You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

2405 lines
108 KiB

#!/usr/bin/env python
# coding: utf-8
from __future__ import absolute_import, unicode_literals
import collections
import contextlib
import copy
import datetime
import errno
import fileinput
import io
import itertools
import json
import locale
import operator
import os
import platform
import re
import shutil
import subprocess
import socket
import sys
import time
import tokenize
import traceback
import random
from string import ascii_letters
from .compat import (
from .utils import (
from .cache import Cache
from .extractor import get_info_extractor, gen_extractor_classes, _LAZY_LOADER
from .extractor.openload import PhantomJSwrapper
from .downloader import get_suitable_downloader
from .downloader.rtmp import rtmpdump_version
from .postprocessor import (
from .version import __version__
if compat_os_name == 'nt':
import ctypes
class YoutubeDL(object):
"""YoutubeDL class.
YoutubeDL objects are the ones responsible of downloading the
actual video file and writing it to disk if the user has requested
it, among some other tasks. In most cases there should be one per
program. As, given a video URL, the downloader doesn't know how to
extract all the needed information, task that InfoExtractors do, it
has to pass the URL to one of them.
For this, YoutubeDL objects have a method that allows
InfoExtractors to be registered in a given order. When it is passed
a URL, the YoutubeDL object handles it to the first InfoExtractor it
finds that reports being able to handle it. The InfoExtractor extracts
all the information about the video or videos the URL refers to, and
YoutubeDL process the extracted information, possibly using a File
Downloader to download the video.
YoutubeDL objects accept a lot of parameters. In order not to saturate
the object constructor with arguments, it receives a dictionary of
options instead. These options are available through the params
attribute for the InfoExtractors to use. The YoutubeDL also
registers itself as the downloader in charge for the InfoExtractors
that are added to it, so this is a "mutual registration".
Available options:
username: Username for authentication purposes.
password: Password for authentication purposes.
videopassword: Password for accessing a video.
ap_mso: Adobe Pass multiple-system operator identifier.
ap_username: Multiple-system operator account username.
ap_password: Multiple-system operator account password.
usenetrc: Use netrc for authentication instead.
verbose: Print additional info to stdout.
quiet: Do not print messages to stdout.
no_warnings: Do not print out anything for warnings.
forceurl: Force printing final URL.
forcetitle: Force printing title.
forceid: Force printing ID.
forcethumbnail: Force printing thumbnail URL.
forcedescription: Force printing description.
forcefilename: Force printing final filename.
forceduration: Force printing duration.
forcejson: Force printing info_dict as JSON.
dump_single_json: Force printing the info_dict of the whole playlist
(or video) as a single JSON line.
simulate: Do not download the video files.
format: Video format code. See for more information.
outtmpl: Template for output names.
restrictfilenames: Do not allow "&" and spaces in file names
ignoreerrors: Do not stop on download errors.
force_generic_extractor: Force downloader to use the generic extractor
nooverwrites: Prevent overwriting files.
playliststart: Playlist item to start at.
playlistend: Playlist item to end at.
playlist_items: Specific indices of playlist to download.
playlistreverse: Download playlist items in reverse order.
playlistrandom: Download playlist items in random order.
matchtitle: Download only matching titles.
rejecttitle: Reject downloads for matching titles.
logger: Log messages to a logging.Logger instance.
logtostderr: Log messages to stderr instead of stdout.
writedescription: Write the video description to a .description file
writeinfojson: Write the video description to a .info.json file
writeannotations: Write the video annotations to a .annotations.xml file
writethumbnail: Write the thumbnail image to a file
write_all_thumbnails: Write all thumbnail formats to files
writesubtitles: Write the video subtitles to a file
writeautomaticsub: Write the automatically generated subtitles to a file
allsubtitles: Downloads all the subtitles of the video
(requires writesubtitles or writeautomaticsub)
listsubtitles: Lists all available subtitles for the video
subtitlesformat: The format code for subtitles
subtitleslangs: List of languages of the subtitles to download
keepvideo: Keep the video file after post-processing
daterange: A DateRange object, download only if the upload_date is in the range.
skip_download: Skip the actual download of the video file
cachedir: Location of the cache files in the filesystem.
False to disable filesystem cache.
noplaylist: Download single video instead of a playlist if in doubt.
age_limit: An integer representing the user's age in years.
Unsuitable videos for the given age are skipped.
min_views: An integer representing the minimum view count the video
must have in order to not be skipped.
Videos without view count information are always
downloaded. None for no limit.
max_views: An integer representing the maximum view count.
Videos that are more popular than that are not
Videos without view count information are always
downloaded. None for no limit.
download_archive: File name of a file where all downloads are recorded.
Videos already present in the file are not downloaded
cookiefile: File name where cookies should be read from and dumped to.
nocheckcertificate:Do not verify SSL certificates
prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
At the moment, this is only supported by YouTube.
proxy: URL of the proxy server to use
geo_verification_proxy: URL of the proxy to use for IP address verification
on geo-restricted sites.
socket_timeout: Time to wait for unresponsive hosts, in seconds
bidi_workaround: Work around buggy terminals without bidirectional text
support, using fridibi
debug_printtraffic:Print out sent and received HTTP traffic
include_ads: Download ads as well
default_search: Prepend this string if an input url is not valid.
'auto' for elaborate guessing
encoding: Use this encoding instead of the system-specified.
extract_flat: Do not resolve URLs, return the immediate result.
Pass in 'in_playlist' to only show this behavior for
playlist items.
postprocessors: A list of dictionaries, each with an entry
* key: The name of the postprocessor. See
youtube_dl/postprocessor/ for a list.
as well as any further keyword arguments for the
progress_hooks: A list of functions that get called on download
progress, with a dictionary with the entries
* status: One of "downloading", "error", or "finished".
Check this first and ignore unknown values.
If status is one of "downloading", or "finished", the
following properties may also be present:
* filename: The final filename (always present)
* tmpfilename: The filename we're currently writing to
* downloaded_bytes: Bytes on disk
* total_bytes: Size of the whole file, None if unknown
* total_bytes_estimate: Guess of the eventual file size,
None if unavailable.
* elapsed: The number of seconds since download started.
* eta: The estimated time in seconds, None if unknown
* speed: The download speed in bytes/second, None if
* fragment_index: The counter of the currently
downloaded video fragment.
* fragment_count: The number of fragments (= individual
files that will be merged)
Progress hooks are guaranteed to be called at least once
(with status "finished") if the download is successful.
merge_output_format: Extension to use when merging formats.
fixup: Automatically correct known faults of the file.
