]> jfr.im git - yt-dlp.git/blob - yt_dlp/YoutubeDL.py
fix motherless
[yt-dlp.git] / yt_dlp / YoutubeDL.py
1 import collections
2 import contextlib
3 import copy
4 import datetime as dt
5 import errno
6 import fileinput
7 import functools
8 import http.cookiejar
9 import io
10 import itertools
11 import json
12 import locale
13 import operator
14 import os
15 import random
16 import re
17 import shutil
18 import string
19 import subprocess
20 import sys
21 import tempfile
22 import time
23 import tokenize
24 import traceback
25 import unicodedata
26
27 from .cache import Cache
28 from .compat import urllib # isort: split
29 from .compat import compat_os_name, urllib_req_to_req
30 from .cookies import LenientSimpleCookie, load_cookies
31 from .downloader import FFmpegFD, get_suitable_downloader, shorten_protocol_name
32 from .downloader.rtmp import rtmpdump_version
33 from .extractor import gen_extractor_classes, get_info_extractor
34 from .extractor.common import UnsupportedURLIE
35 from .extractor.openload import PhantomJSwrapper
36 from .minicurses import format_text
37 from .networking import HEADRequest, Request, RequestDirector
38 from .networking.common import _REQUEST_HANDLERS, _RH_PREFERENCES
39 from .networking.exceptions import (
40 HTTPError,
41 NoSupportingHandlers,
42 RequestError,
43 SSLError,
44 network_exceptions,
45 )
46 from .networking.impersonate import ImpersonateRequestHandler
47 from .plugins import directories as plugin_directories
48 from .postprocessor import _PLUGIN_CLASSES as plugin_pps
49 from .postprocessor import (
50 EmbedThumbnailPP,
51 FFmpegFixupDuplicateMoovPP,
52 FFmpegFixupDurationPP,
53 FFmpegFixupM3u8PP,
54 FFmpegFixupM4aPP,
55 FFmpegFixupStretchedPP,
56 FFmpegFixupTimestampPP,
57 FFmpegMergerPP,
58 FFmpegPostProcessor,
59 FFmpegVideoConvertorPP,
60 MoveFilesAfterDownloadPP,
61 get_postprocessor,
62 )
63 from .postprocessor.ffmpeg import resolve_mapping as resolve_recode_mapping
64 from .update import (
65 REPOSITORY,
66 _get_system_deprecation,
67 _make_label,
68 current_git_head,
69 detect_variant,
70 )
71 from .utils import (
72 DEFAULT_OUTTMPL,
73 IDENTITY,
74 LINK_TEMPLATES,
75 MEDIA_EXTENSIONS,
76 NO_DEFAULT,
77 NUMBER_RE,
78 OUTTMPL_TYPES,
79 POSTPROCESS_WHEN,
80 STR_FORMAT_RE_TMPL,
81 STR_FORMAT_TYPES,
82 ContentTooShortError,
83 DateRange,
84 DownloadCancelled,
85 DownloadError,
86 EntryNotInPlaylist,
87 ExistingVideoReached,
88 ExtractorError,
89 FormatSorter,
90 GeoRestrictedError,
91 ISO3166Utils,
92 LazyList,
93 MaxDownloadsReached,
94 Namespace,
95 PagedList,
96 PlaylistEntries,
97 Popen,
98 PostProcessingError,
99 ReExtractInfo,
100 RejectedVideoReached,
101 SameFileError,
102 UnavailableVideoError,
103 UserNotLive,
104 YoutubeDLError,
105 age_restricted,
106 bug_reports_message,
107 date_from_str,
108 deprecation_warning,
109 determine_ext,
110 determine_protocol,
111 encode_compat_str,
112 encodeFilename,
113 escapeHTML,
114 expand_path,
115 extract_basic_auth,
116 filter_dict,
117 float_or_none,
118 format_bytes,
119 format_decimal_suffix,
120 format_field,
121 formatSeconds,
122 get_compatible_ext,
123 get_domain,
124 int_or_none,
125 iri_to_uri,
126 is_path_like,
127 join_nonempty,
128 locked_file,
129 make_archive_id,
130 make_dir,
131 number_of_digits,
132 orderedSet,
133 orderedSet_from_options,
134 parse_filesize,
135 preferredencoding,
136 prepend_extension,
137 remove_terminal_sequences,
138 render_table,
139 replace_extension,
140 sanitize_filename,
141 sanitize_path,
142 sanitize_url,
143 shell_quote,
144 str_or_none,
145 strftime_or_none,
146 subtitles_filename,
147 supports_terminal_sequences,
148 system_identifier,
149 filesize_from_tbr,
150 timetuple_from_msec,
151 to_high_limit_path,
152 traverse_obj,
153 try_call,
154 try_get,
155 url_basename,
156 variadic,
157 version_tuple,
158 windows_enable_vt_mode,
159 write_json_file,
160 write_string,
161 )
162 from .utils._utils import _UnsafeExtensionError, _YDLLogger
163 from .utils.networking import (
164 HTTPHeaderDict,
165 clean_headers,
166 clean_proxies,
167 std_headers,
168 )
169 from .version import CHANNEL, ORIGIN, RELEASE_GIT_HEAD, VARIANT, __version__
170
171 if compat_os_name == 'nt':
172 import ctypes
173
174
175 def _catch_unsafe_extension_error(func):
176 @functools.wraps(func)
177 def wrapper(self, *args, **kwargs):
178 try:
179 return func(self, *args, **kwargs)
180 except _UnsafeExtensionError as error:
181 self.report_error(
182 f'The extracted extension ({error.extension!r}) is unusual '
183 'and will be skipped for safety reasons. '
184 f'If you believe this is an error{bug_reports_message(",")}')
185
186 return wrapper
187
188
189 class YoutubeDL:
190 """YoutubeDL class.
191
192 YoutubeDL objects are the ones responsible of downloading the
193 actual video file and writing it to disk if the user has requested
194 it, among some other tasks. In most cases there should be one per
195 program. As, given a video URL, the downloader doesn't know how to
196 extract all the needed information, task that InfoExtractors do, it
197 has to pass the URL to one of them.
198
199 For this, YoutubeDL objects have a method that allows
200 InfoExtractors to be registered in a given order. When it is passed
201 a URL, the YoutubeDL object handles it to the first InfoExtractor it
202 finds that reports being able to handle it. The InfoExtractor extracts
203 all the information about the video or videos the URL refers to, and
204 YoutubeDL process the extracted information, possibly using a File
205 Downloader to download the video.
206
207 YoutubeDL objects accept a lot of parameters. In order not to saturate
208 the object constructor with arguments, it receives a dictionary of
209 options instead. These options are available through the params
210 attribute for the InfoExtractors to use. The YoutubeDL also
211 registers itself as the downloader in charge for the InfoExtractors
212 that are added to it, so this is a "mutual registration".
213
214 Available options:
215
216 username: Username for authentication purposes.
217 password: Password for authentication purposes.
218 videopassword: Password for accessing a video.
219 ap_mso: Adobe Pass multiple-system operator identifier.
220 ap_username: Multiple-system operator account username.
221 ap_password: Multiple-system operator account password.
222 usenetrc: Use netrc for authentication instead.
223 netrc_location: Location of the netrc file. Defaults to ~/.netrc.
224 netrc_cmd: Use a shell command to get credentials
225 verbose: Print additional info to stdout.
226 quiet: Do not print messages to stdout.
227 no_warnings: Do not print out anything for warnings.
228 forceprint: A dict with keys WHEN mapped to a list of templates to
229 print to stdout. The allowed keys are video or any of the
230 items in utils.POSTPROCESS_WHEN.
231 For compatibility, a single list is also accepted
232 print_to_file: A dict with keys WHEN (same as forceprint) mapped to
233 a list of tuples with (template, filename)
234 forcejson: Force printing info_dict as JSON.
235 dump_single_json: Force printing the info_dict of the whole playlist
236 (or video) as a single JSON line.
237 force_write_download_archive: Force writing download archive regardless
238 of 'skip_download' or 'simulate'.
239 simulate: Do not download the video files. If unset (or None),
240 simulate only if listsubtitles, listformats or list_thumbnails is used
241 format: Video format code. see "FORMAT SELECTION" for more details.
242 You can also pass a function. The function takes 'ctx' as
243 argument and returns the formats to download.
244 See "build_format_selector" for an implementation
245 allow_unplayable_formats: Allow unplayable formats to be extracted and downloaded.
246 ignore_no_formats_error: Ignore "No video formats" error. Usefull for
247 extracting metadata even if the video is not actually
248 available for download (experimental)
249 format_sort: A list of fields by which to sort the video formats.
250 See "Sorting Formats" for more details.
251 format_sort_force: Force the given format_sort. see "Sorting Formats"
252 for more details.
253 prefer_free_formats: Whether to prefer video formats with free containers
254 over non-free ones of same quality.
255 allow_multiple_video_streams: Allow multiple video streams to be merged
256 into a single file
257 allow_multiple_audio_streams: Allow multiple audio streams to be merged
258 into a single file
259 check_formats Whether to test if the formats are downloadable.
260 Can be True (check all), False (check none),
261 'selected' (check selected formats),
262 or None (check only if requested by extractor)
263 paths: Dictionary of output paths. The allowed keys are 'home'
264 'temp' and the keys of OUTTMPL_TYPES (in utils/_utils.py)
265 outtmpl: Dictionary of templates for output names. Allowed keys
266 are 'default' and the keys of OUTTMPL_TYPES (in utils/_utils.py).
267 For compatibility with youtube-dl, a single string can also be used
268 outtmpl_na_placeholder: Placeholder for unavailable meta fields.
269 restrictfilenames: Do not allow "&" and spaces in file names
270 trim_file_name: Limit length of filename (extension excluded)
271 windowsfilenames: Force the filenames to be windows compatible
272 ignoreerrors: Do not stop on download/postprocessing errors.
273 Can be 'only_download' to ignore only download errors.
274 Default is 'only_download' for CLI, but False for API
275 skip_playlist_after_errors: Number of allowed failures until the rest of
276 the playlist is skipped
277 allowed_extractors: List of regexes to match against extractor names that are allowed
278 overwrites: Overwrite all video and metadata files if True,
279 overwrite only non-video files if None
280 and don't overwrite any file if False
281 playlist_items: Specific indices of playlist to download.
282 playlistrandom: Download playlist items in random order.
283 lazy_playlist: Process playlist entries as they are received.
284 matchtitle: Download only matching titles.
285 rejecttitle: Reject downloads for matching titles.
286 logger: Log messages to a logging.Logger instance.
287 logtostderr: Print everything to stderr instead of stdout.
288 consoletitle: Display progress in console window's titlebar.
289 writedescription: Write the video description to a .description file
290 writeinfojson: Write the video description to a .info.json file
291 clean_infojson: Remove internal metadata from the infojson
292 getcomments: Extract video comments. This will not be written to disk
293 unless writeinfojson is also given
294 writeannotations: Write the video annotations to a .annotations.xml file
295 writethumbnail: Write the thumbnail image to a file
296 allow_playlist_files: Whether to write playlists' description, infojson etc
297 also to disk when using the 'write*' options
298 write_all_thumbnails: Write all thumbnail formats to files
299 writelink: Write an internet shortcut file, depending on the
300 current platform (.url/.webloc/.desktop)
301 writeurllink: Write a Windows internet shortcut file (.url)
302 writewebloclink: Write a macOS internet shortcut file (.webloc)
303 writedesktoplink: Write a Linux internet shortcut file (.desktop)
304 writesubtitles: Write the video subtitles to a file
305 writeautomaticsub: Write the automatically generated subtitles to a file
306 listsubtitles: Lists all available subtitles for the video
307 subtitlesformat: The format code for subtitles
308 subtitleslangs: List of languages of the subtitles to download (can be regex).
309 The list may contain "all" to refer to all the available
310 subtitles. The language can be prefixed with a "-" to
311 exclude it from the requested languages, e.g. ['all', '-live_chat']
312 keepvideo: Keep the video file after post-processing
313 daterange: A utils.DateRange object, download only if the upload_date is in the range.
314 skip_download: Skip the actual download of the video file
315 cachedir: Location of the cache files in the filesystem.
316 False to disable filesystem cache.
317 noplaylist: Download single video instead of a playlist if in doubt.
318 age_limit: An integer representing the user's age in years.
319 Unsuitable videos for the given age are skipped.
320 min_views: An integer representing the minimum view count the video
321 must have in order to not be skipped.
322 Videos without view count information are always
323 downloaded. None for no limit.
324 max_views: An integer representing the maximum view count.
325 Videos that are more popular than that are not
326 downloaded.
327 Videos without view count information are always
328 downloaded. None for no limit.
329 download_archive: A set, or the name of a file where all downloads are recorded.
330 Videos already present in the file are not downloaded again.
331 break_on_existing: Stop the download process after attempting to download a
332 file that is in the archive.
333 break_per_url: Whether break_on_reject and break_on_existing
334 should act on each input URL as opposed to for the entire queue
335 cookiefile: File name or text stream from where cookies should be read and dumped to
336 cookiesfrombrowser: A tuple containing the name of the browser, the profile
337 name/path from where cookies are loaded, the name of the keyring,
338 and the container name, e.g. ('chrome', ) or
339 ('vivaldi', 'default', 'BASICTEXT') or ('firefox', 'default', None, 'Meta')
340 legacyserverconnect: Explicitly allow HTTPS connection to servers that do not
341 support RFC 5746 secure renegotiation
342 nocheckcertificate: Do not verify SSL certificates
343 client_certificate: Path to client certificate file in PEM format. May include the private key
344 client_certificate_key: Path to private key file for client certificate
345 client_certificate_password: Password for client certificate private key, if encrypted.
346 If not provided and the key is encrypted, yt-dlp will ask interactively
347 prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
348 (Only supported by some extractors)
349 enable_file_urls: Enable file:// URLs. This is disabled by default for security reasons.
350 http_headers: A dictionary of custom headers to be used for all requests
351 proxy: URL of the proxy server to use
352 geo_verification_proxy: URL of the proxy to use for IP address verification
353 on geo-restricted sites.
354 socket_timeout: Time to wait for unresponsive hosts, in seconds
355 bidi_workaround: Work around buggy terminals without bidirectional text
356 support, using fridibi
357 debug_printtraffic:Print out sent and received HTTP traffic
358 default_search: Prepend this string if an input url is not valid.
359 'auto' for elaborate guessing
360 encoding: Use this encoding instead of the system-specified.
361 extract_flat: Whether to resolve and process url_results further
362 * False: Always process. Default for API
363 * True: Never process
364 * 'in_playlist': Do not process inside playlist/multi_video
365 * 'discard': Always process, but don't return the result
366 from inside playlist/multi_video
367 * 'discard_in_playlist': Same as "discard", but only for
368 playlists (not multi_video). Default for CLI
369 wait_for_video: If given, wait for scheduled streams to become available.
370 The value should be a tuple containing the range
371 (min_secs, max_secs) to wait between retries
372 postprocessors: A list of dictionaries, each with an entry
373 * key: The name of the postprocessor. See
374 yt_dlp/postprocessor/__init__.py for a list.
375 * when: When to run the postprocessor. Allowed values are
376 the entries of utils.POSTPROCESS_WHEN
377 Assumed to be 'post_process' if not given
378 progress_hooks: A list of functions that get called on download
379 progress, with a dictionary with the entries
380 * status: One of "downloading", "error", or "finished".
381 Check this first and ignore unknown values.
382 * info_dict: The extracted info_dict
383
384 If status is one of "downloading", or "finished", the
385 following properties may also be present:
386 * filename: The final filename (always present)
387 * tmpfilename: The filename we're currently writing to
388 * downloaded_bytes: Bytes on disk
389 * total_bytes: Size of the whole file, None if unknown
390 * total_bytes_estimate: Guess of the eventual file size,
391 None if unavailable.
392 * elapsed: The number of seconds since download started.
393 * eta: The estimated time in seconds, None if unknown
394 * speed: The download speed in bytes/second, None if
395 unknown
396 * fragment_index: The counter of the currently
397 downloaded video fragment.
398 * fragment_count: The number of fragments (= individual
399 files that will be merged)
400
401 Progress hooks are guaranteed to be called at least once
402 (with status "finished") if the download is successful.
403 postprocessor_hooks: A list of functions that get called on postprocessing
404 progress, with a dictionary with the entries
405 * status: One of "started", "processing", or "finished".
406 Check this first and ignore unknown values.
407 * postprocessor: Name of the postprocessor
408 * info_dict: The extracted info_dict
409
410 Progress hooks are guaranteed to be called at least twice
411 (with status "started" and "finished") if the processing is successful.
412 merge_output_format: "/" separated list of extensions to use when merging formats.
413 final_ext: Expected final extension; used to detect when the file was
414 already downloaded and converted
415 fixup: Automatically correct known faults of the file.
416 One of:
417 - "never": do nothing
418 - "warn": only emit a warning
419 - "detect_or_warn": check whether we can do anything
420 about it, warn otherwise (default)
421 source_address: Client-side IP address to bind to.
422 impersonate: Client to impersonate for requests.
423 An ImpersonateTarget (from yt_dlp.networking.impersonate)
424 sleep_interval_requests: Number of seconds to sleep between requests
425 during extraction
426 sleep_interval: Number of seconds to sleep before each download when
427 used alone or a lower bound of a range for randomized
428 sleep before each download (minimum possible number
429 of seconds to sleep) when used along with
430 max_sleep_interval.
431 max_sleep_interval:Upper bound of a range for randomized sleep before each
432 download (maximum possible number of seconds to sleep).
433 Must only be used along with sleep_interval.
434 Actual sleep time will be a random float from range
435 [sleep_interval; max_sleep_interval].
436 sleep_interval_subtitles: Number of seconds to sleep before each subtitle download
437 listformats: Print an overview of available video formats and exit.
438 list_thumbnails: Print a table of all thumbnails and exit.
439 match_filter: A function that gets called for every video with the signature
440 (info_dict, *, incomplete: bool) -> Optional[str]
441 For backward compatibility with youtube-dl, the signature
442 (info_dict) -> Optional[str] is also allowed.
443 - If it returns a message, the video is ignored.
444 - If it returns None, the video is downloaded.
445 - If it returns utils.NO_DEFAULT, the user is interactively
446 asked whether to download the video.
447 - Raise utils.DownloadCancelled(msg) to abort remaining
448 downloads when a video is rejected.
449 match_filter_func in utils/_utils.py is one example for this.
450 color: A Dictionary with output stream names as keys
451 and their respective color policy as values.
452 Can also just be a single color policy,
453 in which case it applies to all outputs.
454 Valid stream names are 'stdout' and 'stderr'.
455 Valid color policies are one of 'always', 'auto', 'no_color' or 'never'.
456 geo_bypass: Bypass geographic restriction via faking X-Forwarded-For
457 HTTP header
458 geo_bypass_country:
459 Two-letter ISO 3166-2 country code that will be used for
460 explicit geographic restriction bypassing via faking
461 X-Forwarded-For HTTP header
462 geo_bypass_ip_block:
463 IP range in CIDR notation that will be used similarly to
464 geo_bypass_country
465 external_downloader: A dictionary of protocol keys and the executable of the
466 external downloader to use for it. The allowed protocols
467 are default|http|ftp|m3u8|dash|rtsp|rtmp|mms.
468 Set the value to 'native' to use the native downloader
469 compat_opts: Compatibility options. See "Differences in default behavior".
470 The following options do not work when used through the API:
471 filename, abort-on-error, multistreams, no-live-chat,
472 format-sort, no-clean-infojson, no-playlist-metafiles,
473 no-keep-subs, no-attach-info-json, allow-unsafe-ext.
474 Refer __init__.py for their implementation
475 progress_template: Dictionary of templates for progress outputs.
476 Allowed keys are 'download', 'postprocess',
477 'download-title' (console title) and 'postprocess-title'.
478 The template is mapped on a dictionary with keys 'progress' and 'info'
479 retry_sleep_functions: Dictionary of functions that takes the number of attempts
480 as argument and returns the time to sleep in seconds.
481 Allowed keys are 'http', 'fragment', 'file_access'
482 download_ranges: A callback function that gets called for every video with
483 the signature (info_dict, ydl) -> Iterable[Section].
484 Only the returned sections will be downloaded.
485 Each Section is a dict with the following keys:
486 * start_time: Start time of the section in seconds
487 * end_time: End time of the section in seconds
488 * title: Section title (Optional)
489 * index: Section number (Optional)
490 force_keyframes_at_cuts: Re-encode the video when downloading ranges to get precise cuts
491 noprogress: Do not print the progress bar
492 live_from_start: Whether to download livestreams videos from the start
493
494 The following parameters are not used by YoutubeDL itself, they are used by
495 the downloader (see yt_dlp/downloader/common.py):
496 nopart, updatetime, buffersize, ratelimit, throttledratelimit, min_filesize,
497 max_filesize, test, noresizebuffer, retries, file_access_retries, fragment_retries,
498 continuedl, xattr_set_filesize, hls_use_mpegts, http_chunk_size,
499 external_downloader_args, concurrent_fragment_downloads, progress_delta.
500
501 The following options are used by the post processors:
502 ffmpeg_location: Location of the ffmpeg/avconv binary; either the path
503 to the binary or its containing directory.
504 postprocessor_args: A dictionary of postprocessor/executable keys (in lower case)
505 and a list of additional command-line arguments for the
506 postprocessor/executable. The dict can also have "PP+EXE" keys
507 which are used when the given exe is used by the given PP.
508 Use 'default' as the name for arguments to passed to all PP
509 For compatibility with youtube-dl, a single list of args
510 can also be used
511
512 The following options are used by the extractors:
513 extractor_retries: Number of times to retry for known errors (default: 3)
514 dynamic_mpd: Whether to process dynamic DASH manifests (default: True)
515 hls_split_discontinuity: Split HLS playlists to different formats at
516 discontinuities such as ad breaks (default: False)
517 extractor_args: A dictionary of arguments to be passed to the extractors.
518 See "EXTRACTOR ARGUMENTS" for details.
519 E.g. {'youtube': {'skip': ['dash', 'hls']}}
520 mark_watched: Mark videos watched (even with --simulate). Only for YouTube
521
522 The following options are deprecated and may be removed in the future:
523
524 break_on_reject: Stop the download process when encountering a video that
525 has been filtered out.
526 - `raise DownloadCancelled(msg)` in match_filter instead
527 force_generic_extractor: Force downloader to use the generic extractor
528 - Use allowed_extractors = ['generic', 'default']
529 playliststart: - Use playlist_items
530 Playlist item to start at.
531 playlistend: - Use playlist_items
532 Playlist item to end at.
533 playlistreverse: - Use playlist_items
534 Download playlist items in reverse order.
535 forceurl: - Use forceprint
536 Force printing final URL.
537 forcetitle: - Use forceprint
538 Force printing title.
539 forceid: - Use forceprint
540 Force printing ID.
541 forcethumbnail: - Use forceprint
542 Force printing thumbnail URL.
543 forcedescription: - Use forceprint
544 Force printing description.
545 forcefilename: - Use forceprint
546 Force printing final filename.
547 forceduration: - Use forceprint
548 Force printing duration.
549 allsubtitles: - Use subtitleslangs = ['all']
550 Downloads all the subtitles of the video
551 (requires writesubtitles or writeautomaticsub)
552 include_ads: - Doesn't work
553 Download ads as well
554 call_home: - Not implemented
555 Boolean, true iff we are allowed to contact the
556 yt-dlp servers for debugging.
557 post_hooks: - Register a custom postprocessor
558 A list of functions that get called as the final step
559 for each video file, after all postprocessors have been
560 called. The filename will be passed as the only argument.
561 hls_prefer_native: - Use external_downloader = {'m3u8': 'native'} or {'m3u8': 'ffmpeg'}.
562 Use the native HLS downloader instead of ffmpeg/avconv
563 if True, otherwise use ffmpeg/avconv if False, otherwise
564 use downloader suggested by extractor if None.
