import collections
import contextlib
import copy
-import datetime
+import datetime as dt
import errno
import fileinput
import http.cookiejar
from .cache import Cache
from .compat import functools, urllib # isort: split
-from .compat import compat_os_name, compat_shlex_quote, urllib_req_to_req
+from .compat import compat_os_name, urllib_req_to_req
from .cookies import LenientSimpleCookie, load_cookies
from .downloader import FFmpegFD, get_suitable_downloader, shorten_protocol_name
from .downloader.rtmp import rtmpdump_version
SSLError,
network_exceptions,
)
+from .networking.impersonate import ImpersonateRequestHandler
from .plugins import directories as plugin_directories
from .postprocessor import _PLUGIN_CLASSES as plugin_pps
from .postprocessor import (
SameFileError,
UnavailableVideoError,
UserNotLive,
+ YoutubeDLError,
age_restricted,
- args_to_str,
bug_reports_message,
date_from_str,
deprecation_warning,
sanitize_filename,
sanitize_path,
sanitize_url,
+ shell_quote,
str_or_none,
strftime_or_none,
subtitles_filename,
supports_terminal_sequences,
system_identifier,
+ filesize_from_tbr,
timetuple_from_msec,
to_high_limit_path,
traverse_obj,
- "detect_or_warn": check whether we can do anything
about it, warn otherwise (default)
source_address: Client-side IP address to bind to.
+ impersonate: Client to impersonate for requests.
+ An ImpersonateTarget (from yt_dlp.networking.impersonate)
sleep_interval_requests: Number of seconds to sleep between requests
during extraction
sleep_interval: Number of seconds to sleep before each download when
nopart, updatetime, buffersize, ratelimit, throttledratelimit, min_filesize,
max_filesize, test, noresizebuffer, retries, file_access_retries, fragment_retries,
continuedl, xattr_set_filesize, hls_use_mpegts, http_chunk_size,
- external_downloader_args, concurrent_fragment_downloads.
+ external_downloader_args, concurrent_fragment_downloads, progress_delta.
The following options are used by the post processors:
ffmpeg_location: Location of the ffmpeg/avconv binary; either the path
'url', 'manifest_url', 'manifest_stream_number', 'ext', 'format', 'format_id', 'format_note',
'width', 'height', 'aspect_ratio', 'resolution', 'dynamic_range', 'tbr', 'abr', 'acodec', 'asr', 'audio_channels',
'vbr', 'fps', 'vcodec', 'container', 'filesize', 'filesize_approx', 'rows', 'columns',
- 'player_url', 'protocol', 'fragment_base_url', 'fragments', 'is_from_start',
+ 'player_url', 'protocol', 'fragment_base_url', 'fragments', 'is_from_start', 'is_dash_periods', 'request_data',
'preference', 'language', 'language_preference', 'quality', 'source_preference', 'cookies',
'http_headers', 'stretched_ratio', 'no_resume', 'has_drm', 'extra_param_to_segment_url', 'hls_aes', 'downloader_options',
'page_url', 'app', 'play_path', 'tc_url', 'flash_version', 'rtmp_live', 'rtmp_conn', 'rtmp_protocol', 'rtmp_real_time'
for msg in self.params.get('_deprecation_warnings', []):
self.deprecated_feature(msg)
+ if impersonate_target := self.params.get('impersonate'):
+ if not self._impersonate_target_available(impersonate_target):
+ raise YoutubeDLError(
+ f'Impersonate target "{impersonate_target}" is not available. '
+ f'Use --list-impersonate-targets to see available targets. '
+ f'You may be missing dependencies required to support this target.')
+
if 'list-formats' in self.params['compat_opts']:
self.params['listformats_table'] = False
self.report_warning(
'Long argument string detected. '
'Use -- to separate parameters and URLs, like this:\n%s' %
- args_to_str(correct_argv))
+ shell_quote(correct_argv))
def add_info_extractor(self, ie):
"""Add an InfoExtractor object to the end of the list."""