One of:
- "never": do nothing
- "warn": only emit a warning
- "detect_or_warn": check whether we can do anything
about it, warn otherwise (default)
source_address: Client-side IP address to bind to.
call_home: Boolean, true iff we are allowed to contact the
youtube-dl servers for debugging.
sleep_interval: Number of seconds to sleep before each download when
used alone or a lower bound of a range for randomized
sleep before each download (minimum possible number
of seconds to sleep) when used along with
max_sleep_interval:Upper bound of a range for randomized sleep before each
download (maximum possible number of seconds to sleep).
Must only be used along with sleep_interval.
Actual sleep time will be a random float from range
[sleep_interval; max_sleep_interval].
listformats: Print an overview of available video formats and exit.
list_thumbnails: Print a table of all thumbnails and exit.
match_filter: A function that gets called with the info_dict of
every video.
If it returns a message, the video is ignored.
If it returns None, the video is downloaded.
match_filter_func in is one example for this.
no_color: Do not emit color codes in output.
geo_bypass: Bypass geographic restriction via faking X-Forwarded-For
HTTP header
Two-letter ISO 3166-2 country code that will be used for
explicit geographic restriction bypassing via faking
X-Forwarded-For HTTP header
IP range in CIDR notation that will be used similarly to
The following options determine which downloader is picked:
external_downloader: Executable of the external downloader to call.
None or unset for standard (built-in) downloader.
hls_prefer_native: Use the native HLS downloader instead of ffmpeg/avconv
if True, otherwise use ffmpeg/avconv if False, otherwise
use downloader suggested by extractor if None.
The following parameters are not used by YoutubeDL itself, they are used by
the downloader (see youtube_dl/downloader/
nopart, updatetime, buffersize, ratelimit, min_filesize, max_filesize, test,
noresizebuffer, retries, continuedl, noprogress, consoletitle,
xattr_set_filesize, external_downloader_args, hls_use_mpegts,
The following options are used by the post processors:
prefer_ffmpeg: If False, use avconv instead of ffmpeg if both are available,
otherwise prefer ffmpeg.
postprocessor_args: A list of additional command-line arguments for the
The following options are used by the Youtube extractor:
youtube_include_dash_manifest: If True (default), DASH manifests and related
data will be downloaded and processed by extractor.
You can reduce network I/O by disabling it if you don't
care about DASH.
'width', 'height', 'tbr', 'abr', 'asr', 'vbr', 'fps', 'filesize', 'filesize_approx',
'timestamp', 'upload_year', 'upload_month', 'upload_day',
'duration', 'view_count', 'like_count', 'dislike_count', 'repost_count',
'average_rating', 'comment_count', 'age_limit',
'start_time', 'end_time',
'chapter_number', 'season_number', 'episode_number',
'track_number', 'disc_number', 'release_year',
params = None
_ies = []
_pps = []
_download_retcode = None
_num_downloads = None
_screen_file = None
def __init__(self, params=None, auto_init=True):
"""Create a FileDownloader object with the given options."""
if params is None:
params = {}
self._ies = []
self._ies_instances = {}
self._pps = []
self._progress_hooks = []
self._download_retcode = 0
self._num_downloads = 0
self._screen_file = [sys.stdout, sys.stderr][params.get('logtostderr', False)]
self._err_file = sys.stderr
self.params = {
# Default parameters
'nocheckcertificate': False,
self.cache = Cache(self)
def check_deprecated(param, option, suggestion):
if self.params.get(param) is not None:
'%s is deprecated. Use %s instead.' % (option, suggestion))
return True
return False
if check_deprecated('cn_verification_proxy', '--cn-verification-proxy', '--geo-verification-proxy'):
if self.params.get('geo_verification_proxy') is None:
self.params['geo_verification_proxy'] = self.params['cn_verification_proxy']
check_deprecated('autonumber_size', '--autonumber-size', 'output template with %(autonumber)0Nd, where N in the number of digits')
check_deprecated('autonumber', '--auto-number', '-o "%(autonumber)s-%(title)s.%(ext)s"')
check_deprecated('usetitle', '--title', '-o "%(title)s-%(id)s.%(ext)s"')
if params.get('bidi_workaround', False):
import pty
master, slave = pty.openpty()
width = compat_get_terminal_size().columns
if width is None:
width_args = []
width_args = ['-w', str(width)]
sp_kwargs = dict(
self._output_process = subprocess.Popen(
['bidiv'] + width_args, **sp_kwargs
except OSError:
self._output_process = subprocess.Popen(
['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
self._output_channel = os.fdopen(master, 'rb')
except OSError as ose:
if ose.errno == errno.ENOENT:
self.report_warning('Could not find fribidi executable, ignoring --bidi-workaround . Make sure that fribidi is an executable file in one of the directories in your $PATH.')
if (sys.platform != 'win32' and
sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968'] and
not params.get('restrictfilenames', False)):
# Unicode filesystem API will throw errors (#1474, #13027)
'Assuming --restrict-filenames since file system encoding '
'cannot encode all characters. '
'Set the LC_ALL environment variable to fix this.')
self.params['restrictfilenames'] = True
if isinstance(params.get('outtmpl'), bytes):
'Parameter outtmpl is bytes, but should be a unicode string. '
'Put from __future__ import unicode_literals at the top of your code file or consider switching to Python 3.x.')
if auto_init:
for pp_def_raw in self.params.get('postprocessors', []):
pp_class = get_postprocessor(pp_def_raw['key'])
pp_def = dict(pp_def_raw)
del pp_def['key']
pp = pp_class(self, **compat_kwargs(pp_def))
for ph in self.params.get('progress_hooks', []):
def warn_if_short_id(self, argv):
# short YouTube ID starting with dash?
idxs = [
i for i, a in enumerate(argv)
if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
if idxs:
correct_argv = (
['youtube-dl'] +
[a for i, a in enumerate(argv) if i not in idxs] +
['--'] + [argv[i] for i in idxs]
'Long argument string detected. '
'Use -- to separate parameters and URLs, like this:\n%s\n' %
def add_info_extractor(self, ie):
"""Add an InfoExtractor object to the end of the list."""
if not isinstance(ie, type):
self._ies_instances[ie.ie_key()] = ie
def get_info_extractor(self, ie_key):
Get an instance of an IE with name ie_key, it will try to get one from
the _ies list, if there's no instance it will create a new one and add
it to the extractor list.
ie = self._ies_instances.get(ie_key)
if ie is None:
ie = get_info_extractor(ie_key)()
return ie
def add_default_info_extractors(self):
Add the InfoExtractors returned by gen_extractors to the end of the list
for ie in gen_extractor_classes():
def add_post_processor(self, pp):
"""Add a PostProcessor object to the end of the chain."""