565 prefer_ffmpeg: - avconv support is deprecated
566 If False, use avconv instead of ffmpeg if both are available,
567 otherwise prefer ffmpeg.
568 youtube_include_dash_manifest: - Use extractor_args
569 If True (default), DASH manifests and related
570 data will be downloaded and processed by extractor.
571 You can reduce network I/O by disabling it if you don't
572 care about DASH. (only for youtube)
573 youtube_include_hls_manifest: - Use extractor_args
574 If True (default), HLS manifests and related
575 data will be downloaded and processed by extractor.
576 You can reduce network I/O by disabling it if you don't
577 care about HLS. (only for youtube)
578 no_color: Same as `color='no_color'`
579 no_overwrites: Same as `overwrites=False`
580 """
581
582 _NUMERIC_FIELDS = {
583 'width', 'height', 'asr', 'audio_channels', 'fps',
584 'tbr', 'abr', 'vbr', 'filesize', 'filesize_approx',
585 'timestamp', 'release_timestamp',
586 'duration', 'view_count', 'like_count', 'dislike_count', 'repost_count',
587 'average_rating', 'comment_count', 'age_limit',
588 'start_time', 'end_time',
589 'chapter_number', 'season_number', 'episode_number',
590 'track_number', 'disc_number', 'release_year',
591 }
592
593 _format_fields = {
594 # NB: Keep in sync with the docstring of extractor/common.py
595 'url', 'manifest_url', 'manifest_stream_number', 'ext', 'format', 'format_id', 'format_note',
596 'width', 'height', 'aspect_ratio', 'resolution', 'dynamic_range', 'tbr', 'abr', 'acodec', 'asr', 'audio_channels',
597 'vbr', 'fps', 'vcodec', 'container', 'filesize', 'filesize_approx', 'rows', 'columns',
598 'player_url', 'protocol', 'fragment_base_url', 'fragments', 'is_from_start', 'is_dash_periods', 'request_data',
599 'preference', 'language', 'language_preference', 'quality', 'source_preference', 'cookies',
600 'http_headers', 'stretched_ratio', 'no_resume', 'has_drm', 'extra_param_to_segment_url', 'extra_param_to_key_url',
601 'hls_aes', 'downloader_options', 'page_url', 'app', 'play_path', 'tc_url', 'flash_version',
602 'rtmp_live', 'rtmp_conn', 'rtmp_protocol', 'rtmp_real_time',
603 }
604 _deprecated_multivalue_fields = {
605 'album_artist': 'album_artists',
606 'artist': 'artists',
607 'composer': 'composers',
608 'creator': 'creators',
609 'genre': 'genres',
610 }
611 _format_selection_exts = {
612 'audio': set(MEDIA_EXTENSIONS.common_audio),
613 'video': {*MEDIA_EXTENSIONS.common_video, '3gp'},
614 'storyboards': set(MEDIA_EXTENSIONS.storyboards),
615 }
616
617 def __init__(self, params=None, auto_init=True):
618 """Create a FileDownloader object with the given options.
619 @param auto_init Whether to load the default extractors and print header (if verbose).
620 Set to 'no_verbose_header' to not print the header
621 """
622 if params is None:
623 params = {}
624 self.params = params
625 self._ies = {}
626 self._ies_instances = {}
627 self._pps = {k: [] for k in POSTPROCESS_WHEN}
628 self._printed_messages = set()
629 self._first_webpage_request = True
630 self._post_hooks = []
631 self._progress_hooks = []
632 self._postprocessor_hooks = []
633 self._download_retcode = 0
634 self._num_downloads = 0
635 self._num_videos = 0
636 self._playlist_level = 0
637 self._playlist_urls = set()
638 self.cache = Cache(self)
639 self.__header_cookies = []
640
641 stdout = sys.stderr if self.params.get('logtostderr') else sys.stdout
642 self._out_files = Namespace(
643 out=stdout,
644 error=sys.stderr,
645 screen=sys.stderr if self.params.get('quiet') else stdout,
646 console=None if compat_os_name == 'nt' else next(
647 filter(supports_terminal_sequences, (sys.stderr, sys.stdout)), None),
648 )
649
650 try:
651 windows_enable_vt_mode()
652 except Exception as e:
653 self.write_debug(f'Failed to enable VT mode: {e}')
654
655 if self.params.get('no_color'):
656 if self.params.get('color') is not None:
657 self.params.setdefault('_warnings', []).append(
658 'Overwriting params from "color" with "no_color"')
659 self.params['color'] = 'no_color'
660
661 term_allow_color = os.getenv('TERM', '').lower() != 'dumb'
662 no_color = bool(os.getenv('NO_COLOR'))
663
664 def process_color_policy(stream):
665 stream_name = {sys.stdout: 'stdout', sys.stderr: 'stderr'}[stream]
666 policy = traverse_obj(self.params, ('color', (stream_name, None), {str}), get_all=False)
667 if policy in ('auto', None):
668 if term_allow_color and supports_terminal_sequences(stream):
669 return 'no_color' if no_color else True
670 return False
671 assert policy in ('always', 'never', 'no_color'), policy
672 return {'always': True, 'never': False}.get(policy, policy)
673
674 self._allow_colors = Namespace(**{
675 name: process_color_policy(stream)
676 for name, stream in self._out_files.items_ if name != 'console'
677 })
678
679 system_deprecation = _get_system_deprecation()
680 if system_deprecation:
681 self.deprecated_feature(system_deprecation.replace('\n', '\n '))
682
683 if self.params.get('allow_unplayable_formats'):
684 self.report_warning(
685 f'You have asked for {self._format_err("UNPLAYABLE", self.Styles.EMPHASIS)} formats to be listed/downloaded. '
686 'This is a developer option intended for debugging. \n'
687 ' If you experience any issues while using this option, '
688 f'{self._format_err("DO NOT", self.Styles.ERROR)} open a bug report')
689
690 if self.params.get('bidi_workaround', False):
691 try:
692 import pty
693 master, slave = pty.openpty()
694 width = shutil.get_terminal_size().columns
695 width_args = [] if width is None else ['-w', str(width)]
696 sp_kwargs = {'stdin': subprocess.PIPE, 'stdout': slave, 'stderr': self._out_files.error}
697 try:
698 self._output_process = Popen(['bidiv', *width_args], **sp_kwargs)
699 except OSError:
700 self._output_process = Popen(['fribidi', '-c', 'UTF-8', *width_args], **sp_kwargs)
701 self._output_channel = os.fdopen(master, 'rb')
702 except OSError as ose:
703 if ose.errno == errno.ENOENT:
704 self.report_warning(
705 'Could not find fribidi executable, ignoring --bidi-workaround. '
706 'Make sure that fribidi is an executable file in one of the directories in your $PATH.')
707 else:
708 raise
709
710 self.params['compat_opts'] = set(self.params.get('compat_opts', ()))
711 self.params['http_headers'] = HTTPHeaderDict(std_headers, self.params.get('http_headers'))
712 self._load_cookies(self.params['http_headers'].get('Cookie')) # compat
713 self.params['http_headers'].pop('Cookie', None)
714
715 if auto_init and auto_init != 'no_verbose_header':
716 self.print_debug_header()
717
718 def check_deprecated(param, option, suggestion):
719 if self.params.get(param) is not None:
720 self.report_warning(f'{option} is deprecated. Use {suggestion} instead')
721 return True
722 return False
723
724 if check_deprecated('cn_verification_proxy', '--cn-verification-proxy', '--geo-verification-proxy'):
725 if self.params.get('geo_verification_proxy') is None:
726 self.params['geo_verification_proxy'] = self.params['cn_verification_proxy']
727
728 check_deprecated('autonumber', '--auto-number', '-o "%(autonumber)s-%(title)s.%(ext)s"')
729 check_deprecated('usetitle', '--title', '-o "%(title)s-%(id)s.%(ext)s"')
730 check_deprecated('useid', '--id', '-o "%(id)s.%(ext)s"')
731
732 for msg in self.params.get('_warnings', []):
733 self.report_warning(msg)
734 for msg in self.params.get('_deprecation_warnings', []):
735 self.deprecated_feature(msg)
736
737 if impersonate_target := self.params.get('impersonate'):
738 if not self._impersonate_target_available(impersonate_target):
739 raise YoutubeDLError(
740 f'Impersonate target "{impersonate_target}" is not available. '
741 f'Use --list-impersonate-targets to see available targets. '
742 f'You may be missing dependencies required to support this target.')
743
744 if 'list-formats' in self.params['compat_opts']:
745 self.params['listformats_table'] = False
746
747 if 'overwrites' not in self.params and self.params.get('nooverwrites') is not None:
748 # nooverwrites was unnecessarily changed to overwrites
749 # in 0c3d0f51778b153f65c21906031c2e091fcfb641
750 # This ensures compatibility with both keys
751 self.params['overwrites'] = not self.params['nooverwrites']
752 elif self.params.get('overwrites') is None:
753 self.params.pop('overwrites', None)
754 else:
755 self.params['nooverwrites'] = not self.params['overwrites']
756
757 if self.params.get('simulate') is None and any((
758 self.params.get('list_thumbnails'),
759 self.params.get('listformats'),
760 self.params.get('listsubtitles'),
761 )):
762 self.params['simulate'] = 'list_only'
763
764 self.params.setdefault('forceprint', {})
765 self.params.setdefault('print_to_file', {})
766
767 # Compatibility with older syntax
768 if not isinstance(params['forceprint'], dict):
769 self.params['forceprint'] = {'video': params['forceprint']}
770
771 if auto_init:
772 self.add_default_info_extractors()
773
774 if (sys.platform != 'win32'
775 and sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968']
776 and not self.params.get('restrictfilenames', False)):
777 # Unicode filesystem API will throw errors (#1474, #13027)
778 self.report_warning(
779 'Assuming --restrict-filenames since file system encoding '
780 'cannot encode all characters. '
781 'Set the LC_ALL environment variable to fix this.')
782 self.params['restrictfilenames'] = True
783
784 self._parse_outtmpl()
785
786 # Creating format selector here allows us to catch syntax errors before the extraction
787 self.format_selector = (
788 self.params.get('format') if self.params.get('format') in (None, '-')
789 else self.params['format'] if callable(self.params['format'])
790 else self.build_format_selector(self.params['format']))
791
792 hooks = {
793 'post_hooks': self.add_post_hook,
794 'progress_hooks': self.add_progress_hook,
795 'postprocessor_hooks': self.add_postprocessor_hook,
796 }
797 for opt, fn in hooks.items():
798 for ph in self.params.get(opt, []):
799 fn(ph)
800
801 for pp_def_raw in self.params.get('postprocessors', []):
802 pp_def = dict(pp_def_raw)
803 when = pp_def.pop('when', 'post_process')
804 self.add_post_processor(
805 get_postprocessor(pp_def.pop('key'))(self, **pp_def),
806 when=when)
807
808 def preload_download_archive(fn):
809 """Preload the archive, if any is specified"""
810 archive = set()
811 if fn is None:
812 return archive
813 elif not is_path_like(fn):
814 return fn
815
816 self.write_debug(f'Loading archive file {fn!r}')
817 try:
818 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
819 for line in archive_file:
820 archive.add(line.strip())
821 except OSError as ioe:
822 if ioe.errno != errno.ENOENT:
823 raise
824 return archive
825
826 self.archive = preload_download_archive(self.params.get('download_archive'))
827
828 def warn_if_short_id(self, argv):
829 # short YouTube ID starting with dash?
830 idxs = [
831 i for i, a in enumerate(argv)
832 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
833 if idxs:
834 correct_argv = (
835 ['yt-dlp']
836 + [a for i, a in enumerate(argv) if i not in idxs]
837 + ['--'] + [argv[i] for i in idxs]
838 )
839 self.report_warning(
840 'Long argument string detected. '
841 f'Use -- to separate parameters and URLs, like this:\n{shell_quote(correct_argv)}')
842
843 def add_info_extractor(self, ie):
844 """Add an InfoExtractor object to the end of the list."""
845 ie_key = ie.ie_key()
846 self._ies[ie_key] = ie
847 if not isinstance(ie, type):
848 self._ies_instances[ie_key] = ie
849 ie.set_downloader(self)
850
851 def get_info_extractor(self, ie_key):
852 """
853 Get an instance of an IE with name ie_key, it will try to get one from
854 the _ies list, if there's no instance it will create a new one and add
855 it to the extractor list.
856 """
857 ie = self._ies_instances.get(ie_key)
858 if ie is None:
859 ie = get_info_extractor(ie_key)()
860 self.add_info_extractor(ie)
861 return ie
862
863 def add_default_info_extractors(self):
864 """
865 Add the InfoExtractors returned by gen_extractors to the end of the list
866 """
867 all_ies = {ie.IE_NAME.lower(): ie for ie in gen_extractor_classes()}
868 all_ies['end'] = UnsupportedURLIE()
869 try:
870 ie_names = orderedSet_from_options(
871 self.params.get('allowed_extractors', ['default']), {
872 'all': list(all_ies),
873 'default': [name for name, ie in all_ies.items() if ie._ENABLED],
874 }, use_regex=True)
875 except re.error as e:
876 raise ValueError(f'Wrong regex for allowed_extractors: {e.pattern}')
877 for name in ie_names:
878 self.add_info_extractor(all_ies[name])
879 self.write_debug(f'Loaded {len(ie_names)} extractors')
880
881 def add_post_processor(self, pp, when='post_process'):
882 """Add a PostProcessor object to the end of the chain."""
883 assert when in POSTPROCESS_WHEN, f'Invalid when={when}'
884 self._pps[when].append(pp)
885 pp.set_downloader(self)
886
887 def add_post_hook(self, ph):
888 """Add the post hook"""
889 self._post_hooks.append(ph)
890
891 def add_progress_hook(self, ph):
892 """Add the download progress hook"""
893 self._progress_hooks.append(ph)
894
895 def add_postprocessor_hook(self, ph):
896 """Add the postprocessing progress hook"""
897 self._postprocessor_hooks.append(ph)
898 for pps in self._pps.values():
899 for pp in pps:
900 pp.add_progress_hook(ph)
901
902 def _bidi_workaround(self, message):
903 if not hasattr(self, '_output_channel'):
904 return message
905
906 assert hasattr(self, '_output_process')
907 assert isinstance(message, str)
908 line_count = message.count('\n') + 1
909 self._output_process.stdin.write((message + '\n').encode())
910 self._output_process.stdin.flush()
911 res = ''.join(self._output_channel.readline().decode()
912 for _ in range(line_count))
913 return res[:-len('\n')]
914
915 def _write_string(self, message, out=None, only_once=False):
916 if only_once:
917 if message in self._printed_messages:
918 return
919 self._printed_messages.add(message)
920 write_string(message, out=out, encoding=self.params.get('encoding'))
921
922 def to_stdout(self, message, skip_eol=False, quiet=None):
923 """Print message to stdout"""
924 if quiet is not None:
925 self.deprecation_warning('"YoutubeDL.to_stdout" no longer accepts the argument quiet. '
926 'Use "YoutubeDL.to_screen" instead')
927 if skip_eol is not False:
928 self.deprecation_warning('"YoutubeDL.to_stdout" no longer accepts the argument skip_eol. '
929 'Use "YoutubeDL.to_screen" instead')
930 self._write_string(f'{self._bidi_workaround(message)}\n', self._out_files.out)
931
932 def to_screen(self, message, skip_eol=False, quiet=None, only_once=False):
933 """Print message to screen if not in quiet mode"""
934 if self.params.get('logger'):
935 self.params['logger'].debug(message)
936 return
937 if (self.params.get('quiet') if quiet is None else quiet) and not self.params.get('verbose'):
938 return
939 self._write_string(
940 '{}{}'.format(self._bidi_workaround(message), ('' if skip_eol else '\n')),
941 self._out_files.screen, only_once=only_once)
942
943 def to_stderr(self, message, only_once=False):
944 """Print message to stderr"""
945 assert isinstance(message, str)
946 if self.params.get('logger'):
947 self.params['logger'].error(message)
948 else:
949 self._write_string(f'{self._bidi_workaround(message)}\n', self._out_files.error, only_once=only_once)
950
951 def _send_console_code(self, code):
952 if compat_os_name == 'nt' or not self._out_files.console:
953 return
954 self._write_string(code, self._out_files.console)
955
956 def to_console_title(self, message):
957 if not self.params.get('consoletitle', False):
958 return
959 message = remove_terminal_sequences(message)
960 if compat_os_name == 'nt':
961 if ctypes.windll.kernel32.GetConsoleWindow():
962 # c_wchar_p() might not be necessary if `message` is
963 # already of type unicode()
964 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
965 else:
966 self._send_console_code(f'\033]0;{message}\007')
967
968 def save_console_title(self):
969 if not self.params.get('consoletitle') or self.params.get('simulate'):
970 return
971 self._send_console_code('\033[22;0t') # Save the title on stack
972
973 def restore_console_title(self):
974 if not self.params.get('consoletitle') or self.params.get('simulate'):
975 return
976 self._send_console_code('\033[23;0t') # Restore the title from stack
977
978 def __enter__(self):
979 self.save_console_title()
980 return self
981
982 def save_cookies(self):
983 if self.params.get('cookiefile') is not None:
984 self.cookiejar.save()
985
986 def __exit__(self, *args):
987 self.restore_console_title()
988 self.close()
989
990 def close(self):
991 self.save_cookies()
992 if '_request_director' in self.__dict__:
993 self._request_director.close()
994 del self._request_director
995
996 def trouble(self, message=None, tb=None, is_error=True):
997 """Determine action to take when a download problem appears.
998
999 Depending on if the downloader has been configured to ignore
1000 download errors or not, this method may throw an exception or
1001 not when errors are found, after printing the message.
1002
1003 @param tb If given, is additional traceback information
1004 @param is_error Whether to raise error according to ignorerrors
1005 """
1006 if message is not None:
1007 self.to_stderr(message)
1008 if self.params.get('verbose'):
1009 if tb is None:
1010 if sys.exc_info()[0]: # if .trouble has been called from an except block
1011 tb = ''
1012 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
1013 tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
1014 tb += encode_compat_str(traceback.format_exc())
1015 else:
1016 tb_data = traceback.format_list(traceback.extract_stack())
1017 tb = ''.join(tb_data)
1018 if tb:
1019 self.to_stderr(tb)
1020 if not is_error:
1021 return
1022 if not self.params.get('ignoreerrors'):
1023 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
1024 exc_info = sys.exc_info()[1].exc_info
1025 else:
1026 exc_info = sys.exc_info()
1027 raise DownloadError(message, exc_info)
1028 self._download_retcode = 1
1029
1030 Styles = Namespace(
1031 HEADERS='yellow',
1032 EMPHASIS='light blue',
1033 FILENAME='green',
1034 ID='green',
1035 DELIM='blue',
1036 ERROR='red',
1037 BAD_FORMAT='light red',
1038 WARNING='yellow',
1039 SUPPRESS='light black',
1040 )
1041
1042 def _format_text(self, handle, allow_colors, text, f, fallback=None, *, test_encoding=False):
1043 text = str(text)
1044 if test_encoding:
1045 original_text = text
1046 # handle.encoding can be None. See https://github.com/yt-dlp/yt-dlp/issues/2711
1047 encoding = self.params.get('encoding') or getattr(handle, 'encoding', None) or 'ascii'
1048 text = text.encode(encoding, 'ignore').decode(encoding)
1049 if fallback is not None and text != original_text:
1050 text = fallback
1051 return format_text(text, f) if allow_colors is True else text if fallback is None else fallback
1052
1053 def _format_out(self, *args, **kwargs):
1054 return self._format_text(self._out_files.out, self._allow_colors.out, *args, **kwargs)
1055
1056 def _format_screen(self, *args, **kwargs):
1057 return self._format_text(self._out_files.screen, self._allow_colors.screen, *args, **kwargs)
1058
1059 def _format_err(self, *args, **kwargs):
1060 return self._format_text(self._out_files.error, self._allow_colors.error, *args, **kwargs)
1061
1062 def report_warning(self, message, only_once=False):
1063 """
1064 Print the message to stderr, it will be prefixed with 'WARNING:'
1065 If stderr is a tty file the 'WARNING:' will be colored
1066 """
1067 if self.params.get('logger') is not None:
1068 self.params['logger'].warning(message)
1069 else:
1070 if self.params.get('no_warnings'):
1071 return
1072 self.to_stderr(f'{self._format_err("WARNING:", self.Styles.WARNING)} {message}', only_once)
1073
1074 def deprecation_warning(self, message, *, stacklevel=0):
1075 deprecation_warning(
1076 message, stacklevel=stacklevel + 1, printer=self.report_error, is_error=False)
1077
1078 def deprecated_feature(self, message):
1079 if self.params.get('logger') is not None:
1080 self.params['logger'].warning(f'Deprecated Feature: {message}')
1081 self.to_stderr(f'{self._format_err("Deprecated Feature:", self.Styles.ERROR)} {message}', True)
1082
1083 def report_error(self, message, *args, **kwargs):
1084 """
1085 Do the same as trouble, but prefixes the message with 'ERROR:', colored
1086 in red if stderr is a tty file.
1087 """
1088 self.trouble(f'{self._format_err("ERROR:", self.Styles.ERROR)} {message}', *args, **kwargs)
1089
1090 def write_debug(self, message, only_once=False):
1091 """Log debug message or Print message to stderr"""
1092 if not self.params.get('verbose', False):
1093 return
1094 message = f'[debug] {message}'
1095 if self.params.get('logger'):
1096 self.params['logger'].debug(message)
1097 else:
1098 self.to_stderr(message, only_once)
1099
1100 def report_file_already_downloaded(self, file_name):
1101 """Report file has already been fully downloaded."""
1102 try:
1103 self.to_screen(f'[download] {file_name} has already been downloaded')
1104 except UnicodeEncodeError:
1105 self.to_screen('[download] The file has already been downloaded')
1106
1107 def report_file_delete(self, file_name):
1108 """Report that existing file will be deleted."""
1109 try:
1110 self.to_screen(f'Deleting existing file {file_name}')
1111 except UnicodeEncodeError:
1112 self.to_screen('Deleting existing file')
1113
1114 def raise_no_formats(self, info, forced=False, *, msg=None):
1115 has_drm = info.get('_has_drm')
1116 ignored, expected = self.params.get('ignore_no_formats_error'), bool(msg)
1117 msg = msg or has_drm and 'This video is DRM protected' or 'No video formats found!'
1118 if forced or not ignored:
1119 raise ExtractorError(msg, video_id=info['id'], ie=info['extractor'],
1120 expected=has_drm or ignored or expected)
1121 else:
1122 self.report_warning(msg)
1123
1124 def parse_outtmpl(self):
1125 self.deprecation_warning('"YoutubeDL.parse_outtmpl" is deprecated and may be removed in a future version')
1126 self._parse_outtmpl()
1127 return self.params['outtmpl']
1128
1129 def _parse_outtmpl(self):
1130 sanitize = IDENTITY
1131 if self.params.get('restrictfilenames'): # Remove spaces in the default template
1132 sanitize = lambda x: x.replace(' - ', ' ').replace(' ', '-')
1133
1134 outtmpl = self.params.setdefault('outtmpl', {})
1135 if not isinstance(outtmpl, dict):
1136 self.params['outtmpl'] = outtmpl = {'default': outtmpl}
1137 outtmpl.update({k: sanitize(v) for k, v in DEFAULT_OUTTMPL.items() if outtmpl.get(k) is None})
1138
1139 def get_output_path(self, dir_type='', filename=None):
1140 paths = self.params.get('paths', {})
1141 assert isinstance(paths, dict), '"paths" parameter must be a dictionary'
1142 path = os.path.join(
1143 expand_path(paths.get('home', '').strip()),
1144 expand_path(paths.get(dir_type, '').strip()) if dir_type else '',
1145 filename or '')
1146 return sanitize_path(path, force=self.params.get('windowsfilenames'))
1147
1148 @staticmethod
1149 def _outtmpl_expandpath(outtmpl):
1150 # expand_path translates '%%' into '%' and '$$' into '$'
1151 # correspondingly that is not what we want since we need to keep
1152 # '%%' intact for template dict substitution step. Working around
1153 # with boundary-alike separator hack.