value, fmt = escapeHTML(str(value)), str_fmt
elif fmt[-1] == 'q': # quoted
value = map(str, variadic(value) if '#' in flags else [value])
- value, fmt = ' '.join(map(compat_shlex_quote, value)), str_fmt
+ value, fmt = shell_quote(value, shell=True), str_fmt
elif fmt[-1] == 'B': # bytes
value = f'%{str_fmt}'.encode() % str(value).encode()
value, fmt = value.decode('utf-8', 'ignore'), 's'
def _check_formats(self, formats):
for f in formats:
+ working = f.get('__working')
+ if working is not None:
+ if working:
+ yield f
+ continue
self.to_screen('[info] Testing format %s' % f['format_id'])
path = self.get_output_path('temp')
if not self._ensure_dir_exists(f'{path}/'):
os.remove(temp_file.name)
except OSError:
self.report_warning('Unable to delete temporary file "%s"' % temp_file.name)
+ f['__working'] = success
if success:
yield f
else:
self.to_screen('[info] Unable to download format %s. Skipping...' % f['format_id'])
+ def _select_formats(self, formats, selector):
+ return list(selector({
+ 'formats': formats,
+ 'has_merged_format': any('none' not in (f.get('acodec'), f.get('vcodec')) for f in formats),
+ 'incomplete_formats': (all(f.get('vcodec') == 'none' for f in formats) # No formats with video
+ or all(f.get('acodec') == 'none' for f in formats)), # OR, No formats with audio
+ }))
+
def _default_format_spec(self, info_dict, download=True):
+ download = download and not self.params.get('simulate')
+ prefer_best = download and (
+ self.params['outtmpl']['default'] == '-'
+ or info_dict.get('is_live') and not self.params.get('live_from_start'))
def can_merge():
merger = FFmpegMergerPP(self)
return merger.available and merger.can_merge()
- prefer_best = (
- not self.params.get('simulate')
- and download
- and (
- not can_merge()
- or info_dict.get('is_live') and not self.params.get('live_from_start')
- or self.params['outtmpl']['default'] == '-'))
- compat = (
- prefer_best
- or self.params.get('allow_multiple_audio_streams', False)
- or 'format-spec' in self.params['compat_opts'])
-
- return (
- 'best/bestvideo+bestaudio' if prefer_best
- else 'bestvideo*+bestaudio/best' if not compat
- else 'bestvideo+bestaudio/best')
+ if not prefer_best and download and not can_merge():
+ prefer_best = True
+ formats = self._get_formats(info_dict)
+ evaluate_formats = lambda spec: self._select_formats(formats, self.build_format_selector(spec))
+ if evaluate_formats('b/bv+ba') != evaluate_formats('bv*+ba/b'):
+ self.report_warning('ffmpeg not found. The downloaded format may not be the best available. '
+ 'Installing ffmpeg is strongly recommended: https://github.com/yt-dlp/yt-dlp#dependencies')
+
+ compat = (self.params.get('allow_multiple_audio_streams')
+ or 'format-spec' in self.params['compat_opts'])
+
+ return ('best/bestvideo+bestaudio' if prefer_best
+ else 'bestvideo+bestaudio/best' if compat
+ else 'bestvideo*+bestaudio/best')
def build_format_selector(self, format_spec):
def syntax_error(note, start):
selectors = []
current_selector = None
for type, string_, start, _, _ in tokens:
- # ENCODING is only defined in python 3.x
+ # ENCODING is only defined in Python 3.x
if type == getattr(tokenize, 'ENCODING', None):
continue
elif type in [tokenize.NAME, tokenize.NUMBER]:
# Working around out-of-range timestamp values (e.g. negative ones on Windows,
# see http://bugs.python.org/issue1646728)
with contextlib.suppress(ValueError, OverflowError, OSError):
- upload_date = datetime.datetime.fromtimestamp(info_dict[ts_key], datetime.timezone.utc)
+ upload_date = dt.datetime.fromtimestamp(info_dict[ts_key], dt.timezone.utc)
info_dict[date_key] = upload_date.strftime('%Y%m%d')
if not info_dict.get('release_year'):
get_from_start = not info_dict.get('is_live') or bool(self.params.get('live_from_start'))
if not get_from_start:
- info_dict['title'] += ' ' + datetime.datetime.now().strftime('%Y-%m-%d %H:%M')
+ info_dict['title'] += ' ' + dt.datetime.now().strftime('%Y-%m-%d %H:%M')
if info_dict.get('is_live') and formats:
formats = [f for f in formats if bool(f.get('is_from_start')) == get_from_start]
if get_from_start and not formats:
format['url'] = sanitize_url(format['url'])
if format.get('ext') is None:
format['ext'] = determine_ext(format['url']).lower()
+ if format['ext'] in ('aac', 'opus', 'mp3', 'flac', 'vorbis'):
+ if format.get('acodec') is None:
+ format['acodec'] = format['ext']
if format.get('protocol') is None:
