]> Raphaƫl G. Git Repositories - youtubedl/blob - youtube_dl/YoutubeDL.py
Initiate new release.
[youtubedl] / youtube_dl / YoutubeDL.py
1 #!/usr/bin/env python
2 # coding: utf-8
3
4 from __future__ import absolute_import, unicode_literals
5
6 import collections
7 import contextlib
8 import copy
9 import datetime
10 import errno
11 import fileinput
12 import io
13 import itertools
14 import json
15 import locale
16 import operator
17 import os
18 import platform
19 import re
20 import shutil
21 import subprocess
22 import socket
23 import sys
24 import time
25 import tokenize
26 import traceback
27 import random
28
29 from string import ascii_letters
30
31 from .compat import (
32 compat_basestring,
33 compat_cookiejar,
34 compat_get_terminal_size,
35 compat_http_client,
36 compat_kwargs,
37 compat_numeric_types,
38 compat_os_name,
39 compat_str,
40 compat_tokenize_tokenize,
41 compat_urllib_error,
42 compat_urllib_request,
43 compat_urllib_request_DataHandler,
44 )
45 from .utils import (
46 age_restricted,
47 args_to_str,
48 ContentTooShortError,
49 date_from_str,
50 DateRange,
51 DEFAULT_OUTTMPL,
52 determine_ext,
53 determine_protocol,
54 DownloadError,
55 encode_compat_str,
56 encodeFilename,
57 error_to_compat_str,
58 expand_path,
59 ExtractorError,
60 format_bytes,
61 formatSeconds,
62 GeoRestrictedError,
63 int_or_none,
64 ISO3166Utils,
65 locked_file,
66 make_HTTPS_handler,
67 MaxDownloadsReached,
68 orderedSet,
69 PagedList,
70 parse_filesize,
71 PerRequestProxyHandler,
72 platform_name,
73 PostProcessingError,
74 preferredencoding,
75 prepend_extension,
76 register_socks_protocols,
77 render_table,
78 replace_extension,
79 SameFileError,
80 sanitize_filename,
81 sanitize_path,
82 sanitize_url,
83 sanitized_Request,
84 std_headers,
85 str_or_none,
86 subtitles_filename,
87 UnavailableVideoError,
88 url_basename,
89 version_tuple,
90 write_json_file,
91 write_string,
92 YoutubeDLCookieJar,
93 YoutubeDLCookieProcessor,
94 YoutubeDLHandler,
95 YoutubeDLRedirectHandler,
96 )
97 from .cache import Cache
98 from .extractor import get_info_extractor, gen_extractor_classes, _LAZY_LOADER
99 from .extractor.openload import PhantomJSwrapper
100 from .downloader import get_suitable_downloader
101 from .downloader.rtmp import rtmpdump_version
102 from .postprocessor import (
103 FFmpegFixupM3u8PP,
104 FFmpegFixupM4aPP,
105 FFmpegFixupStretchedPP,
106 FFmpegMergerPP,
107 FFmpegPostProcessor,
108 get_postprocessor,
109 )
110 from .version import __version__
111
112 if compat_os_name == 'nt':
113 import ctypes
114
115
116 class YoutubeDL(object):
117 """YoutubeDL class.
118
119 YoutubeDL objects are the ones responsible of downloading the
120 actual video file and writing it to disk if the user has requested
121 it, among some other tasks. In most cases there should be one per
122 program. As, given a video URL, the downloader doesn't know how to
123 extract all the needed information, task that InfoExtractors do, it
124 has to pass the URL to one of them.
125
126 For this, YoutubeDL objects have a method that allows
127 InfoExtractors to be registered in a given order. When it is passed
128 a URL, the YoutubeDL object handles it to the first InfoExtractor it
129 finds that reports being able to handle it. The InfoExtractor extracts
130 all the information about the video or videos the URL refers to, and
131 YoutubeDL process the extracted information, possibly using a File
132 Downloader to download the video.
133
134 YoutubeDL objects accept a lot of parameters. In order not to saturate
135 the object constructor with arguments, it receives a dictionary of
136 options instead. These options are available through the params
137 attribute for the InfoExtractors to use. The YoutubeDL also
138 registers itself as the downloader in charge for the InfoExtractors
139 that are added to it, so this is a "mutual registration".
140
141 Available options:
142
143 username: Username for authentication purposes.
144 password: Password for authentication purposes.
145 videopassword: Password for accessing a video.
146 ap_mso: Adobe Pass multiple-system operator identifier.
147 ap_username: Multiple-system operator account username.
148 ap_password: Multiple-system operator account password.
149 usenetrc: Use netrc for authentication instead.
150 verbose: Print additional info to stdout.
151 quiet: Do not print messages to stdout.
152 no_warnings: Do not print out anything for warnings.
153 forceurl: Force printing final URL.
154 forcetitle: Force printing title.
155 forceid: Force printing ID.
156 forcethumbnail: Force printing thumbnail URL.
157 forcedescription: Force printing description.
158 forcefilename: Force printing final filename.
159 forceduration: Force printing duration.
160 forcejson: Force printing info_dict as JSON.
161 dump_single_json: Force printing the info_dict of the whole playlist
162 (or video) as a single JSON line.
163 simulate: Do not download the video files.
164 format: Video format code. See options.py for more information.
165 outtmpl: Template for output names.
166 restrictfilenames: Do not allow "&" and spaces in file names
167 ignoreerrors: Do not stop on download errors.
168 force_generic_extractor: Force downloader to use the generic extractor
169 nooverwrites: Prevent overwriting files.
170 playliststart: Playlist item to start at.
171 playlistend: Playlist item to end at.
172 playlist_items: Specific indices of playlist to download.
173 playlistreverse: Download playlist items in reverse order.
174 playlistrandom: Download playlist items in random order.
175 matchtitle: Download only matching titles.
176 rejecttitle: Reject downloads for matching titles.
177 logger: Log messages to a logging.Logger instance.
178 logtostderr: Log messages to stderr instead of stdout.
179 writedescription: Write the video description to a .description file
180 writeinfojson: Write the video description to a .info.json file
181 writeannotations: Write the video annotations to a .annotations.xml file
182 writethumbnail: Write the thumbnail image to a file
183 write_all_thumbnails: Write all thumbnail formats to files
184 writesubtitles: Write the video subtitles to a file
185 writeautomaticsub: Write the automatically generated subtitles to a file
186 allsubtitles: Downloads all the subtitles of the video
187 (requires writesubtitles or writeautomaticsub)
188 listsubtitles: Lists all available subtitles for the video
189 subtitlesformat: The format code for subtitles
190 subtitleslangs: List of languages of the subtitles to download
191 keepvideo: Keep the video file after post-processing
192 daterange: A DateRange object, download only if the upload_date is in the range.
193 skip_download: Skip the actual download of the video file
194 cachedir: Location of the cache files in the filesystem.
195 False to disable filesystem cache.
196 noplaylist: Download single video instead of a playlist if in doubt.
197 age_limit: An integer representing the user's age in years.
198 Unsuitable videos for the given age are skipped.
199 min_views: An integer representing the minimum view count the video
200 must have in order to not be skipped.
201 Videos without view count information are always
202 downloaded. None for no limit.
203 max_views: An integer representing the maximum view count.
204 Videos that are more popular than that are not
205 downloaded.
206 Videos without view count information are always
207 downloaded. None for no limit.
208 download_archive: File name of a file where all downloads are recorded.
209 Videos already present in the file are not downloaded
210 again.
211 cookiefile: File name where cookies should be read from and dumped to.
212 nocheckcertificate:Do not verify SSL certificates
213 prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
214 At the moment, this is only supported by YouTube.
215 proxy: URL of the proxy server to use
216 geo_verification_proxy: URL of the proxy to use for IP address verification
217 on geo-restricted sites.
218 socket_timeout: Time to wait for unresponsive hosts, in seconds
219 bidi_workaround: Work around buggy terminals without bidirectional text
220 support, using fridibi
221 debug_printtraffic:Print out sent and received HTTP traffic
222 include_ads: Download ads as well
223 default_search: Prepend this string if an input url is not valid.
224 'auto' for elaborate guessing
225 encoding: Use this encoding instead of the system-specified.
226 extract_flat: Do not resolve URLs, return the immediate result.
227 Pass in 'in_playlist' to only show this behavior for
228 playlist items.
229 postprocessors: A list of dictionaries, each with an entry
230 * key: The name of the postprocessor. See
231 youtube_dl/postprocessor/__init__.py for a list.
232 as well as any further keyword arguments for the
233 postprocessor.
234 progress_hooks: A list of functions that get called on download
235 progress, with a dictionary with the entries
236 * status: One of "downloading", "error", or "finished".
237 Check this first and ignore unknown values.
238
239 If status is one of "downloading", or "finished", the
240 following properties may also be present:
241 * filename: The final filename (always present)
242 * tmpfilename: The filename we're currently writing to
243 * downloaded_bytes: Bytes on disk
244 * total_bytes: Size of the whole file, None if unknown
245 * total_bytes_estimate: Guess of the eventual file size,
246 None if unavailable.
247 * elapsed: The number of seconds since download started.
248 * eta: The estimated time in seconds, None if unknown
249 * speed: The download speed in bytes/second, None if
250 unknown
251 * fragment_index: The counter of the currently
252 downloaded video fragment.
253 * fragment_count: The number of fragments (= individual
254 files that will be merged)
255
256 Progress hooks are guaranteed to be called at least once
257 (with status "finished") if the download is successful.
258 merge_output_format: Extension to use when merging formats.
259 fixup: Automatically correct known faults of the file.
260 One of:
261 - "never": do nothing
262 - "warn": only emit a warning
263 - "detect_or_warn": check whether we can do anything
264 about it, warn otherwise (default)
265 source_address: Client-side IP address to bind to.
266 call_home: Boolean, true iff we are allowed to contact the
267 youtube-dl servers for debugging.
268 sleep_interval: Number of seconds to sleep before each download when
269 used alone or a lower bound of a range for randomized
270 sleep before each download (minimum possible number
271 of seconds to sleep) when used along with
272 max_sleep_interval.
273 max_sleep_interval:Upper bound of a range for randomized sleep before each
274 download (maximum possible number of seconds to sleep).
275 Must only be used along with sleep_interval.
276 Actual sleep time will be a random float from range
277 [sleep_interval; max_sleep_interval].
278 listformats: Print an overview of available video formats and exit.
279 list_thumbnails: Print a table of all thumbnails and exit.
280 match_filter: A function that gets called with the info_dict of
281 every video.
282 If it returns a message, the video is ignored.
283 If it returns None, the video is downloaded.
284 match_filter_func in utils.py is one example for this.
285 no_color: Do not emit color codes in output.
286 geo_bypass: Bypass geographic restriction via faking X-Forwarded-For
287 HTTP header
288 geo_bypass_country:
289 Two-letter ISO 3166-2 country code that will be used for
290 explicit geographic restriction bypassing via faking
291 X-Forwarded-For HTTP header
292 geo_bypass_ip_block:
293 IP range in CIDR notation that will be used similarly to
294 geo_bypass_country
295
296 The following options determine which downloader is picked:
297 external_downloader: Executable of the external downloader to call.
298 None or unset for standard (built-in) downloader.
299 hls_prefer_native: Use the native HLS downloader instead of ffmpeg/avconv
300 if True, otherwise use ffmpeg/avconv if False, otherwise
301 use downloader suggested by extractor if None.
302
303 The following parameters are not used by YoutubeDL itself, they are used by
304 the downloader (see youtube_dl/downloader/common.py):
305 nopart, updatetime, buffersize, ratelimit, min_filesize, max_filesize, test,
306 noresizebuffer, retries, continuedl, noprogress, consoletitle,
307 xattr_set_filesize, external_downloader_args, hls_use_mpegts,
308 http_chunk_size.
