]> jfr.im git - yt-dlp.git/blob - youtube_dlc/YoutubeDL.py
#29 New option `-P`/`--paths` to give different paths for different types of files
[yt-dlp.git] / youtube_dlc / 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 DOT_DESKTOP_LINK_TEMPLATE,
55 DOT_URL_LINK_TEMPLATE,
56 DOT_WEBLOC_LINK_TEMPLATE,
57 DownloadError,
58 encode_compat_str,
59 encodeFilename,
60 error_to_compat_str,
61 ExistingVideoReached,
62 expand_path,
63 ExtractorError,
64 format_bytes,
65 format_field,
66 formatSeconds,
67 GeoRestrictedError,
68 int_or_none,
69 iri_to_uri,
70 ISO3166Utils,
71 locked_file,
72 make_dir,
73 make_HTTPS_handler,
74 MaxDownloadsReached,
75 orderedSet,
76 PagedList,
77 parse_filesize,
78 PerRequestProxyHandler,
79 platform_name,
80 PostProcessingError,
81 preferredencoding,
82 prepend_extension,
83 register_socks_protocols,
84 render_table,
85 replace_extension,
86 RejectedVideoReached,
87 SameFileError,
88 sanitize_filename,
89 sanitize_path,
90 sanitize_url,
91 sanitized_Request,
92 std_headers,
93 str_or_none,
94 subtitles_filename,
95 to_high_limit_path,
96 UnavailableVideoError,
97 url_basename,
98 version_tuple,
99 write_json_file,
100 write_string,
101 YoutubeDLCookieJar,
102 YoutubeDLCookieProcessor,
103 YoutubeDLHandler,
104 YoutubeDLRedirectHandler,
105 process_communicate_or_kill,
106 )
107 from .cache import Cache
108 from .extractor import get_info_extractor, gen_extractor_classes, _LAZY_LOADER
109 from .extractor.openload import PhantomJSwrapper
110 from .downloader import get_suitable_downloader
111 from .downloader.rtmp import rtmpdump_version
112 from .postprocessor import (
113 FFmpegFixupM3u8PP,
114 FFmpegFixupM4aPP,
115 FFmpegFixupStretchedPP,
116 FFmpegMergerPP,
117 FFmpegPostProcessor,
118 # FFmpegSubtitlesConvertorPP,
119 get_postprocessor,
120 MoveFilesAfterDownloadPP,
121 )
122 from .version import __version__
123
124 if compat_os_name == 'nt':
125 import ctypes
126
127
128 class YoutubeDL(object):
129 """YoutubeDL class.
130
131 YoutubeDL objects are the ones responsible of downloading the
132 actual video file and writing it to disk if the user has requested
133 it, among some other tasks. In most cases there should be one per
134 program. As, given a video URL, the downloader doesn't know how to
135 extract all the needed information, task that InfoExtractors do, it
136 has to pass the URL to one of them.
137
138 For this, YoutubeDL objects have a method that allows
139 InfoExtractors to be registered in a given order. When it is passed
140 a URL, the YoutubeDL object handles it to the first InfoExtractor it
141 finds that reports being able to handle it. The InfoExtractor extracts
142 all the information about the video or videos the URL refers to, and
143 YoutubeDL process the extracted information, possibly using a File
144 Downloader to download the video.
145
146 YoutubeDL objects accept a lot of parameters. In order not to saturate
147 the object constructor with arguments, it receives a dictionary of
148 options instead. These options are available through the params
149 attribute for the InfoExtractors to use. The YoutubeDL also
150 registers itself as the downloader in charge for the InfoExtractors
151 that are added to it, so this is a "mutual registration".
152
153 Available options:
154
155 username: Username for authentication purposes.
156 password: Password for authentication purposes.
157 videopassword: Password for accessing a video.
158 ap_mso: Adobe Pass multiple-system operator identifier.
159 ap_username: Multiple-system operator account username.
160 ap_password: Multiple-system operator account password.
161 usenetrc: Use netrc for authentication instead.
162 verbose: Print additional info to stdout.
163 quiet: Do not print messages to stdout.
164 no_warnings: Do not print out anything for warnings.
165 forceurl: Force printing final URL.
166 forcetitle: Force printing title.
167 forceid: Force printing ID.
168 forcethumbnail: Force printing thumbnail URL.
169 forcedescription: Force printing description.
170 forcefilename: Force printing final filename.
171 forceduration: Force printing duration.
172 forcejson: Force printing info_dict as JSON.
173 dump_single_json: Force printing the info_dict of the whole playlist
174 (or video) as a single JSON line.
175 force_write_download_archive: Force writing download archive regardless of
176 'skip_download' or 'simulate'.
177 simulate: Do not download the video files.
178 format: Video format code. see "FORMAT SELECTION" for more details.
179 format_sort: How to sort the video formats. see "Sorting Formats" for more details.
180 format_sort_force: Force the given format_sort. see "Sorting Formats" for more details.
181 allow_multiple_video_streams: Allow multiple video streams to be merged into a single file
182 allow_multiple_audio_streams: Allow multiple audio streams to be merged into a single file
183 outtmpl: Template for output names.
184 restrictfilenames: Do not allow "&" and spaces in file names.
185 trim_file_name: Limit length of filename (extension excluded).
186 ignoreerrors: Do not stop on download errors. (Default True when running youtube-dlc, but False when directly accessing YoutubeDL class)
187 force_generic_extractor: Force downloader to use the generic extractor
188 overwrites: Overwrite all video and metadata files if True,
189 overwrite only non-video files if None
190 and don't overwrite any file if False
191 playliststart: Playlist item to start at.
192 playlistend: Playlist item to end at.
193 playlist_items: Specific indices of playlist to download.
194 playlistreverse: Download playlist items in reverse order.
195 playlistrandom: Download playlist items in random order.
196 matchtitle: Download only matching titles.
197 rejecttitle: Reject downloads for matching titles.
198 logger: Log messages to a logging.Logger instance.
199 logtostderr: Log messages to stderr instead of stdout.
200 writedescription: Write the video description to a .description file
201 writeinfojson: Write the video description to a .info.json file
202 writeannotations: Write the video annotations to a .annotations.xml file
203 writethumbnail: Write the thumbnail image to a file
204 write_all_thumbnails: Write all thumbnail formats to files
205 writelink: Write an internet shortcut file, depending on the
206 current platform (.url/.webloc/.desktop)
207 writeurllink: Write a Windows internet shortcut file (.url)
208 writewebloclink: Write a macOS internet shortcut file (.webloc)
209 writedesktoplink: Write a Linux internet shortcut file (.desktop)
210 writesubtitles: Write the video subtitles to a file
211 writeautomaticsub: Write the automatically generated subtitles to a file
212 allsubtitles: Downloads all the subtitles of the video
213 (requires writesubtitles or writeautomaticsub)
214 listsubtitles: Lists all available subtitles for the video
215 subtitlesformat: The format code for subtitles
216 subtitleslangs: List of languages of the subtitles to download
217 keepvideo: Keep the video file after post-processing
218 daterange: A DateRange object, download only if the upload_date is in the range.
219 skip_download: Skip the actual download of the video file
220 cachedir: Location of the cache files in the filesystem.
221 False to disable filesystem cache.
222 noplaylist: Download single video instead of a playlist if in doubt.
223 age_limit: An integer representing the user's age in years.
224 Unsuitable videos for the given age are skipped.
225 min_views: An integer representing the minimum view count the video
226 must have in order to not be skipped.
227 Videos without view count information are always
228 downloaded. None for no limit.
229 max_views: An integer representing the maximum view count.
230 Videos that are more popular than that are not
231 downloaded.
232 Videos without view count information are always
233 downloaded. None for no limit.
234 download_archive: File name of a file where all downloads are recorded.
235 Videos already present in the file are not downloaded
236 again.
237 break_on_existing: Stop the download process after attempting to download a
238 file that is in the archive.
239 break_on_reject: Stop the download process when encountering a video that
240 has been filtered out.
241 cookiefile: File name where cookies should be read from and dumped to
242 nocheckcertificate:Do not verify SSL certificates
243 prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
244 At the moment, this is only supported by YouTube.
245 proxy: URL of the proxy server to use
246 geo_verification_proxy: URL of the proxy to use for IP address verification
247 on geo-restricted sites.
248 socket_timeout: Time to wait for unresponsive hosts, in seconds
249 bidi_workaround: Work around buggy terminals without bidirectional text
250 support, using fridibi
251 debug_printtraffic:Print out sent and received HTTP traffic
252 include_ads: Download ads as well
253 default_search: Prepend this string if an input url is not valid.
254 'auto' for elaborate guessing
255 encoding: Use this encoding instead of the system-specified.
256 extract_flat: Do not resolve URLs, return the immediate result.
257 Pass in 'in_playlist' to only show this behavior for
258 playlist items.
259 postprocessors: A list of dictionaries, each with an entry
260 * key: The name of the postprocessor. See
261 youtube_dlc/postprocessor/__init__.py for a list.
262 * _after_move: Optional. If True, run this post_processor
263 after 'MoveFilesAfterDownload'
264 as well as any further keyword arguments for the
265 postprocessor.
266 post_hooks: A list of functions that get called as the final step
267 for each video file, after all postprocessors have been
268 called. The filename will be passed as the only argument.
269 progress_hooks: A list of functions that get called on download
270 progress, with a dictionary with the entries
271 * status: One of "downloading", "error", or "finished".
272 Check this first and ignore unknown values.
273
274 If status is one of "downloading", or "finished", the
275 following properties may also be present:
276 * filename: The final filename (always present)
277 * tmpfilename: The filename we're currently writing to
278 * downloaded_bytes: Bytes on disk
279 * total_bytes: Size of the whole file, None if unknown
280 * total_bytes_estimate: Guess of the eventual file size,
281 None if unavailable.
282 * elapsed: The number of seconds since download started.
283 * eta: The estimated time in seconds, None if unknown
284 * speed: The download speed in bytes/second, None if
285 unknown
286 * fragment_index: The counter of the currently
287 downloaded video fragment.
288 * fragment_count: The number of fragments (= individual
289 files that will be merged)
290
291 Progress hooks are guaranteed to be called at least once
292 (with status "finished") if the download is successful.
293 merge_output_format: Extension to use when merging formats.
294 fixup: Automatically correct known faults of the file.
295 One of:
296 - "never": do nothing
297 - "warn": only emit a warning
298 - "detect_or_warn": check whether we can do anything
299 about it, warn otherwise (default)
300 source_address: Client-side IP address to bind to.
301 call_home: Boolean, true iff we are allowed to contact the
302 youtube-dlc servers for debugging.
303 sleep_interval: Number of seconds to sleep before each download when
304 used alone or a lower bound of a range for randomized
305 sleep before each download (minimum possible number
306 of seconds to sleep) when used along with
307 max_sleep_interval.
308 max_sleep_interval:Upper bound of a range for randomized sleep before each
309 download (maximum possible number of seconds to sleep).
310 Must only be used along with sleep_interval.
311 Actual sleep time will be a random float from range
312 [sleep_interval; max_sleep_interval].
313 listformats: Print an overview of available video formats and exit.
314 list_thumbnails: Print a table of all thumbnails and exit.
315 match_filter: A function that gets called with the info_dict of
316 every video.
317 If it returns a message, the video is ignored.
318 If it returns None, the video is downloaded.
319 match_filter_func in utils.py is one example for this.
320 no_color: Do not emit color codes in output.
321 geo_bypass: Bypass geographic restriction via faking X-Forwarded-For
322 HTTP header
323 geo_bypass_country:
324 Two-letter ISO 3166-2 country code that will be used for
325 explicit geographic restriction bypassing via faking
326 X-Forwarded-For HTTP header
327 geo_bypass_ip_block:
328 IP range in CIDR notation that will be used similarly to
329 geo_bypass_country
330
331 The following options determine which downloader is picked:
332 external_downloader: Executable of the external downloader to call.
333 None or unset for standard (built-in) downloader.
334 hls_prefer_native: Use the native HLS downloader instead of ffmpeg/avconv
335 if True, otherwise use ffmpeg/avconv if False, otherwise
336 use downloader suggested by extractor if None.
337
338 The following parameters are not used by YoutubeDL itself, they are used by
339 the downloader (see youtube_dlc/downloader/common.py):
340 nopart, updatetime, buffersize, ratelimit, min_filesize, max_filesize, test,
341 noresizebuffer, retries, continuedl, noprogress, consoletitle,
342 xattr_set_filesize, external_downloader_args, hls_use_mpegts,
343 http_chunk_size.
344
345 The following options are used by the post processors:
346 prefer_ffmpeg: If False, use avconv instead of ffmpeg if both are available,
347 otherwise prefer ffmpeg.
348 ffmpeg_location: Location of the ffmpeg/avconv binary; either the path
349 to the binary or its containing directory.
350 postprocessor_args: A dictionary of postprocessor/executable keys (in lower case)
351 and a list of additional command-line arguments for the
352 postprocessor/executable. The dict can also have "PP+EXE" keys
353 which are used when the given exe is used by the given PP.
354 Use 'default' as the name for arguments to passed to all PP
355 The following options are used by the Youtube extractor:
356 youtube_include_dash_manifest: If True (default), DASH manifests and related
357 data will be downloaded and processed by extractor.
358 You can reduce network I/O by disabling it if you don't
359 care about DASH.
360 """
361
362 _NUMERIC_FIELDS = set((
363 'width', 'height', 'tbr', 'abr', 'asr', 'vbr', 'fps', 'filesize', 'filesize_approx',
364 'timestamp', 'upload_year', 'upload_month', 'upload_day',
365 'duration', 'view_count', 'like_count', 'dislike_count', 'repost_count',
366 'average_rating', 'comment_count', 'age_limit',
367 'start_time', 'end_time',
368 'chapter_number', 'season_number', 'episode_number',
369 'track_number', 'disc_number', 'release_year',
370 'playlist_index',
371 ))
372
373 params = None
374 _ies = []
375 _pps = []
376 _pps_end = []
377 __prepare_filename_warned = False
378 _download_retcode = None
379 _num_downloads = None
380 _playlist_level = 0
381 _playlist_urls = set()
382 _screen_file = None
383
384 def __init__(self, params=None, auto_init=True):
385 """Create a FileDownloader object with the given options."""