1154 sep = ''.join(random.choices(string.ascii_letters, k=32))
1155 outtmpl = outtmpl.replace('%%', f'%{sep}%').replace('$$', f'${sep}$')
1156
1157 # outtmpl should be expand_path'ed before template dict substitution
1158 # because meta fields may contain env variables we don't want to
1159 # be expanded. E.g. for outtmpl "%(title)s.%(ext)s" and
1160 # title "Hello $PATH", we don't want `$PATH` to be expanded.
1161 return expand_path(outtmpl).replace(sep, '')
1162
1163 @staticmethod
1164 def escape_outtmpl(outtmpl):
1165 """ Escape any remaining strings like %s, %abc% etc. """
1166 return re.sub(
1167 STR_FORMAT_RE_TMPL.format('', '(?![%(\0])'),
1168 lambda mobj: ('' if mobj.group('has_key') else '%') + mobj.group(0),
1169 outtmpl)
1170
1171 @classmethod
1172 def validate_outtmpl(cls, outtmpl):
1173 """ @return None or Exception object """
1174 outtmpl = re.sub(
1175 STR_FORMAT_RE_TMPL.format('[^)]*', '[ljhqBUDS]'),
1176 lambda mobj: f'{mobj.group(0)[:-1]}s',
1177 cls._outtmpl_expandpath(outtmpl))
1178 try:
1179 cls.escape_outtmpl(outtmpl) % collections.defaultdict(int)
1180 return None
1181 except ValueError as err:
1182 return err
1183
1184 @staticmethod
1185 def _copy_infodict(info_dict):
1186 info_dict = dict(info_dict)
1187 info_dict.pop('__postprocessors', None)
1188 info_dict.pop('__pending_error', None)
1189 return info_dict
1190
1191 def prepare_outtmpl(self, outtmpl, info_dict, sanitize=False):
1192 """ Make the outtmpl and info_dict suitable for substitution: ydl.escape_outtmpl(outtmpl) % info_dict
1193 @param sanitize Whether to sanitize the output as a filename.
1194 For backward compatibility, a function can also be passed
1195 """
1196
1197 info_dict.setdefault('epoch', int(time.time())) # keep epoch consistent once set
1198
1199 info_dict = self._copy_infodict(info_dict)
1200 info_dict['duration_string'] = ( # %(duration>%H-%M-%S)s is wrong if duration > 24hrs
1201 formatSeconds(info_dict['duration'], '-' if sanitize else ':')
1202 if info_dict.get('duration', None) is not None
1203 else None)
1204 info_dict['autonumber'] = int(self.params.get('autonumber_start', 1) - 1 + self._num_downloads)
1205 info_dict['video_autonumber'] = self._num_videos
1206 if info_dict.get('resolution') is None:
1207 info_dict['resolution'] = self.format_resolution(info_dict, default=None)
1208
1209 # For fields playlist_index, playlist_autonumber and autonumber convert all occurrences
1210 # of %(field)s to %(field)0Nd for backward compatibility
1211 field_size_compat_map = {
1212 'playlist_index': number_of_digits(info_dict.get('__last_playlist_index') or 0),
1213 'playlist_autonumber': number_of_digits(info_dict.get('n_entries') or 0),
1214 'autonumber': self.params.get('autonumber_size') or 5,
1215 }
1216
1217 TMPL_DICT = {}
1218 EXTERNAL_FORMAT_RE = re.compile(STR_FORMAT_RE_TMPL.format('[^)]*', f'[{STR_FORMAT_TYPES}ljhqBUDS]'))
1219 MATH_FUNCTIONS = {
1220 '+': float.__add__,
1221 '-': float.__sub__,
1222 '*': float.__mul__,
1223 }
1224 # Field is of the form key1.key2...
1225 # where keys (except first) can be string, int, slice or "{field, ...}"
1226 FIELD_INNER_RE = r'(?:\w+|%(num)s|%(num)s?(?::%(num)s?){1,2})' % {'num': r'(?:-?\d+)'} # noqa: UP031
1227 FIELD_RE = r'\w*(?:\.(?:%(inner)s|{%(field)s(?:,%(field)s)*}))*' % { # noqa: UP031
1228 'inner': FIELD_INNER_RE,
1229 'field': rf'\w*(?:\.{FIELD_INNER_RE})*',
1230 }
1231 MATH_FIELD_RE = rf'(?:{FIELD_RE}|-?{NUMBER_RE})'
1232 MATH_OPERATORS_RE = r'(?:{})'.format('|'.join(map(re.escape, MATH_FUNCTIONS.keys())))
1233 INTERNAL_FORMAT_RE = re.compile(rf'''(?xs)
1234 (?P<negate>-)?
1235 (?P<fields>{FIELD_RE})
1236 (?P<maths>(?:{MATH_OPERATORS_RE}{MATH_FIELD_RE})*)
1237 (?:>(?P<strf_format>.+?))?
1238 (?P<remaining>
1239 (?P<alternate>(?<!\\),[^|&)]+)?
1240 (?:&(?P<replacement>.*?))?
1241 (?:\|(?P<default>.*?))?
1242 )$''')
1243
1244 def _from_user_input(field):
1245 if field == ':':
1246 return ...
1247 elif ':' in field:
1248 return slice(*map(int_or_none, field.split(':')))
1249 elif int_or_none(field) is not None:
1250 return int(field)
1251 return field
1252
1253 def _traverse_infodict(fields):
1254 fields = [f for x in re.split(r'\.({.+?})\.?', fields)
1255 for f in ([x] if x.startswith('{') else x.split('.'))]
1256 for i in (0, -1):
1257 if fields and not fields[i]:
1258 fields.pop(i)
1259
1260 for i, f in enumerate(fields):
1261 if not f.startswith('{'):
1262 fields[i] = _from_user_input(f)
1263 continue
1264 assert f.endswith('}'), f'No closing brace for {f} in {fields}'
1265 fields[i] = {k: list(map(_from_user_input, k.split('.'))) for k in f[1:-1].split(',')}
1266
1267 return traverse_obj(info_dict, fields, traverse_string=True)
1268
1269 def get_value(mdict):
1270 # Object traversal
1271 value = _traverse_infodict(mdict['fields'])
1272 # Negative
1273 if mdict['negate']:
1274 value = float_or_none(value)
1275 if value is not None:
1276 value *= -1
1277 # Do maths
1278 offset_key = mdict['maths']
1279 if offset_key:
1280 value = float_or_none(value)
1281 operator = None
1282 while offset_key:
1283 item = re.match(
1284 MATH_FIELD_RE if operator else MATH_OPERATORS_RE,
1285 offset_key).group(0)
1286 offset_key = offset_key[len(item):]
1287 if operator is None:
1288 operator = MATH_FUNCTIONS[item]
1289 continue
1290 item, multiplier = (item[1:], -1) if item[0] == '-' else (item, 1)
1291 offset = float_or_none(item)
1292 if offset is None:
1293 offset = float_or_none(_traverse_infodict(item))
1294 try:
1295 value = operator(value, multiplier * offset)
1296 except (TypeError, ZeroDivisionError):
1297 return None
1298 operator = None
1299 # Datetime formatting
1300 if mdict['strf_format']:
1301 value = strftime_or_none(value, mdict['strf_format'].replace('\\,', ','))
1302
1303 # XXX: Workaround for https://github.com/yt-dlp/yt-dlp/issues/4485
1304 if sanitize and value == '':
1305 value = None
1306 return value
1307
1308 na = self.params.get('outtmpl_na_placeholder', 'NA')
1309
1310 def filename_sanitizer(key, value, restricted=self.params.get('restrictfilenames')):
1311 return sanitize_filename(str(value), restricted=restricted, is_id=(
1312 bool(re.search(r'(^|[_.])id(\.|$)', key))
1313 if 'filename-sanitization' in self.params['compat_opts']
1314 else NO_DEFAULT))
1315
1316 sanitizer = sanitize if callable(sanitize) else filename_sanitizer
1317 sanitize = bool(sanitize)
1318
1319 def _dumpjson_default(obj):
1320 if isinstance(obj, (set, LazyList)):
1321 return list(obj)
1322 return repr(obj)
1323
1324 class _ReplacementFormatter(string.Formatter):
1325 def get_field(self, field_name, args, kwargs):
1326 if field_name.isdigit():
1327 return args[0], -1
1328 raise ValueError('Unsupported field')
1329
1330 replacement_formatter = _ReplacementFormatter()
1331
1332 def create_key(outer_mobj):
1333 if not outer_mobj.group('has_key'):
1334 return outer_mobj.group(0)
1335 key = outer_mobj.group('key')
1336 mobj = re.match(INTERNAL_FORMAT_RE, key)
1337 value, replacement, default, last_field = None, None, na, ''
1338 while mobj:
1339 mobj = mobj.groupdict()
1340 default = mobj['default'] if mobj['default'] is not None else default
1341 value = get_value(mobj)
1342 last_field, replacement = mobj['fields'], mobj['replacement']
1343 if value is None and mobj['alternate']:
1344 mobj = re.match(INTERNAL_FORMAT_RE, mobj['remaining'][1:])
1345 else:
1346 break
1347
1348 if None not in (value, replacement):
1349 try:
1350 value = replacement_formatter.format(replacement, value)
1351 except ValueError:
1352 value, default = None, na
1353
1354 fmt = outer_mobj.group('format')
1355 if fmt == 's' and last_field in field_size_compat_map and isinstance(value, int):
1356 fmt = f'0{field_size_compat_map[last_field]:d}d'
1357
1358 flags = outer_mobj.group('conversion') or ''
1359 str_fmt = f'{fmt[:-1]}s'
1360 if value is None:
1361 value, fmt = default, 's'
1362 elif fmt[-1] == 'l': # list
1363 delim = '\n' if '#' in flags else ', '
1364 value, fmt = delim.join(map(str, variadic(value, allowed_types=(str, bytes)))), str_fmt
1365 elif fmt[-1] == 'j': # json
1366 value, fmt = json.dumps(
1367 value, default=_dumpjson_default,
1368 indent=4 if '#' in flags else None, ensure_ascii='+' not in flags), str_fmt
1369 elif fmt[-1] == 'h': # html
1370 value, fmt = escapeHTML(str(value)), str_fmt
1371 elif fmt[-1] == 'q': # quoted
1372 value = map(str, variadic(value) if '#' in flags else [value])
1373 value, fmt = shell_quote(value, shell=True), str_fmt
1374 elif fmt[-1] == 'B': # bytes
1375 value = f'%{str_fmt}'.encode() % str(value).encode()
1376 value, fmt = value.decode('utf-8', 'ignore'), 's'
1377 elif fmt[-1] == 'U': # unicode normalized
1378 value, fmt = unicodedata.normalize(
1379 # "+" = compatibility equivalence, "#" = NFD
1380 'NF{}{}'.format('K' if '+' in flags else '', 'D' if '#' in flags else 'C'),
1381 value), str_fmt
1382 elif fmt[-1] == 'D': # decimal suffix
1383 num_fmt, fmt = fmt[:-1].replace('#', ''), 's'
1384 value = format_decimal_suffix(value, f'%{num_fmt}f%s' if num_fmt else '%d%s',
1385 factor=1024 if '#' in flags else 1000)
1386 elif fmt[-1] == 'S': # filename sanitization
1387 value, fmt = filename_sanitizer(last_field, value, restricted='#' in flags), str_fmt
1388 elif fmt[-1] == 'c':
1389 if value:
1390 value = str(value)[0]
1391 else:
1392 fmt = str_fmt
1393 elif fmt[-1] not in 'rsa': # numeric
1394 value = float_or_none(value)
1395 if value is None:
1396 value, fmt = default, 's'
1397
1398 if sanitize:
1399 # If value is an object, sanitize might convert it to a string
1400 # So we convert it to repr first
1401 if fmt[-1] == 'r':
1402 value, fmt = repr(value), str_fmt
1403 elif fmt[-1] == 'a':
1404 value, fmt = ascii(value), str_fmt
1405 if fmt[-1] in 'csra':
1406 value = sanitizer(last_field, value)
1407
1408 key = '{}\0{}'.format(key.replace('%', '%\0'), outer_mobj.group('format'))
1409 TMPL_DICT[key] = value
1410 return '{prefix}%({key}){fmt}'.format(key=key, fmt=fmt, prefix=outer_mobj.group('prefix'))
1411
1412 return EXTERNAL_FORMAT_RE.sub(create_key, outtmpl), TMPL_DICT
1413
1414 def evaluate_outtmpl(self, outtmpl, info_dict, *args, **kwargs):
1415 outtmpl, info_dict = self.prepare_outtmpl(outtmpl, info_dict, *args, **kwargs)
1416 return self.escape_outtmpl(outtmpl) % info_dict
1417
1418 @_catch_unsafe_extension_error
1419 def _prepare_filename(self, info_dict, *, outtmpl=None, tmpl_type=None):
1420 assert None in (outtmpl, tmpl_type), 'outtmpl and tmpl_type are mutually exclusive'
1421 if outtmpl is None:
1422 outtmpl = self.params['outtmpl'].get(tmpl_type or 'default', self.params['outtmpl']['default'])
1423 try:
1424 outtmpl = self._outtmpl_expandpath(outtmpl)
1425 filename = self.evaluate_outtmpl(outtmpl, info_dict, True)
1426 if not filename:
1427 return None
1428
1429 if tmpl_type in ('', 'temp'):
1430 final_ext, ext = self.params.get('final_ext'), info_dict.get('ext')
1431 if final_ext and ext and final_ext != ext and filename.endswith(f'.{final_ext}'):
1432 filename = replace_extension(filename, ext, final_ext)
1433 elif tmpl_type:
1434 force_ext = OUTTMPL_TYPES[tmpl_type]
1435 if force_ext:
1436 filename = replace_extension(filename, force_ext, info_dict.get('ext'))
1437
1438 # https://github.com/blackjack4494/youtube-dlc/issues/85
1439 trim_file_name = self.params.get('trim_file_name', False)
1440 if trim_file_name:
1441 no_ext, *ext = filename.rsplit('.', 2)
1442 filename = join_nonempty(no_ext[:trim_file_name], *ext, delim='.')
1443
1444 return filename
1445 except ValueError as err:
1446 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
1447 return None
1448
1449 def prepare_filename(self, info_dict, dir_type='', *, outtmpl=None, warn=False):
1450 """Generate the output filename"""
1451 if outtmpl:
1452 assert not dir_type, 'outtmpl and dir_type are mutually exclusive'
1453 dir_type = None
1454 filename = self._prepare_filename(info_dict, tmpl_type=dir_type, outtmpl=outtmpl)
1455 if not filename and dir_type not in ('', 'temp'):
1456 return ''
1457
1458 if warn:
1459 if not self.params.get('paths'):
1460 pass
1461 elif filename == '-':
1462 self.report_warning('--paths is ignored when an outputting to stdout', only_once=True)
1463 elif os.path.isabs(filename):
1464 self.report_warning('--paths is ignored since an absolute path is given in output template', only_once=True)
1465 if filename == '-' or not filename:
1466 return filename
1467
1468 return self.get_output_path(dir_type, filename)
1469
1470 def _match_entry(self, info_dict, incomplete=False, silent=False):
1471 """Returns None if the file should be downloaded"""
1472 _type = 'video' if 'playlist-match-filter' in self.params['compat_opts'] else info_dict.get('_type', 'video')
1473 assert incomplete or _type == 'video', 'Only video result can be considered complete'
1474
1475 video_title = info_dict.get('title', info_dict.get('id', 'entry'))
1476
1477 def check_filter():
1478 if _type in ('playlist', 'multi_video'):
1479 return
1480 elif _type in ('url', 'url_transparent') and not try_call(
1481 lambda: self.get_info_extractor(info_dict['ie_key']).is_single_video(info_dict['url'])):
1482 return
1483
1484 if 'title' in info_dict:
1485 # This can happen when we're just evaluating the playlist
1486 title = info_dict['title']
1487 matchtitle = self.params.get('matchtitle', False)
1488 if matchtitle:
1489 if not re.search(matchtitle, title, re.IGNORECASE):
1490 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
1491 rejecttitle = self.params.get('rejecttitle', False)
1492 if rejecttitle:
1493 if re.search(rejecttitle, title, re.IGNORECASE):
1494 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
1495
1496 date = info_dict.get('upload_date')
1497 if date is not None:
1498 date_range = self.params.get('daterange', DateRange())
1499 if date not in date_range:
1500 return f'{date_from_str(date).isoformat()} upload date is not in range {date_range}'
1501 view_count = info_dict.get('view_count')
1502 if view_count is not None:
1503 min_views = self.params.get('min_views')
1504 if min_views is not None and view_count < min_views:
1505 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
1506 max_views = self.params.get('max_views')
1507 if max_views is not None and view_count > max_views:
1508 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
1509 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
1510 return f'Skipping "{video_title}" because it is age restricted'
1511
1512 match_filter = self.params.get('match_filter')
1513 if match_filter is None:
1514 return None
1515
1516 cancelled = None
1517 try:
1518 try:
1519 ret = match_filter(info_dict, incomplete=incomplete)
1520 except TypeError:
1521 # For backward compatibility
1522 ret = None if incomplete else match_filter(info_dict)
1523 except DownloadCancelled as err:
1524 if err.msg is not NO_DEFAULT:
1525 raise
1526 ret, cancelled = err.msg, err
1527
1528 if ret is NO_DEFAULT:
1529 while True:
1530 filename = self._format_screen(self.prepare_filename(info_dict), self.Styles.FILENAME)
1531 reply = input(self._format_screen(
1532 f'Download "{filename}"? (Y/n): ', self.Styles.EMPHASIS)).lower().strip()
1533 if reply in {'y', ''}:
1534 return None
1535 elif reply == 'n':
1536 if cancelled:
1537 raise type(cancelled)(f'Skipping {video_title}')
1538 return f'Skipping {video_title}'
1539 return ret
1540
1541 if self.in_download_archive(info_dict):
1542 reason = ''.join((
1543 format_field(info_dict, 'id', f'{self._format_screen("%s", self.Styles.ID)}: '),
1544 format_field(info_dict, 'title', f'{self._format_screen("%s", self.Styles.EMPHASIS)} '),
1545 'has already been recorded in the archive'))
1546 break_opt, break_err = 'break_on_existing', ExistingVideoReached
1547 else:
1548 try:
1549 reason = check_filter()
1550 except DownloadCancelled as e:
1551 reason, break_opt, break_err = e.msg, 'match_filter', type(e)
1552 else:
1553 break_opt, break_err = 'break_on_reject', RejectedVideoReached
1554 if reason is not None:
1555 if not silent:
1556 self.to_screen('[download] ' + reason)
1557 if self.params.get(break_opt, False):
1558 raise break_err()
1559 return reason
1560
1561 @staticmethod
1562 def add_extra_info(info_dict, extra_info):
1563 """Set the keys from extra_info in info dict if they are missing"""
1564 for key, value in extra_info.items():
1565 info_dict.setdefault(key, value)
1566
1567 def extract_info(self, url, download=True, ie_key=None, extra_info=None,
1568 process=True, force_generic_extractor=False):
1569 """
1570 Extract and return the information dictionary of the URL
1571
1572 Arguments:
1573 @param url URL to extract
1574
1575 Keyword arguments:
1576 @param download Whether to download videos
1577 @param process Whether to resolve all unresolved references (URLs, playlist items).
1578 Must be True for download to work
1579 @param ie_key Use only the extractor with this key
1580
1581 @param extra_info Dictionary containing the extra values to add to the info (For internal use only)
1582 @force_generic_extractor Force using the generic extractor (Deprecated; use ie_key='Generic')
1583 """
1584
1585 if extra_info is None:
1586 extra_info = {}
1587
1588 if not ie_key and force_generic_extractor:
1589 ie_key = 'Generic'
1590
1591 if ie_key:
1592 ies = {ie_key: self._ies[ie_key]} if ie_key in self._ies else {}
1593 else:
1594 ies = self._ies
1595
1596 for key, ie in ies.items():
1597 if not ie.suitable(url):
1598 continue
1599
1600 if not ie.working():
1601 self.report_warning('The program functionality for this site has been marked as broken, '
1602 'and will probably not work.')
1603
1604 temp_id = ie.get_temp_id(url)
1605 if temp_id is not None and self.in_download_archive({'id': temp_id, 'ie_key': key}):
1606 self.to_screen(f'[download] {self._format_screen(temp_id, self.Styles.ID)}: '
1607 'has already been recorded in the archive')
1608 if self.params.get('break_on_existing', False):
1609 raise ExistingVideoReached
1610 break
1611 return self.__extract_info(url, self.get_info_extractor(key), download, extra_info, process)
1612 else:
1613 extractors_restricted = self.params.get('allowed_extractors') not in (None, ['default'])
1614 self.report_error(f'No suitable extractor{format_field(ie_key, None, " (%s)")} found for URL {url}',
1615 tb=False if extractors_restricted else None)
1616
1617 def _handle_extraction_exceptions(func):
1618 @functools.wraps(func)
1619 def wrapper(self, *args, **kwargs):
1620 while True:
1621 try:
1622 return func(self, *args, **kwargs)
1623 except (DownloadCancelled, LazyList.IndexError, PagedList.IndexError):
1624 raise
1625 except ReExtractInfo as e:
1626 if e.expected:
1627 self.to_screen(f'{e}; Re-extracting data')
1628 else:
1629 self.to_stderr('\r')
1630 self.report_warning(f'{e}; Re-extracting data')
1631 continue
1632 except GeoRestrictedError as e:
1633 msg = e.msg
1634 if e.countries:
1635 msg += '\nThis video is available in {}.'.format(', '.join(
1636 map(ISO3166Utils.short2full, e.countries)))
1637 msg += '\nYou might want to use a VPN or a proxy server (with --proxy) to workaround.'