309
310 The following options are used by the post processors:
311 prefer_ffmpeg: If False, use avconv instead of ffmpeg if both are available,
312 otherwise prefer ffmpeg.
313 ffmpeg_location: Location of the ffmpeg/avconv binary; either the path
314 to the binary or its containing directory.
315 postprocessor_args: A list of additional command-line arguments for the
316 postprocessor.
317
318 The following options are used by the Youtube extractor:
319 youtube_include_dash_manifest: If True (default), DASH manifests and related
320 data will be downloaded and processed by extractor.
321 You can reduce network I/O by disabling it if you don't
322 care about DASH.
323 """
324
325 _NUMERIC_FIELDS = set((
326 'width', 'height', 'tbr', 'abr', 'asr', 'vbr', 'fps', 'filesize', 'filesize_approx',
327 'timestamp', 'upload_year', 'upload_month', 'upload_day',
328 'duration', 'view_count', 'like_count', 'dislike_count', 'repost_count',
329 'average_rating', 'comment_count', 'age_limit',
330 'start_time', 'end_time',
331 'chapter_number', 'season_number', 'episode_number',
332 'track_number', 'disc_number', 'release_year',
333 'playlist_index',
334 ))
335
336 params = None
337 _ies = []
338 _pps = []
339 _download_retcode = None
340 _num_downloads = None
341 _screen_file = None
342
343 def __init__(self, params=None, auto_init=True):
344 """Create a FileDownloader object with the given options."""
345 if params is None:
346 params = {}
347 self._ies = []
348 self._ies_instances = {}
349 self._pps = []
350 self._progress_hooks = []
351 self._download_retcode = 0
352 self._num_downloads = 0
353 self._screen_file = [sys.stdout, sys.stderr][params.get('logtostderr', False)]
354 self._err_file = sys.stderr
355 self.params = {
356 # Default parameters
357 'nocheckcertificate': False,
358 }
359 self.params.update(params)
360 self.cache = Cache(self)
361
362 def check_deprecated(param, option, suggestion):
363 if self.params.get(param) is not None:
364 self.report_warning(
365 '%s is deprecated. Use %s instead.' % (option, suggestion))
366 return True
367 return False
368
369 if check_deprecated('cn_verification_proxy', '--cn-verification-proxy', '--geo-verification-proxy'):
370 if self.params.get('geo_verification_proxy') is None:
371 self.params['geo_verification_proxy'] = self.params['cn_verification_proxy']
372
373 check_deprecated('autonumber_size', '--autonumber-size', 'output template with %(autonumber)0Nd, where N in the number of digits')
374 check_deprecated('autonumber', '--auto-number', '-o "%(autonumber)s-%(title)s.%(ext)s"')
375 check_deprecated('usetitle', '--title', '-o "%(title)s-%(id)s.%(ext)s"')
376
377 if params.get('bidi_workaround', False):
378 try:
379 import pty
380 master, slave = pty.openpty()
381 width = compat_get_terminal_size().columns
382 if width is None:
383 width_args = []
384 else:
385 width_args = ['-w', str(width)]
386 sp_kwargs = dict(
387 stdin=subprocess.PIPE,
388 stdout=slave,
389 stderr=self._err_file)
390 try:
391 self._output_process = subprocess.Popen(
392 ['bidiv'] + width_args, **sp_kwargs
393 )
394 except OSError:
395 self._output_process = subprocess.Popen(
396 ['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
397 self._output_channel = os.fdopen(master, 'rb')
398 except OSError as ose:
399 if ose.errno == errno.ENOENT:
400 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 $PATH.')
401 else:
402 raise
403
404 if (sys.platform != 'win32'
405 and sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968']
406 and not params.get('restrictfilenames', False)):
407 # Unicode filesystem API will throw errors (#1474, #13027)
408 self.report_warning(
409 'Assuming --restrict-filenames since file system encoding '
410 'cannot encode all characters. '
411 'Set the LC_ALL environment variable to fix this.')
412 self.params['restrictfilenames'] = True
413
414 if isinstance(params.get('outtmpl'), bytes):
415 self.report_warning(
416 'Parameter outtmpl is bytes, but should be a unicode string. '
417 'Put from __future__ import unicode_literals at the top of your code file or consider switching to Python 3.x.')
418
419 self._setup_opener()
420
421 if auto_init:
422 self.print_debug_header()
423 self.add_default_info_extractors()
424
425 for pp_def_raw in self.params.get('postprocessors', []):
426 pp_class = get_postprocessor(pp_def_raw['key'])
427 pp_def = dict(pp_def_raw)
428 del pp_def['key']
429 pp = pp_class(self, **compat_kwargs(pp_def))
430 self.add_post_processor(pp)
431
432 for ph in self.params.get('progress_hooks', []):
433 self.add_progress_hook(ph)
434
435 register_socks_protocols()
436
437 def warn_if_short_id(self, argv):
438 # short YouTube ID starting with dash?
439 idxs = [
440 i for i, a in enumerate(argv)
441 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
442 if idxs:
443 correct_argv = (
444 ['youtube-dl']
445 + [a for i, a in enumerate(argv) if i not in idxs]
446 + ['--'] + [argv[i] for i in idxs]
447 )
448 self.report_warning(
449 'Long argument string detected. '
450 'Use -- to separate parameters and URLs, like this:\n%s\n' %
451 args_to_str(correct_argv))
452
453 def add_info_extractor(self, ie):
454 """Add an InfoExtractor object to the end of the list."""
455 self._ies.append(ie)
456 if not isinstance(ie, type):
457 self._ies_instances[ie.ie_key()] = ie
458 ie.set_downloader(self)
459
460 def get_info_extractor(self, ie_key):
461 """
462 Get an instance of an IE with name ie_key, it will try to get one from
463 the _ies list, if there's no instance it will create a new one and add
464 it to the extractor list.
465 """
466 ie = self._ies_instances.get(ie_key)
467 if ie is None:
468 ie = get_info_extractor(ie_key)()
469 self.add_info_extractor(ie)
470 return ie
471
472 def add_default_info_extractors(self):
473 """
474 Add the InfoExtractors returned by gen_extractors to the end of the list
475 """
476 for ie in gen_extractor_classes():
477 self.add_info_extractor(ie)
478
479 def add_post_processor(self, pp):
480 """Add a PostProcessor object to the end of the chain."""
481 self._pps.append(pp)
482 pp.set_downloader(self)
483
484 def add_progress_hook(self, ph):
485 """Add the progress hook (currently only for the file downloader)"""
486 self._progress_hooks.append(ph)
487
488 def _bidi_workaround(self, message):
489 if not hasattr(self, '_output_channel'):
490 return message
491
492 assert hasattr(self, '_output_process')
493 assert isinstance(message, compat_str)
494 line_count = message.count('\n') + 1
495 self._output_process.stdin.write((message + '\n').encode('utf-8'))
496 self._output_process.stdin.flush()
497 res = ''.join(self._output_channel.readline().decode('utf-8')
498 for _ in range(line_count))
499 return res[:-len('\n')]
500
501 def to_screen(self, message, skip_eol=False):
502 """Print message to stdout if not in quiet mode."""
503 return self.to_stdout(message, skip_eol, check_quiet=True)
504
505 def _write_string(self, s, out=None):
506 write_string(s, out=out, encoding=self.params.get('encoding'))
507
508 def to_stdout(self, message, skip_eol=False, check_quiet=False):
509 """Print message to stdout if not in quiet mode."""
510 if self.params.get('logger'):
511 self.params['logger'].debug(message)
512 elif not check_quiet or not self.params.get('quiet', False):
513 message = self._bidi_workaround(message)
514 terminator = ['\n', ''][skip_eol]
515 output = message + terminator
516
517 self._write_string(output, self._screen_file)
518
519 def to_stderr(self, message):
520 """Print message to stderr."""
521 assert isinstance(message, compat_str)
522 if self.params.get('logger'):
523 self.params['logger'].error(message)
524 else:
525 message = self._bidi_workaround(message)
526 output = message + '\n'
527 self._write_string(output, self._err_file)
528
529 def to_console_title(self, message):
530 if not self.params.get('consoletitle', False):
531 return
532 if compat_os_name == 'nt':
533 if ctypes.windll.kernel32.GetConsoleWindow():
534 # c_wchar_p() might not be necessary if `message` is
535 # already of type unicode()
536 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
537 elif 'TERM' in os.environ:
538 self._write_string('\033]0;%s\007' % message, self._screen_file)
539
540 def save_console_title(self):
541 if not self.params.get('consoletitle', False):
542 return
543 if self.params.get('simulate', False):
544 return
545 if compat_os_name != 'nt' and 'TERM' in os.environ:
546 # Save the title on stack
547 self._write_string('\033[22;0t', self._screen_file)
548
549 def restore_console_title(self):
550 if not self.params.get('consoletitle', False):
551 return
552 if self.params.get('simulate', False):
553 return
554 if compat_os_name != 'nt' and 'TERM' in os.environ:
555 # Restore the title from stack
556 self._write_string('\033[23;0t', self._screen_file)
557
558 def __enter__(self):
559 self.save_console_title()
560 return self
561
562 def __exit__(self, *args):
563 self.restore_console_title()
564
565 if self.params.get('cookiefile') is not None:
566 self.cookiejar.save(ignore_discard=True, ignore_expires=True)
567
568 def trouble(self, message=None, tb=None):
569 """Determine action to take when a download problem appears.
570
571 Depending on if the downloader has been configured to ignore
572 download errors or not, this method may throw an exception or
573 not when errors are found, after printing the message.
574
575 tb, if given, is additional traceback information.
576 """
577 if message is not None:
578 self.to_stderr(message)
579 if self.params.get('verbose'):
580 if tb is None:
581 if sys.exc_info()[0]: # if .trouble has been called from an except block
582 tb = ''
583 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
584 tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
585 tb += encode_compat_str(traceback.format_exc())
586 else:
587 tb_data = traceback.format_list(traceback.extract_stack())
588 tb = ''.join(tb_data)
589 self.to_stderr(tb)
590 if not self.params.get('ignoreerrors', False):
591 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
592 exc_info = sys.exc_info()[1].exc_info
593 else:
594 exc_info = sys.exc_info()
595 raise DownloadError(message, exc_info)
596 self._download_retcode = 1
597
598 def report_warning(self, message):
599 '''
600 Print the message to stderr, it will be prefixed with 'WARNING:'
601 If stderr is a tty file the 'WARNING:' will be colored
602 '''
603 if self.params.get('logger') is not None:
604 self.params['logger'].warning(message)
605 else:
606 if self.params.get('no_warnings'):
607 return
608 if not self.params.get('no_color') and self._err_file.isatty() and compat_os_name != 'nt':
609 _msg_header = '\033[0;33mWARNING:\033[0m'
610 else:
611 _msg_header = 'WARNING:'
612 warning_message = '%s %s' % (_msg_header, message)
613 self.to_stderr(warning_message)
614
615 def report_error(self, message, tb=None):
616 '''
617 Do the same as trouble, but prefixes the message with 'ERROR:', colored
618 in red if stderr is a tty file.
619 '''
620 if not self.params.get('no_color') and self._err_file.isatty() and compat_os_name != 'nt':
621 _msg_header = '\033[0;31mERROR:\033[0m'
622 else:
623 _msg_header = 'ERROR:'
624 error_message = '%s %s' % (_msg_header, message)
625 self.trouble(error_message, tb)
626
627 def report_file_already_downloaded(self, file_name):
628 """Report file has already been fully downloaded."""
629 try:
630 self.to_screen('[download] %s has already been downloaded' % file_name)
631 except UnicodeEncodeError:
632 self.to_screen('[download] The file has already been downloaded')
633
634 def prepare_filename(self, info_dict):
635 """Generate the output filename."""