def add_progress_hook(self, ph):
"""Add the progress hook (currently only for the file downloader)"""
def _bidi_workaround(self, message):
if not hasattr(self, '_output_channel'):
return message
assert hasattr(self, '_output_process')
assert isinstance(message, compat_str)
line_count = message.count('\n') + 1
self._output_process.stdin.write((message + '\n').encode('utf-8'))
res = ''.join(self._output_channel.readline().decode('utf-8')
for _ in range(line_count))
return res[:-len('\n')]
def to_screen(self, message, skip_eol=False):
"""Print message to stdout if not in quiet mode."""
return self.to_stdout(message, skip_eol, check_quiet=True)
def _write_string(self, s, out=None):
write_string(s, out=out, encoding=self.params.get('encoding'))
def to_stdout(self, message, skip_eol=False, check_quiet=False):
"""Print message to stdout if not in quiet mode."""
if self.params.get('logger'):
elif not check_quiet or not self.params.get('quiet', False):
message = self._bidi_workaround(message)
terminator = ['\n', ''][skip_eol]
output = message + terminator
self._write_string(output, self._screen_file)
def to_stderr(self, message):
"""Print message to stderr."""
assert isinstance(message, compat_str)
if self.params.get('logger'):
message = self._bidi_workaround(message)
output = message + '\n'
self._write_string(output, self._err_file)
def to_console_title(self, message):
if not self.params.get('consoletitle', False):
if compat_os_name == 'nt':
if ctypes.windll.kernel32.GetConsoleWindow():
# c_wchar_p() might not be necessary if `message` is
# already of type unicode()
elif 'TERM' in os.environ:
self._write_string('\033]0;%s\007' % message, self._screen_file)
def save_console_title(self):
if not self.params.get('consoletitle', False):
if self.params.get('simulate', False):
if compat_os_name != 'nt' and 'TERM' in os.environ:
# Save the title on stack
self._write_string('\033[22;0t', self._screen_file)
def restore_console_title(self):
if not self.params.get('consoletitle', False):
if self.params.get('simulate', False):
if compat_os_name != 'nt' and 'TERM' in os.environ:
# Restore the title from stack
self._write_string('\033[23;0t', self._screen_file)
def __enter__(self):
return self
def __exit__(self, *args):
if self.params.get('cookiefile') is not None:, ignore_expires=True)
def trouble(self, message=None, tb=None):
"""Determine action to take when a download problem appears.
Depending on if the downloader has been configured to ignore
download errors or not, this method may throw an exception or
not when errors are found, after printing the message.
tb, if given, is additional traceback information.
if message is not None:
if self.params.get('verbose'):
if tb is None:
if sys.exc_info()[0]: # if .trouble has been called from an except block
tb = ''
if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
tb += encode_compat_str(traceback.format_exc())
tb_data = traceback.format_list(traceback.extract_stack())
tb = ''.join(tb_data)
if not self.params.get('ignoreerrors', False):
if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
exc_info = sys.exc_info()[1].exc_info
exc_info = sys.exc_info()
raise DownloadError(message, exc_info)
self._download_retcode = 1
def report_warning(self, message):
Print the message to stderr, it will be prefixed with 'WARNING:'
If stderr is a tty file the 'WARNING:' will be colored
if self.params.get('logger') is not None:
if self.params.get('no_warnings'):
if not self.params.get('no_color') and self._err_file.isatty() and compat_os_name != 'nt':
_msg_header = '\033[0;33mWARNING:\033[0m'
_msg_header = 'WARNING:'
warning_message = '%s %s' % (_msg_header, message)
def report_error(self, message, tb=None):
Do the same as trouble, but prefixes the message with 'ERROR:', colored
in red if stderr is a tty file.
if not self.params.get('no_color') and self._err_file.isatty() and compat_os_name != 'nt':
_msg_header = '\033[0;31mERROR:\033[0m'
_msg_header = 'ERROR:'
error_message = '%s %s' % (_msg_header, message)
self.trouble(error_message, tb)
def report_file_already_downloaded(self, file_name):
"""Report file has already been fully downloaded."""
self.to_screen('[download] %s has already been downloaded' % file_name)
except UnicodeEncodeError:
self.to_screen('[download] The file has already been downloaded')
def prepare_filename(self, info_dict):
"""Generate the output filename."""
template_dict = dict(info_dict)
template_dict['epoch'] = int(time.time())
autonumber_size = self.params.get('autonumber_size')
if autonumber_size is None:
autonumber_size = 5
template_dict['autonumber'] = self.params.get('autonumber_start', 1) - 1 + self._num_downloads
if template_dict.get('resolution') is None:
if template_dict.get('width') and template_dict.get('height'):
template_dict['resolution'] = '%dx%d' % (template_dict['width'], template_dict['height'])
elif template_dict.get('height'):
template_dict['resolution'] = '%sp' % template_dict['height']
elif template_dict.get('width'):
template_dict['resolution'] = '%dx?' % template_dict['width']
sanitize = lambda k, v: sanitize_filename(
is_id=(k == 'id' or k.endswith('_id')))
template_dict = dict((k, v if isinstance(v, compat_numeric_types) else sanitize(k, v))
for k, v in template_dict.items()
if v is not None and not isinstance(v, (list, tuple, dict)))
template_dict = collections.defaultdict(lambda: 'NA', template_dict)
outtmpl = self.params.get('outtmpl', DEFAULT_OUTTMPL)
# For fields playlist_index and autonumber convert all occurrences
# of %(field)s to %(field)0Nd for backward compatibility
field_size_compat_map = {
'playlist_index': len(str(template_dict['n_entries'])),
'autonumber': autonumber_size,
FIELD_SIZE_COMPAT_RE = r'(?<!%)%\((?P<field>autonumber|playlist_index)\)s'
mobj =, outtmpl)
if mobj:
outtmpl = re.sub(
r'%%(\1)0%dd' % field_size_compat_map['field')],
# Missing numeric fields used together with integer presentation types
# in format specification will break the argument substitution since
# string 'NA' is returned for missing fields. We will patch output
# template for missing fields to meet string presentation type.
for numeric_field in self._NUMERIC_FIELDS:
if numeric_field not in template_dict:
