name} "{max_val}" must be must be greater than or equal to {min_name} "{min_val}"')
# Usernames and passwords
- validate(not opts.usenetrc or (opts.username is None and opts.password is None),
- '.netrc', msg='using {name} conflicts with giving username/password')
+ validate(sum(map(bool, (opts.usenetrc, opts.netrc_cmd, opts.username))) <= 1, '.netrc',
+ msg='{name}, netrc command and username/password are mutually exclusive options')
validate(opts.password is None or opts.username is not None, 'account username', msg='{name} missing')
validate(opts.ap_password is None or opts.ap_username is not None,
'TV Provider account username', msg='{name} missing')
# Format sort
for f in opts.format_sort:
- validate_regex('format sorting', f, InfoExtractor.FormatSort.regex)
+ validate_regex('format sorting', f, FormatSorter.regex)
# Postprocessor formats
validate_regex('merge output format', opts.merge_output_format,
raise ValueError(f'invalid {key} retry sleep expression {expr!r}')
# Bytes
- def parse_bytes(name, value):
+ def validate_bytes(name, value):
if value is None:
return None
- numeric_limit = FileDownloader.parse_bytes(value)
+ numeric_limit = parse_bytes(value)
validate(numeric_limit is not None, 'rate limit', value)
return numeric_limit
- opts.ratelimit = parse_bytes('rate limit', opts.ratelimit)
- opts.throttledratelimit = parse_bytes('throttled rate limit', opts.throttledratelimit)
- opts.min_filesize = parse_bytes('min filesize', opts.min_filesize)
- opts.max_filesize = parse_bytes('max filesize', opts.max_filesize)
- opts.buffersize = parse_bytes('buffer size', opts.buffersize)
- opts.http_chunk_size = parse_bytes('http chunk size', opts.http_chunk_size)
+ opts.ratelimit = validate_bytes('rate limit', opts.ratelimit)
+ opts.throttledratelimit = validate_bytes('throttled rate limit', opts.throttledratelimit)
+ opts.min_filesize = validate_bytes('min filesize', opts.min_filesize)
+ opts.max_filesize = validate_bytes('max filesize', opts.max_filesize)
+ opts.buffersize = validate_bytes('buffer size', opts.buffersize)
+ opts.http_chunk_size = validate_bytes('http chunk size', opts.http_chunk_size)
# Output templates
def validate_outtmpl(tmpl, msg):
if outtmpl_default == '':
opts.skip_download = None
del opts.outtmpl['default']
- if outtmpl_default and not os.path.splitext(outtmpl_default)[1] and opts.extractaudio:
- raise ValueError(
- 'Cannot download a video and extract audio into the same file! '
- f'Use "{outtmpl_default}.%(ext)s" instead of "{outtmpl_default}" as the output template')
- def parse_chapters(name, value):
- chapters, ranges = [], []
+ def parse_chapters(name, value, advanced=False):
parse_timestamp = lambda x: float('inf') if x in ('inf', 'infinite') else parse_duration(x)
+ TIMESTAMP_RE = r'''(?x)(?:
+ (?P<start_sign>-?)(?P<start>[^-]+)
+ )?\s*-\s*(?:
+ (?P<end_sign>-?)(?P<end>[^-]+)
+ )?'''
