#!/usr/bin/env python
-# -*- coding: utf-8 -*-
+# coding: utf-8
from __future__ import absolute_import, unicode_literals
import collections
import contextlib
+import copy
import datetime
import errno
import fileinput
import time
import tokenize
import traceback
+import random
from .compat import (
compat_basestring,
compat_get_terminal_size,
compat_http_client,
compat_kwargs,
+ compat_numeric_types,
compat_os_name,
compat_str,
compat_tokenize_tokenize,
ExtractorError,
format_bytes,
formatSeconds,
+ GeoRestrictedError,
+ ISO3166Utils,
locked_file,
make_HTTPS_handler,
MaxDownloadsReached,
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.
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.
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
- cn_verification_proxy: URL of the proxy to use for IP address verification
- on Chinese sites. (Experimental)
+ geo_verification_proxy: URL of the proxy to use for IP address verification
+ on geo-restricted sites. (Experimental)
socket_timeout: Time to wait for unresponsive hosts, in seconds
bidi_workaround: Work around buggy terminals without bidirectional text
support, using fridibi
source_address: (Experimental) 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.
+ 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.
+ 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
If it returns None, the video is downloaded.
match_filter_func in utils.py 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 (experimental)
+ geo_bypass_country:
+ Two-letter ISO 3166-2 country code that will be used for
+ explicit geographic restriction bypassing via faking
+ X-Forwarded-For HTTP header (experimental)
The following options determine which downloader is picked:
external_downloader: Executable of the external downloader to call.
self.params.update(params)
self.cache = Cache(self)
+ def check_deprecated(param, option, suggestion):
+ if self.params.get(param) is not None:
+ self.report_warning(
+ '%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):
try:
import pty
autonumber_size = self.params.get('autonumber_size')
if autonumber_size is None:
autonumber_size = 5
- autonumber_templ = '%0' + str(autonumber_size) + 'd'
- template_dict['autonumber'] = autonumber_templ % self._num_downloads
- if template_dict.get('playlist_index') is not None:
- template_dict['playlist_index'] = '%0*d' % (len(str(template_dict['n_entries'])), template_dict['playlist_index'])
+ 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'])
sanitize = lambda k, v: sanitize_filename(
compat_str(v),
restricted=self.params.get('restrictfilenames'),
- is_id=(k == 'id'))
- template_dict = dict((k, sanitize(k, v))
+ 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 = re.search(FIELD_SIZE_COMPAT_RE, outtmpl)
+ if mobj:
+ outtmpl = re.sub(
+ FIELD_SIZE_COMPAT_RE,
+ r'%%(\1)0%dd' % field_size_compat_map[mobj.group('field')],
+ outtmpl)
+
+ NUMERIC_FIELDS = set((
+ 'width', 'height', 'tbr', 'abr', 'asr', 'vbr', 'fps', 'filesize', 'filesize_approx',
+ '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',
+ 'playlist_index',
+ ))
+
+ # 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 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. https://docs.python.org/2/library/stdtypes.html#string-formatting
+ 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(
+ FORMAT_RE.format(numeric_field),
+ r'%({0})s'.format(numeric_field), outtmpl)
+
tmpl = compat_expanduser(outtmpl)
filename = tmpl % template_dict
# Temporary fix for #4787
return self.process_ie_result(ie_result, download, extra_info)
else:
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.'
+ self.report_error(msg)
+ break
except ExtractorError as e: # An error we somewhat expected
self.report_error(compat_str(e), e.format_traceback())
break
if self.params.get('playlistreverse', False):
entries = entries[::-1]
+ if self.params.get('playlistrandom', False):
+ random.shuffle(entries)
+
+ 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,
if isinstance(selector, list):
fs = [_build_selector_function(s) for s in selector]
- def selector_function(formats):
+ def selector_function(ctx):
for f in fs:
- for format in f(formats):
+ for format in f(ctx):
yield format
return selector_function
elif selector.type == GROUP:
elif selector.type == PICKFIRST:
fs = [_build_selector_function(s) for s in selector.selector]
- def selector_function(formats):
+ def selector_function(ctx):
for f in fs:
- picked_formats = list(f(formats))
+ picked_formats = list(f(ctx))
if picked_formats:
return picked_formats
return []
elif selector.type == SINGLE:
format_spec = selector.selector
- def selector_function(formats):
- formats = list(formats)
+ def selector_function(ctx):
+ formats = list(ctx['formats'])
if not formats:
return
if format_spec == 'all':
if f.get('vcodec') != 'none' and f.get('acodec') != 'none']
if audiovideo_formats:
yield audiovideo_formats[format_idx]
- # for audio only (soundcloud) or video only (imgur) urls, select the best/worst audio format
- elif (all(f.get('acodec') != 'none' for f in formats) or
- all(f.get('vcodec') != 'none' for f in formats)):
+ # 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 = [
}
video_selector, audio_selector = map(_build_selector_function, selector.selector)
- def selector_function(formats):
- formats = list(formats)
- for pair in itertools.product(video_selector(formats), audio_selector(formats)):
+ 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(formats):
+ def final_selector(ctx):
+ ctx_copy = copy.deepcopy(ctx)
for _filter in filters:
- formats = list(filter(_filter, formats))
- return selector_function(formats)
+ 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'))
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):
info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
if thumbnails:
thumbnails.sort(key=lambda t: (
- t.get('preference'), t.get('width'), t.get('height'),
- t.get('id'), t.get('url')))
+ 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'):
for subtitle_format in subtitle:
if subtitle_format.get('url'):
subtitle_format['url'] = sanitize_url(subtitle_format['url'])
- if 'ext' not in subtitle_format:
+ if subtitle_format.get('ext') is None:
subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
if self.params.get('listsubtitles', False):
format['format_id'] = compat_str(i)
else:
# Sanitize format_id from characters used in format selector expression
- format['format_id'] = re.sub('[\s,/+\[\]()]', '_', format['format_id'])
+ 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] = []
note=' ({0})'.format(format['format_note']) if format.get('format_note') is not None else '',
)
# Automatically determine file extension if missing
- if 'ext' not in format:
+ if format.get('ext') is None:
format['ext'] = determine_ext(format['url']).lower()
# Automatically determine protocol if missing (useful for format
# selection purposes)
- if 'protocol' not in format:
+ 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()
full_format_info.update(format)
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
req_format_list.append('best')
req_format = '/'.join(req_format_list)
format_selector = self.build_format_selector(req_format)
- formats_to_download = list(format_selector(formats))
+
+ # 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
+ # https://github.com/rg3/youtube-dl/pull/5556).
+ # 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
+ # https://github.com/rg3/youtube-dl/issues/10083).
+ 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',
expected=True)
self.to_screen('[info] Video subtitle %s.%s is already_present' % (sub_lang, sub_format))
else:
self.to_screen('[info] Writing video subtitles to: ' + sub_filename)
- with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8') as subfile:
+ # Use newline='' to prevent conversion of newline characters
+ # See https://github.com/rg3/youtube-dl/issues/10268
+ with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8', newline='') as subfile:
subfile.write(sub_data)
except (OSError, IOError):
self.report_error('Cannot write subtitles file ' + sub_filename)
video_ext, audio_ext = audio.get('ext'), video.get('ext')
if video_ext and audio_ext:
COMPATIBLE_EXTS = (
- ('mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v'),
+ ('mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'ismv', 'isma'),
('webm')
)
for exts in COMPATIBLE_EXTS:
Rapsys Git / youtubedl / blobdiff