# As of [1] format syntax is:
# %[mapping_key][conversion_flags][minimum_width][.precision][length_modifier]type
# 1.
FORMAT_RE = r'''(?x)
\({0}\) # mapping key
(?:[#0\-+ ]+)? # conversion flags (optional)
(?:\d+)? # minimum field width (optional)
(?:\.\d+)? # precision (optional)
[hlL]? # length modifier (optional)
[diouxXeEfFgGcrs%] # conversion type
outtmpl = re.sub(
r'%({0})s'.format(numeric_field), outtmpl)
# expand_path translates '%%' into '%' and '$$' into '$'
# correspondingly that is not what we want since we need to keep
# '%%' intact for template dict substitution step. Working around
# with boundary-alike separator hack.
sep = ''.join([random.choice(ascii_letters) for _ in range(32)])
outtmpl = outtmpl.replace('%%', '%{0}%'.format(sep)).replace('$$', '${0}$'.format(sep))
# outtmpl should be expand_path'ed before template dict substitution
# because meta fields may contain env variables we don't want to
# be expanded. For example, for outtmpl "%(title)s.%(ext)s" and
# title "Hello $PATH", we don't want `$PATH` to be expanded.
filename = expand_path(outtmpl).replace(sep, '') % template_dict
# Temporary fix for #4787
# 'Treat' all problem characters by passing filename through preferredencoding
# to workaround encoding issues with subprocess on python2 @ Windows
if sys.version_info < (3, 0) and sys.platform == 'win32':
filename = encodeFilename(filename, True).decode(preferredencoding())
return sanitize_path(filename)
except ValueError as err:
self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
return None
def _match_entry(self, info_dict, incomplete):
""" Returns None iff the file should be downloaded """
video_title = info_dict.get('title', info_dict.get('id', 'video'))
if 'title' in info_dict:
# This can happen when we're just evaluating the playlist
title = info_dict['title']
matchtitle = self.params.get('matchtitle', False)
if matchtitle:
if not, title, re.IGNORECASE):
return '"' + title + '" title did not match pattern "' + matchtitle + '"'
rejecttitle = self.params.get('rejecttitle', False)
if rejecttitle:
if, title, re.IGNORECASE):
return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
date = info_dict.get('upload_date')
if date is not None:
dateRange = self.params.get('daterange', DateRange())
if date not in dateRange:
return '%s upload date is not in range %s' % (date_from_str(date).isoformat(), dateRange)
view_count = info_dict.get('view_count')
if view_count is not None:
min_views = self.params.get('min_views')
if min_views is not None and view_count < min_views:
return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
max_views = self.params.get('max_views')
if max_views is not None and view_count > max_views:
return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
return 'Skipping "%s" because it is age restricted' % video_title
if self.in_download_archive(info_dict):
return '%s has already been recorded in archive' % video_title
if not incomplete:
match_filter = self.params.get('match_filter')
if match_filter is not None:
ret = match_filter(info_dict)
if ret is not None:
return ret
return None
def add_extra_info(info_dict, extra_info):
'''Set the keys from extra_info in info dict if they are missing'''
for key, value in extra_info.items():
info_dict.setdefault(key, value)
def extract_info(self, url, download=True, ie_key=None, extra_info={},
process=True, force_generic_extractor=False):
Returns a list with a dictionary for each video we find.
If 'download', also downloads the videos.
extra_info is a dict containing the extra values to add to each result
if not ie_key and force_generic_extractor:
ie_key = 'Generic'
if ie_key:
ies = [self.get_info_extractor(ie_key)]
ies = self._ies
for ie in ies:
if not ie.suitable(url):
ie = self.get_info_extractor(ie.ie_key())
if not ie.working():
self.report_warning('The program functionality for this site has been marked as broken, '
'and will probably not work.')
ie_result = ie.extract(url)
if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
if isinstance(ie_result, list):
# Backwards compatibility: old IE result format
ie_result = {
'_type': 'compat_list',
'entries': ie_result,
self.add_default_extra_info(ie_result, ie, url)
if process:
return self.process_ie_result(ie_result, download, extra_info)
return ie_result
except GeoRestrictedError as e:
msg = e.msg
if e.countries:
msg += '\nThis video is available in %s.' % ', '.join(
map(ISO3166Utils.short2full, e.countries))
msg += '\nYou might want to use a VPN or a proxy server (with --proxy) to workaround.'
except ExtractorError as e: # An error we somewhat expected
self.report_error(compat_str(e), e.format_traceback())
except MaxDownloadsReached:
except Exception as e:
if self.params.get('ignoreerrors', False):
self.report_error(error_to_compat_str(e), tb=encode_compat_str(traceback.format_exc()))
self.report_error('no suitable InfoExtractor for URL %s' % url)
def add_default_extra_info(self, ie_result, ie, url):
self.add_extra_info(ie_result, {
'extractor': ie.IE_NAME,
'webpage_url': url,
'webpage_url_basename': url_basename(url),
'extractor_key': ie.ie_key(),
def process_ie_result(self, ie_result, download=True, extra_info={}):
Take the result of the ie(may be modified) and resolve all unresolved
references (URLs, playlist items).
It will also download the videos if 'download'.
Returns the resolved ie_result.
result_type = ie_result.get('_type', 'video')
if result_type in ('url', 'url_transparent'):
ie_result['url'] = sanitize_url(ie_result['url'])
extract_flat = self.params.get('extract_flat', False)
if ((extract_flat == 'in_playlist' and 'playlist' in extra_info) or
extract_flat is True):
if self.params.get('forcejson', False):
return ie_result
if result_type == 'video':
self.add_extra_info(ie_result, extra_info)
return self.process_video_result(ie_result, download=download)
elif result_type == 'url':
# We have to add extra_info to the results because it may be
# contained in a playlist
return self.extract_info(ie_result['url'],
elif result_type == 'url_transparent':
# Use the information from the embedding page
info = self.extract_info(
ie_result['url'], ie_key=ie_result.get('ie_key'),
extra_info=extra_info, download=False, process=False)
# extract_info may return None when ignoreerrors is enabled and
# extraction failed with an error, don't crash and return early
# in this case
if not info:
return info
force_properties = dict(
(k, v) for k, v in ie_result.items() if v is not None)
for f in ('_type', 'url', 'id', 'extractor', 'extractor_key', 'ie_key'):
if f in force_properties:
del force_properties[f]
new_result = info.copy()