1638 self.report_error(msg)
1639 except ExtractorError as e: # An error we somewhat expected
1640 self.report_error(str(e), e.format_traceback())
1641 except Exception as e:
1642 if self.params.get('ignoreerrors'):
1643 self.report_error(str(e), tb=encode_compat_str(traceback.format_exc()))
1644 else:
1645 raise
1646 break
1647 return wrapper
1648
1649 def _wait_for_video(self, ie_result={}):
1650 if (not self.params.get('wait_for_video')
1651 or ie_result.get('_type', 'video') != 'video'
1652 or ie_result.get('formats') or ie_result.get('url')):
1653 return
1654
1655 format_dur = lambda dur: '%02d:%02d:%02d' % timetuple_from_msec(dur * 1000)[:-1]
1656 last_msg = ''
1657
1658 def progress(msg):
1659 nonlocal last_msg
1660 full_msg = f'{msg}\n'
1661 if not self.params.get('noprogress'):
1662 full_msg = msg + ' ' * (len(last_msg) - len(msg)) + '\r'
1663 elif last_msg:
1664 return
1665 self.to_screen(full_msg, skip_eol=True)
1666 last_msg = msg
1667
1668 min_wait, max_wait = self.params.get('wait_for_video')
1669 diff = try_get(ie_result, lambda x: x['release_timestamp'] - time.time())
1670 if diff is None and ie_result.get('live_status') == 'is_upcoming':
1671 diff = round(random.uniform(min_wait, max_wait) if (max_wait and min_wait) else (max_wait or min_wait), 0)
1672 self.report_warning('Release time of video is not known')
1673 elif ie_result and (diff or 0) <= 0:
1674 self.report_warning('Video should already be available according to extracted info')
1675 diff = min(max(diff or 0, min_wait or 0), max_wait or float('inf'))
1676 self.to_screen(f'[wait] Waiting for {format_dur(diff)} - Press Ctrl+C to try now')
1677
1678 wait_till = time.time() + diff
1679 try:
1680 while True:
1681 diff = wait_till - time.time()
1682 if diff <= 0:
1683 progress('')
1684 raise ReExtractInfo('[wait] Wait period ended', expected=True)
1685 progress(f'[wait] Remaining time until next attempt: {self._format_screen(format_dur(diff), self.Styles.EMPHASIS)}')
1686 time.sleep(1)
1687 except KeyboardInterrupt:
1688 progress('')
1689 raise ReExtractInfo('[wait] Interrupted by user', expected=True)
1690 except BaseException as e:
1691 if not isinstance(e, ReExtractInfo):
1692 self.to_screen('')
1693 raise
1694
1695 def _load_cookies(self, data, *, autoscope=True):
1696 """Loads cookies from a `Cookie` header
1697
1698 This tries to work around the security vulnerability of passing cookies to every domain.
1699 See: https://github.com/yt-dlp/yt-dlp/security/advisories/GHSA-v8mc-9377-rwjj
1700
1701 @param data The Cookie header as string to load the cookies from
1702 @param autoscope If `False`, scope cookies using Set-Cookie syntax and error for cookie without domains
1703 If `True`, save cookies for later to be stored in the jar with a limited scope
1704 If a URL, save cookies in the jar with the domain of the URL
1705 """
1706 for cookie in LenientSimpleCookie(data).values():
1707 if autoscope and any(cookie.values()):
1708 raise ValueError('Invalid syntax in Cookie Header')
1709
1710 domain = cookie.get('domain') or ''
1711 expiry = cookie.get('expires')
1712 if expiry == '': # 0 is valid
1713 expiry = None
1714 prepared_cookie = http.cookiejar.Cookie(
1715 cookie.get('version') or 0, cookie.key, cookie.value, None, False,
1716 domain, True, True, cookie.get('path') or '', bool(cookie.get('path')),
1717 cookie.get('secure') or False, expiry, False, None, None, {})
1718
1719 if domain:
1720 self.cookiejar.set_cookie(prepared_cookie)
1721 elif autoscope is True:
1722 self.deprecated_feature(
1723 'Passing cookies as a header is a potential security risk; '
1724 'they will be scoped to the domain of the downloaded urls. '
1725 'Please consider loading cookies from a file or browser instead.')
1726 self.__header_cookies.append(prepared_cookie)
1727 elif autoscope:
1728 self.report_warning(
1729 'The extractor result contains an unscoped cookie as an HTTP header. '
1730 f'If you are using yt-dlp with an input URL{bug_reports_message(before=",")}',
1731 only_once=True)
1732 self._apply_header_cookies(autoscope, [prepared_cookie])
1733 else:
1734 self.report_error('Unscoped cookies are not allowed; please specify some sort of scoping',
1735 tb=False, is_error=False)
1736
1737 def _apply_header_cookies(self, url, cookies=None):
1738 """Applies stray header cookies to the provided url
1739
1740 This loads header cookies and scopes them to the domain provided in `url`.
1741 While this is not ideal, it helps reduce the risk of them being sent
1742 to an unintended destination while mostly maintaining compatibility.
1743 """
1744 parsed = urllib.parse.urlparse(url)
1745 if not parsed.hostname:
1746 return
1747
1748 for cookie in map(copy.copy, cookies or self.__header_cookies):
1749 cookie.domain = f'.{parsed.hostname}'
1750 self.cookiejar.set_cookie(cookie)
1751
1752 @_handle_extraction_exceptions
1753 def __extract_info(self, url, ie, download, extra_info, process):
1754 self._apply_header_cookies(url)
1755
1756 try:
1757 ie_result = ie.extract(url)
1758 except UserNotLive as e:
1759 if process:
1760 if self.params.get('wait_for_video'):
1761 self.report_warning(e)
1762 self._wait_for_video()
1763 raise
1764 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
1765 self.report_warning(f'Extractor {ie.IE_NAME} returned nothing{bug_reports_message()}')
1766 return
1767 if isinstance(ie_result, list):
1768 # Backwards compatibility: old IE result format
1769 ie_result = {
1770 '_type': 'compat_list',
1771 'entries': ie_result,
1772 }
1773 if extra_info.get('original_url'):
1774 ie_result.setdefault('original_url', extra_info['original_url'])
1775 self.add_default_extra_info(ie_result, ie, url)
1776 if process:
1777 self._wait_for_video(ie_result)
1778 return self.process_ie_result(ie_result, download, extra_info)
1779 else:
1780 return ie_result
1781
1782 def add_default_extra_info(self, ie_result, ie, url):
1783 if url is not None:
1784 self.add_extra_info(ie_result, {
1785 'webpage_url': url,
1786 'original_url': url,
1787 })
1788 webpage_url = ie_result.get('webpage_url')
1789 if webpage_url:
1790 self.add_extra_info(ie_result, {
1791 'webpage_url_basename': url_basename(webpage_url),
1792 'webpage_url_domain': get_domain(webpage_url),
1793 })
1794 if ie is not None:
1795 self.add_extra_info(ie_result, {
1796 'extractor': ie.IE_NAME,
1797 'extractor_key': ie.ie_key(),
1798 })
1799
1800 def process_ie_result(self, ie_result, download=True, extra_info=None):
1801 """
1802 Take the result of the ie(may be modified) and resolve all unresolved
1803 references (URLs, playlist items).
1804
1805 It will also download the videos if 'download'.
1806 Returns the resolved ie_result.
1807 """
1808 if extra_info is None:
1809 extra_info = {}
1810 result_type = ie_result.get('_type', 'video')
1811
1812 if result_type in ('url', 'url_transparent'):
1813 ie_result['url'] = sanitize_url(
1814 ie_result['url'], scheme='http' if self.params.get('prefer_insecure') else 'https')
1815 if ie_result.get('original_url') and not extra_info.get('original_url'):
1816 extra_info = {'original_url': ie_result['original_url'], **extra_info}
1817
1818 extract_flat = self.params.get('extract_flat', False)
1819 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info)
1820 or extract_flat is True):
1821 info_copy = ie_result.copy()
1822 ie = try_get(ie_result.get('ie_key'), self.get_info_extractor)
1823 if ie and not ie_result.get('id'):
1824 info_copy['id'] = ie.get_temp_id(ie_result['url'])
1825 self.add_default_extra_info(info_copy, ie, ie_result['url'])
1826 self.add_extra_info(info_copy, extra_info)
1827 info_copy, _ = self.pre_process(info_copy)
1828 self._fill_common_fields(info_copy, False)
1829 self.__forced_printings(info_copy)
1830 self._raise_pending_errors(info_copy)
1831 if self.params.get('force_write_download_archive', False):
1832 self.record_download_archive(info_copy)
1833 return ie_result
1834
1835 if result_type == 'video':
1836 self.add_extra_info(ie_result, extra_info)
1837 ie_result = self.process_video_result(ie_result, download=download)
1838 self._raise_pending_errors(ie_result)
1839 additional_urls = (ie_result or {}).get('additional_urls')
1840 if additional_urls:
1841 # TODO: Improve MetadataParserPP to allow setting a list
1842 if isinstance(additional_urls, str):
1843 additional_urls = [additional_urls]
1844 self.to_screen(
1845 '[info] {}: {} additional URL(s) requested'.format(ie_result['id'], len(additional_urls)))
1846 self.write_debug('Additional URLs: "{}"'.format('", "'.join(additional_urls)))
1847 ie_result['additional_entries'] = [
1848 self.extract_info(
1849 url, download, extra_info=extra_info,
1850 force_generic_extractor=self.params.get('force_generic_extractor'))
1851 for url in additional_urls
1852 ]
1853 return ie_result
1854 elif result_type == 'url':
1855 # We have to add extra_info to the results because it may be
1856 # contained in a playlist
1857 return self.extract_info(
1858 ie_result['url'], download,
1859 ie_key=ie_result.get('ie_key'),
1860 extra_info=extra_info)
1861 elif result_type == 'url_transparent':
1862 # Use the information from the embedding page
1863 info = self.extract_info(
1864 ie_result['url'], ie_key=ie_result.get('ie_key'),
1865 extra_info=extra_info, download=False, process=False)
1866
1867 # extract_info may return None when ignoreerrors is enabled and
1868 # extraction failed with an error, don't crash and return early
1869 # in this case
1870 if not info:
1871 return info
1872
1873 exempted_fields = {'_type', 'url', 'ie_key'}
1874 if not ie_result.get('section_end') and ie_result.get('section_start') is None:
1875 # For video clips, the id etc of the clip extractor should be used
1876 exempted_fields |= {'id', 'extractor', 'extractor_key'}
1877
1878 new_result = info.copy()
1879 new_result.update(filter_dict(ie_result, lambda k, v: v is not None and k not in exempted_fields))
1880
1881 # Extracted info may not be a video result (i.e.
1882 # info.get('_type', 'video') != video) but rather an url or
1883 # url_transparent. In such cases outer metadata (from ie_result)
1884 # should be propagated to inner one (info). For this to happen
1885 # _type of info should be overridden with url_transparent. This
1886 # fixes issue from https://github.com/ytdl-org/youtube-dl/pull/11163.
1887 if new_result.get('_type') == 'url':
1888 new_result['_type'] = 'url_transparent'
1889
1890 return self.process_ie_result(
1891 new_result, download=download, extra_info=extra_info)
1892 elif result_type in ('playlist', 'multi_video'):
1893 # Protect from infinite recursion due to recursively nested playlists
1894 # (see https://github.com/ytdl-org/youtube-dl/issues/27833)
1895 webpage_url = ie_result.get('webpage_url') # Playlists maynot have webpage_url
1896 if webpage_url and webpage_url in self._playlist_urls:
1897 self.to_screen(
1898 '[download] Skipping already downloaded playlist: {}'.format(
1899 ie_result.get('title')) or ie_result.get('id'))
1900 return
1901
1902 self._playlist_level += 1
1903 self._playlist_urls.add(webpage_url)
1904 self._fill_common_fields(ie_result, False)
1905 self._sanitize_thumbnails(ie_result)
1906 try:
1907 return self.__process_playlist(ie_result, download)
1908 finally:
1909 self._playlist_level -= 1
1910 if not self._playlist_level:
1911 self._playlist_urls.clear()
1912 elif result_type == 'compat_list':
1913 self.report_warning(
1914 'Extractor {} returned a compat_list result. '
1915 'It needs to be updated.'.format(ie_result.get('extractor')))
1916
1917 def _fixup(r):
1918 self.add_extra_info(r, {
1919 'extractor': ie_result['extractor'],
1920 'webpage_url': ie_result['webpage_url'],
1921 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1922 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1923 'extractor_key': ie_result['extractor_key'],
1924 })
1925 return r
1926 ie_result['entries'] = [
1927 self.process_ie_result(_fixup(r), download, extra_info)
1928 for r in ie_result['entries']
1929 ]
1930 return ie_result
1931 else:
1932 raise Exception(f'Invalid result type: {result_type}')
1933
1934 def _ensure_dir_exists(self, path):
1935 return make_dir(path, self.report_error)
1936
1937 @staticmethod
1938 def _playlist_infodict(ie_result, strict=False, **kwargs):
1939 info = {
1940 'playlist_count': ie_result.get('playlist_count'),
1941 'playlist': ie_result.get('title') or ie_result.get('id'),
1942 'playlist_id': ie_result.get('id'),
1943 'playlist_title': ie_result.get('title'),
1944 'playlist_uploader': ie_result.get('uploader'),
1945 'playlist_uploader_id': ie_result.get('uploader_id'),
1946 'playlist_channel': ie_result.get('channel'),
1947 'playlist_channel_id': ie_result.get('channel_id'),
1948 **kwargs,
1949 }
1950 if strict:
1951 return info
1952 if ie_result.get('webpage_url'):
1953 info.update({
1954 'webpage_url': ie_result['webpage_url'],
1955 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1956 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1957 })
1958 return {
1959 **info,
1960 'playlist_index': 0,
1961 '__last_playlist_index': max(ie_result.get('requested_entries') or (0, 0)),
1962 'extractor': ie_result['extractor'],
1963 'extractor_key': ie_result['extractor_key'],
1964 }
1965
1966 def __process_playlist(self, ie_result, download):
1967 """Process each entry in the playlist"""
1968 assert ie_result['_type'] in ('playlist', 'multi_video')
1969
1970 common_info = self._playlist_infodict(ie_result, strict=True)
1971 title = common_info.get('playlist') or '<Untitled>'
1972 if self._match_entry(common_info, incomplete=True) is not None:
1973 return
1974 self.to_screen(f'[download] Downloading {ie_result["_type"]}: {title}')
1975
1976 all_entries = PlaylistEntries(self, ie_result)
1977 entries = orderedSet(all_entries.get_requested_items(), lazy=True)
1978
1979 lazy = self.params.get('lazy_playlist')
1980 if lazy:
1981 resolved_entries, n_entries = [], 'N/A'
1982 ie_result['requested_entries'], ie_result['entries'] = None, None
1983 else:
1984 entries = resolved_entries = list(entries)
1985 n_entries = len(resolved_entries)
1986 ie_result['requested_entries'], ie_result['entries'] = tuple(zip(*resolved_entries)) or ([], [])
1987 if not ie_result.get('playlist_count'):
1988 # Better to do this after potentially exhausting entries
1989 ie_result['playlist_count'] = all_entries.get_full_count()
1990
1991 extra = self._playlist_infodict(ie_result, n_entries=int_or_none(n_entries))
1992 ie_copy = collections.ChainMap(ie_result, extra)
1993
1994 _infojson_written = False
1995 write_playlist_files = self.params.get('allow_playlist_files', True)
1996 if write_playlist_files and self.params.get('list_thumbnails'):
1997 self.list_thumbnails(ie_result)
1998 if write_playlist_files and not self.params.get('simulate'):
1999 _infojson_written = self._write_info_json(
2000 'playlist', ie_result, self.prepare_filename(ie_copy, 'pl_infojson'))
2001 if _infojson_written is None:
2002 return
2003 if self._write_description('playlist', ie_result,
2004 self.prepare_filename(ie_copy, 'pl_description')) is None:
2005 return
2006 # TODO: This should be passed to ThumbnailsConvertor if necessary
2007 self._write_thumbnails('playlist', ie_result, self.prepare_filename(ie_copy, 'pl_thumbnail'))
2008
2009 if lazy:
2010 if self.params.get('playlistreverse') or self.params.get('playlistrandom'):
2011 self.report_warning('playlistreverse and playlistrandom are not supported with lazy_playlist', only_once=True)
2012 elif self.params.get('playlistreverse'):
2013 entries.reverse()
2014 elif self.params.get('playlistrandom'):
2015 random.shuffle(entries)
2016
2017 self.to_screen(f'[{ie_result["extractor"]}] Playlist {title}: Downloading {n_entries} items'
2018 f'{format_field(ie_result, "playlist_count", " of %s")}')
2019
2020 keep_resolved_entries = self.params.get('extract_flat') != 'discard'
2021 if self.params.get('extract_flat') == 'discard_in_playlist':
2022 keep_resolved_entries = ie_result['_type'] != 'playlist'
2023 if keep_resolved_entries:
2024 self.write_debug('The information of all playlist entries will be held in memory')
2025
2026 failures = 0
2027 max_failures = self.params.get('skip_playlist_after_errors') or float('inf')
2028 for i, (playlist_index, entry) in enumerate(entries):
2029 if lazy:
2030 resolved_entries.append((playlist_index, entry))
2031 if not entry:
2032 continue
2033
2034 entry['__x_forwarded_for_ip'] = ie_result.get('__x_forwarded_for_ip')
2035 if not lazy and 'playlist-index' in self.params['compat_opts']:
2036 playlist_index = ie_result['requested_entries'][i]
2037
2038 entry_copy = collections.ChainMap(entry, {
2039 **common_info,
2040 'n_entries': int_or_none(n_entries),
2041 'playlist_index': playlist_index,
2042 'playlist_autonumber': i + 1,
2043 })
2044
2045 if self._match_entry(entry_copy, incomplete=True) is not None:
2046 # For compatabilty with youtube-dl. See https://github.com/yt-dlp/yt-dlp/issues/4369
2047 resolved_entries[i] = (playlist_index, NO_DEFAULT)
2048 continue
2049
2050 self.to_screen(
2051 f'[download] Downloading item {self._format_screen(i + 1, self.Styles.ID)} '
2052 f'of {self._format_screen(n_entries, self.Styles.EMPHASIS)}')
2053
2054 entry_result = self.__process_iterable_entry(entry, download, collections.ChainMap({
2055 'playlist_index': playlist_index,
2056 'playlist_autonumber': i + 1,
2057 }, extra))
2058 if not entry_result:
2059 failures += 1
2060 if failures >= max_failures:
2061 self.report_error(
2062 f'Skipping the remaining entries in playlist "{title}" since {failures} items failed extraction')
2063 break
2064 if keep_resolved_entries:
2065 resolved_entries[i] = (playlist_index, entry_result)
2066
2067 # Update with processed data
2068 ie_result['entries'] = [e for _, e in resolved_entries if e is not NO_DEFAULT]
2069 ie_result['requested_entries'] = [i for i, e in resolved_entries if e is not NO_DEFAULT]
2070 if ie_result['requested_entries'] == try_call(lambda: list(range(1, ie_result['playlist_count'] + 1))):
2071 # Do not set for full playlist
2072 ie_result.pop('requested_entries')
2073
2074 # Write the updated info to json
2075 if _infojson_written is True and self._write_info_json(
2076 'updated playlist', ie_result,
2077 self.prepare_filename(ie_copy, 'pl_infojson'), overwrite=True) is None:
2078 return
2079
2080 ie_result = self.run_all_pps('playlist', ie_result)
2081 self.to_screen(f'[download] Finished downloading playlist: {title}')
2082 return ie_result
2083
2084 @_handle_extraction_exceptions
2085 def __process_iterable_entry(self, entry, download, extra_info):
2086 return self.process_ie_result(
2087 entry, download=download, extra_info=extra_info)
2088
2089 def _build_format_filter(self, filter_spec):
2090 " Returns a function to filter the formats according to the filter_spec "
2091
2092 OPERATORS = {
2093 '<': operator.lt,
2094 '<=': operator.le,
2095 '>': operator.gt,
2096 '>=': operator.ge,
2097 '=': operator.eq,
2098 '!=': operator.ne,
2099 }
2100 operator_rex = re.compile(r'''(?x)\s*
2101 (?P<key>[\w.-]+)\s*
2102 (?P<op>{})(?P<none_inclusive>\s*\?)?\s*
2103 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)\s*
2104 '''.format('|'.join(map(re.escape, OPERATORS.keys()))))
2105 m = operator_rex.fullmatch(filter_spec)
2106 if m:
2107 try:
2108 comparison_value = int(m.group('value'))
2109 except ValueError:
2110 comparison_value = parse_filesize(m.group('value'))
2111 if comparison_value is None:
2112 comparison_value = parse_filesize(m.group('value') + 'B')
2113 if comparison_value is None:
2114 raise ValueError(
2115 'Invalid value {!r} in format specification {!r}'.format(
2116 m.group('value'), filter_spec))
2117 op = OPERATORS[m.group('op')]
2118
2119 if not m:
2120 STR_OPERATORS = {
2121 '=': operator.eq,
2122 '^=': lambda attr, value: attr.startswith(value),
2123 '$=': lambda attr, value: attr.endswith(value),
2124 '*=': lambda attr, value: value in attr,
2125 '~=': lambda attr, value: value.search(attr) is not None,
2126 }
2127 str_operator_rex = re.compile(r'''(?x)\s*
2128 (?P<key>[a-zA-Z0-9._-]+)\s*
2129 (?P<negation>!\s*)?(?P<op>{})\s*(?P<none_inclusive>\?\s*)?
2130 (?P<quote>["'])?
2131 (?P<value>(?(quote)(?:(?!(?P=quote))[^\\]|\\.)+|[\w.-]+))
2132 (?(quote)(?P=quote))\s*
2133 '''.format('|'.join(map(re.escape, STR_OPERATORS.keys()))))
2134 m = str_operator_rex.fullmatch(filter_spec)
2135 if m:
2136 if m.group('op') == '~=':
2137 comparison_value = re.compile(m.group('value'))
2138 else:
2139 comparison_value = re.sub(r'''\\([\\"'])''', r'\1', m.group('value'))
2140 str_op = STR_OPERATORS[m.group('op')]
2141 if m.group('negation'):
2142 op = lambda attr, value: not str_op(attr, value)
2143 else:
2144 op = str_op
2145
2146 if not m:
2147 raise SyntaxError(f'Invalid filter specification {filter_spec!r}')
2148
2149 def _filter(f):
2150 actual_value = f.get(m.group('key'))
2151 if actual_value is None:
2152 return m.group('none_inclusive')
2153 return op(actual_value, comparison_value)
2154 return _filter
2155
2156 def _check_formats(self, formats):
2157 for f in formats:
2158 working = f.get('__working')
2159 if working is not None:
2160 if working:
2161 yield f
2162 continue
2163 self.to_screen('[info] Testing format {}'.format(f['format_id']))
2164 path = self.get_output_path('temp')
2165 if not self._ensure_dir_exists(f'{path}/'):
2166 continue
2167 temp_file = tempfile.NamedTemporaryFile(suffix='.tmp', delete=False, dir=path or None)
2168 temp_file.close()
2169 try:
2170 success, _ = self.dl(temp_file.name, f, test=True)
2171 except (DownloadError, OSError, ValueError, *network_exceptions):
2172 success = False
2173 finally:
2174 if os.path.exists(temp_file.name):
2175 try:
2176 os.remove(temp_file.name)
2177 except OSError:
2178 self.report_warning(f'Unable to delete temporary file "{temp_file.name}"')
2179 f['__working'] = success
2180 if success:
2181 yield f
2182 else:
2183 self.to_screen('[info] Unable to download format {}. Skipping...'.format(f['format_id']))
2184
2185 def _select_formats(self, formats, selector):
2186 return list(selector({
2187 'formats': formats,
2188 'has_merged_format': any('none' not in (f.get('acodec'), f.get('vcodec')) for f in formats),
2189 'incomplete_formats': (all(f.get('vcodec') == 'none' for f in formats) # No formats with video
2190 or all(f.get('acodec') == 'none' for f in formats)), # OR, No formats with audio
2191 }))
2192
2193 def _default_format_spec(self, info_dict, download=True):
2194 download = download and not self.params.get('simulate')
2195 prefer_best = download and (
2196 self.params['outtmpl']['default'] == '-'
2197 or info_dict.get('is_live') and not self.params.get('live_from_start'))
2198
2199 def can_merge():
2200 merger = FFmpegMergerPP(self)
2201 return merger.available and merger.can_merge()
2202
2203 if not prefer_best and download and not can_merge():
2204 prefer_best = True
2205 formats = self._get_formats(info_dict)
2206 evaluate_formats = lambda spec: self._select_formats(formats, self.build_format_selector(spec))
2207 if evaluate_formats('b/bv+ba') != evaluate_formats('bv*+ba/b'):
2208 self.report_warning('ffmpeg not found. The downloaded format may not be the best available. '
2209 'Installing ffmpeg is strongly recommended: https://github.com/yt-dlp/yt-dlp#dependencies')
2210
2211 compat = (self.params.get('allow_multiple_audio_streams')
2212 or 'format-spec' in self.params['compat_opts'])
2213
2214 return ('best/bestvideo+bestaudio' if prefer_best
2215 else 'bestvideo+bestaudio/best' if compat
2216 else 'bestvideo*+bestaudio/best')
2217
2218 def build_format_selector(self, format_spec):
2219 def syntax_error(note, start):
2220 message = (
2221 'Invalid format specification: '
2222 '{}\n\t{}\n\t{}^'.format(note, format_spec, ' ' * start[1]))
2223 return SyntaxError(message)
2224
2225 PICKFIRST = 'PICKFIRST'
2226 MERGE = 'MERGE'
2227 SINGLE = 'SINGLE'
2228 GROUP = 'GROUP'
2229 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
2230
2231 allow_multiple_streams = {'audio': self.params.get('allow_multiple_audio_streams', False),
2232 'video': self.params.get('allow_multiple_video_streams', False)}
2233
2234 def _parse_filter(tokens):
2235 filter_parts = []
2236 for type_, string_, _start, _, _ in tokens:
2237 if type_ == tokenize.OP and string_ == ']':
2238 return ''.join(filter_parts)
2239 else:
2240 filter_parts.append(string_)
2241
2242 def _remove_unused_ops(tokens):
2243 # Remove operators that we don't use and join them with the surrounding strings.