636 try:
637 template_dict = dict(info_dict)
638
639 template_dict['epoch'] = int(time.time())
640 autonumber_size = self.params.get('autonumber_size')
641 if autonumber_size is None:
642 autonumber_size = 5
643 template_dict['autonumber'] = self.params.get('autonumber_start', 1) - 1 + self._num_downloads
644 if template_dict.get('resolution') is None:
645 if template_dict.get('width') and template_dict.get('height'):
646 template_dict['resolution'] = '%dx%d' % (template_dict['width'], template_dict['height'])
647 elif template_dict.get('height'):
648 template_dict['resolution'] = '%sp' % template_dict['height']
649 elif template_dict.get('width'):
650 template_dict['resolution'] = '%dx?' % template_dict['width']
651
652 sanitize = lambda k, v: sanitize_filename(
653 compat_str(v),
654 restricted=self.params.get('restrictfilenames'),
655 is_id=(k == 'id' or k.endswith('_id')))
656 template_dict = dict((k, v if isinstance(v, compat_numeric_types) else sanitize(k, v))
657 for k, v in template_dict.items()
658 if v is not None and not isinstance(v, (list, tuple, dict)))
659 template_dict = collections.defaultdict(lambda: 'NA', template_dict)
660
661 outtmpl = self.params.get('outtmpl', DEFAULT_OUTTMPL)
662
663 # For fields playlist_index and autonumber convert all occurrences
664 # of %(field)s to %(field)0Nd for backward compatibility
665 field_size_compat_map = {
666 'playlist_index': len(str(template_dict['n_entries'])),
667 'autonumber': autonumber_size,
668 }
669 FIELD_SIZE_COMPAT_RE = r'(?<!%)%\((?P<field>autonumber|playlist_index)\)s'
670 mobj = re.search(FIELD_SIZE_COMPAT_RE, outtmpl)
671 if mobj:
672 outtmpl = re.sub(
673 FIELD_SIZE_COMPAT_RE,
674 r'%%(\1)0%dd' % field_size_compat_map[mobj.group('field')],
675 outtmpl)
676
677 # Missing numeric fields used together with integer presentation types
678 # in format specification will break the argument substitution since
679 # string 'NA' is returned for missing fields. We will patch output
680 # template for missing fields to meet string presentation type.
681 for numeric_field in self._NUMERIC_FIELDS:
682 if numeric_field not in template_dict:
683 # As of [1] format syntax is:
684 # %[mapping_key][conversion_flags][minimum_width][.precision][length_modifier]type
685 # 1. https://docs.python.org/2/library/stdtypes.html#string-formatting
686 FORMAT_RE = r'''(?x)
687 (?<!%)
688 %
689 \({0}\) # mapping key
690 (?:[#0\-+ ]+)? # conversion flags (optional)
691 (?:\d+)? # minimum field width (optional)
692 (?:\.\d+)? # precision (optional)
693 [hlL]? # length modifier (optional)
694 [diouxXeEfFgGcrs%] # conversion type
695 '''
696 outtmpl = re.sub(
697 FORMAT_RE.format(numeric_field),
698 r'%({0})s'.format(numeric_field), outtmpl)
699
700 # expand_path translates '%%' into '%' and '$$' into '$'
701 # correspondingly that is not what we want since we need to keep
702 # '%%' intact for template dict substitution step. Working around
703 # with boundary-alike separator hack.
704 sep = ''.join([random.choice(ascii_letters) for _ in range(32)])
705 outtmpl = outtmpl.replace('%%', '%{0}%'.format(sep)).replace('$$', '${0}$'.format(sep))
706
707 # outtmpl should be expand_path'ed before template dict substitution
708 # because meta fields may contain env variables we don't want to
709 # be expanded. For example, for outtmpl "%(title)s.%(ext)s" and
710 # title "Hello $PATH", we don't want `$PATH` to be expanded.
711 filename = expand_path(outtmpl).replace(sep, '') % template_dict
712
713 # Temporary fix for #4787
714 # 'Treat' all problem characters by passing filename through preferredencoding
715 # to workaround encoding issues with subprocess on python2 @ Windows
716 if sys.version_info < (3, 0) and sys.platform == 'win32':
717 filename = encodeFilename(filename, True).decode(preferredencoding())
718 return sanitize_path(filename)
719 except ValueError as err:
720 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
721 return None
722
723 def _match_entry(self, info_dict, incomplete):
724 """ Returns None iff the file should be downloaded """
725
726 video_title = info_dict.get('title', info_dict.get('id', 'video'))
727 if 'title' in info_dict:
728 # This can happen when we're just evaluating the playlist
729 title = info_dict['title']
730 matchtitle = self.params.get('matchtitle', False)
731 if matchtitle:
732 if not re.search(matchtitle, title, re.IGNORECASE):
733 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
734 rejecttitle = self.params.get('rejecttitle', False)
735 if rejecttitle:
736 if re.search(rejecttitle, title, re.IGNORECASE):
737 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
738 date = info_dict.get('upload_date')
739 if date is not None:
740 dateRange = self.params.get('daterange', DateRange())
741 if date not in dateRange:
742 return '%s upload date is not in range %s' % (date_from_str(date).isoformat(), dateRange)
743 view_count = info_dict.get('view_count')
744 if view_count is not None:
745 min_views = self.params.get('min_views')
746 if min_views is not None and view_count < min_views:
747 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
748 max_views = self.params.get('max_views')
749 if max_views is not None and view_count > max_views:
750 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
751 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
752 return 'Skipping "%s" because it is age restricted' % video_title
753 if self.in_download_archive(info_dict):
754 return '%s has already been recorded in archive' % video_title
755
756 if not incomplete:
757 match_filter = self.params.get('match_filter')
758 if match_filter is not None:
759 ret = match_filter(info_dict)
760 if ret is not None:
761 return ret
762
763 return None
764
765 @staticmethod
766 def add_extra_info(info_dict, extra_info):
767 '''Set the keys from extra_info in info dict if they are missing'''
768 for key, value in extra_info.items():
769 info_dict.setdefault(key, value)
770
771 def extract_info(self, url, download=True, ie_key=None, extra_info={},
772 process=True, force_generic_extractor=False):
773 '''
774 Returns a list with a dictionary for each video we find.
775 If 'download', also downloads the videos.
776 extra_info is a dict containing the extra values to add to each result
777 '''
778
779 if not ie_key and force_generic_extractor:
780 ie_key = 'Generic'
781
782 if ie_key:
783 ies = [self.get_info_extractor(ie_key)]
784 else:
785 ies = self._ies
786
787 for ie in ies:
788 if not ie.suitable(url):
789 continue
790
791 ie = self.get_info_extractor(ie.ie_key())
792 if not ie.working():
793 self.report_warning('The program functionality for this site has been marked as broken, '
794 'and will probably not work.')
795
796 try:
797 ie_result = ie.extract(url)
798 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
799 break
800 if isinstance(ie_result, list):
801 # Backwards compatibility: old IE result format
802 ie_result = {
803 '_type': 'compat_list',
804 'entries': ie_result,
805 }
806 self.add_default_extra_info(ie_result, ie, url)
807 if process:
808 return self.process_ie_result(ie_result, download, extra_info)
809 else:
810 return ie_result
811 except GeoRestrictedError as e:
812 msg = e.msg
813 if e.countries:
814 msg += '\nThis video is available in %s.' % ', '.join(
815 map(ISO3166Utils.short2full, e.countries))
816 msg += '\nYou might want to use a VPN or a proxy server (with --proxy) to workaround.'
817 self.report_error(msg)
818 break
819 except ExtractorError as e: # An error we somewhat expected
820 self.report_error(compat_str(e), e.format_traceback())
821 break
822 except MaxDownloadsReached:
823 raise
824 except Exception as e:
825 if self.params.get('ignoreerrors', False):
826 self.report_error(error_to_compat_str(e), tb=encode_compat_str(traceback.format_exc()))
827 break
828 else:
829 raise
830 else:
831 self.report_error('no suitable InfoExtractor for URL %s' % url)
832
833 def add_default_extra_info(self, ie_result, ie, url):
834 self.add_extra_info(ie_result, {
835 'extractor': ie.IE_NAME,
836 'webpage_url': url,
837 'webpage_url_basename': url_basename(url),
838 'extractor_key': ie.ie_key(),
839 })
840
841 def process_ie_result(self, ie_result, download=True, extra_info={}):
842 """
843 Take the result of the ie(may be modified) and resolve all unresolved
844 references (URLs, playlist items).
845
846 It will also download the videos if 'download'.
847 Returns the resolved ie_result.
848 """
849 result_type = ie_result.get('_type', 'video')
850
851 if result_type in ('url', 'url_transparent'):
852 ie_result['url'] = sanitize_url(ie_result['url'])
853 extract_flat = self.params.get('extract_flat', False)
854 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info)
855 or extract_flat is True):
856 self.__forced_printings(
857 ie_result, self.prepare_filename(ie_result),
858 incomplete=True)
859 return ie_result
860
861 if result_type == 'video':
862 self.add_extra_info(ie_result, extra_info)
863 return self.process_video_result(ie_result, download=download)
864 elif result_type == 'url':
865 # We have to add extra_info to the results because it may be
866 # contained in a playlist
867 return self.extract_info(ie_result['url'],
868 download,
869 ie_key=ie_result.get('ie_key'),
870 extra_info=extra_info)
871 elif result_type == 'url_transparent':
872 # Use the information from the embedding page
873 info = self.extract_info(
874 ie_result['url'], ie_key=ie_result.get('ie_key'),
875 extra_info=extra_info, download=False, process=False)
876
877 # extract_info may return None when ignoreerrors is enabled and
878 # extraction failed with an error, don't crash and return early
879 # in this case
880 if not info:
881 return info
882
883 force_properties = dict(
884 (k, v) for k, v in ie_result.items() if v is not None)
885 for f in ('_type', 'url', 'id', 'extractor', 'extractor_key', 'ie_key'):
886 if f in force_properties:
887 del force_properties[f]
888 new_result = info.copy()
889 new_result.update(force_properties)
890
891 # Extracted info may not be a video result (i.e.
892 # info.get('_type', 'video') != video) but rather an url or
893 # url_transparent. In such cases outer metadata (from ie_result)
894 # should be propagated to inner one (info). For this to happen
895 # _type of info should be overridden with url_transparent. This
896 # fixes issue from https://github.com/ytdl-org/youtube-dl/pull/11163.