+
+ chapters, ranges, from_url = [], [], False
for regex in value or []:
- if regex.startswith('*'):
- for range_ in map(str.strip, regex[1:].split(',')):
- mobj = range_ != '-' and re.fullmatch(r'([^-]+)?\s*-\s*([^-]+)?', range_)
- dur = mobj and (parse_timestamp(mobj.group(1) or '0'), parse_timestamp(mobj.group(2) or 'inf'))
- if None in (dur or [None]):
- raise ValueError(f'invalid {name} time range "{regex}". Must be of the form *start-end')
- ranges.append(dur)
+ if advanced and regex == '*from-url':
+ from_url = True
+ continue
+ elif not regex.startswith('*'):
+ try:
+ chapters.append(re.compile(regex))
+ except re.error as err:
+ raise ValueError(f'invalid {name} regex "{regex}" - {err}')
continue
- try:
- chapters.append(re.compile(regex))
- except re.error as err:
- raise ValueError(f'invalid {name} regex "{regex}" - {err}')
- return chapters, ranges
- opts.remove_chapters, opts.remove_ranges = parse_chapters('--remove-chapters', opts.remove_chapters)
- opts.download_ranges = download_range_func(*parse_chapters('--download-sections', opts.download_ranges))
+ for range_ in map(str.strip, regex[1:].split(',')):
+ mobj = range_ != '-' and re.fullmatch(TIMESTAMP_RE, range_)
+ dur = mobj and [parse_timestamp(mobj.group('start') or '0'), parse_timestamp(mobj.group('end') or 'inf')]
+ signs = mobj and (mobj.group('start_sign'), mobj.group('end_sign'))
+
+ err = None
+ if None in (dur or [None]):
+ err = 'Must be of the form "*start-end"'
+ elif not advanced and any(signs):
+ err = 'Negative timestamps are not allowed'
+ else:
+ dur[0] *= -1 if signs[0] else 1
+ dur[1] *= -1 if signs[1] else 1
+ if dur[1] == float('-inf'):
+ err = '"-inf" is not a valid end'
+ if err:
+ raise ValueError(f'invalid {name} time range "{regex}". {err}')
+ ranges.append(dur)
+
+ return chapters, ranges, from_url
+
+ opts.remove_chapters, opts.remove_ranges, _ = parse_chapters('--remove-chapters', opts.remove_chapters)
+ opts.download_ranges = download_range_func(*parse_chapters('--download-sections', opts.download_ranges, True))
# Cookies from browser
if opts.cookiesfrombrowser:
mobj = re.fullmatch(r'''(?x)
(?P<name>[^+:]+)
(?:\s*\+\s*(?P<keyring>[^:]+))?
- (?:\s*:\s*(?P<profile>.+?))?
+ (?:\s*:\s*(?!:)(?P<profile>.+?))?
(?:\s*::\s*(?P<container>.+))?
''', opts.cookiesfrombrowser)
if mobj is None:
f'Supported keyrings are: {", ".join(sorted(SUPPORTED_KEYRINGS))}')
opts.cookiesfrombrowser = (browser_name, profile, keyring, container)
+ if opts.impersonate is not None:
+ opts.impersonate = ImpersonateTarget.from_str(opts.impersonate.lower())
+
# MetadataParser
def metadataparser_actions(f):
if isinstance(f, str):
raise ValueError(f'{cmd} is invalid; {err}')
yield action
- parse_metadata = opts.parse_metadata or []
if opts.metafromtitle is not None:
- parse_metadata.append('title:%s' % opts.metafromtitle)
- opts.parse_metadata = list(itertools.chain(*map(metadataparser_actions, parse_metadata)))
+ opts.parse_metadata.setdefault('pre_process', []).append('title:%s' % opts.metafromtitle)
+ opts.parse_metadata = {
+ k: list(itertools.chain(*map(metadataparser_actions, v)))
+ for k, v in opts.parse_metadata.items()
+ }
# Other options
if opts.playlist_items is not None:
except Exception as err:
raise ValueError(f'Invalid playlist-items {opts.playlist_items!r}: {err}')
- geo_bypass_code = opts.geo_bypass_ip_block or opts.geo_bypass_country
- if geo_bypass_code is not None:
+ opts.geo_bypass_country, opts.geo_bypass_ip_block = None, None
+ if opts.geo_bypass.lower() not in ('default', 'never'):
try:
- GeoUtils.random_ipv4(geo_bypass_code)
+ GeoUtils.random_ipv4(opts.geo_bypass)
except Exception:
- raise ValueError('unsupported geo-bypass country or ip-block')
+ raise ValueError(f'Unsupported --xff "{opts.geo_bypass}"')
+ if len(opts.geo_bypass) == 2:
+ opts.geo_bypass_country = opts.geo_bypass
+ else:
+ opts.geo_bypass_ip_block = opts.geo_bypass
+ opts.geo_bypass = opts.geo_bypass.lower() != 'never'
- opts.match_filter = match_filter_func(opts.match_filter)
+ opts.match_filter = match_filter_func(opts.match_filter, opts.breaking_match_filter)
if opts.download_archive is not None:
opts.download_archive = expand_path(opts.download_archive)
elif ed and proto == 'default':
default_downloader = ed.get_basename()
+ for policy in opts.color.values():
+ if policy not in ('always', 'auto', 'no_color', 'never'):
+ raise ValueError(f'"{policy}" is not a valid color policy')
+
warnings, deprecation_warnings = [], []
# Common mistake: -f best
val1=opts.sponskrub and opts.sponskrub_cut)