# Extracted info may not be a video result (i.e.
# info.get('_type', 'video') != video) but rather an url or
# url_transparent. In such cases outer metadata (from ie_result)
# should be propagated to inner one (info). For this to happen
# _type of info should be overridden with url_transparent. This
# fixes issue from
if new_result.get('_type') == 'url':
new_result['_type'] = 'url_transparent'
return self.process_ie_result(
new_result, download=download, extra_info=extra_info)
elif result_type in ('playlist', 'multi_video'):
# We process each entry in the playlist
playlist = ie_result.get('title') or ie_result.get('id')
self.to_screen('[download] Downloading playlist: %s' % playlist)
playlist_results = []
playliststart = self.params.get('playliststart', 1) - 1
playlistend = self.params.get('playlistend')
# For backwards compatibility, interpret -1 as whole list
if playlistend == -1:
playlistend = None
playlistitems_str = self.params.get('playlist_items')
playlistitems = None
if playlistitems_str is not None:
def iter_playlistitems(format):
for string_segment in format.split(','):
if '-' in string_segment:
start, end = string_segment.split('-')
for item in range(int(start), int(end) + 1):
yield int(item)
yield int(string_segment)
playlistitems = orderedSet(iter_playlistitems(playlistitems_str))
ie_entries = ie_result['entries']
def make_playlistitems_entries(list_ie_entries):
num_entries = len(list_ie_entries)
return [
list_ie_entries[i - 1] for i in playlistitems
if -num_entries <= i - 1 < num_entries]
def report_download(num_entries):
'[%s] playlist %s: Downloading %d videos' %
(ie_result['extractor'], playlist, num_entries))
if isinstance(ie_entries, list):
n_all_entries = len(ie_entries)
if playlistitems:
entries = make_playlistitems_entries(ie_entries)
entries = ie_entries[playliststart:playlistend]
n_entries = len(entries)
'[%s] playlist %s: Collected %d video ids (downloading %d of them)' %
(ie_result['extractor'], playlist, n_all_entries, n_entries))
elif isinstance(ie_entries, PagedList):
if playlistitems:
entries = []
for item in playlistitems:
item - 1, item
entries = ie_entries.getslice(
playliststart, playlistend)
n_entries = len(entries)
else: # iterable
if playlistitems:
entries = make_playlistitems_entries(list(itertools.islice(
ie_entries, 0, max(playlistitems))))
entries = list(itertools.islice(
ie_entries, playliststart, playlistend))
n_entries = len(entries)
if self.params.get('playlistreverse', False):
entries = entries[::-1]
if self.params.get('playlistrandom', False):
x_forwarded_for = ie_result.get('__x_forwarded_for_ip')
for i, entry in enumerate(entries, 1):
self.to_screen('[download] Downloading video %s of %s' % (i, n_entries))
# This __x_forwarded_for_ip thing is a bit ugly but requires
# minimal changes
if x_forwarded_for:
entry['__x_forwarded_for_ip'] = x_forwarded_for
extra = {
'n_entries': n_entries,
'playlist': playlist,
'playlist_id': ie_result.get('id'),
'playlist_title': ie_result.get('title'),
'playlist_uploader': ie_result.get('uploader'),
'playlist_uploader_id': ie_result.get('uploader_id'),
'playlist_index': i + playliststart,
'extractor': ie_result['extractor'],
'webpage_url': ie_result['webpage_url'],
'webpage_url_basename': url_basename(ie_result['webpage_url']),
'extractor_key': ie_result['extractor_key'],
reason = self._match_entry(entry, incomplete=True)
if reason is not None:
self.to_screen('[download] ' + reason)
entry_result = self.process_ie_result(entry,
ie_result['entries'] = playlist_results
self.to_screen('[download] Finished downloading playlist: %s' % playlist)
return ie_result
elif result_type == 'compat_list':
'Extractor %s returned a compat_list result. '
'It needs to be updated.' % ie_result.get('extractor'))
def _fixup(r):
'extractor': ie_result['extractor'],
'webpage_url': ie_result['webpage_url'],
'webpage_url_basename': url_basename(ie_result['webpage_url']),
'extractor_key': ie_result['extractor_key'],
return r
ie_result['entries'] = [
self.process_ie_result(_fixup(r), download, extra_info)
for r in ie_result['entries']
return ie_result
raise Exception('Invalid result type: %s' % result_type)
def _build_format_filter(self, filter_spec):
" Returns a function to filter the formats according to the filter_spec "
'<=': operator.le,
'=': operator.eq,
operator_rex = re.compile(r'''(?x)\s*
''' % '|'.join(map(re.escape, OPERATORS.keys())))
m =
if m:
comparison_value = int('value'))
except ValueError:
comparison_value = parse_filesize('value'))
if comparison_value is None:
comparison_value = parse_filesize('value') + 'B')
if comparison_value is None:
raise ValueError(
'Invalid value %r in format specification %r' % ('value'), filter_spec))
op = OPERATORS['op')]
if not m:
'=': operator.eq,
'^=': lambda attr, value: attr.startswith(value),
'$=': lambda attr, value: attr.endswith(value),
'*=': lambda attr, value: value in attr,
str_operator_rex = re.compile(r'''(?x)
''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
m =
if m:
comparison_value ='value')
str_op = STR_OPERATORS['op')]
op = lambda attr, value: not str_op(attr, value)
op = str_op
if not m:
raise ValueError('Invalid filter specification %r' % filter_spec)
def _filter(f):
actual_value = f.get('key'))
if actual_value is None:
return op(actual_value, comparison_value)
return _filter
def _default_format_spec(self, info_dict, download=True):
def can_merge():
merger = FFmpegMergerPP(self)
return merger.available and merger.can_merge()
def prefer_best():
if self.params.get('simulate', False):
return False
if not download:
return False
if self.params.get('outtmpl', DEFAULT_OUTTMPL) == '-':
return True
if info_dict.get('is_live'):
return True
if not can_merge():
return True
return False
req_format_list = ['bestvideo+bestaudio', 'best']
if prefer_best():
return '/'.join(req_format_list)
def build_format_selector(self, format_spec):
def syntax_error(note, start):
message = (
'Invalid format specification: '
'{0}\n\t{1}\n\t{2}^'.format(note, format_spec, ' ' * start[1]))
return SyntaxError(message)
FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
def _parse_filter(tokens):
filter_parts = []
for type, string, start, _, _ in tokens:
if type == tokenize.OP and string == ']':
return ''.join(filter_parts)
def _remove_unused_ops(tokens):
# Remove operators that we don't use and join them with the surrounding strings
# for example: 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
ALLOWED_OPS = ('/', '+', ',', '(', ')')
last_string, last_start, last_end, last_line = None, None, None, None
for type, string, start, end, line in tokens:
if type == tokenize.OP and string == '[':
if last_string:
yield tokenize.NAME, last_string, last_start, last_end, last_line
last_string = None
yield type, string, start, end, line