897 if new_result.get('_type') == 'url':
898 new_result['_type'] = 'url_transparent'
899
900 return self.process_ie_result(
901 new_result, download=download, extra_info=extra_info)
902 elif result_type in ('playlist', 'multi_video'):
903 # We process each entry in the playlist
904 playlist = ie_result.get('title') or ie_result.get('id')
905 self.to_screen('[download] Downloading playlist: %s' % playlist)
906
907 playlist_results = []
908
909 playliststart = self.params.get('playliststart', 1) - 1
910 playlistend = self.params.get('playlistend')
911 # For backwards compatibility, interpret -1 as whole list
912 if playlistend == -1:
913 playlistend = None
914
915 playlistitems_str = self.params.get('playlist_items')
916 playlistitems = None
917 if playlistitems_str is not None:
918 def iter_playlistitems(format):
919 for string_segment in format.split(','):
920 if '-' in string_segment:
921 start, end = string_segment.split('-')
922 for item in range(int(start), int(end) + 1):
923 yield int(item)
924 else:
925 yield int(string_segment)
926 playlistitems = orderedSet(iter_playlistitems(playlistitems_str))
927
928 ie_entries = ie_result['entries']
929
930 def make_playlistitems_entries(list_ie_entries):
931 num_entries = len(list_ie_entries)
932 return [
933 list_ie_entries[i - 1] for i in playlistitems
934 if -num_entries <= i - 1 < num_entries]
935
936 def report_download(num_entries):
937 self.to_screen(
938 '[%s] playlist %s: Downloading %d videos' %
939 (ie_result['extractor'], playlist, num_entries))
940
941 if isinstance(ie_entries, list):
942 n_all_entries = len(ie_entries)
943 if playlistitems:
944 entries = make_playlistitems_entries(ie_entries)
945 else:
946 entries = ie_entries[playliststart:playlistend]
947 n_entries = len(entries)
948 self.to_screen(
949 '[%s] playlist %s: Collected %d video ids (downloading %d of them)' %
950 (ie_result['extractor'], playlist, n_all_entries, n_entries))
951 elif isinstance(ie_entries, PagedList):
952 if playlistitems:
953 entries = []
954 for item in playlistitems:
955 entries.extend(ie_entries.getslice(
956 item - 1, item
957 ))
958 else:
959 entries = ie_entries.getslice(
960 playliststart, playlistend)
961 n_entries = len(entries)
962 report_download(n_entries)
963 else: # iterable
964 if playlistitems:
965 entries = make_playlistitems_entries(list(itertools.islice(
966 ie_entries, 0, max(playlistitems))))
967 else:
968 entries = list(itertools.islice(
969 ie_entries, playliststart, playlistend))
970 n_entries = len(entries)
971 report_download(n_entries)
972
973 if self.params.get('playlistreverse', False):
974 entries = entries[::-1]
975
976 if self.params.get('playlistrandom', False):
977 random.shuffle(entries)
978
979 x_forwarded_for = ie_result.get('__x_forwarded_for_ip')
980
981 for i, entry in enumerate(entries, 1):
982 self.to_screen('[download] Downloading video %s of %s' % (i, n_entries))
983 # This __x_forwarded_for_ip thing is a bit ugly but requires
984 # minimal changes
985 if x_forwarded_for:
986 entry['__x_forwarded_for_ip'] = x_forwarded_for
987 extra = {
988 'n_entries': n_entries,
989 'playlist': playlist,
990 'playlist_id': ie_result.get('id'),
991 'playlist_title': ie_result.get('title'),
992 'playlist_uploader': ie_result.get('uploader'),
993 'playlist_uploader_id': ie_result.get('uploader_id'),
994 'playlist_index': playlistitems[i - 1] if playlistitems else i + playliststart,
995 'extractor': ie_result['extractor'],
996 'webpage_url': ie_result['webpage_url'],
997 'webpage_url_basename': url_basename(ie_result['webpage_url']),
998 'extractor_key': ie_result['extractor_key'],
999 }
1000
1001 reason = self._match_entry(entry, incomplete=True)
1002 if reason is not None:
1003 self.to_screen('[download] ' + reason)
1004 continue
1005
1006 entry_result = self.process_ie_result(entry,
1007 download=download,
1008 extra_info=extra)
1009 playlist_results.append(entry_result)
1010 ie_result['entries'] = playlist_results
1011 self.to_screen('[download] Finished downloading playlist: %s' % playlist)
1012 return ie_result
1013 elif result_type == 'compat_list':
1014 self.report_warning(
1015 'Extractor %s returned a compat_list result. '
1016 'It needs to be updated.' % ie_result.get('extractor'))
1017
1018 def _fixup(r):
1019 self.add_extra_info(
1020 r,
1021 {
1022 'extractor': ie_result['extractor'],
1023 'webpage_url': ie_result['webpage_url'],
1024 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1025 'extractor_key': ie_result['extractor_key'],
1026 }
1027 )
1028 return r
1029 ie_result['entries'] = [
1030 self.process_ie_result(_fixup(r), download, extra_info)
1031 for r in ie_result['entries']
1032 ]
1033 return ie_result
1034 else:
1035 raise Exception('Invalid result type: %s' % result_type)
1036
1037 def _build_format_filter(self, filter_spec):
1038 " Returns a function to filter the formats according to the filter_spec "
1039
1040 OPERATORS = {
1041 '<': operator.lt,
1042 '<=': operator.le,
1043 '>': operator.gt,
1044 '>=': operator.ge,
1045 '=': operator.eq,
1046 '!=': operator.ne,
1047 }
1048 operator_rex = re.compile(r'''(?x)\s*
1049 (?P<key>width|height|tbr|abr|vbr|asr|filesize|filesize_approx|fps)
1050 \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
1051 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)
1052 $
1053 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
1054 m = operator_rex.search(filter_spec)
1055 if m:
1056 try:
1057 comparison_value = int(m.group('value'))
1058 except ValueError:
1059 comparison_value = parse_filesize(m.group('value'))
1060 if comparison_value is None:
1061 comparison_value = parse_filesize(m.group('value') + 'B')
1062 if comparison_value is None:
1063 raise ValueError(
1064 'Invalid value %r in format specification %r' % (
1065 m.group('value'), filter_spec))
1066 op = OPERATORS[m.group('op')]
1067
1068 if not m:
1069 STR_OPERATORS = {
1070 '=': operator.eq,
1071 '^=': lambda attr, value: attr.startswith(value),
1072 '$=': lambda attr, value: attr.endswith(value),
1073 '*=': lambda attr, value: value in attr,
1074 }
1075 str_operator_rex = re.compile(r'''(?x)
1076 \s*(?P<key>ext|acodec|vcodec|container|protocol|format_id)
1077 \s*(?P<negation>!\s*)?(?P<op>%s)(?P<none_inclusive>\s*\?)?
1078 \s*(?P<value>[a-zA-Z0-9._-]+)
1079 \s*$
1080 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
1081 m = str_operator_rex.search(filter_spec)
1082 if m:
1083 comparison_value = m.group('value')
1084 str_op = STR_OPERATORS[m.group('op')]
1085 if m.group('negation'):
1086 op = lambda attr, value: not str_op(attr, value)
1087 else:
1088 op = str_op
1089
1090 if not m:
1091 raise ValueError('Invalid filter specification %r' % filter_spec)
1092
1093 def _filter(f):
1094 actual_value = f.get(m.group('key'))
1095 if actual_value is None:
1096 return m.group('none_inclusive')
1097 return op(actual_value, comparison_value)
1098 return _filter
1099
1100 def _default_format_spec(self, info_dict, download=True):
1101
1102 def can_merge():
1103 merger = FFmpegMergerPP(self)
1104 return merger.available and merger.can_merge()
1105
1106 def prefer_best():
1107 if self.params.get('simulate', False):
1108 return False
1109 if not download:
1110 return False
1111 if self.params.get('outtmpl', DEFAULT_OUTTMPL) == '-':
1112 return True
1113 if info_dict.get('is_live'):
1114 return True
1115 if not can_merge():
1116 return True
1117 return False
1118
1119 req_format_list = ['bestvideo+bestaudio', 'best']
1120 if prefer_best():
1121 req_format_list.reverse()
1122 return '/'.join(req_format_list)
1123
1124 def build_format_selector(self, format_spec):
1125 def syntax_error(note, start):
1126 message = (
1127 'Invalid format specification: '
1128 '{0}\n\t{1}\n\t{2}^'.format(note, format_spec, ' ' * start[1]))
1129 return SyntaxError(message)
1130
1131 PICKFIRST = 'PICKFIRST'
1132 MERGE = 'MERGE'
1133 SINGLE = 'SINGLE'
1134 GROUP = 'GROUP'
1135 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
1136
1137 def _parse_filter(tokens):
1138 filter_parts = []
1139 for type, string, start, _, _ in tokens:
1140 if type == tokenize.OP and string == ']':
1141 return ''.join(filter_parts)
1142 else:
1143 filter_parts.append(string)
1144
1145 def _remove_unused_ops(tokens):
1146 # Remove operators that we don't use and join them with the surrounding strings
1147 # for example: 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
1148 ALLOWED_OPS = ('/', '+', ',', '(', ')')
1149 last_string, last_start, last_end, last_line = None, None, None, None
1150 for type, string, start, end, line in tokens:
1151 if type == tokenize.OP and string == '[':
1152 if last_string:
1153 yield tokenize.NAME, last_string, last_start, last_end, last_line
1154 last_string = None
1155 yield type, string, start, end, line
1156 # everything inside brackets will be handled by _parse_filter
1157 for type, string, start, end, line in tokens:
1158 yield type, string, start, end, line
1159 if type == tokenize.OP and string == ']':
1160 break
1161 elif type == tokenize.OP and string in ALLOWED_OPS:
1162 if last_string:
1163 yield tokenize.NAME, last_string, last_start, last_end, last_line
1164 last_string = None
1165 yield type, string, start, end, line
1166 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
1167 if not last_string:
1168 last_string = string
1169 last_start = start
1170 last_end = end
1171 else:
1172 last_string += string
1173 if last_string:
1174 yield tokenize.NAME, last_string, last_start, last_end, last_line
1175
1176 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
1177 selectors = []
1178 current_selector = None
1179 for type, string, start, _, _ in tokens:
1180 # ENCODING is only defined in python 3.x
1181 if type == getattr(tokenize, 'ENCODING', None):
1182 continue
1183 elif type in [tokenize.NAME, tokenize.NUMBER]:
1184 current_selector = FormatSelector(SINGLE, string, [])
1185 elif type == tokenize.OP:
1186 if string == ')':
1187 if not inside_group:
1188 # ')' will be handled by the parentheses group
1189 tokens.restore_last_token()
1190 break
1191 elif inside_merge and string in ['/', ',']:
1192 tokens.restore_last_token()
1193 break
1194 elif inside_choice and string == ',':
1195 tokens.restore_last_token()
1196 break
1197 elif string == ',':
1198 if not current_selector:
1199 raise syntax_error('"," must follow a format selector', start)
1200 selectors.append(current_selector)
1201 current_selector = None
1202 elif string == '/':
1203 if not current_selector:
1204 raise syntax_error('"/" must follow a format selector', start)
1205 first_choice = current_selector
1206 second_choice = _parse_format_selection(tokens, inside_choice=True)
1207 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
1208 elif string == '[':
1209 if not current_selector:
1210 current_selector = FormatSelector(SINGLE, 'best', [])
1211 format_filter = _parse_filter(tokens)
1212 current_selector.filters.append(format_filter)
1213 elif string == '(':
1214 if current_selector:
1215 raise syntax_error('Unexpected "("', start)
1216 group = _parse_format_selection(tokens, inside_group=True)