2244 # E.g. 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
2245 ALLOWED_OPS = ('/', '+', ',', '(', ')')
2246 last_string, last_start, last_end, last_line = None, None, None, None
2247 for type_, string_, start, end, line in tokens:
2248 if type_ == tokenize.OP and string_ == '[':
2249 if last_string:
2250 yield tokenize.NAME, last_string, last_start, last_end, last_line
2251 last_string = None
2252 yield type_, string_, start, end, line
2253 # everything inside brackets will be handled by _parse_filter
2254 for type_, string_, start, end, line in tokens:
2255 yield type_, string_, start, end, line
2256 if type_ == tokenize.OP and string_ == ']':
2257 break
2258 elif type_ == tokenize.OP and string_ in ALLOWED_OPS:
2259 if last_string:
2260 yield tokenize.NAME, last_string, last_start, last_end, last_line
2261 last_string = None
2262 yield type_, string_, start, end, line
2263 elif type_ in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
2264 if not last_string:
2265 last_string = string_
2266 last_start = start
2267 last_end = end
2268 else:
2269 last_string += string_
2270 if last_string:
2271 yield tokenize.NAME, last_string, last_start, last_end, last_line
2272
2273 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
2274 selectors = []
2275 current_selector = None
2276 for type_, string_, start, _, _ in tokens:
2277 # ENCODING is only defined in Python 3.x
2278 if type_ == getattr(tokenize, 'ENCODING', None):
2279 continue
2280 elif type_ in [tokenize.NAME, tokenize.NUMBER]:
2281 current_selector = FormatSelector(SINGLE, string_, [])
2282 elif type_ == tokenize.OP:
2283 if string_ == ')':
2284 if not inside_group:
2285 # ')' will be handled by the parentheses group
2286 tokens.restore_last_token()
2287 break
2288 elif inside_merge and string_ in ['/', ',']:
2289 tokens.restore_last_token()
2290 break
2291 elif inside_choice and string_ == ',':
2292 tokens.restore_last_token()
2293 break
2294 elif string_ == ',':
2295 if not current_selector:
2296 raise syntax_error('"," must follow a format selector', start)
2297 selectors.append(current_selector)
2298 current_selector = None
2299 elif string_ == '/':
2300 if not current_selector:
2301 raise syntax_error('"/" must follow a format selector', start)
2302 first_choice = current_selector
2303 second_choice = _parse_format_selection(tokens, inside_choice=True)
2304 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
2305 elif string_ == '[':
2306 if not current_selector:
2307 current_selector = FormatSelector(SINGLE, 'best', [])
2308 format_filter = _parse_filter(tokens)
2309 current_selector.filters.append(format_filter)
2310 elif string_ == '(':
2311 if current_selector:
2312 raise syntax_error('Unexpected "("', start)
2313 group = _parse_format_selection(tokens, inside_group=True)
2314 current_selector = FormatSelector(GROUP, group, [])
2315 elif string_ == '+':
2316 if not current_selector:
2317 raise syntax_error('Unexpected "+"', start)
2318 selector_1 = current_selector
2319 selector_2 = _parse_format_selection(tokens, inside_merge=True)
2320 if not selector_2:
2321 raise syntax_error('Expected a selector', start)
2322 current_selector = FormatSelector(MERGE, (selector_1, selector_2), [])
2323 else:
2324 raise syntax_error(f'Operator not recognized: "{string_}"', start)
2325 elif type_ == tokenize.ENDMARKER:
2326 break
2327 if current_selector:
2328 selectors.append(current_selector)
2329 return selectors
2330
2331 def _merge(formats_pair):
2332 format_1, format_2 = formats_pair
2333
2334 formats_info = []
2335 formats_info.extend(format_1.get('requested_formats', (format_1,)))
2336 formats_info.extend(format_2.get('requested_formats', (format_2,)))
2337
2338 if not allow_multiple_streams['video'] or not allow_multiple_streams['audio']:
2339 get_no_more = {'video': False, 'audio': False}
2340 for (i, fmt_info) in enumerate(formats_info):
2341 if fmt_info.get('acodec') == fmt_info.get('vcodec') == 'none':
2342 formats_info.pop(i)
2343 continue
2344 for aud_vid in ['audio', 'video']:
2345 if not allow_multiple_streams[aud_vid] and fmt_info.get(aud_vid[0] + 'codec') != 'none':
2346 if get_no_more[aud_vid]:
2347 formats_info.pop(i)
2348 break
2349 get_no_more[aud_vid] = True
2350
2351 if len(formats_info) == 1:
2352 return formats_info[0]
2353
2354 video_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('vcodec') != 'none']
2355 audio_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('acodec') != 'none']
2356
2357 the_only_video = video_fmts[0] if len(video_fmts) == 1 else None
2358 the_only_audio = audio_fmts[0] if len(audio_fmts) == 1 else None
2359
2360 output_ext = get_compatible_ext(
2361 vcodecs=[f.get('vcodec') for f in video_fmts],
2362 acodecs=[f.get('acodec') for f in audio_fmts],
2363 vexts=[f['ext'] for f in video_fmts],
2364 aexts=[f['ext'] for f in audio_fmts],
2365 preferences=(try_call(lambda: self.params['merge_output_format'].split('/'))
2366 or self.params.get('prefer_free_formats') and ('webm', 'mkv')))
2367
2368 filtered = lambda *keys: filter(None, (traverse_obj(fmt, *keys) for fmt in formats_info))
2369
2370 new_dict = {
2371 'requested_formats': formats_info,
2372 'format': '+'.join(filtered('format')),
2373 'format_id': '+'.join(filtered('format_id')),
2374 'ext': output_ext,
2375 'protocol': '+'.join(map(determine_protocol, formats_info)),
2376 'language': '+'.join(orderedSet(filtered('language'))) or None,
2377 'format_note': '+'.join(orderedSet(filtered('format_note'))) or None,
2378 'filesize_approx': sum(filtered('filesize', 'filesize_approx')) or None,
2379 'tbr': sum(filtered('tbr', 'vbr', 'abr')),
2380 }
2381
2382 if the_only_video:
2383 new_dict.update({
2384 'width': the_only_video.get('width'),
2385 'height': the_only_video.get('height'),
2386 'resolution': the_only_video.get('resolution') or self.format_resolution(the_only_video),
2387 'fps': the_only_video.get('fps'),
2388 'dynamic_range': the_only_video.get('dynamic_range'),
2389 'vcodec': the_only_video.get('vcodec'),
2390 'vbr': the_only_video.get('vbr'),
2391 'stretched_ratio': the_only_video.get('stretched_ratio'),
2392 'aspect_ratio': the_only_video.get('aspect_ratio'),
2393 })
2394
2395 if the_only_audio:
2396 new_dict.update({
2397 'acodec': the_only_audio.get('acodec'),
2398 'abr': the_only_audio.get('abr'),
2399 'asr': the_only_audio.get('asr'),
2400 'audio_channels': the_only_audio.get('audio_channels'),
2401 })
2402
2403 return new_dict
2404
2405 def _check_formats(formats):
2406 if self.params.get('check_formats') == 'selected':
2407 yield from self._check_formats(formats)
2408 return
2409 elif (self.params.get('check_formats') is not None
2410 or self.params.get('allow_unplayable_formats')):
2411 yield from formats
2412 return
2413
2414 for f in formats:
2415 if f.get('has_drm') or f.get('__needs_testing'):
2416 yield from self._check_formats([f])
2417 else:
2418 yield f
2419
2420 def _build_selector_function(selector):
2421 if isinstance(selector, list): # ,
2422 fs = [_build_selector_function(s) for s in selector]
2423
2424 def selector_function(ctx):
2425 for f in fs:
2426 yield from f(ctx)
2427 return selector_function
2428
2429 elif selector.type == GROUP: # ()
2430 selector_function = _build_selector_function(selector.selector)
2431
2432 elif selector.type == PICKFIRST: # /
2433 fs = [_build_selector_function(s) for s in selector.selector]
2434
2435 def selector_function(ctx):
2436 for f in fs:
2437 picked_formats = list(f(ctx))
2438 if picked_formats:
2439 return picked_formats
2440 return []
2441
2442 elif selector.type == MERGE: # +
2443 selector_1, selector_2 = map(_build_selector_function, selector.selector)
2444
2445 def selector_function(ctx):
2446 for pair in itertools.product(selector_1(ctx), selector_2(ctx)):
2447 yield _merge(pair)
2448
2449 elif selector.type == SINGLE: # atom
2450 format_spec = selector.selector or 'best'
2451
2452 # TODO: Add allvideo, allaudio etc by generalizing the code with best/worst selector
2453 if format_spec == 'all':
2454 def selector_function(ctx):
2455 yield from _check_formats(ctx['formats'][::-1])
2456 elif format_spec == 'mergeall':
2457 def selector_function(ctx):
2458 formats = list(_check_formats(
2459 f for f in ctx['formats'] if f.get('vcodec') != 'none' or f.get('acodec') != 'none'))
2460 if not formats:
2461 return
2462 merged_format = formats[-1]
2463 for f in formats[-2::-1]:
2464 merged_format = _merge((merged_format, f))
2465 yield merged_format
2466
2467 else:
2468 format_fallback, seperate_fallback, format_reverse, format_idx = False, None, True, 1
2469 mobj = re.match(
2470 r'(?P<bw>best|worst|b|w)(?P<type>video|audio|v|a)?(?P<mod>\*)?(?:\.(?P<n>[1-9]\d*))?$',
2471 format_spec)
2472 if mobj is not None:
2473 format_idx = int_or_none(mobj.group('n'), default=1)
2474 format_reverse = mobj.group('bw')[0] == 'b'
2475 format_type = (mobj.group('type') or [None])[0]
2476 not_format_type = {'v': 'a', 'a': 'v'}.get(format_type)
2477 format_modified = mobj.group('mod') is not None
2478
2479 format_fallback = not format_type and not format_modified # for b, w
2480 _filter_f = (
2481 (lambda f: f.get(f'{format_type}codec') != 'none')
2482 if format_type and format_modified # bv*, ba*, wv*, wa*
2483 else (lambda f: f.get(f'{not_format_type}codec') == 'none')
2484 if format_type # bv, ba, wv, wa
2485 else (lambda f: f.get('vcodec') != 'none' and f.get('acodec') != 'none')
2486 if not format_modified # b, w
2487 else lambda f: True) # b*, w*
2488 filter_f = lambda f: _filter_f(f) and (
2489 f.get('vcodec') != 'none' or f.get('acodec') != 'none')
2490 else:
2491 if format_spec in self._format_selection_exts['audio']:
2492 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none'
2493 elif format_spec in self._format_selection_exts['video']:
2494 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none' and f.get('vcodec') != 'none'
2495 seperate_fallback = lambda f: f.get('ext') == format_spec and f.get('vcodec') != 'none'
2496 elif format_spec in self._format_selection_exts['storyboards']:
2497 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') == 'none' and f.get('vcodec') == 'none'
2498 else:
2499 filter_f = lambda f: f.get('format_id') == format_spec # id
2500
2501 def selector_function(ctx):
2502 formats = list(ctx['formats'])
2503 matches = list(filter(filter_f, formats)) if filter_f is not None else formats
2504 if not matches:
2505 if format_fallback and ctx['incomplete_formats']:
2506 # for extractors with incomplete formats (audio only (soundcloud)
2507 # or video only (imgur)) best/worst will fallback to
2508 # best/worst {video,audio}-only format
2509 matches = list(filter(lambda f: f.get('vcodec') != 'none' or f.get('acodec') != 'none', formats))
2510 elif seperate_fallback and not ctx['has_merged_format']:
2511 # for compatibility with youtube-dl when there is no pre-merged format
2512 matches = list(filter(seperate_fallback, formats))
2513 matches = LazyList(_check_formats(matches[::-1 if format_reverse else 1]))
2514 try:
2515 yield matches[format_idx - 1]
2516 except LazyList.IndexError:
2517 return
2518
2519 filters = [self._build_format_filter(f) for f in selector.filters]
2520
2521 def final_selector(ctx):
2522 ctx_copy = dict(ctx)
2523 for _filter in filters:
2524 ctx_copy['formats'] = list(filter(_filter, ctx_copy['formats']))
2525 return selector_function(ctx_copy)
2526 return final_selector
2527
2528 # HACK: Python 3.12 changed the underlying parser, rendering '7_a' invalid
2529 # Prefix numbers with random letters to avoid it being classified as a number
2530 # See: https://github.com/yt-dlp/yt-dlp/pulls/8797
2531 # TODO: Implement parser not reliant on tokenize.tokenize
2532 prefix = ''.join(random.choices(string.ascii_letters, k=32))
2533 stream = io.BytesIO(re.sub(r'\d[_\d]*', rf'{prefix}\g<0>', format_spec).encode())
2534 try:
2535 tokens = list(_remove_unused_ops(
2536 token._replace(string=token.string.replace(prefix, ''))
2537 for token in tokenize.tokenize(stream.readline)))
2538 except tokenize.TokenError:
2539 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
2540
2541 class TokenIterator:
2542 def __init__(self, tokens):
2543 self.tokens = tokens
2544 self.counter = 0
2545
2546 def __iter__(self):
2547 return self
2548
2549 def __next__(self):
2550 if self.counter >= len(self.tokens):
2551 raise StopIteration
2552 value = self.tokens[self.counter]
2553 self.counter += 1
2554 return value
2555
2556 next = __next__
2557
2558 def restore_last_token(self):
2559 self.counter -= 1
2560
2561 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
2562 return _build_selector_function(parsed_selector)
2563
2564 def _calc_headers(self, info_dict, load_cookies=False):
2565 res = HTTPHeaderDict(self.params['http_headers'], info_dict.get('http_headers'))
2566 clean_headers(res)
2567
2568 if load_cookies: # For --load-info-json
2569 self._load_cookies(res.get('Cookie'), autoscope=info_dict['url']) # compat
2570 self._load_cookies(info_dict.get('cookies'), autoscope=False)
2571 # The `Cookie` header is removed to prevent leaks and unscoped cookies.
2572 # See: https://github.com/yt-dlp/yt-dlp/security/advisories/GHSA-v8mc-9377-rwjj
2573 res.pop('Cookie', None)
2574 cookies = self.cookiejar.get_cookies_for_url(info_dict['url'])
2575 if cookies:
2576 encoder = LenientSimpleCookie()
2577 values = []
2578 for cookie in cookies:
2579 _, value = encoder.value_encode(cookie.value)
2580 values.append(f'{cookie.name}={value}')
2581 if cookie.domain:
2582 values.append(f'Domain={cookie.domain}')
2583 if cookie.path:
2584 values.append(f'Path={cookie.path}')
2585 if cookie.secure:
2586 values.append('Secure')
2587 if cookie.expires:
2588 values.append(f'Expires={cookie.expires}')
2589 if cookie.version:
2590 values.append(f'Version={cookie.version}')
2591 info_dict['cookies'] = '; '.join(values)
2592
2593 if 'X-Forwarded-For' not in res:
2594 x_forwarded_for_ip = info_dict.get('__x_forwarded_for_ip')
2595 if x_forwarded_for_ip:
2596 res['X-Forwarded-For'] = x_forwarded_for_ip
2597
2598 return res
2599
2600 def _calc_cookies(self, url):
2601 self.deprecation_warning('"YoutubeDL._calc_cookies" is deprecated and may be removed in a future version')
2602 return self.cookiejar.get_cookie_header(url)
2603
2604 def _sort_thumbnails(self, thumbnails):
2605 thumbnails.sort(key=lambda t: (
2606 t.get('preference') if t.get('preference') is not None else -1,
2607 t.get('width') if t.get('width') is not None else -1,
2608 t.get('height') if t.get('height') is not None else -1,
2609 t.get('id') if t.get('id') is not None else '',
2610 t.get('url')))
2611
2612 def _sanitize_thumbnails(self, info_dict):
2613 thumbnails = info_dict.get('thumbnails')
2614 if thumbnails is None:
2615 thumbnail = info_dict.get('thumbnail')
2616 if thumbnail:
2617 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
2618 if not thumbnails:
2619 return
2620
2621 def check_thumbnails(thumbnails):
2622 for t in thumbnails:
2623 self.to_screen(f'[info] Testing thumbnail {t["id"]}')
2624 try:
2625 self.urlopen(HEADRequest(t['url']))
2626 except network_exceptions as err:
2627 self.to_screen(f'[info] Unable to connect to thumbnail {t["id"]} URL {t["url"]!r} - {err}. Skipping...')
2628 continue
2629 yield t
2630
2631 self._sort_thumbnails(thumbnails)
2632 for i, t in enumerate(thumbnails):
2633 if t.get('id') is None:
2634 t['id'] = str(i)
2635 if t.get('width') and t.get('height'):
2636 t['resolution'] = '%dx%d' % (t['width'], t['height'])
2637 t['url'] = sanitize_url(t['url'])
2638
2639 if self.params.get('check_formats') is True:
2640 info_dict['thumbnails'] = LazyList(check_thumbnails(thumbnails[::-1]), reverse=True)
2641 else:
2642 info_dict['thumbnails'] = thumbnails
2643
2644 def _fill_common_fields(self, info_dict, final=True):
2645 # TODO: move sanitization here
2646 if final:
2647 title = info_dict['fulltitle'] = info_dict.get('title')
2648 if not title:
2649 if title == '':
2650 self.write_debug('Extractor gave empty title. Creating a generic title')
2651 else:
2652 self.report_warning('Extractor failed to obtain "title". Creating a generic title instead')
2653 info_dict['title'] = f'{info_dict["extractor"].replace(":", "-")} video #{info_dict["id"]}'
2654
2655 if info_dict.get('duration') is not None:
2656 info_dict['duration_string'] = formatSeconds(info_dict['duration'])
2657
2658 for ts_key, date_key in (
2659 ('timestamp', 'upload_date'),
2660 ('release_timestamp', 'release_date'),
2661 ('modified_timestamp', 'modified_date'),
2662 ):
2663 if info_dict.get(date_key) is None and info_dict.get(ts_key) is not None:
2664 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
2665 # see http://bugs.python.org/issue1646728)
2666 with contextlib.suppress(ValueError, OverflowError, OSError):
2667 upload_date = dt.datetime.fromtimestamp(info_dict[ts_key], dt.timezone.utc)
2668 info_dict[date_key] = upload_date.strftime('%Y%m%d')
2669
2670 if not info_dict.get('release_year'):
2671 info_dict['release_year'] = traverse_obj(info_dict, ('release_date', {lambda x: int(x[:4])}))
2672
2673 live_keys = ('is_live', 'was_live')
2674 live_status = info_dict.get('live_status')
2675 if live_status is None:
2676 for key in live_keys:
2677 if info_dict.get(key) is False:
2678 continue
2679 if info_dict.get(key):
2680 live_status = key
2681 break
2682 if all(info_dict.get(key) is False for key in live_keys):
2683 live_status = 'not_live'
2684 if live_status:
2685 info_dict['live_status'] = live_status
2686 for key in live_keys:
2687 if info_dict.get(key) is None:
2688 info_dict[key] = (live_status == key)
2689 if live_status == 'post_live':
2690 info_dict['was_live'] = True
2691
2692 # Auto generate title fields corresponding to the *_number fields when missing
2693 # in order to always have clean titles. This is very common for TV series.