386 if params is None:
387 params = {}
388 self._ies = []
389 self._ies_instances = {}
390 self._pps = []
391 self._pps_end = []
392 self.__prepare_filename_warned = False
393 self._post_hooks = []
394 self._progress_hooks = []
395 self._download_retcode = 0
396 self._num_downloads = 0
397 self._screen_file = [sys.stdout, sys.stderr][params.get('logtostderr', False)]
398 self._err_file = sys.stderr
399 self.params = {
400 # Default parameters
401 'nocheckcertificate': False,
402 }
403 self.params.update(params)
404 self.cache = Cache(self)
405 self.archive = set()
406
407 """Preload the archive, if any is specified"""
408 def preload_download_archive(self):
409 fn = self.params.get('download_archive')
410 if fn is None:
411 return False
412 try:
413 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
414 for line in archive_file:
415 self.archive.add(line.strip())
416 except IOError as ioe:
417 if ioe.errno != errno.ENOENT:
418 raise
419 return False
420 return True
421
422 def check_deprecated(param, option, suggestion):
423 if self.params.get(param) is not None:
424 self.report_warning(
425 '%s is deprecated. Use %s instead.' % (option, suggestion))
426 return True
427 return False
428
429 if self.params.get('verbose'):
430 self.to_stdout('[debug] Loading archive file %r' % self.params.get('download_archive'))
431
432 preload_download_archive(self)
433
434 if check_deprecated('cn_verification_proxy', '--cn-verification-proxy', '--geo-verification-proxy'):
435 if self.params.get('geo_verification_proxy') is None:
436 self.params['geo_verification_proxy'] = self.params['cn_verification_proxy']
437
438 check_deprecated('autonumber_size', '--autonumber-size', 'output template with %(autonumber)0Nd, where N in the number of digits')
439 check_deprecated('autonumber', '--auto-number', '-o "%(autonumber)s-%(title)s.%(ext)s"')
440 check_deprecated('usetitle', '--title', '-o "%(title)s-%(id)s.%(ext)s"')
441
442 if params.get('bidi_workaround', False):
443 try:
444 import pty
445 master, slave = pty.openpty()
446 width = compat_get_terminal_size().columns
447 if width is None:
448 width_args = []
449 else:
450 width_args = ['-w', str(width)]
451 sp_kwargs = dict(
452 stdin=subprocess.PIPE,
453 stdout=slave,
454 stderr=self._err_file)
455 try:
456 self._output_process = subprocess.Popen(
457 ['bidiv'] + width_args, **sp_kwargs
458 )
459 except OSError:
460 self._output_process = subprocess.Popen(
461 ['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
462 self._output_channel = os.fdopen(master, 'rb')
463 except OSError as ose:
464 if ose.errno == errno.ENOENT:
465 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 $PATH.')
466 else:
467 raise
468
469 if (sys.platform != 'win32'
470 and sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968']
471 and not params.get('restrictfilenames', False)):
472 # Unicode filesystem API will throw errors (#1474, #13027)
473 self.report_warning(
474 'Assuming --restrict-filenames since file system encoding '
475 'cannot encode all characters. '
476 'Set the LC_ALL environment variable to fix this.')
477 self.params['restrictfilenames'] = True
478
479 if isinstance(params.get('outtmpl'), bytes):
480 self.report_warning(
481 'Parameter outtmpl is bytes, but should be a unicode string. '
482 'Put from __future__ import unicode_literals at the top of your code file or consider switching to Python 3.x.')
483
484 self._setup_opener()
485
486 if auto_init:
487 self.print_debug_header()
488 self.add_default_info_extractors()
489
490 for pp_def_raw in self.params.get('postprocessors', []):
491 pp_class = get_postprocessor(pp_def_raw['key'])
492 pp_def = dict(pp_def_raw)
493 del pp_def['key']
494 after_move = pp_def.get('_after_move', False)
495 if '_after_move' in pp_def:
496 del pp_def['_after_move']
497 pp = pp_class(self, **compat_kwargs(pp_def))
498 self.add_post_processor(pp, after_move=after_move)
499
500 for ph in self.params.get('post_hooks', []):
501 self.add_post_hook(ph)
502
503 for ph in self.params.get('progress_hooks', []):
504 self.add_progress_hook(ph)
505
506 register_socks_protocols()
507
508 def warn_if_short_id(self, argv):
509 # short YouTube ID starting with dash?
510 idxs = [
511 i for i, a in enumerate(argv)
512 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
513 if idxs:
514 correct_argv = (
515 ['youtube-dlc']
516 + [a for i, a in enumerate(argv) if i not in idxs]
517 + ['--'] + [argv[i] for i in idxs]
518 )
519 self.report_warning(
520 'Long argument string detected. '
521 'Use -- to separate parameters and URLs, like this:\n%s\n' %
522 args_to_str(correct_argv))
523
524 def add_info_extractor(self, ie):
525 """Add an InfoExtractor object to the end of the list."""
526 self._ies.append(ie)
527 if not isinstance(ie, type):
528 self._ies_instances[ie.ie_key()] = ie
529 ie.set_downloader(self)
530
531 def get_info_extractor(self, ie_key):
532 """
533 Get an instance of an IE with name ie_key, it will try to get one from
534 the _ies list, if there's no instance it will create a new one and add
535 it to the extractor list.
536 """
537 ie = self._ies_instances.get(ie_key)
538 if ie is None:
539 ie = get_info_extractor(ie_key)()
540 self.add_info_extractor(ie)
541 return ie
542
543 def add_default_info_extractors(self):
544 """
545 Add the InfoExtractors returned by gen_extractors to the end of the list
546 """
547 for ie in gen_extractor_classes():
548 self.add_info_extractor(ie)
549
550 def add_post_processor(self, pp, after_move=False):
551 """Add a PostProcessor object to the end of the chain."""
552 if after_move:
553 self._pps_end.append(pp)
554 else:
555 self._pps.append(pp)
556 pp.set_downloader(self)
557
558 def add_post_hook(self, ph):
559 """Add the post hook"""
560 self._post_hooks.append(ph)
561
562 def add_progress_hook(self, ph):
563 """Add the progress hook (currently only for the file downloader)"""
564 self._progress_hooks.append(ph)
565
566 def _bidi_workaround(self, message):
567 if not hasattr(self, '_output_channel'):
568 return message
569
570 assert hasattr(self, '_output_process')
571 assert isinstance(message, compat_str)
572 line_count = message.count('\n') + 1
573 self._output_process.stdin.write((message + '\n').encode('utf-8'))
574 self._output_process.stdin.flush()
575 res = ''.join(self._output_channel.readline().decode('utf-8')
576 for _ in range(line_count))
577 return res[:-len('\n')]
578
579 def to_screen(self, message, skip_eol=False):
580 """Print message to stdout if not in quiet mode."""
581 return self.to_stdout(message, skip_eol, check_quiet=True)
582
583 def _write_string(self, s, out=None):
584 write_string(s, out=out, encoding=self.params.get('encoding'))
585
586 def to_stdout(self, message, skip_eol=False, check_quiet=False):
587 """Print message to stdout if not in quiet mode."""
588 if self.params.get('logger'):
589 self.params['logger'].debug(message)
590 elif not check_quiet or not self.params.get('quiet', False):
591 message = self._bidi_workaround(message)
592 terminator = ['\n', ''][skip_eol]
593 output = message + terminator
594
595 self._write_string(output, self._screen_file)
596
597 def to_stderr(self, message):
598 """Print message to stderr."""
599 assert isinstance(message, compat_str)
600 if self.params.get('logger'):
601 self.params['logger'].error(message)
602 else:
603 message = self._bidi_workaround(message)
604 output = message + '\n'
605 self._write_string(output, self._err_file)
606
607 def to_console_title(self, message):
608 if not self.params.get('consoletitle', False):
609 return
610 if compat_os_name == 'nt':
611 if ctypes.windll.kernel32.GetConsoleWindow():
612 # c_wchar_p() might not be necessary if `message` is
613 # already of type unicode()
614 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
615 elif 'TERM' in os.environ:
616 self._write_string('\033]0;%s\007' % message, self._screen_file)
617
618 def save_console_title(self):
619 if not self.params.get('consoletitle', False):
620 return
621 if self.params.get('simulate', False):
622 return
623 if compat_os_name != 'nt' and 'TERM' in os.environ:
624 # Save the title on stack
625 self._write_string('\033[22;0t', self._screen_file)
626
627 def restore_console_title(self):
628 if not self.params.get('consoletitle', False):
629 return
630 if self.params.get('simulate', False):
631 return
632 if compat_os_name != 'nt' and 'TERM' in os.environ:
633 # Restore the title from stack
634 self._write_string('\033[23;0t', self._screen_file)
635
636 def __enter__(self):
637 self.save_console_title()
638 return self
639
640 def __exit__(self, *args):
641 self.restore_console_title()
642
643 if self.params.get('cookiefile') is not None:
644 self.cookiejar.save(ignore_discard=True, ignore_expires=True)
645
646 def trouble(self, message=None, tb=None):
647 """Determine action to take when a download problem appears.
648
649 Depending on if the downloader has been configured to ignore
650 download errors or not, this method may throw an exception or
651 not when errors are found, after printing the message.
652
653 tb, if given, is additional traceback information.
654 """
655 if message is not None:
656 self.to_stderr(message)
657 if self.params.get('verbose'):
658 if tb is None:
659 if sys.exc_info()[0]: # if .trouble has been called from an except block
660 tb = ''
661 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
662 tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
663 tb += encode_compat_str(traceback.format_exc())
664 else:
665 tb_data = traceback.format_list(traceback.extract_stack())
666 tb = ''.join(tb_data)
667 self.to_stderr(tb)
668 if not self.params.get('ignoreerrors', False):
669 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
670 exc_info = sys.exc_info()[1].exc_info
671 else:
672 exc_info = sys.exc_info()
673 raise DownloadError(message, exc_info)
674 self._download_retcode = 1
675
676 def report_warning(self, message):
677 '''
678 Print the message to stderr, it will be prefixed with 'WARNING:'
679 If stderr is a tty file the 'WARNING:' will be colored
680 '''
681 if self.params.get('logger') is not None:
682 self.params['logger'].warning(message)
683 else:
684 if self.params.get('no_warnings'):
685 return
686 if not self.params.get('no_color') and self._err_file.isatty() and compat_os_name != 'nt':
687 _msg_header = '\033[0;33mWARNING:\033[0m'
688 else:
689 _msg_header = 'WARNING:'
690 warning_message = '%s %s' % (_msg_header, message)
691 self.to_stderr(warning_message)
692
693 def report_error(self, message, tb=None):
694 '''
695 Do the same as trouble, but prefixes the message with 'ERROR:', colored
696 in red if stderr is a tty file.
697 '''
698 if not self.params.get('no_color') and self._err_file.isatty() and compat_os_name != 'nt':
699 _msg_header = '\033[0;31mERROR:\033[0m'
700 else:
701 _msg_header = 'ERROR:'
702 error_message = '%s %s' % (_msg_header, message)
703 self.trouble(error_message, tb)
704
705 def report_file_already_downloaded(self, file_name):
706 """Report file has already been fully downloaded."""
707 try:
708 self.to_screen('[download] %s has already been downloaded' % file_name)
709 except UnicodeEncodeError:
710 self.to_screen('[download] The file has already been downloaded')
711
712 def report_file_delete(self, file_name):
713 """Report that existing file will be deleted."""
714 try:
715 self.to_screen('Deleting already existent file %s' % file_name)
716 except UnicodeEncodeError:
717 self.to_screen('Deleting already existent file')
718
719 def prepare_filename(self, info_dict, warn=False):
720 """Generate the output filename."""
721 try:
722 template_dict = dict(info_dict)
723
724 template_dict['epoch'] = int(time.time())
725 autonumber_size = self.params.get('autonumber_size')
726 if autonumber_size is None:
727 autonumber_size = 5
728 template_dict['autonumber'] = self.params.get('autonumber_start', 1) - 1 + self._num_downloads
729 if template_dict.get('resolution') is None:
730 if template_dict.get('width') and template_dict.get('height'):
731 template_dict['resolution'] = '%dx%d' % (template_dict['width'], template_dict['height'])
732 elif template_dict.get('height'):
733 template_dict['resolution'] = '%sp' % template_dict['height']
734 elif template_dict.get('width'):
735 template_dict['resolution'] = '%dx?' % template_dict['width']
736
737 sanitize = lambda k, v: sanitize_filename(
738 compat_str(v),
739 restricted=self.params.get('restrictfilenames'),
740 is_id=(k == 'id' or k.endswith('_id')))
741 template_dict = dict((k, v if isinstance(v, compat_numeric_types) else sanitize(k, v))
742 for k, v in template_dict.items()
743 if v is not None and not isinstance(v, (list, tuple, dict)))
744 template_dict = collections.defaultdict(lambda: 'NA', template_dict)
745
746 outtmpl = self.params.get('outtmpl', DEFAULT_OUTTMPL)
747
748 # For fields playlist_index and autonumber convert all occurrences
749 # of %(field)s to %(field)0Nd for backward compatibility
750 field_size_compat_map = {
751 'playlist_index': len(str(template_dict['n_entries'])),
752 'autonumber': autonumber_size,
753 }
754 FIELD_SIZE_COMPAT_RE = r'(?<!%)%\((?P<field>autonumber|playlist_index)\)s'
755 mobj = re.search(FIELD_SIZE_COMPAT_RE, outtmpl)
756 if mobj:
757 outtmpl = re.sub(
758 FIELD_SIZE_COMPAT_RE,
759 r'%%(\1)0%dd' % field_size_compat_map[mobj.group('field')],
760 outtmpl)
761
762 # Missing numeric fields used together with integer presentation types
763 # in format specification will break the argument substitution since
764 # string 'NA' is returned for missing fields. We will patch output
765 # template for missing fields to meet string presentation type.