1217 current_selector = FormatSelector(GROUP, group, [])
1218 elif string == '+':
1219 video_selector = current_selector
1220 audio_selector = _parse_format_selection(tokens, inside_merge=True)
1221 if not video_selector or not audio_selector:
1222 raise syntax_error('"+" must be between two format selectors', start)
1223 current_selector = FormatSelector(MERGE, (video_selector, audio_selector), [])
1224 else:
1225 raise syntax_error('Operator not recognized: "{0}"'.format(string), start)
1226 elif type == tokenize.ENDMARKER:
1227 break
1228 if current_selector:
1229 selectors.append(current_selector)
1230 return selectors
1231
1232 def _build_selector_function(selector):
1233 if isinstance(selector, list):
1234 fs = [_build_selector_function(s) for s in selector]
1235
1236 def selector_function(ctx):
1237 for f in fs:
1238 for format in f(ctx):
1239 yield format
1240 return selector_function
1241 elif selector.type == GROUP:
1242 selector_function = _build_selector_function(selector.selector)
1243 elif selector.type == PICKFIRST:
1244 fs = [_build_selector_function(s) for s in selector.selector]
1245
1246 def selector_function(ctx):
1247 for f in fs:
1248 picked_formats = list(f(ctx))
1249 if picked_formats:
1250 return picked_formats
1251 return []
1252 elif selector.type == SINGLE:
1253 format_spec = selector.selector
1254
1255 def selector_function(ctx):
1256 formats = list(ctx['formats'])
1257 if not formats:
1258 return
1259 if format_spec == 'all':
1260 for f in formats:
1261 yield f
1262 elif format_spec in ['best', 'worst', None]:
1263 format_idx = 0 if format_spec == 'worst' else -1
1264 audiovideo_formats = [
1265 f for f in formats
1266 if f.get('vcodec') != 'none' and f.get('acodec') != 'none']
1267 if audiovideo_formats:
1268 yield audiovideo_formats[format_idx]
1269 # for extractors with incomplete formats (audio only (soundcloud)
1270 # or video only (imgur)) we will fallback to best/worst
1271 # {video,audio}-only format
1272 elif ctx['incomplete_formats']:
1273 yield formats[format_idx]
1274 elif format_spec == 'bestaudio':
1275 audio_formats = [
1276 f for f in formats
1277 if f.get('vcodec') == 'none']
1278 if audio_formats:
1279 yield audio_formats[-1]
1280 elif format_spec == 'worstaudio':
1281 audio_formats = [
1282 f for f in formats
1283 if f.get('vcodec') == 'none']
1284 if audio_formats:
1285 yield audio_formats[0]
1286 elif format_spec == 'bestvideo':
1287 video_formats = [
1288 f for f in formats
1289 if f.get('acodec') == 'none']
1290 if video_formats:
1291 yield video_formats[-1]
1292 elif format_spec == 'worstvideo':
1293 video_formats = [
1294 f for f in formats
1295 if f.get('acodec') == 'none']
1296 if video_formats:
1297 yield video_formats[0]
1298 else:
1299 extensions = ['mp4', 'flv', 'webm', '3gp', 'm4a', 'mp3', 'ogg', 'aac', 'wav']
1300 if format_spec in extensions:
1301 filter_f = lambda f: f['ext'] == format_spec
1302 else:
1303 filter_f = lambda f: f['format_id'] == format_spec
1304 matches = list(filter(filter_f, formats))
1305 if matches:
1306 yield matches[-1]
1307 elif selector.type == MERGE:
1308 def _merge(formats_info):
1309 format_1, format_2 = [f['format_id'] for f in formats_info]
1310 # The first format must contain the video and the
1311 # second the audio
1312 if formats_info[0].get('vcodec') == 'none':
1313 self.report_error('The first format must '
1314 'contain the video, try using '
1315 '"-f %s+%s"' % (format_2, format_1))
1316 return
1317 # Formats must be opposite (video+audio)
1318 if formats_info[0].get('acodec') == 'none' and formats_info[1].get('acodec') == 'none':
1319 self.report_error(
1320 'Both formats %s and %s are video-only, you must specify "-f video+audio"'
1321 % (format_1, format_2))
1322 return
1323 output_ext = (
1324 formats_info[0]['ext']
1325 if self.params.get('merge_output_format') is None
1326 else self.params['merge_output_format'])
1327 return {
1328 'requested_formats': formats_info,
1329 'format': '%s+%s' % (formats_info[0].get('format'),
1330 formats_info[1].get('format')),
1331 'format_id': '%s+%s' % (formats_info[0].get('format_id'),
1332 formats_info[1].get('format_id')),
1333 'width': formats_info[0].get('width'),
1334 'height': formats_info[0].get('height'),
1335 'resolution': formats_info[0].get('resolution'),
1336 'fps': formats_info[0].get('fps'),
1337 'vcodec': formats_info[0].get('vcodec'),
1338 'vbr': formats_info[0].get('vbr'),
1339 'stretched_ratio': formats_info[0].get('stretched_ratio'),
1340 'acodec': formats_info[1].get('acodec'),
1341 'abr': formats_info[1].get('abr'),
1342 'ext': output_ext,
1343 }
1344 video_selector, audio_selector = map(_build_selector_function, selector.selector)
1345
1346 def selector_function(ctx):
1347 for pair in itertools.product(
1348 video_selector(copy.deepcopy(ctx)), audio_selector(copy.deepcopy(ctx))):
1349 yield _merge(pair)
1350
1351 filters = [self._build_format_filter(f) for f in selector.filters]
1352
1353 def final_selector(ctx):
1354 ctx_copy = copy.deepcopy(ctx)
1355 for _filter in filters:
1356 ctx_copy['formats'] = list(filter(_filter, ctx_copy['formats']))
1357 return selector_function(ctx_copy)
1358 return final_selector
1359
1360 stream = io.BytesIO(format_spec.encode('utf-8'))
1361 try:
1362 tokens = list(_remove_unused_ops(compat_tokenize_tokenize(stream.readline)))
1363 except tokenize.TokenError:
1364 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
1365
1366 class TokenIterator(object):
1367 def __init__(self, tokens):
1368 self.tokens = tokens
1369 self.counter = 0
1370
1371 def __iter__(self):
1372 return self
1373
1374 def __next__(self):
1375 if self.counter >= len(self.tokens):
1376 raise StopIteration()
1377 value = self.tokens[self.counter]
1378 self.counter += 1
1379 return value
1380
1381 next = __next__
1382
1383 def restore_last_token(self):
1384 self.counter -= 1
1385
1386 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
1387 return _build_selector_function(parsed_selector)
1388
1389 def _calc_headers(self, info_dict):
1390 res = std_headers.copy()
1391
1392 add_headers = info_dict.get('http_headers')
1393 if add_headers:
1394 res.update(add_headers)
1395
1396 cookies = self._calc_cookies(info_dict)
1397 if cookies:
1398 res['Cookie'] = cookies
1399
1400 if 'X-Forwarded-For' not in res:
1401 x_forwarded_for_ip = info_dict.get('__x_forwarded_for_ip')
1402 if x_forwarded_for_ip:
1403 res['X-Forwarded-For'] = x_forwarded_for_ip
1404
1405 return res
1406
1407 def _calc_cookies(self, info_dict):
1408 pr = sanitized_Request(info_dict['url'])
1409 self.cookiejar.add_cookie_header(pr)
1410 return pr.get_header('Cookie')
1411
1412 def process_video_result(self, info_dict, download=True):
1413 assert info_dict.get('_type', 'video') == 'video'
1414
1415 if 'id' not in info_dict:
1416 raise ExtractorError('Missing "id" field in extractor result')
1417 if 'title' not in info_dict:
1418 raise ExtractorError('Missing "title" field in extractor result')
1419
1420 def report_force_conversion(field, field_not, conversion):
1421 self.report_warning(
1422 '"%s" field is not %s - forcing %s conversion, there is an error in extractor'
1423 % (field, field_not, conversion))
1424
1425 def sanitize_string_field(info, string_field):
1426 field = info.get(string_field)
1427 if field is None or isinstance(field, compat_str):
1428 return
1429 report_force_conversion(string_field, 'a string', 'string')
1430 info[string_field] = compat_str(field)
1431
1432 def sanitize_numeric_fields(info):
1433 for numeric_field in self._NUMERIC_FIELDS:
1434 field = info.get(numeric_field)
1435 if field is None or isinstance(field, compat_numeric_types):
1436 continue
1437 report_force_conversion(numeric_field, 'numeric', 'int')
1438 info[numeric_field] = int_or_none(field)
1439
1440 sanitize_string_field(info_dict, 'id')
1441 sanitize_numeric_fields(info_dict)
1442
1443 if 'playlist' not in info_dict:
1444 # It isn't part of a playlist
1445 info_dict['playlist'] = None
1446 info_dict['playlist_index'] = None
1447
1448 thumbnails = info_dict.get('thumbnails')
1449 if thumbnails is None:
1450 thumbnail = info_dict.get('thumbnail')
1451 if thumbnail:
1452 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
1453 if thumbnails:
1454 thumbnails.sort(key=lambda t: (
1455 t.get('preference') if t.get('preference') is not None else -1,
1456 t.get('width') if t.get('width') is not None else -1,
1457 t.get('height') if t.get('height') is not None else -1,
1458 t.get('id') if t.get('id') is not None else '', t.get('url')))
1459 for i, t in enumerate(thumbnails):
1460 t['url'] = sanitize_url(t['url'])
1461 if t.get('width') and t.get('height'):
1462 t['resolution'] = '%dx%d' % (t['width'], t['height'])
1463 if t.get('id') is None:
1464 t['id'] = '%d' % i
1465
1466 if self.params.get('list_thumbnails'):
1467 self.list_thumbnails(info_dict)
1468 return
1469
1470 thumbnail = info_dict.get('thumbnail')
1471 if thumbnail:
1472 info_dict['thumbnail'] = sanitize_url(thumbnail)
1473 elif thumbnails:
1474 info_dict['thumbnail'] = thumbnails[-1]['url']
1475
1476 if 'display_id' not in info_dict and 'id' in info_dict:
1477 info_dict['display_id'] = info_dict['id']
1478
1479 if info_dict.get('upload_date') is None and info_dict.get('timestamp') is not None:
1480 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
1481 # see http://bugs.python.org/issue1646728)
1482 try:
1483 upload_date = datetime.datetime.utcfromtimestamp(info_dict['timestamp'])
1484 info_dict['upload_date'] = upload_date.strftime('%Y%m%d')
1485 except (ValueError, OverflowError, OSError):
1486 pass
1487
1488 # Auto generate title fields corresponding to the *_number fields when missing
1489 # in order to always have clean titles. This is very common for TV series.
1490 for field in ('chapter', 'season', 'episode'):
1491 if info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
1492 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
1493
1494 for cc_kind in ('subtitles', 'automatic_captions'):
1495 cc = info_dict.get(cc_kind)
1496 if cc:
1497 for _, subtitle in cc.items():
1498 for subtitle_format in subtitle:
1499 if subtitle_format.get('url'):
1500 subtitle_format['url'] = sanitize_url(subtitle_format['url'])
1501 if subtitle_format.get('ext') is None:
1502 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
1503
1504 automatic_captions = info_dict.get('automatic_captions')
1505 subtitles = info_dict.get('subtitles')
1506
1507 if self.params.get('listsubtitles', False):
1508 if 'automatic_captions' in info_dict:
1509 self.list_subtitles(
1510 info_dict['id'], automatic_captions, 'automatic captions')
1511 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
1512 return
1513
1514 info_dict['requested_subtitles'] = self.process_subtitles(
1515 info_dict['id'], subtitles, automatic_captions)
1516
1517 # We now pick which formats have to be downloaded
1518 if info_dict.get('formats') is None:
1519 # There's only one format available
1520 formats = [info_dict]
1521 else:
1522 formats = info_dict['formats']
1523
1524 if not formats:
1525 raise ExtractorError('No video formats found!')
