]> jfr.im git - yt-dlp.git/blob - yt_dlp/YoutubeDL.py
[ie/mlbtv] Fix extraction (#10296)
[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 http.cookiejar
8 import io
9 import itertools
10 import json
11 import locale
12 import operator
13 import os
14 import random
15 import re
16 import shutil
17 import string
18 import subprocess
19 import sys
20 import tempfile
21 import time
22 import tokenize
23 import traceback
24 import unicodedata
25
26 from .cache import Cache
27 from .compat import functools, urllib # isort: split
28 from .compat import compat_os_name, urllib_req_to_req
29 from .cookies import LenientSimpleCookie, load_cookies
30 from .downloader import FFmpegFD, get_suitable_downloader, shorten_protocol_name
31 from .downloader.rtmp import rtmpdump_version
32 from .extractor import gen_extractor_classes, get_info_extractor
33 from .extractor.common import UnsupportedURLIE
34 from .extractor.openload import PhantomJSwrapper
35 from .minicurses import format_text
36 from .networking import HEADRequest, Request, RequestDirector
37 from .networking.common import _REQUEST_HANDLERS, _RH_PREFERENCES
38 from .networking.exceptions import (
39 HTTPError,
40 NoSupportingHandlers,
41 RequestError,
42 SSLError,
43 network_exceptions,
44 )
45 from .networking.impersonate import ImpersonateRequestHandler
46 from .plugins import directories as plugin_directories
47 from .postprocessor import _PLUGIN_CLASSES as plugin_pps
48 from .postprocessor import (
49 EmbedThumbnailPP,
50 FFmpegFixupDuplicateMoovPP,
51 FFmpegFixupDurationPP,
52 FFmpegFixupM3u8PP,
53 FFmpegFixupM4aPP,
54 FFmpegFixupStretchedPP,
55 FFmpegFixupTimestampPP,
56 FFmpegMergerPP,
57 FFmpegPostProcessor,
58 FFmpegVideoConvertorPP,
59 MoveFilesAfterDownloadPP,
60 get_postprocessor,
61 )
62 from .postprocessor.ffmpeg import resolve_mapping as resolve_recode_mapping
63 from .update import (
64 REPOSITORY,
65 _get_system_deprecation,
66 _make_label,
67 current_git_head,
68 detect_variant,
69 )
70 from .utils import (
71 DEFAULT_OUTTMPL,
72 IDENTITY,
73 LINK_TEMPLATES,
74 MEDIA_EXTENSIONS,
75 NO_DEFAULT,
76 NUMBER_RE,
77 OUTTMPL_TYPES,
78 POSTPROCESS_WHEN,
79 STR_FORMAT_RE_TMPL,
80 STR_FORMAT_TYPES,
81 ContentTooShortError,
82 DateRange,
83 DownloadCancelled,
84 DownloadError,
85 EntryNotInPlaylist,
86 ExistingVideoReached,
87 ExtractorError,
88 FormatSorter,
89 GeoRestrictedError,
90 ISO3166Utils,
91 LazyList,
92 MaxDownloadsReached,
93 Namespace,
94 PagedList,
95 PlaylistEntries,
96 Popen,
97 PostProcessingError,
98 ReExtractInfo,
99 RejectedVideoReached,
100 SameFileError,
101 UnavailableVideoError,
102 UserNotLive,
103 YoutubeDLError,
104 age_restricted,
105 bug_reports_message,
106 date_from_str,
107 deprecation_warning,
108 determine_ext,
109 determine_protocol,
110 encode_compat_str,
111 encodeFilename,
112 escapeHTML,
113 expand_path,
114 extract_basic_auth,
115 filter_dict,
116 float_or_none,
117 format_bytes,
118 format_decimal_suffix,
119 format_field,
120 formatSeconds,
121 get_compatible_ext,
122 get_domain,
123 int_or_none,
124 iri_to_uri,
125 is_path_like,
126 join_nonempty,
127 locked_file,
128 make_archive_id,
129 make_dir,
130 number_of_digits,
131 orderedSet,
132 orderedSet_from_options,
133 parse_filesize,
134 preferredencoding,
135 prepend_extension,
136 remove_terminal_sequences,
137 render_table,
138 replace_extension,
139 sanitize_filename,
140 sanitize_path,
141 sanitize_url,
142 shell_quote,
143 str_or_none,
144 strftime_or_none,
145 subtitles_filename,
146 supports_terminal_sequences,
147 system_identifier,
148 filesize_from_tbr,
149 timetuple_from_msec,
150 to_high_limit_path,
151 traverse_obj,
152 try_call,
153 try_get,
154 url_basename,
155 variadic,
156 version_tuple,
157 windows_enable_vt_mode,
158 write_json_file,
159 write_string,
160 )
161 from .utils._utils import _YDLLogger
162 from .utils.networking import (
163 HTTPHeaderDict,
164 clean_headers,
165 clean_proxies,
166 std_headers,
167 )
168 from .version import CHANNEL, ORIGIN, RELEASE_GIT_HEAD, VARIANT, __version__
169
170 if compat_os_name == 'nt':
171 import ctypes
172
173
174 class YoutubeDL:
175 """YoutubeDL class.
176
177 YoutubeDL objects are the ones responsible of downloading the
178 actual video file and writing it to disk if the user has requested
179 it, among some other tasks. In most cases there should be one per
180 program. As, given a video URL, the downloader doesn't know how to
181 extract all the needed information, task that InfoExtractors do, it
182 has to pass the URL to one of them.
183
184 For this, YoutubeDL objects have a method that allows
185 InfoExtractors to be registered in a given order. When it is passed
186 a URL, the YoutubeDL object handles it to the first InfoExtractor it
187 finds that reports being able to handle it. The InfoExtractor extracts
188 all the information about the video or videos the URL refers to, and
189 YoutubeDL process the extracted information, possibly using a File
190 Downloader to download the video.
191
192 YoutubeDL objects accept a lot of parameters. In order not to saturate
193 the object constructor with arguments, it receives a dictionary of
194 options instead. These options are available through the params
195 attribute for the InfoExtractors to use. The YoutubeDL also
196 registers itself as the downloader in charge for the InfoExtractors
197 that are added to it, so this is a "mutual registration".
198
199 Available options:
200
201 username: Username for authentication purposes.
202 password: Password for authentication purposes.
203 videopassword: Password for accessing a video.
204 ap_mso: Adobe Pass multiple-system operator identifier.
205 ap_username: Multiple-system operator account username.
206 ap_password: Multiple-system operator account password.
207 usenetrc: Use netrc for authentication instead.
208 netrc_location: Location of the netrc file. Defaults to ~/.netrc.
209 netrc_cmd: Use a shell command to get credentials
210 verbose: Print additional info to stdout.
211 quiet: Do not print messages to stdout.
212 no_warnings: Do not print out anything for warnings.
213 forceprint: A dict with keys WHEN mapped to a list of templates to
214 print to stdout. The allowed keys are video or any of the
215 items in utils.POSTPROCESS_WHEN.
216 For compatibility, a single list is also accepted
217 print_to_file: A dict with keys WHEN (same as forceprint) mapped to
218 a list of tuples with (template, filename)
219 forcejson: Force printing info_dict as JSON.
220 dump_single_json: Force printing the info_dict of the whole playlist
221 (or video) as a single JSON line.
222 force_write_download_archive: Force writing download archive regardless
223 of 'skip_download' or 'simulate'.
224 simulate: Do not download the video files. If unset (or None),
225 simulate only if listsubtitles, listformats or list_thumbnails is used
226 format: Video format code. see "FORMAT SELECTION" for more details.
227 You can also pass a function. The function takes 'ctx' as
228 argument and returns the formats to download.
229 See "build_format_selector" for an implementation
230 allow_unplayable_formats: Allow unplayable formats to be extracted and downloaded.
231 ignore_no_formats_error: Ignore "No video formats" error. Usefull for
232 extracting metadata even if the video is not actually
233 available for download (experimental)
234 format_sort: A list of fields by which to sort the video formats.
235 See "Sorting Formats" for more details.
236 format_sort_force: Force the given format_sort. see "Sorting Formats"
237 for more details.
238 prefer_free_formats: Whether to prefer video formats with free containers
239 over non-free ones of same quality.
240 allow_multiple_video_streams: Allow multiple video streams to be merged
241 into a single file
242 allow_multiple_audio_streams: Allow multiple audio streams to be merged
243 into a single file
244 check_formats Whether to test if the formats are downloadable.
245 Can be True (check all), False (check none),
246 'selected' (check selected formats),
247 or None (check only if requested by extractor)
248 paths: Dictionary of output paths. The allowed keys are 'home'
249 'temp' and the keys of OUTTMPL_TYPES (in utils/_utils.py)
250 outtmpl: Dictionary of templates for output names. Allowed keys
251 are 'default' and the keys of OUTTMPL_TYPES (in utils/_utils.py).
252 For compatibility with youtube-dl, a single string can also be used
253 outtmpl_na_placeholder: Placeholder for unavailable meta fields.
254 restrictfilenames: Do not allow "&" and spaces in file names
255 trim_file_name: Limit length of filename (extension excluded)
256 windowsfilenames: Force the filenames to be windows compatible
257 ignoreerrors: Do not stop on download/postprocessing errors.
258 Can be 'only_download' to ignore only download errors.
259 Default is 'only_download' for CLI, but False for API
260 skip_playlist_after_errors: Number of allowed failures until the rest of
261 the playlist is skipped
262 allowed_extractors: List of regexes to match against extractor names that are allowed
263 overwrites: Overwrite all video and metadata files if True,
264 overwrite only non-video files if None
265 and don't overwrite any file if False
266 playlist_items: Specific indices of playlist to download.
267 playlistrandom: Download playlist items in random order.
268 lazy_playlist: Process playlist entries as they are received.
269 matchtitle: Download only matching titles.
270 rejecttitle: Reject downloads for matching titles.
271 logger: Log messages to a logging.Logger instance.
272 logtostderr: Print everything to stderr instead of stdout.
273 consoletitle: Display progress in console window's titlebar.
274 writedescription: Write the video description to a .description file
275 writeinfojson: Write the video description to a .info.json file
276 clean_infojson: Remove internal metadata from the infojson
277 getcomments: Extract video comments. This will not be written to disk
278 unless writeinfojson is also given
279 writeannotations: Write the video annotations to a .annotations.xml file
280 writethumbnail: Write the thumbnail image to a file
281 allow_playlist_files: Whether to write playlists' description, infojson etc
282 also to disk when using the 'write*' options
283 write_all_thumbnails: Write all thumbnail formats to files
284 writelink: Write an internet shortcut file, depending on the
285 current platform (.url/.webloc/.desktop)
286 writeurllink: Write a Windows internet shortcut file (.url)
287 writewebloclink: Write a macOS internet shortcut file (.webloc)
288 writedesktoplink: Write a Linux internet shortcut file (.desktop)
289 writesubtitles: Write the video subtitles to a file
290 writeautomaticsub: Write the automatically generated subtitles to a file
291 listsubtitles: Lists all available subtitles for the video
292 subtitlesformat: The format code for subtitles
293 subtitleslangs: List of languages of the subtitles to download (can be regex).
294 The list may contain "all" to refer to all the available
295 subtitles. The language can be prefixed with a "-" to
296 exclude it from the requested languages, e.g. ['all', '-live_chat']
297 keepvideo: Keep the video file after post-processing
298 daterange: A utils.DateRange object, download only if the upload_date is in the range.
299 skip_download: Skip the actual download of the video file
300 cachedir: Location of the cache files in the filesystem.
301 False to disable filesystem cache.
302 noplaylist: Download single video instead of a playlist if in doubt.
303 age_limit: An integer representing the user's age in years.
304 Unsuitable videos for the given age are skipped.
305 min_views: An integer representing the minimum view count the video
306 must have in order to not be skipped.
307 Videos without view count information are always
308 downloaded. None for no limit.
309 max_views: An integer representing the maximum view count.
310 Videos that are more popular than that are not
311 downloaded.
312 Videos without view count information are always
313 downloaded. None for no limit.
314 download_archive: A set, or the name of a file where all downloads are recorded.
315 Videos already present in the file are not downloaded again.
316 break_on_existing: Stop the download process after attempting to download a
317 file that is in the archive.
318 break_per_url: Whether break_on_reject and break_on_existing
319 should act on each input URL as opposed to for the entire queue
320 cookiefile: File name or text stream from where cookies should be read and dumped to
321 cookiesfrombrowser: A tuple containing the name of the browser, the profile
322 name/path from where cookies are loaded, the name of the keyring,
323 and the container name, e.g. ('chrome', ) or
324 ('vivaldi', 'default', 'BASICTEXT') or ('firefox', 'default', None, 'Meta')
325 legacyserverconnect: Explicitly allow HTTPS connection to servers that do not
326 support RFC 5746 secure renegotiation
327 nocheckcertificate: Do not verify SSL certificates
328 client_certificate: Path to client certificate file in PEM format. May include the private key
329 client_certificate_key: Path to private key file for client certificate
330 client_certificate_password: Password for client certificate private key, if encrypted.
331 If not provided and the key is encrypted, yt-dlp will ask interactively
332 prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
333 (Only supported by some extractors)
334 enable_file_urls: Enable file:// URLs. This is disabled by default for security reasons.
335 http_headers: A dictionary of custom headers to be used for all requests
336 proxy: URL of the proxy server to use
337 geo_verification_proxy: URL of the proxy to use for IP address verification
338 on geo-restricted sites.
339 socket_timeout: Time to wait for unresponsive hosts, in seconds
340 bidi_workaround: Work around buggy terminals without bidirectional text
341 support, using fridibi
342 debug_printtraffic:Print out sent and received HTTP traffic
343 default_search: Prepend this string if an input url is not valid.
344 'auto' for elaborate guessing
345 encoding: Use this encoding instead of the system-specified.
346 extract_flat: Whether to resolve and process url_results further
347 * False: Always process. Default for API
348 * True: Never process
349 * 'in_playlist': Do not process inside playlist/multi_video
350 * 'discard': Always process, but don't return the result
351 from inside playlist/multi_video
352 * 'discard_in_playlist': Same as "discard", but only for
353 playlists (not multi_video). Default for CLI
354 wait_for_video: If given, wait for scheduled streams to become available.