766 for numeric_field in self._NUMERIC_FIELDS:
767 if numeric_field not in template_dict:
768 # As of [1] format syntax is:
769 # %[mapping_key][conversion_flags][minimum_width][.precision][length_modifier]type
770 # 1. https://docs.python.org/2/library/stdtypes.html#string-formatting
771 FORMAT_RE = r'''(?x)
772 (?<!%)
773 %
774 \({0}\) # mapping key
775 (?:[#0\-+ ]+)? # conversion flags (optional)
776 (?:\d+)? # minimum field width (optional)
777 (?:\.\d+)? # precision (optional)
778 [hlL]? # length modifier (optional)
779 [diouxXeEfFgGcrs%] # conversion type
780 '''
781 outtmpl = re.sub(
782 FORMAT_RE.format(numeric_field),
783 r'%({0})s'.format(numeric_field), outtmpl)
784
785 # expand_path translates '%%' into '%' and '$$' into '$'
786 # correspondingly that is not what we want since we need to keep
787 # '%%' intact for template dict substitution step. Working around
788 # with boundary-alike separator hack.
789 sep = ''.join([random.choice(ascii_letters) for _ in range(32)])
790 outtmpl = outtmpl.replace('%%', '%{0}%'.format(sep)).replace('$$', '${0}$'.format(sep))
791
792 # outtmpl should be expand_path'ed before template dict substitution
793 # because meta fields may contain env variables we don't want to
794 # be expanded. For example, for outtmpl "%(title)s.%(ext)s" and
795 # title "Hello $PATH", we don't want `$PATH` to be expanded.
796 filename = expand_path(outtmpl).replace(sep, '') % template_dict
797
798 # https://github.com/blackjack4494/youtube-dlc/issues/85
799 trim_file_name = self.params.get('trim_file_name', False)
800 if trim_file_name:
801 fn_groups = filename.rsplit('.')
802 ext = fn_groups[-1]
803 sub_ext = ''
804 if len(fn_groups) > 2:
805 sub_ext = fn_groups[-2]
806 filename = '.'.join(filter(None, [fn_groups[0][:trim_file_name], sub_ext, ext]))
807
808 # Temporary fix for #4787
809 # 'Treat' all problem characters by passing filename through preferredencoding
810 # to workaround encoding issues with subprocess on python2 @ Windows
811 if sys.version_info < (3, 0) and sys.platform == 'win32':
812 filename = encodeFilename(filename, True).decode(preferredencoding())
813 filename = sanitize_path(filename)
814
815 if warn and not self.__prepare_filename_warned:
816 if not self.params.get('paths'):
817 pass
818 elif filename == '-':
819 self.report_warning('--paths is ignored when an outputting to stdout')
820 elif os.path.isabs(filename):
821 self.report_warning('--paths is ignored since an absolute path is given in output template')
822 self.__prepare_filename_warned = True
823
824 return filename
825 except ValueError as err:
826 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
827 return None
828
829 def prepare_filepath(self, filename, dir_type=''):
830 if filename == '-':
831 return filename
832 paths = self.params.get('paths', {})
833 assert isinstance(paths, dict)
834 homepath = expand_path(paths.get('home', '').strip())
835 assert isinstance(homepath, compat_str)
836 subdir = expand_path(paths.get(dir_type, '').strip()) if dir_type else ''
837 assert isinstance(subdir, compat_str)
838 return sanitize_path(os.path.join(homepath, subdir, filename))
839
840 def _match_entry(self, info_dict, incomplete):
841 """ Returns None if the file should be downloaded """
842
843 def check_filter():
844 video_title = info_dict.get('title', info_dict.get('id', 'video'))
845 if 'title' in info_dict:
846 # This can happen when we're just evaluating the playlist
847 title = info_dict['title']
848 matchtitle = self.params.get('matchtitle', False)
849 if matchtitle:
850 if not re.search(matchtitle, title, re.IGNORECASE):
851 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
852 rejecttitle = self.params.get('rejecttitle', False)
853 if rejecttitle:
854 if re.search(rejecttitle, title, re.IGNORECASE):
855 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
856 date = info_dict.get('upload_date')
857 if date is not None:
858 dateRange = self.params.get('daterange', DateRange())
859 if date not in dateRange:
860 return '%s upload date is not in range %s' % (date_from_str(date).isoformat(), dateRange)
861 view_count = info_dict.get('view_count')
862 if view_count is not None:
863 min_views = self.params.get('min_views')
864 if min_views is not None and view_count < min_views:
865 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
866 max_views = self.params.get('max_views')
867 if max_views is not None and view_count > max_views:
868 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
869 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
870 return 'Skipping "%s" because it is age restricted' % video_title
871 if self.in_download_archive(info_dict):
872 return '%s has already been recorded in archive' % video_title
873
874 if not incomplete:
875 match_filter = self.params.get('match_filter')
876 if match_filter is not None:
877 ret = match_filter(info_dict)
878 if ret is not None:
879 return ret
880 return None
881
882 reason = check_filter()
883 if reason is not None:
884 self.to_screen('[download] ' + reason)
885 if reason.endswith('has already been recorded in the archive') and self.params.get('break_on_existing', False):
886 raise ExistingVideoReached()
887 elif self.params.get('break_on_reject', False):
888 raise RejectedVideoReached()
889 return reason
890
891 @staticmethod
892 def add_extra_info(info_dict, extra_info):
893 '''Set the keys from extra_info in info dict if they are missing'''
894 for key, value in extra_info.items():
895 info_dict.setdefault(key, value)
896
897 def extract_info(self, url, download=True, ie_key=None, info_dict=None, extra_info={},
898 process=True, force_generic_extractor=False):
899 '''
900 Returns a list with a dictionary for each video we find.
901 If 'download', also downloads the videos.
902 extra_info is a dict containing the extra values to add to each result
903 '''
904
905 if not ie_key and force_generic_extractor:
906 ie_key = 'Generic'
907
908 if ie_key:
909 ies = [self.get_info_extractor(ie_key)]
910 else:
911 ies = self._ies
912
913 for ie in ies:
914 if not ie.suitable(url):
915 continue
916
917 ie_key = ie.ie_key()
918 ie = self.get_info_extractor(ie_key)
919 if not ie.working():
920 self.report_warning('The program functionality for this site has been marked as broken, '
921 'and will probably not work.')
922
923 try:
924 temp_id = str_or_none(
925 ie.extract_id(url) if callable(getattr(ie, 'extract_id', None))
926 else ie._match_id(url))
927 except (AssertionError, IndexError, AttributeError):
928 temp_id = None
929 if temp_id is not None and self.in_download_archive({'id': temp_id, 'ie_key': ie_key}):
930 self.to_screen("[%s] %s: has already been recorded in archive" % (
931 ie_key, temp_id))
932 break
933
934 return self.__extract_info(url, ie, download, extra_info, process, info_dict)
935
936 else:
937 self.report_error('no suitable InfoExtractor for URL %s' % url)
938
939 def __handle_extraction_exceptions(func):
940 def wrapper(self, *args, **kwargs):
941 try:
942 return func(self, *args, **kwargs)
943 except GeoRestrictedError as e:
944 msg = e.msg
945 if e.countries:
946 msg += '\nThis video is available in %s.' % ', '.join(
947 map(ISO3166Utils.short2full, e.countries))
948 msg += '\nYou might want to use a VPN or a proxy server (with --proxy) to workaround.'
949 self.report_error(msg)
950 except ExtractorError as e: # An error we somewhat expected
951 self.report_error(compat_str(e), e.format_traceback())
952 except (MaxDownloadsReached, ExistingVideoReached, RejectedVideoReached):
953 raise
954 except Exception as e:
955 if self.params.get('ignoreerrors', False):
956 self.report_error(error_to_compat_str(e), tb=encode_compat_str(traceback.format_exc()))
957 else:
958 raise
959 return wrapper
960
961 @__handle_extraction_exceptions
962 def __extract_info(self, url, ie, download, extra_info, process, info_dict):
963 ie_result = ie.extract(url)
964 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
965 return
966 if isinstance(ie_result, list):
967 # Backwards compatibility: old IE result format
968 ie_result = {
969 '_type': 'compat_list',
970 'entries': ie_result,
971 }
972 if info_dict:
973 if info_dict.get('id'):
974 ie_result['id'] = info_dict['id']
975 if info_dict.get('title'):
976 ie_result['title'] = info_dict['title']
977 self.add_default_extra_info(ie_result, ie, url)
978 if process:
979 return self.process_ie_result(ie_result, download, extra_info)
980 else:
981 return ie_result
982
983 def add_default_extra_info(self, ie_result, ie, url):
984 self.add_extra_info(ie_result, {
985 'extractor': ie.IE_NAME,
986 'webpage_url': url,
987 'duration_string': (
988 formatSeconds(ie_result['duration'], '-')
989 if ie_result.get('duration', None) is not None
990 else None),
991 'webpage_url_basename': url_basename(url),
992 'extractor_key': ie.ie_key(),
993 })
994
995 def process_ie_result(self, ie_result, download=True, extra_info={}):
996 """
997 Take the result of the ie(may be modified) and resolve all unresolved
998 references (URLs, playlist items).
999
1000 It will also download the videos if 'download'.
1001 Returns the resolved ie_result.
1002 """
1003 result_type = ie_result.get('_type', 'video')
1004
1005 if result_type in ('url', 'url_transparent'):
1006 ie_result['url'] = sanitize_url(ie_result['url'])
1007 extract_flat = self.params.get('extract_flat', False)
1008 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info)
1009 or extract_flat is True):
1010 self.__forced_printings(
1011 ie_result,
1012 self.prepare_filepath(self.prepare_filename(ie_result)),
1013 incomplete=True)
1014 return ie_result
1015
1016 if result_type == 'video':
1017 self.add_extra_info(ie_result, extra_info)
1018 return self.process_video_result(ie_result, download=download)
1019 elif result_type == 'url':
1020 # We have to add extra_info to the results because it may be
1021 # contained in a playlist
1022 return self.extract_info(ie_result['url'],
1023 download, info_dict=ie_result,
1024 ie_key=ie_result.get('ie_key'),
1025 extra_info=extra_info)
1026 elif result_type == 'url_transparent':
1027 # Use the information from the embedding page
1028 info = self.extract_info(
1029 ie_result['url'], ie_key=ie_result.get('ie_key'),
1030 extra_info=extra_info, download=False, process=False)
1031
1032 # extract_info may return None when ignoreerrors is enabled and
1033 # extraction failed with an error, don't crash and return early
1034 # in this case
1035 if not info:
1036 return info
1037
1038 force_properties = dict(
1039 (k, v) for k, v in ie_result.items() if v is not None)
1040 for f in ('_type', 'url', 'id', 'extractor', 'extractor_key', 'ie_key'):
1041 if f in force_properties:
1042 del force_properties[f]
1043 new_result = info.copy()
1044 new_result.update(force_properties)
1045
1046 # Extracted info may not be a video result (i.e.
1047 # info.get('_type', 'video') != video) but rather an url or
1048 # url_transparent. In such cases outer metadata (from ie_result)
1049 # should be propagated to inner one (info). For this to happen
1050 # _type of info should be overridden with url_transparent. This
1051 # fixes issue from https://github.com/ytdl-org/youtube-dl/pull/11163.