1526
1527 def is_wellformed(f):
1528 url = f.get('url')
1529 if not url:
1530 self.report_warning(
1531 '"url" field is missing or empty - skipping format, '
1532 'there is an error in extractor')
1533 return False
1534 if isinstance(url, bytes):
1535 sanitize_string_field(f, 'url')
1536 return True
1537
1538 # Filter out malformed formats for better extraction robustness
1539 formats = list(filter(is_wellformed, formats))
1540
1541 formats_dict = {}
1542
1543 # We check that all the formats have the format and format_id fields
1544 for i, format in enumerate(formats):
1545 sanitize_string_field(format, 'format_id')
1546 sanitize_numeric_fields(format)
1547 format['url'] = sanitize_url(format['url'])
1548 if not format.get('format_id'):
1549 format['format_id'] = compat_str(i)
1550 else:
1551 # Sanitize format_id from characters used in format selector expression
1552 format['format_id'] = re.sub(r'[\s,/+\[\]()]', '_', format['format_id'])
1553 format_id = format['format_id']
1554 if format_id not in formats_dict:
1555 formats_dict[format_id] = []
1556 formats_dict[format_id].append(format)
1557
1558 # Make sure all formats have unique format_id
1559 for format_id, ambiguous_formats in formats_dict.items():
1560 if len(ambiguous_formats) > 1:
1561 for i, format in enumerate(ambiguous_formats):
1562 format['format_id'] = '%s-%d' % (format_id, i)
1563
1564 for i, format in enumerate(formats):
1565 if format.get('format') is None:
1566 format['format'] = '{id} - {res}{note}'.format(
1567 id=format['format_id'],
1568 res=self.format_resolution(format),
1569 note=' ({0})'.format(format['format_note']) if format.get('format_note') is not None else '',
1570 )
1571 # Automatically determine file extension if missing
1572 if format.get('ext') is None:
1573 format['ext'] = determine_ext(format['url']).lower()
1574 # Automatically determine protocol if missing (useful for format
1575 # selection purposes)
1576 if format.get('protocol') is None:
1577 format['protocol'] = determine_protocol(format)
1578 # Add HTTP headers, so that external programs can use them from the
1579 # json output
1580 full_format_info = info_dict.copy()
1581 full_format_info.update(format)
1582 format['http_headers'] = self._calc_headers(full_format_info)
1583 # Remove private housekeeping stuff
1584 if '__x_forwarded_for_ip' in info_dict:
1585 del info_dict['__x_forwarded_for_ip']
1586
1587 # TODO Central sorting goes here
1588
1589 if formats[0] is not info_dict:
1590 # only set the 'formats' fields if the original info_dict list them
1591 # otherwise we end up with a circular reference, the first (and unique)
1592 # element in the 'formats' field in info_dict is info_dict itself,
1593 # which can't be exported to json
1594 info_dict['formats'] = formats
1595 if self.params.get('listformats'):
1596 self.list_formats(info_dict)
1597 return
1598
1599 req_format = self.params.get('format')
1600 if req_format is None:
1601 req_format = self._default_format_spec(info_dict, download=download)
1602 if self.params.get('verbose'):
1603 self.to_stdout('[debug] Default format spec: %s' % req_format)
1604
1605 format_selector = self.build_format_selector(req_format)
1606
1607 # While in format selection we may need to have an access to the original
1608 # format set in order to calculate some metrics or do some processing.
1609 # For now we need to be able to guess whether original formats provided
1610 # by extractor are incomplete or not (i.e. whether extractor provides only
1611 # video-only or audio-only formats) for proper formats selection for
1612 # extractors with such incomplete formats (see
1613 # https://github.com/ytdl-org/youtube-dl/pull/5556).
1614 # Since formats may be filtered during format selection and may not match
1615 # the original formats the results may be incorrect. Thus original formats
1616 # or pre-calculated metrics should be passed to format selection routines
1617 # as well.
1618 # We will pass a context object containing all necessary additional data
1619 # instead of just formats.
1620 # This fixes incorrect format selection issue (see
1621 # https://github.com/ytdl-org/youtube-dl/issues/10083).
1622 incomplete_formats = (
1623 # All formats are video-only or
1624 all(f.get('vcodec') != 'none' and f.get('acodec') == 'none' for f in formats)
1625 # all formats are audio-only
1626 or all(f.get('vcodec') == 'none' and f.get('acodec') != 'none' for f in formats))
1627
1628 ctx = {
1629 'formats': formats,
1630 'incomplete_formats': incomplete_formats,
1631 }
1632
1633 formats_to_download = list(format_selector(ctx))
1634 if not formats_to_download:
1635 raise ExtractorError('requested format not available',
1636 expected=True)
1637
1638 if download:
1639 if len(formats_to_download) > 1:
1640 self.to_screen('[info] %s: downloading video in %s formats' % (info_dict['id'], len(formats_to_download)))
1641 for format in formats_to_download:
1642 new_info = dict(info_dict)
1643 new_info.update(format)
1644 self.process_info(new_info)
1645 # We update the info dict with the best quality format (backwards compatibility)
1646 info_dict.update(formats_to_download[-1])
1647 return info_dict
1648
1649 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
1650 """Select the requested subtitles and their format"""
1651 available_subs = {}
1652 if normal_subtitles and self.params.get('writesubtitles'):
1653 available_subs.update(normal_subtitles)
1654 if automatic_captions and self.params.get('writeautomaticsub'):
1655 for lang, cap_info in automatic_captions.items():
1656 if lang not in available_subs:
1657 available_subs[lang] = cap_info
1658
1659 if (not self.params.get('writesubtitles') and not
1660 self.params.get('writeautomaticsub') or not
1661 available_subs):
1662 return None
1663
1664 if self.params.get('allsubtitles', False):
1665 requested_langs = available_subs.keys()
1666 else:
1667 if self.params.get('subtitleslangs', False):
1668 requested_langs = self.params.get('subtitleslangs')
1669 elif 'en' in available_subs:
1670 requested_langs = ['en']
1671 else:
1672 requested_langs = [list(available_subs.keys())[0]]
1673
1674 formats_query = self.params.get('subtitlesformat', 'best')
1675 formats_preference = formats_query.split('/') if formats_query else []
1676 subs = {}
1677 for lang in requested_langs:
1678 formats = available_subs.get(lang)
1679 if formats is None:
1680 self.report_warning('%s subtitles not available for %s' % (lang, video_id))
1681 continue
1682 for ext in formats_preference:
1683 if ext == 'best':
1684 f = formats[-1]
1685 break
1686 matches = list(filter(lambda f: f['ext'] == ext, formats))
1687 if matches:
1688 f = matches[-1]
1689 break
1690 else:
1691 f = formats[-1]
1692 self.report_warning(
1693 'No subtitle format found matching "%s" for language %s, '
1694 'using %s' % (formats_query, lang, f['ext']))
1695 subs[lang] = f
1696 return subs
1697
1698 def __forced_printings(self, info_dict, filename, incomplete):
1699 def print_mandatory(field):
1700 if (self.params.get('force%s' % field, False)
1701 and (not incomplete or info_dict.get(field) is not None)):
1702 self.to_stdout(info_dict[field])
1703
1704 def print_optional(field):
1705 if (self.params.get('force%s' % field, False)
1706 and info_dict.get(field) is not None):
1707 self.to_stdout(info_dict[field])
1708
1709 print_mandatory('title')
1710 print_mandatory('id')
1711 if self.params.get('forceurl', False) and not incomplete:
1712 if info_dict.get('requested_formats') is not None:
1713 for f in info_dict['requested_formats']:
1714 self.to_stdout(f['url'] + f.get('play_path', ''))
1715 else:
1716 # For RTMP URLs, also include the playpath
1717 self.to_stdout(info_dict['url'] + info_dict.get('play_path', ''))
1718 print_optional('thumbnail')
1719 print_optional('description')
1720 if self.params.get('forcefilename', False) and filename is not None:
1721 self.to_stdout(filename)
1722 if self.params.get('forceduration', False) and info_dict.get('duration') is not None:
1723 self.to_stdout(formatSeconds(info_dict['duration']))
1724 print_mandatory('format')
1725 if self.params.get('forcejson', False):
1726 self.to_stdout(json.dumps(info_dict))
1727
1728 def process_info(self, info_dict):
1729 """Process a single resolved IE result."""
1730
1731 assert info_dict.get('_type', 'video') == 'video'
1732
1733 max_downloads = self.params.get('max_downloads')
1734 if max_downloads is not None:
1735 if self._num_downloads >= int(max_downloads):
1736 raise MaxDownloadsReached()
1737
1738 # TODO: backward compatibility, to be removed
1739 info_dict['fulltitle'] = info_dict['title']
1740
1741 if 'format' not in info_dict:
1742 info_dict['format'] = info_dict['ext']
1743
1744 reason = self._match_entry(info_dict, incomplete=False)
1745 if reason is not None:
1746 self.to_screen('[download] ' + reason)
1747 return
1748
1749 self._num_downloads += 1
1750
1751 info_dict['_filename'] = filename = self.prepare_filename(info_dict)
1752
1753 # Forced printings
1754 self.__forced_printings(info_dict, filename, incomplete=False)
1755
1756 # Do nothing else if in simulate mode
1757 if self.params.get('simulate', False):
1758 return
1759
1760 if filename is None:
1761 return
1762
1763 def ensure_dir_exists(path):
1764 try:
1765 dn = os.path.dirname(path)
1766 if dn and not os.path.exists(dn):
1767 os.makedirs(dn)
1768 return True
1769 except (OSError, IOError) as err:
1770 self.report_error('unable to create directory ' + error_to_compat_str(err))
1771 return False
1772
1773 if not ensure_dir_exists(sanitize_path(encodeFilename(filename))):
1774 return
1775
1776 if self.params.get('writedescription', False):
1777 descfn = replace_extension(filename, 'description', info_dict.get('ext'))
1778 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(descfn)):
1779 self.to_screen('[info] Video description is already present')
1780 elif info_dict.get('description') is None:
1781 self.report_warning('There\'s no description to write.')
1782 else:
1783 try:
1784 self.to_screen('[info] Writing video description to: ' + descfn)
1785 with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
1786 descfile.write(info_dict['description'])
1787 except (OSError, IOError):
1788 self.report_error('Cannot write description file ' + descfn)
1789 return
1790
1791 if self.params.get('writeannotations', False):
1792 annofn = replace_extension(filename, 'annotations.xml', info_dict.get('ext'))
1793 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(annofn)):
1794 self.to_screen('[info] Video annotations are already present')
1795 elif not info_dict.get('annotations'):
1796 self.report_warning('There are no annotations to write.')
1797 else:
1798 try:
1799 self.to_screen('[info] Writing video annotations to: ' + annofn)
1800 with io.open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
1801 annofile.write(info_dict['annotations'])
1802 except (KeyError, TypeError):
1803 self.report_warning('There are no annotations to write.')
1804 except (OSError, IOError):
1805 self.report_error('Cannot write annotations file: ' + annofn)
1806 return
1807
1808 subtitles_are_requested = any([self.params.get('writesubtitles', False),
1809 self.params.get('writeautomaticsub')])
1810
1811 if subtitles_are_requested and info_dict.get('requested_subtitles'):
1812 # subtitles download errors are already managed as troubles in relevant IE
1813 # that way it will silently go on when used with unsupporting IE
1814 subtitles = info_dict['requested_subtitles']
1815 ie = self.get_info_extractor(info_dict['extractor_key'])
1816 for sub_lang, sub_info in subtitles.items():
1817 sub_format = sub_info['ext']
1818 sub_filename = subtitles_filename(filename, sub_lang, sub_format, info_dict.get('ext'))
1819 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(sub_filename)):
1820 self.to_screen('[info] Video subtitle %s.%s is already present' % (sub_lang, sub_format))
1821 else:
1822 self.to_screen('[info] Writing video subtitles to: ' + sub_filename)
1823 if sub_info.get('data') is not None:
1824 try:
1825 # Use newline='' to prevent conversion of newline characters
1826 # See https://github.com/ytdl-org/youtube-dl/issues/10268
1827 with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8', newline='') as subfile:
1828 subfile.write(sub_info['data'])
1829 except (OSError, IOError):
1830 self.report_error('Cannot write subtitles file ' + sub_filename)
1831 return
1832 else:
1833 try:
1834 sub_data = ie._request_webpage(
1835 sub_info['url'], info_dict['id'], note=False).read()
1836 with io.open(encodeFilename(sub_filename), 'wb') as subfile:
1837 subfile.write(sub_data)
1838 except (ExtractorError, IOError, OSError, ValueError) as err:
1839 self.report_warning('Unable to download subtitle for "%s": %s' %
1840 (sub_lang, error_to_compat_str(err)))
1841 continue
1842
1843 if self.params.get('writeinfojson', False):
1844 infofn = replace_extension(filename, 'info.json', info_dict.get('ext'))
1845 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(infofn)):
1846 self.to_screen('[info] Video description metadata is already present')
1847 else:
1848 self.to_screen('[info] Writing video description metadata as JSON to: ' + infofn)
1849 try:
1850 write_json_file(self.filter_requested_info(info_dict), infofn)
1851 except (OSError, IOError):
1852 self.report_error('Cannot write metadata to JSON file ' + infofn)
1853 return
1854
1855 self._write_thumbnails(info_dict, filename)
1856
1857 if not self.params.get('skip_download', False):
1858 try:
1859 def dl(name, info):
1860 fd = get_suitable_downloader(info, self.params)(self, self.params)
1861 for ph in self._progress_hooks:
1862 fd.add_progress_hook(ph)
1863 if self.params.get('verbose'):
1864 self.to_stdout('[debug] Invoking downloader on %r' % info.get('url'))
1865 return fd.download(name, info)
1866
1867 if info_dict.get('requested_formats') is not None:
1868 downloaded = []
1869 success = True
1870 merger = FFmpegMergerPP(self)
1871 if not merger.available:
1872 postprocessors = []
1873 self.report_warning('You have requested multiple '
1874 'formats but ffmpeg or avconv are not installed.'