355 The value should be a tuple containing the range
356 (min_secs, max_secs) to wait between retries
357 postprocessors: A list of dictionaries, each with an entry
358 * key: The name of the postprocessor. See
359 yt_dlp/postprocessor/__init__.py for a list.
360 * when: When to run the postprocessor. Allowed values are
361 the entries of utils.POSTPROCESS_WHEN
362 Assumed to be 'post_process' if not given
363 progress_hooks: A list of functions that get called on download
364 progress, with a dictionary with the entries
365 * status: One of "downloading", "error", or "finished".
366 Check this first and ignore unknown values.
367 * info_dict: The extracted info_dict
368
369 If status is one of "downloading", or "finished", the
370 following properties may also be present:
371 * filename: The final filename (always present)
372 * tmpfilename: The filename we're currently writing to
373 * downloaded_bytes: Bytes on disk
374 * total_bytes: Size of the whole file, None if unknown
375 * total_bytes_estimate: Guess of the eventual file size,
376 None if unavailable.
377 * elapsed: The number of seconds since download started.
378 * eta: The estimated time in seconds, None if unknown
379 * speed: The download speed in bytes/second, None if
380 unknown
381 * fragment_index: The counter of the currently
382 downloaded video fragment.
383 * fragment_count: The number of fragments (= individual
384 files that will be merged)
385
386 Progress hooks are guaranteed to be called at least once
387 (with status "finished") if the download is successful.
388 postprocessor_hooks: A list of functions that get called on postprocessing
389 progress, with a dictionary with the entries
390 * status: One of "started", "processing", or "finished".
391 Check this first and ignore unknown values.
392 * postprocessor: Name of the postprocessor
393 * info_dict: The extracted info_dict
394
395 Progress hooks are guaranteed to be called at least twice
396 (with status "started" and "finished") if the processing is successful.
397 merge_output_format: "/" separated list of extensions to use when merging formats.
398 final_ext: Expected final extension; used to detect when the file was
399 already downloaded and converted
400 fixup: Automatically correct known faults of the file.
401 One of:
402 - "never": do nothing
403 - "warn": only emit a warning
404 - "detect_or_warn": check whether we can do anything
405 about it, warn otherwise (default)
406 source_address: Client-side IP address to bind to.
407 impersonate: Client to impersonate for requests.
408 An ImpersonateTarget (from yt_dlp.networking.impersonate)
409 sleep_interval_requests: Number of seconds to sleep between requests
410 during extraction
411 sleep_interval: Number of seconds to sleep before each download when
412 used alone or a lower bound of a range for randomized
413 sleep before each download (minimum possible number
414 of seconds to sleep) when used along with
415 max_sleep_interval.
416 max_sleep_interval:Upper bound of a range for randomized sleep before each
417 download (maximum possible number of seconds to sleep).
418 Must only be used along with sleep_interval.
419 Actual sleep time will be a random float from range
420 [sleep_interval; max_sleep_interval].
421 sleep_interval_subtitles: Number of seconds to sleep before each subtitle download
422 listformats: Print an overview of available video formats and exit.
423 list_thumbnails: Print a table of all thumbnails and exit.
424 match_filter: A function that gets called for every video with the signature
425 (info_dict, *, incomplete: bool) -> Optional[str]
426 For backward compatibility with youtube-dl, the signature
427 (info_dict) -> Optional[str] is also allowed.
428 - If it returns a message, the video is ignored.
429 - If it returns None, the video is downloaded.
430 - If it returns utils.NO_DEFAULT, the user is interactively
431 asked whether to download the video.
432 - Raise utils.DownloadCancelled(msg) to abort remaining
433 downloads when a video is rejected.
434 match_filter_func in utils/_utils.py is one example for this.
435 color: A Dictionary with output stream names as keys
436 and their respective color policy as values.
437 Can also just be a single color policy,
438 in which case it applies to all outputs.
439 Valid stream names are 'stdout' and 'stderr'.
440 Valid color policies are one of 'always', 'auto', 'no_color' or 'never'.
441 geo_bypass: Bypass geographic restriction via faking X-Forwarded-For
442 HTTP header
443 geo_bypass_country:
444 Two-letter ISO 3166-2 country code that will be used for
445 explicit geographic restriction bypassing via faking
446 X-Forwarded-For HTTP header
447 geo_bypass_ip_block:
448 IP range in CIDR notation that will be used similarly to
449 geo_bypass_country
450 external_downloader: A dictionary of protocol keys and the executable of the
451 external downloader to use for it. The allowed protocols
452 are default|http|ftp|m3u8|dash|rtsp|rtmp|mms.
453 Set the value to 'native' to use the native downloader
454 compat_opts: Compatibility options. See "Differences in default behavior".
455 The following options do not work when used through the API:
456 filename, abort-on-error, multistreams, no-live-chat, format-sort
457 no-clean-infojson, no-playlist-metafiles, no-keep-subs, no-attach-info-json.
458 Refer __init__.py for their implementation
459 progress_template: Dictionary of templates for progress outputs.
460 Allowed keys are 'download', 'postprocess',
461 'download-title' (console title) and 'postprocess-title'.
462 The template is mapped on a dictionary with keys 'progress' and 'info'
463 retry_sleep_functions: Dictionary of functions that takes the number of attempts
464 as argument and returns the time to sleep in seconds.
465 Allowed keys are 'http', 'fragment', 'file_access'
466 download_ranges: A callback function that gets called for every video with
467 the signature (info_dict, ydl) -> Iterable[Section].
468 Only the returned sections will be downloaded.
469 Each Section is a dict with the following keys:
470 * start_time: Start time of the section in seconds
471 * end_time: End time of the section in seconds
472 * title: Section title (Optional)
473 * index: Section number (Optional)
474 force_keyframes_at_cuts: Re-encode the video when downloading ranges to get precise cuts
475 noprogress: Do not print the progress bar
476 live_from_start: Whether to download livestreams videos from the start
477
478 The following parameters are not used by YoutubeDL itself, they are used by
479 the downloader (see yt_dlp/downloader/common.py):
480 nopart, updatetime, buffersize, ratelimit, throttledratelimit, min_filesize,
481 max_filesize, test, noresizebuffer, retries, file_access_retries, fragment_retries,
482 continuedl, xattr_set_filesize, hls_use_mpegts, http_chunk_size,
483 external_downloader_args, concurrent_fragment_downloads, progress_delta.
484
485 The following options are used by the post processors:
486 ffmpeg_location: Location of the ffmpeg/avconv binary; either the path
487 to the binary or its containing directory.
488 postprocessor_args: A dictionary of postprocessor/executable keys (in lower case)
489 and a list of additional command-line arguments for the
490 postprocessor/executable. The dict can also have "PP+EXE" keys
491 which are used when the given exe is used by the given PP.
492 Use 'default' as the name for arguments to passed to all PP
493 For compatibility with youtube-dl, a single list of args
494 can also be used
495
496 The following options are used by the extractors:
497 extractor_retries: Number of times to retry for known errors (default: 3)
498 dynamic_mpd: Whether to process dynamic DASH manifests (default: True)
499 hls_split_discontinuity: Split HLS playlists to different formats at
500 discontinuities such as ad breaks (default: False)
501 extractor_args: A dictionary of arguments to be passed to the extractors.
502 See "EXTRACTOR ARGUMENTS" for details.
503 E.g. {'youtube': {'skip': ['dash', 'hls']}}
504 mark_watched: Mark videos watched (even with --simulate). Only for YouTube
505
506 The following options are deprecated and may be removed in the future:
507
508 break_on_reject: Stop the download process when encountering a video that
509 has been filtered out.
510 - `raise DownloadCancelled(msg)` in match_filter instead
511 force_generic_extractor: Force downloader to use the generic extractor
512 - Use allowed_extractors = ['generic', 'default']
513 playliststart: - Use playlist_items
514 Playlist item to start at.
515 playlistend: - Use playlist_items
516 Playlist item to end at.
517 playlistreverse: - Use playlist_items
518 Download playlist items in reverse order.
519 forceurl: - Use forceprint
520 Force printing final URL.
521 forcetitle: - Use forceprint
522 Force printing title.
523 forceid: - Use forceprint
524 Force printing ID.
525 forcethumbnail: - Use forceprint
526 Force printing thumbnail URL.
527 forcedescription: - Use forceprint
528 Force printing description.
529 forcefilename: - Use forceprint
530 Force printing final filename.
531 forceduration: - Use forceprint
532 Force printing duration.
533 allsubtitles: - Use subtitleslangs = ['all']
534 Downloads all the subtitles of the video
535 (requires writesubtitles or writeautomaticsub)
536 include_ads: - Doesn't work
537 Download ads as well
538 call_home: - Not implemented
539 Boolean, true iff we are allowed to contact the
540 yt-dlp servers for debugging.
541 post_hooks: - Register a custom postprocessor
542 A list of functions that get called as the final step
543 for each video file, after all postprocessors have been
544 called. The filename will be passed as the only argument.
545 hls_prefer_native: - Use external_downloader = {'m3u8': 'native'} or {'m3u8': 'ffmpeg'}.
546 Use the native HLS downloader instead of ffmpeg/avconv
547 if True, otherwise use ffmpeg/avconv if False, otherwise
548 use downloader suggested by extractor if None.
549 prefer_ffmpeg: - avconv support is deprecated
550 If False, use avconv instead of ffmpeg if both are available,
551 otherwise prefer ffmpeg.
552 youtube_include_dash_manifest: - Use extractor_args
553 If True (default), DASH manifests and related
554 data will be downloaded and processed by extractor.
555 You can reduce network I/O by disabling it if you don't
556 care about DASH. (only for youtube)
557 youtube_include_hls_manifest: - Use extractor_args
558 If True (default), HLS manifests and related
559 data will be downloaded and processed by extractor.
560 You can reduce network I/O by disabling it if you don't
561 care about HLS. (only for youtube)
562 no_color: Same as `color='no_color'`
563 no_overwrites: Same as `overwrites=False`
564 """
565
566 _NUMERIC_FIELDS = {
567 'width', 'height', 'asr', 'audio_channels', 'fps',
568 'tbr', 'abr', 'vbr', 'filesize', 'filesize_approx',
569 'timestamp', 'release_timestamp',
570 'duration', 'view_count', 'like_count', 'dislike_count', 'repost_count',
571 'average_rating', 'comment_count', 'age_limit',
572 'start_time', 'end_time',
573 'chapter_number', 'season_number', 'episode_number',
574 'track_number', 'disc_number', 'release_year',
575 }
576
577 _format_fields = {
578 # NB: Keep in sync with the docstring of extractor/common.py
579 'url', 'manifest_url', 'manifest_stream_number', 'ext', 'format', 'format_id', 'format_note',
580 'width', 'height', 'aspect_ratio', 'resolution', 'dynamic_range', 'tbr', 'abr', 'acodec', 'asr', 'audio_channels',
581 'vbr', 'fps', 'vcodec', 'container', 'filesize', 'filesize_approx', 'rows', 'columns',
582 'player_url', 'protocol', 'fragment_base_url', 'fragments', 'is_from_start', 'is_dash_periods', 'request_data',
583 'preference', 'language', 'language_preference', 'quality', 'source_preference', 'cookies',
584 'http_headers', 'stretched_ratio', 'no_resume', 'has_drm', 'extra_param_to_segment_url', 'extra_param_to_key_url',
585 'hls_aes', 'downloader_options', 'page_url', 'app', 'play_path', 'tc_url', 'flash_version',
586 'rtmp_live', 'rtmp_conn', 'rtmp_protocol', 'rtmp_real_time',
587 }
588 _deprecated_multivalue_fields = {
589 'album_artist': 'album_artists',
590 'artist': 'artists',
591 'composer': 'composers',
592 'creator': 'creators',
593 'genre': 'genres',
594 }
595 _format_selection_exts = {
596 'audio': set(MEDIA_EXTENSIONS.common_audio),
597 'video': {*MEDIA_EXTENSIONS.common_video, '3gp'},
598 'storyboards': set(MEDIA_EXTENSIONS.storyboards),
599 }
600
601 def __init__(self, params=None, auto_init=True):
602 """Create a FileDownloader object with the given options.
603 @param auto_init Whether to load the default extractors and print header (if verbose).
604 Set to 'no_verbose_header' to not print the header
605 """
606 if params is None:
607 params = {}
608 self.params = params
609 self._ies = {}
610 self._ies_instances = {}
611 self._pps = {k: [] for k in POSTPROCESS_WHEN}
612 self._printed_messages = set()
613 self._first_webpage_request = True
614 self._post_hooks = []
615 self._progress_hooks = []
616 self._postprocessor_hooks = []
617 self._download_retcode = 0
618 self._num_downloads = 0
619 self._num_videos = 0
620 self._playlist_level = 0
621 self._playlist_urls = set()
622 self.cache = Cache(self)
623 self.__header_cookies = []
624
625 stdout = sys.stderr if self.params.get('logtostderr') else sys.stdout
626 self._out_files = Namespace(
627 out=stdout,
628 error=sys.stderr,
629 screen=sys.stderr if self.params.get('quiet') else stdout,
630 console=None if compat_os_name == 'nt' else next(
631 filter(supports_terminal_sequences, (sys.stderr, sys.stdout)), None),
632 )
633
634 try:
635 windows_enable_vt_mode()
636 except Exception as e:
637 self.write_debug(f'Failed to enable VT mode: {e}')
638
639 if self.params.get('no_color'):
640 if self.params.get('color') is not None:
641 self.params.setdefault('_warnings', []).append(
642 'Overwriting params from "color" with "no_color"')
643 self.params['color'] = 'no_color'
644
645 term_allow_color = os.getenv('TERM', '').lower() != 'dumb'
646 no_color = bool(os.getenv('NO_COLOR'))
647
648 def process_color_policy(stream):
649 stream_name = {sys.stdout: 'stdout', sys.stderr: 'stderr'}[stream]
650 policy = traverse_obj(self.params, ('color', (stream_name, None), {str}), get_all=False)
651 if policy in ('auto', None):
652 if term_allow_color and supports_terminal_sequences(stream):
653 return 'no_color' if no_color else True
654 return False
655 assert policy in ('always', 'never', 'no_color'), policy
656 return {'always': True, 'never': False}.get(policy, policy)
657
658 self._allow_colors = Namespace(**{
659 name: process_color_policy(stream)
660 for name, stream in self._out_files.items_ if name != 'console'
661 })
662
663 system_deprecation = _get_system_deprecation()
664 if system_deprecation:
665 self.deprecated_feature(system_deprecation.replace('\n', '\n '))
666
667 if self.params.get('allow_unplayable_formats'):
668 self.report_warning(
669 f'You have asked for {self._format_err("UNPLAYABLE", self.Styles.EMPHASIS)} formats to be listed/downloaded. '
670 'This is a developer option intended for debugging. \n'
671 ' If you experience any issues while using this option, '
672 f'{self._format_err("DO NOT", self.Styles.ERROR)} open a bug report')
673
674 if self.params.get('bidi_workaround', False):
675 try:
676 import pty
677 master, slave = pty.openpty()
678 width = shutil.get_terminal_size().columns
679 width_args = [] if width is None else ['-w', str(width)]
680 sp_kwargs = {'stdin': subprocess.PIPE, 'stdout': slave, 'stderr': self._out_files.error}
681 try:
682 self._output_process = Popen(['bidiv', *width_args], **sp_kwargs)
683 except OSError:
684 self._output_process = Popen(['fribidi', '-c', 'UTF-8', *width_args], **sp_kwargs)
685 self._output_channel = os.fdopen(master, 'rb')
686 except OSError as ose:
687 if ose.errno == errno.ENOENT:
688 self.report_warning(
689 'Could not find fribidi executable, ignoring --bidi-workaround. '
690 'Make sure that fribidi is an executable file in one of the directories in your $PATH.')