# everything inside brackets will be handled by _parse_filter
for type, string, start, end, line in tokens:
yield type, string, start, end, line
if type == tokenize.OP and string == ']':
elif type == tokenize.OP and string in ALLOWED_OPS:
if last_string:
yield tokenize.NAME, last_string, last_start, last_end, last_line
last_string = None
yield type, string, start, end, line
elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
if not last_string:
last_string = string
last_start = start
last_end = end
last_string += string
if last_string:
yield tokenize.NAME, last_string, last_start, last_end, last_line
def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
selectors = []
current_selector = None
for type, string, start, _, _ in tokens:
# ENCODING is only defined in python 3.x
if type == getattr(tokenize, 'ENCODING', None):
elif type in [tokenize.NAME, tokenize.NUMBER]:
current_selector = FormatSelector(SINGLE, string, [])
elif type == tokenize.OP:
if string == ')':
if not inside_group:
# ')' will be handled by the parentheses group
elif inside_merge and string in ['/', ',']:
elif inside_choice and string == ',':
elif string == ',':
if not current_selector:
raise syntax_error('"," must follow a format selector', start)
current_selector = None
elif string == '/':
if not current_selector:
raise syntax_error('"/" must follow a format selector', start)
first_choice = current_selector
second_choice = _parse_format_selection(tokens, inside_choice=True)
current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
elif string == '[':
if not current_selector:
current_selector = FormatSelector(SINGLE, 'best', [])
format_filter = _parse_filter(tokens)
elif string == '(':
if current_selector:
raise syntax_error('Unexpected "("', start)
group = _parse_format_selection(tokens, inside_group=True)
current_selector = FormatSelector(GROUP, group, [])
elif string == '+':
video_selector = current_selector
audio_selector = _parse_format_selection(tokens, inside_merge=True)
if not video_selector or not audio_selector:
raise syntax_error('"+" must be between two format selectors', start)
current_selector = FormatSelector(MERGE, (video_selector, audio_selector), [])
raise syntax_error('Operator not recognized: "{0}"'.format(string), start)
elif type == tokenize.ENDMARKER:
if current_selector:
return selectors
def _build_selector_function(selector):
if isinstance(selector, list):
fs = [_build_selector_function(s) for s in selector]
def selector_function(ctx):
for f in fs:
for format in f(ctx):
yield format
return selector_function
elif selector.type == GROUP:
selector_function = _build_selector_function(selector.selector)
elif selector.type == PICKFIRST:
fs = [_build_selector_function(s) for s in selector.selector]
def selector_function(ctx):
for f in fs:
picked_formats = list(f(ctx))
if picked_formats:
return picked_formats
return []
elif selector.type == SINGLE:
format_spec = selector.selector
def selector_function(ctx):
formats = list(ctx['formats'])
if not formats:
if format_spec == 'all':
for f in formats:
yield f
elif format_spec in ['best', 'worst', None]:
format_idx = 0 if format_spec == 'worst' else -1
audiovideo_formats = [
f for f in formats
if f.get('vcodec') != 'none' and f.get('acodec') != 'none']
if audiovideo_formats:
yield audiovideo_formats[format_idx]
# for extractors with incomplete formats (audio only (soundcloud)
# or video only (imgur)) we will fallback to best/worst
# {video,audio}-only format
elif ctx['incomplete_formats']:
yield formats[format_idx]
elif format_spec == 'bestaudio':
audio_formats = [
f for f in formats
if f.get('vcodec') == 'none']
if audio_formats:
yield audio_formats[-1]
elif format_spec == 'worstaudio':
audio_formats = [
f for f in formats
if f.get('vcodec') == 'none']
if audio_formats:
yield audio_formats[0]
elif format_spec == 'bestvideo':
video_formats = [
f for f in formats
if f.get('acodec') == 'none']
if video_formats:
yield video_formats[-1]
elif format_spec == 'worstvideo':
video_formats = [
f for f in formats
if f.get('acodec') == 'none']
if video_formats:
yield video_formats[0]
extensions = ['mp4', 'flv', 'webm', '3gp', 'm4a', 'mp3', 'ogg', 'aac', 'wav']
if format_spec in extensions:
filter_f = lambda f: f['ext'] == format_spec
filter_f = lambda f: f['format_id'] == format_spec
matches = list(filter(filter_f, formats))
if matches:
yield matches[-1]
elif selector.type == MERGE:
def _merge(formats_info):
format_1, format_2 = [f['format_id'] for f in formats_info]
# The first format must contain the video and the
# second the audio
if formats_info[0].get('vcodec') == 'none':
self.report_error('The first format must '
'contain the video, try using '
'"-f %s+%s"' % (format_2, format_1))
# Formats must be opposite (video+audio)
if formats_info[0].get('acodec') == 'none' and formats_info[1].get('acodec') == 'none':
'Both formats %s and %s are video-only, you must specify "-f video+audio"'
% (format_1, format_2))
output_ext = (
if self.params.get('merge_output_format') is None
else self.params['merge_output_format'])
return {
'requested_formats': formats_info,
'format': '%s+%s' % (formats_info[0].get('format'),
'format_id': '%s+%s' % (formats_info[0].get('format_id'),
'width': formats_info[0].get('width'),
'height': formats_info[0].get('height'),
'resolution': formats_info[0].get('resolution'),
'fps': formats_info[0].get('fps'),
'vcodec': formats_info[0].get('vcodec'),
'vbr': formats_info[0].get('vbr'),
'stretched_ratio': formats_info[0].get('stretched_ratio'),
'acodec': formats_info[1].get('acodec'),
'abr': formats_info[1].get('abr'),
'ext': output_ext,
video_selector, audio_selector = map(_build_selector_function, selector.selector)
def selector_function(ctx):
for pair in itertools.product(
video_selector(copy.deepcopy(ctx)), audio_selector(copy.deepcopy(ctx))):
yield _merge(pair)
filters = [self._build_format_filter(f) for f in selector.filters]
def final_selector(ctx):
ctx_copy = copy.deepcopy(ctx)
for _filter in filters:
ctx_copy['formats'] = list(filter(_filter, ctx_copy['formats']))
return selector_function(ctx_copy)
return final_selector
stream = io.BytesIO(format_spec.encode('utf-8'))
tokens = list(_remove_unused_ops(compat_tokenize_tokenize(stream.readline)))
except tokenize.TokenError:
raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
class TokenIterator(object):
def __init__(self, tokens):
self.tokens = tokens
self.counter = 0
def __iter__(self):
return self
def __next__(self):
if self.counter >= len(self.tokens):
raise StopIteration()
value = self.tokens[self.counter]
self.counter += 1
return value
next = __next__
def restore_last_token(self):
self.counter -= 1
parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
return _build_selector_function(parsed_selector)