# Conflicts with --allow-unplayable-formats
- report_conflict('--add-metadata', 'addmetadata')
+ report_conflict('--embed-metadata', 'addmetadata')
report_conflict('--embed-chapters', 'addchapters')
report_conflict('--embed-info-json', 'embed_infojson')
report_conflict('--embed-subs', 'embedsubtitles')
def get_postprocessors(opts):
yield from opts.add_postprocessors
- if opts.parse_metadata:
+ for when, actions in opts.parse_metadata.items():
yield {
'key': 'MetadataParser',
- 'actions': opts.parse_metadata,
- 'when': 'pre_process'
+ 'actions': actions,
+ 'when': when
}
sponsorblock_query = opts.sponsorblock_mark | opts.sponsorblock_remove
if sponsorblock_query:
def parse_options(argv=None):
"""@returns ParsedOptions(parser, opts, urls, ydl_opts)"""
parser, opts, urls = parseOpts(argv)
- urls = get_urls(urls, opts.batchfile, opts.verbose)
+ urls = get_urls(urls, opts.batchfile, -1 if opts.quiet and not opts.verbose else opts.verbose)
set_compat_opts(opts)
try:
postprocessors = list(get_postprocessors(opts))
- print_only = bool(opts.forceprint) and all(k not in opts.forceprint for k in POSTPROCESS_WHEN[2:])
+ print_only = bool(opts.forceprint) and all(k not in opts.forceprint for k in POSTPROCESS_WHEN[3:])
any_getting = any(getattr(opts, k) for k in (
'dumpjson', 'dump_single_json', 'getdescription', 'getduration', 'getfilename',
'getformat', 'getid', 'getthumbnail', 'gettitle', 'geturl'
))
+ if opts.quiet is None:
+ opts.quiet = any_getting or opts.print_json or bool(opts.forceprint)
playlist_pps = [pp for pp in postprocessors if pp.get('when') == 'playlist']
write_playlist_infojson = (opts.writeinfojson and not opts.clean_infojson
return ParsedOptions(parser, opts, urls, {
'usenetrc': opts.usenetrc,
'netrc_location': opts.netrc_location,
+ 'netrc_cmd': opts.netrc_cmd,
'username': opts.username,
'password': opts.password,
'twofactor': opts.twofactor,
'client_certificate': opts.client_certificate,
'client_certificate_key': opts.client_certificate_key,
'client_certificate_password': opts.client_certificate_password,
- 'quiet': opts.quiet or any_getting or opts.print_json or bool(opts.forceprint),
+ 'quiet': opts.quiet,
'no_warnings': opts.no_warnings,
'forceurl': opts.geturl,
'forcetitle': opts.gettitle,
'noprogress': opts.quiet if opts.noprogress is None else opts.noprogress,
'progress_with_newline': opts.progress_with_newline,
'progress_template': opts.progress_template,
+ 'progress_delta': opts.progress_delta,
'playliststart': opts.playliststart,
'playlistend': opts.playlistend,
'playlistreverse': opts.playlist_reverse,
'legacyserverconnect': opts.legacy_server_connect,
'nocheckcertificate': opts.no_check_certificate,
'prefer_insecure': opts.prefer_insecure,
+ 'enable_file_urls': opts.enable_file_urls,
'http_headers': opts.headers,
'proxy': opts.proxy,
'socket_timeout': opts.socket_timeout,
'postprocessors': postprocessors,
'fixup': opts.fixup,
'source_address': opts.source_address,
+ 'impersonate': opts.impersonate,
'call_home': opts.call_home,
'sleep_interval_requests': opts.sleep_interval_requests,
'sleep_interval': opts.sleep_interval,
'playlist_items': opts.playlist_items,
'xattr_set_filesize': opts.xattr_set_filesize,
'match_filter': opts.match_filter,
- 'no_color': opts.no_color,
+ 'color': opts.color,
'ffmpeg_location': opts.ffmpeg_location,
'hls_prefer_native': opts.hls_prefer_native,
'hls_use_mpegts': opts.hls_use_mpegts,
if opts.rm_cachedir:
ydl.cache.remove()
- updater = Updater(ydl)
- if opts.update_self and updater.update() and actual_use:
- if updater.cmd:
- return updater.restart()
- # This code is reachable only for zip variant in py < 3.10
- # It makes sense to exit here, but the old behavior is to continue
- ydl.report_warning('Restart yt-dlp to use the updated version')
- # return 100, 'ERROR: The program must exit for the update to complete'
+ try:
+ updater = Updater(ydl, opts.update_self)
+ if opts.update_self and updater.update() and actual_use:
+ if updater.cmd:
+ return updater.restart()
+ # This code is reachable only for zip variant in py < 3.10
+ # It makes sense to exit here, but the old behavior is to continue
+ ydl.report_warning('Restart yt-dlp to use the updated version')
+ # return 100, 'ERROR: The program must exit for the update to complete'
+ except Exception:
+ traceback.print_exc()
+ ydl._download_retcode = 100
+
+ if opts.list_impersonate_targets:
+
+ known_targets = [
+ # List of simplified targets we know are supported,
+ # to help users know what dependencies may be required.