691 else:
692 raise
693
694 self.params['compat_opts'] = set(self.params.get('compat_opts', ()))
695 self.params['http_headers'] = HTTPHeaderDict(std_headers, self.params.get('http_headers'))
696 self._load_cookies(self.params['http_headers'].get('Cookie')) # compat
697 self.params['http_headers'].pop('Cookie', None)
698
699 if auto_init and auto_init != 'no_verbose_header':
700 self.print_debug_header()
701
702 def check_deprecated(param, option, suggestion):
703 if self.params.get(param) is not None:
704 self.report_warning(f'{option} is deprecated. Use {suggestion} instead')
705 return True
706 return False
707
708 if check_deprecated('cn_verification_proxy', '--cn-verification-proxy', '--geo-verification-proxy'):
709 if self.params.get('geo_verification_proxy') is None:
710 self.params['geo_verification_proxy'] = self.params['cn_verification_proxy']
711
712 check_deprecated('autonumber', '--auto-number', '-o "%(autonumber)s-%(title)s.%(ext)s"')
713 check_deprecated('usetitle', '--title', '-o "%(title)s-%(id)s.%(ext)s"')
714 check_deprecated('useid', '--id', '-o "%(id)s.%(ext)s"')
715
716 for msg in self.params.get('_warnings', []):
717 self.report_warning(msg)
718 for msg in self.params.get('_deprecation_warnings', []):
719 self.deprecated_feature(msg)
720
721 if impersonate_target := self.params.get('impersonate'):
722 if not self._impersonate_target_available(impersonate_target):
723 raise YoutubeDLError(
724 f'Impersonate target "{impersonate_target}" is not available. '
725 f'Use --list-impersonate-targets to see available targets. '
726 f'You may be missing dependencies required to support this target.')
727
728 if 'list-formats' in self.params['compat_opts']:
729 self.params['listformats_table'] = False
730
731 if 'overwrites' not in self.params and self.params.get('nooverwrites') is not None:
732 # nooverwrites was unnecessarily changed to overwrites
733 # in 0c3d0f51778b153f65c21906031c2e091fcfb641
734 # This ensures compatibility with both keys
735 self.params['overwrites'] = not self.params['nooverwrites']
736 elif self.params.get('overwrites') is None:
737 self.params.pop('overwrites', None)
738 else:
739 self.params['nooverwrites'] = not self.params['overwrites']
740
741 if self.params.get('simulate') is None and any((
742 self.params.get('list_thumbnails'),
743 self.params.get('listformats'),
744 self.params.get('listsubtitles'),
745 )):
746 self.params['simulate'] = 'list_only'
747
748 self.params.setdefault('forceprint', {})
749 self.params.setdefault('print_to_file', {})
750
751 # Compatibility with older syntax
752 if not isinstance(params['forceprint'], dict):
753 self.params['forceprint'] = {'video': params['forceprint']}
754
755 if auto_init:
756 self.add_default_info_extractors()
757
758 if (sys.platform != 'win32'
759 and sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968']
760 and not self.params.get('restrictfilenames', False)):
761 # Unicode filesystem API will throw errors (#1474, #13027)
762 self.report_warning(
763 'Assuming --restrict-filenames since file system encoding '
764 'cannot encode all characters. '
765 'Set the LC_ALL environment variable to fix this.')
766 self.params['restrictfilenames'] = True
767
768 self._parse_outtmpl()
769
770 # Creating format selector here allows us to catch syntax errors before the extraction
771 self.format_selector = (
772 self.params.get('format') if self.params.get('format') in (None, '-')
773 else self.params['format'] if callable(self.params['format'])
774 else self.build_format_selector(self.params['format']))
775
776 hooks = {
777 'post_hooks': self.add_post_hook,
778 'progress_hooks': self.add_progress_hook,
779 'postprocessor_hooks': self.add_postprocessor_hook,
780 }
781 for opt, fn in hooks.items():
782 for ph in self.params.get(opt, []):
783 fn(ph)
784
785 for pp_def_raw in self.params.get('postprocessors', []):
786 pp_def = dict(pp_def_raw)
787 when = pp_def.pop('when', 'post_process')
788 self.add_post_processor(
789 get_postprocessor(pp_def.pop('key'))(self, **pp_def),
790 when=when)
791
792 def preload_download_archive(fn):
793 """Preload the archive, if any is specified"""
794 archive = set()
795 if fn is None:
796 return archive
797 elif not is_path_like(fn):
798 return fn
799
800 self.write_debug(f'Loading archive file {fn!r}')
801 try:
802 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
803 for line in archive_file:
804 archive.add(line.strip())
805 except OSError as ioe:
806 if ioe.errno != errno.ENOENT:
807 raise
808 return archive
809
810 self.archive = preload_download_archive(self.params.get('download_archive'))
811
812 def warn_if_short_id(self, argv):
813 # short YouTube ID starting with dash?
814 idxs = [
815 i for i, a in enumerate(argv)
816 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
817 if idxs:
818 correct_argv = (
819 ['yt-dlp']
820 + [a for i, a in enumerate(argv) if i not in idxs]
821 + ['--'] + [argv[i] for i in idxs]
822 )
823 self.report_warning(
824 'Long argument string detected. '
825 f'Use -- to separate parameters and URLs, like this:\n{shell_quote(correct_argv)}')
826
827 def add_info_extractor(self, ie):
828 """Add an InfoExtractor object to the end of the list."""
829 ie_key = ie.ie_key()
830 self._ies[ie_key] = ie
831 if not isinstance(ie, type):
832 self._ies_instances[ie_key] = ie
833 ie.set_downloader(self)
834
835 def get_info_extractor(self, ie_key):
836 """
837 Get an instance of an IE with name ie_key, it will try to get one from
838 the _ies list, if there's no instance it will create a new one and add
839 it to the extractor list.
840 """
841 ie = self._ies_instances.get(ie_key)
842 if ie is None:
843 ie = get_info_extractor(ie_key)()
844 self.add_info_extractor(ie)
845 return ie
846
847 def add_default_info_extractors(self):
848 """
849 Add the InfoExtractors returned by gen_extractors to the end of the list
850 """
851 all_ies = {ie.IE_NAME.lower(): ie for ie in gen_extractor_classes()}
852 all_ies['end'] = UnsupportedURLIE()
853 try:
854 ie_names = orderedSet_from_options(
855 self.params.get('allowed_extractors', ['default']), {
856 'all': list(all_ies),
857 'default': [name for name, ie in all_ies.items() if ie._ENABLED],
858 }, use_regex=True)
859 except re.error as e:
860 raise ValueError(f'Wrong regex for allowed_extractors: {e.pattern}')
861 for name in ie_names:
862 self.add_info_extractor(all_ies[name])
863 self.write_debug(f'Loaded {len(ie_names)} extractors')
864
865 def add_post_processor(self, pp, when='post_process'):
866 """Add a PostProcessor object to the end of the chain."""
867 assert when in POSTPROCESS_WHEN, f'Invalid when={when}'
868 self._pps[when].append(pp)
869 pp.set_downloader(self)
870
871 def add_post_hook(self, ph):
872 """Add the post hook"""
873 self._post_hooks.append(ph)
874
875 def add_progress_hook(self, ph):
876 """Add the download progress hook"""
877 self._progress_hooks.append(ph)
878
879 def add_postprocessor_hook(self, ph):
880 """Add the postprocessing progress hook"""
881 self._postprocessor_hooks.append(ph)
882 for pps in self._pps.values():
883 for pp in pps:
884 pp.add_progress_hook(ph)
885
886 def _bidi_workaround(self, message):
887 if not hasattr(self, '_output_channel'):
888 return message
889
890 assert hasattr(self, '_output_process')
891 assert isinstance(message, str)
892 line_count = message.count('\n') + 1
893 self._output_process.stdin.write((message + '\n').encode())
894 self._output_process.stdin.flush()
895 res = ''.join(self._output_channel.readline().decode()
896 for _ in range(line_count))
897 return res[:-len('\n')]
898
899 def _write_string(self, message, out=None, only_once=False):
900 if only_once:
901 if message in self._printed_messages:
902 return
903 self._printed_messages.add(message)
904 write_string(message, out=out, encoding=self.params.get('encoding'))
905
906 def to_stdout(self, message, skip_eol=False, quiet=None):
907 """Print message to stdout"""
908 if quiet is not None:
909 self.deprecation_warning('"YoutubeDL.to_stdout" no longer accepts the argument quiet. '
910 'Use "YoutubeDL.to_screen" instead')
911 if skip_eol is not False:
912 self.deprecation_warning('"YoutubeDL.to_stdout" no longer accepts the argument skip_eol. '
913 'Use "YoutubeDL.to_screen" instead')
914 self._write_string(f'{self._bidi_workaround(message)}\n', self._out_files.out)
915
916 def to_screen(self, message, skip_eol=False, quiet=None, only_once=False):
917 """Print message to screen if not in quiet mode"""
918 if self.params.get('logger'):
919 self.params['logger'].debug(message)
920 return
921 if (self.params.get('quiet') if quiet is None else quiet) and not self.params.get('verbose'):
922 return
923 self._write_string(
924 '{}{}'.format(self._bidi_workaround(message), ('' if skip_eol else '\n')),
925 self._out_files.screen, only_once=only_once)
926
927 def to_stderr(self, message, only_once=False):
928 """Print message to stderr"""
929 assert isinstance(message, str)
930 if self.params.get('logger'):
931 self.params['logger'].error(message)
932 else:
933 self._write_string(f'{self._bidi_workaround(message)}\n', self._out_files.error, only_once=only_once)
934
935 def _send_console_code(self, code):
936 if compat_os_name == 'nt' or not self._out_files.console:
937 return
938 self._write_string(code, self._out_files.console)
939
940 def to_console_title(self, message):
941 if not self.params.get('consoletitle', False):
942 return
943 message = remove_terminal_sequences(message)
944 if compat_os_name == 'nt':
945 if ctypes.windll.kernel32.GetConsoleWindow():
946 # c_wchar_p() might not be necessary if `message` is
947 # already of type unicode()
948 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
949 else:
950 self._send_console_code(f'\033]0;{message}\007')
951
952 def save_console_title(self):
953 if not self.params.get('consoletitle') or self.params.get('simulate'):
954 return
955 self._send_console_code('\033[22;0t') # Save the title on stack
956
957 def restore_console_title(self):
958 if not self.params.get('consoletitle') or self.params.get('simulate'):
959 return
960 self._send_console_code('\033[23;0t') # Restore the title from stack
961
962 def __enter__(self):
963 self.save_console_title()
964 return self
965
966 def save_cookies(self):
967 if self.params.get('cookiefile') is not None:
968 self.cookiejar.save()
969
970 def __exit__(self, *args):
971 self.restore_console_title()
972 self.close()
973
974 def close(self):
975 self.save_cookies()
976 if '_request_director' in self.__dict__:
977 self._request_director.close()
978 del self._request_director
979
980 def trouble(self, message=None, tb=None, is_error=True):
981 """Determine action to take when a download problem appears.
982
983 Depending on if the downloader has been configured to ignore
984 download errors or not, this method may throw an exception or
985 not when errors are found, after printing the message.