def _calc_headers(self, info_dict):
res = std_headers.copy()
add_headers = info_dict.get('http_headers')
if add_headers:
cookies = self._calc_cookies(info_dict)
if cookies:
res['Cookie'] = cookies
if 'X-Forwarded-For' not in res:
x_forwarded_for_ip = info_dict.get('__x_forwarded_for_ip')
if x_forwarded_for_ip:
res['X-Forwarded-For'] = x_forwarded_for_ip
return res
def _calc_cookies(self, info_dict):
pr = sanitized_Request(info_dict['url'])
return pr.get_header('Cookie')
def process_video_result(self, info_dict, download=True):
assert info_dict.get('_type', 'video') == 'video'
if 'id' not in info_dict:
raise ExtractorError('Missing "id" field in extractor result')
if 'title' not in info_dict:
raise ExtractorError('Missing "title" field in extractor result')
def report_force_conversion(field, field_not, conversion):
'"%s" field is not %s - forcing %s conversion, there is an error in extractor'
% (field, field_not, conversion))
def sanitize_string_field(info, string_field):
field = info.get(string_field)
if field is None or isinstance(field, compat_str):
report_force_conversion(string_field, 'a string', 'string')
info[string_field] = compat_str(field)
def sanitize_numeric_fields(info):
for numeric_field in self._NUMERIC_FIELDS:
field = info.get(numeric_field)
if field is None or isinstance(field, compat_numeric_types):
report_force_conversion(numeric_field, 'numeric', 'int')
info[numeric_field] = int_or_none(field)
sanitize_string_field(info_dict, 'id')
if 'playlist' not in info_dict:
# It isn't part of a playlist
info_dict['playlist'] = None
info_dict['playlist_index'] = None
thumbnails = info_dict.get('thumbnails')
if thumbnails is None:
thumbnail = info_dict.get('thumbnail')
if thumbnail:
info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
if thumbnails:
thumbnails.sort(key=lambda t: (
t.get('preference') if t.get('preference') is not None else -1,
t.get('width') if t.get('width') is not None else -1,
t.get('height') if t.get('height') is not None else -1,
t.get('id') if t.get('id') is not None else '', t.get('url')))
for i, t in enumerate(thumbnails):
t['url'] = sanitize_url(t['url'])
if t.get('width') and t.get('height'):
t['resolution'] = '%dx%d' % (t['width'], t['height'])
if t.get('id') is None:
t['id'] = '%d' % i
if self.params.get('list_thumbnails'):
thumbnail = info_dict.get('thumbnail')
if thumbnail:
info_dict['thumbnail'] = sanitize_url(thumbnail)
elif thumbnails:
info_dict['thumbnail'] = thumbnails[-1]['url']
if 'display_id' not in info_dict and 'id' in info_dict:
info_dict['display_id'] = info_dict['id']
if info_dict.get('upload_date') is None and info_dict.get('timestamp') is not None:
# Working around out-of-range timestamp values (e.g. negative ones on Windows,
# see
upload_date = datetime.datetime.utcfromtimestamp(info_dict['timestamp'])
info_dict['upload_date'] = upload_date.strftime('%Y%m%d')
except (ValueError, OverflowError, OSError):
# Auto generate title fields corresponding to the *_number fields when missing
# in order to always have clean titles. This is very common for TV series.
for field in ('chapter', 'season', 'episode'):
if info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
for cc_kind in ('subtitles', 'automatic_captions'):
cc = info_dict.get(cc_kind)
if cc:
for _, subtitle in cc.items():
for subtitle_format in subtitle:
if subtitle_format.get('url'):
subtitle_format['url'] = sanitize_url(subtitle_format['url'])
if subtitle_format.get('ext') is None:
subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
automatic_captions = info_dict.get('automatic_captions')
subtitles = info_dict.get('subtitles')
if self.params.get('listsubtitles', False):
if 'automatic_captions' in info_dict:
info_dict['id'], automatic_captions, 'automatic captions')
self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
info_dict['requested_subtitles'] = self.process_subtitles(
info_dict['id'], subtitles, automatic_captions)
# We now pick which formats have to be downloaded
if info_dict.get('formats') is None:
# There's only one format available
formats = [info_dict]
formats = info_dict['formats']
if not formats:
raise ExtractorError('No video formats found!')
def is_wellformed(f):
url = f.get('url')
if not url:
'"url" field is missing or empty - skipping format, '
'there is an error in extractor')
return False
if isinstance(url, bytes):
sanitize_string_field(f, 'url')
return True
# Filter out malformed formats for better extraction robustness
formats = list(filter(is_wellformed, formats))
formats_dict = {}
# We check that all the formats have the format and format_id fields
for i, format in enumerate(formats):
sanitize_string_field(format, 'format_id')
format['url'] = sanitize_url(format['url'])
if not format.get('format_id'):
format['format_id'] = compat_str(i)
# Sanitize format_id from characters used in format selector expression
format['format_id'] = re.sub(r'[\s,/+\[\]()]', '_', format['format_id'])
format_id = format['format_id']
if format_id not in formats_dict:
formats_dict[format_id] = []
# Make sure all formats have unique format_id
for format_id, ambiguous_formats in formats_dict.items():
if len(ambiguous_formats) > 1:
for i, format in enumerate(ambiguous_formats):
format['format_id'] = '%s-%d' % (format_id, i)
for i, format in enumerate(formats):
if format.get('format') is None:
format['format'] = '{id} - {res}{note}'.format(
note=' ({0})'.format(format['format_note']) if format.get('format_note') is not None else '',
# Automatically determine file extension if missing
if format.get('ext') is None:
format['ext'] = determine_ext(format['url']).lower()
# Automatically determine protocol if missing (useful for format
# selection purposes)
if format.get('protocol') is None:
format['protocol'] = determine_protocol(format)
# Add HTTP headers, so that external programs can use them from the
# json output
full_format_info = info_dict.copy()
format['http_headers'] = self._calc_headers(full_format_info)
# Remove private housekeeping stuff
if '__x_forwarded_for_ip' in info_dict:
del info_dict['__x_forwarded_for_ip']
# TODO Central sorting goes here
if formats[0] is not info_dict:
# only set the 'formats' fields if the original info_dict list them
# otherwise we end up with a circular reference, the first (and unique)
# element in the 'formats' field in info_dict is info_dict itself,
# which can't be exported to json
info_dict['formats'] = formats
if self.params.get('listformats'):
req_format = self.params.get('format')
if req_format is None:
req_format = self._default_format_spec(info_dict, download=download)
if self.params.get('verbose'):
self.to_stdout('[debug] Default format spec: %s' % req_format)
format_selector = self.build_format_selector(req_format)