2694 for field in ('chapter', 'season', 'episode'):
2695 if final and info_dict.get(f'{field}_number') is not None and not info_dict.get(field):
2696 info_dict[field] = '%s %d' % (field.capitalize(), info_dict[f'{field}_number'])
2697
2698 for old_key, new_key in self._deprecated_multivalue_fields.items():
2699 if new_key in info_dict and old_key in info_dict:
2700 if '_version' not in info_dict: # HACK: Do not warn when using --load-info-json
2701 self.deprecation_warning(f'Do not return {old_key!r} when {new_key!r} is present')
2702 elif old_value := info_dict.get(old_key):
2703 info_dict[new_key] = old_value.split(', ')
2704 elif new_value := info_dict.get(new_key):
2705 info_dict[old_key] = ', '.join(v.replace(',', '\N{FULLWIDTH COMMA}') for v in new_value)
2706
2707 def _raise_pending_errors(self, info):
2708 err = info.pop('__pending_error', None)
2709 if err:
2710 self.report_error(err, tb=False)
2711
2712 def sort_formats(self, info_dict):
2713 formats = self._get_formats(info_dict)
2714 formats.sort(key=FormatSorter(
2715 self, info_dict.get('_format_sort_fields') or []).calculate_preference)
2716
2717 def process_video_result(self, info_dict, download=True):
2718 assert info_dict.get('_type', 'video') == 'video'
2719 self._num_videos += 1
2720
2721 if 'id' not in info_dict:
2722 raise ExtractorError('Missing "id" field in extractor result', ie=info_dict['extractor'])
2723 elif not info_dict.get('id'):
2724 raise ExtractorError('Extractor failed to obtain "id"', ie=info_dict['extractor'])
2725
2726 def report_force_conversion(field, field_not, conversion):
2727 self.report_warning(
2728 f'"{field}" field is not {field_not} - forcing {conversion} conversion, '
2729 'there is an error in extractor')
2730
2731 def sanitize_string_field(info, string_field):
2732 field = info.get(string_field)
2733 if field is None or isinstance(field, str):
2734 return
2735 report_force_conversion(string_field, 'a string', 'string')
2736 info[string_field] = str(field)
2737
2738 def sanitize_numeric_fields(info):
2739 for numeric_field in self._NUMERIC_FIELDS:
2740 field = info.get(numeric_field)
2741 if field is None or isinstance(field, (int, float)):
2742 continue
2743 report_force_conversion(numeric_field, 'numeric', 'int')
2744 info[numeric_field] = int_or_none(field)
2745
2746 sanitize_string_field(info_dict, 'id')
2747 sanitize_numeric_fields(info_dict)
2748 if info_dict.get('section_end') and info_dict.get('section_start') is not None:
2749 info_dict['duration'] = round(info_dict['section_end'] - info_dict['section_start'], 3)
2750 if (info_dict.get('duration') or 0) <= 0 and info_dict.pop('duration', None):
2751 self.report_warning('"duration" field is negative, there is an error in extractor')
2752
2753 chapters = info_dict.get('chapters') or []
2754 if chapters and chapters[0].get('start_time'):
2755 chapters.insert(0, {'start_time': 0})
2756
2757 dummy_chapter = {'end_time': 0, 'start_time': info_dict.get('duration')}
2758 for idx, (prev, current, next_) in enumerate(zip(
2759 (dummy_chapter, *chapters), chapters, (*chapters[1:], dummy_chapter)), 1):
2760 if current.get('start_time') is None:
2761 current['start_time'] = prev.get('end_time')
2762 if not current.get('end_time'):
2763 current['end_time'] = next_.get('start_time')
2764 if not current.get('title'):
2765 current['title'] = f'<Untitled Chapter {idx}>'
2766
2767 if 'playlist' not in info_dict:
2768 # It isn't part of a playlist
2769 info_dict['playlist'] = None
2770 info_dict['playlist_index'] = None
2771
2772 self._sanitize_thumbnails(info_dict)
2773
2774 thumbnail = info_dict.get('thumbnail')
2775 thumbnails = info_dict.get('thumbnails')
2776 if thumbnail:
2777 info_dict['thumbnail'] = sanitize_url(thumbnail)
2778 elif thumbnails:
2779 info_dict['thumbnail'] = thumbnails[-1]['url']
2780
2781 if info_dict.get('display_id') is None and 'id' in info_dict:
2782 info_dict['display_id'] = info_dict['id']
2783
2784 self._fill_common_fields(info_dict)
2785
2786 for cc_kind in ('subtitles', 'automatic_captions'):
2787 cc = info_dict.get(cc_kind)
2788 if cc:
2789 for _, subtitle in cc.items():
2790 for subtitle_format in subtitle:
2791 if subtitle_format.get('url'):
2792 subtitle_format['url'] = sanitize_url(subtitle_format['url'])
2793 if subtitle_format.get('ext') is None:
2794 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
2795
2796 automatic_captions = info_dict.get('automatic_captions')
2797 subtitles = info_dict.get('subtitles')
2798
2799 info_dict['requested_subtitles'] = self.process_subtitles(
2800 info_dict['id'], subtitles, automatic_captions)
2801
2802 formats = self._get_formats(info_dict)
2803
2804 # Backward compatibility with InfoExtractor._sort_formats
2805 field_preference = (formats or [{}])[0].pop('__sort_fields', None)
2806 if field_preference:
2807 info_dict['_format_sort_fields'] = field_preference
2808
2809 info_dict['_has_drm'] = any( # or None ensures --clean-infojson removes it
2810 f.get('has_drm') and f['has_drm'] != 'maybe' for f in formats) or None
2811 if not self.params.get('allow_unplayable_formats'):
2812 formats = [f for f in formats if not f.get('has_drm') or f['has_drm'] == 'maybe']
2813
2814 if formats and all(f.get('acodec') == f.get('vcodec') == 'none' for f in formats):
2815 self.report_warning(
2816 f'{"This video is DRM protected and " if info_dict["_has_drm"] else ""}'
2817 'only images are available for download. Use --list-formats to see them'.capitalize())
2818
2819 get_from_start = not info_dict.get('is_live') or bool(self.params.get('live_from_start'))
2820 if not get_from_start:
2821 info_dict['title'] += ' ' + dt.datetime.now().strftime('%Y-%m-%d %H:%M')
2822 if info_dict.get('is_live') and formats:
2823 formats = [f for f in formats if bool(f.get('is_from_start')) == get_from_start]
2824 if get_from_start and not formats:
2825 self.raise_no_formats(info_dict, msg=(
2826 '--live-from-start is passed, but there are no formats that can be downloaded from the start. '
2827 'If you want to download from the current time, use --no-live-from-start'))
2828
2829 def is_wellformed(f):
2830 url = f.get('url')
2831 if not url:
2832 self.report_warning(
2833 '"url" field is missing or empty - skipping format, '
2834 'there is an error in extractor')
2835 return False
2836 if isinstance(url, bytes):
2837 sanitize_string_field(f, 'url')
2838 return True
2839
2840 # Filter out malformed formats for better extraction robustness
2841 formats = list(filter(is_wellformed, formats or []))
2842
2843 if not formats:
2844 self.raise_no_formats(info_dict)
2845
2846 for fmt in formats:
2847 sanitize_string_field(fmt, 'format_id')
2848 sanitize_numeric_fields(fmt)
2849 fmt['url'] = sanitize_url(fmt['url'])
2850 if fmt.get('ext') is None:
2851 fmt['ext'] = determine_ext(fmt['url']).lower()
2852 if fmt['ext'] in ('aac', 'opus', 'mp3', 'flac', 'vorbis'):
2853 if fmt.get('acodec') is None:
2854 fmt['acodec'] = fmt['ext']
2855 if fmt.get('protocol') is None:
2856 fmt['protocol'] = determine_protocol(fmt)
2857 if fmt.get('resolution') is None:
2858 fmt['resolution'] = self.format_resolution(fmt, default=None)
2859 if fmt.get('dynamic_range') is None and fmt.get('vcodec') != 'none':
2860 fmt['dynamic_range'] = 'SDR'
2861 if fmt.get('aspect_ratio') is None:
2862 fmt['aspect_ratio'] = try_call(lambda: round(fmt['width'] / fmt['height'], 2))
2863 # For fragmented formats, "tbr" is often max bitrate and not average
2864 if (('manifest-filesize-approx' in self.params['compat_opts'] or not fmt.get('manifest_url'))
2865 and not fmt.get('filesize') and not fmt.get('filesize_approx')):
2866 fmt['filesize_approx'] = filesize_from_tbr(fmt.get('tbr'), info_dict.get('duration'))
2867 fmt['http_headers'] = self._calc_headers(collections.ChainMap(fmt, info_dict), load_cookies=True)
2868
2869 # Safeguard against old/insecure infojson when using --load-info-json
2870 if info_dict.get('http_headers'):
2871 info_dict['http_headers'] = HTTPHeaderDict(info_dict['http_headers'])
2872 info_dict['http_headers'].pop('Cookie', None)
2873
2874 # This is copied to http_headers by the above _calc_headers and can now be removed
2875 if '__x_forwarded_for_ip' in info_dict:
2876 del info_dict['__x_forwarded_for_ip']
2877
2878 self.sort_formats({
2879 'formats': formats,
2880 '_format_sort_fields': info_dict.get('_format_sort_fields'),
2881 })
2882
2883 # Sanitize and group by format_id
2884 formats_dict = {}
2885 for i, fmt in enumerate(formats):
2886 if not fmt.get('format_id'):
2887 fmt['format_id'] = str(i)
2888 else:
2889 # Sanitize format_id from characters used in format selector expression
2890 fmt['format_id'] = re.sub(r'[\s,/+\[\]()]', '_', fmt['format_id'])
2891 formats_dict.setdefault(fmt['format_id'], []).append(fmt)
2892
2893 # Make sure all formats have unique format_id
2894 common_exts = set(itertools.chain(*self._format_selection_exts.values()))
2895 for format_id, ambiguous_formats in formats_dict.items():
2896 ambigious_id = len(ambiguous_formats) > 1
2897 for i, fmt in enumerate(ambiguous_formats):
2898 if ambigious_id:
2899 fmt['format_id'] = f'{format_id}-{i}'
2900 # Ensure there is no conflict between id and ext in format selection
2901 # See https://github.com/yt-dlp/yt-dlp/issues/1282
2902 if fmt['format_id'] != fmt['ext'] and fmt['format_id'] in common_exts:
2903 fmt['format_id'] = 'f{}'.format(fmt['format_id'])
2904
2905 if fmt.get('format') is None:
2906 fmt['format'] = '{id} - {res}{note}'.format(
2907 id=fmt['format_id'],
2908 res=self.format_resolution(fmt),
2909 note=format_field(fmt, 'format_note', ' (%s)'),
2910 )
2911
2912 if self.params.get('check_formats') is True:
2913 formats = LazyList(self._check_formats(formats[::-1]), reverse=True)
2914
2915 if not formats or formats[0] is not info_dict:
2916 # only set the 'formats' fields if the original info_dict list them
2917 # otherwise we end up with a circular reference, the first (and unique)
2918 # element in the 'formats' field in info_dict is info_dict itself,
2919 # which can't be exported to json
2920 info_dict['formats'] = formats
2921
2922 info_dict, _ = self.pre_process(info_dict)
2923
2924 if self._match_entry(info_dict, incomplete=self._format_fields) is not None:
2925 return info_dict
2926
2927 self.post_extract(info_dict)
2928 info_dict, _ = self.pre_process(info_dict, 'after_filter')
2929
2930 # The pre-processors may have modified the formats
2931 formats = self._get_formats(info_dict)
2932
2933 list_only = self.params.get('simulate') == 'list_only'
2934 interactive_format_selection = not list_only and self.format_selector == '-'
2935 if self.params.get('list_thumbnails'):
2936 self.list_thumbnails(info_dict)
2937 if self.params.get('listsubtitles'):
2938 if 'automatic_captions' in info_dict:
2939 self.list_subtitles(
2940 info_dict['id'], automatic_captions, 'automatic captions')
2941 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
2942 if self.params.get('listformats') or interactive_format_selection:
2943 self.list_formats(info_dict)
2944 if list_only:
2945 # Without this printing, -F --print-json will not work
2946 self.__forced_printings(info_dict)
2947 return info_dict
2948
2949 format_selector = self.format_selector
2950 while True:
2951 if interactive_format_selection:
2952 req_format = input(self._format_screen('\nEnter format selector ', self.Styles.EMPHASIS)
2953 + '(Press ENTER for default, or Ctrl+C to quit)'
2954 + self._format_screen(': ', self.Styles.EMPHASIS))
2955 try:
2956 format_selector = self.build_format_selector(req_format) if req_format else None
2957 except SyntaxError as err:
2958 self.report_error(err, tb=False, is_error=False)
2959 continue
2960
2961 if format_selector is None:
2962 req_format = self._default_format_spec(info_dict, download=download)
2963 self.write_debug(f'Default format spec: {req_format}')
2964 format_selector = self.build_format_selector(req_format)
2965
2966 formats_to_download = self._select_formats(formats, format_selector)
2967 if interactive_format_selection and not formats_to_download:
2968 self.report_error('Requested format is not available', tb=False, is_error=False)
2969 continue
2970 break
2971
2972 if not formats_to_download:
2973 if not self.params.get('ignore_no_formats_error'):
2974 raise ExtractorError(
2975 'Requested format is not available. Use --list-formats for a list of available formats',
2976 expected=True, video_id=info_dict['id'], ie=info_dict['extractor'])
2977 self.report_warning('Requested format is not available')
2978 # Process what we can, even without any available formats.
2979 formats_to_download = [{}]
2980
2981 requested_ranges = tuple(self.params.get('download_ranges', lambda *_: [{}])(info_dict, self))
2982 best_format, downloaded_formats = formats_to_download[-1], []
2983 if download:
2984 if best_format and requested_ranges:
2985 def to_screen(*msg):
2986 self.to_screen(f'[info] {info_dict["id"]}: {" ".join(", ".join(variadic(m)) for m in msg)}')
2987
2988 to_screen(f'Downloading {len(formats_to_download)} format(s):',
2989 (f['format_id'] for f in formats_to_download))
2990 if requested_ranges != ({}, ):
2991 to_screen(f'Downloading {len(requested_ranges)} time ranges:',
2992 (f'{c["start_time"]:.1f}-{c["end_time"]:.1f}' for c in requested_ranges))
2993 max_downloads_reached = False
2994
2995 for fmt, chapter in itertools.product(formats_to_download, requested_ranges):
2996 new_info = self._copy_infodict(info_dict)
2997 new_info.update(fmt)
2998 offset, duration = info_dict.get('section_start') or 0, info_dict.get('duration') or float('inf')
2999 end_time = offset + min(chapter.get('end_time', duration), duration)
3000 # duration may not be accurate. So allow deviations <1sec
3001 if end_time == float('inf') or end_time > offset + duration + 1:
3002 end_time = None
3003 if chapter or offset:
3004 new_info.update({
3005 'section_start': offset + chapter.get('start_time', 0),
3006 'section_end': end_time,
3007 'section_title': chapter.get('title'),
3008 'section_number': chapter.get('index'),
3009 })
3010 downloaded_formats.append(new_info)
3011 try:
3012 self.process_info(new_info)
3013 except MaxDownloadsReached:
3014 max_downloads_reached = True
3015 self._raise_pending_errors(new_info)
3016 # Remove copied info
3017 for key, val in tuple(new_info.items()):
3018 if info_dict.get(key) == val:
3019 new_info.pop(key)
3020 if max_downloads_reached:
3021 break
3022
3023 write_archive = {f.get('__write_download_archive', False) for f in downloaded_formats}
3024 assert write_archive.issubset({True, False, 'ignore'})
3025 if True in write_archive and False not in write_archive:
3026 self.record_download_archive(info_dict)
3027
3028 info_dict['requested_downloads'] = downloaded_formats
3029 info_dict = self.run_all_pps('after_video', info_dict)
3030 if max_downloads_reached:
3031 raise MaxDownloadsReached
3032
3033 # We update the info dict with the selected best quality format (backwards compatibility)
3034 info_dict.update(best_format)
3035 return info_dict
3036
3037 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
3038 """Select the requested subtitles and their format"""
3039 available_subs, normal_sub_langs = {}, []
3040 if normal_subtitles and self.params.get('writesubtitles'):
3041 available_subs.update(normal_subtitles)
3042 normal_sub_langs = tuple(normal_subtitles.keys())
3043 if automatic_captions and self.params.get('writeautomaticsub'):
3044 for lang, cap_info in automatic_captions.items():
3045 if lang not in available_subs:
3046 available_subs[lang] = cap_info
3047
3048 if not available_subs or (
3049 not self.params.get('writesubtitles')
3050 and not self.params.get('writeautomaticsub')):
3051 return None
3052
3053 all_sub_langs = tuple(available_subs.keys())
3054 if self.params.get('allsubtitles', False):
3055 requested_langs = all_sub_langs
3056 elif self.params.get('subtitleslangs', False):
3057 try:
3058 requested_langs = orderedSet_from_options(
3059 self.params.get('subtitleslangs'), {'all': all_sub_langs}, use_regex=True)
3060 except re.error as e:
3061 raise ValueError(f'Wrong regex for subtitlelangs: {e.pattern}')
3062 else:
3063 requested_langs = LazyList(itertools.chain(
3064 ['en'] if 'en' in normal_sub_langs else [],
3065 filter(lambda f: f.startswith('en'), normal_sub_langs),
3066 ['en'] if 'en' in all_sub_langs else [],
3067 filter(lambda f: f.startswith('en'), all_sub_langs),
3068 normal_sub_langs, all_sub_langs,
3069 ))[:1]
3070 if requested_langs:
3071 self.to_screen(f'[info] {video_id}: Downloading subtitles: {", ".join(requested_langs)}')
3072
3073 formats_query = self.params.get('subtitlesformat', 'best')
3074 formats_preference = formats_query.split('/') if formats_query else []
3075 subs = {}
3076 for lang in requested_langs:
3077 formats = available_subs.get(lang)
3078 if formats is None:
3079 self.report_warning(f'{lang} subtitles not available for {video_id}')
3080 continue
3081 for ext in formats_preference:
3082 if ext == 'best':
3083 f = formats[-1]
3084 break
3085 matches = list(filter(lambda f: f['ext'] == ext, formats))
3086 if matches:
3087 f = matches[-1]
3088 break
3089 else:
3090 f = formats[-1]
3091 self.report_warning(
3092 'No subtitle format found matching "{}" for language {}, '
3093 'using {}. Use --list-subs for a list of available subtitles'.format(formats_query, lang, f['ext']))
3094 subs[lang] = f
3095 return subs
3096
3097 def _forceprint(self, key, info_dict):
3098 if info_dict is None:
3099 return
3100 info_copy = info_dict.copy()
3101 info_copy.setdefault('filename', self.prepare_filename(info_dict))
3102 if info_dict.get('requested_formats') is not None:
3103 # For RTMP URLs, also include the playpath
3104 info_copy['urls'] = '\n'.join(f['url'] + f.get('play_path', '') for f in info_dict['requested_formats'])
3105 elif info_dict.get('url'):
3106 info_copy['urls'] = info_dict['url'] + info_dict.get('play_path', '')
3107 info_copy['formats_table'] = self.render_formats_table(info_dict)
3108 info_copy['thumbnails_table'] = self.render_thumbnails_table(info_dict)
3109 info_copy['subtitles_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('subtitles'))
3110 info_copy['automatic_captions_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('automatic_captions'))
3111
3112 def format_tmpl(tmpl):
3113 mobj = re.fullmatch(r'([\w.:,]|-\d|(?P<dict>{([\w.:,]|-\d)+}))+=?', tmpl)
3114 if not mobj:
3115 return tmpl
3116
3117 fmt = '%({})s'
3118 if tmpl.startswith('{'):
3119 tmpl, fmt = f'.{tmpl}', '%({})j'
3120 if tmpl.endswith('='):
3121 tmpl, fmt = tmpl[:-1], '{0} = %({0})#j'
3122 return '\n'.join(map(fmt.format, [tmpl] if mobj.group('dict') else tmpl.split(',')))
3123
3124 for tmpl in self.params['forceprint'].get(key, []):
3125 self.to_stdout(self.evaluate_outtmpl(format_tmpl(tmpl), info_copy))
3126
3127 for tmpl, file_tmpl in self.params['print_to_file'].get(key, []):
3128 filename = self.prepare_filename(info_dict, outtmpl=file_tmpl)
3129 tmpl = format_tmpl(tmpl)
3130 self.to_screen(f'[info] Writing {tmpl!r} to: {filename}')
3131 if self._ensure_dir_exists(filename):
3132 with open(filename, 'a', encoding='utf-8', newline='') as f:
3133 f.write(self.evaluate_outtmpl(tmpl, info_copy) + os.linesep)
3134
3135 return info_copy
3136
3137 def __forced_printings(self, info_dict, filename=None, incomplete=True):
3138 if (self.params.get('forcejson')
3139 or self.params['forceprint'].get('video')
3140 or self.params['print_to_file'].get('video')):
3141 self.post_extract(info_dict)
3142 if filename:
3143 info_dict['filename'] = filename
3144 info_copy = self._forceprint('video', info_dict)
3145
3146 def print_field(field, actual_field=None, optional=False):
3147 if actual_field is None:
3148 actual_field = field
3149 if self.params.get(f'force{field}') and (
3150 info_copy.get(field) is not None or (not optional and not incomplete)):
3151 self.to_stdout(info_copy[actual_field])
3152
3153 print_field('title')
3154 print_field('id')
3155 print_field('url', 'urls')
3156 print_field('thumbnail', optional=True)
3157 print_field('description', optional=True)
3158 print_field('filename')
3159 if self.params.get('forceduration') and info_copy.get('duration') is not None:
3160 self.to_stdout(formatSeconds(info_copy['duration']))
3161 print_field('format')
3162
3163 if self.params.get('forcejson'):
3164 self.to_stdout(json.dumps(self.sanitize_info(info_dict)))
3165
3166 def dl(self, name, info, subtitle=False, test=False):
3167 if not info.get('url'):
3168 self.raise_no_formats(info, True)
3169
3170 if test:
3171 verbose = self.params.get('verbose')
3172 params = {
3173 'test': True,
3174 'quiet': self.params.get('quiet') or not verbose,
3175 'verbose': verbose,
3176 'noprogress': not verbose,
3177 'nopart': True,
3178 'skip_unavailable_fragments': False,
3179 'keep_fragments': False,
3180 'overwrites': True,
3181 '_no_ytdl_file': True,
3182 }
3183 else:
3184 params = self.params
3185 fd = get_suitable_downloader(info, params, to_stdout=(name == '-'))(self, params)
3186 if not test:
3187 for ph in self._progress_hooks:
3188 fd.add_progress_hook(ph)
3189 urls = '", "'.join(
3190 (f['url'].split(',')[0] + ',<data>' if f['url'].startswith('data:') else f['url'])
3191 for f in info.get('requested_formats', []) or [info])
3192 self.write_debug(f'Invoking {fd.FD_NAME} downloader on "{urls}"')
3193
3194 # Note: Ideally info should be a deep-copied so that hooks cannot modify it.
3195 # But it may contain objects that are not deep-copyable
3196 new_info = self._copy_infodict(info)
3197 if new_info.get('http_headers') is None:
3198 new_info['http_headers'] = self._calc_headers(new_info)
3199 return fd.download(name, new_info, subtitle)
3200
3201 def existing_file(self, filepaths, *, default_overwrite=True):
3202 existing_files = list(filter(os.path.exists, orderedSet(filepaths)))
3203 if existing_files and not self.params.get('overwrites', default_overwrite):
3204 return existing_files[0]
3205
3206 for file in existing_files:
3207 self.report_file_delete(file)
3208 os.remove(file)
3209 return None
3210
3211 @_catch_unsafe_extension_error
3212 def process_info(self, info_dict):
3213 """Process a single resolved IE result. (Modifies it in-place)"""
3214
3215 assert info_dict.get('_type', 'video') == 'video'
3216 original_infodict = info_dict
3217
3218 if 'format' not in info_dict and 'ext' in info_dict:
3219 info_dict['format'] = info_dict['ext']
3220
3221 if self._match_entry(info_dict) is not None:
3222 info_dict['__write_download_archive'] = 'ignore'
3223 return
3224
3225 # Does nothing under normal operation - for backward compatibility of process_info
3226 self.post_extract(info_dict)
3227
3228 def replace_info_dict(new_info):
3229 nonlocal info_dict
3230 if new_info == info_dict:
3231 return
3232 info_dict.clear()
3233 info_dict.update(new_info)
3234
3235 new_info, _ = self.pre_process(info_dict, 'video')
3236 replace_info_dict(new_info)
3237 self._num_downloads += 1
3238
3239 # info_dict['_filename'] needs to be set for backward compatibility
3240 info_dict['_filename'] = full_filename = self.prepare_filename(info_dict, warn=True)
3241 temp_filename = self.prepare_filename(info_dict, 'temp')
3242 files_to_move = {}
3243
3244 # Forced printings
3245 self.__forced_printings(info_dict, full_filename, incomplete=('format' not in info_dict))
3246
3247 def check_max_downloads():
3248 if self._num_downloads >= float(self.params.get('max_downloads') or 'inf'):
3249 raise MaxDownloadsReached
3250
3251 if self.params.get('simulate'):
3252 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
3253 check_max_downloads()
3254 return
3255
3256 if full_filename is None:
3257 return
3258 if not self._ensure_dir_exists(encodeFilename(full_filename)):
3259 return
3260 if not self._ensure_dir_exists(encodeFilename(temp_filename)):
3261 return
3262
3263 if self._write_description('video', info_dict,
3264 self.prepare_filename(info_dict, 'description')) is None:
3265 return
3266
3267 sub_files = self._write_subtitles(info_dict, temp_filename)
3268 if sub_files is None:
3269 return
3270 files_to_move.update(dict(sub_files))
3271
3272 thumb_files = self._write_thumbnails(
3273 'video', info_dict, temp_filename, self.prepare_filename(info_dict, 'thumbnail'))
3274 if thumb_files is None:
3275 return
3276 files_to_move.update(dict(thumb_files))
3277
3278 infofn = self.prepare_filename(info_dict, 'infojson')
3279 _infojson_written = self._write_info_json('video', info_dict, infofn)
3280 if _infojson_written:
3281 info_dict['infojson_filename'] = infofn
3282 # For backward compatibility, even though it was a private field
3283 info_dict['__infojson_filename'] = infofn
3284 elif _infojson_written is None:
3285 return
3286
3287 # Note: Annotations are deprecated
3288 annofn = None
3289 if self.params.get('writeannotations', False):
3290 annofn = self.prepare_filename(info_dict, 'annotation')
3291 if annofn:
3292 if not self._ensure_dir_exists(encodeFilename(annofn)):
3293 return
3294 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(annofn)):
3295 self.to_screen('[info] Video annotations are already present')
3296 elif not info_dict.get('annotations'):
3297 self.report_warning('There are no annotations to write.')
3298 else:
3299 try:
3300 self.to_screen('[info] Writing video annotations to: ' + annofn)
3301 with open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
3302 annofile.write(info_dict['annotations'])
3303 except (KeyError, TypeError):
3304 self.report_warning('There are no annotations to write.')