986
987 @param tb If given, is additional traceback information
988 @param is_error Whether to raise error according to ignorerrors
989 """
990 if message is not None:
991 self.to_stderr(message)
992 if self.params.get('verbose'):
993 if tb is None:
994 if sys.exc_info()[0]: # if .trouble has been called from an except block
995 tb = ''
996 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
997 tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
998 tb += encode_compat_str(traceback.format_exc())
999 else:
1000 tb_data = traceback.format_list(traceback.extract_stack())
1001 tb = ''.join(tb_data)
1002 if tb:
1003 self.to_stderr(tb)
1004 if not is_error:
1005 return
1006 if not self.params.get('ignoreerrors'):
1007 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
1008 exc_info = sys.exc_info()[1].exc_info
1009 else:
1010 exc_info = sys.exc_info()
1011 raise DownloadError(message, exc_info)
1012 self._download_retcode = 1
1013
1014 Styles = Namespace(
1015 HEADERS='yellow',
1016 EMPHASIS='light blue',
1017 FILENAME='green',
1018 ID='green',
1019 DELIM='blue',
1020 ERROR='red',
1021 BAD_FORMAT='light red',
1022 WARNING='yellow',
1023 SUPPRESS='light black',
1024 )
1025
1026 def _format_text(self, handle, allow_colors, text, f, fallback=None, *, test_encoding=False):
1027 text = str(text)
1028 if test_encoding:
1029 original_text = text
1030 # handle.encoding can be None. See https://github.com/yt-dlp/yt-dlp/issues/2711
1031 encoding = self.params.get('encoding') or getattr(handle, 'encoding', None) or 'ascii'
1032 text = text.encode(encoding, 'ignore').decode(encoding)
1033 if fallback is not None and text != original_text:
1034 text = fallback
1035 return format_text(text, f) if allow_colors is True else text if fallback is None else fallback
1036
1037 def _format_out(self, *args, **kwargs):
1038 return self._format_text(self._out_files.out, self._allow_colors.out, *args, **kwargs)
1039
1040 def _format_screen(self, *args, **kwargs):
1041 return self._format_text(self._out_files.screen, self._allow_colors.screen, *args, **kwargs)
1042
1043 def _format_err(self, *args, **kwargs):
1044 return self._format_text(self._out_files.error, self._allow_colors.error, *args, **kwargs)
1045
1046 def report_warning(self, message, only_once=False):
1047 """
1048 Print the message to stderr, it will be prefixed with 'WARNING:'
1049 If stderr is a tty file the 'WARNING:' will be colored
1050 """
1051 if self.params.get('logger') is not None:
1052 self.params['logger'].warning(message)
1053 else:
1054 if self.params.get('no_warnings'):
1055 return
1056 self.to_stderr(f'{self._format_err("WARNING:", self.Styles.WARNING)} {message}', only_once)
1057
1058 def deprecation_warning(self, message, *, stacklevel=0):
1059 deprecation_warning(
1060 message, stacklevel=stacklevel + 1, printer=self.report_error, is_error=False)
1061
1062 def deprecated_feature(self, message):
1063 if self.params.get('logger') is not None:
1064 self.params['logger'].warning(f'Deprecated Feature: {message}')
1065 self.to_stderr(f'{self._format_err("Deprecated Feature:", self.Styles.ERROR)} {message}', True)
1066
1067 def report_error(self, message, *args, **kwargs):
1068 """
1069 Do the same as trouble, but prefixes the message with 'ERROR:', colored
1070 in red if stderr is a tty file.
1071 """
1072 self.trouble(f'{self._format_err("ERROR:", self.Styles.ERROR)} {message}', *args, **kwargs)
1073
1074 def write_debug(self, message, only_once=False):
1075 """Log debug message or Print message to stderr"""
1076 if not self.params.get('verbose', False):
1077 return
1078 message = f'[debug] {message}'
1079 if self.params.get('logger'):
1080 self.params['logger'].debug(message)
1081 else:
1082 self.to_stderr(message, only_once)
1083
1084 def report_file_already_downloaded(self, file_name):
1085 """Report file has already been fully downloaded."""
1086 try:
1087 self.to_screen(f'[download] {file_name} has already been downloaded')
1088 except UnicodeEncodeError:
1089 self.to_screen('[download] The file has already been downloaded')
1090
1091 def report_file_delete(self, file_name):
1092 """Report that existing file will be deleted."""
1093 try:
1094 self.to_screen(f'Deleting existing file {file_name}')
1095 except UnicodeEncodeError:
1096 self.to_screen('Deleting existing file')
1097
1098 def raise_no_formats(self, info, forced=False, *, msg=None):
1099 has_drm = info.get('_has_drm')
1100 ignored, expected = self.params.get('ignore_no_formats_error'), bool(msg)
1101 msg = msg or has_drm and 'This video is DRM protected' or 'No video formats found!'
1102 if forced or not ignored:
1103 raise ExtractorError(msg, video_id=info['id'], ie=info['extractor'],
1104 expected=has_drm or ignored or expected)
1105 else:
1106 self.report_warning(msg)
1107
1108 def parse_outtmpl(self):
1109 self.deprecation_warning('"YoutubeDL.parse_outtmpl" is deprecated and may be removed in a future version')
1110 self._parse_outtmpl()
1111 return self.params['outtmpl']
1112
1113 def _parse_outtmpl(self):
1114 sanitize = IDENTITY
1115 if self.params.get('restrictfilenames'): # Remove spaces in the default template
1116 sanitize = lambda x: x.replace(' - ', ' ').replace(' ', '-')
1117
1118 outtmpl = self.params.setdefault('outtmpl', {})
1119 if not isinstance(outtmpl, dict):
1120 self.params['outtmpl'] = outtmpl = {'default': outtmpl}
1121 outtmpl.update({k: sanitize(v) for k, v in DEFAULT_OUTTMPL.items() if outtmpl.get(k) is None})
1122
1123 def get_output_path(self, dir_type='', filename=None):
1124 paths = self.params.get('paths', {})
1125 assert isinstance(paths, dict), '"paths" parameter must be a dictionary'
1126 path = os.path.join(
1127 expand_path(paths.get('home', '').strip()),
1128 expand_path(paths.get(dir_type, '').strip()) if dir_type else '',
1129 filename or '')
1130 return sanitize_path(path, force=self.params.get('windowsfilenames'))
1131
1132 @staticmethod
1133 def _outtmpl_expandpath(outtmpl):
1134 # expand_path translates '%%' into '%' and '$$' into '$'
1135 # correspondingly that is not what we want since we need to keep
1136 # '%%' intact for template dict substitution step. Working around
1137 # with boundary-alike separator hack.
1138 sep = ''.join(random.choices(string.ascii_letters, k=32))
1139 outtmpl = outtmpl.replace('%%', f'%{sep}%').replace('$$', f'${sep}$')
1140
1141 # outtmpl should be expand_path'ed before template dict substitution
1142 # because meta fields may contain env variables we don't want to
1143 # be expanded. E.g. for outtmpl "%(title)s.%(ext)s" and
1144 # title "Hello $PATH", we don't want `$PATH` to be expanded.
1145 return expand_path(outtmpl).replace(sep, '')
1146
1147 @staticmethod
1148 def escape_outtmpl(outtmpl):
1149 """ Escape any remaining strings like %s, %abc% etc. """
1150 return re.sub(
1151 STR_FORMAT_RE_TMPL.format('', '(?![%(\0])'),
1152 lambda mobj: ('' if mobj.group('has_key') else '%') + mobj.group(0),
1153 outtmpl)
1154
1155 @classmethod
1156 def validate_outtmpl(cls, outtmpl):
1157 """ @return None or Exception object """
1158 outtmpl = re.sub(
1159 STR_FORMAT_RE_TMPL.format('[^)]*', '[ljhqBUDS]'),
1160 lambda mobj: f'{mobj.group(0)[:-1]}s',
1161 cls._outtmpl_expandpath(outtmpl))
1162 try:
1163 cls.escape_outtmpl(outtmpl) % collections.defaultdict(int)
1164 return None
1165 except ValueError as err:
1166 return err
1167
1168 @staticmethod
1169 def _copy_infodict(info_dict):
1170 info_dict = dict(info_dict)
1171 info_dict.pop('__postprocessors', None)
1172 info_dict.pop('__pending_error', None)
1173 return info_dict
1174
1175 def prepare_outtmpl(self, outtmpl, info_dict, sanitize=False):
1176 """ Make the outtmpl and info_dict suitable for substitution: ydl.escape_outtmpl(outtmpl) % info_dict
1177 @param sanitize Whether to sanitize the output as a filename.
1178 For backward compatibility, a function can also be passed
1179 """
1180
1181 info_dict.setdefault('epoch', int(time.time())) # keep epoch consistent once set
1182
1183 info_dict = self._copy_infodict(info_dict)
1184 info_dict['duration_string'] = ( # %(duration>%H-%M-%S)s is wrong if duration > 24hrs
1185 formatSeconds(info_dict['duration'], '-' if sanitize else ':')
1186 if info_dict.get('duration', None) is not None
1187 else None)
1188 info_dict['autonumber'] = int(self.params.get('autonumber_start', 1) - 1 + self._num_downloads)
1189 info_dict['video_autonumber'] = self._num_videos
1190 if info_dict.get('resolution') is None:
1191 info_dict['resolution'] = self.format_resolution(info_dict, default=None)
1192
1193 # For fields playlist_index, playlist_autonumber and autonumber convert all occurrences
1194 # of %(field)s to %(field)0Nd for backward compatibility
1195 field_size_compat_map = {
1196 'playlist_index': number_of_digits(info_dict.get('__last_playlist_index') or 0),
1197 'playlist_autonumber': number_of_digits(info_dict.get('n_entries') or 0),
1198 'autonumber': self.params.get('autonumber_size') or 5,
1199 }
1200
1201 TMPL_DICT = {}
1202 EXTERNAL_FORMAT_RE = re.compile(STR_FORMAT_RE_TMPL.format('[^)]*', f'[{STR_FORMAT_TYPES}ljhqBUDS]'))
1203 MATH_FUNCTIONS = {
1204 '+': float.__add__,
1205 '-': float.__sub__,
1206 '*': float.__mul__,
1207 }
1208 # Field is of the form key1.key2...
1209 # where keys (except first) can be string, int, slice or "{field, ...}"
1210 FIELD_INNER_RE = r'(?:\w+|%(num)s|%(num)s?(?::%(num)s?){1,2})' % {'num': r'(?:-?\d+)'} # noqa: UP031
1211 FIELD_RE = r'\w*(?:\.(?:%(inner)s|{%(field)s(?:,%(field)s)*}))*' % { # noqa: UP031
1212 'inner': FIELD_INNER_RE,
1213 'field': rf'\w*(?:\.{FIELD_INNER_RE})*',
1214 }
1215 MATH_FIELD_RE = rf'(?:{FIELD_RE}|-?{NUMBER_RE})'
1216 MATH_OPERATORS_RE = r'(?:{})'.format('|'.join(map(re.escape, MATH_FUNCTIONS.keys())))
1217 INTERNAL_FORMAT_RE = re.compile(rf'''(?xs)
1218 (?P<negate>-)?
1219 (?P<fields>{FIELD_RE})
1220 (?P<maths>(?:{MATH_OPERATORS_RE}{MATH_FIELD_RE})*)
1221 (?:>(?P<strf_format>.+?))?
1222 (?P<remaining>
1223 (?P<alternate>(?<!\\),[^|&)]+)?
1224 (?:&(?P<replacement>.*?))?
1225 (?:\|(?P<default>.*?))?
1226 )$''')
1227
1228 def _from_user_input(field):
1229 if field == ':':
1230 return ...
1231 elif ':' in field:
1232 return slice(*map(int_or_none, field.split(':')))
1233 elif int_or_none(field) is not None:
1234 return int(field)
1235 return field
1236
1237 def _traverse_infodict(fields):
1238 fields = [f for x in re.split(r'\.({.+?})\.?', fields)
1239 for f in ([x] if x.startswith('{') else x.split('.'))]