1052 if new_result.get('_type') == 'url':
1053 new_result['_type'] = 'url_transparent'
1054
1055 return self.process_ie_result(
1056 new_result, download=download, extra_info=extra_info)
1057 elif result_type in ('playlist', 'multi_video'):
1058 # Protect from infinite recursion due to recursively nested playlists
1059 # (see https://github.com/ytdl-org/youtube-dl/issues/27833)
1060 webpage_url = ie_result['webpage_url']
1061 if webpage_url in self._playlist_urls:
1062 self.to_screen(
1063 '[download] Skipping already downloaded playlist: %s'
1064 % ie_result.get('title') or ie_result.get('id'))
1065 return
1066
1067 self._playlist_level += 1
1068 self._playlist_urls.add(webpage_url)
1069 try:
1070 return self.__process_playlist(ie_result, download)
1071 finally:
1072 self._playlist_level -= 1
1073 if not self._playlist_level:
1074 self._playlist_urls.clear()
1075 elif result_type == 'compat_list':
1076 self.report_warning(
1077 'Extractor %s returned a compat_list result. '
1078 'It needs to be updated.' % ie_result.get('extractor'))
1079
1080 def _fixup(r):
1081 self.add_extra_info(
1082 r,
1083 {
1084 'extractor': ie_result['extractor'],
1085 'webpage_url': ie_result['webpage_url'],
1086 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1087 'extractor_key': ie_result['extractor_key'],
1088 }
1089 )
1090 return r
1091 ie_result['entries'] = [
1092 self.process_ie_result(_fixup(r), download, extra_info)
1093 for r in ie_result['entries']
1094 ]
1095 return ie_result
1096 else:
1097 raise Exception('Invalid result type: %s' % result_type)
1098
1099 def __process_playlist(self, ie_result, download):
1100 # We process each entry in the playlist
1101 playlist = ie_result.get('title') or ie_result.get('id')
1102 self.to_screen('[download] Downloading playlist: %s' % playlist)
1103
1104 playlist_results = []
1105
1106 playliststart = self.params.get('playliststart', 1) - 1
1107 playlistend = self.params.get('playlistend')
1108 # For backwards compatibility, interpret -1 as whole list
1109 if playlistend == -1:
1110 playlistend = None
1111
1112 playlistitems_str = self.params.get('playlist_items')
1113 playlistitems = None
1114 if playlistitems_str is not None:
1115 def iter_playlistitems(format):
1116 for string_segment in format.split(','):
1117 if '-' in string_segment:
1118 start, end = string_segment.split('-')
1119 for item in range(int(start), int(end) + 1):
1120 yield int(item)
1121 else:
1122 yield int(string_segment)
1123 playlistitems = orderedSet(iter_playlistitems(playlistitems_str))
1124
1125 ie_entries = ie_result['entries']
1126
1127 def make_playlistitems_entries(list_ie_entries):
1128 num_entries = len(list_ie_entries)
1129 return [
1130 list_ie_entries[i - 1] for i in playlistitems
1131 if -num_entries <= i - 1 < num_entries]
1132
1133 def report_download(num_entries):
1134 self.to_screen(
1135 '[%s] playlist %s: Downloading %d videos' %
1136 (ie_result['extractor'], playlist, num_entries))
1137
1138 if isinstance(ie_entries, list):
1139 n_all_entries = len(ie_entries)
1140 if playlistitems:
1141 entries = make_playlistitems_entries(ie_entries)
1142 else:
1143 entries = ie_entries[playliststart:playlistend]
1144 n_entries = len(entries)
1145 self.to_screen(
1146 '[%s] playlist %s: Collected %d video ids (downloading %d of them)' %
1147 (ie_result['extractor'], playlist, n_all_entries, n_entries))
1148 elif isinstance(ie_entries, PagedList):
1149 if playlistitems:
1150 entries = []
1151 for item in playlistitems:
1152 entries.extend(ie_entries.getslice(
1153 item - 1, item
1154 ))
1155 else:
1156 entries = ie_entries.getslice(
1157 playliststart, playlistend)
1158 n_entries = len(entries)
1159 report_download(n_entries)
1160 else: # iterable
1161 if playlistitems:
1162 entries = make_playlistitems_entries(list(itertools.islice(
1163 ie_entries, 0, max(playlistitems))))
1164 else:
1165 entries = list(itertools.islice(
1166 ie_entries, playliststart, playlistend))
1167 n_entries = len(entries)
1168 report_download(n_entries)
1169
1170 if self.params.get('playlistreverse', False):
1171 entries = entries[::-1]
1172
1173 if self.params.get('playlistrandom', False):
1174 random.shuffle(entries)
1175
1176 x_forwarded_for = ie_result.get('__x_forwarded_for_ip')
1177
1178 for i, entry in enumerate(entries, 1):
1179 self.to_screen('[download] Downloading video %s of %s' % (i, n_entries))
1180 # This __x_forwarded_for_ip thing is a bit ugly but requires
1181 # minimal changes
1182 if x_forwarded_for:
1183 entry['__x_forwarded_for_ip'] = x_forwarded_for
1184 extra = {
1185 'n_entries': n_entries,
1186 'playlist': playlist,
1187 'playlist_id': ie_result.get('id'),
1188 'playlist_title': ie_result.get('title'),
1189 'playlist_uploader': ie_result.get('uploader'),
1190 'playlist_uploader_id': ie_result.get('uploader_id'),
1191 'playlist_index': playlistitems[i - 1] if playlistitems else i + playliststart,
1192 'extractor': ie_result['extractor'],
1193 'webpage_url': ie_result['webpage_url'],
1194 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1195 'extractor_key': ie_result['extractor_key'],
1196 }
1197
1198 if self._match_entry(entry, incomplete=True) is not None:
1199 continue
1200
1201 entry_result = self.__process_iterable_entry(entry, download, extra)
1202 # TODO: skip failed (empty) entries?
1203 playlist_results.append(entry_result)
1204 ie_result['entries'] = playlist_results
1205 self.to_screen('[download] Finished downloading playlist: %s' % playlist)
1206 return ie_result
1207
1208 @__handle_extraction_exceptions
1209 def __process_iterable_entry(self, entry, download, extra_info):
1210 return self.process_ie_result(
1211 entry, download=download, extra_info=extra_info)
1212
1213 def _build_format_filter(self, filter_spec):
1214 " Returns a function to filter the formats according to the filter_spec "
1215
1216 OPERATORS = {
1217 '<': operator.lt,
1218 '<=': operator.le,
1219 '>': operator.gt,
1220 '>=': operator.ge,
1221 '=': operator.eq,
1222 '!=': operator.ne,
1223 }
1224 operator_rex = re.compile(r'''(?x)\s*
1225 (?P<key>width|height|tbr|abr|vbr|asr|filesize|filesize_approx|fps)
1226 \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
1227 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)
1228 $
1229 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
1230 m = operator_rex.search(filter_spec)
1231 if m:
1232 try:
1233 comparison_value = int(m.group('value'))
1234 except ValueError:
1235 comparison_value = parse_filesize(m.group('value'))
1236 if comparison_value is None:
1237 comparison_value = parse_filesize(m.group('value') + 'B')
1238 if comparison_value is None:
1239 raise ValueError(
1240 'Invalid value %r in format specification %r' % (
1241 m.group('value'), filter_spec))
1242 op = OPERATORS[m.group('op')]
1243
1244 if not m:
1245 STR_OPERATORS = {
1246 '=': operator.eq,
1247 '^=': lambda attr, value: attr.startswith(value),
1248 '$=': lambda attr, value: attr.endswith(value),
1249 '*=': lambda attr, value: value in attr,
1250 }
1251 str_operator_rex = re.compile(r'''(?x)
1252 \s*(?P<key>[a-zA-Z0-9._-]+)
1253 \s*(?P<negation>!\s*)?(?P<op>%s)(?P<none_inclusive>\s*\?)?
1254 \s*(?P<value>[a-zA-Z0-9._-]+)
1255 \s*$
1256 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
1257 m = str_operator_rex.search(filter_spec)
1258 if m:
1259 comparison_value = m.group('value')
1260 str_op = STR_OPERATORS[m.group('op')]
1261 if m.group('negation'):
1262 op = lambda attr, value: not str_op(attr, value)
1263 else:
1264 op = str_op
1265
1266 if not m:
1267 raise ValueError('Invalid filter specification %r' % filter_spec)
1268
1269 def _filter(f):
1270 actual_value = f.get(m.group('key'))
1271 if actual_value is None:
1272 return m.group('none_inclusive')
1273 return op(actual_value, comparison_value)
1274 return _filter
1275
1276 def _default_format_spec(self, info_dict, download=True):
1277
1278 def can_merge():
1279 merger = FFmpegMergerPP(self)
1280 return merger.available and merger.can_merge()
1281
1282 prefer_best = (
1283 not self.params.get('simulate', False)
1284 and download
1285 and (
1286 not can_merge()
1287 or info_dict.get('is_live', False)
1288 or self.params.get('outtmpl', DEFAULT_OUTTMPL) == '-'))
1289
1290 return (
1291 'best/bestvideo+bestaudio'
1292 if prefer_best
1293 else 'bestvideo*+bestaudio/best'
1294 if not self.params.get('allow_multiple_audio_streams', False)
1295 else 'bestvideo+bestaudio/best')
1296
1297 def build_format_selector(self, format_spec):
1298 def syntax_error(note, start):
1299 message = (
1300 'Invalid format specification: '
1301 '{0}\n\t{1}\n\t{2}^'.format(note, format_spec, ' ' * start[1]))
1302 return SyntaxError(message)
1303
1304 PICKFIRST = 'PICKFIRST'
1305 MERGE = 'MERGE'
1306 SINGLE = 'SINGLE'
1307 GROUP = 'GROUP'
1308 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
1309
1310 allow_multiple_streams = {'audio': self.params.get('allow_multiple_audio_streams', False),
1311 'video': self.params.get('allow_multiple_video_streams', False)}
1312
1313 def _parse_filter(tokens):
1314 filter_parts = []
1315 for type, string, start, _, _ in tokens:
1316 if type == tokenize.OP and string == ']':
1317 return ''.join(filter_parts)
1318 else:
1319 filter_parts.append(string)
1320
1321 def _remove_unused_ops(tokens):
1322 # Remove operators that we don't use and join them with the surrounding strings
1323 # for example: 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
1324 ALLOWED_OPS = ('/', '+', ',', '(', ')')
1325 last_string, last_start, last_end, last_line = None, None, None, None
1326 for type, string, start, end, line in tokens:
1327 if type == tokenize.OP and string == '[':
1328 if last_string:
1329 yield tokenize.NAME, last_string, last_start, last_end, last_line
1330 last_string = None
1331 yield type, string, start, end, line
1332 # everything inside brackets will be handled by _parse_filter
1333 for type, string, start, end, line in tokens:
1334 yield type, string, start, end, line
1335 if type == tokenize.OP and string == ']':
1336 break
1337 elif type == tokenize.OP and string in ALLOWED_OPS:
1338 if last_string:
1339 yield tokenize.NAME, last_string, last_start, last_end, last_line
1340 last_string = None
1341 yield type, string, start, end, line
1342 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
1343 if not last_string:
1344 last_string = string
1345 last_start = start
1346 last_end = end
1347 else:
1348 last_string += string
1349 if last_string:
1350 yield tokenize.NAME, last_string, last_start, last_end, last_line
1351
1352 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
1353 selectors = []
1354 current_selector = None
1355 for type, string, start, _, _ in tokens:
1356 # ENCODING is only defined in python 3.x
1357 if type == getattr(tokenize, 'ENCODING', None):
1358 continue
1359 elif type in [tokenize.NAME, tokenize.NUMBER]:
1360 current_selector = FormatSelector(SINGLE, string, [])
1361 elif type == tokenize.OP:
1362 if string == ')':
1363 if not inside_group:
1364 # ')' will be handled by the parentheses group
1365 tokens.restore_last_token()
1366 break
1367 elif inside_merge and string in ['/', ',']:
1368 tokens.restore_last_token()
1369 break
1370 elif inside_choice and string == ',':
1371 tokens.restore_last_token()
1372 break
1373 elif string == ',':
1374 if not current_selector:
1375 raise syntax_error('"," must follow a format selector', start)
1376 selectors.append(current_selector)
1377 current_selector = None
1378 elif string == '/':
1379 if not current_selector:
1380 raise syntax_error('"/" must follow a format selector', start)
1381 first_choice = current_selector
1382 second_choice = _parse_format_selection(tokens, inside_choice=True)
1383 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
1384 elif string == '[':
1385 if not current_selector:
1386 current_selector = FormatSelector(SINGLE, 'best', [])
1387 format_filter = _parse_filter(tokens)
1388 current_selector.filters.append(format_filter)
1389 elif string == '(':
1390 if current_selector:
1391 raise syntax_error('Unexpected "("', start)
1392 group = _parse_format_selection(tokens, inside_group=True)
1393 current_selector = FormatSelector(GROUP, group, [])
1394 elif string == '+':
1395 if not current_selector:
1396 raise syntax_error('Unexpected "+"', start)
1397 selector_1 = current_selector
1398 selector_2 = _parse_format_selection(tokens, inside_merge=True)
1399 if not selector_2:
1400 raise syntax_error('Expected a selector', start)
1401 current_selector = FormatSelector(MERGE, (selector_1, selector_2), [])
1402 else:
1403 raise syntax_error('Operator not recognized: "{0}"'.format(string), start)
1404 elif type == tokenize.ENDMARKER:
1405 break
1406 if current_selector:
1407 selectors.append(current_selector)
1408 return selectors
1409
1410 def _build_selector_function(selector):
1411 if isinstance(selector, list): # ,
1412 fs = [_build_selector_function(s) for s in selector]
1413
1414 def selector_function(ctx):
1415 for f in fs:
1416 for format in f(ctx):
1417 yield format
1418 return selector_function
1419
1420 elif selector.type == GROUP: # ()
1421 selector_function = _build_selector_function(selector.selector)
1422
1423 elif selector.type == PICKFIRST: # /
1424 fs = [_build_selector_function(s) for s in selector.selector]
1425
1426 def selector_function(ctx):
1427 for f in fs:
1428 picked_formats = list(f(ctx))
1429 if picked_formats:
1430 return picked_formats
1431 return []
1432
1433 elif selector.type == SINGLE: # atom
1434 format_spec = selector.selector if selector.selector is not None else 'best'
1435
1436 if format_spec == 'all':
1437 def selector_function(ctx):
1438 formats = list(ctx['formats'])
1439 if formats:
1440 for f in formats:
1441 yield f
1442
1443 else:
1444 format_fallback = False
1445 format_spec_obj = re.match(r'(best|worst|b|w)(video|audio|v|a)?(\*)?$', format_spec)
1446 if format_spec_obj is not None:
1447 format_idx = 0 if format_spec_obj.group(1)[0] == 'w' else -1
1448 format_type = format_spec_obj.group(2)[0] if format_spec_obj.group(2) else False