1875 ' The formats won\'t be merged.')
1876 else:
1877 postprocessors = [merger]
1878
1879 def compatible_formats(formats):
1880 video, audio = formats
1881 # Check extension
1882 video_ext, audio_ext = video.get('ext'), audio.get('ext')
1883 if video_ext and audio_ext:
1884 COMPATIBLE_EXTS = (
1885 ('mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'ismv', 'isma'),
1886 ('webm')
1887 )
1888 for exts in COMPATIBLE_EXTS:
1889 if video_ext in exts and audio_ext in exts:
1890 return True
1891 # TODO: Check acodec/vcodec
1892 return False
1893
1894 filename_real_ext = os.path.splitext(filename)[1][1:]
1895 filename_wo_ext = (
1896 os.path.splitext(filename)[0]
1897 if filename_real_ext == info_dict['ext']
1898 else filename)
1899 requested_formats = info_dict['requested_formats']
1900 if self.params.get('merge_output_format') is None and not compatible_formats(requested_formats):
1901 info_dict['ext'] = 'mkv'
1902 self.report_warning(
1903 'Requested formats are incompatible for merge and will be merged into mkv.')
1904 # Ensure filename always has a correct extension for successful merge
1905 filename = '%s.%s' % (filename_wo_ext, info_dict['ext'])
1906 if os.path.exists(encodeFilename(filename)):
1907 self.to_screen(
1908 '[download] %s has already been downloaded and '
1909 'merged' % filename)
1910 else:
1911 for f in requested_formats:
1912 new_info = dict(info_dict)
1913 new_info.update(f)
1914 fname = prepend_extension(
1915 self.prepare_filename(new_info),
1916 'f%s' % f['format_id'], new_info['ext'])
1917 if not ensure_dir_exists(fname):
1918 return
1919 downloaded.append(fname)
1920 partial_success = dl(fname, new_info)
1921 success = success and partial_success
1922 info_dict['__postprocessors'] = postprocessors
1923 info_dict['__files_to_merge'] = downloaded
1924 else:
1925 # Just a single file
1926 success = dl(filename, info_dict)
1927 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
1928 self.report_error('unable to download video data: %s' % error_to_compat_str(err))
1929 return
1930 except (OSError, IOError) as err:
1931 raise UnavailableVideoError(err)
1932 except (ContentTooShortError, ) as err:
1933 self.report_error('content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded))
1934 return
1935
1936 if success and filename != '-':
1937 # Fixup content
1938 fixup_policy = self.params.get('fixup')
1939 if fixup_policy is None:
1940 fixup_policy = 'detect_or_warn'
1941
1942 INSTALL_FFMPEG_MESSAGE = 'Install ffmpeg or avconv to fix this automatically.'
1943
1944 stretched_ratio = info_dict.get('stretched_ratio')
1945 if stretched_ratio is not None and stretched_ratio != 1:
1946 if fixup_policy == 'warn':
1947 self.report_warning('%s: Non-uniform pixel ratio (%s)' % (
1948 info_dict['id'], stretched_ratio))
1949 elif fixup_policy == 'detect_or_warn':
1950 stretched_pp = FFmpegFixupStretchedPP(self)
1951 if stretched_pp.available:
1952 info_dict.setdefault('__postprocessors', [])
1953 info_dict['__postprocessors'].append(stretched_pp)
1954 else:
1955 self.report_warning(
1956 '%s: Non-uniform pixel ratio (%s). %s'
1957 % (info_dict['id'], stretched_ratio, INSTALL_FFMPEG_MESSAGE))
1958 else:
1959 assert fixup_policy in ('ignore', 'never')
1960
1961 if (info_dict.get('requested_formats') is None
1962 and info_dict.get('container') == 'm4a_dash'):
1963 if fixup_policy == 'warn':
1964 self.report_warning(
1965 '%s: writing DASH m4a. '
1966 'Only some players support this container.'
1967 % info_dict['id'])
1968 elif fixup_policy == 'detect_or_warn':
1969 fixup_pp = FFmpegFixupM4aPP(self)
1970 if fixup_pp.available:
1971 info_dict.setdefault('__postprocessors', [])
1972 info_dict['__postprocessors'].append(fixup_pp)
1973 else:
1974 self.report_warning(
1975 '%s: writing DASH m4a. '
1976 'Only some players support this container. %s'
1977 % (info_dict['id'], INSTALL_FFMPEG_MESSAGE))
1978 else:
1979 assert fixup_policy in ('ignore', 'never')
1980
1981 if (info_dict.get('protocol') == 'm3u8_native'
1982 or info_dict.get('protocol') == 'm3u8'
1983 and self.params.get('hls_prefer_native')):
1984 if fixup_policy == 'warn':
1985 self.report_warning('%s: malformed AAC bitstream detected.' % (
1986 info_dict['id']))
1987 elif fixup_policy == 'detect_or_warn':
1988 fixup_pp = FFmpegFixupM3u8PP(self)
1989 if fixup_pp.available:
1990 info_dict.setdefault('__postprocessors', [])
1991 info_dict['__postprocessors'].append(fixup_pp)
1992 else:
1993 self.report_warning(
1994 '%s: malformed AAC bitstream detected. %s'
1995 % (info_dict['id'], INSTALL_FFMPEG_MESSAGE))
1996 else:
1997 assert fixup_policy in ('ignore', 'never')
1998
1999 try:
2000 self.post_process(filename, info_dict)
2001 except (PostProcessingError) as err:
2002 self.report_error('postprocessing: %s' % str(err))
2003 return
2004 self.record_download_archive(info_dict)
2005
2006 def download(self, url_list):
2007 """Download a given list of URLs."""
2008 outtmpl = self.params.get('outtmpl', DEFAULT_OUTTMPL)
2009 if (len(url_list) > 1
2010 and outtmpl != '-'
2011 and '%' not in outtmpl
2012 and self.params.get('max_downloads') != 1):
2013 raise SameFileError(outtmpl)
2014
2015 for url in url_list:
2016 try:
2017 # It also downloads the videos
2018 res = self.extract_info(
2019 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
2020 except UnavailableVideoError:
2021 self.report_error('unable to download video')
2022 except MaxDownloadsReached:
2023 self.to_screen('[info] Maximum number of downloaded files reached.')
2024 raise
2025 else:
2026 if self.params.get('dump_single_json', False):
2027 self.to_stdout(json.dumps(res))
2028
2029 return self._download_retcode
2030
2031 def download_with_info_file(self, info_filename):
2032 with contextlib.closing(fileinput.FileInput(
2033 [info_filename], mode='r',
2034 openhook=fileinput.hook_encoded('utf-8'))) as f:
2035 # FileInput doesn't have a read method, we can't call json.load
2036 info = self.filter_requested_info(json.loads('\n'.join(f)))
2037 try:
2038 self.process_ie_result(info, download=True)
2039 except DownloadError:
2040 webpage_url = info.get('webpage_url')
2041 if webpage_url is not None:
2042 self.report_warning('The info failed to download, trying with "%s"' % webpage_url)
2043 return self.download([webpage_url])
2044 else:
2045 raise
2046 return self._download_retcode
2047
2048 @staticmethod
2049 def filter_requested_info(info_dict):
2050 return dict(
2051 (k, v) for k, v in info_dict.items()
2052 if k not in ['requested_formats', 'requested_subtitles'])
2053
2054 def post_process(self, filename, ie_info):
2055 """Run all the postprocessors on the given file."""