1240 for i in (0, -1):
1241 if fields and not fields[i]:
1242 fields.pop(i)
1243
1244 for i, f in enumerate(fields):
1245 if not f.startswith('{'):
1246 fields[i] = _from_user_input(f)
1247 continue
1248 assert f.endswith('}'), f'No closing brace for {f} in {fields}'
1249 fields[i] = {k: list(map(_from_user_input, k.split('.'))) for k in f[1:-1].split(',')}
1250
1251 return traverse_obj(info_dict, fields, traverse_string=True)
1252
1253 def get_value(mdict):
1254 # Object traversal
1255 value = _traverse_infodict(mdict['fields'])
1256 # Negative
1257 if mdict['negate']:
1258 value = float_or_none(value)
1259 if value is not None:
1260 value *= -1
1261 # Do maths
1262 offset_key = mdict['maths']
1263 if offset_key:
1264 value = float_or_none(value)
1265 operator = None
1266 while offset_key:
1267 item = re.match(
1268 MATH_FIELD_RE if operator else MATH_OPERATORS_RE,
1269 offset_key).group(0)
1270 offset_key = offset_key[len(item):]
1271 if operator is None:
1272 operator = MATH_FUNCTIONS[item]
1273 continue
1274 item, multiplier = (item[1:], -1) if item[0] == '-' else (item, 1)
1275 offset = float_or_none(item)
1276 if offset is None:
1277 offset = float_or_none(_traverse_infodict(item))
1278 try:
1279 value = operator(value, multiplier * offset)
1280 except (TypeError, ZeroDivisionError):
1281 return None
1282 operator = None
1283 # Datetime formatting
1284 if mdict['strf_format']:
1285 value = strftime_or_none(value, mdict['strf_format'].replace('\\,', ','))
1286
1287 # XXX: Workaround for https://github.com/yt-dlp/yt-dlp/issues/4485
1288 if sanitize and value == '':
1289 value = None
1290 return value
1291
1292 na = self.params.get('outtmpl_na_placeholder', 'NA')
1293
1294 def filename_sanitizer(key, value, restricted=self.params.get('restrictfilenames')):
1295 return sanitize_filename(str(value), restricted=restricted, is_id=(
1296 bool(re.search(r'(^|[_.])id(\.|$)', key))
1297 if 'filename-sanitization' in self.params['compat_opts']
1298 else NO_DEFAULT))
1299
1300 sanitizer = sanitize if callable(sanitize) else filename_sanitizer
1301 sanitize = bool(sanitize)
1302
1303 def _dumpjson_default(obj):
1304 if isinstance(obj, (set, LazyList)):
1305 return list(obj)
1306 return repr(obj)
1307
1308 class _ReplacementFormatter(string.Formatter):
1309 def get_field(self, field_name, args, kwargs):
1310 if field_name.isdigit():
1311 return args[0], -1
1312 raise ValueError('Unsupported field')
1313
1314 replacement_formatter = _ReplacementFormatter()
1315
1316 def create_key(outer_mobj):
1317 if not outer_mobj.group('has_key'):
1318 return outer_mobj.group(0)
1319 key = outer_mobj.group('key')
1320 mobj = re.match(INTERNAL_FORMAT_RE, key)
1321 value, replacement, default, last_field = None, None, na, ''
1322 while mobj:
1323 mobj = mobj.groupdict()
1324 default = mobj['default'] if mobj['default'] is not None else default
1325 value = get_value(mobj)
1326 last_field, replacement = mobj['fields'], mobj['replacement']
1327 if value is None and mobj['alternate']:
1328 mobj = re.match(INTERNAL_FORMAT_RE, mobj['remaining'][1:])
1329 else:
1330 break
1331
1332 if None not in (value, replacement):
1333 try:
1334 value = replacement_formatter.format(replacement, value)
1335 except ValueError:
1336 value, default = None, na
1337
1338 fmt = outer_mobj.group('format')
1339 if fmt == 's' and last_field in field_size_compat_map and isinstance(value, int):
1340 fmt = f'0{field_size_compat_map[last_field]:d}d'
1341
1342 flags = outer_mobj.group('conversion') or ''
1343 str_fmt = f'{fmt[:-1]}s'
1344 if value is None:
1345 value, fmt = default, 's'
1346 elif fmt[-1] == 'l': # list
1347 delim = '\n' if '#' in flags else ', '
1348 value, fmt = delim.join(map(str, variadic(value, allowed_types=(str, bytes)))), str_fmt
1349 elif fmt[-1] == 'j': # json
1350 value, fmt = json.dumps(
1351 value, default=_dumpjson_default,
1352 indent=4 if '#' in flags else None, ensure_ascii='+' not in flags), str_fmt
1353 elif fmt[-1] == 'h': # html
1354 value, fmt = escapeHTML(str(value)), str_fmt
1355 elif fmt[-1] == 'q': # quoted
1356 value = map(str, variadic(value) if '#' in flags else [value])
1357 value, fmt = shell_quote(value, shell=True), str_fmt
1358 elif fmt[-1] == 'B': # bytes
1359 value = f'%{str_fmt}'.encode() % str(value).encode()
1360 value, fmt = value.decode('utf-8', 'ignore'), 's'
1361 elif fmt[-1] == 'U': # unicode normalized
1362 value, fmt = unicodedata.normalize(
1363 # "+" = compatibility equivalence, "#" = NFD
1364 'NF{}{}'.format('K' if '+' in flags else '', 'D' if '#' in flags else 'C'),
1365 value), str_fmt
1366 elif fmt[-1] == 'D': # decimal suffix
1367 num_fmt, fmt = fmt[:-1].replace('#', ''), 's'
1368 value = format_decimal_suffix(value, f'%{num_fmt}f%s' if num_fmt else '%d%s',
1369 factor=1024 if '#' in flags else 1000)
1370 elif fmt[-1] == 'S': # filename sanitization
1371 value, fmt = filename_sanitizer(last_field, value, restricted='#' in flags), str_fmt
1372 elif fmt[-1] == 'c':
1373 if value:
1374 value = str(value)[0]
1375 else:
1376 fmt = str_fmt
1377 elif fmt[-1] not in 'rsa': # numeric
1378 value = float_or_none(value)
1379 if value is None:
1380 value, fmt = default, 's'
1381
1382 if sanitize:
1383 # If value is an object, sanitize might convert it to a string
1384 # So we convert it to repr first
1385 if fmt[-1] == 'r':
1386 value, fmt = repr(value), str_fmt
1387 elif fmt[-1] == 'a':
1388 value, fmt = ascii(value), str_fmt
1389 if fmt[-1] in 'csra':
1390 value = sanitizer(last_field, value)
1391
1392 key = '{}\0{}'.format(key.replace('%', '%\0'), outer_mobj.group('format'))
1393 TMPL_DICT[key] = value
1394 return '{prefix}%({key}){fmt}'.format(key=key, fmt=fmt, prefix=outer_mobj.group('prefix'))
1395
1396 return EXTERNAL_FORMAT_RE.sub(create_key, outtmpl), TMPL_DICT
1397
1398 def evaluate_outtmpl(self, outtmpl, info_dict, *args, **kwargs):
1399 outtmpl, info_dict = self.prepare_outtmpl(outtmpl, info_dict, *args, **kwargs)
1400 return self.escape_outtmpl(outtmpl) % info_dict
1401
1402 def _prepare_filename(self, info_dict, *, outtmpl=None, tmpl_type=None):
1403 assert None in (outtmpl, tmpl_type), 'outtmpl and tmpl_type are mutually exclusive'
1404 if outtmpl is None:
1405 outtmpl = self.params['outtmpl'].get(tmpl_type or 'default', self.params['outtmpl']['default'])
1406 try:
1407 outtmpl = self._outtmpl_expandpath(outtmpl)
1408 filename = self.evaluate_outtmpl(outtmpl, info_dict, True)
1409 if not filename:
1410 return None
1411
1412 if tmpl_type in ('', 'temp'):
1413 final_ext, ext = self.params.get('final_ext'), info_dict.get('ext')
1414 if final_ext and ext and final_ext != ext and filename.endswith(f'.{final_ext}'):
1415 filename = replace_extension(filename, ext, final_ext)
1416 elif tmpl_type:
1417 force_ext = OUTTMPL_TYPES[tmpl_type]
1418 if force_ext:
1419 filename = replace_extension(filename, force_ext, info_dict.get('ext'))
1420
1421 # https://github.com/blackjack4494/youtube-dlc/issues/85
1422 trim_file_name = self.params.get('trim_file_name', False)
1423 if trim_file_name:
1424 no_ext, *ext = filename.rsplit('.', 2)
1425 filename = join_nonempty(no_ext[:trim_file_name], *ext, delim='.')
1426
1427 return filename
1428 except ValueError as err:
1429 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
1430 return None
1431
1432 def prepare_filename(self, info_dict, dir_type='', *, outtmpl=None, warn=False):
1433 """Generate the output filename"""
1434 if outtmpl:
1435 assert not dir_type, 'outtmpl and dir_type are mutually exclusive'
1436 dir_type = None
1437 filename = self._prepare_filename(info_dict, tmpl_type=dir_type, outtmpl=outtmpl)
1438 if not filename and dir_type not in ('', 'temp'):
1439 return ''
1440
1441 if warn:
1442 if not self.params.get('paths'):
1443 pass
1444 elif filename == '-':
1445 self.report_warning('--paths is ignored when an outputting to stdout', only_once=True)
1446 elif os.path.isabs(filename):
1447 self.report_warning('--paths is ignored since an absolute path is given in output template', only_once=True)
1448 if filename == '-' or not filename:
1449 return filename
1450
1451 return self.get_output_path(dir_type, filename)
1452
1453 def _match_entry(self, info_dict, incomplete=False, silent=False):
1454 """Returns None if the file should be downloaded"""
1455 _type = 'video' if 'playlist-match-filter' in self.params['compat_opts'] else info_dict.get('_type', 'video')
1456 assert incomplete or _type == 'video', 'Only video result can be considered complete'
1457
1458 video_title = info_dict.get('title', info_dict.get('id', 'entry'))
1459
1460 def check_filter():
1461 if _type in ('playlist', 'multi_video'):
1462 return
1463 elif _type in ('url', 'url_transparent') and not try_call(
1464 lambda: self.get_info_extractor(info_dict['ie_key']).is_single_video(info_dict['url'])):
1465 return
1466
1467 if 'title' in info_dict:
1468 # This can happen when we're just evaluating the playlist
1469 title = info_dict['title']
1470 matchtitle = self.params.get('matchtitle', False)
1471 if matchtitle:
1472 if not re.search(matchtitle, title, re.IGNORECASE):
1473 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
1474 rejecttitle = self.params.get('rejecttitle', False)
1475 if rejecttitle:
1476 if re.search(rejecttitle, title, re.IGNORECASE):
1477 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
1478
1479 date = info_dict.get('upload_date')
1480 if date is not None:
1481 date_range = self.params.get('daterange', DateRange())
1482 if date not in date_range:
1483 return f'{date_from_str(date).isoformat()} upload date is not in range {date_range}'
1484 view_count = info_dict.get('view_count')
1485 if view_count is not None:
1486 min_views = self.params.get('min_views')
1487 if min_views is not None and view_count < min_views:
1488 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
1489 max_views = self.params.get('max_views')
1490 if max_views is not None and view_count > max_views:
1491 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
1492 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
1493 return f'Skipping "{video_title}" because it is age restricted'
1494
1495 match_filter = self.params.get('match_filter')
1496 if match_filter is None:
1497 return None
1498
1499 cancelled = None
1500 try:
1501 try:
1502 ret = match_filter(info_dict, incomplete=incomplete)
1503 except TypeError:
1504 # For backward compatibility
1505 ret = None if incomplete else match_filter(info_dict)
1506 except DownloadCancelled as err:
1507 if err.msg is not NO_DEFAULT:
1508 raise
1509 ret, cancelled = err.msg, err
1510
1511 if ret is NO_DEFAULT:
1512 while True:
1513 filename = self._format_screen(self.prepare_filename(info_dict), self.Styles.FILENAME)
1514 reply = input(self._format_screen(
1515 f'Download "{filename}"? (Y/n): ', self.Styles.EMPHASIS)).lower().strip()
1516 if reply in {'y', ''}:
1517 return None
1518 elif reply == 'n':
1519 if cancelled:
1520 raise type(cancelled)(f'Skipping {video_title}')
1521 return f'Skipping {video_title}'
1522 return ret
1523
1524 if self.in_download_archive(info_dict):
1525 reason = ''.join((
1526 format_field(info_dict, 'id', f'{self._format_screen("%s", self.Styles.ID)}: '),
1527 format_field(info_dict, 'title', f'{self._format_screen("%s", self.Styles.EMPHASIS)} '),
1528 'has already been recorded in the archive'))
1529 break_opt, break_err = 'break_on_existing', ExistingVideoReached
1530 else:
1531 try:
1532 reason = check_filter()
1533 except DownloadCancelled as e:
1534 reason, break_opt, break_err = e.msg, 'match_filter', type(e)
1535 else:
1536 break_opt, break_err = 'break_on_reject', RejectedVideoReached
1537 if reason is not None:
1538 if not silent:
1539 self.to_screen('[download] ' + reason)
1540 if self.params.get(break_opt, False):
1541 raise break_err()
1542 return reason
1543
1544 @staticmethod
1545 def add_extra_info(info_dict, extra_info):
1546 """Set the keys from extra_info in info dict if they are missing"""
1547 for key, value in extra_info.items():
1548 info_dict.setdefault(key, value)
1549
1550 def extract_info(self, url, download=True, ie_key=None, extra_info=None,
1551 process=True, force_generic_extractor=False):
1552 """
1553 Extract and return the information dictionary of the URL
1554
1555 Arguments:
1556 @param url URL to extract
1557
1558 Keyword arguments:
1559 @param download Whether to download videos
1560 @param process Whether to resolve all unresolved references (URLs, playlist items).
1561 Must be True for download to work
1562 @param ie_key Use only the extractor with this key
1563
1564 @param extra_info Dictionary containing the extra values to add to the info (For internal use only)
1565 @force_generic_extractor Force using the generic extractor (Deprecated; use ie_key='Generic')
1566 """
1567
1568 if extra_info is None:
1569 extra_info = {}
1570
1571 if not ie_key and force_generic_extractor:
1572 ie_key = 'Generic'
1573
1574 if ie_key:
1575 ies = {ie_key: self._ies[ie_key]} if ie_key in self._ies else {}
1576 else:
1577 ies = self._ies
1578
1579 for key, ie in ies.items():
1580 if not ie.suitable(url):
1581 continue
1582
1583 if not ie.working():
1584 self.report_warning('The program functionality for this site has been marked as broken, '
1585 'and will probably not work.')
1586
1587 temp_id = ie.get_temp_id(url)
1588 if temp_id is not None and self.in_download_archive({'id': temp_id, 'ie_key': key}):
1589 self.to_screen(f'[download] {self._format_screen(temp_id, self.Styles.ID)}: '
1590 'has already been recorded in the archive')
1591 if self.params.get('break_on_existing', False):
1592 raise ExistingVideoReached
1593 break
1594 return self.__extract_info(url, self.get_info_extractor(key), download, extra_info, process)
1595 else:
1596 extractors_restricted = self.params.get('allowed_extractors') not in (None, ['default'])
1597 self.report_error(f'No suitable extractor{format_field(ie_key, None, " (%s)")} found for URL {url}',
1598 tb=False if extractors_restricted else None)
1599
1600 def _handle_extraction_exceptions(func):
1601 @functools.wraps(func)
1602 def wrapper(self, *args, **kwargs):
1603 while True:
1604 try:
1605 return func(self, *args, **kwargs)
1606 except (DownloadCancelled, LazyList.IndexError, PagedList.IndexError):
1607 raise
1608 except ReExtractInfo as e:
1609 if e.expected:
1610 self.to_screen(f'{e}; Re-extracting data')
1611 else:
1612 self.to_stderr('\r')
1613 self.report_warning(f'{e}; Re-extracting data')
1614 continue
1615 except GeoRestrictedError as e:
1616 msg = e.msg
1617 if e.countries:
1618 msg += '\nThis video is available in {}.'.format(', '.join(
1619 map(ISO3166Utils.short2full, e.countries)))
1620 msg += '\nYou might want to use a VPN or a proxy server (with --proxy) to workaround.'