1449 not_format_type = 'v' if format_type == 'a' else 'a'
1450 format_modified = format_spec_obj.group(3) is not None
1451
1452 format_fallback = not format_type and not format_modified # for b, w
1453 filter_f = ((lambda f: f.get(format_type + 'codec') != 'none')
1454 if format_type and format_modified # bv*, ba*, wv*, wa*
1455 else (lambda f: f.get(not_format_type + 'codec') == 'none')
1456 if format_type # bv, ba, wv, wa
1457 else (lambda f: f.get('vcodec') != 'none' and f.get('acodec') != 'none')
1458 if not format_modified # b, w
1459 else None) # b*, w*
1460 else:
1461 format_idx = -1
1462 filter_f = ((lambda f: f.get('ext') == format_spec)
1463 if format_spec in ['mp4', 'flv', 'webm', '3gp', 'm4a', 'mp3', 'ogg', 'aac', 'wav'] # extension
1464 else (lambda f: f.get('format_id') == format_spec)) # id
1465
1466 def selector_function(ctx):
1467 formats = list(ctx['formats'])
1468 if not formats:
1469 return
1470 matches = list(filter(filter_f, formats)) if filter_f is not None else formats
1471 if matches:
1472 yield matches[format_idx]
1473 elif format_fallback == 'force' or (format_fallback and ctx['incomplete_formats']):
1474 # for extractors with incomplete formats (audio only (soundcloud)
1475 # or video only (imgur)) best/worst will fallback to
1476 # best/worst {video,audio}-only format
1477 yield formats[format_idx]
1478
1479 elif selector.type == MERGE: # +
1480 def _merge(formats_pair):
1481 format_1, format_2 = formats_pair
1482
1483 formats_info = []
1484 formats_info.extend(format_1.get('requested_formats', (format_1,)))
1485 formats_info.extend(format_2.get('requested_formats', (format_2,)))
1486
1487 if not allow_multiple_streams['video'] or not allow_multiple_streams['audio']:
1488 get_no_more = {"video": False, "audio": False}
1489 for (i, fmt_info) in enumerate(formats_info):
1490 for aud_vid in ["audio", "video"]:
1491 if not allow_multiple_streams[aud_vid] and fmt_info.get(aud_vid[0] + 'codec') != 'none':
1492 if get_no_more[aud_vid]:
1493 formats_info.pop(i)
1494 get_no_more[aud_vid] = True
1495
1496 if len(formats_info) == 1:
1497 return formats_info[0]
1498
1499 video_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('vcodec') != 'none']
1500 audio_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('acodec') != 'none']
1501
1502 the_only_video = video_fmts[0] if len(video_fmts) == 1 else None
1503 the_only_audio = audio_fmts[0] if len(audio_fmts) == 1 else None
1504
1505 output_ext = self.params.get('merge_output_format')
1506 if not output_ext:
1507 if the_only_video:
1508 output_ext = the_only_video['ext']
1509 elif the_only_audio and not video_fmts:
1510 output_ext = the_only_audio['ext']
1511 else:
1512 output_ext = 'mkv'
1513
1514 new_dict = {
1515 'requested_formats': formats_info,
1516 'format': '+'.join(fmt_info.get('format') for fmt_info in formats_info),
1517 'format_id': '+'.join(fmt_info.get('format_id') for fmt_info in formats_info),
1518 'ext': output_ext,
1519 }
1520
1521 if the_only_video:
1522 new_dict.update({
1523 'width': the_only_video.get('width'),
1524 'height': the_only_video.get('height'),
1525 'resolution': the_only_video.get('resolution'),
1526 'fps': the_only_video.get('fps'),
1527 'vcodec': the_only_video.get('vcodec'),
1528 'vbr': the_only_video.get('vbr'),
1529 'stretched_ratio': the_only_video.get('stretched_ratio'),
1530 })
1531
1532 if the_only_audio:
1533 new_dict.update({
1534 'acodec': the_only_audio.get('acodec'),
1535 'abr': the_only_audio.get('abr'),
1536 })
1537
1538 return new_dict
1539
1540 selector_1, selector_2 = map(_build_selector_function, selector.selector)
1541
1542 def selector_function(ctx):
1543 for pair in itertools.product(
1544 selector_1(copy.deepcopy(ctx)), selector_2(copy.deepcopy(ctx))):
1545 yield _merge(pair)
1546
1547 filters = [self._build_format_filter(f) for f in selector.filters]
1548
1549 def final_selector(ctx):
1550 ctx_copy = copy.deepcopy(ctx)
1551 for _filter in filters:
1552 ctx_copy['formats'] = list(filter(_filter, ctx_copy['formats']))
1553 return selector_function(ctx_copy)
1554 return final_selector
1555
1556 stream = io.BytesIO(format_spec.encode('utf-8'))
1557 try:
1558 tokens = list(_remove_unused_ops(compat_tokenize_tokenize(stream.readline)))
1559 except tokenize.TokenError:
1560 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
1561
1562 class TokenIterator(object):
1563 def __init__(self, tokens):
1564 self.tokens = tokens
1565 self.counter = 0
1566
1567 def __iter__(self):
1568 return self
1569
1570 def __next__(self):
1571 if self.counter >= len(self.tokens):
1572 raise StopIteration()
1573 value = self.tokens[self.counter]
1574 self.counter += 1
1575 return value
1576
1577 next = __next__
1578
1579 def restore_last_token(self):
1580 self.counter -= 1
1581
1582 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
1583 return _build_selector_function(parsed_selector)
1584
1585 def _calc_headers(self, info_dict):
1586 res = std_headers.copy()
1587
1588 add_headers = info_dict.get('http_headers')
1589 if add_headers:
1590 res.update(add_headers)
1591
1592 cookies = self._calc_cookies(info_dict)
1593 if cookies:
1594 res['Cookie'] = cookies
1595
1596 if 'X-Forwarded-For' not in res:
1597 x_forwarded_for_ip = info_dict.get('__x_forwarded_for_ip')
1598 if x_forwarded_for_ip:
1599 res['X-Forwarded-For'] = x_forwarded_for_ip
1600
1601 return res
1602
1603 def _calc_cookies(self, info_dict):
1604 pr = sanitized_Request(info_dict['url'])
1605 self.cookiejar.add_cookie_header(pr)
1606 return pr.get_header('Cookie')
1607
1608 def process_video_result(self, info_dict, download=True):
1609 assert info_dict.get('_type', 'video') == 'video'
1610
1611 if 'id' not in info_dict:
1612 raise ExtractorError('Missing "id" field in extractor result')
1613 if 'title' not in info_dict:
1614 raise ExtractorError('Missing "title" field in extractor result')
1615
1616 def report_force_conversion(field, field_not, conversion):
1617 self.report_warning(
1618 '"%s" field is not %s - forcing %s conversion, there is an error in extractor'
1619 % (field, field_not, conversion))
1620
1621 def sanitize_string_field(info, string_field):
1622 field = info.get(string_field)
1623 if field is None or isinstance(field, compat_str):
1624 return
1625 report_force_conversion(string_field, 'a string', 'string')
1626 info[string_field] = compat_str(field)
1627
1628 def sanitize_numeric_fields(info):
1629 for numeric_field in self._NUMERIC_FIELDS:
1630 field = info.get(numeric_field)
1631 if field is None or isinstance(field, compat_numeric_types):
1632 continue
1633 report_force_conversion(numeric_field, 'numeric', 'int')
1634 info[numeric_field] = int_or_none(field)
1635
1636 sanitize_string_field(info_dict, 'id')
1637 sanitize_numeric_fields(info_dict)
1638
1639 if 'playlist' not in info_dict:
1640 # It isn't part of a playlist
1641 info_dict['playlist'] = None
1642 info_dict['playlist_index'] = None
1643
1644 thumbnails = info_dict.get('thumbnails')
1645 if thumbnails is None:
1646 thumbnail = info_dict.get('thumbnail')
1647 if thumbnail:
1648 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
1649 if thumbnails:
1650 thumbnails.sort(key=lambda t: (
1651 t.get('preference') if t.get('preference') is not None else -1,
1652 t.get('width') if t.get('width') is not None else -1,
1653 t.get('height') if t.get('height') is not None else -1,
1654 t.get('id') if t.get('id') is not None else '', t.get('url')))
1655 for i, t in enumerate(thumbnails):
1656 t['url'] = sanitize_url(t['url'])
1657 if t.get('width') and t.get('height'):
1658 t['resolution'] = '%dx%d' % (t['width'], t['height'])
1659 if t.get('id') is None:
1660 t['id'] = '%d' % i
1661
1662 if self.params.get('list_thumbnails'):
1663 self.list_thumbnails(info_dict)
1664 return
1665
1666 thumbnail = info_dict.get('thumbnail')
1667 if thumbnail:
1668 info_dict['thumbnail'] = sanitize_url(thumbnail)
1669 elif thumbnails:
1670 info_dict['thumbnail'] = thumbnails[-1]['url']
1671
1672 if 'display_id' not in info_dict and 'id' in info_dict:
1673 info_dict['display_id'] = info_dict['id']
1674
1675 if info_dict.get('upload_date') is None and info_dict.get('timestamp') is not None:
1676 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
1677 # see http://bugs.python.org/issue1646728)
1678 try:
1679 upload_date = datetime.datetime.utcfromtimestamp(info_dict['timestamp'])
1680 info_dict['upload_date'] = upload_date.strftime('%Y%m%d')
1681 except (ValueError, OverflowError, OSError):
1682 pass
1683
1684 # Auto generate title fields corresponding to the *_number fields when missing
1685 # in order to always have clean titles. This is very common for TV series.
1686 for field in ('chapter', 'season', 'episode'):
1687 if info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
1688 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
1689
1690 for cc_kind in ('subtitles', 'automatic_captions'):
1691 cc = info_dict.get(cc_kind)
1692 if cc:
1693 for _, subtitle in cc.items():
1694 for subtitle_format in subtitle:
1695 if subtitle_format.get('url'):
1696 subtitle_format['url'] = sanitize_url(subtitle_format['url'])
1697 if subtitle_format.get('ext') is None:
1698 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
1699
1700 automatic_captions = info_dict.get('automatic_captions')
1701 subtitles = info_dict.get('subtitles')
1702
1703 if self.params.get('listsubtitles', False):
1704 if 'automatic_captions' in info_dict:
1705 self.list_subtitles(
1706 info_dict['id'], automatic_captions, 'automatic captions')
1707 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
1708 return
1709
1710 info_dict['requested_subtitles'] = self.process_subtitles(
1711 info_dict['id'], subtitles, automatic_captions)
1712
1713 # We now pick which formats have to be downloaded
1714 if info_dict.get('formats') is None:
1715 # There's only one format available
1716 formats = [info_dict]
1717 else:
1718 formats = info_dict['formats']
1719
1720 if not formats:
1721 raise ExtractorError('No video formats found!')
1722
1723 def is_wellformed(f):
1724 url = f.get('url')
1725 if not url:
1726 self.report_warning(
1727 '"url" field is missing or empty - skipping format, '
1728 'there is an error in extractor')
1729 return False
1730 if isinstance(url, bytes):
1731 sanitize_string_field(f, 'url')
1732 return True
1733
1734 # Filter out malformed formats for better extraction robustness
1735 formats = list(filter(is_wellformed, formats))
1736
1737 formats_dict = {}
1738
1739 # We check that all the formats have the format and format_id fields
1740 for i, format in enumerate(formats):
1741 sanitize_string_field(format, 'format_id')
1742 sanitize_numeric_fields(format)
1743 format['url'] = sanitize_url(format['url'])
1744 if not format.get('format_id'):
1745 format['format_id'] = compat_str(i)
1746 else:
1747 # Sanitize format_id from characters used in format selector expression
1748 format['format_id'] = re.sub(r'[\s,/+\[\]()]', '_', format['format_id'])
1749 format_id = format['format_id']
1750 if format_id not in formats_dict:
1751 formats_dict[format_id] = []
1752 formats_dict[format_id].append(format)
1753
1754 # Make sure all formats have unique format_id
1755 for format_id, ambiguous_formats in formats_dict.items():
1756 if len(ambiguous_formats) > 1:
1757 for i, format in enumerate(ambiguous_formats):
1758 format['format_id'] = '%s-%d' % (format_id, i)
1759
1760 for i, format in enumerate(formats):
1761 if format.get('format') is None:
1762 format['format'] = '{id} - {res}{note}'.format(
1763 id=format['format_id'],
1764 res=self.format_resolution(format),
1765 note=' ({0})'.format(format['format_note']) if format.get('format_note') is not None else '',
1766 )
1767 # Automatically determine file extension if missing
1768 if format.get('ext') is None:
1769 format['ext'] = determine_ext(format['url']).lower()
1770 # Automatically determine protocol if missing (useful for format
1771 # selection purposes)
1772 if format.get('protocol') is None:
1773 format['protocol'] = determine_protocol(format)
1774 # Add HTTP headers, so that external programs can use them from the
1775 # json output
1776 full_format_info = info_dict.copy()
1777 full_format_info.update(format)
1778 format['http_headers'] = self._calc_headers(full_format_info)
1779 # Remove private housekeeping stuff
1780 if '__x_forwarded_for_ip' in info_dict:
1781 del info_dict['__x_forwarded_for_ip']
1782
1783 # TODO Central sorting goes here
1784
1785 if formats[0] is not info_dict:
1786 # only set the 'formats' fields if the original info_dict list them
1787 # otherwise we end up with a circular reference, the first (and unique)
1788 # element in the 'formats' field in info_dict is info_dict itself,
1789 # which can't be exported to json
1790 info_dict['formats'] = formats
1791 if self.params.get('listformats'):
1792 self.list_formats(info_dict)
1793 return
1794
1795 req_format = self.params.get('format')
1796 if req_format is None:
1797 req_format = self._default_format_spec(info_dict, download=download)
1798 if self.params.get('verbose'):
1799 self._write_string('[debug] Default format spec: %s\n' % req_format)
1800
1801 format_selector = self.build_format_selector(req_format)
1802
1803 # While in format selection we may need to have an access to the original
1804 # format set in order to calculate some metrics or do some processing.
1805 # For now we need to be able to guess whether original formats provided
1806 # by extractor are incomplete or not (i.e. whether extractor provides only
1807 # video-only or audio-only formats) for proper formats selection for
1808 # extractors with such incomplete formats (see
1809 # https://github.com/ytdl-org/youtube-dl/pull/5556).
1810 # Since formats may be filtered during format selection and may not match
1811 # the original formats the results may be incorrect. Thus original formats
1812 # or pre-calculated metrics should be passed to format selection routines
1813 # as well.
1814 # We will pass a context object containing all necessary additional data
1815 # instead of just formats.
1816 # This fixes incorrect format selection issue (see
1817 # https://github.com/ytdl-org/youtube-dl/issues/10083).
1818 incomplete_formats = (
1819 # All formats are video-only or
1820 all(f.get('vcodec') != 'none' and f.get('acodec') == 'none' for f in formats)
1821 # all formats are audio-only
1822 or all(f.get('vcodec') == 'none' and f.get('acodec') != 'none' for f in formats))
1823
1824 ctx = {
1825 'formats': formats,
1826 'incomplete_formats': incomplete_formats,
1827 }
1828
1829 formats_to_download = list(format_selector(ctx))
1830 if not formats_to_download:
1831 raise ExtractorError('requested format not available',
1832 expected=True)
1833
1834 if download:
1835 self.to_screen('[info] Downloading format(s) %s' % ", ".join([f['format_id'] for f in formats_to_download]))
1836 if len(formats_to_download) > 1:
1837 self.to_screen('[info] %s: downloading video in %s formats' % (info_dict['id'], len(formats_to_download)))
1838 for format in formats_to_download:
1839 new_info = dict(info_dict)
1840 new_info.update(format)
1841 self.process_info(new_info)
1842 # We update the info dict with the best quality format (backwards compatibility)
1843 info_dict.update(formats_to_download[-1])
1844 return info_dict
1845
1846 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
1847 """Select the requested subtitles and their format"""
1848 available_subs = {}
1849 if normal_subtitles and self.params.get('writesubtitles'):
1850 available_subs.update(normal_subtitles)
1851 if automatic_captions and self.params.get('writeautomaticsub'):
1852 for lang, cap_info in automatic_captions.items():
1853 if lang not in available_subs:
1854 available_subs[lang] = cap_info
1855
1856 if (not self.params.get('writesubtitles') and not
1857 self.params.get('writeautomaticsub') or not
1858 available_subs):
1859 return None
1860
1861 if self.params.get('allsubtitles', False):
1862 requested_langs = available_subs.keys()
1863 else:
1864 if self.params.get('subtitleslangs', False):
1865 requested_langs = self.params.get('subtitleslangs')
1866 elif 'en' in available_subs:
1867 requested_langs = ['en']
1868 else:
1869 requested_langs = [list(available_subs.keys())[0]]
1870
1871 formats_query = self.params.get('subtitlesformat', 'best')
1872 formats_preference = formats_query.split('/') if formats_query else []
1873 subs = {}
1874 for lang in requested_langs:
1875 formats = available_subs.get(lang)
1876 if formats is None:
1877 self.report_warning('%s subtitles not available for %s' % (lang, video_id))
1878 continue
1879 for ext in formats_preference:
1880 if ext == 'best':
1881 f = formats[-1]
1882 break
1883 matches = list(filter(lambda f: f['ext'] == ext, formats))
1884 if matches:
1885 f = matches[-1]
1886 break
1887 else:
1888 f = formats[-1]
1889 self.report_warning(
1890 'No subtitle format found matching "%s" for language %s, '
1891 'using %s' % (formats_query, lang, f['ext']))
1892 subs[lang] = f
1893 return subs
1894
1895 def __forced_printings(self, info_dict, filename, incomplete):
1896 def print_mandatory(field):
1897 if (self.params.get('force%s' % field, False)
1898 and (not incomplete or info_dict.get(field) is not None)):
1899 self.to_stdout(info_dict[field])
1900
1901 def print_optional(field):
1902 if (self.params.get('force%s' % field, False)
1903 and info_dict.get(field) is not None):
1904 self.to_stdout(info_dict[field])
1905
1906 print_mandatory('title')
1907 print_mandatory('id')
1908 if self.params.get('forceurl', False) and not incomplete:
1909 if info_dict.get('requested_formats') is not None:
1910 for f in info_dict['requested_formats']:
1911 self.to_stdout(f['url'] + f.get('play_path', ''))
1912 else:
1913 # For RTMP URLs, also include the playpath
1914 self.to_stdout(info_dict['url'] + info_dict.get('play_path', ''))
1915 print_optional('thumbnail')
1916 print_optional('description')
1917 if self.params.get('forcefilename', False) and filename is not None:
1918 self.to_stdout(filename)
1919 if self.params.get('forceduration', False) and info_dict.get('duration') is not None:
1920 self.to_stdout(formatSeconds(info_dict['duration']))
1921 print_mandatory('format')
1922 if self.params.get('forcejson', False):
1923 self.to_stdout(json.dumps(info_dict))
1924
1925 def process_info(self, info_dict):
1926 """Process a single resolved IE result."""
1927
1928 assert info_dict.get('_type', 'video') == 'video'
1929
1930 info_dict.setdefault('__postprocessors', [])
1931
1932 max_downloads = self.params.get('max_downloads')
1933 if max_downloads is not None:
1934 if self._num_downloads >= int(max_downloads):
1935 raise MaxDownloadsReached()
1936
1937 # TODO: backward compatibility, to be removed
1938 info_dict['fulltitle'] = info_dict['title']
1939
1940 if 'format' not in info_dict:
1941 info_dict['format'] = info_dict['ext']
1942
1943 if self._match_entry(info_dict, incomplete=False) is not None:
1944 return
1945
1946 self._num_downloads += 1
1947
1948 filename = self.prepare_filename(info_dict, warn=True)
1949 info_dict['_filename'] = full_filename = self.prepare_filepath(filename)
1950 temp_filename = self.prepare_filepath(filename, 'temp')
1951 files_to_move = {}
1952
1953 # Forced printings
1954 self.__forced_printings(info_dict, full_filename, incomplete=False)
1955
1956 if self.params.get('simulate', False):
1957 if self.params.get('force_write_download_archive', False):
1958 self.record_download_archive(info_dict)
1959
1960 # Do nothing else if in simulate mode
1961 return
1962
1963 if filename is None:
1964 return
1965
1966 def ensure_dir_exists(path):
1967 return make_dir(path, self.report_error)
1968
1969 if not ensure_dir_exists(encodeFilename(full_filename)):
1970 return
1971 if not ensure_dir_exists(encodeFilename(temp_filename)):
1972 return
1973
1974 if self.params.get('writedescription', False):
1975 descfn = replace_extension(
1976 self.prepare_filepath(filename, 'description'),
1977 'description', info_dict.get('ext'))
1978 if not ensure_dir_exists(encodeFilename(descfn)):
1979 return
1980 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(descfn)):
1981 self.to_screen('[info] Video description is already present')
1982 elif info_dict.get('description') is None:
1983 self.report_warning('There\'s no description to write.')
1984 else:
1985 try:
1986 self.to_screen('[info] Writing video description to: ' + descfn)
1987 with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
1988 descfile.write(info_dict['description'])
1989 except (OSError, IOError):
1990 self.report_error('Cannot write description file ' + descfn)
1991 return
1992
1993 if self.params.get('writeannotations', False):
1994 annofn = replace_extension(
1995 self.prepare_filepath(filename, 'annotation'),
1996 'annotations.xml', info_dict.get('ext'))
1997 if not ensure_dir_exists(encodeFilename(annofn)):
1998 return
1999 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(annofn)):
2000 self.to_screen('[info] Video annotations are already present')
2001 elif not info_dict.get('annotations'):
2002 self.report_warning('There are no annotations to write.')
2003 else:
2004 try:
2005 self.to_screen('[info] Writing video annotations to: ' + annofn)
2006 with io.open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
2007 annofile.write(info_dict['annotations'])
2008 except (KeyError, TypeError):
2009 self.report_warning('There are no annotations to write.')
2010 except (OSError, IOError):
2011 self.report_error('Cannot write annotations file: ' + annofn)
2012 return
2013
2014 def dl(name, info, subtitle=False):
2015 fd = get_suitable_downloader(info, self.params)(self, self.params)
2016 for ph in self._progress_hooks:
2017 fd.add_progress_hook(ph)
2018 if self.params.get('verbose'):
2019 self.to_screen('[debug] Invoking downloader on %r' % info.get('url'))
2020 return fd.download(name, info, subtitle)
2021
2022 subtitles_are_requested = any([self.params.get('writesubtitles', False),
2023 self.params.get('writeautomaticsub')])
2024
2025 if subtitles_are_requested and info_dict.get('requested_subtitles'):
2026 # subtitles download errors are already managed as troubles in relevant IE
2027 # that way it will silently go on when used with unsupporting IE
2028 subtitles = info_dict['requested_subtitles']
2029 # ie = self.get_info_extractor(info_dict['extractor_key'])
2030 for sub_lang, sub_info in subtitles.items():
2031 sub_format = sub_info['ext']
2032 sub_filename = subtitles_filename(temp_filename, sub_lang, sub_format, info_dict.get('ext'))
2033 sub_filename_final = subtitles_filename(
2034 self.prepare_filepath(filename, 'subtitle'),
2035 sub_lang, sub_format, info_dict.get('ext'))
2036 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(sub_filename)):
2037 self.to_screen('[info] Video subtitle %s.%s is already present' % (sub_lang, sub_format))
2038 files_to_move[sub_filename] = sub_filename_final
2039 else:
2040 self.to_screen('[info] Writing video subtitles to: ' + sub_filename)
2041 if sub_info.get('data') is not None:
2042 try:
2043 # Use newline='' to prevent conversion of newline characters
2044 # See https://github.com/ytdl-org/youtube-dl/issues/10268
2045 with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8', newline='') as subfile:
2046 subfile.write(sub_info['data'])
2047 files_to_move[sub_filename] = sub_filename_final
2048 except (OSError, IOError):
2049 self.report_error('Cannot write subtitles file ' + sub_filename)
2050 return
2051 else:
2052 try:
2053 dl(sub_filename, sub_info, subtitle=True)
2054 '''
2055 if self.params.get('sleep_interval_subtitles', False):
2056 dl(sub_filename, sub_info)
2057 else:
2058 sub_data = ie._request_webpage(
2059 sub_info['url'], info_dict['id'], note=False).read()
2060 with io.open(encodeFilename(sub_filename), 'wb') as subfile:
2061 subfile.write(sub_data)
2062 '''
2063 files_to_move[sub_filename] = sub_filename_final
2064 except (ExtractorError, IOError, OSError, ValueError, compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
2065 self.report_warning('Unable to download subtitle for "%s": %s' %
2066 (sub_lang, error_to_compat_str(err)))
2067 continue
2068
2069 if self.params.get('skip_download', False):
2070 if self.params.get('convertsubtitles', False):
2071 # subconv = FFmpegSubtitlesConvertorPP(self, format=self.params.get('convertsubtitles'))
2072 filename_real_ext = os.path.splitext(filename)[1][1:]
2073 filename_wo_ext = (
2074 os.path.splitext(full_filename)[0]
2075 if filename_real_ext == info_dict['ext']
2076 else full_filename)
2077 afilename = '%s.%s' % (filename_wo_ext, self.params.get('convertsubtitles'))
2078 # if subconv.available:
2079 # info_dict['__postprocessors'].append(subconv)
2080 if os.path.exists(encodeFilename(afilename)):
2081 self.to_screen(
2082 '[download] %s has already been downloaded and '
2083 'converted' % afilename)
2084 else:
2085 try:
2086 self.post_process(full_filename, info_dict, files_to_move)
2087 except (PostProcessingError) as err:
2088 self.report_error('postprocessing: %s' % str(err))
2089 return
2090
2091 if self.params.get('writeinfojson', False):
2092 infofn = replace_extension(
2093 self.prepare_filepath(filename, 'infojson'),
2094 'info.json', info_dict.get('ext'))
2095 if not ensure_dir_exists(encodeFilename(infofn)):
2096 return
2097 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(infofn)):
2098 self.to_screen('[info] Video description metadata is already present')
2099 else:
2100 self.to_screen('[info] Writing video description metadata as JSON to: ' + infofn)
2101 try:
2102 write_json_file(self.filter_requested_info(info_dict), infofn)
2103 except (OSError, IOError):
2104 self.report_error('Cannot write metadata to JSON file ' + infofn)
2105 return
2106
2107 thumbdir = os.path.dirname(self.prepare_filepath(filename, 'thumbnail'))
2108 for thumbfn in self._write_thumbnails(info_dict, temp_filename):
2109 files_to_move[thumbfn] = os.path.join(thumbdir, os.path.basename(thumbfn))
2110
2111 # Write internet shortcut files
2112 url_link = webloc_link = desktop_link = False
2113 if self.params.get('writelink', False):
2114 if sys.platform == "darwin": # macOS.