# While in format selection we may need to have an access to the original
# format set in order to calculate some metrics or do some processing.
# For now we need to be able to guess whether original formats provided
# by extractor are incomplete or not (i.e. whether extractor provides only
# video-only or audio-only formats) for proper formats selection for
# extractors with such incomplete formats (see
# Since formats may be filtered during format selection and may not match
# the original formats the results may be incorrect. Thus original formats
# or pre-calculated metrics should be passed to format selection routines
# as well.
# We will pass a context object containing all necessary additional data
# instead of just formats.
# This fixes incorrect format selection issue (see
incomplete_formats = (
# All formats are video-only or
all(f.get('vcodec') != 'none' and f.get('acodec') == 'none' for f in formats) or
# all formats are audio-only
all(f.get('vcodec') == 'none' and f.get('acodec') != 'none' for f in formats))
ctx = {
'formats': formats,
'incomplete_formats': incomplete_formats,
formats_to_download = list(format_selector(ctx))
if not formats_to_download:
raise ExtractorError('requested format not available',
if download:
if len(formats_to_download) > 1:
self.to_screen('[info] %s: downloading video in %s formats' % (info_dict['id'], len(formats_to_download)))
for format in formats_to_download:
new_info = dict(info_dict)
# We update the info dict with the best quality format (backwards compatibility)
return info_dict
def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
"""Select the requested subtitles and their format"""
available_subs = {}
if normal_subtitles and self.params.get('writesubtitles'):
if automatic_captions and self.params.get('writeautomaticsub'):
for lang, cap_info in automatic_captions.items():
if lang not in available_subs:
available_subs[lang] = cap_info
if (not self.params.get('writesubtitles') and not
self.params.get('writeautomaticsub') or not
return None
if self.params.get('allsubtitles', False):
requested_langs = available_subs.keys()
if self.params.get('subtitleslangs', False):
requested_langs = self.params.get('subtitleslangs')
elif 'en' in available_subs:
requested_langs = ['en']
requested_langs = [list(available_subs.keys())[0]]
formats_query = self.params.get('subtitlesformat', 'best')
formats_preference = formats_query.split('/') if formats_query else []
subs = {}
for lang in requested_langs:
formats = available_subs.get(lang)
if formats is None:
self.report_warning('%s subtitles not available for %s' % (lang, video_id))
for ext in formats_preference:
if ext == 'best':
f = formats[-1]
matches = list(filter(lambda f: f['ext'] == ext, formats))
if matches:
f = matches[-1]
f = formats[-1]
'No subtitle format found matching "%s" for language %s, '
'using %s' % (formats_query, lang, f['ext']))
subs[lang] = f
return subs
def process_info(self, info_dict):
"""Process a single resolved IE result."""
assert info_dict.get('_type', 'video') == 'video'
max_downloads = self.params.get('max_downloads')
if max_downloads is not None:
if self._num_downloads >= int(max_downloads):
raise MaxDownloadsReached()
info_dict['fulltitle'] = info_dict['title']
if len(info_dict['title']) > 200:
info_dict['title'] = info_dict['title'][:197] + '...'
if 'format' not in info_dict:
info_dict['format'] = info_dict['ext']
reason = self._match_entry(info_dict, incomplete=False)
if reason is not None:
self.to_screen('[download] ' + reason)
self._num_downloads += 1
info_dict['_filename'] = filename = self.prepare_filename(info_dict)
# Forced printings
if self.params.get('forcetitle', False):
if self.params.get('forceid', False):
if self.params.get('forceurl', False):
if info_dict.get('requested_formats') is not None:
for f in info_dict['requested_formats']:
self.to_stdout(f['url'] + f.get('play_path', ''))
# For RTMP URLs, also include the playpath
self.to_stdout(info_dict['url'] + info_dict.get('play_path', ''))
if self.params.get('forcethumbnail', False) and info_dict.get('thumbnail') is not None:
if self.params.get('forcedescription', False) and info_dict.get('description') is not None:
if self.params.get('forcefilename', False) and filename is not None:
if self.params.get('forceduration', False) and info_dict.get('duration') is not None:
if self.params.get('forceformat', False):
if self.params.get('forcejson', False):
# Do nothing else if in simulate mode
if self.params.get('simulate', False):
if filename is None:
def ensure_dir_exists(path):
dn = os.path.dirname(path)
if dn and not os.path.exists(dn):
return True
except (OSError, IOError) as err:
self.report_error('unable to create directory ' + error_to_compat_str(err))
return False
if not ensure_dir_exists(sanitize_path(encodeFilename(filename))):
if self.params.get('writedescription', False):
descfn = replace_extension(filename, 'description', info_dict.get('ext'))
if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(descfn)):
self.to_screen('[info] Video description is already present')
elif info_dict.get('description') is None:
self.report_warning('There\'s no description to write.')
self.to_screen('[info] Writing video description to: ' + descfn)
with, 'w', encoding='utf-8') as descfile:
except (OSError, IOError):
self.report_error('Cannot write description file ' + descfn)
if self.params.get('writeannotations', False):
annofn = replace_extension(filename, 'annotations.xml', info_dict.get('ext'))
if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(annofn)):
self.to_screen('[info] Video annotations are already present')
self.to_screen('[info] Writing video annotations to: ' + annofn)
with, 'w', encoding='utf-8') as annofile:
except (KeyError, TypeError):
self.report_warning('There are no annotations to write.')
except (OSError, IOError):
self.report_error('Cannot write annotations file: ' + annofn)
subtitles_are_requested = any([self.params.get('writesubtitles', False),
if subtitles_are_requested and info_dict.get('requested_subtitles'):
# subtitles download errors are already managed as troubles in relevant IE
# that way it will silently go on when used with unsupporting IE
subtitles = info_dict['requested_subtitles']
ie = self.get_info_extractor(info_dict['extractor_key'])
for sub_lang, sub_info in subtitles.items():
sub_format = sub_info['ext']
sub_filename = subtitles_filename(filename, sub_lang, sub_format)
if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(sub_filename)):
self.to_screen('[info] Video subtitle %s.%s is already present' % (sub_lang, sub_format))
self.to_screen('[info] Writing video subtitles to: ' + sub_filename)
if sub_info.get('data') is not None:
# Use newline='' to prevent conversion of newline characters
# See
with, 'w', encoding='utf-8', newline='') as subfile:
except (OSError, IOError):
self.report_error('Cannot write subtitles file ' + sub_filename)
sub_data = ie._request_webpage(
sub_info['url'], info_dict['id'], note=False).read()
with, 'wb') as subfile:
except (ExtractorError, IOError, OSError, ValueError) as err:
self.report_warning('Unable to download subtitle for "%s": %s' %
(sub_lang, error_to_compat_str(err)))
if self.params.get('writeinfojson', False):
infofn = replace_extension(filename, 'info.json', info_dict.get('ext'))
if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(infofn)):