+ (ImpersonateTarget('chrome'), 'curl_cffi'),
+ (ImpersonateTarget('edge'), 'curl_cffi'),
+ (ImpersonateTarget('safari'), 'curl_cffi'),
+ ]
+
+ available_targets = ydl._get_available_impersonate_targets()
+
+ def make_row(target, handler):
+ return [
+ join_nonempty(target.client.title(), target.version, delim='-') or '-',
+ join_nonempty((target.os or "").title(), target.os_version, delim='-') or '-',
+ handler,
+ ]
+
+ rows = [make_row(target, handler) for target, handler in available_targets]
+
+ for known_target, known_handler in known_targets:
+ if not any(
+ known_target in target and handler == known_handler
+ for target, handler in available_targets
+ ):
+ rows.append([
+ ydl._format_out(text, ydl.Styles.SUPPRESS)
+ for text in make_row(known_target, f'{known_handler} (not available)')
+ ])
+
+ ydl.to_screen('[info] Available impersonate targets')
+ ydl.to_stdout(render_table(['Client', 'OS', 'Source'], rows, extra_gap=2, delim='-'))
+ return
if not actual_use:
if pre_process:
return ydl._download_retcode
- ydl.warn_if_short_id(sys.argv[1:] if argv is None else argv)
+ args = sys.argv[1:] if argv is None else argv
+ ydl.warn_if_short_id(args)
+
+ # Show a useful error message and wait for keypress if not launched from shell on Windows
+ if not args and compat_os_name == 'nt' and getattr(sys, 'frozen', False):
+ import ctypes.wintypes
+ import msvcrt
+
+ kernel32 = ctypes.WinDLL('Kernel32')
+
+ buffer = (1 * ctypes.wintypes.DWORD)()
+ attached_processes = kernel32.GetConsoleProcessList(buffer, 1)
+ # If we only have a single process attached, then the executable was double clicked
+ # When using `pyinstaller` with `--onefile`, two processes get attached
+ is_onefile = hasattr(sys, '_MEIPASS') and os.path.basename(sys._MEIPASS).startswith('_MEI')
+ if attached_processes == 1 or is_onefile and attached_processes == 2:
+ print(parser._generate_error_message(
+ 'Do not double-click the executable, instead call it from a command line.\n'
+ 'Please read the README for further information on how to use yt-dlp: '
+ 'https://github.com/yt-dlp/yt-dlp#readme'))
+ msvcrt.getch()
+ _exit(2)
parser.error(
'You must provide at least one URL.\n'
'Type yt-dlp --help to see a list of all options.')
parser.destroy()
try:
if opts.load_info_filename is not None:
+ if all_urls:
+ ydl.report_warning('URLs are ignored due to --load-info-json')
return ydl.download_with_info_file(expand_path(opts.load_info_filename))
else:
return ydl.download(all_urls)
def main(argv=None):
+ global _IN_CLI
+ _IN_CLI = True
try:
_exit(*variadic(_real_main(argv)))
except DownloadError: