]> jfr.im git - yt-dlp.git/blob - yt_dlp/YoutubeDL.py
8878d710f4622cc3056e1bb39689add955a550a4
[yt-dlp.git] / yt_dlp / YoutubeDL.py
1 #!/usr/bin/env python3
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 functools
13 import io
14 import itertools
15 import json
16 import locale
17 import operator
18 import os
19 import platform
20 import re
21 import shutil
22 import subprocess
23 import sys
24 import tempfile
25 import time
26 import tokenize
27 import traceback
28 import random
29 import unicodedata
30
31 from string import ascii_letters
32
33 from .compat import (
34 compat_basestring,
35 compat_get_terminal_size,
36 compat_kwargs,
37 compat_numeric_types,
38 compat_os_name,
39 compat_pycrypto_AES,
40 compat_shlex_quote,
41 compat_str,
42 compat_tokenize_tokenize,
43 compat_urllib_error,
44 compat_urllib_request,
45 compat_urllib_request_DataHandler,
46 windows_enable_vt_mode,
47 )
48 from .cookies import load_cookies
49 from .utils import (
50 age_restricted,
51 args_to_str,
52 ContentTooShortError,
53 date_from_str,
54 DateRange,
55 DEFAULT_OUTTMPL,
56 determine_ext,
57 determine_protocol,
58 DOT_DESKTOP_LINK_TEMPLATE,
59 DOT_URL_LINK_TEMPLATE,
60 DOT_WEBLOC_LINK_TEMPLATE,
61 DownloadError,
62 encode_compat_str,
63 encodeFilename,
64 EntryNotInPlaylist,
65 error_to_compat_str,
66 ExistingVideoReached,
67 expand_path,
68 ExtractorError,
69 float_or_none,
70 format_bytes,
71 format_field,
72 formatSeconds,
73 GeoRestrictedError,
74 HEADRequest,
75 int_or_none,
76 iri_to_uri,
77 ISO3166Utils,
78 LazyList,
79 locked_file,
80 make_dir,
81 make_HTTPS_handler,
82 MaxDownloadsReached,
83 network_exceptions,
84 orderedSet,
85 OUTTMPL_TYPES,
86 PagedList,
87 parse_filesize,
88 PerRequestProxyHandler,
89 platform_name,
90 PostProcessingError,
91 preferredencoding,
92 prepend_extension,
93 process_communicate_or_kill,
94 register_socks_protocols,
95 RejectedVideoReached,
96 render_table,
97 replace_extension,
98 SameFileError,
99 sanitize_filename,
100 sanitize_path,
101 sanitize_url,
102 sanitized_Request,
103 std_headers,
104 STR_FORMAT_RE_TMPL,
105 STR_FORMAT_TYPES,
106 str_or_none,
107 strftime_or_none,
108 subtitles_filename,
109 supports_terminal_sequences,
110 TERMINAL_SEQUENCES,
111 ThrottledDownload,
112 to_high_limit_path,
113 traverse_obj,
114 try_get,
115 UnavailableVideoError,
116 url_basename,
117 variadic,
118 version_tuple,
119 write_json_file,
120 write_string,
121 YoutubeDLCookieProcessor,
122 YoutubeDLHandler,
123 YoutubeDLRedirectHandler,
124 )
125 from .cache import Cache
126 from .extractor import (
127 gen_extractor_classes,
128 get_info_extractor,
129 _LAZY_LOADER,
130 _PLUGIN_CLASSES as plugin_extractors
131 )
132 from .extractor.openload import PhantomJSwrapper
133 from .downloader import (
134 FFmpegFD,
135 get_suitable_downloader,
136 shorten_protocol_name
137 )
138 from .downloader.rtmp import rtmpdump_version
139 from .postprocessor import (
140 get_postprocessor,
141 EmbedThumbnailPP,
142 FFmpegFixupDurationPP,
143 FFmpegFixupM3u8PP,
144 FFmpegFixupM4aPP,
145 FFmpegFixupStretchedPP,
146 FFmpegFixupTimestampPP,
147 FFmpegMergerPP,
148 FFmpegPostProcessor,
149 MoveFilesAfterDownloadPP,
150 _PLUGIN_CLASSES as plugin_postprocessors
151 )
152 from .update import detect_variant
153 from .version import __version__
154
155 if compat_os_name == 'nt':
156 import ctypes
157
158
159 class YoutubeDL(object):
160 """YoutubeDL class.
161
162 YoutubeDL objects are the ones responsible of downloading the
163 actual video file and writing it to disk if the user has requested
164 it, among some other tasks. In most cases there should be one per
165 program. As, given a video URL, the downloader doesn't know how to
166 extract all the needed information, task that InfoExtractors do, it
167 has to pass the URL to one of them.
168
169 For this, YoutubeDL objects have a method that allows
170 InfoExtractors to be registered in a given order. When it is passed
171 a URL, the YoutubeDL object handles it to the first InfoExtractor it
172 finds that reports being able to handle it. The InfoExtractor extracts
173 all the information about the video or videos the URL refers to, and
174 YoutubeDL process the extracted information, possibly using a File
175 Downloader to download the video.
176
177 YoutubeDL objects accept a lot of parameters. In order not to saturate
178 the object constructor with arguments, it receives a dictionary of
179 options instead. These options are available through the params
180 attribute for the InfoExtractors to use. The YoutubeDL also
181 registers itself as the downloader in charge for the InfoExtractors
182 that are added to it, so this is a "mutual registration".
183
184 Available options:
185
186 username: Username for authentication purposes.
187 password: Password for authentication purposes.
188 videopassword: Password for accessing a video.
189 ap_mso: Adobe Pass multiple-system operator identifier.
190 ap_username: Multiple-system operator account username.
191 ap_password: Multiple-system operator account password.
192 usenetrc: Use netrc for authentication instead.
193 verbose: Print additional info to stdout.
194 quiet: Do not print messages to stdout.
195 no_warnings: Do not print out anything for warnings.
196 forceprint: A list of templates to force print
197 forceurl: Force printing final URL. (Deprecated)
198 forcetitle: Force printing title. (Deprecated)
199 forceid: Force printing ID. (Deprecated)
200 forcethumbnail: Force printing thumbnail URL. (Deprecated)
201 forcedescription: Force printing description. (Deprecated)
202 forcefilename: Force printing final filename. (Deprecated)
203 forceduration: Force printing duration. (Deprecated)
204 forcejson: Force printing info_dict as JSON.
205 dump_single_json: Force printing the info_dict of the whole playlist
206 (or video) as a single JSON line.
207 force_write_download_archive: Force writing download archive regardless
208 of 'skip_download' or 'simulate'.
209 simulate: Do not download the video files. If unset (or None),
210 simulate only if listsubtitles, listformats or list_thumbnails is used
211 format: Video format code. see "FORMAT SELECTION" for more details.
212 allow_unplayable_formats: Allow unplayable formats to be extracted and downloaded.
213 ignore_no_formats_error: Ignore "No video formats" error. Usefull for
214 extracting metadata even if the video is not actually
215 available for download (experimental)
216 format_sort: How to sort the video formats. see "Sorting Formats"
217 for more details.
218 format_sort_force: Force the given format_sort. see "Sorting Formats"
219 for more details.
220 allow_multiple_video_streams: Allow multiple video streams to be merged
221 into a single file
222 allow_multiple_audio_streams: Allow multiple audio streams to be merged
223 into a single file
224 check_formats Whether to test if the formats are downloadable.
225 Can be True (check all), False (check none)
226 or None (check only if requested by extractor)
227 paths: Dictionary of output paths. The allowed keys are 'home'
228 'temp' and the keys of OUTTMPL_TYPES (in utils.py)
229 outtmpl: Dictionary of templates for output names. Allowed keys
230 are 'default' and the keys of OUTTMPL_TYPES (in utils.py).
231 For compatibility with youtube-dl, a single string can also be used
232 outtmpl_na_placeholder: Placeholder for unavailable meta fields.
233 restrictfilenames: Do not allow "&" and spaces in file names
234 trim_file_name: Limit length of filename (extension excluded)
235 windowsfilenames: Force the filenames to be windows compatible
236 ignoreerrors: Do not stop on download/postprocessing errors.
237 Can be 'only_download' to ignore only download errors.
238 Default is 'only_download' for CLI, but False for API
239 skip_playlist_after_errors: Number of allowed failures until the rest of
240 the playlist is skipped
241 force_generic_extractor: Force downloader to use the generic extractor
242 overwrites: Overwrite all video and metadata files if True,
243 overwrite only non-video files if None
244 and don't overwrite any file if False
245 For compatibility with youtube-dl,
246 "nooverwrites" may also be used instead
247 playliststart: Playlist item to start at.
248 playlistend: Playlist item to end at.
249 playlist_items: Specific indices of playlist to download.
250 playlistreverse: Download playlist items in reverse order.
251 playlistrandom: Download playlist items in random order.
252 matchtitle: Download only matching titles.
253 rejecttitle: Reject downloads for matching titles.
254 logger: Log messages to a logging.Logger instance.
255 logtostderr: Log messages to stderr instead of stdout.
256 consoletitle: Display progress in console window's titlebar.
257 writedescription: Write the video description to a .description file
258 writeinfojson: Write the video description to a .info.json file
259 clean_infojson: Remove private fields from the infojson
260 getcomments: Extract video comments. This will not be written to disk
261 unless writeinfojson is also given
262 writeannotations: Write the video annotations to a .annotations.xml file
263 writethumbnail: Write the thumbnail image to a file
264 allow_playlist_files: Whether to write playlists' description, infojson etc
265 also to disk when using the 'write*' options
266 write_all_thumbnails: Write all thumbnail formats to files
267 writelink: Write an internet shortcut file, depending on the
268 current platform (.url/.webloc/.desktop)
269 writeurllink: Write a Windows internet shortcut file (.url)
270 writewebloclink: Write a macOS internet shortcut file (.webloc)
271 writedesktoplink: Write a Linux internet shortcut file (.desktop)
272 writesubtitles: Write the video subtitles to a file
273 writeautomaticsub: Write the automatically generated subtitles to a file
274 allsubtitles: Deprecated - Use subtitleslangs = ['all']
275 Downloads all the subtitles of the video
276 (requires writesubtitles or writeautomaticsub)
277 listsubtitles: Lists all available subtitles for the video
278 subtitlesformat: The format code for subtitles
279 subtitleslangs: List of languages of the subtitles to download (can be regex).
280 The list may contain "all" to refer to all the available
281 subtitles. The language can be prefixed with a "-" to
282 exclude it from the requested languages. Eg: ['all', '-live_chat']
283 keepvideo: Keep the video file after post-processing
284 daterange: A DateRange object, download only if the upload_date is in the range.
285 skip_download: Skip the actual download of the video file
286 cachedir: Location of the cache files in the filesystem.
287 False to disable filesystem cache.
288 noplaylist: Download single video instead of a playlist if in doubt.
289 age_limit: An integer representing the user's age in years.
290 Unsuitable videos for the given age are skipped.
291 min_views: An integer representing the minimum view count the video
292 must have in order to not be skipped.
293 Videos without view count information are always
294 downloaded. None for no limit.
295 max_views: An integer representing the maximum view count.
296 Videos that are more popular than that are not
297 downloaded.
298 Videos without view count information are always
299 downloaded. None for no limit.
300 download_archive: File name of a file where all downloads are recorded.
301 Videos already present in the file are not downloaded
302 again.
303 break_on_existing: Stop the download process after attempting to download a
304 file that is in the archive.
305 break_on_reject: Stop the download process when encountering a video that
306 has been filtered out.
307 cookiefile: File name where cookies should be read from and dumped to
308 cookiesfrombrowser: A tuple containing the name of the browser and the profile
309 name/path from where cookies are loaded.
310 Eg: ('chrome', ) or (vivaldi, 'default')
311 nocheckcertificate:Do not verify SSL certificates
312 prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
313 At the moment, this is only supported by YouTube.
314 proxy: URL of the proxy server to use
315 geo_verification_proxy: URL of the proxy to use for IP address verification
316 on geo-restricted sites.
317 socket_timeout: Time to wait for unresponsive hosts, in seconds
318 bidi_workaround: Work around buggy terminals without bidirectional text
319 support, using fridibi
320 debug_printtraffic:Print out sent and received HTTP traffic
321 include_ads: Download ads as well
322 default_search: Prepend this string if an input url is not valid.
323 'auto' for elaborate guessing
324 encoding: Use this encoding instead of the system-specified.
325 extract_flat: Do not resolve URLs, return the immediate result.
326 Pass in 'in_playlist' to only show this behavior for
327 playlist items.
328 postprocessors: A list of dictionaries, each with an entry
329 * key: The name of the postprocessor. See
330 yt_dlp/postprocessor/__init__.py for a list.
331 * when: When to run the postprocessor. Can be one of
332 pre_process|before_dl|post_process|after_move.
333 Assumed to be 'post_process' if not given
334 post_hooks: Deprecated - Register a custom postprocessor instead
335 A list of functions that get called as the final step
336 for each video file, after all postprocessors have been
337 called. The filename will be passed as the only argument.
338 progress_hooks: A list of functions that get called on download
339 progress, with a dictionary with the entries
340 * status: One of "downloading", "error", or "finished".
341 Check this first and ignore unknown values.
342 * info_dict: The extracted info_dict
343
344 If status is one of "downloading", or "finished", the
345 following properties may also be present:
346 * filename: The final filename (always present)
347 * tmpfilename: The filename we're currently writing to
348 * downloaded_bytes: Bytes on disk
349 * total_bytes: Size of the whole file, None if unknown
350 * total_bytes_estimate: Guess of the eventual file size,
351 None if unavailable.
352 * elapsed: The number of seconds since download started.
353 * eta: The estimated time in seconds, None if unknown
354 * speed: The download speed in bytes/second, None if
355 unknown
356 * fragment_index: The counter of the currently
357 downloaded video fragment.
358 * fragment_count: The number of fragments (= individual
359 files that will be merged)
360
361 Progress hooks are guaranteed to be called at least once
362 (with status "finished") if the download is successful.
363 postprocessor_hooks: A list of functions that get called on postprocessing
364 progress, with a dictionary with the entries
365 * status: One of "started", "processing", or "finished".
366 Check this first and ignore unknown values.
367 * postprocessor: Name of the postprocessor
368 * info_dict: The extracted info_dict
369
370 Progress hooks are guaranteed to be called at least twice
371 (with status "started" and "finished") if the processing is successful.
372 merge_output_format: Extension to use when merging formats.
373 final_ext: Expected final extension; used to detect when the file was
374 already downloaded and converted. "merge_output_format" is
375 replaced by this extension when given
376 fixup: Automatically correct known faults of the file.
377 One of:
378 - "never": do nothing
379 - "warn": only emit a warning
380 - "detect_or_warn": check whether we can do anything
381 about it, warn otherwise (default)
382 source_address: Client-side IP address to bind to.
383 call_home: Boolean, true iff we are allowed to contact the
384 yt-dlp servers for debugging. (BROKEN)
385 sleep_interval_requests: Number of seconds to sleep between requests
386 during extraction
387 sleep_interval: Number of seconds to sleep before each download when
388 used alone or a lower bound of a range for randomized
389 sleep before each download (minimum possible number
390 of seconds to sleep) when used along with
391 max_sleep_interval.
392 max_sleep_interval:Upper bound of a range for randomized sleep before each
393 download (maximum possible number of seconds to sleep).
394 Must only be used along with sleep_interval.
395 Actual sleep time will be a random float from range
396 [sleep_interval; max_sleep_interval].
397 sleep_interval_subtitles: Number of seconds to sleep before each subtitle download
398 listformats: Print an overview of available video formats and exit.
399 list_thumbnails: Print a table of all thumbnails and exit.
400 match_filter: A function that gets called with the info_dict of
401 every video.
402 If it returns a message, the video is ignored.
403 If it returns None, the video is downloaded.
404 match_filter_func in utils.py is one example for this.
405 no_color: Do not emit color codes in output.
406 geo_bypass: Bypass geographic restriction via faking X-Forwarded-For
407 HTTP header
408 geo_bypass_country:
409 Two-letter ISO 3166-2 country code that will be used for
410 explicit geographic restriction bypassing via faking
411 X-Forwarded-For HTTP header
412 geo_bypass_ip_block:
413 IP range in CIDR notation that will be used similarly to
414 geo_bypass_country
415
416 The following options determine which downloader is picked:
417 external_downloader: A dictionary of protocol keys and the executable of the
418 external downloader to use for it. The allowed protocols
419 are default|http|ftp|m3u8|dash|rtsp|rtmp|mms.
420 Set the value to 'native' to use the native downloader
421 hls_prefer_native: Deprecated - Use external_downloader = {'m3u8': 'native'}
422 or {'m3u8': 'ffmpeg'} instead.
423 Use the native HLS downloader instead of ffmpeg/avconv
424 if True, otherwise use ffmpeg/avconv if False, otherwise
425 use downloader suggested by extractor if None.
426 compat_opts: Compatibility options. See "Differences in default behavior".
427 The following options do not work when used through the API:
428 filename, abort-on-error, multistreams, no-live-chat, format-sort
429 no-clean-infojson, no-playlist-metafiles, no-keep-subs.
430 Refer __init__.py for their implementation
431 progress_template: Dictionary of templates for progress outputs.
432 Allowed keys are 'download', 'postprocess',
433 'download-title' (console title) and 'postprocess-title'.
434 The template is mapped on a dictionary with keys 'progress' and 'info'
435
436 The following parameters are not used by YoutubeDL itself, they are used by
437 the downloader (see yt_dlp/downloader/common.py):
438 nopart, updatetime, buffersize, ratelimit, throttledratelimit, min_filesize,
439 max_filesize, test, noresizebuffer, retries, fragment_retries, continuedl,
440 noprogress, xattr_set_filesize, hls_use_mpegts, http_chunk_size,
441 external_downloader_args.
442
443 The following options are used by the post processors:
444 prefer_ffmpeg: If False, use avconv instead of ffmpeg if both are available,
445 otherwise prefer ffmpeg. (avconv support is deprecated)
446 ffmpeg_location: Location of the ffmpeg/avconv binary; either the path
447 to the binary or its containing directory.
448 postprocessor_args: A dictionary of postprocessor/executable keys (in lower case)
449 and a list of additional command-line arguments for the
450 postprocessor/executable. The dict can also have "PP+EXE" keys
451 which are used when the given exe is used by the given PP.
452 Use 'default' as the name for arguments to passed to all PP
453 For compatibility with youtube-dl, a single list of args
454 can also be used
455
456 The following options are used by the extractors:
457 extractor_retries: Number of times to retry for known errors
458 dynamic_mpd: Whether to process dynamic DASH manifests (default: True)
459 hls_split_discontinuity: Split HLS playlists to different formats at
460 discontinuities such as ad breaks (default: False)
461 extractor_args: A dictionary of arguments to be passed to the extractors.
462 See "EXTRACTOR ARGUMENTS" for details.
463 Eg: {'youtube': {'skip': ['dash', 'hls']}}
464 youtube_include_dash_manifest: Deprecated - Use extractor_args instead.
465 If True (default), DASH manifests and related
466 data will be downloaded and processed by extractor.
467 You can reduce network I/O by disabling it if you don't
468 care about DASH. (only for youtube)
469 youtube_include_hls_manifest: Deprecated - Use extractor_args instead.
470 If True (default), HLS manifests and related
471 data will be downloaded and processed by extractor.
472 You can reduce network I/O by disabling it if you don't
473 care about HLS. (only for youtube)
474 """
475
476 _NUMERIC_FIELDS = set((
477 'width', 'height', 'tbr', 'abr', 'asr', 'vbr', 'fps', 'filesize', 'filesize_approx',
478 'timestamp', 'release_timestamp',
479 'duration', 'view_count', 'like_count', 'dislike_count', 'repost_count',
480 'average_rating', 'comment_count', 'age_limit',
481 'start_time', 'end_time',
482 'chapter_number', 'season_number', 'episode_number',
483 'track_number', 'disc_number', 'release_year',
484 ))
485
486 params = None
487 _ies = {}
488 _pps = {'pre_process': [], 'before_dl': [], 'after_move': [], 'post_process': []}
489 _printed_messages = set()
490 _first_webpage_request = True
491 _download_retcode = None
492 _num_downloads = None
493 _playlist_level = 0
494 _playlist_urls = set()
495 _screen_file = None
496
497 def __init__(self, params=None, auto_init=True):
498 """Create a FileDownloader object with the given options."""
499 if params is None:
500 params = {}
501 self._ies = {}
502 self._ies_instances = {}
503 self._pps = {'pre_process': [], 'before_dl': [], 'after_move': [], 'post_process': []}
504 self._printed_messages = set()
505 self._first_webpage_request = True
506 self._post_hooks = []
507 self._progress_hooks = []
508 self._postprocessor_hooks = []
509 self._download_retcode = 0
510 self._num_downloads = 0
511 self._screen_file = [sys.stdout, sys.stderr][params.get('logtostderr', False)]
512 self._err_file = sys.stderr
513 self.params = params
514 self.cache = Cache(self)
515
516 windows_enable_vt_mode()
517 # FIXME: This will break if we ever print color to stdout
518 self.params['no_color'] = self.params.get('no_color') or not supports_terminal_sequences(self._err_file)
519
520 if sys.version_info < (3, 6):
521 self.report_warning(
522 'Python version %d.%d is not supported! Please update to Python 3.6 or above' % sys.version_info[:2])
523
524 if self.params.get('allow_unplayable_formats'):
525 self.report_warning(
526 f'You have asked for {self._color_text("unplayable formats", "blue")} to be listed/downloaded. '
527 'This is a developer option intended for debugging. \n'
528 ' If you experience any issues while using this option, '
529 f'{self._color_text("DO NOT", "red")} open a bug report')
530
531 def check_deprecated(param, option, suggestion):
532 if self.params.get(param) is not None:
533 self.report_warning('%s is deprecated. Use %s instead' % (option, suggestion))
534 return True
535 return False
536
537 if check_deprecated('cn_verification_proxy', '--cn-verification-proxy', '--geo-verification-proxy'):
538 if self.params.get('geo_verification_proxy') is None:
539 self.params['geo_verification_proxy'] = self.params['cn_verification_proxy']
540
541 check_deprecated('autonumber', '--auto-number', '-o "%(autonumber)s-%(title)s.%(ext)s"')
542 check_deprecated('usetitle', '--title', '-o "%(title)s-%(id)s.%(ext)s"')
543 check_deprecated('useid', '--id', '-o "%(id)s.%(ext)s"')
544
545 for msg in self.params.get('warnings', []):
546 self.report_warning(msg)
547
548 if 'overwrites' not in self.params and self.params.get('nooverwrites') is not None:
549 # nooverwrites was unnecessarily changed to overwrites
550 # in 0c3d0f51778b153f65c21906031c2e091fcfb641
551 # This ensures compatibility with both keys
552 self.params['overwrites'] = not self.params['nooverwrites']
553 elif self.params.get('overwrites') is None:
554 self.params.pop('overwrites', None)
555 else:
556 self.params['nooverwrites'] = not self.params['overwrites']
557
558 if params.get('bidi_workaround', False):
559 try:
560 import pty
561 master, slave = pty.openpty()
562 width = compat_get_terminal_size().columns
563 if width is None:
564 width_args = []
565 else:
566 width_args = ['-w', str(width)]
567 sp_kwargs = dict(
568 stdin=subprocess.PIPE,
569 stdout=slave,
570 stderr=self._err_file)
571 try:
572 self._output_process = subprocess.Popen(
573 ['bidiv'] + width_args, **sp_kwargs
574 )
575 except OSError:
576 self._output_process = subprocess.Popen(
577 ['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
578 self._output_channel = os.fdopen(master, 'rb')
579 except OSError as ose:
580 if ose.errno == errno.ENOENT:
581 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.')
582 else:
583 raise
584
585 if (sys.platform != 'win32'
586 and sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968']
587 and not params.get('restrictfilenames', False)):
588 # Unicode filesystem API will throw errors (#1474, #13027)
589 self.report_warning(
590 'Assuming --restrict-filenames since file system encoding '
591 'cannot encode all characters. '
592 'Set the LC_ALL environment variable to fix this.')
593 self.params['restrictfilenames'] = True
594
595 self.outtmpl_dict = self.parse_outtmpl()
596
597 # Creating format selector here allows us to catch syntax errors before the extraction
598 self.format_selector = (
599 None if self.params.get('format') is None
600 else self.build_format_selector(self.params['format']))
601
602 self._setup_opener()
603
604 if auto_init:
605 self.print_debug_header()
606 self.add_default_info_extractors()
607
608 for pp_def_raw in self.params.get('postprocessors', []):
609 pp_def = dict(pp_def_raw)
610 when = pp_def.pop('when', 'post_process')
611 pp_class = get_postprocessor(pp_def.pop('key'))
612 pp = pp_class(self, **compat_kwargs(pp_def))
613 self.add_post_processor(pp, when=when)
614
615 for ph in self.params.get('post_hooks', []):
616 self.add_post_hook(ph)
617
618 for ph in self.params.get('progress_hooks', []):
619 self.add_progress_hook(ph)
620
621 register_socks_protocols()
622
623 def preload_download_archive(fn):
624 """Preload the archive, if any is specified"""
625 if fn is None:
626 return False
627 self.write_debug('Loading archive file %r\n' % fn)
628 try:
629 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
630 for line in archive_file:
631 self.archive.add(line.strip())
632 except IOError as ioe:
633 if ioe.errno != errno.ENOENT:
634 raise
635 return False
636 return True
637
638 self.archive = set()
639 preload_download_archive(self.params.get('download_archive'))
640
641 def warn_if_short_id(self, argv):
642 # short YouTube ID starting with dash?
643 idxs = [
644 i for i, a in enumerate(argv)
645 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
646 if idxs:
647 correct_argv = (
648 ['yt-dlp']
649 + [a for i, a in enumerate(argv) if i not in idxs]
650 + ['--'] + [argv[i] for i in idxs]
651 )
652 self.report_warning(
653 'Long argument string detected. '
654 'Use -- to separate parameters and URLs, like this:\n%s\n' %
655 args_to_str(correct_argv))
656
657 def add_info_extractor(self, ie):
658 """Add an InfoExtractor object to the end of the list."""
659 ie_key = ie.ie_key()
660 self._ies[ie_key] = ie
661 if not isinstance(ie, type):
662 self._ies_instances[ie_key] = ie
663 ie.set_downloader(self)
664
665 def _get_info_extractor_class(self, ie_key):
666 ie = self._ies.get(ie_key)
667 if ie is None:
668 ie = get_info_extractor(ie_key)
669 self.add_info_extractor(ie)
670 return ie
671
672 def get_info_extractor(self, ie_key):
673 """
674 Get an instance of an IE with name ie_key, it will try to get one from
675 the _ies list, if there's no instance it will create a new one and add
676 it to the extractor list.
677 """
678 ie = self._ies_instances.get(ie_key)
679 if ie is None:
680 ie = get_info_extractor(ie_key)()
681 self.add_info_extractor(ie)
682 return ie
683
684 def add_default_info_extractors(self):
685 """
686 Add the InfoExtractors returned by gen_extractors to the end of the list
687 """
688 for ie in gen_extractor_classes():
689 self.add_info_extractor(ie)
690
691 def add_post_processor(self, pp, when='post_process'):
692 """Add a PostProcessor object to the end of the chain."""
693 self._pps[when].append(pp)
694 pp.set_downloader(self)
695
696 def add_post_hook(self, ph):
697 """Add the post hook"""
698 self._post_hooks.append(ph)
699
700 def add_progress_hook(self, ph):
701 """Add the download progress hook"""
702 self._progress_hooks.append(ph)
703
704 def add_postprocessor_hook(self, ph):
705 """Add the postprocessing progress hook"""
706 self._postprocessor_hooks.append(ph)
707
708 def _bidi_workaround(self, message):
709 if not hasattr(self, '_output_channel'):
710 return message
711
712 assert hasattr(self, '_output_process')
713 assert isinstance(message, compat_str)
714 line_count = message.count('\n') + 1
715 self._output_process.stdin.write((message + '\n').encode('utf-8'))
716 self._output_process.stdin.flush()
717 res = ''.join(self._output_channel.readline().decode('utf-8')
718 for _ in range(line_count))
719 return res[:-len('\n')]
720
721 def _write_string(self, message, out=None, only_once=False):
722 if only_once:
723 if message in self._printed_messages:
724 return
725 self._printed_messages.add(message)
726 write_string(message, out=out, encoding=self.params.get('encoding'))
727
728 def to_stdout(self, message, skip_eol=False, quiet=False):
729 """Print message to stdout"""
730 if self.params.get('logger'):
731 self.params['logger'].debug(message)
732 elif not quiet or self.params.get('verbose'):
733 self._write_string(
734 '%s%s' % (self._bidi_workaround(message), ('' if skip_eol else '\n')),
735 self._err_file if quiet else self._screen_file)
736
737 def to_stderr(self, message, only_once=False):
738 """Print message to stderr"""
739 assert isinstance(message, compat_str)
740 if self.params.get('logger'):
741 self.params['logger'].error(message)
742 else:
743 self._write_string('%s\n' % self._bidi_workaround(message), self._err_file, only_once=only_once)
744
745 def to_console_title(self, message):
746 if not self.params.get('consoletitle', False):
747 return
748 if compat_os_name == 'nt':
749 if ctypes.windll.kernel32.GetConsoleWindow():
750 # c_wchar_p() might not be necessary if `message` is
751 # already of type unicode()
752 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
753 elif 'TERM' in os.environ:
754 self._write_string('\033]0;%s\007' % message, self._screen_file)
755
756 def save_console_title(self):
757 if not self.params.get('consoletitle', False):
758 return
759 if self.params.get('simulate'):
760 return
761 if compat_os_name != 'nt' and 'TERM' in os.environ:
762 # Save the title on stack
763 self._write_string('\033[22;0t', self._screen_file)
764
765 def restore_console_title(self):
766 if not self.params.get('consoletitle', False):
767 return
768 if self.params.get('simulate'):
769 return
770 if compat_os_name != 'nt' and 'TERM' in os.environ:
771 # Restore the title from stack
772 self._write_string('\033[23;0t', self._screen_file)
773
774 def __enter__(self):
775 self.save_console_title()
776 return self
777
778 def __exit__(self, *args):
779 self.restore_console_title()
780
781 if self.params.get('cookiefile') is not None:
782 self.cookiejar.save(ignore_discard=True, ignore_expires=True)
783
784 def trouble(self, message=None, tb=None):
785 """Determine action to take when a download problem appears.
786
787 Depending on if the downloader has been configured to ignore
788 download errors or not, this method may throw an exception or
789 not when errors are found, after printing the message.
790
791 tb, if given, is additional traceback information.
792 """
793 if message is not None:
794 self.to_stderr(message)
795 if self.params.get('verbose'):
796 if tb is None:
797 if sys.exc_info()[0]: # if .trouble has been called from an except block
798 tb = ''
799 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
800 tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
801 tb += encode_compat_str(traceback.format_exc())
802 else:
803 tb_data = traceback.format_list(traceback.extract_stack())
804 tb = ''.join(tb_data)
805 if tb:
806 self.to_stderr(tb)
807 if not self.params.get('ignoreerrors'):
808 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
809 exc_info = sys.exc_info()[1].exc_info
810 else:
811 exc_info = sys.exc_info()
812 raise DownloadError(message, exc_info)
813 self._download_retcode = 1
814
815 def to_screen(self, message, skip_eol=False):
816 """Print message to stdout if not in quiet mode"""
817 self.to_stdout(
818 message, skip_eol, quiet=self.params.get('quiet', False))
819
820 def _color_text(self, text, color):
821 if self.params.get('no_color'):
822 return text
823 return f'{TERMINAL_SEQUENCES[color.upper()]}{text}{TERMINAL_SEQUENCES["RESET_STYLE"]}'
824
825 def report_warning(self, message, only_once=False):
826 '''
827 Print the message to stderr, it will be prefixed with 'WARNING:'
828 If stderr is a tty file the 'WARNING:' will be colored
829 '''
830 if self.params.get('logger') is not None:
831 self.params['logger'].warning(message)
832 else:
833 if self.params.get('no_warnings'):
834 return
835 self.to_stderr(f'{self._color_text("WARNING:", "yellow")} {message}', only_once)
836
837 def report_error(self, message, tb=None):
838 '''
839 Do the same as trouble, but prefixes the message with 'ERROR:', colored
840 in red if stderr is a tty file.
841 '''
842 self.trouble(f'{self._color_text("ERROR:", "red")} {message}', tb)
843
844 def write_debug(self, message, only_once=False):
845 '''Log debug message or Print message to stderr'''
846 if not self.params.get('verbose', False):
847 return
848 message = '[debug] %s' % message
849 if self.params.get('logger'):
850 self.params['logger'].debug(message)
851 else:
852 self.to_stderr(message, only_once)
853
854 def report_file_already_downloaded(self, file_name):
855 """Report file has already been fully downloaded."""
856 try:
857 self.to_screen('[download] %s has already been downloaded' % file_name)
858 except UnicodeEncodeError:
859 self.to_screen('[download] The file has already been downloaded')
860
861 def report_file_delete(self, file_name):
862 """Report that existing file will be deleted."""
863 try:
864 self.to_screen('Deleting existing file %s' % file_name)
865 except UnicodeEncodeError:
866 self.to_screen('Deleting existing file')
867
868 def raise_no_formats(self, info, forced=False):
869 has_drm = info.get('__has_drm')
870 msg = 'This video is DRM protected' if has_drm else 'No video formats found!'
871 expected = self.params.get('ignore_no_formats_error')
872 if forced or not expected:
873 raise ExtractorError(msg, video_id=info['id'], ie=info['extractor'],
874 expected=has_drm or expected)
875 else:
876 self.report_warning(msg)
877
878 def parse_outtmpl(self):
879 outtmpl_dict = self.params.get('outtmpl', {})
880 if not isinstance(outtmpl_dict, dict):
881 outtmpl_dict = {'default': outtmpl_dict}
882 outtmpl_dict.update({
883 k: v for k, v in DEFAULT_OUTTMPL.items()
884 if outtmpl_dict.get(k) is None})
885 for key, val in outtmpl_dict.items():
886 if isinstance(val, bytes):
887 self.report_warning(
888 'Parameter outtmpl is bytes, but should be a unicode string. '
889 'Put from __future__ import unicode_literals at the top of your code file or consider switching to Python 3.x.')
890 return outtmpl_dict
891
892 def get_output_path(self, dir_type='', filename=None):
893 paths = self.params.get('paths', {})
894 assert isinstance(paths, dict)
895 path = os.path.join(
896 expand_path(paths.get('home', '').strip()),
897 expand_path(paths.get(dir_type, '').strip()) if dir_type else '',
898 filename or '')
899
900 # Temporary fix for #4787
901 # 'Treat' all problem characters by passing filename through preferredencoding
902 # to workaround encoding issues with subprocess on python2 @ Windows
903 if sys.version_info < (3, 0) and sys.platform == 'win32':
904 path = encodeFilename(path, True).decode(preferredencoding())
905 return sanitize_path(path, force=self.params.get('windowsfilenames'))
906
907 @staticmethod
908 def _outtmpl_expandpath(outtmpl):
909 # expand_path translates '%%' into '%' and '$$' into '$'
910 # correspondingly that is not what we want since we need to keep
911 # '%%' intact for template dict substitution step. Working around
912 # with boundary-alike separator hack.
913 sep = ''.join([random.choice(ascii_letters) for _ in range(32)])
914 outtmpl = outtmpl.replace('%%', '%{0}%'.format(sep)).replace('$$', '${0}$'.format(sep))
915
916 # outtmpl should be expand_path'ed before template dict substitution
917 # because meta fields may contain env variables we don't want to
918 # be expanded. For example, for outtmpl "%(title)s.%(ext)s" and
919 # title "Hello $PATH", we don't want `$PATH` to be expanded.
920 return expand_path(outtmpl).replace(sep, '')
921
922 @staticmethod
923 def escape_outtmpl(outtmpl):
924 ''' Escape any remaining strings like %s, %abc% etc. '''
925 return re.sub(
926 STR_FORMAT_RE_TMPL.format('', '(?![%(\0])'),
927 lambda mobj: ('' if mobj.group('has_key') else '%') + mobj.group(0),
928 outtmpl)
929
930 @classmethod
931 def validate_outtmpl(cls, outtmpl):
932 ''' @return None or Exception object '''
933 outtmpl = re.sub(
934 STR_FORMAT_RE_TMPL.format('[^)]*', '[ljqBU]'),
935 lambda mobj: f'{mobj.group(0)[:-1]}s',
936 cls._outtmpl_expandpath(outtmpl))
937 try:
938 cls.escape_outtmpl(outtmpl) % collections.defaultdict(int)
939 return None
940 except ValueError as err:
941 return err
942
943 def prepare_outtmpl(self, outtmpl, info_dict, sanitize=None):
944 """ Make the outtmpl and info_dict suitable for substitution: ydl.escape_outtmpl(outtmpl) % info_dict """
945 info_dict.setdefault('epoch', int(time.time())) # keep epoch consistent once set
946
947 info_dict = dict(info_dict) # Do not sanitize so as not to consume LazyList
948 for key in ('__original_infodict', '__postprocessors'):
949 info_dict.pop(key, None)
950 info_dict['duration_string'] = ( # %(duration>%H-%M-%S)s is wrong if duration > 24hrs
951 formatSeconds(info_dict['duration'], '-' if sanitize else ':')
952 if info_dict.get('duration', None) is not None
953 else None)
954 info_dict['autonumber'] = self.params.get('autonumber_start', 1) - 1 + self._num_downloads
955 if info_dict.get('resolution') is None:
956 info_dict['resolution'] = self.format_resolution(info_dict, default=None)
957
958 # For fields playlist_index, playlist_autonumber and autonumber convert all occurrences
959 # of %(field)s to %(field)0Nd for backward compatibility
960 field_size_compat_map = {
961 'playlist_index': len(str(info_dict.get('_last_playlist_index') or '')),
962 'playlist_autonumber': len(str(info_dict.get('n_entries') or '')),
963 'autonumber': self.params.get('autonumber_size') or 5,
964 }
965
966 TMPL_DICT = {}
967 EXTERNAL_FORMAT_RE = re.compile(STR_FORMAT_RE_TMPL.format('[^)]*', f'[{STR_FORMAT_TYPES}ljqBU]'))
968 MATH_FUNCTIONS = {
969 '+': float.__add__,
970 '-': float.__sub__,
971 }
972 # Field is of the form key1.key2...
973 # where keys (except first) can be string, int or slice
974 FIELD_RE = r'\w*(?:\.(?:\w+|{num}|{num}?(?::{num}?){{1,2}}))*'.format(num=r'(?:-?\d+)')
975 MATH_FIELD_RE = r'''{field}|{num}'''.format(field=FIELD_RE, num=r'-?\d+(?:.\d+)?')
976 MATH_OPERATORS_RE = r'(?:%s)' % '|'.join(map(re.escape, MATH_FUNCTIONS.keys()))
977 INTERNAL_FORMAT_RE = re.compile(r'''(?x)
978 (?P<negate>-)?
979 (?P<fields>{field})
980 (?P<maths>(?:{math_op}{math_field})*)
981 (?:>(?P<strf_format>.+?))?
982 (?P<alternate>(?<!\\),[^|)]+)?
983 (?:\|(?P<default>.*?))?
984 $'''.format(field=FIELD_RE, math_op=MATH_OPERATORS_RE, math_field=MATH_FIELD_RE))
985
986 def _traverse_infodict(k):
987 k = k.split('.')
988 if k[0] == '':
989 k.pop(0)
990 return traverse_obj(info_dict, k, is_user_input=True, traverse_string=True)
991
992 def get_value(mdict):
993 # Object traversal
994 value = _traverse_infodict(mdict['fields'])
995 # Negative
996 if mdict['negate']:
997 value = float_or_none(value)
998 if value is not None:
999 value *= -1
1000 # Do maths
1001 offset_key = mdict['maths']
1002 if offset_key:
1003 value = float_or_none(value)
1004 operator = None
1005 while offset_key:
1006 item = re.match(
1007 MATH_FIELD_RE if operator else MATH_OPERATORS_RE,
1008 offset_key).group(0)
1009 offset_key = offset_key[len(item):]
1010 if operator is None:
1011 operator = MATH_FUNCTIONS[item]
1012 continue
1013 item, multiplier = (item[1:], -1) if item[0] == '-' else (item, 1)
1014 offset = float_or_none(item)
1015 if offset is None:
1016 offset = float_or_none(_traverse_infodict(item))
1017 try:
1018 value = operator(value, multiplier * offset)
1019 except (TypeError, ZeroDivisionError):
1020 return None
1021 operator = None
1022 # Datetime formatting
1023 if mdict['strf_format']:
1024 value = strftime_or_none(value, mdict['strf_format'].replace('\\,', ','))
1025
1026 return value
1027
1028 na = self.params.get('outtmpl_na_placeholder', 'NA')
1029
1030 def _dumpjson_default(obj):
1031 if isinstance(obj, (set, LazyList)):
1032 return list(obj)
1033 raise TypeError(f'Object of type {type(obj).__name__} is not JSON serializable')
1034
1035 def create_key(outer_mobj):
1036 if not outer_mobj.group('has_key'):
1037 return f'%{outer_mobj.group(0)}'
1038 key = outer_mobj.group('key')
1039 mobj = re.match(INTERNAL_FORMAT_RE, key)
1040 initial_field = mobj.group('fields').split('.')[-1] if mobj else ''
1041 value, default = None, na
1042 while mobj:
1043 mobj = mobj.groupdict()
1044 default = mobj['default'] if mobj['default'] is not None else default
1045 value = get_value(mobj)
1046 if value is None and mobj['alternate']:
1047 mobj = re.match(INTERNAL_FORMAT_RE, mobj['alternate'][1:])
1048 else:
1049 break
1050
1051 fmt = outer_mobj.group('format')
1052 if fmt == 's' and value is not None and key in field_size_compat_map.keys():
1053 fmt = '0{:d}d'.format(field_size_compat_map[key])
1054
1055 value = default if value is None else value
1056
1057 str_fmt = f'{fmt[:-1]}s'
1058 if fmt[-1] == 'l': # list
1059 delim = '\n' if '#' in (outer_mobj.group('conversion') or '') else ', '
1060 value, fmt = delim.join(variadic(value)), str_fmt
1061 elif fmt[-1] == 'j': # json
1062 value, fmt = json.dumps(value, default=_dumpjson_default), str_fmt
1063 elif fmt[-1] == 'q': # quoted
1064 value, fmt = compat_shlex_quote(str(value)), str_fmt
1065 elif fmt[-1] == 'B': # bytes
1066 value = f'%{str_fmt}'.encode('utf-8') % str(value).encode('utf-8')
1067 value, fmt = value.decode('utf-8', 'ignore'), 's'
1068 elif fmt[-1] == 'U': # unicode normalized
1069 opts = outer_mobj.group('conversion') or ''
1070 value, fmt = unicodedata.normalize(
1071 # "+" = compatibility equivalence, "#" = NFD
1072 'NF%s%s' % ('K' if '+' in opts else '', 'D' if '#' in opts else 'C'),
1073 value), str_fmt
1074 elif fmt[-1] == 'c':
1075 if value:
1076 value = str(value)[0]
1077 else:
1078 fmt = str_fmt
1079 elif fmt[-1] not in 'rs': # numeric
1080 value = float_or_none(value)
1081 if value is None:
1082 value, fmt = default, 's'
1083
1084 if sanitize:
1085 if fmt[-1] == 'r':
1086 # If value is an object, sanitize might convert it to a string
1087 # So we convert it to repr first
1088 value, fmt = repr(value), str_fmt
1089 if fmt[-1] in 'csr':
1090 value = sanitize(initial_field, value)
1091
1092 key = '%s\0%s' % (key.replace('%', '%\0'), outer_mobj.group('format'))
1093 TMPL_DICT[key] = value
1094 return '{prefix}%({key}){fmt}'.format(key=key, fmt=fmt, prefix=outer_mobj.group('prefix'))
1095
1096 return EXTERNAL_FORMAT_RE.sub(create_key, outtmpl), TMPL_DICT
1097
1098 def evaluate_outtmpl(self, outtmpl, info_dict, *args, **kwargs):
1099 outtmpl, info_dict = self.prepare_outtmpl(outtmpl, info_dict, *args, **kwargs)
1100 return self.escape_outtmpl(outtmpl) % info_dict
1101
1102 def _prepare_filename(self, info_dict, tmpl_type='default'):
1103 try:
1104 sanitize = lambda k, v: sanitize_filename(
1105 compat_str(v),
1106 restricted=self.params.get('restrictfilenames'),
1107 is_id=(k == 'id' or k.endswith('_id')))
1108 outtmpl = self.outtmpl_dict.get(tmpl_type, self.outtmpl_dict['default'])
1109 outtmpl, template_dict = self.prepare_outtmpl(outtmpl, info_dict, sanitize)
1110 outtmpl = self.escape_outtmpl(self._outtmpl_expandpath(outtmpl))
1111 filename = outtmpl % template_dict
1112
1113 force_ext = OUTTMPL_TYPES.get(tmpl_type)
1114 if filename and force_ext is not None:
1115 filename = replace_extension(filename, force_ext, info_dict.get('ext'))
1116
1117 # https://github.com/blackjack4494/youtube-dlc/issues/85
1118 trim_file_name = self.params.get('trim_file_name', False)
1119 if trim_file_name:
1120 fn_groups = filename.rsplit('.')
1121 ext = fn_groups[-1]
1122 sub_ext = ''
1123 if len(fn_groups) > 2:
1124 sub_ext = fn_groups[-2]
1125 filename = '.'.join(filter(None, [fn_groups[0][:trim_file_name], sub_ext, ext]))
1126
1127 return filename
1128 except ValueError as err:
1129 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
1130 return None
1131
1132 def prepare_filename(self, info_dict, dir_type='', warn=False):
1133 """Generate the output filename."""
1134
1135 filename = self._prepare_filename(info_dict, dir_type or 'default')
1136 if not filename and dir_type not in ('', 'temp'):
1137 return ''
1138
1139 if warn:
1140 if not self.params.get('paths'):
1141 pass
1142 elif filename == '-':
1143 self.report_warning('--paths is ignored when an outputting to stdout', only_once=True)
1144 elif os.path.isabs(filename):
1145 self.report_warning('--paths is ignored since an absolute path is given in output template', only_once=True)
1146 if filename == '-' or not filename:
1147 return filename
1148
1149 return self.get_output_path(dir_type, filename)
1150
1151 def _match_entry(self, info_dict, incomplete=False, silent=False):
1152 """ Returns None if the file should be downloaded """
1153
1154 video_title = info_dict.get('title', info_dict.get('id', 'video'))
1155
1156 def check_filter():
1157 if 'title' in info_dict:
1158 # This can happen when we're just evaluating the playlist
1159 title = info_dict['title']
1160 matchtitle = self.params.get('matchtitle', False)
1161 if matchtitle:
1162 if not re.search(matchtitle, title, re.IGNORECASE):
1163 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
1164 rejecttitle = self.params.get('rejecttitle', False)
1165 if rejecttitle:
1166 if re.search(rejecttitle, title, re.IGNORECASE):
1167 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
1168 date = info_dict.get('upload_date')
1169 if date is not None:
1170 dateRange = self.params.get('daterange', DateRange())
1171 if date not in dateRange:
1172 return '%s upload date is not in range %s' % (date_from_str(date).isoformat(), dateRange)
1173 view_count = info_dict.get('view_count')
1174 if view_count is not None:
1175 min_views = self.params.get('min_views')
1176 if min_views is not None and view_count < min_views:
1177 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
1178 max_views = self.params.get('max_views')
1179 if max_views is not None and view_count > max_views:
1180 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
1181 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
1182 return 'Skipping "%s" because it is age restricted' % video_title
1183
1184 match_filter = self.params.get('match_filter')
1185 if match_filter is not None:
1186 try:
1187 ret = match_filter(info_dict, incomplete=incomplete)
1188 except TypeError:
1189 # For backward compatibility
1190 ret = None if incomplete else match_filter(info_dict)
1191 if ret is not None:
1192 return ret
1193 return None
1194
1195 if self.in_download_archive(info_dict):
1196 reason = '%s has already been recorded in the archive' % video_title
1197 break_opt, break_err = 'break_on_existing', ExistingVideoReached
1198 else:
1199 reason = check_filter()
1200 break_opt, break_err = 'break_on_reject', RejectedVideoReached
1201 if reason is not None:
1202 if not silent:
1203 self.to_screen('[download] ' + reason)
1204 if self.params.get(break_opt, False):
1205 raise break_err()
1206 return reason
1207
1208 @staticmethod
1209 def add_extra_info(info_dict, extra_info):
1210 '''Set the keys from extra_info in info dict if they are missing'''
1211 for key, value in extra_info.items():
1212 info_dict.setdefault(key, value)
1213
1214 def extract_info(self, url, download=True, ie_key=None, extra_info=None,
1215 process=True, force_generic_extractor=False):
1216 """
1217 Return a list with a dictionary for each video extracted.
1218
1219 Arguments:
1220 url -- URL to extract
1221
1222 Keyword arguments:
1223 download -- whether to download videos during extraction
1224 ie_key -- extractor key hint
1225 extra_info -- dictionary containing the extra values to add to each result
1226 process -- whether to resolve all unresolved references (URLs, playlist items),
1227 must be True for download to work.
1228 force_generic_extractor -- force using the generic extractor
1229 """
1230
1231 if extra_info is None:
1232 extra_info = {}
1233
1234 if not ie_key and force_generic_extractor:
1235 ie_key = 'Generic'
1236
1237 if ie_key:
1238 ies = {ie_key: self._get_info_extractor_class(ie_key)}
1239 else:
1240 ies = self._ies
1241
1242 for ie_key, ie in ies.items():
1243 if not ie.suitable(url):
1244 continue
1245
1246 if not ie.working():
1247 self.report_warning('The program functionality for this site has been marked as broken, '
1248 'and will probably not work.')
1249
1250 temp_id = ie.get_temp_id(url)
1251 if temp_id is not None and self.in_download_archive({'id': temp_id, 'ie_key': ie_key}):
1252 self.to_screen("[%s] %s: has already been recorded in archive" % (
1253 ie_key, temp_id))
1254 break
1255 return self.__extract_info(url, self.get_info_extractor(ie_key), download, extra_info, process)
1256 else:
1257 self.report_error('no suitable InfoExtractor for URL %s' % url)
1258
1259 def __handle_extraction_exceptions(func):
1260 @functools.wraps(func)
1261 def wrapper(self, *args, **kwargs):
1262 try:
1263 return func(self, *args, **kwargs)
1264 except GeoRestrictedError as e:
1265 msg = e.msg
1266 if e.countries:
1267 msg += '\nThis video is available in %s.' % ', '.join(
1268 map(ISO3166Utils.short2full, e.countries))
1269 msg += '\nYou might want to use a VPN or a proxy server (with --proxy) to workaround.'
1270 self.report_error(msg)
1271 except ExtractorError as e: # An error we somewhat expected
1272 self.report_error(compat_str(e), e.format_traceback())
1273 except ThrottledDownload:
1274 self.to_stderr('\r')
1275 self.report_warning('The download speed is below throttle limit. Re-extracting data')
1276 return wrapper(self, *args, **kwargs)
1277 except (MaxDownloadsReached, ExistingVideoReached, RejectedVideoReached, LazyList.IndexError):
1278 raise
1279 except Exception as e:
1280 if self.params.get('ignoreerrors'):
1281 self.report_error(error_to_compat_str(e), tb=encode_compat_str(traceback.format_exc()))
1282 else:
1283 raise
1284 return wrapper
1285
1286 @__handle_extraction_exceptions
1287 def __extract_info(self, url, ie, download, extra_info, process):
1288 ie_result = ie.extract(url)
1289 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
1290 return
1291 if isinstance(ie_result, list):
1292 # Backwards compatibility: old IE result format
1293 ie_result = {
1294 '_type': 'compat_list',
1295 'entries': ie_result,
1296 }
1297 if extra_info.get('original_url'):
1298 ie_result.setdefault('original_url', extra_info['original_url'])
1299 self.add_default_extra_info(ie_result, ie, url)
1300 if process:
1301 return self.process_ie_result(ie_result, download, extra_info)
1302 else:
1303 return ie_result
1304
1305 def add_default_extra_info(self, ie_result, ie, url):
1306 if url is not None:
1307 self.add_extra_info(ie_result, {
1308 'webpage_url': url,
1309 'original_url': url,
1310 'webpage_url_basename': url_basename(url),
1311 })
1312 if ie is not None:
1313 self.add_extra_info(ie_result, {
1314 'extractor': ie.IE_NAME,
1315 'extractor_key': ie.ie_key(),
1316 })
1317
1318 def process_ie_result(self, ie_result, download=True, extra_info=None):
1319 """
1320 Take the result of the ie(may be modified) and resolve all unresolved
1321 references (URLs, playlist items).
1322
1323 It will also download the videos if 'download'.
1324 Returns the resolved ie_result.
1325 """
1326 if extra_info is None:
1327 extra_info = {}
1328 result_type = ie_result.get('_type', 'video')
1329
1330 if result_type in ('url', 'url_transparent'):
1331 ie_result['url'] = sanitize_url(ie_result['url'])
1332 if ie_result.get('original_url'):
1333 extra_info.setdefault('original_url', ie_result['original_url'])
1334
1335 extract_flat = self.params.get('extract_flat', False)
1336 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info)
1337 or extract_flat is True):
1338 info_copy = ie_result.copy()
1339 ie = try_get(ie_result.get('ie_key'), self.get_info_extractor)
1340 if ie and not ie_result.get('id'):
1341 info_copy['id'] = ie.get_temp_id(ie_result['url'])
1342 self.add_default_extra_info(info_copy, ie, ie_result['url'])
1343 self.add_extra_info(info_copy, extra_info)
1344 self.__forced_printings(info_copy, self.prepare_filename(info_copy), incomplete=True)
1345 if self.params.get('force_write_download_archive', False):
1346 self.record_download_archive(info_copy)
1347 return ie_result
1348
1349 if result_type == 'video':
1350 self.add_extra_info(ie_result, extra_info)
1351 ie_result = self.process_video_result(ie_result, download=download)
1352 additional_urls = (ie_result or {}).get('additional_urls')
1353 if additional_urls:
1354 # TODO: Improve MetadataParserPP to allow setting a list
1355 if isinstance(additional_urls, compat_str):
1356 additional_urls = [additional_urls]
1357 self.to_screen(
1358 '[info] %s: %d additional URL(s) requested' % (ie_result['id'], len(additional_urls)))
1359 self.write_debug('Additional URLs: "%s"' % '", "'.join(additional_urls))
1360 ie_result['additional_entries'] = [
1361 self.extract_info(
1362 url, download, extra_info,
1363 force_generic_extractor=self.params.get('force_generic_extractor'))
1364 for url in additional_urls
1365 ]
1366 return ie_result
1367 elif result_type == 'url':
1368 # We have to add extra_info to the results because it may be
1369 # contained in a playlist
1370 return self.extract_info(
1371 ie_result['url'], download,
1372 ie_key=ie_result.get('ie_key'),
1373 extra_info=extra_info)
1374 elif result_type == 'url_transparent':
1375 # Use the information from the embedding page
1376 info = self.extract_info(
1377 ie_result['url'], ie_key=ie_result.get('ie_key'),
1378 extra_info=extra_info, download=False, process=False)
1379
1380 # extract_info may return None when ignoreerrors is enabled and
1381 # extraction failed with an error, don't crash and return early
1382 # in this case
1383 if not info:
1384 return info
1385
1386 force_properties = dict(
1387 (k, v) for k, v in ie_result.items() if v is not None)
1388 for f in ('_type', 'url', 'id', 'extractor', 'extractor_key', 'ie_key'):
1389 if f in force_properties:
1390 del force_properties[f]
1391 new_result = info.copy()
1392 new_result.update(force_properties)
1393
1394 # Extracted info may not be a video result (i.e.
1395 # info.get('_type', 'video') != video) but rather an url or
1396 # url_transparent. In such cases outer metadata (from ie_result)
1397 # should be propagated to inner one (info). For this to happen
1398 # _type of info should be overridden with url_transparent. This
1399 # fixes issue from https://github.com/ytdl-org/youtube-dl/pull/11163.
1400 if new_result.get('_type') == 'url':
1401 new_result['_type'] = 'url_transparent'
1402
1403 return self.process_ie_result(
1404 new_result, download=download, extra_info=extra_info)
1405 elif result_type in ('playlist', 'multi_video'):
1406 # Protect from infinite recursion due to recursively nested playlists
1407 # (see https://github.com/ytdl-org/youtube-dl/issues/27833)
1408 webpage_url = ie_result['webpage_url']
1409 if webpage_url in self._playlist_urls:
1410 self.to_screen(
1411 '[download] Skipping already downloaded playlist: %s'
1412 % ie_result.get('title') or ie_result.get('id'))
1413 return
1414
1415 self._playlist_level += 1
1416 self._playlist_urls.add(webpage_url)
1417 self._sanitize_thumbnails(ie_result)
1418 try:
1419 return self.__process_playlist(ie_result, download)
1420 finally:
1421 self._playlist_level -= 1
1422 if not self._playlist_level:
1423 self._playlist_urls.clear()
1424 elif result_type == 'compat_list':
1425 self.report_warning(
1426 'Extractor %s returned a compat_list result. '
1427 'It needs to be updated.' % ie_result.get('extractor'))
1428
1429 def _fixup(r):
1430 self.add_extra_info(r, {
1431 'extractor': ie_result['extractor'],
1432 'webpage_url': ie_result['webpage_url'],
1433 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1434 'extractor_key': ie_result['extractor_key'],
1435 })
1436 return r
1437 ie_result['entries'] = [
1438 self.process_ie_result(_fixup(r), download, extra_info)
1439 for r in ie_result['entries']
1440 ]
1441 return ie_result
1442 else:
1443 raise Exception('Invalid result type: %s' % result_type)
1444
1445 def _ensure_dir_exists(self, path):
1446 return make_dir(path, self.report_error)
1447
1448 def __process_playlist(self, ie_result, download):
1449 # We process each entry in the playlist
1450 playlist = ie_result.get('title') or ie_result.get('id')
1451 self.to_screen('[download] Downloading playlist: %s' % playlist)
1452
1453 if 'entries' not in ie_result:
1454 raise EntryNotInPlaylist()
1455 incomplete_entries = bool(ie_result.get('requested_entries'))
1456 if incomplete_entries:
1457 def fill_missing_entries(entries, indexes):
1458 ret = [None] * max(*indexes)
1459 for i, entry in zip(indexes, entries):
1460 ret[i - 1] = entry
1461 return ret
1462 ie_result['entries'] = fill_missing_entries(ie_result['entries'], ie_result['requested_entries'])
1463
1464 playlist_results = []
1465
1466 playliststart = self.params.get('playliststart', 1)
1467 playlistend = self.params.get('playlistend')
1468 # For backwards compatibility, interpret -1 as whole list
1469 if playlistend == -1:
1470 playlistend = None
1471
1472 playlistitems_str = self.params.get('playlist_items')
1473 playlistitems = None
1474 if playlistitems_str is not None:
1475 def iter_playlistitems(format):
1476 for string_segment in format.split(','):
1477 if '-' in string_segment:
1478 start, end = string_segment.split('-')
1479 for item in range(int(start), int(end) + 1):
1480 yield int(item)
1481 else:
1482 yield int(string_segment)
1483 playlistitems = orderedSet(iter_playlistitems(playlistitems_str))
1484
1485 ie_entries = ie_result['entries']
1486 msg = (
1487 'Downloading %d videos' if not isinstance(ie_entries, list)
1488 else 'Collected %d videos; downloading %%d of them' % len(ie_entries))
1489
1490 if isinstance(ie_entries, list):
1491 def get_entry(i):
1492 return ie_entries[i - 1]
1493 else:
1494 if not isinstance(ie_entries, PagedList):
1495 ie_entries = LazyList(ie_entries)
1496
1497 def get_entry(i):
1498 return YoutubeDL.__handle_extraction_exceptions(
1499 lambda self, i: ie_entries[i - 1]
1500 )(self, i)
1501
1502 entries = []
1503 items = playlistitems if playlistitems is not None else itertools.count(playliststart)
1504 for i in items:
1505 if i == 0:
1506 continue
1507 if playlistitems is None and playlistend is not None and playlistend < i:
1508 break
1509 entry = None
1510 try:
1511 entry = get_entry(i)
1512 if entry is None:
1513 raise EntryNotInPlaylist()
1514 except (IndexError, EntryNotInPlaylist):
1515 if incomplete_entries:
1516 raise EntryNotInPlaylist()
1517 elif not playlistitems:
1518 break
1519 entries.append(entry)
1520 try:
1521 if entry is not None:
1522 self._match_entry(entry, incomplete=True, silent=True)
1523 except (ExistingVideoReached, RejectedVideoReached):
1524 break
1525 ie_result['entries'] = entries
1526
1527 # Save playlist_index before re-ordering
1528 entries = [
1529 ((playlistitems[i - 1] if playlistitems else i + playliststart - 1), entry)
1530 for i, entry in enumerate(entries, 1)
1531 if entry is not None]
1532 n_entries = len(entries)
1533
1534 if not playlistitems and (playliststart or playlistend):
1535 playlistitems = list(range(playliststart, playliststart + n_entries))
1536 ie_result['requested_entries'] = playlistitems
1537
1538 if self.params.get('allow_playlist_files', True):
1539 ie_copy = {
1540 'playlist': playlist,
1541 'playlist_id': ie_result.get('id'),
1542 'playlist_title': ie_result.get('title'),
1543 'playlist_uploader': ie_result.get('uploader'),
1544 'playlist_uploader_id': ie_result.get('uploader_id'),
1545 'playlist_index': 0,
1546 }
1547 ie_copy.update(dict(ie_result))
1548
1549 if self._write_info_json('playlist', ie_result,
1550 self.prepare_filename(ie_copy, 'pl_infojson')) is None:
1551 return
1552 if self._write_description('playlist', ie_result,
1553 self.prepare_filename(ie_copy, 'pl_description')) is None:
1554 return
1555 # TODO: This should be passed to ThumbnailsConvertor if necessary
1556 self._write_thumbnails('playlist', ie_copy, self.prepare_filename(ie_copy, 'pl_thumbnail'))
1557
1558 if self.params.get('playlistreverse', False):
1559 entries = entries[::-1]
1560 if self.params.get('playlistrandom', False):
1561 random.shuffle(entries)
1562
1563 x_forwarded_for = ie_result.get('__x_forwarded_for_ip')
1564
1565 self.to_screen('[%s] playlist %s: %s' % (ie_result['extractor'], playlist, msg % n_entries))
1566 failures = 0
1567 max_failures = self.params.get('skip_playlist_after_errors') or float('inf')
1568 for i, entry_tuple in enumerate(entries, 1):
1569 playlist_index, entry = entry_tuple
1570 if 'playlist-index' in self.params.get('compat_opts', []):
1571 playlist_index = playlistitems[i - 1] if playlistitems else i + playliststart - 1
1572 self.to_screen('[download] Downloading video %s of %s' % (i, n_entries))
1573 # This __x_forwarded_for_ip thing is a bit ugly but requires
1574 # minimal changes
1575 if x_forwarded_for:
1576 entry['__x_forwarded_for_ip'] = x_forwarded_for
1577 extra = {
1578 'n_entries': n_entries,
1579 '_last_playlist_index': max(playlistitems) if playlistitems else (playlistend or n_entries),
1580 'playlist_index': playlist_index,
1581 'playlist_autonumber': i,
1582 'playlist': playlist,
1583 'playlist_id': ie_result.get('id'),
1584 'playlist_title': ie_result.get('title'),
1585 'playlist_uploader': ie_result.get('uploader'),
1586 'playlist_uploader_id': ie_result.get('uploader_id'),
1587 'extractor': ie_result['extractor'],
1588 'webpage_url': ie_result['webpage_url'],
1589 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1590 'extractor_key': ie_result['extractor_key'],
1591 }
1592
1593 if self._match_entry(entry, incomplete=True) is not None:
1594 continue
1595
1596 entry_result = self.__process_iterable_entry(entry, download, extra)
1597 if not entry_result:
1598 failures += 1
1599 if failures >= max_failures:
1600 self.report_error(
1601 'Skipping the remaining entries in playlist "%s" since %d items failed extraction' % (playlist, failures))
1602 break
1603 # TODO: skip failed (empty) entries?
1604 playlist_results.append(entry_result)
1605 ie_result['entries'] = playlist_results
1606 self.to_screen('[download] Finished downloading playlist: %s' % playlist)
1607 return ie_result
1608
1609 @__handle_extraction_exceptions
1610 def __process_iterable_entry(self, entry, download, extra_info):
1611 return self.process_ie_result(
1612 entry, download=download, extra_info=extra_info)
1613
1614 def _build_format_filter(self, filter_spec):
1615 " Returns a function to filter the formats according to the filter_spec "
1616
1617 OPERATORS = {
1618 '<': operator.lt,
1619 '<=': operator.le,
1620 '>': operator.gt,
1621 '>=': operator.ge,
1622 '=': operator.eq,
1623 '!=': operator.ne,
1624 }
1625 operator_rex = re.compile(r'''(?x)\s*
1626 (?P<key>width|height|tbr|abr|vbr|asr|filesize|filesize_approx|fps)\s*
1627 (?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
1628 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)\s*
1629 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
1630 m = operator_rex.fullmatch(filter_spec)
1631 if m:
1632 try:
1633 comparison_value = int(m.group('value'))
1634 except ValueError:
1635 comparison_value = parse_filesize(m.group('value'))
1636 if comparison_value is None:
1637 comparison_value = parse_filesize(m.group('value') + 'B')
1638 if comparison_value is None:
1639 raise ValueError(
1640 'Invalid value %r in format specification %r' % (
1641 m.group('value'), filter_spec))
1642 op = OPERATORS[m.group('op')]
1643
1644 if not m:
1645 STR_OPERATORS = {
1646 '=': operator.eq,
1647 '^=': lambda attr, value: attr.startswith(value),
1648 '$=': lambda attr, value: attr.endswith(value),
1649 '*=': lambda attr, value: value in attr,
1650 }
1651 str_operator_rex = re.compile(r'''(?x)\s*
1652 (?P<key>[a-zA-Z0-9._-]+)\s*
1653 (?P<negation>!\s*)?(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
1654 (?P<value>[a-zA-Z0-9._-]+)\s*
1655 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
1656 m = str_operator_rex.fullmatch(filter_spec)
1657 if m:
1658 comparison_value = m.group('value')
1659 str_op = STR_OPERATORS[m.group('op')]
1660 if m.group('negation'):
1661 op = lambda attr, value: not str_op(attr, value)
1662 else:
1663 op = str_op
1664
1665 if not m:
1666 raise SyntaxError('Invalid filter specification %r' % filter_spec)
1667
1668 def _filter(f):
1669 actual_value = f.get(m.group('key'))
1670 if actual_value is None:
1671 return m.group('none_inclusive')
1672 return op(actual_value, comparison_value)
1673 return _filter
1674
1675 def _default_format_spec(self, info_dict, download=True):
1676
1677 def can_merge():
1678 merger = FFmpegMergerPP(self)
1679 return merger.available and merger.can_merge()
1680
1681 prefer_best = (
1682 not self.params.get('simulate')
1683 and download
1684 and (
1685 not can_merge()
1686 or info_dict.get('is_live', False)
1687 or self.outtmpl_dict['default'] == '-'))
1688 compat = (
1689 prefer_best
1690 or self.params.get('allow_multiple_audio_streams', False)
1691 or 'format-spec' in self.params.get('compat_opts', []))
1692
1693 return (
1694 'best/bestvideo+bestaudio' if prefer_best
1695 else 'bestvideo*+bestaudio/best' if not compat
1696 else 'bestvideo+bestaudio/best')
1697
1698 def build_format_selector(self, format_spec):
1699 def syntax_error(note, start):
1700 message = (
1701 'Invalid format specification: '
1702 '{0}\n\t{1}\n\t{2}^'.format(note, format_spec, ' ' * start[1]))
1703 return SyntaxError(message)
1704
1705 PICKFIRST = 'PICKFIRST'
1706 MERGE = 'MERGE'
1707 SINGLE = 'SINGLE'
1708 GROUP = 'GROUP'
1709 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
1710
1711 allow_multiple_streams = {'audio': self.params.get('allow_multiple_audio_streams', False),
1712 'video': self.params.get('allow_multiple_video_streams', False)}
1713
1714 check_formats = self.params.get('check_formats')
1715
1716 def _parse_filter(tokens):
1717 filter_parts = []
1718 for type, string, start, _, _ in tokens:
1719 if type == tokenize.OP and string == ']':
1720 return ''.join(filter_parts)
1721 else:
1722 filter_parts.append(string)
1723
1724 def _remove_unused_ops(tokens):
1725 # Remove operators that we don't use and join them with the surrounding strings
1726 # for example: 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
1727 ALLOWED_OPS = ('/', '+', ',', '(', ')')
1728 last_string, last_start, last_end, last_line = None, None, None, None
1729 for type, string, start, end, line in tokens:
1730 if type == tokenize.OP and string == '[':
1731 if last_string:
1732 yield tokenize.NAME, last_string, last_start, last_end, last_line
1733 last_string = None
1734 yield type, string, start, end, line
1735 # everything inside brackets will be handled by _parse_filter
1736 for type, string, start, end, line in tokens:
1737 yield type, string, start, end, line
1738 if type == tokenize.OP and string == ']':
1739 break
1740 elif type == tokenize.OP and string in ALLOWED_OPS:
1741 if last_string:
1742 yield tokenize.NAME, last_string, last_start, last_end, last_line
1743 last_string = None
1744 yield type, string, start, end, line
1745 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
1746 if not last_string:
1747 last_string = string
1748 last_start = start
1749 last_end = end
1750 else:
1751 last_string += string
1752 if last_string:
1753 yield tokenize.NAME, last_string, last_start, last_end, last_line
1754
1755 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
1756 selectors = []
1757 current_selector = None
1758 for type, string, start, _, _ in tokens:
1759 # ENCODING is only defined in python 3.x
1760 if type == getattr(tokenize, 'ENCODING', None):
1761 continue
1762 elif type in [tokenize.NAME, tokenize.NUMBER]:
1763 current_selector = FormatSelector(SINGLE, string, [])
1764 elif type == tokenize.OP:
1765 if string == ')':
1766 if not inside_group:
1767 # ')' will be handled by the parentheses group
1768 tokens.restore_last_token()
1769 break
1770 elif inside_merge and string in ['/', ',']:
1771 tokens.restore_last_token()
1772 break
1773 elif inside_choice and string == ',':
1774 tokens.restore_last_token()
1775 break
1776 elif string == ',':
1777 if not current_selector:
1778 raise syntax_error('"," must follow a format selector', start)
1779 selectors.append(current_selector)
1780 current_selector = None
1781 elif string == '/':
1782 if not current_selector:
1783 raise syntax_error('"/" must follow a format selector', start)
1784 first_choice = current_selector
1785 second_choice = _parse_format_selection(tokens, inside_choice=True)
1786 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
1787 elif string == '[':
1788 if not current_selector:
1789 current_selector = FormatSelector(SINGLE, 'best', [])
1790 format_filter = _parse_filter(tokens)
1791 current_selector.filters.append(format_filter)
1792 elif string == '(':
1793 if current_selector:
1794 raise syntax_error('Unexpected "("', start)
1795 group = _parse_format_selection(tokens, inside_group=True)
1796 current_selector = FormatSelector(GROUP, group, [])
1797 elif string == '+':
1798 if not current_selector:
1799 raise syntax_error('Unexpected "+"', start)
1800 selector_1 = current_selector
1801 selector_2 = _parse_format_selection(tokens, inside_merge=True)
1802 if not selector_2:
1803 raise syntax_error('Expected a selector', start)
1804 current_selector = FormatSelector(MERGE, (selector_1, selector_2), [])
1805 else:
1806 raise syntax_error('Operator not recognized: "{0}"'.format(string), start)
1807 elif type == tokenize.ENDMARKER:
1808 break
1809 if current_selector:
1810 selectors.append(current_selector)
1811 return selectors
1812
1813 def _merge(formats_pair):
1814 format_1, format_2 = formats_pair
1815
1816 formats_info = []
1817 formats_info.extend(format_1.get('requested_formats', (format_1,)))
1818 formats_info.extend(format_2.get('requested_formats', (format_2,)))
1819
1820 if not allow_multiple_streams['video'] or not allow_multiple_streams['audio']:
1821 get_no_more = {'video': False, 'audio': False}
1822 for (i, fmt_info) in enumerate(formats_info):
1823 if fmt_info.get('acodec') == fmt_info.get('vcodec') == 'none':
1824 formats_info.pop(i)
1825 continue
1826 for aud_vid in ['audio', 'video']:
1827 if not allow_multiple_streams[aud_vid] and fmt_info.get(aud_vid[0] + 'codec') != 'none':
1828 if get_no_more[aud_vid]:
1829 formats_info.pop(i)
1830 break
1831 get_no_more[aud_vid] = True
1832
1833 if len(formats_info) == 1:
1834 return formats_info[0]
1835
1836 video_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('vcodec') != 'none']
1837 audio_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('acodec') != 'none']
1838
1839 the_only_video = video_fmts[0] if len(video_fmts) == 1 else None
1840 the_only_audio = audio_fmts[0] if len(audio_fmts) == 1 else None
1841
1842 output_ext = self.params.get('merge_output_format')
1843 if not output_ext:
1844 if the_only_video:
1845 output_ext = the_only_video['ext']
1846 elif the_only_audio and not video_fmts:
1847 output_ext = the_only_audio['ext']
1848 else:
1849 output_ext = 'mkv'
1850
1851 new_dict = {
1852 'requested_formats': formats_info,
1853 'format': '+'.join(fmt_info.get('format') for fmt_info in formats_info),
1854 'format_id': '+'.join(fmt_info.get('format_id') for fmt_info in formats_info),
1855 'ext': output_ext,
1856 }
1857
1858 if the_only_video:
1859 new_dict.update({
1860 'width': the_only_video.get('width'),
1861 'height': the_only_video.get('height'),
1862 'resolution': the_only_video.get('resolution') or self.format_resolution(the_only_video),
1863 'fps': the_only_video.get('fps'),
1864 'vcodec': the_only_video.get('vcodec'),
1865 'vbr': the_only_video.get('vbr'),
1866 'stretched_ratio': the_only_video.get('stretched_ratio'),
1867 })
1868
1869 if the_only_audio:
1870 new_dict.update({
1871 'acodec': the_only_audio.get('acodec'),
1872 'abr': the_only_audio.get('abr'),
1873 })
1874
1875 return new_dict
1876
1877 def _check_formats(formats):
1878 if not check_formats:
1879 yield from formats
1880 return
1881 for f in formats:
1882 self.to_screen('[info] Testing format %s' % f['format_id'])
1883 temp_file = tempfile.NamedTemporaryFile(
1884 suffix='.tmp', delete=False,
1885 dir=self.get_output_path('temp') or None)
1886 temp_file.close()
1887 try:
1888 success, _ = self.dl(temp_file.name, f, test=True)
1889 except (DownloadError, IOError, OSError, ValueError) + network_exceptions:
1890 success = False
1891 finally:
1892 if os.path.exists(temp_file.name):
1893 try:
1894 os.remove(temp_file.name)
1895 except OSError:
1896 self.report_warning('Unable to delete temporary file "%s"' % temp_file.name)
1897 if success:
1898 yield f
1899 else:
1900 self.to_screen('[info] Unable to download format %s. Skipping...' % f['format_id'])
1901
1902 def _build_selector_function(selector):
1903 if isinstance(selector, list): # ,
1904 fs = [_build_selector_function(s) for s in selector]
1905
1906 def selector_function(ctx):
1907 for f in fs:
1908 yield from f(ctx)
1909 return selector_function
1910
1911 elif selector.type == GROUP: # ()
1912 selector_function = _build_selector_function(selector.selector)
1913
1914 elif selector.type == PICKFIRST: # /
1915 fs = [_build_selector_function(s) for s in selector.selector]
1916
1917 def selector_function(ctx):
1918 for f in fs:
1919 picked_formats = list(f(ctx))
1920 if picked_formats:
1921 return picked_formats
1922 return []
1923
1924 elif selector.type == MERGE: # +
1925 selector_1, selector_2 = map(_build_selector_function, selector.selector)
1926
1927 def selector_function(ctx):
1928 for pair in itertools.product(
1929 selector_1(copy.deepcopy(ctx)), selector_2(copy.deepcopy(ctx))):
1930 yield _merge(pair)
1931
1932 elif selector.type == SINGLE: # atom
1933 format_spec = selector.selector or 'best'
1934
1935 # TODO: Add allvideo, allaudio etc by generalizing the code with best/worst selector
1936 if format_spec == 'all':
1937 def selector_function(ctx):
1938 yield from _check_formats(ctx['formats'])
1939 elif format_spec == 'mergeall':
1940 def selector_function(ctx):
1941 formats = list(_check_formats(ctx['formats']))
1942 if not formats:
1943 return
1944 merged_format = formats[-1]
1945 for f in formats[-2::-1]:
1946 merged_format = _merge((merged_format, f))
1947 yield merged_format
1948
1949 else:
1950 format_fallback, format_reverse, format_idx = False, True, 1
1951 mobj = re.match(
1952 r'(?P<bw>best|worst|b|w)(?P<type>video|audio|v|a)?(?P<mod>\*)?(?:\.(?P<n>[1-9]\d*))?$',
1953 format_spec)
1954 if mobj is not None:
1955 format_idx = int_or_none(mobj.group('n'), default=1)
1956 format_reverse = mobj.group('bw')[0] == 'b'
1957 format_type = (mobj.group('type') or [None])[0]
1958 not_format_type = {'v': 'a', 'a': 'v'}.get(format_type)
1959 format_modified = mobj.group('mod') is not None
1960
1961 format_fallback = not format_type and not format_modified # for b, w
1962 _filter_f = (
1963 (lambda f: f.get('%scodec' % format_type) != 'none')
1964 if format_type and format_modified # bv*, ba*, wv*, wa*
1965 else (lambda f: f.get('%scodec' % not_format_type) == 'none')
1966 if format_type # bv, ba, wv, wa
1967 else (lambda f: f.get('vcodec') != 'none' and f.get('acodec') != 'none')
1968 if not format_modified # b, w
1969 else lambda f: True) # b*, w*
1970 filter_f = lambda f: _filter_f(f) and (
1971 f.get('vcodec') != 'none' or f.get('acodec') != 'none')
1972 else:
1973 if format_spec in ('m4a', 'mp3', 'ogg', 'aac'): # audio extension
1974 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none'
1975 elif format_spec in ('mp4', 'flv', 'webm', '3gp'): # video extension
1976 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none' and f.get('vcodec') != 'none'
1977 elif format_spec in ('mhtml', ): # storyboards extension
1978 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') == 'none' and f.get('vcodec') == 'none'
1979 else:
1980 filter_f = lambda f: f.get('format_id') == format_spec # id
1981
1982 def selector_function(ctx):
1983 formats = list(ctx['formats'])
1984 matches = list(filter(filter_f, formats)) if filter_f is not None else formats
1985 if format_fallback and ctx['incomplete_formats'] and not matches:
1986 # for extractors with incomplete formats (audio only (soundcloud)
1987 # or video only (imgur)) best/worst will fallback to
1988 # best/worst {video,audio}-only format
1989 matches = formats
1990 matches = LazyList(_check_formats(matches[::-1 if format_reverse else 1]))
1991 try:
1992 yield matches[format_idx - 1]
1993 except IndexError:
1994 return
1995
1996 filters = [self._build_format_filter(f) for f in selector.filters]
1997
1998 def final_selector(ctx):
1999 ctx_copy = copy.deepcopy(ctx)
2000 for _filter in filters:
2001 ctx_copy['formats'] = list(filter(_filter, ctx_copy['formats']))
2002 return selector_function(ctx_copy)
2003 return final_selector
2004
2005 stream = io.BytesIO(format_spec.encode('utf-8'))
2006 try:
2007 tokens = list(_remove_unused_ops(compat_tokenize_tokenize(stream.readline)))
2008 except tokenize.TokenError:
2009 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
2010
2011 class TokenIterator(object):
2012 def __init__(self, tokens):
2013 self.tokens = tokens
2014 self.counter = 0
2015
2016 def __iter__(self):
2017 return self
2018
2019 def __next__(self):
2020 if self.counter >= len(self.tokens):
2021 raise StopIteration()
2022 value = self.tokens[self.counter]
2023 self.counter += 1
2024 return value
2025
2026 next = __next__
2027
2028 def restore_last_token(self):
2029 self.counter -= 1
2030
2031 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
2032 return _build_selector_function(parsed_selector)
2033
2034 def _calc_headers(self, info_dict):
2035 res = std_headers.copy()
2036
2037 add_headers = info_dict.get('http_headers')
2038 if add_headers:
2039 res.update(add_headers)
2040
2041 cookies = self._calc_cookies(info_dict)
2042 if cookies:
2043 res['Cookie'] = cookies
2044
2045 if 'X-Forwarded-For' not in res:
2046 x_forwarded_for_ip = info_dict.get('__x_forwarded_for_ip')
2047 if x_forwarded_for_ip:
2048 res['X-Forwarded-For'] = x_forwarded_for_ip
2049
2050 return res
2051
2052 def _calc_cookies(self, info_dict):
2053 pr = sanitized_Request(info_dict['url'])
2054 self.cookiejar.add_cookie_header(pr)
2055 return pr.get_header('Cookie')
2056
2057 def _sanitize_thumbnails(self, info_dict):
2058 thumbnails = info_dict.get('thumbnails')
2059 if thumbnails is None:
2060 thumbnail = info_dict.get('thumbnail')
2061 if thumbnail:
2062 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
2063 if thumbnails:
2064 thumbnails.sort(key=lambda t: (
2065 t.get('preference') if t.get('preference') is not None else -1,
2066 t.get('width') if t.get('width') is not None else -1,
2067 t.get('height') if t.get('height') is not None else -1,
2068 t.get('id') if t.get('id') is not None else '',
2069 t.get('url')))
2070
2071 def thumbnail_tester():
2072 if self.params.get('check_formats'):
2073 test_all = True
2074 to_screen = lambda msg: self.to_screen(f'[info] {msg}')
2075 else:
2076 test_all = False
2077 to_screen = self.write_debug
2078
2079 def test_thumbnail(t):
2080 if not test_all and not t.get('_test_url'):
2081 return True
2082 to_screen('Testing thumbnail %s' % t['id'])
2083 try:
2084 self.urlopen(HEADRequest(t['url']))
2085 except network_exceptions as err:
2086 to_screen('Unable to connect to thumbnail %s URL "%s" - %s. Skipping...' % (
2087 t['id'], t['url'], error_to_compat_str(err)))
2088 return False
2089 return True
2090
2091 return test_thumbnail
2092
2093 for i, t in enumerate(thumbnails):
2094 if t.get('id') is None:
2095 t['id'] = '%d' % i
2096 if t.get('width') and t.get('height'):
2097 t['resolution'] = '%dx%d' % (t['width'], t['height'])
2098 t['url'] = sanitize_url(t['url'])
2099
2100 if self.params.get('check_formats') is not False:
2101 info_dict['thumbnails'] = LazyList(filter(thumbnail_tester(), thumbnails[::-1])).reverse()
2102 else:
2103 info_dict['thumbnails'] = thumbnails
2104
2105 def process_video_result(self, info_dict, download=True):
2106 assert info_dict.get('_type', 'video') == 'video'
2107
2108 if 'id' not in info_dict:
2109 raise ExtractorError('Missing "id" field in extractor result')
2110 if 'title' not in info_dict:
2111 raise ExtractorError('Missing "title" field in extractor result',
2112 video_id=info_dict['id'], ie=info_dict['extractor'])
2113
2114 def report_force_conversion(field, field_not, conversion):
2115 self.report_warning(
2116 '"%s" field is not %s - forcing %s conversion, there is an error in extractor'
2117 % (field, field_not, conversion))
2118
2119 def sanitize_string_field(info, string_field):
2120 field = info.get(string_field)
2121 if field is None or isinstance(field, compat_str):
2122 return
2123 report_force_conversion(string_field, 'a string', 'string')
2124 info[string_field] = compat_str(field)
2125
2126 def sanitize_numeric_fields(info):
2127 for numeric_field in self._NUMERIC_FIELDS:
2128 field = info.get(numeric_field)
2129 if field is None or isinstance(field, compat_numeric_types):
2130 continue
2131 report_force_conversion(numeric_field, 'numeric', 'int')
2132 info[numeric_field] = int_or_none(field)
2133
2134 sanitize_string_field(info_dict, 'id')
2135 sanitize_numeric_fields(info_dict)
2136
2137 if 'playlist' not in info_dict:
2138 # It isn't part of a playlist
2139 info_dict['playlist'] = None
2140 info_dict['playlist_index'] = None
2141
2142 self._sanitize_thumbnails(info_dict)
2143
2144 thumbnail = info_dict.get('thumbnail')
2145 thumbnails = info_dict.get('thumbnails')
2146 if thumbnail:
2147 info_dict['thumbnail'] = sanitize_url(thumbnail)
2148 elif thumbnails:
2149 info_dict['thumbnail'] = thumbnails[-1]['url']
2150
2151 if info_dict.get('display_id') is None and 'id' in info_dict:
2152 info_dict['display_id'] = info_dict['id']
2153
2154 for ts_key, date_key in (
2155 ('timestamp', 'upload_date'),
2156 ('release_timestamp', 'release_date'),
2157 ):
2158 if info_dict.get(date_key) is None and info_dict.get(ts_key) is not None:
2159 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
2160 # see http://bugs.python.org/issue1646728)
2161 try:
2162 upload_date = datetime.datetime.utcfromtimestamp(info_dict[ts_key])
2163 info_dict[date_key] = upload_date.strftime('%Y%m%d')
2164 except (ValueError, OverflowError, OSError):
2165 pass
2166
2167 live_keys = ('is_live', 'was_live')
2168 live_status = info_dict.get('live_status')
2169 if live_status is None:
2170 for key in live_keys:
2171 if info_dict.get(key) is False:
2172 continue
2173 if info_dict.get(key):
2174 live_status = key
2175 break
2176 if all(info_dict.get(key) is False for key in live_keys):
2177 live_status = 'not_live'
2178 if live_status:
2179 info_dict['live_status'] = live_status
2180 for key in live_keys:
2181 if info_dict.get(key) is None:
2182 info_dict[key] = (live_status == key)
2183
2184 # Auto generate title fields corresponding to the *_number fields when missing
2185 # in order to always have clean titles. This is very common for TV series.
2186 for field in ('chapter', 'season', 'episode'):
2187 if info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
2188 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
2189
2190 for cc_kind in ('subtitles', 'automatic_captions'):
2191 cc = info_dict.get(cc_kind)
2192 if cc:
2193 for _, subtitle in cc.items():
2194 for subtitle_format in subtitle:
2195 if subtitle_format.get('url'):
2196 subtitle_format['url'] = sanitize_url(subtitle_format['url'])
2197 if subtitle_format.get('ext') is None:
2198 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
2199
2200 automatic_captions = info_dict.get('automatic_captions')
2201 subtitles = info_dict.get('subtitles')
2202
2203 info_dict['requested_subtitles'] = self.process_subtitles(
2204 info_dict['id'], subtitles, automatic_captions)
2205
2206 # We now pick which formats have to be downloaded
2207 if info_dict.get('formats') is None:
2208 # There's only one format available
2209 formats = [info_dict]
2210 else:
2211 formats = info_dict['formats']
2212
2213 info_dict['__has_drm'] = any(f.get('has_drm') for f in formats)
2214 if not self.params.get('allow_unplayable_formats'):
2215 formats = [f for f in formats if not f.get('has_drm')]
2216
2217 if not formats:
2218 self.raise_no_formats(info_dict)
2219
2220 def is_wellformed(f):
2221 url = f.get('url')
2222 if not url:
2223 self.report_warning(
2224 '"url" field is missing or empty - skipping format, '
2225 'there is an error in extractor')
2226 return False
2227 if isinstance(url, bytes):
2228 sanitize_string_field(f, 'url')
2229 return True
2230
2231 # Filter out malformed formats for better extraction robustness
2232 formats = list(filter(is_wellformed, formats))
2233
2234 formats_dict = {}
2235
2236 # We check that all the formats have the format and format_id fields
2237 for i, format in enumerate(formats):
2238 sanitize_string_field(format, 'format_id')
2239 sanitize_numeric_fields(format)
2240 format['url'] = sanitize_url(format['url'])
2241 if not format.get('format_id'):
2242 format['format_id'] = compat_str(i)
2243 else:
2244 # Sanitize format_id from characters used in format selector expression
2245 format['format_id'] = re.sub(r'[\s,/+\[\]()]', '_', format['format_id'])
2246 format_id = format['format_id']
2247 if format_id not in formats_dict:
2248 formats_dict[format_id] = []
2249 formats_dict[format_id].append(format)
2250
2251 # Make sure all formats have unique format_id
2252 for format_id, ambiguous_formats in formats_dict.items():
2253 if len(ambiguous_formats) > 1:
2254 for i, format in enumerate(ambiguous_formats):
2255 format['format_id'] = '%s-%d' % (format_id, i)
2256
2257 for i, format in enumerate(formats):
2258 if format.get('format') is None:
2259 format['format'] = '{id} - {res}{note}'.format(
2260 id=format['format_id'],
2261 res=self.format_resolution(format),
2262 note=format_field(format, 'format_note', ' (%s)'),
2263 )
2264 # Automatically determine file extension if missing
2265 if format.get('ext') is None:
2266 format['ext'] = determine_ext(format['url']).lower()
2267 # Automatically determine protocol if missing (useful for format
2268 # selection purposes)
2269 if format.get('protocol') is None:
2270 format['protocol'] = determine_protocol(format)
2271 # Add HTTP headers, so that external programs can use them from the
2272 # json output
2273 full_format_info = info_dict.copy()
2274 full_format_info.update(format)
2275 format['http_headers'] = self._calc_headers(full_format_info)
2276 # Remove private housekeeping stuff
2277 if '__x_forwarded_for_ip' in info_dict:
2278 del info_dict['__x_forwarded_for_ip']
2279
2280 # TODO Central sorting goes here
2281
2282 if not formats or formats[0] is not info_dict:
2283 # only set the 'formats' fields if the original info_dict list them
2284 # otherwise we end up with a circular reference, the first (and unique)
2285 # element in the 'formats' field in info_dict is info_dict itself,
2286 # which can't be exported to json
2287 info_dict['formats'] = formats
2288
2289 info_dict, _ = self.pre_process(info_dict)
2290
2291 if self.params.get('list_thumbnails'):
2292 self.list_thumbnails(info_dict)
2293 if self.params.get('listformats'):
2294 if not info_dict.get('formats') and not info_dict.get('url'):
2295 self.to_screen('%s has no formats' % info_dict['id'])
2296 else:
2297 self.list_formats(info_dict)
2298 if self.params.get('listsubtitles'):
2299 if 'automatic_captions' in info_dict:
2300 self.list_subtitles(
2301 info_dict['id'], automatic_captions, 'automatic captions')
2302 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
2303 list_only = self.params.get('simulate') is None and (
2304 self.params.get('list_thumbnails') or self.params.get('listformats') or self.params.get('listsubtitles'))
2305 if list_only:
2306 # Without this printing, -F --print-json will not work
2307 self.__forced_printings(info_dict, self.prepare_filename(info_dict), incomplete=True)
2308 return
2309
2310 format_selector = self.format_selector
2311 if format_selector is None:
2312 req_format = self._default_format_spec(info_dict, download=download)
2313 self.write_debug('Default format spec: %s' % req_format)
2314 format_selector = self.build_format_selector(req_format)
2315
2316 # While in format selection we may need to have an access to the original
2317 # format set in order to calculate some metrics or do some processing.
2318 # For now we need to be able to guess whether original formats provided
2319 # by extractor are incomplete or not (i.e. whether extractor provides only
2320 # video-only or audio-only formats) for proper formats selection for
2321 # extractors with such incomplete formats (see
2322 # https://github.com/ytdl-org/youtube-dl/pull/5556).
2323 # Since formats may be filtered during format selection and may not match
2324 # the original formats the results may be incorrect. Thus original formats
2325 # or pre-calculated metrics should be passed to format selection routines
2326 # as well.
2327 # We will pass a context object containing all necessary additional data
2328 # instead of just formats.
2329 # This fixes incorrect format selection issue (see
2330 # https://github.com/ytdl-org/youtube-dl/issues/10083).
2331 incomplete_formats = (
2332 # All formats are video-only or
2333 all(f.get('vcodec') != 'none' and f.get('acodec') == 'none' for f in formats)
2334 # all formats are audio-only
2335 or all(f.get('vcodec') == 'none' and f.get('acodec') != 'none' for f in formats))
2336
2337 ctx = {
2338 'formats': formats,
2339 'incomplete_formats': incomplete_formats,
2340 }
2341
2342 formats_to_download = list(format_selector(ctx))
2343 if not formats_to_download:
2344 if not self.params.get('ignore_no_formats_error'):
2345 raise ExtractorError('Requested format is not available', expected=True,
2346 video_id=info_dict['id'], ie=info_dict['extractor'])
2347 else:
2348 self.report_warning('Requested format is not available')
2349 # Process what we can, even without any available formats.
2350 self.process_info(dict(info_dict))
2351 elif download:
2352 self.to_screen(
2353 '[info] %s: Downloading %d format(s): %s' % (
2354 info_dict['id'], len(formats_to_download),
2355 ", ".join([f['format_id'] for f in formats_to_download])))
2356 for fmt in formats_to_download:
2357 new_info = dict(info_dict)
2358 # Save a reference to the original info_dict so that it can be modified in process_info if needed
2359 new_info['__original_infodict'] = info_dict
2360 new_info.update(fmt)
2361 self.process_info(new_info)
2362 # We update the info dict with the best quality format (backwards compatibility)
2363 if formats_to_download:
2364 info_dict.update(formats_to_download[-1])
2365 return info_dict
2366
2367 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
2368 """Select the requested subtitles and their format"""
2369 available_subs = {}
2370 if normal_subtitles and self.params.get('writesubtitles'):
2371 available_subs.update(normal_subtitles)
2372 if automatic_captions and self.params.get('writeautomaticsub'):
2373 for lang, cap_info in automatic_captions.items():
2374 if lang not in available_subs:
2375 available_subs[lang] = cap_info
2376
2377 if (not self.params.get('writesubtitles') and not
2378 self.params.get('writeautomaticsub') or not
2379 available_subs):
2380 return None
2381
2382 all_sub_langs = available_subs.keys()
2383 if self.params.get('allsubtitles', False):
2384 requested_langs = all_sub_langs
2385 elif self.params.get('subtitleslangs', False):
2386 # A list is used so that the order of languages will be the same as
2387 # given in subtitleslangs. See https://github.com/yt-dlp/yt-dlp/issues/1041
2388 requested_langs = []
2389 for lang_re in self.params.get('subtitleslangs'):
2390 if lang_re == 'all':
2391 requested_langs.extend(all_sub_langs)
2392 continue
2393 discard = lang_re[0] == '-'
2394 if discard:
2395 lang_re = lang_re[1:]
2396 current_langs = filter(re.compile(lang_re + '$').match, all_sub_langs)
2397 if discard:
2398 for lang in current_langs:
2399 while lang in requested_langs:
2400 requested_langs.remove(lang)
2401 else:
2402 requested_langs.extend(current_langs)
2403 requested_langs = orderedSet(requested_langs)
2404 elif 'en' in available_subs:
2405 requested_langs = ['en']
2406 else:
2407 requested_langs = [list(all_sub_langs)[0]]
2408 if requested_langs:
2409 self.write_debug('Downloading subtitles: %s' % ', '.join(requested_langs))
2410
2411 formats_query = self.params.get('subtitlesformat', 'best')
2412 formats_preference = formats_query.split('/') if formats_query else []
2413 subs = {}
2414 for lang in requested_langs:
2415 formats = available_subs.get(lang)
2416 if formats is None:
2417 self.report_warning('%s subtitles not available for %s' % (lang, video_id))
2418 continue
2419 for ext in formats_preference:
2420 if ext == 'best':
2421 f = formats[-1]
2422 break
2423 matches = list(filter(lambda f: f['ext'] == ext, formats))
2424 if matches:
2425 f = matches[-1]
2426 break
2427 else:
2428 f = formats[-1]
2429 self.report_warning(
2430 'No subtitle format found matching "%s" for language %s, '
2431 'using %s' % (formats_query, lang, f['ext']))
2432 subs[lang] = f
2433 return subs
2434
2435 def __forced_printings(self, info_dict, filename, incomplete):
2436 def print_mandatory(field, actual_field=None):
2437 if actual_field is None:
2438 actual_field = field
2439 if (self.params.get('force%s' % field, False)
2440 and (not incomplete or info_dict.get(actual_field) is not None)):
2441 self.to_stdout(info_dict[actual_field])
2442
2443 def print_optional(field):
2444 if (self.params.get('force%s' % field, False)
2445 and info_dict.get(field) is not None):
2446 self.to_stdout(info_dict[field])
2447
2448 info_dict = info_dict.copy()
2449 if filename is not None:
2450 info_dict['filename'] = filename
2451 if info_dict.get('requested_formats') is not None:
2452 # For RTMP URLs, also include the playpath
2453 info_dict['urls'] = '\n'.join(f['url'] + f.get('play_path', '') for f in info_dict['requested_formats'])
2454 elif 'url' in info_dict:
2455 info_dict['urls'] = info_dict['url'] + info_dict.get('play_path', '')
2456
2457 if self.params.get('forceprint') or self.params.get('forcejson'):
2458 self.post_extract(info_dict)
2459 for tmpl in self.params.get('forceprint', []):
2460 mobj = re.match(r'\w+(=?)$', tmpl)
2461 if mobj and mobj.group(1):
2462 tmpl = f'{tmpl[:-1]} = %({tmpl[:-1]})s'
2463 elif mobj:
2464 tmpl = '%({})s'.format(tmpl)
2465 self.to_stdout(self.evaluate_outtmpl(tmpl, info_dict))
2466
2467 print_mandatory('title')
2468 print_mandatory('id')
2469 print_mandatory('url', 'urls')
2470 print_optional('thumbnail')
2471 print_optional('description')
2472 print_optional('filename')
2473 if self.params.get('forceduration') and info_dict.get('duration') is not None:
2474 self.to_stdout(formatSeconds(info_dict['duration']))
2475 print_mandatory('format')
2476
2477 if self.params.get('forcejson'):
2478 self.to_stdout(json.dumps(self.sanitize_info(info_dict)))
2479
2480 def dl(self, name, info, subtitle=False, test=False):
2481 if not info.get('url'):
2482 self.raise_no_formats(info, True)
2483
2484 if test:
2485 verbose = self.params.get('verbose')
2486 params = {
2487 'test': True,
2488 'quiet': self.params.get('quiet') or not verbose,
2489 'verbose': verbose,
2490 'noprogress': not verbose,
2491 'nopart': True,
2492 'skip_unavailable_fragments': False,
2493 'keep_fragments': False,
2494 'overwrites': True,
2495 '_no_ytdl_file': True,
2496 }
2497 else:
2498 params = self.params
2499 fd = get_suitable_downloader(info, params, to_stdout=(name == '-'))(self, params)
2500 if not test:
2501 for ph in self._progress_hooks:
2502 fd.add_progress_hook(ph)
2503 urls = '", "'.join([f['url'] for f in info.get('requested_formats', [])] or [info['url']])
2504 self.write_debug('Invoking downloader on "%s"' % urls)
2505 new_info = dict(info)
2506 if new_info.get('http_headers') is None:
2507 new_info['http_headers'] = self._calc_headers(new_info)
2508 return fd.download(name, new_info, subtitle)
2509
2510 def process_info(self, info_dict):
2511 """Process a single resolved IE result."""
2512
2513 assert info_dict.get('_type', 'video') == 'video'
2514
2515 max_downloads = self.params.get('max_downloads')
2516 if max_downloads is not None:
2517 if self._num_downloads >= int(max_downloads):
2518 raise MaxDownloadsReached()
2519
2520 # TODO: backward compatibility, to be removed
2521 info_dict['fulltitle'] = info_dict['title']
2522
2523 if 'format' not in info_dict and 'ext' in info_dict:
2524 info_dict['format'] = info_dict['ext']
2525
2526 if self._match_entry(info_dict) is not None:
2527 return
2528
2529 self.post_extract(info_dict)
2530 self._num_downloads += 1
2531
2532 # info_dict['_filename'] needs to be set for backward compatibility
2533 info_dict['_filename'] = full_filename = self.prepare_filename(info_dict, warn=True)
2534 temp_filename = self.prepare_filename(info_dict, 'temp')
2535 files_to_move = {}
2536
2537 # Forced printings
2538 self.__forced_printings(info_dict, full_filename, incomplete=('format' not in info_dict))
2539
2540 if self.params.get('simulate'):
2541 if self.params.get('force_write_download_archive', False):
2542 self.record_download_archive(info_dict)
2543 # Do nothing else if in simulate mode
2544 return
2545
2546 if full_filename is None:
2547 return
2548 if not self._ensure_dir_exists(encodeFilename(full_filename)):
2549 return
2550 if not self._ensure_dir_exists(encodeFilename(temp_filename)):
2551 return
2552
2553 if self._write_description('video', info_dict,
2554 self.prepare_filename(info_dict, 'description')) is None:
2555 return
2556
2557 sub_files = self._write_subtitles(info_dict, temp_filename)
2558 if sub_files is None:
2559 return
2560 files_to_move.update(dict(sub_files))
2561
2562 thumb_files = self._write_thumbnails(
2563 'video', info_dict, temp_filename, self.prepare_filename(info_dict, 'thumbnail'))
2564 if thumb_files is None:
2565 return
2566 files_to_move.update(dict(thumb_files))
2567
2568 infofn = self.prepare_filename(info_dict, 'infojson')
2569 _infojson_written = self._write_info_json('video', info_dict, infofn)
2570 if _infojson_written:
2571 info_dict['__infojson_filename'] = infofn
2572 elif _infojson_written is None:
2573 return
2574
2575 # Note: Annotations are deprecated
2576 annofn = None
2577 if self.params.get('writeannotations', False):
2578 annofn = self.prepare_filename(info_dict, 'annotation')
2579 if annofn:
2580 if not self._ensure_dir_exists(encodeFilename(annofn)):
2581 return
2582 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(annofn)):
2583 self.to_screen('[info] Video annotations are already present')
2584 elif not info_dict.get('annotations'):
2585 self.report_warning('There are no annotations to write.')
2586 else:
2587 try:
2588 self.to_screen('[info] Writing video annotations to: ' + annofn)
2589 with io.open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
2590 annofile.write(info_dict['annotations'])
2591 except (KeyError, TypeError):
2592 self.report_warning('There are no annotations to write.')
2593 except (OSError, IOError):
2594 self.report_error('Cannot write annotations file: ' + annofn)
2595 return
2596
2597 # Write internet shortcut files
2598 url_link = webloc_link = desktop_link = False
2599 if self.params.get('writelink', False):
2600 if sys.platform == "darwin": # macOS.
2601 webloc_link = True
2602 elif sys.platform.startswith("linux"):
2603 desktop_link = True
2604 else: # if sys.platform in ['win32', 'cygwin']:
2605 url_link = True
2606 if self.params.get('writeurllink', False):
2607 url_link = True
2608 if self.params.get('writewebloclink', False):
2609 webloc_link = True
2610 if self.params.get('writedesktoplink', False):
2611 desktop_link = True
2612
2613 if url_link or webloc_link or desktop_link:
2614 if 'webpage_url' not in info_dict:
2615 self.report_error('Cannot write internet shortcut file because the "webpage_url" field is missing in the media information')
2616 return
2617 ascii_url = iri_to_uri(info_dict['webpage_url'])
2618
2619 def _write_link_file(extension, template, newline, embed_filename):
2620 linkfn = replace_extension(full_filename, extension, info_dict.get('ext'))
2621 if self.params.get('overwrites', True) and os.path.exists(encodeFilename(linkfn)):
2622 self.to_screen('[info] Internet shortcut is already present')
2623 else:
2624 try:
2625 self.to_screen('[info] Writing internet shortcut to: ' + linkfn)
2626 with io.open(encodeFilename(to_high_limit_path(linkfn)), 'w', encoding='utf-8', newline=newline) as linkfile:
2627 template_vars = {'url': ascii_url}
2628 if embed_filename:
2629 template_vars['filename'] = linkfn[:-(len(extension) + 1)]
2630 linkfile.write(template % template_vars)
2631 except (OSError, IOError):
2632 self.report_error('Cannot write internet shortcut ' + linkfn)
2633 return False
2634 return True
2635
2636 if url_link:
2637 if not _write_link_file('url', DOT_URL_LINK_TEMPLATE, '\r\n', embed_filename=False):
2638 return
2639 if webloc_link:
2640 if not _write_link_file('webloc', DOT_WEBLOC_LINK_TEMPLATE, '\n', embed_filename=False):
2641 return
2642 if desktop_link:
2643 if not _write_link_file('desktop', DOT_DESKTOP_LINK_TEMPLATE, '\n', embed_filename=True):
2644 return
2645
2646 try:
2647 info_dict, files_to_move = self.pre_process(info_dict, 'before_dl', files_to_move)
2648 except PostProcessingError as err:
2649 self.report_error('Preprocessing: %s' % str(err))
2650 return
2651
2652 must_record_download_archive = False
2653 if self.params.get('skip_download', False):
2654 info_dict['filepath'] = temp_filename
2655 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
2656 info_dict['__files_to_move'] = files_to_move
2657 info_dict = self.run_pp(MoveFilesAfterDownloadPP(self, False), info_dict)
2658 else:
2659 # Download
2660 info_dict.setdefault('__postprocessors', [])
2661 try:
2662
2663 def existing_file(*filepaths):
2664 ext = info_dict.get('ext')
2665 final_ext = self.params.get('final_ext', ext)
2666 existing_files = []
2667 for file in orderedSet(filepaths):
2668 if final_ext != ext:
2669 converted = replace_extension(file, final_ext, ext)
2670 if os.path.exists(encodeFilename(converted)):
2671 existing_files.append(converted)
2672 if os.path.exists(encodeFilename(file)):
2673 existing_files.append(file)
2674
2675 if not existing_files or self.params.get('overwrites', False):
2676 for file in orderedSet(existing_files):
2677 self.report_file_delete(file)
2678 os.remove(encodeFilename(file))
2679 return None
2680
2681 info_dict['ext'] = os.path.splitext(existing_files[0])[1][1:]
2682 return existing_files[0]
2683
2684 success = True
2685 if info_dict.get('requested_formats') is not None:
2686
2687 def compatible_formats(formats):
2688 # TODO: some formats actually allow this (mkv, webm, ogg, mp4), but not all of them.
2689 video_formats = [format for format in formats if format.get('vcodec') != 'none']
2690 audio_formats = [format for format in formats if format.get('acodec') != 'none']
2691 if len(video_formats) > 2 or len(audio_formats) > 2:
2692 return False
2693
2694 # Check extension
2695 exts = set(format.get('ext') for format in formats)
2696 COMPATIBLE_EXTS = (
2697 set(('mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'ismv', 'isma')),
2698 set(('webm',)),
2699 )
2700 for ext_sets in COMPATIBLE_EXTS:
2701 if ext_sets.issuperset(exts):
2702 return True
2703 # TODO: Check acodec/vcodec
2704 return False
2705
2706 requested_formats = info_dict['requested_formats']
2707 old_ext = info_dict['ext']
2708 if self.params.get('merge_output_format') is None:
2709 if not compatible_formats(requested_formats):
2710 info_dict['ext'] = 'mkv'
2711 self.report_warning(
2712 'Requested formats are incompatible for merge and will be merged into mkv')
2713 if (info_dict['ext'] == 'webm'
2714 and info_dict.get('thumbnails')
2715 # check with type instead of pp_key, __name__, or isinstance
2716 # since we dont want any custom PPs to trigger this
2717 and any(type(pp) == EmbedThumbnailPP for pp in self._pps['post_process'])):
2718 info_dict['ext'] = 'mkv'
2719 self.report_warning(
2720 'webm doesn\'t support embedding a thumbnail, mkv will be used')
2721 new_ext = info_dict['ext']
2722
2723 def correct_ext(filename, ext=new_ext):
2724 if filename == '-':
2725 return filename
2726 filename_real_ext = os.path.splitext(filename)[1][1:]
2727 filename_wo_ext = (
2728 os.path.splitext(filename)[0]
2729 if filename_real_ext in (old_ext, new_ext)
2730 else filename)
2731 return '%s.%s' % (filename_wo_ext, ext)
2732
2733 # Ensure filename always has a correct extension for successful merge
2734 full_filename = correct_ext(full_filename)
2735 temp_filename = correct_ext(temp_filename)
2736 dl_filename = existing_file(full_filename, temp_filename)
2737 info_dict['__real_download'] = False
2738
2739 _protocols = set(determine_protocol(f) for f in requested_formats)
2740 if len(_protocols) == 1: # All requested formats have same protocol
2741 info_dict['protocol'] = _protocols.pop()
2742 directly_mergable = FFmpegFD.can_merge_formats(info_dict, self.params)
2743 if dl_filename is not None:
2744 self.report_file_already_downloaded(dl_filename)
2745 elif (directly_mergable and get_suitable_downloader(
2746 info_dict, self.params, to_stdout=(temp_filename == '-')) == FFmpegFD):
2747 info_dict['url'] = '\n'.join(f['url'] for f in requested_formats)
2748 success, real_download = self.dl(temp_filename, info_dict)
2749 info_dict['__real_download'] = real_download
2750 else:
2751 downloaded = []
2752 merger = FFmpegMergerPP(self)
2753 if self.params.get('allow_unplayable_formats'):
2754 self.report_warning(
2755 'You have requested merging of multiple formats '
2756 'while also allowing unplayable formats to be downloaded. '
2757 'The formats won\'t be merged to prevent data corruption.')
2758 elif not merger.available:
2759 self.report_warning(
2760 'You have requested merging of multiple formats but ffmpeg is not installed. '
2761 'The formats won\'t be merged.')
2762
2763 if temp_filename == '-':
2764 reason = ('using a downloader other than ffmpeg' if directly_mergable
2765 else 'but the formats are incompatible for simultaneous download' if merger.available
2766 else 'but ffmpeg is not installed')
2767 self.report_warning(
2768 f'You have requested downloading multiple formats to stdout {reason}. '
2769 'The formats will be streamed one after the other')
2770 fname = temp_filename
2771 for f in requested_formats:
2772 new_info = dict(info_dict)
2773 del new_info['requested_formats']
2774 new_info.update(f)
2775 if temp_filename != '-':
2776 fname = prepend_extension(
2777 correct_ext(temp_filename, new_info['ext']),
2778 'f%s' % f['format_id'], new_info['ext'])
2779 if not self._ensure_dir_exists(fname):
2780 return
2781 f['filepath'] = fname
2782 downloaded.append(fname)
2783 partial_success, real_download = self.dl(fname, new_info)
2784 info_dict['__real_download'] = info_dict['__real_download'] or real_download
2785 success = success and partial_success
2786 if merger.available and not self.params.get('allow_unplayable_formats'):
2787 info_dict['__postprocessors'].append(merger)
2788 info_dict['__files_to_merge'] = downloaded
2789 # Even if there were no downloads, it is being merged only now
2790 info_dict['__real_download'] = True
2791 else:
2792 for file in downloaded:
2793 files_to_move[file] = None
2794 else:
2795 # Just a single file
2796 dl_filename = existing_file(full_filename, temp_filename)
2797 if dl_filename is None or dl_filename == temp_filename:
2798 # dl_filename == temp_filename could mean that the file was partially downloaded with --no-part.
2799 # So we should try to resume the download
2800 success, real_download = self.dl(temp_filename, info_dict)
2801 info_dict['__real_download'] = real_download
2802 else:
2803 self.report_file_already_downloaded(dl_filename)
2804
2805 dl_filename = dl_filename or temp_filename
2806 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
2807
2808 except network_exceptions as err:
2809 self.report_error('unable to download video data: %s' % error_to_compat_str(err))
2810 return
2811 except (OSError, IOError) as err:
2812 raise UnavailableVideoError(err)
2813 except (ContentTooShortError, ) as err:
2814 self.report_error('content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded))
2815 return
2816
2817 if success and full_filename != '-':
2818
2819 def fixup():
2820 do_fixup = True
2821 fixup_policy = self.params.get('fixup')
2822 vid = info_dict['id']
2823
2824 if fixup_policy in ('ignore', 'never'):
2825 return
2826 elif fixup_policy == 'warn':
2827 do_fixup = False
2828 elif fixup_policy != 'force':
2829 assert fixup_policy in ('detect_or_warn', None)
2830 if not info_dict.get('__real_download'):
2831 do_fixup = False
2832
2833 def ffmpeg_fixup(cndn, msg, cls):
2834 if not cndn:
2835 return
2836 if not do_fixup:
2837 self.report_warning(f'{vid}: {msg}')
2838 return
2839 pp = cls(self)
2840 if pp.available:
2841 info_dict['__postprocessors'].append(pp)
2842 else:
2843 self.report_warning(f'{vid}: {msg}. Install ffmpeg to fix this automatically')
2844
2845 stretched_ratio = info_dict.get('stretched_ratio')
2846 ffmpeg_fixup(
2847 stretched_ratio not in (1, None),
2848 f'Non-uniform pixel ratio {stretched_ratio}',
2849 FFmpegFixupStretchedPP)
2850
2851 ffmpeg_fixup(
2852 (info_dict.get('requested_formats') is None
2853 and info_dict.get('container') == 'm4a_dash'
2854 and info_dict.get('ext') == 'm4a'),
2855 'writing DASH m4a. Only some players support this container',
2856 FFmpegFixupM4aPP)
2857
2858 downloader = (get_suitable_downloader(info_dict, self.params).__name__
2859 if 'protocol' in info_dict else None)
2860 ffmpeg_fixup(info_dict.get('requested_formats') is None and downloader == 'HlsFD',
2861 'malformed AAC bitstream detected', FFmpegFixupM3u8PP)
2862 ffmpeg_fixup(downloader == 'WebSocketFragmentFD', 'malformed timestamps detected', FFmpegFixupTimestampPP)
2863 ffmpeg_fixup(downloader == 'WebSocketFragmentFD', 'malformed duration detected', FFmpegFixupDurationPP)
2864
2865 fixup()
2866 try:
2867 info_dict = self.post_process(dl_filename, info_dict, files_to_move)
2868 except PostProcessingError as err:
2869 self.report_error('Postprocessing: %s' % str(err))
2870 return
2871 try:
2872 for ph in self._post_hooks:
2873 ph(info_dict['filepath'])
2874 except Exception as err:
2875 self.report_error('post hooks: %s' % str(err))
2876 return
2877 must_record_download_archive = True
2878
2879 if must_record_download_archive or self.params.get('force_write_download_archive', False):
2880 self.record_download_archive(info_dict)
2881 max_downloads = self.params.get('max_downloads')
2882 if max_downloads is not None and self._num_downloads >= int(max_downloads):
2883 raise MaxDownloadsReached()
2884
2885 def download(self, url_list):
2886 """Download a given list of URLs."""
2887 outtmpl = self.outtmpl_dict['default']
2888 if (len(url_list) > 1
2889 and outtmpl != '-'
2890 and '%' not in outtmpl
2891 and self.params.get('max_downloads') != 1):
2892 raise SameFileError(outtmpl)
2893
2894 for url in url_list:
2895 try:
2896 # It also downloads the videos
2897 res = self.extract_info(
2898 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
2899 except UnavailableVideoError:
2900 self.report_error('unable to download video')
2901 except MaxDownloadsReached:
2902 self.to_screen('[info] Maximum number of downloads reached')
2903 raise
2904 except ExistingVideoReached:
2905 self.to_screen('[info] Encountered a video that is already in the archive, stopping due to --break-on-existing')
2906 raise
2907 except RejectedVideoReached:
2908 self.to_screen('[info] Encountered a video that did not match filter, stopping due to --break-on-reject')
2909 raise
2910 else:
2911 if self.params.get('dump_single_json', False):
2912 self.post_extract(res)
2913 self.to_stdout(json.dumps(self.sanitize_info(res)))
2914
2915 return self._download_retcode
2916
2917 def download_with_info_file(self, info_filename):
2918 with contextlib.closing(fileinput.FileInput(
2919 [info_filename], mode='r',
2920 openhook=fileinput.hook_encoded('utf-8'))) as f:
2921 # FileInput doesn't have a read method, we can't call json.load
2922 info = self.sanitize_info(json.loads('\n'.join(f)), self.params.get('clean_infojson', True))
2923 try:
2924 self.process_ie_result(info, download=True)
2925 except (DownloadError, EntryNotInPlaylist, ThrottledDownload):
2926 webpage_url = info.get('webpage_url')
2927 if webpage_url is not None:
2928 self.report_warning('The info failed to download, trying with "%s"' % webpage_url)
2929 return self.download([webpage_url])
2930 else:
2931 raise
2932 return self._download_retcode
2933
2934 @staticmethod
2935 def sanitize_info(info_dict, remove_private_keys=False):
2936 ''' Sanitize the infodict for converting to json '''
2937 if info_dict is None:
2938 return info_dict
2939 info_dict.setdefault('epoch', int(time.time()))
2940 remove_keys = {'__original_infodict'} # Always remove this since this may contain a copy of the entire dict
2941 keep_keys = ['_type'], # Always keep this to facilitate load-info-json
2942 if remove_private_keys:
2943 remove_keys |= {
2944 'requested_formats', 'requested_subtitles', 'requested_entries',
2945 'filepath', 'entries', 'original_url', 'playlist_autonumber',
2946 }
2947 empty_values = (None, {}, [], set(), tuple())
2948 reject = lambda k, v: k not in keep_keys and (
2949 k.startswith('_') or k in remove_keys or v in empty_values)
2950 else:
2951 reject = lambda k, v: k in remove_keys
2952 filter_fn = lambda obj: (
2953 list(map(filter_fn, obj)) if isinstance(obj, (LazyList, list, tuple, set))
2954 else obj if not isinstance(obj, dict)
2955 else dict((k, filter_fn(v)) for k, v in obj.items() if not reject(k, v)))
2956 return filter_fn(info_dict)
2957
2958 @staticmethod
2959 def filter_requested_info(info_dict, actually_filter=True):
2960 ''' Alias of sanitize_info for backward compatibility '''
2961 return YoutubeDL.sanitize_info(info_dict, actually_filter)
2962
2963 def run_pp(self, pp, infodict):
2964 files_to_delete = []
2965 if '__files_to_move' not in infodict:
2966 infodict['__files_to_move'] = {}
2967 try:
2968 files_to_delete, infodict = pp.run(infodict)
2969 except PostProcessingError as e:
2970 # Must be True and not 'only_download'
2971 if self.params.get('ignoreerrors') is True:
2972 self.report_error(e)
2973 return infodict
2974 raise
2975
2976 if not files_to_delete:
2977 return infodict
2978 if self.params.get('keepvideo', False):
2979 for f in files_to_delete:
2980 infodict['__files_to_move'].setdefault(f, '')
2981 else:
2982 for old_filename in set(files_to_delete):
2983 self.to_screen('Deleting original file %s (pass -k to keep)' % old_filename)
2984 try:
2985 os.remove(encodeFilename(old_filename))
2986 except (IOError, OSError):
2987 self.report_warning('Unable to remove downloaded original file')
2988 if old_filename in infodict['__files_to_move']:
2989 del infodict['__files_to_move'][old_filename]
2990 return infodict
2991
2992 @staticmethod
2993 def post_extract(info_dict):
2994 def actual_post_extract(info_dict):
2995 if info_dict.get('_type') in ('playlist', 'multi_video'):
2996 for video_dict in info_dict.get('entries', {}):
2997 actual_post_extract(video_dict or {})
2998 return
2999
3000 post_extractor = info_dict.get('__post_extractor') or (lambda: {})
3001 extra = post_extractor().items()
3002 info_dict.update(extra)
3003 info_dict.pop('__post_extractor', None)
3004
3005 original_infodict = info_dict.get('__original_infodict') or {}
3006 original_infodict.update(extra)
3007 original_infodict.pop('__post_extractor', None)
3008
3009 actual_post_extract(info_dict or {})
3010
3011 def pre_process(self, ie_info, key='pre_process', files_to_move=None):
3012 info = dict(ie_info)
3013 info['__files_to_move'] = files_to_move or {}
3014 for pp in self._pps[key]:
3015 info = self.run_pp(pp, info)
3016 return info, info.pop('__files_to_move', None)
3017
3018 def post_process(self, filename, ie_info, files_to_move=None):
3019 """Run all the postprocessors on the given file."""
3020 info = dict(ie_info)
3021 info['filepath'] = filename
3022 info['__files_to_move'] = files_to_move or {}
3023
3024 for pp in ie_info.get('__postprocessors', []) + self._pps['post_process']:
3025 info = self.run_pp(pp, info)
3026 info = self.run_pp(MoveFilesAfterDownloadPP(self), info)
3027 del info['__files_to_move']
3028 for pp in self._pps['after_move']:
3029 info = self.run_pp(pp, info)
3030 return info
3031
3032 def _make_archive_id(self, info_dict):
3033 video_id = info_dict.get('id')
3034 if not video_id:
3035 return
3036 # Future-proof against any change in case
3037 # and backwards compatibility with prior versions
3038 extractor = info_dict.get('extractor_key') or info_dict.get('ie_key') # key in a playlist
3039 if extractor is None:
3040 url = str_or_none(info_dict.get('url'))
3041 if not url:
3042 return
3043 # Try to find matching extractor for the URL and take its ie_key
3044 for ie_key, ie in self._ies.items():
3045 if ie.suitable(url):
3046 extractor = ie_key
3047 break
3048 else:
3049 return
3050 return '%s %s' % (extractor.lower(), video_id)
3051
3052 def in_download_archive(self, info_dict):
3053 fn = self.params.get('download_archive')
3054 if fn is None:
3055 return False
3056
3057 vid_id = self._make_archive_id(info_dict)
3058 if not vid_id:
3059 return False # Incomplete video information
3060
3061 return vid_id in self.archive
3062
3063 def record_download_archive(self, info_dict):
3064 fn = self.params.get('download_archive')
3065 if fn is None:
3066 return
3067 vid_id = self._make_archive_id(info_dict)
3068 assert vid_id
3069 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
3070 archive_file.write(vid_id + '\n')
3071 self.archive.add(vid_id)
3072
3073 @staticmethod
3074 def format_resolution(format, default='unknown'):
3075 is_images = format.get('vcodec') == 'none' and format.get('acodec') == 'none'
3076 if format.get('vcodec') == 'none' and format.get('acodec') != 'none':
3077 return 'audio only'
3078 if format.get('resolution') is not None:
3079 return format['resolution']
3080 if format.get('width') and format.get('height'):
3081 res = '%dx%d' % (format['width'], format['height'])
3082 elif format.get('height'):
3083 res = '%sp' % format['height']
3084 elif format.get('width'):
3085 res = '%dx?' % format['width']
3086 elif is_images:
3087 return 'images'
3088 else:
3089 return default
3090 return f'{res} images' if is_images else res
3091
3092 def _format_note(self, fdict):
3093 res = ''
3094 if fdict.get('ext') in ['f4f', 'f4m']:
3095 res += '(unsupported) '
3096 if fdict.get('language'):
3097 if res:
3098 res += ' '
3099 res += '[%s] ' % fdict['language']
3100 if fdict.get('format_note') is not None:
3101 res += fdict['format_note'] + ' '
3102 if fdict.get('tbr') is not None:
3103 res += '%4dk ' % fdict['tbr']
3104 if fdict.get('container') is not None:
3105 if res:
3106 res += ', '
3107 res += '%s container' % fdict['container']
3108 if (fdict.get('vcodec') is not None
3109 and fdict.get('vcodec') != 'none'):
3110 if res:
3111 res += ', '
3112 res += fdict['vcodec']
3113 if fdict.get('vbr') is not None:
3114 res += '@'
3115 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
3116 res += 'video@'
3117 if fdict.get('vbr') is not None:
3118 res += '%4dk' % fdict['vbr']
3119 if fdict.get('fps') is not None:
3120 if res:
3121 res += ', '
3122 res += '%sfps' % fdict['fps']
3123 if fdict.get('acodec') is not None:
3124 if res:
3125 res += ', '
3126 if fdict['acodec'] == 'none':
3127 res += 'video only'
3128 else:
3129 res += '%-5s' % fdict['acodec']
3130 elif fdict.get('abr') is not None:
3131 if res:
3132 res += ', '
3133 res += 'audio'
3134 if fdict.get('abr') is not None:
3135 res += '@%3dk' % fdict['abr']
3136 if fdict.get('asr') is not None:
3137 res += ' (%5dHz)' % fdict['asr']
3138 if fdict.get('filesize') is not None:
3139 if res:
3140 res += ', '
3141 res += format_bytes(fdict['filesize'])
3142 elif fdict.get('filesize_approx') is not None:
3143 if res:
3144 res += ', '
3145 res += '~' + format_bytes(fdict['filesize_approx'])
3146 return res
3147
3148 def list_formats(self, info_dict):
3149 formats = info_dict.get('formats', [info_dict])
3150 new_format = (
3151 'list-formats' not in self.params.get('compat_opts', [])
3152 and self.params.get('listformats_table', True) is not False)
3153 if new_format:
3154 table = [
3155 [
3156 format_field(f, 'format_id'),
3157 format_field(f, 'ext'),
3158 self.format_resolution(f),
3159 format_field(f, 'fps', '%d'),
3160 '|',
3161 format_field(f, 'filesize', ' %s', func=format_bytes) + format_field(f, 'filesize_approx', '~%s', func=format_bytes),
3162 format_field(f, 'tbr', '%4dk'),
3163 shorten_protocol_name(f.get('protocol', '').replace("native", "n")),
3164 '|',
3165 format_field(f, 'vcodec', default='unknown').replace('none', ''),
3166 format_field(f, 'vbr', '%4dk'),
3167 format_field(f, 'acodec', default='unknown').replace('none', ''),
3168 format_field(f, 'abr', '%3dk'),
3169 format_field(f, 'asr', '%5dHz'),
3170 ', '.join(filter(None, (
3171 'UNSUPPORTED' if f.get('ext') in ('f4f', 'f4m') else '',
3172 format_field(f, 'language', '[%s]'),
3173 format_field(f, 'format_note'),
3174 format_field(f, 'container', ignore=(None, f.get('ext'))),
3175 ))),
3176 ] for f in formats if f.get('preference') is None or f['preference'] >= -1000]
3177 header_line = ['ID', 'EXT', 'RESOLUTION', 'FPS', '|', ' FILESIZE', ' TBR', 'PROTO',
3178 '|', 'VCODEC', ' VBR', 'ACODEC', ' ABR', ' ASR', 'MORE INFO']
3179 else:
3180 table = [
3181 [
3182 format_field(f, 'format_id'),
3183 format_field(f, 'ext'),
3184 self.format_resolution(f),
3185 self._format_note(f)]
3186 for f in formats
3187 if f.get('preference') is None or f['preference'] >= -1000]
3188 header_line = ['format code', 'extension', 'resolution', 'note']
3189
3190 self.to_screen(
3191 '[info] Available formats for %s:' % info_dict['id'])
3192 self.to_stdout(render_table(
3193 header_line, table, delim=new_format, extraGap=(0 if new_format else 1), hideEmpty=new_format))
3194
3195 def list_thumbnails(self, info_dict):
3196 thumbnails = list(info_dict.get('thumbnails'))
3197 if not thumbnails:
3198 self.to_screen('[info] No thumbnails present for %s' % info_dict['id'])
3199 return
3200
3201 self.to_screen(
3202 '[info] Thumbnails for %s:' % info_dict['id'])
3203 self.to_stdout(render_table(
3204 ['ID', 'width', 'height', 'URL'],
3205 [[t['id'], t.get('width', 'unknown'), t.get('height', 'unknown'), t['url']] for t in thumbnails]))
3206
3207 def list_subtitles(self, video_id, subtitles, name='subtitles'):
3208 if not subtitles:
3209 self.to_screen('%s has no %s' % (video_id, name))
3210 return
3211 self.to_screen(
3212 'Available %s for %s:' % (name, video_id))
3213
3214 def _row(lang, formats):
3215 exts, names = zip(*((f['ext'], f.get('name') or 'unknown') for f in reversed(formats)))
3216 if len(set(names)) == 1:
3217 names = [] if names[0] == 'unknown' else names[:1]
3218 return [lang, ', '.join(names), ', '.join(exts)]
3219
3220 self.to_stdout(render_table(
3221 ['Language', 'Name', 'Formats'],
3222 [_row(lang, formats) for lang, formats in subtitles.items()],
3223 hideEmpty=True))
3224
3225 def urlopen(self, req):
3226 """ Start an HTTP download """
3227 if isinstance(req, compat_basestring):
3228 req = sanitized_Request(req)
3229 return self._opener.open(req, timeout=self._socket_timeout)
3230
3231 def print_debug_header(self):
3232 if not self.params.get('verbose'):
3233 return
3234
3235 stdout_encoding = getattr(
3236 sys.stdout, 'encoding', 'missing (%s)' % type(sys.stdout).__name__)
3237 encoding_str = (
3238 '[debug] Encodings: locale %s, fs %s, out %s, pref %s\n' % (
3239 locale.getpreferredencoding(),
3240 sys.getfilesystemencoding(),
3241 stdout_encoding,
3242 self.get_encoding()))
3243 write_string(encoding_str, encoding=None)
3244
3245 source = detect_variant()
3246 self._write_string('[debug] yt-dlp version %s%s\n' % (__version__, '' if source == 'unknown' else f' ({source})'))
3247 if _LAZY_LOADER:
3248 self._write_string('[debug] Lazy loading extractors enabled\n')
3249 if plugin_extractors or plugin_postprocessors:
3250 self._write_string('[debug] Plugins: %s\n' % [
3251 '%s%s' % (klass.__name__, '' if klass.__name__ == name else f' as {name}')
3252 for name, klass in itertools.chain(plugin_extractors.items(), plugin_postprocessors.items())])
3253 if self.params.get('compat_opts'):
3254 self._write_string(
3255 '[debug] Compatibility options: %s\n' % ', '.join(self.params.get('compat_opts')))
3256 try:
3257 sp = subprocess.Popen(
3258 ['git', 'rev-parse', '--short', 'HEAD'],
3259 stdout=subprocess.PIPE, stderr=subprocess.PIPE,
3260 cwd=os.path.dirname(os.path.abspath(__file__)))
3261 out, err = process_communicate_or_kill(sp)
3262 out = out.decode().strip()
3263 if re.match('[0-9a-f]+', out):
3264 self._write_string('[debug] Git HEAD: %s\n' % out)
3265 except Exception:
3266 try:
3267 sys.exc_clear()
3268 except Exception:
3269 pass
3270
3271 def python_implementation():
3272 impl_name = platform.python_implementation()
3273 if impl_name == 'PyPy' and hasattr(sys, 'pypy_version_info'):
3274 return impl_name + ' version %d.%d.%d' % sys.pypy_version_info[:3]
3275 return impl_name
3276
3277 self._write_string('[debug] Python version %s (%s %s) - %s\n' % (
3278 platform.python_version(),
3279 python_implementation(),
3280 platform.architecture()[0],
3281 platform_name()))
3282
3283 exe_versions = FFmpegPostProcessor.get_versions(self)
3284 exe_versions['rtmpdump'] = rtmpdump_version()
3285 exe_versions['phantomjs'] = PhantomJSwrapper._version()
3286 exe_str = ', '.join(
3287 f'{exe} {v}' for exe, v in sorted(exe_versions.items()) if v
3288 ) or 'none'
3289 self._write_string('[debug] exe versions: %s\n' % exe_str)
3290
3291 from .downloader.websocket import has_websockets
3292 from .postprocessor.embedthumbnail import has_mutagen
3293 from .cookies import SQLITE_AVAILABLE, KEYRING_AVAILABLE
3294
3295 lib_str = ', '.join(sorted(filter(None, (
3296 compat_pycrypto_AES and compat_pycrypto_AES.__name__.split('.')[0],
3297 has_websockets and 'websockets',
3298 has_mutagen and 'mutagen',
3299 SQLITE_AVAILABLE and 'sqlite',
3300 KEYRING_AVAILABLE and 'keyring',
3301 )))) or 'none'
3302 self._write_string('[debug] Optional libraries: %s\n' % lib_str)
3303 self._write_string('[debug] ANSI escape support: stdout = %s, stderr = %s\n' % (
3304 supports_terminal_sequences(self._screen_file),
3305 supports_terminal_sequences(self._err_file)))
3306
3307 proxy_map = {}
3308 for handler in self._opener.handlers:
3309 if hasattr(handler, 'proxies'):
3310 proxy_map.update(handler.proxies)
3311 self._write_string('[debug] Proxy map: ' + compat_str(proxy_map) + '\n')
3312
3313 if self.params.get('call_home', False):
3314 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode('utf-8')
3315 self._write_string('[debug] Public IP address: %s\n' % ipaddr)
3316 return
3317 latest_version = self.urlopen(
3318 'https://yt-dl.org/latest/version').read().decode('utf-8')
3319 if version_tuple(latest_version) > version_tuple(__version__):
3320 self.report_warning(
3321 'You are using an outdated version (newest version: %s)! '
3322 'See https://yt-dl.org/update if you need help updating.' %
3323 latest_version)
3324
3325 def _setup_opener(self):
3326 timeout_val = self.params.get('socket_timeout')
3327 self._socket_timeout = 600 if timeout_val is None else float(timeout_val)
3328
3329 opts_cookiesfrombrowser = self.params.get('cookiesfrombrowser')
3330 opts_cookiefile = self.params.get('cookiefile')
3331 opts_proxy = self.params.get('proxy')
3332
3333 self.cookiejar = load_cookies(opts_cookiefile, opts_cookiesfrombrowser, self)
3334
3335 cookie_processor = YoutubeDLCookieProcessor(self.cookiejar)
3336 if opts_proxy is not None:
3337 if opts_proxy == '':
3338 proxies = {}
3339 else:
3340 proxies = {'http': opts_proxy, 'https': opts_proxy}
3341 else:
3342 proxies = compat_urllib_request.getproxies()
3343 # Set HTTPS proxy to HTTP one if given (https://github.com/ytdl-org/youtube-dl/issues/805)
3344 if 'http' in proxies and 'https' not in proxies:
3345 proxies['https'] = proxies['http']
3346 proxy_handler = PerRequestProxyHandler(proxies)
3347
3348 debuglevel = 1 if self.params.get('debug_printtraffic') else 0
3349 https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
3350 ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
3351 redirect_handler = YoutubeDLRedirectHandler()
3352 data_handler = compat_urllib_request_DataHandler()
3353
3354 # When passing our own FileHandler instance, build_opener won't add the
3355 # default FileHandler and allows us to disable the file protocol, which
3356 # can be used for malicious purposes (see
3357 # https://github.com/ytdl-org/youtube-dl/issues/8227)
3358 file_handler = compat_urllib_request.FileHandler()
3359
3360 def file_open(*args, **kwargs):
3361 raise compat_urllib_error.URLError('file:// scheme is explicitly disabled in yt-dlp for security reasons')
3362 file_handler.file_open = file_open
3363
3364 opener = compat_urllib_request.build_opener(
3365 proxy_handler, https_handler, cookie_processor, ydlh, redirect_handler, data_handler, file_handler)
3366
3367 # Delete the default user-agent header, which would otherwise apply in
3368 # cases where our custom HTTP handler doesn't come into play
3369 # (See https://github.com/ytdl-org/youtube-dl/issues/1309 for details)
3370 opener.addheaders = []
3371 self._opener = opener
3372
3373 def encode(self, s):
3374 if isinstance(s, bytes):
3375 return s # Already encoded
3376
3377 try:
3378 return s.encode(self.get_encoding())
3379 except UnicodeEncodeError as err:
3380 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
3381 raise
3382
3383 def get_encoding(self):
3384 encoding = self.params.get('encoding')
3385 if encoding is None:
3386 encoding = preferredencoding()
3387 return encoding
3388
3389 def _write_info_json(self, label, ie_result, infofn):
3390 ''' Write infojson and returns True = written, False = skip, None = error '''
3391 if not self.params.get('writeinfojson'):
3392 return False
3393 elif not infofn:
3394 self.write_debug(f'Skipping writing {label} infojson')
3395 return False
3396 elif not self._ensure_dir_exists(infofn):
3397 return None
3398 elif not self.params.get('overwrites', True) and os.path.exists(infofn):
3399 self.to_screen(f'[info] {label.title()} metadata is already present')
3400 else:
3401 self.to_screen(f'[info] Writing {label} metadata as JSON to: {infofn}')
3402 try:
3403 write_json_file(self.sanitize_info(ie_result, self.params.get('clean_infojson', True)), infofn)
3404 except (OSError, IOError):
3405 self.report_error(f'Cannot write {label} metadata to JSON file {infofn}')
3406 return None
3407 return True
3408
3409 def _write_description(self, label, ie_result, descfn):
3410 ''' Write description and returns True = written, False = skip, None = error '''
3411 if not self.params.get('writedescription'):
3412 return False
3413 elif not descfn:
3414 self.write_debug(f'Skipping writing {label} description')
3415 return False
3416 elif not self._ensure_dir_exists(descfn):
3417 return None
3418 elif not self.params.get('overwrites', True) and os.path.exists(descfn):
3419 self.to_screen(f'[info] {label.title()} description is already present')
3420 elif ie_result.get('description') is None:
3421 self.report_warning(f'There\'s no {label} description to write')
3422 return False
3423 else:
3424 try:
3425 self.to_screen(f'[info] Writing {label} description to: {descfn}')
3426 with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
3427 descfile.write(ie_result['description'])
3428 except (OSError, IOError):
3429 self.report_error(f'Cannot write {label} description file {descfn}')
3430 return None
3431 return True
3432
3433 def _write_subtitles(self, info_dict, filename):
3434 ''' Write subtitles to file and return list of (sub_filename, final_sub_filename); or None if error'''
3435 ret = []
3436 subtitles = info_dict.get('requested_subtitles')
3437 if not subtitles or not (self.params.get('writesubtitles') or self.params.get('writeautomaticsub')):
3438 # subtitles download errors are already managed as troubles in relevant IE
3439 # that way it will silently go on when used with unsupporting IE
3440 return ret
3441
3442 sub_filename_base = self.prepare_filename(info_dict, 'subtitle')
3443 if not sub_filename_base:
3444 self.to_screen('[info] Skipping writing video subtitles')
3445 return ret
3446 for sub_lang, sub_info in subtitles.items():
3447 sub_format = sub_info['ext']
3448 sub_filename = subtitles_filename(filename, sub_lang, sub_format, info_dict.get('ext'))
3449 sub_filename_final = subtitles_filename(sub_filename_base, sub_lang, sub_format, info_dict.get('ext'))
3450 if not self.params.get('overwrites', True) and os.path.exists(sub_filename):
3451 self.to_screen(f'[info] Video subtitle {sub_lang}.{sub_format} is already present')
3452 sub_info['filepath'] = sub_filename
3453 ret.append((sub_filename, sub_filename_final))
3454 continue
3455
3456 self.to_screen(f'[info] Writing video subtitles to: {sub_filename}')
3457 if sub_info.get('data') is not None:
3458 try:
3459 # Use newline='' to prevent conversion of newline characters
3460 # See https://github.com/ytdl-org/youtube-dl/issues/10268
3461 with io.open(sub_filename, 'w', encoding='utf-8', newline='') as subfile:
3462 subfile.write(sub_info['data'])
3463 sub_info['filepath'] = sub_filename
3464 ret.append((sub_filename, sub_filename_final))
3465 continue
3466 except (OSError, IOError):
3467 self.report_error(f'Cannot write video subtitles file {sub_filename}')
3468 return None
3469
3470 try:
3471 sub_copy = sub_info.copy()
3472 sub_copy.setdefault('http_headers', info_dict.get('http_headers'))
3473 self.dl(sub_filename, sub_copy, subtitle=True)
3474 sub_info['filepath'] = sub_filename
3475 ret.append((sub_filename, sub_filename_final))
3476 except (ExtractorError, IOError, OSError, ValueError) + network_exceptions as err:
3477 self.report_warning(f'Unable to download video subtitles for {sub_lang!r}: {err}')
3478 continue
3479 return ret
3480
3481 def _write_thumbnails(self, label, info_dict, filename, thumb_filename_base=None):
3482 ''' Write thumbnails to file and return list of (thumb_filename, final_thumb_filename) '''
3483 write_all = self.params.get('write_all_thumbnails', False)
3484 thumbnails, ret = [], []
3485 if write_all or self.params.get('writethumbnail', False):
3486 thumbnails = info_dict.get('thumbnails') or []
3487 multiple = write_all and len(thumbnails) > 1
3488
3489 if thumb_filename_base is None:
3490 thumb_filename_base = filename
3491 if thumbnails and not thumb_filename_base:
3492 self.write_debug(f'Skipping writing {label} thumbnail')
3493 return ret
3494
3495 for t in thumbnails[::-1]:
3496 thumb_ext = (f'{t["id"]}.' if multiple else '') + determine_ext(t['url'], 'jpg')
3497 thumb_display_id = f'{label} thumbnail' + (f' {t["id"]}' if multiple else '')
3498 thumb_filename = replace_extension(filename, thumb_ext, info_dict.get('ext'))
3499 thumb_filename_final = replace_extension(thumb_filename_base, thumb_ext, info_dict.get('ext'))
3500
3501 if not self.params.get('overwrites', True) and os.path.exists(thumb_filename):
3502 ret.append((thumb_filename, thumb_filename_final))
3503 t['filepath'] = thumb_filename
3504 self.to_screen(f'[info] {thumb_display_id.title()} is already present')
3505 else:
3506 self.to_screen(f'[info] Downloading {thumb_display_id} ...')
3507 try:
3508 uf = self.urlopen(t['url'])
3509 self.to_screen(f'[info] Writing {thumb_display_id} to: {thumb_filename}')
3510 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
3511 shutil.copyfileobj(uf, thumbf)
3512 ret.append((thumb_filename, thumb_filename_final))
3513 t['filepath'] = thumb_filename
3514 except network_exceptions as err:
3515 self.report_warning(f'Unable to download {thumb_display_id}: {err}')
3516 if ret and not write_all:
3517 break
3518 return ret