2115 webloc_link = True
2116 elif sys.platform.startswith("linux"):
2117 desktop_link = True
2118 else: # if sys.platform in ['win32', 'cygwin']:
2119 url_link = True
2120 if self.params.get('writeurllink', False):
2121 url_link = True
2122 if self.params.get('writewebloclink', False):
2123 webloc_link = True
2124 if self.params.get('writedesktoplink', False):
2125 desktop_link = True
2126
2127 if url_link or webloc_link or desktop_link:
2128 if 'webpage_url' not in info_dict:
2129 self.report_error('Cannot write internet shortcut file because the "webpage_url" field is missing in the media information')
2130 return
2131 ascii_url = iri_to_uri(info_dict['webpage_url'])
2132
2133 def _write_link_file(extension, template, newline, embed_filename):
2134 linkfn = replace_extension(full_filename, extension, info_dict.get('ext'))
2135 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(linkfn)):
2136 self.to_screen('[info] Internet shortcut is already present')
2137 else:
2138 try:
2139 self.to_screen('[info] Writing internet shortcut to: ' + linkfn)
2140 with io.open(encodeFilename(to_high_limit_path(linkfn)), 'w', encoding='utf-8', newline=newline) as linkfile:
2141 template_vars = {'url': ascii_url}
2142 if embed_filename:
2143 template_vars['filename'] = linkfn[:-(len(extension) + 1)]
2144 linkfile.write(template % template_vars)
2145 except (OSError, IOError):
2146 self.report_error('Cannot write internet shortcut ' + linkfn)
2147 return False
2148 return True
2149
2150 if url_link:
2151 if not _write_link_file('url', DOT_URL_LINK_TEMPLATE, '\r\n', embed_filename=False):
2152 return
2153 if webloc_link:
2154 if not _write_link_file('webloc', DOT_WEBLOC_LINK_TEMPLATE, '\n', embed_filename=False):
2155 return
2156 if desktop_link:
2157 if not _write_link_file('desktop', DOT_DESKTOP_LINK_TEMPLATE, '\n', embed_filename=True):
2158 return
2159
2160 # Download
2161 must_record_download_archive = False
2162 if not self.params.get('skip_download', False):
2163 try:
2164
2165 def existing_file(filename, temp_filename):
2166 file_exists = os.path.exists(encodeFilename(filename))
2167 tempfile_exists = (
2168 False if temp_filename == filename
2169 else os.path.exists(encodeFilename(temp_filename)))
2170 if not self.params.get('overwrites', False) and (file_exists or tempfile_exists):
2171 existing_filename = temp_filename if tempfile_exists else filename
2172 self.to_screen('[download] %s has already been downloaded and merged' % existing_filename)
2173 return existing_filename
2174 if tempfile_exists:
2175 self.report_file_delete(temp_filename)
2176 os.remove(encodeFilename(temp_filename))
2177 if file_exists:
2178 self.report_file_delete(filename)
2179 os.remove(encodeFilename(filename))
2180 return None
2181
2182 success = True
2183 if info_dict.get('requested_formats') is not None:
2184 downloaded = []
2185 merger = FFmpegMergerPP(self)
2186 if not merger.available:
2187 postprocessors = []
2188 self.report_warning('You have requested multiple '
2189 'formats but ffmpeg or avconv are not installed.'
2190 ' The formats won\'t be merged.')
2191 else:
2192 postprocessors = [merger]
2193
2194 def compatible_formats(formats):
2195 # TODO: some formats actually allow this (mkv, webm, ogg, mp4), but not all of them.
2196 video_formats = [format for format in formats if format.get('vcodec') != 'none']
2197 audio_formats = [format for format in formats if format.get('acodec') != 'none']
2198 if len(video_formats) > 2 or len(audio_formats) > 2:
2199 return False
2200
2201 # Check extension
2202 exts = set(format.get('ext') for format in formats)
2203 COMPATIBLE_EXTS = (
2204 set(('mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'ismv', 'isma')),
2205 set(('webm',)),
2206 )
2207 for ext_sets in COMPATIBLE_EXTS:
2208 if ext_sets.issuperset(exts):
2209 return True
2210 # TODO: Check acodec/vcodec
2211 return False
2212
2213 requested_formats = info_dict['requested_formats']
2214 old_ext = info_dict['ext']
2215 if self.params.get('merge_output_format') is None and not compatible_formats(requested_formats):
2216 info_dict['ext'] = 'mkv'
2217 self.report_warning(
2218 'Requested formats are incompatible for merge and will be merged into mkv.')
2219
2220 def correct_ext(filename):
2221 filename_real_ext = os.path.splitext(filename)[1][1:]
2222 filename_wo_ext = (
2223 os.path.splitext(filename)[0]
2224 if filename_real_ext == old_ext
2225 else filename)
2226 return '%s.%s' % (filename_wo_ext, info_dict['ext'])
2227
2228 # Ensure filename always has a correct extension for successful merge
2229 full_filename = correct_ext(full_filename)
2230 temp_filename = correct_ext(temp_filename)
2231 dl_filename = existing_file(full_filename, temp_filename)
2232 if dl_filename is None:
2233 for f in requested_formats:
2234 new_info = dict(info_dict)
2235 new_info.update(f)
2236 fname = prepend_extension(
2237 self.prepare_filepath(self.prepare_filename(new_info), 'temp'),
2238 'f%s' % f['format_id'], new_info['ext'])
2239 if not ensure_dir_exists(fname):
2240 return
2241 downloaded.append(fname)
2242 partial_success, real_download = dl(fname, new_info)
2243 success = success and partial_success
2244 info_dict['__postprocessors'] = postprocessors
2245 info_dict['__files_to_merge'] = downloaded
2246 # Even if there were no downloads, it is being merged only now
2247 info_dict['__real_download'] = True
2248 else:
2249 # Just a single file
2250 dl_filename = existing_file(full_filename, temp_filename)
2251 if dl_filename is None:
2252 success, real_download = dl(temp_filename, info_dict)
2253 info_dict['__real_download'] = real_download
2254
2255 # info_dict['__temp_filename'] = temp_filename
2256 dl_filename = dl_filename or temp_filename
2257 info_dict['__dl_filename'] = dl_filename
2258 info_dict['__final_filename'] = full_filename
2259
2260 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
2261 self.report_error('unable to download video data: %s' % error_to_compat_str(err))
2262 return
2263 except (OSError, IOError) as err:
2264 raise UnavailableVideoError(err)
2265 except (ContentTooShortError, ) as err:
2266 self.report_error('content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded))
2267 return
2268
2269 if success and filename != '-':
2270 # Fixup content
2271 fixup_policy = self.params.get('fixup')
2272 if fixup_policy is None:
2273 fixup_policy = 'detect_or_warn'
2274
2275 INSTALL_FFMPEG_MESSAGE = 'Install ffmpeg or avconv to fix this automatically.'
2276
2277 stretched_ratio = info_dict.get('stretched_ratio')
2278 if stretched_ratio is not None and stretched_ratio != 1:
2279 if fixup_policy == 'warn':
2280 self.report_warning('%s: Non-uniform pixel ratio (%s)' % (
2281 info_dict['id'], stretched_ratio))
2282 elif fixup_policy == 'detect_or_warn':
2283 stretched_pp = FFmpegFixupStretchedPP(self)
2284 if stretched_pp.available:
2285 info_dict['__postprocessors'].append(stretched_pp)
2286 else:
2287 self.report_warning(
2288 '%s: Non-uniform pixel ratio (%s). %s'
2289 % (info_dict['id'], stretched_ratio, INSTALL_FFMPEG_MESSAGE))
2290 else:
2291 assert fixup_policy in ('ignore', 'never')
2292
2293 if (info_dict.get('requested_formats') is None
2294 and info_dict.get('container') == 'm4a_dash'):
2295 if fixup_policy == 'warn':
2296 self.report_warning(
2297 '%s: writing DASH m4a. '
2298 'Only some players support this container.'
2299 % info_dict['id'])
2300 elif fixup_policy == 'detect_or_warn':
2301 fixup_pp = FFmpegFixupM4aPP(self)
2302 if fixup_pp.available:
2303 info_dict['__postprocessors'].append(fixup_pp)
2304 else:
2305 self.report_warning(
2306 '%s: writing DASH m4a. '
2307 'Only some players support this container. %s'
2308 % (info_dict['id'], INSTALL_FFMPEG_MESSAGE))
2309 else:
2310 assert fixup_policy in ('ignore', 'never')
2311
2312 if (info_dict.get('protocol') == 'm3u8_native'
2313 or info_dict.get('protocol') == 'm3u8'
2314 and self.params.get('hls_prefer_native')):
2315 if fixup_policy == 'warn':
2316 self.report_warning('%s: malformed AAC bitstream detected.' % (
2317 info_dict['id']))
2318 elif fixup_policy == 'detect_or_warn':
2319 fixup_pp = FFmpegFixupM3u8PP(self)
2320 if fixup_pp.available:
2321 info_dict['__postprocessors'].append(fixup_pp)
2322 else:
2323 self.report_warning(
2324 '%s: malformed AAC bitstream detected. %s'
2325 % (info_dict['id'], INSTALL_FFMPEG_MESSAGE))
2326 else:
2327 assert fixup_policy in ('ignore', 'never')
2328
2329 try:
2330 self.post_process(dl_filename, info_dict, files_to_move)
2331 except (PostProcessingError) as err:
2332 self.report_error('postprocessing: %s' % str(err))
2333 return
2334 try:
2335 for ph in self._post_hooks:
2336 ph(full_filename)
2337 except Exception as err:
2338 self.report_error('post hooks: %s' % str(err))
2339 return
2340 must_record_download_archive = True
2341
2342 if must_record_download_archive or self.params.get('force_write_download_archive', False):
2343 self.record_download_archive(info_dict)
2344 max_downloads = self.params.get('max_downloads')
2345 if max_downloads is not None and self._num_downloads >= int(max_downloads):
2346 raise MaxDownloadsReached()
2347
2348 def download(self, url_list):
2349 """Download a given list of URLs."""
2350 outtmpl = self.params.get('outtmpl', DEFAULT_OUTTMPL)
2351 if (len(url_list) > 1
2352 and outtmpl != '-'
2353 and '%' not in outtmpl
2354 and self.params.get('max_downloads') != 1):
2355 raise SameFileError(outtmpl)
2356
2357 for url in url_list:
2358 try:
2359 # It also downloads the videos
2360 res = self.extract_info(
2361 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
2362 except UnavailableVideoError:
2363 self.report_error('unable to download video')
2364 except MaxDownloadsReached:
2365 self.to_screen('[info] Maximum number of downloaded files reached')
2366 raise
2367 except ExistingVideoReached:
2368 self.to_screen('[info] Encountered a file that is already in the archive, stopping due to --break-on-existing')
2369 raise
2370 except RejectedVideoReached:
2371 self.to_screen('[info] Encountered a file that did not match filter, stopping due to --break-on-reject')
2372 raise
2373 else:
2374 if self.params.get('dump_single_json', False):
2375 self.to_stdout(json.dumps(res))
2376
2377 return self._download_retcode
2378
2379 def download_with_info_file(self, info_filename):
2380 with contextlib.closing(fileinput.FileInput(
2381 [info_filename], mode='r',
2382 openhook=fileinput.hook_encoded('utf-8'))) as f:
2383 # FileInput doesn't have a read method, we can't call json.load
2384 info = self.filter_requested_info(json.loads('\n'.join(f)))
2385 try:
2386 self.process_ie_result(info, download=True)
2387 except DownloadError:
2388 webpage_url = info.get('webpage_url')
2389 if webpage_url is not None:
2390 self.report_warning('The info failed to download, trying with "%s"' % webpage_url)
2391 return self.download([webpage_url])
2392 else:
2393 raise
2394 return self._download_retcode
2395
2396 @staticmethod
2397 def filter_requested_info(info_dict):
2398 return dict(
2399 (k, v) for k, v in info_dict.items()
2400 if k not in ['requested_formats', 'requested_subtitles'])
2401
2402 def post_process(self, filename, ie_info, files_to_move={}):
2403 """Run all the postprocessors on the given file."""