3305 except OSError:
3306 self.report_error('Cannot write annotations file: ' + annofn)
3307 return
3308
3309 # Write internet shortcut files
3310 def _write_link_file(link_type):
3311 url = try_get(info_dict['webpage_url'], iri_to_uri)
3312 if not url:
3313 self.report_warning(
3314 f'Cannot write internet shortcut file because the actual URL of "{info_dict["webpage_url"]}" is unknown')
3315 return True
3316 linkfn = replace_extension(self.prepare_filename(info_dict, 'link'), link_type, info_dict.get('ext'))
3317 if not self._ensure_dir_exists(encodeFilename(linkfn)):
3318 return False
3319 if self.params.get('overwrites', True) and os.path.exists(encodeFilename(linkfn)):
3320 self.to_screen(f'[info] Internet shortcut (.{link_type}) is already present')
3321 return True
3322 try:
3323 self.to_screen(f'[info] Writing internet shortcut (.{link_type}) to: {linkfn}')
3324 with open(encodeFilename(to_high_limit_path(linkfn)), 'w', encoding='utf-8',
3325 newline='\r\n' if link_type == 'url' else '\n') as linkfile:
3326 template_vars = {'url': url}
3327 if link_type == 'desktop':
3328 template_vars['filename'] = linkfn[:-(len(link_type) + 1)]
3329 linkfile.write(LINK_TEMPLATES[link_type] % template_vars)
3330 except OSError:
3331 self.report_error(f'Cannot write internet shortcut {linkfn}')
3332 return False
3333 return True
3334
3335 write_links = {
3336 'url': self.params.get('writeurllink'),
3337 'webloc': self.params.get('writewebloclink'),
3338 'desktop': self.params.get('writedesktoplink'),
3339 }
3340 if self.params.get('writelink'):
3341 link_type = ('webloc' if sys.platform == 'darwin'
3342 else 'desktop' if sys.platform.startswith('linux')
3343 else 'url')
3344 write_links[link_type] = True
3345
3346 if any(should_write and not _write_link_file(link_type)
3347 for link_type, should_write in write_links.items()):
3348 return
3349
3350 new_info, files_to_move = self.pre_process(info_dict, 'before_dl', files_to_move)
3351 replace_info_dict(new_info)
3352
3353 if self.params.get('skip_download'):
3354 info_dict['filepath'] = temp_filename
3355 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
3356 info_dict['__files_to_move'] = files_to_move
3357 replace_info_dict(self.run_pp(MoveFilesAfterDownloadPP(self, False), info_dict))
3358 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
3359 else:
3360 # Download
3361 info_dict.setdefault('__postprocessors', [])
3362 try:
3363
3364 def existing_video_file(*filepaths):
3365 ext = info_dict.get('ext')
3366 converted = lambda file: replace_extension(file, self.params.get('final_ext') or ext, ext)
3367 file = self.existing_file(itertools.chain(*zip(map(converted, filepaths), filepaths)),
3368 default_overwrite=False)
3369 if file:
3370 info_dict['ext'] = os.path.splitext(file)[1][1:]
3371 return file
3372
3373 fd, success = None, True
3374 if info_dict.get('protocol') or info_dict.get('url'):
3375 fd = get_suitable_downloader(info_dict, self.params, to_stdout=temp_filename == '-')
3376 if fd != FFmpegFD and 'no-direct-merge' not in self.params['compat_opts'] and (
3377 info_dict.get('section_start') or info_dict.get('section_end')):
3378 msg = ('This format cannot be partially downloaded' if FFmpegFD.available()
3379 else 'You have requested downloading the video partially, but ffmpeg is not installed')
3380 self.report_error(f'{msg}. Aborting')
3381 return
3382
3383 if info_dict.get('requested_formats') is not None:
3384 old_ext = info_dict['ext']
3385 if self.params.get('merge_output_format') is None:
3386 if (info_dict['ext'] == 'webm'
3387 and info_dict.get('thumbnails')
3388 # check with type instead of pp_key, __name__, or isinstance
3389 # since we dont want any custom PPs to trigger this
3390 and any(type(pp) == EmbedThumbnailPP for pp in self._pps['post_process'])): # noqa: E721
3391 info_dict['ext'] = 'mkv'
3392 self.report_warning(
3393 'webm doesn\'t support embedding a thumbnail, mkv will be used')
3394 new_ext = info_dict['ext']
3395
3396 def correct_ext(filename, ext=new_ext):
3397 if filename == '-':
3398 return filename
3399 filename_real_ext = os.path.splitext(filename)[1][1:]
3400 filename_wo_ext = (
3401 os.path.splitext(filename)[0]
3402 if filename_real_ext in (old_ext, new_ext)
3403 else filename)
3404 return f'{filename_wo_ext}.{ext}'
3405
3406 # Ensure filename always has a correct extension for successful merge
3407 full_filename = correct_ext(full_filename)
3408 temp_filename = correct_ext(temp_filename)
3409 dl_filename = existing_video_file(full_filename, temp_filename)
3410
3411 info_dict['__real_download'] = False
3412 # NOTE: Copy so that original format dicts are not modified
3413 info_dict['requested_formats'] = list(map(dict, info_dict['requested_formats']))
3414
3415 merger = FFmpegMergerPP(self)
3416 downloaded = []
3417 if dl_filename is not None:
3418 self.report_file_already_downloaded(dl_filename)
3419 elif fd:
3420 for f in info_dict['requested_formats'] if fd != FFmpegFD else []:
3421 f['filepath'] = fname = prepend_extension(
3422 correct_ext(temp_filename, info_dict['ext']),
3423 'f{}'.format(f['format_id']), info_dict['ext'])
3424 downloaded.append(fname)
3425 info_dict['url'] = '\n'.join(f['url'] for f in info_dict['requested_formats'])
3426 success, real_download = self.dl(temp_filename, info_dict)
3427 info_dict['__real_download'] = real_download
3428 else:
3429 if self.params.get('allow_unplayable_formats'):
3430 self.report_warning(
3431 'You have requested merging of multiple formats '
3432 'while also allowing unplayable formats to be downloaded. '
3433 'The formats won\'t be merged to prevent data corruption.')
3434 elif not merger.available:
3435 msg = 'You have requested merging of multiple formats but ffmpeg is not installed'
3436 if not self.params.get('ignoreerrors'):
3437 self.report_error(f'{msg}. Aborting due to --abort-on-error')
3438 return
3439 self.report_warning(f'{msg}. The formats won\'t be merged')
3440
3441 if temp_filename == '-':
3442 reason = ('using a downloader other than ffmpeg' if FFmpegFD.can_merge_formats(info_dict, self.params)
3443 else 'but the formats are incompatible for simultaneous download' if merger.available
3444 else 'but ffmpeg is not installed')
3445 self.report_warning(
3446 f'You have requested downloading multiple formats to stdout {reason}. '
3447 'The formats will be streamed one after the other')
3448 fname = temp_filename
3449 for f in info_dict['requested_formats']:
3450 new_info = dict(info_dict)
3451 del new_info['requested_formats']
3452 new_info.update(f)
3453 if temp_filename != '-':
3454 fname = prepend_extension(
3455 correct_ext(temp_filename, new_info['ext']),
3456 'f{}'.format(f['format_id']), new_info['ext'])
3457 if not self._ensure_dir_exists(fname):
3458 return
3459 f['filepath'] = fname
3460 downloaded.append(fname)
3461 partial_success, real_download = self.dl(fname, new_info)
3462 info_dict['__real_download'] = info_dict['__real_download'] or real_download
3463 success = success and partial_success
3464
3465 if downloaded and merger.available and not self.params.get('allow_unplayable_formats'):
3466 info_dict['__postprocessors'].append(merger)
3467 info_dict['__files_to_merge'] = downloaded
3468 # Even if there were no downloads, it is being merged only now
3469 info_dict['__real_download'] = True
3470 else:
3471 for file in downloaded:
3472 files_to_move[file] = None
3473 else:
3474 # Just a single file
3475 dl_filename = existing_video_file(full_filename, temp_filename)
3476 if dl_filename is None or dl_filename == temp_filename:
3477 # dl_filename == temp_filename could mean that the file was partially downloaded with --no-part.
3478 # So we should try to resume the download
3479 success, real_download = self.dl(temp_filename, info_dict)
3480 info_dict['__real_download'] = real_download
3481 else:
3482 self.report_file_already_downloaded(dl_filename)
3483
3484 dl_filename = dl_filename or temp_filename
3485 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
3486
3487 except network_exceptions as err:
3488 self.report_error(f'unable to download video data: {err}')
3489 return
3490 except OSError as err:
3491 raise UnavailableVideoError(err)
3492 except ContentTooShortError as err:
3493 self.report_error(f'content too short (expected {err.expected} bytes and served {err.downloaded})')
3494 return
3495
3496 self._raise_pending_errors(info_dict)
3497 if success and full_filename != '-':
3498
3499 def fixup():
3500 do_fixup = True
3501 fixup_policy = self.params.get('fixup')
3502 vid = info_dict['id']
3503
3504 if fixup_policy in ('ignore', 'never'):
3505 return
3506 elif fixup_policy == 'warn':
3507 do_fixup = 'warn'
3508 elif fixup_policy != 'force':
3509 assert fixup_policy in ('detect_or_warn', None)
3510 if not info_dict.get('__real_download'):
3511 do_fixup = False
3512
3513 def ffmpeg_fixup(cndn, msg, cls):
3514 if not (do_fixup and cndn):
3515 return
3516 elif do_fixup == 'warn':
3517 self.report_warning(f'{vid}: {msg}')
3518 return
3519 pp = cls(self)
3520 if pp.available:
3521 info_dict['__postprocessors'].append(pp)
3522 else:
3523 self.report_warning(f'{vid}: {msg}. Install ffmpeg to fix this automatically')
3524
3525 stretched_ratio = info_dict.get('stretched_ratio')
3526 ffmpeg_fixup(stretched_ratio not in (1, None),
3527 f'Non-uniform pixel ratio {stretched_ratio}',
3528 FFmpegFixupStretchedPP)
3529
3530 downloader = get_suitable_downloader(info_dict, self.params) if 'protocol' in info_dict else None
3531 downloader = downloader.FD_NAME if downloader else None
3532
3533 ext = info_dict.get('ext')
3534 postprocessed_by_ffmpeg = info_dict.get('requested_formats') or any((
3535 isinstance(pp, FFmpegVideoConvertorPP)
3536 and resolve_recode_mapping(ext, pp.mapping)[0] not in (ext, None)
3537 ) for pp in self._pps['post_process'])
3538
3539 if not postprocessed_by_ffmpeg:
3540 ffmpeg_fixup(fd != FFmpegFD and ext == 'm4a'
3541 and info_dict.get('container') == 'm4a_dash',
3542 'writing DASH m4a. Only some players support this container',
3543 FFmpegFixupM4aPP)
3544 ffmpeg_fixup(downloader == 'hlsnative' and not self.params.get('hls_use_mpegts')
3545 or info_dict.get('is_live') and self.params.get('hls_use_mpegts') is None,
3546 'Possible MPEG-TS in MP4 container or malformed AAC timestamps',
3547 FFmpegFixupM3u8PP)
3548 ffmpeg_fixup(downloader == 'dashsegments'
3549 and (info_dict.get('is_live') or info_dict.get('is_dash_periods')),
3550 'Possible duplicate MOOV atoms', FFmpegFixupDuplicateMoovPP)
3551
3552 ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed timestamps detected', FFmpegFixupTimestampPP)
3553 ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed duration detected', FFmpegFixupDurationPP)
3554
3555 fixup()
3556 try:
3557 replace_info_dict(self.post_process(dl_filename, info_dict, files_to_move))
3558 except PostProcessingError as err:
3559 self.report_error(f'Postprocessing: {err}')
3560 return
3561 try:
3562 for ph in self._post_hooks:
3563 ph(info_dict['filepath'])
3564 except Exception as err:
3565 self.report_error(f'post hooks: {err}')
3566 return
3567 info_dict['__write_download_archive'] = True
3568
3569 assert info_dict is original_infodict # Make sure the info_dict was modified in-place
3570 if self.params.get('force_write_download_archive'):
3571 info_dict['__write_download_archive'] = True
3572 check_max_downloads()
3573
3574 def __download_wrapper(self, func):
3575 @functools.wraps(func)
3576 def wrapper(*args, **kwargs):
3577 try:
3578 res = func(*args, **kwargs)
3579 except UnavailableVideoError as e:
3580 self.report_error(e)
3581 except DownloadCancelled as e:
3582 self.to_screen(f'[info] {e}')
3583 if not self.params.get('break_per_url'):
3584 raise
3585 self._num_downloads = 0
3586 else:
3587 if self.params.get('dump_single_json', False):
3588 self.post_extract(res)
3589 self.to_stdout(json.dumps(self.sanitize_info(res)))
3590 return wrapper
3591
3592 def download(self, url_list):
3593 """Download a given list of URLs."""
3594 url_list = variadic(url_list) # Passing a single URL is a common mistake
3595 outtmpl = self.params['outtmpl']['default']
3596 if (len(url_list) > 1
3597 and outtmpl != '-'
3598 and '%' not in outtmpl
3599 and self.params.get('max_downloads') != 1):
3600 raise SameFileError(outtmpl)
3601
3602 for url in url_list:
3603 self.__download_wrapper(self.extract_info)(
3604 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
3605
3606 return self._download_retcode
3607
3608 def download_with_info_file(self, info_filename):
3609 with contextlib.closing(fileinput.FileInput(
3610 [info_filename], mode='r',
3611 openhook=fileinput.hook_encoded('utf-8'))) as f:
3612 # FileInput doesn't have a read method, we can't call json.load
3613 infos = [self.sanitize_info(info, self.params.get('clean_infojson', True))
3614 for info in variadic(json.loads('\n'.join(f)))]
3615 for info in infos:
3616 try:
3617 self.__download_wrapper(self.process_ie_result)(info, download=True)
3618 except (DownloadError, EntryNotInPlaylist, ReExtractInfo) as e:
3619 if not isinstance(e, EntryNotInPlaylist):
3620 self.to_stderr('\r')
3621 webpage_url = info.get('webpage_url')
3622 if webpage_url is None:
3623 raise
3624 self.report_warning(f'The info failed to download: {e}; trying with URL {webpage_url}')
3625 self.download([webpage_url])
3626 except ExtractorError as e:
3627 self.report_error(e)
3628 return self._download_retcode
3629
3630 @staticmethod
3631 def sanitize_info(info_dict, remove_private_keys=False):
3632 """ Sanitize the infodict for converting to json """
3633 if info_dict is None:
3634 return info_dict
3635 info_dict.setdefault('epoch', int(time.time()))
3636 info_dict.setdefault('_type', 'video')
3637 info_dict.setdefault('_version', {
3638 'version': __version__,
3639 'current_git_head': current_git_head(),
3640 'release_git_head': RELEASE_GIT_HEAD,
3641 'repository': ORIGIN,
3642 })
3643
3644 if remove_private_keys:
3645 reject = lambda k, v: v is None or k.startswith('__') or k in {
3646 'requested_downloads', 'requested_formats', 'requested_subtitles', 'requested_entries',
3647 'entries', 'filepath', '_filename', 'filename', 'infojson_filename', 'original_url',
3648 'playlist_autonumber',
3649 }
3650 else:
3651 reject = lambda k, v: False
3652
3653 def filter_fn(obj):
3654 if isinstance(obj, dict):
3655 return {k: filter_fn(v) for k, v in obj.items() if not reject(k, v)}
3656 elif isinstance(obj, (list, tuple, set, LazyList)):
3657 return list(map(filter_fn, obj))
3658 elif obj is None or isinstance(obj, (str, int, float, bool)):
3659 return obj
3660 else:
3661 return repr(obj)
3662
3663 return filter_fn(info_dict)
3664
3665 @staticmethod
3666 def filter_requested_info(info_dict, actually_filter=True):
3667 """ Alias of sanitize_info for backward compatibility """
3668 return YoutubeDL.sanitize_info(info_dict, actually_filter)
3669
3670 def _delete_downloaded_files(self, *files_to_delete, info={}, msg=None):
3671 for filename in set(filter(None, files_to_delete)):
3672 if msg:
3673 self.to_screen(msg % filename)
3674 try:
3675 os.remove(filename)
3676 except OSError:
3677 self.report_warning(f'Unable to delete file {filename}')
3678 if filename in info.get('__files_to_move', []): # NB: Delete even if None
3679 del info['__files_to_move'][filename]
3680
3681 @staticmethod
3682 def post_extract(info_dict):
3683 def actual_post_extract(info_dict):
3684 if info_dict.get('_type') in ('playlist', 'multi_video'):
3685 for video_dict in info_dict.get('entries', {}):
3686 actual_post_extract(video_dict or {})
3687 return
3688
3689 post_extractor = info_dict.pop('__post_extractor', None) or dict
3690 info_dict.update(post_extractor())
3691
3692 actual_post_extract(info_dict or {})
3693
3694 def run_pp(self, pp, infodict):
3695 files_to_delete = []
3696 if '__files_to_move' not in infodict:
3697 infodict['__files_to_move'] = {}
3698 try:
3699 files_to_delete, infodict = pp.run(infodict)
3700 except PostProcessingError as e:
3701 # Must be True and not 'only_download'
3702 if self.params.get('ignoreerrors') is True:
3703 self.report_error(e)
3704 return infodict
3705 raise
3706
3707 if not files_to_delete:
3708 return infodict
3709 if self.params.get('keepvideo', False):
3710 for f in files_to_delete:
3711 infodict['__files_to_move'].setdefault(f, '')
3712 else:
3713 self._delete_downloaded_files(
3714 *files_to_delete, info=infodict, msg='Deleting original file %s (pass -k to keep)')
3715 return infodict
3716
3717 def run_all_pps(self, key, info, *, additional_pps=None):
3718 if key != 'video':
3719 self._forceprint(key, info)
3720 for pp in (additional_pps or []) + self._pps[key]:
3721 info = self.run_pp(pp, info)
3722 return info
3723
3724 def pre_process(self, ie_info, key='pre_process', files_to_move=None):
3725 info = dict(ie_info)
3726 info['__files_to_move'] = files_to_move or {}
3727 try:
3728 info = self.run_all_pps(key, info)
3729 except PostProcessingError as err:
3730 msg = f'Preprocessing: {err}'
3731 info.setdefault('__pending_error', msg)
3732 self.report_error(msg, is_error=False)
3733 return info, info.pop('__files_to_move', None)
3734
3735 def post_process(self, filename, info, files_to_move=None):
3736 """Run all the postprocessors on the given file."""