1621 self.report_error(msg)
1622 except ExtractorError as e: # An error we somewhat expected
1623 self.report_error(str(e), e.format_traceback())
1624 except Exception as e:
1625 if self.params.get('ignoreerrors'):
1626 self.report_error(str(e), tb=encode_compat_str(traceback.format_exc()))
1627 else:
1628 raise
1629 break
1630 return wrapper
1631
1632 def _wait_for_video(self, ie_result={}):
1633 if (not self.params.get('wait_for_video')
1634 or ie_result.get('_type', 'video') != 'video'
1635 or ie_result.get('formats') or ie_result.get('url')):
1636 return
1637
1638 format_dur = lambda dur: '%02d:%02d:%02d' % timetuple_from_msec(dur * 1000)[:-1]
1639 last_msg = ''
1640
1641 def progress(msg):
1642 nonlocal last_msg
1643 full_msg = f'{msg}\n'
1644 if not self.params.get('noprogress'):
1645 full_msg = msg + ' ' * (len(last_msg) - len(msg)) + '\r'
1646 elif last_msg:
1647 return
1648 self.to_screen(full_msg, skip_eol=True)
1649 last_msg = msg
1650
1651 min_wait, max_wait = self.params.get('wait_for_video')
1652 diff = try_get(ie_result, lambda x: x['release_timestamp'] - time.time())
1653 if diff is None and ie_result.get('live_status') == 'is_upcoming':
1654 diff = round(random.uniform(min_wait, max_wait) if (max_wait and min_wait) else (max_wait or min_wait), 0)
1655 self.report_warning('Release time of video is not known')
1656 elif ie_result and (diff or 0) <= 0:
1657 self.report_warning('Video should already be available according to extracted info')
1658 diff = min(max(diff or 0, min_wait or 0), max_wait or float('inf'))
1659 self.to_screen(f'[wait] Waiting for {format_dur(diff)} - Press Ctrl+C to try now')
1660
1661 wait_till = time.time() + diff
1662 try:
1663 while True:
1664 diff = wait_till - time.time()
1665 if diff <= 0:
1666 progress('')
1667 raise ReExtractInfo('[wait] Wait period ended', expected=True)
1668 progress(f'[wait] Remaining time until next attempt: {self._format_screen(format_dur(diff), self.Styles.EMPHASIS)}')
1669 time.sleep(1)
1670 except KeyboardInterrupt:
1671 progress('')
1672 raise ReExtractInfo('[wait] Interrupted by user', expected=True)
1673 except BaseException as e:
1674 if not isinstance(e, ReExtractInfo):
1675 self.to_screen('')
1676 raise
1677
1678 def _load_cookies(self, data, *, autoscope=True):
1679 """Loads cookies from a `Cookie` header
1680
1681 This tries to work around the security vulnerability of passing cookies to every domain.
1682 See: https://github.com/yt-dlp/yt-dlp/security/advisories/GHSA-v8mc-9377-rwjj
1683
1684 @param data The Cookie header as string to load the cookies from
1685 @param autoscope If `False`, scope cookies using Set-Cookie syntax and error for cookie without domains
1686 If `True`, save cookies for later to be stored in the jar with a limited scope
1687 If a URL, save cookies in the jar with the domain of the URL
1688 """
1689 for cookie in LenientSimpleCookie(data).values():
1690 if autoscope and any(cookie.values()):
1691 raise ValueError('Invalid syntax in Cookie Header')
1692
1693 domain = cookie.get('domain') or ''
1694 expiry = cookie.get('expires')
1695 if expiry == '': # 0 is valid
1696 expiry = None
1697 prepared_cookie = http.cookiejar.Cookie(
1698 cookie.get('version') or 0, cookie.key, cookie.value, None, False,
1699 domain, True, True, cookie.get('path') or '', bool(cookie.get('path')),
1700 cookie.get('secure') or False, expiry, False, None, None, {})
1701
1702 if domain:
1703 self.cookiejar.set_cookie(prepared_cookie)
1704 elif autoscope is True:
1705 self.deprecated_feature(
1706 'Passing cookies as a header is a potential security risk; '
1707 'they will be scoped to the domain of the downloaded urls. '
1708 'Please consider loading cookies from a file or browser instead.')
1709 self.__header_cookies.append(prepared_cookie)
1710 elif autoscope:
1711 self.report_warning(
1712 'The extractor result contains an unscoped cookie as an HTTP header. '
1713 f'If you are using yt-dlp with an input URL{bug_reports_message(before=",")}',
1714 only_once=True)
1715 self._apply_header_cookies(autoscope, [prepared_cookie])
1716 else:
1717 self.report_error('Unscoped cookies are not allowed; please specify some sort of scoping',
1718 tb=False, is_error=False)
1719
1720 def _apply_header_cookies(self, url, cookies=None):
1721 """Applies stray header cookies to the provided url
1722
1723 This loads header cookies and scopes them to the domain provided in `url`.
1724 While this is not ideal, it helps reduce the risk of them being sent
1725 to an unintended destination while mostly maintaining compatibility.
1726 """
1727 parsed = urllib.parse.urlparse(url)
1728 if not parsed.hostname:
1729 return
1730
1731 for cookie in map(copy.copy, cookies or self.__header_cookies):
1732 cookie.domain = f'.{parsed.hostname}'
1733 self.cookiejar.set_cookie(cookie)
1734
1735 @_handle_extraction_exceptions
1736 def __extract_info(self, url, ie, download, extra_info, process):
1737 self._apply_header_cookies(url)
1738
1739 try:
1740 ie_result = ie.extract(url)
1741 except UserNotLive as e:
1742 if process:
1743 if self.params.get('wait_for_video'):
1744 self.report_warning(e)
1745 self._wait_for_video()
1746 raise
1747 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
1748 self.report_warning(f'Extractor {ie.IE_NAME} returned nothing{bug_reports_message()}')
1749 return
1750 if isinstance(ie_result, list):
1751 # Backwards compatibility: old IE result format
1752 ie_result = {
1753 '_type': 'compat_list',
1754 'entries': ie_result,
1755 }
1756 if extra_info.get('original_url'):
1757 ie_result.setdefault('original_url', extra_info['original_url'])
1758 self.add_default_extra_info(ie_result, ie, url)
1759 if process:
1760 self._wait_for_video(ie_result)
1761 return self.process_ie_result(ie_result, download, extra_info)
1762 else:
1763 return ie_result
1764
1765 def add_default_extra_info(self, ie_result, ie, url):
1766 if url is not None:
1767 self.add_extra_info(ie_result, {
1768 'webpage_url': url,
1769 'original_url': url,
1770 })
1771 webpage_url = ie_result.get('webpage_url')
1772 if webpage_url:
1773 self.add_extra_info(ie_result, {
1774 'webpage_url_basename': url_basename(webpage_url),
1775 'webpage_url_domain': get_domain(webpage_url),
1776 })
1777 if ie is not None:
1778 self.add_extra_info(ie_result, {
1779 'extractor': ie.IE_NAME,
1780 'extractor_key': ie.ie_key(),
1781 })
1782
1783 def process_ie_result(self, ie_result, download=True, extra_info=None):
1784 """
1785 Take the result of the ie(may be modified) and resolve all unresolved
1786 references (URLs, playlist items).
1787
1788 It will also download the videos if 'download'.
1789 Returns the resolved ie_result.
1790 """
1791 if extra_info is None:
1792 extra_info = {}
1793 result_type = ie_result.get('_type', 'video')
1794
1795 if result_type in ('url', 'url_transparent'):
1796 ie_result['url'] = sanitize_url(
1797 ie_result['url'], scheme='http' if self.params.get('prefer_insecure') else 'https')
1798 if ie_result.get('original_url') and not extra_info.get('original_url'):
1799 extra_info = {'original_url': ie_result['original_url'], **extra_info}
1800
1801 extract_flat = self.params.get('extract_flat', False)
1802 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info)
1803 or extract_flat is True):
1804 info_copy = ie_result.copy()
1805 ie = try_get(ie_result.get('ie_key'), self.get_info_extractor)
1806 if ie and not ie_result.get('id'):
1807 info_copy['id'] = ie.get_temp_id(ie_result['url'])
1808 self.add_default_extra_info(info_copy, ie, ie_result['url'])
1809 self.add_extra_info(info_copy, extra_info)
1810 info_copy, _ = self.pre_process(info_copy)
1811 self._fill_common_fields(info_copy, False)
1812 self.__forced_printings(info_copy)
1813 self._raise_pending_errors(info_copy)
1814 if self.params.get('force_write_download_archive', False):
1815 self.record_download_archive(info_copy)
1816 return ie_result
1817
1818 if result_type == 'video':
1819 self.add_extra_info(ie_result, extra_info)
1820 ie_result = self.process_video_result(ie_result, download=download)
1821 self._raise_pending_errors(ie_result)
1822 additional_urls = (ie_result or {}).get('additional_urls')
1823 if additional_urls:
1824 # TODO: Improve MetadataParserPP to allow setting a list
1825 if isinstance(additional_urls, str):
1826 additional_urls = [additional_urls]
1827 self.to_screen(
1828 '[info] {}: {} additional URL(s) requested'.format(ie_result['id'], len(additional_urls)))
1829 self.write_debug('Additional URLs: "{}"'.format('", "'.join(additional_urls)))
1830 ie_result['additional_entries'] = [
1831 self.extract_info(
1832 url, download, extra_info=extra_info,
1833 force_generic_extractor=self.params.get('force_generic_extractor'))
1834 for url in additional_urls
1835 ]
1836 return ie_result
1837 elif result_type == 'url':
1838 # We have to add extra_info to the results because it may be
1839 # contained in a playlist
1840 return self.extract_info(
1841 ie_result['url'], download,
1842 ie_key=ie_result.get('ie_key'),
1843 extra_info=extra_info)
1844 elif result_type == 'url_transparent':
1845 # Use the information from the embedding page
1846 info = self.extract_info(
1847 ie_result['url'], ie_key=ie_result.get('ie_key'),
1848 extra_info=extra_info, download=False, process=False)
1849
1850 # extract_info may return None when ignoreerrors is enabled and
1851 # extraction failed with an error, don't crash and return early
1852 # in this case
1853 if not info:
1854 return info
1855
1856 exempted_fields = {'_type', 'url', 'ie_key'}
1857 if not ie_result.get('section_end') and ie_result.get('section_start') is None:
1858 # For video clips, the id etc of the clip extractor should be used
1859 exempted_fields |= {'id', 'extractor', 'extractor_key'}
1860
1861 new_result = info.copy()
1862 new_result.update(filter_dict(ie_result, lambda k, v: v is not None and k not in exempted_fields))
1863
1864 # Extracted info may not be a video result (i.e.
1865 # info.get('_type', 'video') != video) but rather an url or
1866 # url_transparent. In such cases outer metadata (from ie_result)
1867 # should be propagated to inner one (info). For this to happen
1868 # _type of info should be overridden with url_transparent. This
1869 # fixes issue from https://github.com/ytdl-org/youtube-dl/pull/11163.
1870 if new_result.get('_type') == 'url':
1871 new_result['_type'] = 'url_transparent'
1872
1873 return self.process_ie_result(
1874 new_result, download=download, extra_info=extra_info)
1875 elif result_type in ('playlist', 'multi_video'):
1876 # Protect from infinite recursion due to recursively nested playlists
1877 # (see https://github.com/ytdl-org/youtube-dl/issues/27833)
1878 webpage_url = ie_result.get('webpage_url') # Playlists maynot have webpage_url
1879 if webpage_url and webpage_url in self._playlist_urls:
1880 self.to_screen(
1881 '[download] Skipping already downloaded playlist: {}'.format(
1882 ie_result.get('title')) or ie_result.get('id'))
1883 return
1884
1885 self._playlist_level += 1
1886 self._playlist_urls.add(webpage_url)
1887 self._fill_common_fields(ie_result, False)
1888 self._sanitize_thumbnails(ie_result)
1889 try:
1890 return self.__process_playlist(ie_result, download)
1891 finally:
1892 self._playlist_level -= 1
1893 if not self._playlist_level:
1894 self._playlist_urls.clear()
1895 elif result_type == 'compat_list':
1896 self.report_warning(
1897 'Extractor {} returned a compat_list result. '
1898 'It needs to be updated.'.format(ie_result.get('extractor')))
1899
1900 def _fixup(r):
1901 self.add_extra_info(r, {
1902 'extractor': ie_result['extractor'],
1903 'webpage_url': ie_result['webpage_url'],
1904 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1905 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1906 'extractor_key': ie_result['extractor_key'],
1907 })
1908 return r
1909 ie_result['entries'] = [
1910 self.process_ie_result(_fixup(r), download, extra_info)
1911 for r in ie_result['entries']
1912 ]
1913 return ie_result
1914 else:
1915 raise Exception(f'Invalid result type: {result_type}')
1916
1917 def _ensure_dir_exists(self, path):
1918 return make_dir(path, self.report_error)
1919
1920 @staticmethod