2404 info = dict(ie_info)
2405 info['filepath'] = filename
2406
2407 def run_pp(pp):
2408 files_to_delete = []
2409 infodict = info
2410 try:
2411 files_to_delete, infodict = pp.run(infodict)
2412 except PostProcessingError as e:
2413 self.report_error(e.msg)
2414 if not files_to_delete:
2415 return infodict
2416
2417 if self.params.get('keepvideo', False):
2418 for f in files_to_delete:
2419 files_to_move.setdefault(f, '')
2420 else:
2421 for old_filename in set(files_to_delete):
2422 self.to_screen('Deleting original file %s (pass -k to keep)' % old_filename)
2423 try:
2424 os.remove(encodeFilename(old_filename))
2425 except (IOError, OSError):
2426 self.report_warning('Unable to remove downloaded original file')
2427 if old_filename in files_to_move:
2428 del files_to_move[old_filename]
2429 return infodict
2430
2431 for pp in ie_info.get('__postprocessors', []) + self._pps:
2432 info = run_pp(pp)
2433 info = run_pp(MoveFilesAfterDownloadPP(self, files_to_move))
2434 files_to_move = {}
2435 for pp in self._pps_end:
2436 info = run_pp(pp)
2437
2438 def _make_archive_id(self, info_dict):
2439 video_id = info_dict.get('id')
2440 if not video_id:
2441 return
2442 # Future-proof against any change in case
2443 # and backwards compatibility with prior versions
2444 extractor = info_dict.get('extractor_key') or info_dict.get('ie_key') # key in a playlist
2445 if extractor is None:
2446 url = str_or_none(info_dict.get('url'))
2447 if not url:
2448 return
2449 # Try to find matching extractor for the URL and take its ie_key
2450 for ie in self._ies:
2451 if ie.suitable(url):
2452 extractor = ie.ie_key()
2453 break
2454 else:
2455 return
2456 return '%s %s' % (extractor.lower(), video_id)
2457
2458 def in_download_archive(self, info_dict):
2459 fn = self.params.get('download_archive')
2460 if fn is None:
2461 return False
2462
2463 vid_id = self._make_archive_id(info_dict)
2464 if not vid_id:
2465 return False # Incomplete video information
2466
2467 return vid_id in self.archive
2468
2469 def record_download_archive(self, info_dict):
2470 fn = self.params.get('download_archive')
2471 if fn is None:
2472 return
2473 vid_id = self._make_archive_id(info_dict)
2474 assert vid_id
2475 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
2476 archive_file.write(vid_id + '\n')
2477 self.archive.add(vid_id)
2478
2479 @staticmethod
2480 def format_resolution(format, default='unknown'):
2481 if format.get('vcodec') == 'none':
2482 return 'audio only'
2483 if format.get('resolution') is not None:
2484 return format['resolution']
2485 if format.get('height') is not None:
2486 if format.get('width') is not None:
2487 res = '%sx%s' % (format['width'], format['height'])
2488 else:
2489 res = '%sp' % format['height']
2490 elif format.get('width') is not None:
2491 res = '%dx?' % format['width']
2492 else:
2493 res = default
2494 return res
2495
2496 def _format_note(self, fdict):
2497 res = ''
2498 if fdict.get('ext') in ['f4f', 'f4m']:
2499 res += '(unsupported) '
2500 if fdict.get('language'):
2501 if res:
2502 res += ' '
2503 res += '[%s] ' % fdict['language']
2504 if fdict.get('format_note') is not None:
2505 res += fdict['format_note'] + ' '
2506 if fdict.get('tbr') is not None:
2507 res += '%4dk ' % fdict['tbr']
2508 if fdict.get('container') is not None:
2509 if res:
2510 res += ', '
2511 res += '%s container' % fdict['container']
2512 if (fdict.get('vcodec') is not None
2513 and fdict.get('vcodec') != 'none'):
2514 if res:
2515 res += ', '
2516 res += fdict['vcodec']
2517 if fdict.get('vbr') is not None:
2518 res += '@'
2519 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
2520 res += 'video@'
2521 if fdict.get('vbr') is not None:
2522 res += '%4dk' % fdict['vbr']
2523 if fdict.get('fps') is not None:
2524 if res:
2525 res += ', '
2526 res += '%sfps' % fdict['fps']
2527 if fdict.get('acodec') is not None:
2528 if res:
2529 res += ', '
2530 if fdict['acodec'] == 'none':
2531 res += 'video only'
2532 else:
2533 res += '%-5s' % fdict['acodec']
2534 elif fdict.get('abr') is not None:
2535 if res:
2536 res += ', '
2537 res += 'audio'
2538 if fdict.get('abr') is not None:
2539 res += '@%3dk' % fdict['abr']
2540 if fdict.get('asr') is not None:
2541 res += ' (%5dHz)' % fdict['asr']
2542 if fdict.get('filesize') is not None:
2543 if res:
2544 res += ', '
2545 res += format_bytes(fdict['filesize'])
2546 elif fdict.get('filesize_approx') is not None:
2547 if res:
2548 res += ', '
2549 res += '~' + format_bytes(fdict['filesize_approx'])
2550 return res
2551
2552 def _format_note_table(self, f):
2553 def join_fields(*vargs):
2554 return ', '.join((val for val in vargs if val != ''))
2555
2556 return join_fields(
2557 'UNSUPPORTED' if f.get('ext') in ('f4f', 'f4m') else '',
2558 format_field(f, 'language', '[%s]'),
2559 format_field(f, 'format_note'),
2560 format_field(f, 'container', ignore=(None, f.get('ext'))),
2561 format_field(f, 'asr', '%5dHz'))
2562
2563 def list_formats(self, info_dict):
2564 formats = info_dict.get('formats', [info_dict])
2565 new_format = self.params.get('listformats_table', False)
2566 if new_format:
2567 table = [
2568 [
2569 format_field(f, 'format_id'),
2570 format_field(f, 'ext'),
2571 self.format_resolution(f),
2572 format_field(f, 'fps', '%d'),
2573 '|',
2574 format_field(f, 'filesize', ' %s', func=format_bytes) + format_field(f, 'filesize_approx', '~%s', func=format_bytes),
2575 format_field(f, 'tbr', '%4dk'),
2576 f.get('protocol').replace('http_dash_segments', 'dash').replace("native", "n"),
2577 '|',
2578 format_field(f, 'vcodec', default='unknown').replace('none', ''),
2579 format_field(f, 'vbr', '%4dk'),
2580 format_field(f, 'acodec', default='unknown').replace('none', ''),
2581 format_field(f, 'abr', '%3dk'),
2582 format_field(f, 'asr', '%5dHz'),
2583 self._format_note_table(f)]
2584 for f in formats
2585 if f.get('preference') is None or f['preference'] >= -1000]
2586 header_line = ['ID', 'EXT', 'RESOLUTION', 'FPS', '|', ' FILESIZE', ' TBR', 'PROTO',
2587 '|', 'VCODEC', ' VBR', 'ACODEC', ' ABR', ' ASR', 'NOTE']
2588 else:
2589 table = [
2590 [
2591 format_field(f, 'format_id'),
2592 format_field(f, 'ext'),
2593 self.format_resolution(f),
2594 self._format_note(f)]
2595 for f in formats
2596 if f.get('preference') is None or f['preference'] >= -1000]
2597 header_line = ['format code', 'extension', 'resolution', 'note']
2598
2599 # if len(formats) > 1:
2600 # table[-1][-1] += (' ' if table[-1][-1] else '') + '(best)'
2601 self.to_screen(
2602 '[info] Available formats for %s:\n%s' % (info_dict['id'], render_table(
2603 header_line,
2604 table,
2605 delim=new_format,
2606 extraGap=(0 if new_format else 1),
2607 hideEmpty=new_format)))
2608
2609 def list_thumbnails(self, info_dict):
2610 thumbnails = info_dict.get('thumbnails')
2611 if not thumbnails:
2612 self.to_screen('[info] No thumbnails present for %s' % info_dict['id'])
2613 return
2614
2615 self.to_screen(
2616 '[info] Thumbnails for %s:' % info_dict['id'])
2617 self.to_screen(render_table(
2618 ['ID', 'width', 'height', 'URL'],
2619 [[t['id'], t.get('width', 'unknown'), t.get('height', 'unknown'), t['url']] for t in thumbnails]))
2620
2621 def list_subtitles(self, video_id, subtitles, name='subtitles'):
2622 if not subtitles:
2623 self.to_screen('%s has no %s' % (video_id, name))
2624 return
2625 self.to_screen(
2626 'Available %s for %s:' % (name, video_id))
2627 self.to_screen(render_table(
2628 ['Language', 'formats'],
2629 [[lang, ', '.join(f['ext'] for f in reversed(formats))]
2630 for lang, formats in subtitles.items()]))
2631
2632 def urlopen(self, req):
2633 """ Start an HTTP download """
2634 if isinstance(req, compat_basestring):
2635 req = sanitized_Request(req)
2636 return self._opener.open(req, timeout=self._socket_timeout)
2637
2638 def print_debug_header(self):
2639 if not self.params.get('verbose'):
2640 return
2641
2642 if type('') is not compat_str:
2643 # Python 2.6 on SLES11 SP1 (https://github.com/ytdl-org/youtube-dl/issues/3326)
2644 self.report_warning(
2645 'Your Python is broken! Update to a newer and supported version')
2646
2647 stdout_encoding = getattr(
2648 sys.stdout, 'encoding', 'missing (%s)' % type(sys.stdout).__name__)
2649 encoding_str = (
2650 '[debug] Encodings: locale %s, fs %s, out %s, pref %s\n' % (
2651 locale.getpreferredencoding(),
2652 sys.getfilesystemencoding(),
2653 stdout_encoding,
2654 self.get_encoding()))
2655 write_string(encoding_str, encoding=None)
2656
2657 self._write_string('[debug] yt-dlp version ' + __version__ + '\n')
2658 if _LAZY_LOADER:
2659 self._write_string('[debug] Lazy loading extractors enabled' + '\n')
2660 try:
2661 sp = subprocess.Popen(
2662 ['git', 'rev-parse', '--short', 'HEAD'],
2663 stdout=subprocess.PIPE, stderr=subprocess.PIPE,
2664 cwd=os.path.dirname(os.path.abspath(__file__)))
2665 out, err = process_communicate_or_kill(sp)
2666 out = out.decode().strip()
2667 if re.match('[0-9a-f]+', out):
2668 self._write_string('[debug] Git HEAD: ' + out + '\n')
2669 except Exception:
2670 try:
2671 sys.exc_clear()
2672 except Exception:
2673 pass
2674
2675 def python_implementation():
2676 impl_name = platform.python_implementation()
2677 if impl_name == 'PyPy' and hasattr(sys, 'pypy_version_info'):
2678 return impl_name + ' version %d.%d.%d' % sys.pypy_version_info[:3]
2679 return impl_name
2680
2681 self._write_string('[debug] Python version %s (%s) - %s\n' % (
2682 platform.python_version(), python_implementation(),
2683 platform_name()))
2684
2685 exe_versions = FFmpegPostProcessor.get_versions(self)
2686 exe_versions['rtmpdump'] = rtmpdump_version()
2687 exe_versions['phantomjs'] = PhantomJSwrapper._version()
2688 exe_str = ', '.join(
2689 '%s %s' % (exe, v)
2690 for exe, v in sorted(exe_versions.items())
2691 if v
2692 )
2693 if not exe_str:
2694 exe_str = 'none'
2695 self._write_string('[debug] exe versions: %s\n' % exe_str)
2696
2697 proxy_map = {}
2698 for handler in self._opener.handlers:
2699 if hasattr(handler, 'proxies'):
2700 proxy_map.update(handler.proxies)
2701 self._write_string('[debug] Proxy map: ' + compat_str(proxy_map) + '\n')
2702
2703 if self.params.get('call_home', False):
2704 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode('utf-8')
2705 self._write_string('[debug] Public IP address: %s\n' % ipaddr)
2706 return
2707 latest_version = self.urlopen(
2708 'https://yt-dl.org/latest/version').read().decode('utf-8')
2709 if version_tuple(latest_version) > version_tuple(__version__):
2710 self.report_warning(
2711 'You are using an outdated version (newest version: %s)! '
2712 'See https://yt-dl.org/update if you need help updating.' %
2713 latest_version)
2714
2715 def _setup_opener(self):
2716 timeout_val = self.params.get('socket_timeout')
2717 self._socket_timeout = 600 if timeout_val is None else float(timeout_val)
2718
2719 opts_cookiefile = self.params.get('cookiefile')
2720 opts_proxy = self.params.get('proxy')
2721
2722 if opts_cookiefile is None:
2723 self.cookiejar = compat_cookiejar.CookieJar()
2724 else:
2725 opts_cookiefile = expand_path(opts_cookiefile)
2726 self.cookiejar = YoutubeDLCookieJar(opts_cookiefile)
2727 if os.access(opts_cookiefile, os.R_OK):
2728 self.cookiejar.load(ignore_discard=True, ignore_expires=True)
2729
2730 cookie_processor = YoutubeDLCookieProcessor(self.cookiejar)
2731 if opts_proxy is not None:
2732 if opts_proxy == '':
2733 proxies = {}
2734 else:
2735 proxies = {'http': opts_proxy, 'https': opts_proxy}
2736 else:
2737 proxies = compat_urllib_request.getproxies()
2738 # Set HTTPS proxy to HTTP one if given (https://github.com/ytdl-org/youtube-dl/issues/805)
2739 if 'http' in proxies and 'https' not in proxies:
2740 proxies['https'] = proxies['http']
2741 proxy_handler = PerRequestProxyHandler(proxies)
2742
2743 debuglevel = 1 if self.params.get('debug_printtraffic') else 0
2744 https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
2745 ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
2746 redirect_handler = YoutubeDLRedirectHandler()
2747 data_handler = compat_urllib_request_DataHandler()
2748
2749 # When passing our own FileHandler instance, build_opener won't add the
2750 # default FileHandler and allows us to disable the file protocol, which
2751 # can be used for malicious purposes (see
2752 # https://github.com/ytdl-org/youtube-dl/issues/8227)
2753 file_handler = compat_urllib_request.FileHandler()
2754
2755 def file_open(*args, **kwargs):
2756 raise compat_urllib_error.URLError('file:// scheme is explicitly disabled in youtube-dlc for security reasons')
2757 file_handler.file_open = file_open
2758
2759 opener = compat_urllib_request.build_opener(
2760 proxy_handler, https_handler, cookie_processor, ydlh, redirect_handler, data_handler, file_handler)
2761
2762 # Delete the default user-agent header, which would otherwise apply in
2763 # cases where our custom HTTP handler doesn't come into play
2764 # (See https://github.com/ytdl-org/youtube-dl/issues/1309 for details)
2765 opener.addheaders = []
2766 self._opener = opener
2767
2768 def encode(self, s):
2769 if isinstance(s, bytes):
2770 return s # Already encoded
2771
2772 try:
2773 return s.encode(self.get_encoding())
2774 except UnicodeEncodeError as err:
2775 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
2776 raise
2777
2778 def get_encoding(self):
2779 encoding = self.params.get('encoding')
2780 if encoding is None:
2781 encoding = preferredencoding()
2782 return encoding
2783
2784 def _write_thumbnails(self, info_dict, filename):
2785 if self.params.get('writethumbnail', False):
2786 thumbnails = info_dict.get('thumbnails')
2787 if thumbnails:
2788 thumbnails = [thumbnails[-1]]
2789 elif self.params.get('write_all_thumbnails', False):
2790 thumbnails = info_dict.get('thumbnails') or []
2791 else:
2792 thumbnails = []
2793
2794 ret = []
2795 for t in thumbnails:
2796 thumb_ext = determine_ext(t['url'], 'jpg')
2797 suffix = '_%s' % t['id'] if len(thumbnails) > 1 else ''
2798 thumb_display_id = '%s ' % t['id'] if len(thumbnails) > 1 else ''
2799 t['filename'] = thumb_filename = replace_extension(filename + suffix, thumb_ext, info_dict.get('ext'))
2800
2801 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(thumb_filename)):
2802 ret.append(thumb_filename)
2803 self.to_screen('[%s] %s: Thumbnail %sis already present' %
2804 (info_dict['extractor'], info_dict['id'], thumb_display_id))
2805 else:
2806 self.to_screen('[%s] %s: Downloading thumbnail %s...' %
2807 (info_dict['extractor'], info_dict['id'], thumb_display_id))
2808 try:
2809 uf = self.urlopen(t['url'])
2810 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
2811 shutil.copyfileobj(uf, thumbf)
2812 ret.append(thumb_filename)
2813 self.to_screen('[%s] %s: Writing thumbnail %sto: %s' %
2814 (info_dict['extractor'], info_dict['id'], thumb_display_id, thumb_filename))
2815 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
2816 self.report_warning('Unable to download thumbnail "%s": %s' %
2817 (t['url'], error_to_compat_str(err)))
2818 return ret