2056 info = dict(ie_info)
2057 info['filepath'] = filename
2058 pps_chain = []
2059 if ie_info.get('__postprocessors') is not None:
2060 pps_chain.extend(ie_info['__postprocessors'])
2061 pps_chain.extend(self._pps)
2062 for pp in pps_chain:
2063 files_to_delete = []
2064 try:
2065 files_to_delete, info = pp.run(info)
2066 except PostProcessingError as e:
2067 self.report_error(e.msg)
2068 if files_to_delete and not self.params.get('keepvideo', False):
2069 for old_filename in files_to_delete:
2070 self.to_screen('Deleting original file %s (pass -k to keep)' % old_filename)
2071 try:
2072 os.remove(encodeFilename(old_filename))
2073 except (IOError, OSError):
2074 self.report_warning('Unable to remove downloaded original file')
2075
2076 def _make_archive_id(self, info_dict):
2077 video_id = info_dict.get('id')
2078 if not video_id:
2079 return
2080 # Future-proof against any change in case
2081 # and backwards compatibility with prior versions
2082 extractor = info_dict.get('extractor_key') or info_dict.get('ie_key') # key in a playlist
2083 if extractor is None:
2084 url = str_or_none(info_dict.get('url'))
2085 if not url:
2086 return
2087 # Try to find matching extractor for the URL and take its ie_key
2088 for ie in self._ies:
2089 if ie.suitable(url):
2090 extractor = ie.ie_key()
2091 break
2092 else:
2093 return
2094 return extractor.lower() + ' ' + video_id
2095
2096 def in_download_archive(self, info_dict):
2097 fn = self.params.get('download_archive')
2098 if fn is None:
2099 return False
2100
2101 vid_id = self._make_archive_id(info_dict)
2102 if not vid_id:
2103 return False # Incomplete video information
2104
2105 try:
2106 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
2107 for line in archive_file:
2108 if line.strip() == vid_id:
2109 return True
2110 except IOError as ioe:
2111 if ioe.errno != errno.ENOENT:
2112 raise
2113 return False
2114
2115 def record_download_archive(self, info_dict):
2116 fn = self.params.get('download_archive')
2117 if fn is None:
2118 return
2119 vid_id = self._make_archive_id(info_dict)
2120 assert vid_id
2121 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
2122 archive_file.write(vid_id + '\n')
2123
2124 @staticmethod
2125 def format_resolution(format, default='unknown'):
2126 if format.get('vcodec') == 'none':
2127 return 'audio only'
2128 if format.get('resolution') is not None:
2129 return format['resolution']
2130 if format.get('height') is not None:
2131 if format.get('width') is not None:
2132 res = '%sx%s' % (format['width'], format['height'])
2133 else:
2134 res = '%sp' % format['height']
2135 elif format.get('width') is not None:
2136 res = '%dx?' % format['width']
2137 else:
2138 res = default
2139 return res
2140
2141 def _format_note(self, fdict):
2142 res = ''
2143 if fdict.get('ext') in ['f4f', 'f4m']:
2144 res += '(unsupported) '
2145 if fdict.get('language'):
2146 if res:
2147 res += ' '
2148 res += '[%s] ' % fdict['language']
2149 if fdict.get('format_note') is not None:
2150 res += fdict['format_note'] + ' '
2151 if fdict.get('tbr') is not None:
2152 res += '%4dk ' % fdict['tbr']
2153 if fdict.get('container') is not None:
2154 if res:
2155 res += ', '
2156 res += '%s container' % fdict['container']
2157 if (fdict.get('vcodec') is not None
2158 and fdict.get('vcodec') != 'none'):
2159 if res:
2160 res += ', '
2161 res += fdict['vcodec']
2162 if fdict.get('vbr') is not None:
2163 res += '@'
2164 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
2165 res += 'video@'
2166 if fdict.get('vbr') is not None:
2167 res += '%4dk' % fdict['vbr']
2168 if fdict.get('fps') is not None:
2169 if res:
2170 res += ', '
2171 res += '%sfps' % fdict['fps']
2172 if fdict.get('acodec') is not None:
2173 if res:
2174 res += ', '
2175 if fdict['acodec'] == 'none':
2176 res += 'video only'
2177 else:
2178 res += '%-5s' % fdict['acodec']
2179 elif fdict.get('abr') is not None:
2180 if res:
2181 res += ', '
2182 res += 'audio'
2183 if fdict.get('abr') is not None:
2184 res += '@%3dk' % fdict['abr']
2185 if fdict.get('asr') is not None:
2186 res += ' (%5dHz)' % fdict['asr']
2187 if fdict.get('filesize') is not None:
2188 if res:
2189 res += ', '
2190 res += format_bytes(fdict['filesize'])
2191 elif fdict.get('filesize_approx') is not None:
2192 if res:
2193 res += ', '
2194 res += '~' + format_bytes(fdict['filesize_approx'])
2195 return res
2196
2197 def list_formats(self, info_dict):
2198 formats = info_dict.get('formats', [info_dict])
2199 table = [
2200 [f['format_id'], f['ext'], self.format_resolution(f), self._format_note(f)]
2201 for f in formats
2202 if f.get('preference') is None or f['preference'] >= -1000]
2203 if len(formats) > 1:
2204 table[-1][-1] += (' ' if table[-1][-1] else '') + '(best)'
2205
2206 header_line = ['format code', 'extension', 'resolution', 'note']
2207 self.to_screen(
2208 '[info] Available formats for %s:\n%s' %
2209 (info_dict['id'], render_table(header_line, table)))
2210
2211 def list_thumbnails(self, info_dict):
2212 thumbnails = info_dict.get('thumbnails')
2213 if not thumbnails:
2214 self.to_screen('[info] No thumbnails present for %s' % info_dict['id'])
2215 return
2216
2217 self.to_screen(
2218 '[info] Thumbnails for %s:' % info_dict['id'])
2219 self.to_screen(render_table(
2220 ['ID', 'width', 'height', 'URL'],
2221 [[t['id'], t.get('width', 'unknown'), t.get('height', 'unknown'), t['url']] for t in thumbnails]))
2222
2223 def list_subtitles(self, video_id, subtitles, name='subtitles'):
2224 if not subtitles:
2225 self.to_screen('%s has no %s' % (video_id, name))
2226 return
2227 self.to_screen(
2228 'Available %s for %s:' % (name, video_id))
2229 self.to_screen(render_table(
2230 ['Language', 'formats'],
2231 [[lang, ', '.join(f['ext'] for f in reversed(formats))]
2232 for lang, formats in subtitles.items()]))
2233
2234 def urlopen(self, req):
2235 """ Start an HTTP download """
2236 if isinstance(req, compat_basestring):
2237 req = sanitized_Request(req)
2238 return self._opener.open(req, timeout=self._socket_timeout)
2239
2240 def print_debug_header(self):
2241 if not self.params.get('verbose'):
2242 return
2243
2244 if type('') is not compat_str:
2245 # Python 2.6 on SLES11 SP1 (https://github.com/ytdl-org/youtube-dl/issues/3326)
2246 self.report_warning(
2247 'Your Python is broken! Update to a newer and supported version')
2248
2249 stdout_encoding = getattr(
2250 sys.stdout, 'encoding', 'missing (%s)' % type(sys.stdout).__name__)
2251 encoding_str = (
2252 '[debug] Encodings: locale %s, fs %s, out %s, pref %s\n' % (
2253 locale.getpreferredencoding(),
2254 sys.getfilesystemencoding(),
2255 stdout_encoding,
2256 self.get_encoding()))
2257 write_string(encoding_str, encoding=None)
2258
2259 self._write_string('[debug] youtube-dl version ' + __version__ + '\n')
2260 if _LAZY_LOADER:
2261 self._write_string('[debug] Lazy loading extractors enabled' + '\n')
2262 try:
2263 sp = subprocess.Popen(
2264 ['git', 'rev-parse', '--short', 'HEAD'],
2265 stdout=subprocess.PIPE, stderr=subprocess.PIPE,
2266 cwd=os.path.dirname(os.path.abspath(__file__)))
2267 out, err = sp.communicate()
2268 out = out.decode().strip()
2269 if re.match('[0-9a-f]+', out):
2270 self._write_string('[debug] Git HEAD: ' + out + '\n')
2271 except Exception:
2272 try:
2273 sys.exc_clear()
2274 except Exception:
2275 pass
2276
2277 def python_implementation():
2278 impl_name = platform.python_implementation()
2279 if impl_name == 'PyPy' and hasattr(sys, 'pypy_version_info'):
2280 return impl_name + ' version %d.%d.%d' % sys.pypy_version_info[:3]
2281 return impl_name
2282
2283 self._write_string('[debug] Python version %s (%s) - %s\n' % (
2284 platform.python_version(), python_implementation(),
2285 platform_name()))
2286
2287 exe_versions = FFmpegPostProcessor.get_versions(self)
2288 exe_versions['rtmpdump'] = rtmpdump_version()
2289 exe_versions['phantomjs'] = PhantomJSwrapper._version()
2290 exe_str = ', '.join(
2291 '%s %s' % (exe, v)
2292 for exe, v in sorted(exe_versions.items())
2293 if v
2294 )
2295 if not exe_str:
2296 exe_str = 'none'
2297 self._write_string('[debug] exe versions: %s\n' % exe_str)
2298
2299 proxy_map = {}
2300 for handler in self._opener.handlers:
2301 if hasattr(handler, 'proxies'):
2302 proxy_map.update(handler.proxies)
2303 self._write_string('[debug] Proxy map: ' + compat_str(proxy_map) + '\n')
2304
2305 if self.params.get('call_home', False):
2306 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode('utf-8')
2307 self._write_string('[debug] Public IP address: %s\n' % ipaddr)
2308 latest_version = self.urlopen(
2309 'https://yt-dl.org/latest/version').read().decode('utf-8')
2310 if version_tuple(latest_version) > version_tuple(__version__):
2311 self.report_warning(
2312 'You are using an outdated version (newest version: %s)! '
2313 'See https://yt-dl.org/update if you need help updating.' %
2314 latest_version)
2315
2316 def _setup_opener(self):
2317 timeout_val = self.params.get('socket_timeout')
2318 self._socket_timeout = 600 if timeout_val is None else float(timeout_val)
2319
2320 opts_cookiefile = self.params.get('cookiefile')
2321 opts_proxy = self.params.get('proxy')
2322
2323 if opts_cookiefile is None:
2324 self.cookiejar = compat_cookiejar.CookieJar()
2325 else:
2326 opts_cookiefile = expand_path(opts_cookiefile)
2327 self.cookiejar = YoutubeDLCookieJar(opts_cookiefile)
2328 if os.access(opts_cookiefile, os.R_OK):
2329 self.cookiejar.load(ignore_discard=True, ignore_expires=True)
2330
2331 cookie_processor = YoutubeDLCookieProcessor(self.cookiejar)
2332 if opts_proxy is not None:
2333 if opts_proxy == '':
2334 proxies = {}
2335 else:
2336 proxies = {'http': opts_proxy, 'https': opts_proxy}
2337 else:
2338 proxies = compat_urllib_request.getproxies()
2339 # Set HTTPS proxy to HTTP one if given (https://github.com/ytdl-org/youtube-dl/issues/805)
2340 if 'http' in proxies and 'https' not in proxies:
2341 proxies['https'] = proxies['http']
2342 proxy_handler = PerRequestProxyHandler(proxies)
2343
2344 debuglevel = 1 if self.params.get('debug_printtraffic') else 0
2345 https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
2346 ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
2347 redirect_handler = YoutubeDLRedirectHandler()
2348 data_handler = compat_urllib_request_DataHandler()
2349
2350 # When passing our own FileHandler instance, build_opener won't add the
2351 # default FileHandler and allows us to disable the file protocol, which
2352 # can be used for malicious purposes (see
2353 # https://github.com/ytdl-org/youtube-dl/issues/8227)
2354 file_handler = compat_urllib_request.FileHandler()
2355
2356 def file_open(*args, **kwargs):
2357 raise compat_urllib_error.URLError('file:// scheme is explicitly disabled in youtube-dl for security reasons')
2358 file_handler.file_open = file_open
2359
2360 opener = compat_urllib_request.build_opener(
2361 proxy_handler, https_handler, cookie_processor, ydlh, redirect_handler, data_handler, file_handler)
2362
2363 # Delete the default user-agent header, which would otherwise apply in
2364 # cases where our custom HTTP handler doesn't come into play
2365 # (See https://github.com/ytdl-org/youtube-dl/issues/1309 for details)
2366 opener.addheaders = []
2367 self._opener = opener
2368
2369 def encode(self, s):
2370 if isinstance(s, bytes):
2371 return s # Already encoded
2372
2373 try:
2374 return s.encode(self.get_encoding())
2375 except UnicodeEncodeError as err:
2376 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
2377 raise
2378
2379 def get_encoding(self):
2380 encoding = self.params.get('encoding')
2381 if encoding is None:
2382 encoding = preferredencoding()
2383 return encoding
2384
2385 def _write_thumbnails(self, info_dict, filename):
2386 if self.params.get('writethumbnail', False):
2387 thumbnails = info_dict.get('thumbnails')
2388 if thumbnails:
2389 thumbnails = [thumbnails[-1]]
2390 elif self.params.get('write_all_thumbnails', False):
2391 thumbnails = info_dict.get('thumbnails')
2392 else:
2393 return
2394
2395 if not thumbnails:
2396 # No thumbnails present, so return immediately
2397 return
2398
2399 for t in thumbnails:
2400 thumb_ext = determine_ext(t['url'], 'jpg')
2401 suffix = '_%s' % t['id'] if len(thumbnails) > 1 else ''
2402 thumb_display_id = '%s ' % t['id'] if len(thumbnails) > 1 else ''
2403 t['filename'] = thumb_filename = os.path.splitext(filename)[0] + suffix + '.' + thumb_ext
2404
2405 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(thumb_filename)):
2406 self.to_screen('[%s] %s: Thumbnail %sis already present' %
2407 (info_dict['extractor'], info_dict['id'], thumb_display_id))
2408 else:
2409 self.to_screen('[%s] %s: Downloading thumbnail %s...' %
2410 (info_dict['extractor'], info_dict['id'], thumb_display_id))
2411 try:
2412 uf = self.urlopen(t['url'])
2413 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
2414 shutil.copyfileobj(uf, thumbf)
2415 self.to_screen('[%s] %s: Writing thumbnail %sto: %s' %
2416 (info_dict['extractor'], info_dict['id'], thumb_display_id, thumb_filename))
2417 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
2418 self.report_warning('Unable to download thumbnail "%s": %s' %
2419 (t['url'], error_to_compat_str(err)))