1921 def _playlist_infodict(ie_result, strict=False, **kwargs):
1922 info = {
1923 'playlist_count': ie_result.get('playlist_count'),
1924 'playlist': ie_result.get('title') or ie_result.get('id'),
1925 'playlist_id': ie_result.get('id'),
1926 'playlist_title': ie_result.get('title'),
1927 'playlist_uploader': ie_result.get('uploader'),
1928 'playlist_uploader_id': ie_result.get('uploader_id'),
1929 **kwargs,
1930 }
1931 if strict:
1932 return info
1933 if ie_result.get('webpage_url'):
1934 info.update({
1935 'webpage_url': ie_result['webpage_url'],
1936 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1937 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1938 })
1939 return {
1940 **info,
1941 'playlist_index': 0,
1942 '__last_playlist_index': max(ie_result.get('requested_entries') or (0, 0)),
1943 'extractor': ie_result['extractor'],
1944 'extractor_key': ie_result['extractor_key'],
1945 }
1946
1947 def __process_playlist(self, ie_result, download):
1948 """Process each entry in the playlist"""
1949 assert ie_result['_type'] in ('playlist', 'multi_video')
1950
1951 common_info = self._playlist_infodict(ie_result, strict=True)
1952 title = common_info.get('playlist') or '<Untitled>'
1953 if self._match_entry(common_info, incomplete=True) is not None:
1954 return
1955 self.to_screen(f'[download] Downloading {ie_result["_type"]}: {title}')
1956
1957 all_entries = PlaylistEntries(self, ie_result)
1958 entries = orderedSet(all_entries.get_requested_items(), lazy=True)
1959
1960 lazy = self.params.get('lazy_playlist')
1961 if lazy:
1962 resolved_entries, n_entries = [], 'N/A'
1963 ie_result['requested_entries'], ie_result['entries'] = None, None
1964 else:
1965 entries = resolved_entries = list(entries)
1966 n_entries = len(resolved_entries)
1967 ie_result['requested_entries'], ie_result['entries'] = tuple(zip(*resolved_entries)) or ([], [])
1968 if not ie_result.get('playlist_count'):
1969 # Better to do this after potentially exhausting entries
1970 ie_result['playlist_count'] = all_entries.get_full_count()
1971
1972 extra = self._playlist_infodict(ie_result, n_entries=int_or_none(n_entries))
1973 ie_copy = collections.ChainMap(ie_result, extra)
1974
1975 _infojson_written = False
1976 write_playlist_files = self.params.get('allow_playlist_files', True)
1977 if write_playlist_files and self.params.get('list_thumbnails'):
1978 self.list_thumbnails(ie_result)
1979 if write_playlist_files and not self.params.get('simulate'):
1980 _infojson_written = self._write_info_json(
1981 'playlist', ie_result, self.prepare_filename(ie_copy, 'pl_infojson'))
1982 if _infojson_written is None:
1983 return
1984 if self._write_description('playlist', ie_result,
1985 self.prepare_filename(ie_copy, 'pl_description')) is None:
1986 return
1987 # TODO: This should be passed to ThumbnailsConvertor if necessary
1988 self._write_thumbnails('playlist', ie_result, self.prepare_filename(ie_copy, 'pl_thumbnail'))
1989
1990 if lazy:
1991 if self.params.get('playlistreverse') or self.params.get('playlistrandom'):
1992 self.report_warning('playlistreverse and playlistrandom are not supported with lazy_playlist', only_once=True)
1993 elif self.params.get('playlistreverse'):
1994 entries.reverse()
1995 elif self.params.get('playlistrandom'):
1996 random.shuffle(entries)
1997
1998 self.to_screen(f'[{ie_result["extractor"]}] Playlist {title}: Downloading {n_entries} items'
1999 f'{format_field(ie_result, "playlist_count", " of %s")}')
2000
2001 keep_resolved_entries = self.params.get('extract_flat') != 'discard'
2002 if self.params.get('extract_flat') == 'discard_in_playlist':
2003 keep_resolved_entries = ie_result['_type'] != 'playlist'
2004 if keep_resolved_entries:
2005 self.write_debug('The information of all playlist entries will be held in memory')
2006
2007 failures = 0
2008 max_failures = self.params.get('skip_playlist_after_errors') or float('inf')
2009 for i, (playlist_index, entry) in enumerate(entries):
2010 if lazy:
2011 resolved_entries.append((playlist_index, entry))
2012 if not entry:
2013 continue
2014
2015 entry['__x_forwarded_for_ip'] = ie_result.get('__x_forwarded_for_ip')
2016 if not lazy and 'playlist-index' in self.params['compat_opts']:
2017 playlist_index = ie_result['requested_entries'][i]
2018
2019 entry_copy = collections.ChainMap(entry, {
2020 **common_info,
2021 'n_entries': int_or_none(n_entries),
2022 'playlist_index': playlist_index,
2023 'playlist_autonumber': i + 1,
2024 })
2025
2026 if self._match_entry(entry_copy, incomplete=True) is not None:
2027 # For compatabilty with youtube-dl. See https://github.com/yt-dlp/yt-dlp/issues/4369
2028 resolved_entries[i] = (playlist_index, NO_DEFAULT)
2029 continue
2030
2031 self.to_screen(
2032 f'[download] Downloading item {self._format_screen(i + 1, self.Styles.ID)} '
2033 f'of {self._format_screen(n_entries, self.Styles.EMPHASIS)}')
2034
2035 entry_result = self.__process_iterable_entry(entry, download, collections.ChainMap({
2036 'playlist_index': playlist_index,
2037 'playlist_autonumber': i + 1,
2038 }, extra))
2039 if not entry_result:
2040 failures += 1
2041 if failures >= max_failures:
2042 self.report_error(
2043 f'Skipping the remaining entries in playlist "{title}" since {failures} items failed extraction')
2044 break
2045 if keep_resolved_entries:
2046 resolved_entries[i] = (playlist_index, entry_result)
2047
2048 # Update with processed data
2049 ie_result['entries'] = [e for _, e in resolved_entries if e is not NO_DEFAULT]
2050 ie_result['requested_entries'] = [i for i, e in resolved_entries if e is not NO_DEFAULT]
2051 if ie_result['requested_entries'] == try_call(lambda: list(range(1, ie_result['playlist_count'] + 1))):
2052 # Do not set for full playlist
2053 ie_result.pop('requested_entries')
2054
2055 # Write the updated info to json
2056 if _infojson_written is True and self._write_info_json(
2057 'updated playlist', ie_result,
2058 self.prepare_filename(ie_copy, 'pl_infojson'), overwrite=True) is None:
2059 return
2060
2061 ie_result = self.run_all_pps('playlist', ie_result)
2062 self.to_screen(f'[download] Finished downloading playlist: {title}')
2063 return ie_result
2064
2065 @_handle_extraction_exceptions
2066 def __process_iterable_entry(self, entry, download, extra_info):
2067 return self.process_ie_result(
2068 entry, download=download, extra_info=extra_info)
2069
2070 def _build_format_filter(self, filter_spec):
2071 " Returns a function to filter the formats according to the filter_spec "
2072
2073 OPERATORS = {
2074 '<': operator.lt,
2075 '<=': operator.le,
2076 '>': operator.gt,
2077 '>=': operator.ge,
2078 '=': operator.eq,
2079 '!=': operator.ne,
2080 }
2081 operator_rex = re.compile(r'''(?x)\s*
2082 (?P<key>[\w.-]+)\s*
2083 (?P<op>{})(?P<none_inclusive>\s*\?)?\s*
2084 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)\s*
2085 '''.format('|'.join(map(re.escape, OPERATORS.keys()))))
2086 m = operator_rex.fullmatch(filter_spec)
2087 if m:
2088 try:
2089 comparison_value = int(m.group('value'))
2090 except ValueError:
2091 comparison_value = parse_filesize(m.group('value'))
2092 if comparison_value is None:
2093 comparison_value = parse_filesize(m.group('value') + 'B')
2094 if comparison_value is None:
2095 raise ValueError(
2096 'Invalid value {!r} in format specification {!r}'.format(
2097 m.group('value'), filter_spec))
2098 op = OPERATORS[m.group('op')]
2099
2100 if not m:
2101 STR_OPERATORS = {
2102 '=': operator.eq,
2103 '^=': lambda attr, value: attr.startswith(value),
2104 '$=': lambda attr, value: attr.endswith(value),
2105 '*=': lambda attr, value: value in attr,
2106 '~=': lambda attr, value: value.search(attr) is not None,
2107 }
2108 str_operator_rex = re.compile(r'''(?x)\s*
2109 (?P<key>[a-zA-Z0-9._-]+)\s*
2110 (?P<negation>!\s*)?(?P<op>{})\s*(?P<none_inclusive>\?\s*)?
2111 (?P<quote>["'])?
2112 (?P<value>(?(quote)(?:(?!(?P=quote))[^\\]|\\.)+|[\w.-]+))
2113 (?(quote)(?P=quote))\s*
2114 '''.format('|'.join(map(re.escape, STR_OPERATORS.keys()))))
2115 m = str_operator_rex.fullmatch(filter_spec)
2116 if m:
2117 if m.group('op') == '~=':
2118 comparison_value = re.compile(m.group('value'))
2119 else:
2120 comparison_value = re.sub(r'''\\([\\"'])''', r'\1', m.group('value'))
2121 str_op = STR_OPERATORS[m.group('op')]
2122 if m.group('negation'):
2123 op = lambda attr, value: not str_op(attr, value)
2124 else:
2125 op = str_op
2126
2127 if not m:
2128 raise SyntaxError(f'Invalid filter specification {filter_spec!r}')
2129
2130 def _filter(f):
2131 actual_value = f.get(m.group('key'))
2132 if actual_value is None:
2133 return m.group('none_inclusive')
2134 return op(actual_value, comparison_value)
2135 return _filter
2136
2137 def _check_formats(self, formats):
2138 for f in formats:
2139 working = f.get('__working')
2140 if working is not None:
2141 if working:
2142 yield f
2143 continue
2144 self.to_screen('[info] Testing format {}'.format(f['format_id']))
2145 path = self.get_output_path('temp')
2146 if not self._ensure_dir_exists(f'{path}/'):
2147 continue
2148 temp_file = tempfile.NamedTemporaryFile(suffix='.tmp', delete=False, dir=path or None)
2149 temp_file.close()
2150 try:
2151 success, _ = self.dl(temp_file.name, f, test=True)
2152 except (DownloadError, OSError, ValueError, *network_exceptions):
2153 success = False
2154 finally:
2155 if os.path.exists(temp_file.name):
2156 try:
2157 os.remove(temp_file.name)
2158 except OSError:
2159 self.report_warning(f'Unable to delete temporary file "{temp_file.name}"')
2160 f['__working'] = success
2161 if success:
2162 yield f
2163 else:
2164 self.to_screen('[info] Unable to download format {}. Skipping...'.format(f['format_id']))
2165
2166 def _select_formats(self, formats, selector):
2167 return list(selector({
2168 'formats': formats,
2169 'has_merged_format': any('none' not in (f.get('acodec'), f.get('vcodec')) for f in formats),
2170 'incomplete_formats': (all(f.get('vcodec') == 'none' for f in formats) # No formats with video
2171 or all(f.get('acodec') == 'none' for f in formats)), # OR, No formats with audio
2172 }))
2173
2174 def _default_format_spec(self, info_dict, download=True):
2175 download = download and not self.params.get('simulate')
2176 prefer_best = download and (
2177 self.params['outtmpl']['default'] == '-'
2178 or info_dict.get('is_live') and not self.params.get('live_from_start'))
2179
2180 def can_merge():
2181 merger = FFmpegMergerPP(self)
2182 return merger.available and merger.can_merge()
2183
2184 if not prefer_best and download and not can_merge():
2185 prefer_best = True
2186 formats = self._get_formats(info_dict)
2187 evaluate_formats = lambda spec: self._select_formats(formats, self.build_format_selector(spec))
2188 if evaluate_formats('b/bv+ba') != evaluate_formats('bv*+ba/b'):
2189 self.report_warning('ffmpeg not found. The downloaded format may not be the best available. '
2190 'Installing ffmpeg is strongly recommended: https://github.com/yt-dlp/yt-dlp#dependencies')
2191
2192 compat = (self.params.get('allow_multiple_audio_streams')
2193 or 'format-spec' in self.params['compat_opts'])
2194
2195 return ('best/bestvideo+bestaudio' if prefer_best
2196 else 'bestvideo+bestaudio/best' if compat
2197 else 'bestvideo*+bestaudio/best')
2198
2199 def build_format_selector(self, format_spec):
2200 def syntax_error(note, start):
2201 message = (
2202 'Invalid format specification: '
2203 '{}\n\t{}\n\t{}^'.format(note, format_spec, ' ' * start[1]))
2204 return SyntaxError(message)
2205
2206 PICKFIRST = 'PICKFIRST'
2207 MERGE = 'MERGE'
2208 SINGLE = 'SINGLE'
2209 GROUP = 'GROUP'
2210 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
2211
2212 allow_multiple_streams = {'audio': self.params.get('allow_multiple_audio_streams', False),
2213 'video': self.params.get('allow_multiple_video_streams', False)}
2214
2215 def _parse_filter(tokens):
2216 filter_parts = []
2217 for type_, string_, _start, _, _ in tokens:
2218 if type_ == tokenize.OP and string_ == ']':
2219 return ''.join(filter_parts)
2220 else:
2221 filter_parts.append(string_)
2222
2223 def _remove_unused_ops(tokens):
2224 # Remove operators that we don't use and join them with the surrounding strings.