format['protocol'] = determine_protocol(format)
if format.get('resolution') is None:
format['aspect_ratio'] = try_call(lambda: round(format['width'] / format['height'], 2))
# For fragmented formats, "tbr" is often max bitrate and not average
if (('manifest-filesize-approx' in self.params['compat_opts'] or not format.get('manifest_url'))
- and info_dict.get('duration') and format.get('tbr')
and not format.get('filesize') and not format.get('filesize_approx')):
- format['filesize_approx'] = int(info_dict['duration'] * format['tbr'] * (1024 / 8))
+ format['filesize_approx'] = filesize_from_tbr(format.get('tbr'), info_dict.get('duration'))
format['http_headers'] = self._calc_headers(collections.ChainMap(format, info_dict), load_cookies=True)
# Safeguard against old/insecure infojson when using --load-info-json
self.write_debug(f'Default format spec: {req_format}')
format_selector = self.build_format_selector(req_format)
- formats_to_download = list(format_selector({
- 'formats': formats,
- 'has_merged_format': any('none' not in (f.get('acodec'), f.get('vcodec')) for f in formats),
- 'incomplete_formats': (all(f.get('vcodec') == 'none' for f in formats) # No formats with video
- or all(f.get('acodec') == 'none' for f in formats)), # OR, No formats with audio
- }))
+ formats_to_download = self._select_formats(formats, format_selector)
if interactive_format_selection and not formats_to_download:
self.report_error('Requested format is not available', tb=False, is_error=False)
continue
delim, (
format_field(f, 'filesize', ' \t%s', func=format_bytes)
or format_field(f, 'filesize_approx', '≈\t%s', func=format_bytes)
- or format_field(try_call(lambda: format_bytes(int(info_dict['duration'] * f['tbr'] * (1024 / 8)))),
- None, self._format_out('~\t%s', self.Styles.SUPPRESS))),
+ or format_field(filesize_from_tbr(f.get('tbr'), info_dict.get('duration')), None,
+ self._format_out('~\t%s', self.Styles.SUPPRESS), func=format_bytes)),
format_field(f, 'tbr', '\t%dk', func=round),
shorten_protocol_name(f.get('protocol', '')),
delim,
handler = self._request_director.handlers['Urllib']
return handler._get_instance(cookiejar=self.cookiejar, proxies=self.proxies)
+ def _get_available_impersonate_targets(self):
+ # todo(future): make available as public API
+ return [
+ (target, rh.RH_NAME)
+ for rh in self._request_director.handlers.values()
+ if isinstance(rh, ImpersonateRequestHandler)
+ for target in rh.supported_targets
+ ]
+
+ def _impersonate_target_available(self, target):
+ # todo(future): make available as public API
+ return any(
+ rh.is_supported_target(target)
+ for rh in self._request_director.handlers.values()
+ if isinstance(rh, ImpersonateRequestHandler))
+
def urlopen(self, req):
""" Start an HTTP download """
if isinstance(req, str):
raise RequestError(
'file:// URLs are disabled by default in yt-dlp for security reasons. '
'Use --enable-file-urls to enable at your own risk.', cause=ue) from ue
- if 'unsupported proxy type: "https"' in ue.msg.lower():
+ if (
+ 'unsupported proxy type: "https"' in ue.msg.lower()
+ and 'requests' not in self._request_director.handlers
+ and 'curl_cffi' not in self._request_director.handlers
+ ):
raise RequestError(
- 'To use an HTTPS proxy for this request, one of the following dependencies needs to be installed: requests')
+ 'To use an HTTPS proxy for this request, one of the following dependencies needs to be installed: requests, curl_cffi')
elif (
re.match(r'unsupported url scheme: "wss?"', ue.msg.lower())
'This request requires WebSocket support. '
'Ensure one of the following dependencies are installed: websockets',
cause=ue) from ue
+
+ elif re.match(r'unsupported (?:extensions: impersonate|impersonate target)', ue.msg.lower()):
+ raise RequestError(
+ f'Impersonate target "{req.extensions["impersonate"]}" is not available.'
+ f' See --list-impersonate-targets for available targets.'
+ f' This request requires browser impersonation, however you may be missing dependencies'
+ f' required to support this target.')
raise
except SSLError as e:
if 'UNSAFE_LEGACY_RENEGOTIATION_DISABLED' in str(e):
'timeout': 'socket_timeout',
'legacy_ssl_support': 'legacyserverconnect',
'enable_file_urls': 'enable_file_urls',
+ 'impersonate': 'impersonate',
'client_cert': {
'client_certificate': 'client_certificate',
'client_certificate_key': 'client_certificate_key',