3737 info['filepath'] = filename
3738 info['__files_to_move'] = files_to_move or {}
3739 info = self.run_all_pps('post_process', info, additional_pps=info.get('__postprocessors'))
3740 info = self.run_pp(MoveFilesAfterDownloadPP(self), info)
3741 del info['__files_to_move']
3742 return self.run_all_pps('after_move', info)
3743
3744 def _make_archive_id(self, info_dict):
3745 video_id = info_dict.get('id')
3746 if not video_id:
3747 return
3748 # Future-proof against any change in case
3749 # and backwards compatibility with prior versions
3750 extractor = info_dict.get('extractor_key') or info_dict.get('ie_key') # key in a playlist
3751 if extractor is None:
3752 url = str_or_none(info_dict.get('url'))
3753 if not url:
3754 return
3755 # Try to find matching extractor for the URL and take its ie_key
3756 for ie_key, ie in self._ies.items():
3757 if ie.suitable(url):
3758 extractor = ie_key
3759 break
3760 else:
3761 return
3762 return make_archive_id(extractor, video_id)
3763
3764 def in_download_archive(self, info_dict):
3765 if not self.archive:
3766 return False
3767
3768 vid_ids = [self._make_archive_id(info_dict)]
3769 vid_ids.extend(info_dict.get('_old_archive_ids') or [])
3770 return any(id_ in self.archive for id_ in vid_ids)
3771
3772 def record_download_archive(self, info_dict):
3773 fn = self.params.get('download_archive')
3774 if fn is None:
3775 return
3776 vid_id = self._make_archive_id(info_dict)
3777 assert vid_id
3778
3779 self.write_debug(f'Adding to archive: {vid_id}')
3780 if is_path_like(fn):
3781 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
3782 archive_file.write(vid_id + '\n')
3783 self.archive.add(vid_id)
3784
3785 @staticmethod
3786 def format_resolution(format, default='unknown'):
3787 if format.get('vcodec') == 'none' and format.get('acodec') != 'none':
3788 return 'audio only'
3789 if format.get('resolution') is not None:
3790 return format['resolution']
3791 if format.get('width') and format.get('height'):
3792 return '%dx%d' % (format['width'], format['height'])
3793 elif format.get('height'):
3794 return '{}p'.format(format['height'])
3795 elif format.get('width'):
3796 return '%dx?' % format['width']
3797 return default
3798
3799 def _list_format_headers(self, *headers):
3800 if self.params.get('listformats_table', True) is not False:
3801 return [self._format_out(header, self.Styles.HEADERS) for header in headers]
3802 return headers
3803
3804 def _format_note(self, fdict):
3805 res = ''
3806 if fdict.get('ext') in ['f4f', 'f4m']:
3807 res += '(unsupported)'
3808 if fdict.get('language'):
3809 if res:
3810 res += ' '
3811 res += '[{}]'.format(fdict['language'])
3812 if fdict.get('format_note') is not None:
3813 if res:
3814 res += ' '
3815 res += fdict['format_note']
3816 if fdict.get('tbr') is not None:
3817 if res:
3818 res += ', '
3819 res += '%4dk' % fdict['tbr']
3820 if fdict.get('container') is not None:
3821 if res:
3822 res += ', '
3823 res += '{} container'.format(fdict['container'])
3824 if (fdict.get('vcodec') is not None
3825 and fdict.get('vcodec') != 'none'):
3826 if res:
3827 res += ', '
3828 res += fdict['vcodec']
3829 if fdict.get('vbr') is not None:
3830 res += '@'
3831 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
3832 res += 'video@'
3833 if fdict.get('vbr') is not None:
3834 res += '%4dk' % fdict['vbr']
3835 if fdict.get('fps') is not None:
3836 if res:
3837 res += ', '
3838 res += '{}fps'.format(fdict['fps'])
3839 if fdict.get('acodec') is not None:
3840 if res:
3841 res += ', '
3842 if fdict['acodec'] == 'none':
3843 res += 'video only'
3844 else:
3845 res += '%-5s' % fdict['acodec']
3846 elif fdict.get('abr') is not None:
3847 if res:
3848 res += ', '
3849 res += 'audio'
3850 if fdict.get('abr') is not None:
3851 res += '@%3dk' % fdict['abr']
3852 if fdict.get('asr') is not None:
3853 res += ' (%5dHz)' % fdict['asr']
3854 if fdict.get('filesize') is not None:
3855 if res:
3856 res += ', '
3857 res += format_bytes(fdict['filesize'])
3858 elif fdict.get('filesize_approx') is not None:
3859 if res:
3860 res += ', '
3861 res += '~' + format_bytes(fdict['filesize_approx'])
3862 return res
3863
3864 def _get_formats(self, info_dict):
3865 if info_dict.get('formats') is None:
3866 if info_dict.get('url') and info_dict.get('_type', 'video') == 'video':
3867 return [info_dict]
3868 return []
3869 return info_dict['formats']
3870
3871 def render_formats_table(self, info_dict):
3872 formats = self._get_formats(info_dict)
3873 if not formats:
3874 return
3875 if not self.params.get('listformats_table', True) is not False:
3876 table = [
3877 [
3878 format_field(f, 'format_id'),
3879 format_field(f, 'ext'),
3880 self.format_resolution(f),
3881 self._format_note(f),
3882 ] for f in formats if (f.get('preference') or 0) >= -1000]
3883 return render_table(['format code', 'extension', 'resolution', 'note'], table, extra_gap=1)
3884
3885 def simplified_codec(f, field):
3886 assert field in ('acodec', 'vcodec')
3887 codec = f.get(field)
3888 if not codec:
3889 return 'unknown'
3890 elif codec != 'none':
3891 return '.'.join(codec.split('.')[:4])
3892
3893 if field == 'vcodec' and f.get('acodec') == 'none':
3894 return 'images'
3895 elif field == 'acodec' and f.get('vcodec') == 'none':
3896 return ''
3897 return self._format_out('audio only' if field == 'vcodec' else 'video only',
3898 self.Styles.SUPPRESS)
3899
3900 delim = self._format_out('\u2502', self.Styles.DELIM, '|', test_encoding=True)
3901 table = [
3902 [
3903 self._format_out(format_field(f, 'format_id'), self.Styles.ID),
3904 format_field(f, 'ext'),
3905 format_field(f, func=self.format_resolution, ignore=('audio only', 'images')),
3906 format_field(f, 'fps', '\t%d', func=round),
3907 format_field(f, 'dynamic_range', '%s', ignore=(None, 'SDR')).replace('HDR', ''),
3908 format_field(f, 'audio_channels', '\t%s'),
3909 delim, (
3910 format_field(f, 'filesize', ' \t%s', func=format_bytes)
3911 or format_field(f, 'filesize_approx', '≈\t%s', func=format_bytes)
3912 or format_field(filesize_from_tbr(f.get('tbr'), info_dict.get('duration')), None,
3913 self._format_out('~\t%s', self.Styles.SUPPRESS), func=format_bytes)),
3914 format_field(f, 'tbr', '\t%dk', func=round),
3915 shorten_protocol_name(f.get('protocol', '')),
3916 delim,
3917 simplified_codec(f, 'vcodec'),
3918 format_field(f, 'vbr', '\t%dk', func=round),
3919 simplified_codec(f, 'acodec'),
3920 format_field(f, 'abr', '\t%dk', func=round),
3921 format_field(f, 'asr', '\t%s', func=format_decimal_suffix),
3922 join_nonempty(format_field(f, 'language', '[%s]'), join_nonempty(
3923 self._format_out('UNSUPPORTED', self.Styles.BAD_FORMAT) if f.get('ext') in ('f4f', 'f4m') else None,
3924 (self._format_out('Maybe DRM', self.Styles.WARNING) if f.get('has_drm') == 'maybe'
3925 else self._format_out('DRM', self.Styles.BAD_FORMAT) if f.get('has_drm') else None),
3926 format_field(f, 'format_note'),
3927 format_field(f, 'container', ignore=(None, f.get('ext'))),
3928 delim=', '), delim=' '),
3929 ] for f in formats if f.get('preference') is None or f['preference'] >= -1000]
3930 header_line = self._list_format_headers(
3931 'ID', 'EXT', 'RESOLUTION', '\tFPS', 'HDR', 'CH', delim, '\tFILESIZE', '\tTBR', 'PROTO',
3932 delim, 'VCODEC', '\tVBR', 'ACODEC', '\tABR', '\tASR', 'MORE INFO')
3933
3934 return render_table(
3935 header_line, table, hide_empty=True,
3936 delim=self._format_out('\u2500', self.Styles.DELIM, '-', test_encoding=True))
3937
3938 def render_thumbnails_table(self, info_dict):
3939 thumbnails = list(info_dict.get('thumbnails') or [])
3940 if not thumbnails:
3941 return None
3942 return render_table(
3943 self._list_format_headers('ID', 'Width', 'Height', 'URL'),
3944 [[t.get('id'), t.get('width') or 'unknown', t.get('height') or 'unknown', t['url']] for t in thumbnails])
3945
3946 def render_subtitles_table(self, video_id, subtitles):
3947 def _row(lang, formats):
3948 exts, names = zip(*((f['ext'], f.get('name') or 'unknown') for f in reversed(formats)))
3949 if len(set(names)) == 1:
3950 names = [] if names[0] == 'unknown' else names[:1]
3951 return [lang, ', '.join(names), ', '.join(exts)]
3952
3953 if not subtitles:
3954 return None
3955 return render_table(
3956 self._list_format_headers('Language', 'Name', 'Formats'),
3957 [_row(lang, formats) for lang, formats in subtitles.items()],
3958 hide_empty=True)
3959
3960 def __list_table(self, video_id, name, func, *args):
3961 table = func(*args)
3962 if not table:
3963 self.to_screen(f'{video_id} has no {name}')
3964 return
3965 self.to_screen(f'[info] Available {name} for {video_id}:')
3966 self.to_stdout(table)
3967
3968 def list_formats(self, info_dict):
3969 self.__list_table(info_dict['id'], 'formats', self.render_formats_table, info_dict)
3970
3971 def list_thumbnails(self, info_dict):
3972 self.__list_table(info_dict['id'], 'thumbnails', self.render_thumbnails_table, info_dict)
3973
3974 def list_subtitles(self, video_id, subtitles, name='subtitles'):
3975 self.__list_table(video_id, name, self.render_subtitles_table, video_id, subtitles)
3976
3977 def print_debug_header(self):
3978 if not self.params.get('verbose'):
3979 return
3980
3981 from . import _IN_CLI # Must be delayed import
3982
3983 # These imports can be slow. So import them only as needed
3984 from .extractor.extractors import _LAZY_LOADER
3985 from .extractor.extractors import (
3986 _PLUGIN_CLASSES as plugin_ies,
3987 _PLUGIN_OVERRIDES as plugin_ie_overrides,
3988 )
3989
3990 def get_encoding(stream):
3991 ret = str(getattr(stream, 'encoding', f'missing ({type(stream).__name__})'))
3992 additional_info = []
3993 if os.environ.get('TERM', '').lower() == 'dumb':
3994 additional_info.append('dumb')
3995 if not supports_terminal_sequences(stream):
3996 from .utils import WINDOWS_VT_MODE # Must be imported locally
3997 additional_info.append('No VT' if WINDOWS_VT_MODE is False else 'No ANSI')
3998 if additional_info:
3999 ret = f'{ret} ({",".join(additional_info)})'
4000 return ret
4001
4002 encoding_str = 'Encodings: locale {}, fs {}, pref {}, {}'.format(
4003 locale.getpreferredencoding(),
4004 sys.getfilesystemencoding(),
4005 self.get_encoding(),
4006 ', '.join(
4007 f'{key} {get_encoding(stream)}' for key, stream in self._out_files.items_
4008 if stream is not None and key != 'console'),
4009 )
4010
4011 logger = self.params.get('logger')
4012 if logger:
4013 write_debug = lambda msg: logger.debug(f'[debug] {msg}')
4014 write_debug(encoding_str)
4015 else:
4016 write_string(f'[debug] {encoding_str}\n', encoding=None)
4017 write_debug = lambda msg: self._write_string(f'[debug] {msg}\n')
4018
4019 source = detect_variant()
4020 if VARIANT not in (None, 'pip'):
4021 source += '*'
4022 klass = type(self)
4023 write_debug(join_nonempty(
4024 f'{REPOSITORY.rpartition("/")[2]} version',
4025 _make_label(ORIGIN, CHANNEL.partition('@')[2] or __version__, __version__),
4026 f'[{RELEASE_GIT_HEAD[:9]}]' if RELEASE_GIT_HEAD else '',
4027 '' if source == 'unknown' else f'({source})',
4028 '' if _IN_CLI else 'API' if klass == YoutubeDL else f'API:{self.__module__}.{klass.__qualname__}',
4029 delim=' '))
4030
4031 if not _IN_CLI:
4032 write_debug(f'params: {self.params}')
4033
4034 if not _LAZY_LOADER:
4035 if os.environ.get('YTDLP_NO_LAZY_EXTRACTORS'):
4036 write_debug('Lazy loading extractors is forcibly disabled')
4037 else:
4038 write_debug('Lazy loading extractors is disabled')
4039 if self.params['compat_opts']:
4040 write_debug('Compatibility options: {}'.format(', '.join(self.params['compat_opts'])))
4041
4042 if current_git_head():
4043 write_debug(f'Git HEAD: {current_git_head()}')
4044 write_debug(system_identifier())
4045
4046 exe_versions, ffmpeg_features = FFmpegPostProcessor.get_versions_and_features(self)
4047 ffmpeg_features = {key for key, val in ffmpeg_features.items() if val}
4048 if ffmpeg_features:
4049 exe_versions['ffmpeg'] += ' ({})'.format(','.join(sorted(ffmpeg_features)))
4050
4051 exe_versions['rtmpdump'] = rtmpdump_version()
4052 exe_versions['phantomjs'] = PhantomJSwrapper._version()
4053 exe_str = ', '.join(
4054 f'{exe} {v}' for exe, v in sorted(exe_versions.items()) if v
4055 ) or 'none'
4056 write_debug(f'exe versions: {exe_str}')
4057
4058 from .compat.compat_utils import get_package_info
4059 from .dependencies import available_dependencies
4060
4061 write_debug('Optional libraries: %s' % (', '.join(sorted({
4062 join_nonempty(*get_package_info(m)) for m in available_dependencies.values()
4063 })) or 'none'))
4064
4065 write_debug(f'Proxy map: {self.proxies}')
4066 write_debug(f'Request Handlers: {", ".join(rh.RH_NAME for rh in self._request_director.handlers.values())}')
4067 for plugin_type, plugins in {'Extractor': plugin_ies, 'Post-Processor': plugin_pps}.items():
4068 display_list = ['{}{}'.format(
4069 klass.__name__, '' if klass.__name__ == name else f' as {name}')
4070 for name, klass in plugins.items()]
4071 if plugin_type == 'Extractor':
4072 display_list.extend(f'{plugins[-1].IE_NAME.partition("+")[2]} ({parent.__name__})'
4073 for parent, plugins in plugin_ie_overrides.items())
4074 if not display_list:
4075 continue
4076 write_debug(f'{plugin_type} Plugins: {", ".join(sorted(display_list))}')
4077
4078 plugin_dirs = plugin_directories()
4079 if plugin_dirs:
4080 write_debug(f'Plugin directories: {plugin_dirs}')
4081
4082 # Not implemented
4083 if False and self.params.get('call_home'):
4084 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode()
4085 write_debug(f'Public IP address: {ipaddr}')
4086 latest_version = self.urlopen(
4087 'https://yt-dl.org/latest/version').read().decode()
4088 if version_tuple(latest_version) > version_tuple(__version__):
4089 self.report_warning(
4090 f'You are using an outdated version (newest version: {latest_version})! '
4091 'See https://yt-dl.org/update if you need help updating.')
4092
4093 @functools.cached_property
4094 def proxies(self):
4095 """Global proxy configuration"""
4096 opts_proxy = self.params.get('proxy')
4097 if opts_proxy is not None:
4098 if opts_proxy == '':
4099 opts_proxy = '__noproxy__'
4100 proxies = {'all': opts_proxy}
4101 else:
4102 proxies = urllib.request.getproxies()
4103 # compat. Set HTTPS_PROXY to __noproxy__ to revert
4104 if 'http' in proxies and 'https' not in proxies:
4105 proxies['https'] = proxies['http']
4106
4107 return proxies
4108
4109 @functools.cached_property
4110 def cookiejar(self):
4111 """Global cookiejar instance"""
4112 return load_cookies(
4113 self.params.get('cookiefile'), self.params.get('cookiesfrombrowser'), self)
4114
4115 @property
4116 def _opener(self):
4117 """
4118 Get a urllib OpenerDirector from the Urllib handler (deprecated).
4119 """
4120 self.deprecation_warning('YoutubeDL._opener is deprecated, use YoutubeDL.urlopen()')
4121 handler = self._request_director.handlers['Urllib']
4122 return handler._get_instance(cookiejar=self.cookiejar, proxies=self.proxies)
4123
4124 def _get_available_impersonate_targets(self):
4125 # TODO(future): make available as public API
4126 return [
4127 (target, rh.RH_NAME)
4128 for rh in self._request_director.handlers.values()
4129 if isinstance(rh, ImpersonateRequestHandler)
4130 for target in rh.supported_targets
4131 ]
4132
4133 def _impersonate_target_available(self, target):
4134 # TODO(future): make available as public API
4135 return any(
4136 rh.is_supported_target(target)
4137 for rh in self._request_director.handlers.values()
4138 if isinstance(rh, ImpersonateRequestHandler))
4139
4140 def urlopen(self, req):
4141 """ Start an HTTP download """
4142 if isinstance(req, str):
4143 req = Request(req)
4144 elif isinstance(req, urllib.request.Request):
4145 self.deprecation_warning(
4146 'Passing a urllib.request.Request object to YoutubeDL.urlopen() is deprecated. '
4147 'Use yt_dlp.networking.common.Request instead.')
4148 req = urllib_req_to_req(req)
4149 assert isinstance(req, Request)
4150
4151 # compat: Assume user:pass url params are basic auth
4152 url, basic_auth_header = extract_basic_auth(req.url)
4153 if basic_auth_header:
4154 req.headers['Authorization'] = basic_auth_header
4155 req.url = sanitize_url(url)
4156
4157 clean_proxies(proxies=req.proxies, headers=req.headers)
4158 clean_headers(req.headers)
4159
4160 try:
4161 return self._request_director.send(req)
4162 except NoSupportingHandlers as e:
4163 for ue in e.unsupported_errors:
4164 # FIXME: This depends on the order of errors.
4165 if not (ue.handler and ue.msg):
4166 continue
4167 if ue.handler.RH_KEY == 'Urllib' and 'unsupported url scheme: "file"' in ue.msg.lower():
4168 raise RequestError(
4169 'file:// URLs are disabled by default in yt-dlp for security reasons. '
4170 'Use --enable-file-urls to enable at your own risk.', cause=ue) from ue
4171 if (
4172 'unsupported proxy type: "https"' in ue.msg.lower()
4173 and 'requests' not in self._request_director.handlers
4174 and 'curl_cffi' not in self._request_director.handlers
4175 ):
4176 raise RequestError(
4177 'To use an HTTPS proxy for this request, one of the following dependencies needs to be installed: requests, curl_cffi')
4178
4179 elif (
4180 re.match(r'unsupported url scheme: "wss?"', ue.msg.lower())
4181 and 'websockets' not in self._request_director.handlers
4182 ):
4183 raise RequestError(
4184 'This request requires WebSocket support. '
4185 'Ensure one of the following dependencies are installed: websockets',
4186 cause=ue) from ue
4187
4188 elif re.match(r'unsupported (?:extensions: impersonate|impersonate target)', ue.msg.lower()):
4189 raise RequestError(
4190 f'Impersonate target "{req.extensions["impersonate"]}" is not available.'
4191 f' See --list-impersonate-targets for available targets.'
4192 f' This request requires browser impersonation, however you may be missing dependencies'
4193 f' required to support this target.')
4194 raise
4195 except SSLError as e:
4196 if 'UNSAFE_LEGACY_RENEGOTIATION_DISABLED' in str(e):
4197 raise RequestError('UNSAFE_LEGACY_RENEGOTIATION_DISABLED: Try using --legacy-server-connect', cause=e) from e
4198 elif 'SSLV3_ALERT_HANDSHAKE_FAILURE' in str(e):
4199 raise RequestError(
4200 'SSLV3_ALERT_HANDSHAKE_FAILURE: The server may not support the current cipher list. '
4201 'Try using --legacy-server-connect', cause=e) from e
4202 raise
4203
4204 def build_request_director(self, handlers, preferences=None):
4205 logger = _YDLLogger(self)
4206 headers = self.params['http_headers'].copy()
4207 proxies = self.proxies.copy()
4208 clean_headers(headers)
4209 clean_proxies(proxies, headers)
4210
4211 director = RequestDirector(logger=logger, verbose=self.params.get('debug_printtraffic'))
4212 for handler in handlers:
4213 director.add_handler(handler(
4214 logger=logger,
4215 headers=headers,
4216 cookiejar=self.cookiejar,
4217 proxies=proxies,
4218 prefer_system_certs='no-certifi' in self.params['compat_opts'],
4219 verify=not self.params.get('nocheckcertificate'),
4220 **traverse_obj(self.params, {
4221 'verbose': 'debug_printtraffic',
4222 'source_address': 'source_address',
4223 'timeout': 'socket_timeout',
4224 'legacy_ssl_support': 'legacyserverconnect',
4225 'enable_file_urls': 'enable_file_urls',
4226 'impersonate': 'impersonate',
4227 'client_cert': {
4228 'client_certificate': 'client_certificate',
4229 'client_certificate_key': 'client_certificate_key',
4230 'client_certificate_password': 'client_certificate_password',
4231 },
4232 }),
4233 ))
4234 director.preferences.update(preferences or [])
4235 if 'prefer-legacy-http-handler' in self.params['compat_opts']:
4236 director.preferences.add(lambda rh, _: 500 if rh.RH_KEY == 'Urllib' else 0)
4237 return director
4238
4239 @functools.cached_property
4240 def _request_director(self):
4241 return self.build_request_director(_REQUEST_HANDLERS.values(), _RH_PREFERENCES)
4242
4243 def encode(self, s):
4244 if isinstance(s, bytes):
4245 return s # Already encoded
4246
4247 try:
4248 return s.encode(self.get_encoding())
4249 except UnicodeEncodeError as err:
4250 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
4251 raise
4252
4253 def get_encoding(self):
4254 encoding = self.params.get('encoding')
4255 if encoding is None:
4256 encoding = preferredencoding()
4257 return encoding
4258
4259 def _write_info_json(self, label, ie_result, infofn, overwrite=None):
4260 """ Write infojson and returns True = written, 'exists' = Already exists, False = skip, None = error """
4261 if overwrite is None:
4262 overwrite = self.params.get('overwrites', True)
4263 if not self.params.get('writeinfojson'):
4264 return False
4265 elif not infofn:
4266 self.write_debug(f'Skipping writing {label} infojson')
4267 return False
4268 elif not self._ensure_dir_exists(infofn):
4269 return None
4270 elif not overwrite and os.path.exists(infofn):
4271 self.to_screen(f'[info] {label.title()} metadata is already present')
4272 return 'exists'
4273
4274 self.to_screen(f'[info] Writing {label} metadata as JSON to: {infofn}')
4275 try:
4276 write_json_file(self.sanitize_info(ie_result, self.params.get('clean_infojson', True)), infofn)
4277 return True
4278 except OSError:
4279 self.report_error(f'Cannot write {label} metadata to JSON file {infofn}')
4280 return None
4281
4282 def _write_description(self, label, ie_result, descfn):
4283 """ Write description and returns True = written, False = skip, None = error """
4284 if not self.params.get('writedescription'):
4285 return False
4286 elif not descfn:
4287 self.write_debug(f'Skipping writing {label} description')
4288 return False
4289 elif not self._ensure_dir_exists(descfn):
4290 return None
4291 elif not self.params.get('overwrites', True) and os.path.exists(descfn):
4292 self.to_screen(f'[info] {label.title()} description is already present')
4293 elif ie_result.get('description') is None:
4294 self.to_screen(f'[info] There\'s no {label} description to write')
4295 return False
4296 else:
4297 try:
4298 self.to_screen(f'[info] Writing {label} description to: {descfn}')
4299 with open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
4300 descfile.write(ie_result['description'])
4301 except OSError:
4302 self.report_error(f'Cannot write {label} description file {descfn}')
4303 return None
4304 return True
4305
4306 def _write_subtitles(self, info_dict, filename):
4307 """ Write subtitles to file and return list of (sub_filename, final_sub_filename); or None if error"""
4308 ret = []
4309 subtitles = info_dict.get('requested_subtitles')
4310 if not (self.params.get('writesubtitles') or self.params.get('writeautomaticsub')):
4311 # subtitles download errors are already managed as troubles in relevant IE
4312 # that way it will silently go on when used with unsupporting IE
4313 return ret
4314 elif not subtitles:
4315 self.to_screen('[info] There are no subtitles for the requested languages')
4316 return ret
4317 sub_filename_base = self.prepare_filename(info_dict, 'subtitle')
4318 if not sub_filename_base:
4319 self.to_screen('[info] Skipping writing video subtitles')
4320 return ret
4321
4322 for sub_lang, sub_info in subtitles.items():
4323 sub_format = sub_info['ext']
4324 sub_filename = subtitles_filename(filename, sub_lang, sub_format, info_dict.get('ext'))
4325 sub_filename_final = subtitles_filename(sub_filename_base, sub_lang, sub_format, info_dict.get('ext'))
4326 existing_sub = self.existing_file((sub_filename_final, sub_filename))
4327 if existing_sub:
4328 self.to_screen(f'[info] Video subtitle {sub_lang}.{sub_format} is already present')
4329 sub_info['filepath'] = existing_sub
4330 ret.append((existing_sub, sub_filename_final))
4331 continue
4332
4333 self.to_screen(f'[info] Writing video subtitles to: {sub_filename}')
4334 if sub_info.get('data') is not None:
4335 try:
4336 # Use newline='' to prevent conversion of newline characters
4337 # See https://github.com/ytdl-org/youtube-dl/issues/10268
4338 with open(sub_filename, 'w', encoding='utf-8', newline='') as subfile:
4339 subfile.write(sub_info['data'])
4340 sub_info['filepath'] = sub_filename
4341 ret.append((sub_filename, sub_filename_final))
4342 continue
4343 except OSError:
4344 self.report_error(f'Cannot write video subtitles file {sub_filename}')
4345 return None
4346
4347 try:
4348 sub_copy = sub_info.copy()
4349 sub_copy.setdefault('http_headers', info_dict.get('http_headers'))
4350 self.dl(sub_filename, sub_copy, subtitle=True)
4351 sub_info['filepath'] = sub_filename
4352 ret.append((sub_filename, sub_filename_final))
4353 except (DownloadError, ExtractorError, OSError, ValueError, *network_exceptions) as err:
4354 msg = f'Unable to download video subtitles for {sub_lang!r}: {err}'
4355 if self.params.get('ignoreerrors') is not True: # False or 'only_download'
4356 if not self.params.get('ignoreerrors'):
4357 self.report_error(msg)
4358 raise DownloadError(msg)
4359 self.report_warning(msg)
4360 return ret
4361
4362 def _write_thumbnails(self, label, info_dict, filename, thumb_filename_base=None):
4363 """ Write thumbnails to file and return list of (thumb_filename, final_thumb_filename); or None if error """
4364 write_all = self.params.get('write_all_thumbnails', False)
4365 thumbnails, ret = [], []
4366 if write_all or self.params.get('writethumbnail', False):
4367 thumbnails = info_dict.get('thumbnails') or []
4368 if not thumbnails:
4369 self.to_screen(f'[info] There are no {label} thumbnails to download')
4370 return ret
4371 multiple = write_all and len(thumbnails) > 1
4372
4373 if thumb_filename_base is None:
4374 thumb_filename_base = filename
4375 if thumbnails and not thumb_filename_base:
4376 self.write_debug(f'Skipping writing {label} thumbnail')
4377 return ret
4378
4379 if thumbnails and not self._ensure_dir_exists(filename):
4380 return None
4381
4382 for idx, t in list(enumerate(thumbnails))[::-1]:
4383 thumb_ext = (f'{t["id"]}.' if multiple else '') + determine_ext(t['url'], 'jpg')
4384 thumb_display_id = f'{label} thumbnail {t["id"]}'
4385 thumb_filename = replace_extension(filename, thumb_ext, info_dict.get('ext'))
4386 thumb_filename_final = replace_extension(thumb_filename_base, thumb_ext, info_dict.get('ext'))
4387
4388 existing_thumb = self.existing_file((thumb_filename_final, thumb_filename))
4389 if existing_thumb:
4390 self.to_screen('[info] {} is already present'.format((
4391 thumb_display_id if multiple else f'{label} thumbnail').capitalize()))
4392 t['filepath'] = existing_thumb
4393 ret.append((existing_thumb, thumb_filename_final))
4394 else:
4395 self.to_screen(f'[info] Downloading {thumb_display_id} ...')
4396 try:
4397 uf = self.urlopen(Request(t['url'], headers=t.get('http_headers', {})))
4398 self.to_screen(f'[info] Writing {thumb_display_id} to: {thumb_filename}')
4399 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
4400 shutil.copyfileobj(uf, thumbf)
4401 ret.append((thumb_filename, thumb_filename_final))
4402 t['filepath'] = thumb_filename
4403 except network_exceptions as err:
4404 if isinstance(err, HTTPError) and err.status == 404:
4405 self.to_screen(f'[info] {thumb_display_id.title()} does not exist')
4406 else:
4407 self.report_warning(f'Unable to download {thumb_display_id}: {err}')
4408 thumbnails.pop(idx)
4409 if ret and not write_all:
4410 break
4411 return ret