2225 # E.g. 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
2226 ALLOWED_OPS = ('/', '+', ',', '(', ')')
2227 last_string, last_start, last_end, last_line = None, None, None, None
2228 for type_, string_, start, end, line in tokens:
2229 if type_ == tokenize.OP and string_ == '[':
2230 if last_string:
2231 yield tokenize.NAME, last_string, last_start, last_end, last_line
2232 last_string = None
2233 yield type_, string_, start, end, line
2234 # everything inside brackets will be handled by _parse_filter
2235 for type_, string_, start, end, line in tokens:
2236 yield type_, string_, start, end, line
2237 if type_ == tokenize.OP and string_ == ']':
2238 break
2239 elif type_ == tokenize.OP and string_ in ALLOWED_OPS:
2240 if last_string:
2241 yield tokenize.NAME, last_string, last_start, last_end, last_line
2242 last_string = None
2243 yield type_, string_, start, end, line
2244 elif type_ in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
2245 if not last_string:
2246 last_string = string_
2247 last_start = start
2248 last_end = end
2249 else:
2250 last_string += string_
2251 if last_string:
2252 yield tokenize.NAME, last_string, last_start, last_end, last_line
2253
2254 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
2255 selectors = []
2256 current_selector = None
2257 for type_, string_, start, _, _ in tokens:
2258 # ENCODING is only defined in Python 3.x
2259 if type_ == getattr(tokenize, 'ENCODING', None):
2260 continue
2261 elif type_ in [tokenize.NAME, tokenize.NUMBER]:
2262 current_selector = FormatSelector(SINGLE, string_, [])
2263 elif type_ == tokenize.OP:
2264 if string_ == ')':
2265 if not inside_group:
2266 # ')' will be handled by the parentheses group
2267 tokens.restore_last_token()
2268 break
2269 elif inside_merge and string_ in ['/', ',']:
2270 tokens.restore_last_token()
2271 break
2272 elif inside_choice and string_ == ',':
2273 tokens.restore_last_token()
2274 break
2275 elif string_ == ',':
2276 if not current_selector:
2277 raise syntax_error('"," must follow a format selector', start)
2278 selectors.append(current_selector)
2279 current_selector = None
2280 elif string_ == '/':
2281 if not current_selector:
2282 raise syntax_error('"/" must follow a format selector', start)
2283 first_choice = current_selector
2284 second_choice = _parse_format_selection(tokens, inside_choice=True)
2285 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
2286 elif string_ == '[':
2287 if not current_selector:
2288 current_selector = FormatSelector(SINGLE, 'best', [])
2289 format_filter = _parse_filter(tokens)
2290 current_selector.filters.append(format_filter)
2291 elif string_ == '(':
2292 if current_selector:
2293 raise syntax_error('Unexpected "("', start)
2294 group = _parse_format_selection(tokens, inside_group=True)
2295 current_selector = FormatSelector(GROUP, group, [])
2296 elif string_ == '+':
2297 if not current_selector:
2298 raise syntax_error('Unexpected "+"', start)
2299 selector_1 = current_selector
2300 selector_2 = _parse_format_selection(tokens, inside_merge=True)
2301 if not selector_2:
2302 raise syntax_error('Expected a selector', start)
2303 current_selector = FormatSelector(MERGE, (selector_1, selector_2), [])
2304 else:
2305 raise syntax_error(f'Operator not recognized: "{string_}"', start)
2306 elif type_ == tokenize.ENDMARKER:
2307 break
2308 if current_selector:
2309 selectors.append(current_selector)
2310 return selectors
2311
2312 def _merge(formats_pair):
2313 format_1, format_2 = formats_pair
2314
2315 formats_info = []
2316 formats_info.extend(format_1.get('requested_formats', (format_1,)))
2317 formats_info.extend(format_2.get('requested_formats', (format_2,)))
2318
2319 if not allow_multiple_streams['video'] or not allow_multiple_streams['audio']:
2320 get_no_more = {'video': False, 'audio': False}
2321 for (i, fmt_info) in enumerate(formats_info):
2322 if fmt_info.get('acodec') == fmt_info.get('vcodec') == 'none':
2323 formats_info.pop(i)
2324 continue
2325 for aud_vid in ['audio', 'video']:
2326 if not allow_multiple_streams[aud_vid] and fmt_info.get(aud_vid[0] + 'codec') != 'none':
2327 if get_no_more[aud_vid]:
2328 formats_info.pop(i)
2329 break
2330 get_no_more[aud_vid] = True
2331
2332 if len(formats_info) == 1:
2333 return formats_info[0]
2334
2335 video_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('vcodec') != 'none']
2336 audio_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('acodec') != 'none']
2337
2338 the_only_video = video_fmts[0] if len(video_fmts) == 1 else None
2339 the_only_audio = audio_fmts[0] if len(audio_fmts) == 1 else None
2340
2341 output_ext = get_compatible_ext(
2342 vcodecs=[f.get('vcodec') for f in video_fmts],
2343 acodecs=[f.get('acodec') for f in audio_fmts],
2344 vexts=[f['ext'] for f in video_fmts],
2345 aexts=[f['ext'] for f in audio_fmts],
2346 preferences=(try_call(lambda: self.params['merge_output_format'].split('/'))
2347 or self.params.get('prefer_free_formats') and ('webm', 'mkv')))
2348
2349 filtered = lambda *keys: filter(None, (traverse_obj(fmt, *keys) for fmt in formats_info))
2350
2351 new_dict = {
2352 'requested_formats': formats_info,
2353 'format': '+'.join(filtered('format')),
2354 'format_id': '+'.join(filtered('format_id')),
2355 'ext': output_ext,
2356 'protocol': '+'.join(map(determine_protocol, formats_info)),
2357 'language': '+'.join(orderedSet(filtered('language'))) or None,
2358 'format_note': '+'.join(orderedSet(filtered('format_note'))) or None,
2359 'filesize_approx': sum(filtered('filesize', 'filesize_approx')) or None,
2360 'tbr': sum(filtered('tbr', 'vbr', 'abr')),
2361 }
2362
2363 if the_only_video:
2364 new_dict.update({
2365 'width': the_only_video.get('width'),
2366 'height': the_only_video.get('height'),
2367 'resolution': the_only_video.get('resolution') or self.format_resolution(the_only_video),
2368 'fps': the_only_video.get('fps'),
2369 'dynamic_range': the_only_video.get('dynamic_range'),
2370 'vcodec': the_only_video.get('vcodec'),
2371 'vbr': the_only_video.get('vbr'),
2372 'stretched_ratio': the_only_video.get('stretched_ratio'),
2373 'aspect_ratio': the_only_video.get('aspect_ratio'),
2374 })
2375
2376 if the_only_audio:
2377 new_dict.update({
2378 'acodec': the_only_audio.get('acodec'),
2379 'abr': the_only_audio.get('abr'),
2380 'asr': the_only_audio.get('asr'),
2381 'audio_channels': the_only_audio.get('audio_channels'),
2382 })
2383
2384 return new_dict
2385
2386 def _check_formats(formats):
2387 if self.params.get('check_formats') == 'selected':
2388 yield from self._check_formats(formats)
2389 return
2390 elif (self.params.get('check_formats') is not None
2391 or self.params.get('allow_unplayable_formats')):
2392 yield from formats
2393 return
2394
2395 for f in formats:
2396 if f.get('has_drm') or f.get('__needs_testing'):
2397 yield from self._check_formats([f])
2398 else:
2399 yield f
2400
2401 def _build_selector_function(selector):
2402 if isinstance(selector, list): # ,
2403 fs = [_build_selector_function(s) for s in selector]
2404
2405 def selector_function(ctx):
2406 for f in fs:
2407 yield from f(ctx)
2408 return selector_function
2409
2410 elif selector.type == GROUP: # ()
2411 selector_function = _build_selector_function(selector.selector)
2412
2413 elif selector.type == PICKFIRST: # /
2414 fs = [_build_selector_function(s) for s in selector.selector]
2415
2416 def selector_function(ctx):
2417 for f in fs:
2418 picked_formats = list(f(ctx))
2419 if picked_formats:
2420 return picked_formats
2421 return []
2422
2423 elif selector.type == MERGE: # +
2424 selector_1, selector_2 = map(_build_selector_function, selector.selector)
2425
2426 def selector_function(ctx):
2427 for pair in itertools.product(selector_1(ctx), selector_2(ctx)):
2428 yield _merge(pair)
2429
2430 elif selector.type == SINGLE: # atom
2431 format_spec = selector.selector or 'best'
2432
2433 # TODO: Add allvideo, allaudio etc by generalizing the code with best/worst selector
2434 if format_spec == 'all':
2435 def selector_function(ctx):
2436 yield from _check_formats(ctx['formats'][::-1])
2437 elif format_spec == 'mergeall':
2438 def selector_function(ctx):
2439 formats = list(_check_formats(
2440 f for f in ctx['formats'] if f.get('vcodec') != 'none' or f.get('acodec') != 'none'))
2441 if not formats:
2442 return
2443 merged_format = formats[-1]
2444 for f in formats[-2::-1]:
2445 merged_format = _merge((merged_format, f))
2446 yield merged_format
2447
2448 else:
2449 format_fallback, seperate_fallback, format_reverse, format_idx = False, None, True, 1
2450 mobj = re.match(
2451 r'(?P<bw>best|worst|b|w)(?P<type>video|audio|v|a)?(?P<mod>\*)?(?:\.(?P<n>[1-9]\d*))?$',
2452 format_spec)
2453 if mobj is not None:
2454 format_idx = int_or_none(mobj.group('n'), default=1)
2455 format_reverse = mobj.group('bw')[0] == 'b'
2456 format_type = (mobj.group('type') or [None])[0]
2457 not_format_type = {'v': 'a', 'a': 'v'}.get(format_type)
2458 format_modified = mobj.group('mod') is not None
2459
2460 format_fallback = not format_type and not format_modified # for b, w
2461 _filter_f = (
2462 (lambda f: f.get(f'{format_type}codec') != 'none')
2463 if format_type and format_modified # bv*, ba*, wv*, wa*
2464 else (lambda f: f.get(f'{not_format_type}codec') == 'none')
2465 if format_type # bv, ba, wv, wa
2466 else (lambda f: f.get('vcodec') != 'none' and f.get('acodec') != 'none')
2467 if not format_modified # b, w
2468 else lambda f: True) # b*, w*
2469 filter_f = lambda f: _filter_f(f) and (
2470 f.get('vcodec') != 'none' or f.get('acodec') != 'none')
2471 else:
2472 if format_spec in self._format_selection_exts['audio']:
2473 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none'
2474 elif format_spec in self._format_selection_exts['video']:
2475 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none' and f.get('vcodec') != 'none'
2476 seperate_fallback = lambda f: f.get('ext') == format_spec and f.get('vcodec') != 'none'
2477 elif format_spec in self._format_selection_exts['storyboards']:
2478 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') == 'none' and f.get('vcodec') == 'none'
2479 else:
2480 filter_f = lambda f: f.get('format_id') == format_spec # id
2481
2482 def selector_function(ctx):
2483 formats = list(ctx['formats'])
2484 matches = list(filter(filter_f, formats)) if filter_f is not None else formats
2485 if not matches:
2486 if format_fallback and ctx['incomplete_formats']:
2487 # for extractors with incomplete formats (audio only (soundcloud)
2488 # or video only (imgur)) best/worst will fallback to
2489 # best/worst {video,audio}-only format
2490 matches = list(filter(lambda f: f.get('vcodec') != 'none' or f.get('acodec') != 'none', formats))
2491 elif seperate_fallback and not ctx['has_merged_format']:
2492 # for compatibility with youtube-dl when there is no pre-merged format
2493 matches = list(filter(seperate_fallback, formats))
2494 matches = LazyList(_check_formats(matches[::-1 if format_reverse else 1]))
2495 try:
2496 yield matches[format_idx - 1]
2497 except LazyList.IndexError:
2498 return
2499
2500 filters = [self._build_format_filter(f) for f in selector.filters]
2501
2502 def final_selector(ctx):
2503 ctx_copy = dict(ctx)
2504 for _filter in filters:
2505 ctx_copy['formats'] = list(filter(_filter, ctx_copy['formats']))
2506 return selector_function(ctx_copy)
2507 return final_selector
2508
2509 # HACK: Python 3.12 changed the underlying parser, rendering '7_a' invalid
2510 # Prefix numbers with random letters to avoid it being classified as a number
2511 # See: https://github.com/yt-dlp/yt-dlp/pulls/8797
2512 # TODO: Implement parser not reliant on tokenize.tokenize
2513 prefix = ''.join(random.choices(string.ascii_letters, k=32))
2514 stream = io.BytesIO(re.sub(r'\d[_\d]*', rf'{prefix}\g<0>', format_spec).encode())
2515 try:
2516 tokens = list(_remove_unused_ops(
2517 token._replace(string=token.string.replace(prefix, ''))
2518 for token in tokenize.tokenize(stream.readline)))
2519 except tokenize.TokenError:
2520 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
2521
2522 class TokenIterator:
2523 def __init__(self, tokens):
2524 self.tokens = tokens
2525 self.counter = 0
2526
2527 def __iter__(self):
2528 return self
2529
2530 def __next__(self):
2531 if self.counter >= len(self.tokens):
2532 raise StopIteration
2533 value = self.tokens[self.counter]
2534 self.counter += 1
2535 return value
2536
2537 next = __next__
2538
2539 def restore_last_token(self):
2540 self.counter -= 1
2541
2542 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
2543 return _build_selector_function(parsed_selector)
2544
2545 def _calc_headers(self, info_dict, load_cookies=False):
2546 res = HTTPHeaderDict(self.params['http_headers'], info_dict.get('http_headers'))
2547 clean_headers(res)
2548
2549 if load_cookies: # For --load-info-json
2550 self._load_cookies(res.get('Cookie'), autoscope=info_dict['url']) # compat
2551 self._load_cookies(info_dict.get('cookies'), autoscope=False)
2552 # The `Cookie` header is removed to prevent leaks and unscoped cookies.
2553 # See: https://github.com/yt-dlp/yt-dlp/security/advisories/GHSA-v8mc-9377-rwjj
2554 res.pop('Cookie', None)
2555 cookies = self.cookiejar.get_cookies_for_url(info_dict['url'])
2556 if cookies:
2557 encoder = LenientSimpleCookie()
2558 values = []
2559 for cookie in cookies:
2560 _, value = encoder.value_encode(cookie.value)
2561 values.append(f'{cookie.name}={value}')
2562 if cookie.domain:
2563 values.append(f'Domain={cookie.domain}')
2564 if cookie.path:
2565 values.append(f'Path={cookie.path}')
2566 if cookie.secure:
2567 values.append('Secure')
2568 if cookie.expires:
2569 values.append(f'Expires={cookie.expires}')
2570 if cookie.version:
2571 values.append(f'Version={cookie.version}')
2572 info_dict['cookies'] = '; '.join(values)
2573
2574 if 'X-Forwarded-For' not in res:
2575 x_forwarded_for_ip = info_dict.get('__x_forwarded_for_ip')
2576 if x_forwarded_for_ip:
2577 res['X-Forwarded-For'] = x_forwarded_for_ip
2578
2579 return res
2580
2581 def _calc_cookies(self, url):
2582 self.deprecation_warning('"YoutubeDL._calc_cookies" is deprecated and may be removed in a future version')
2583 return self.cookiejar.get_cookie_header(url)
2584
2585 def _sort_thumbnails(self, thumbnails):
2586 thumbnails.sort(key=lambda t: (
2587 t.get('preference') if t.get('preference') is not None else -1,
2588 t.get('width') if t.get('width') is not None else -1,
2589 t.get('height') if t.get('height') is not None else -1,
2590 t.get('id') if t.get('id') is not None else '',
2591 t.get('url')))
2592
2593 def _sanitize_thumbnails(self, info_dict):
2594 thumbnails = info_dict.get('thumbnails')
2595 if thumbnails is None:
2596 thumbnail = info_dict.get('thumbnail')
2597 if thumbnail:
2598 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
2599 if not thumbnails:
2600 return
2601
2602 def check_thumbnails(thumbnails):
2603 for t in thumbnails:
2604 self.to_screen(f'[info] Testing thumbnail {t["id"]}')
2605 try:
2606 self.urlopen(HEADRequest(t['url']))
2607 except network_exceptions as err:
2608 self.to_screen(f'[info] Unable to connect to thumbnail {t["id"]} URL {t["url"]!r} - {err}. Skipping...')
2609 continue
2610 yield t
2611
2612 self._sort_thumbnails(thumbnails)
2613 for i, t in enumerate(thumbnails):
2614 if t.get('id') is None:
2615 t['id'] = str(i)
2616 if t.get('width') and t.get('height'):
2617 t['resolution'] = '%dx%d' % (t['width'], t['height'])
2618 t['url'] = sanitize_url(t['url'])
2619
2620 if self.params.get('check_formats') is True:
2621 info_dict['thumbnails'] = LazyList(check_thumbnails(thumbnails[::-1]), reverse=True)
2622 else:
2623 info_dict['thumbnails'] = thumbnails
2624
2625 def _fill_common_fields(self, info_dict, final=True):
2626 # TODO: move sanitization here
2627 if final:
2628 title = info_dict['fulltitle'] = info_dict.get('title')
2629 if not title:
2630 if title == '':
2631 self.write_debug('Extractor gave empty title. Creating a generic title')
2632 else:
2633 self.report_warning('Extractor failed to obtain "title". Creating a generic title instead')
2634 info_dict['title'] = f'{info_dict["extractor"].replace(":", "-")} video #{info_dict["id"]}'
2635
2636 if info_dict.get('duration') is not None:
2637 info_dict['duration_string'] = formatSeconds(info_dict['duration'])
2638
2639 for ts_key, date_key in (
2640 ('timestamp', 'upload_date'),
2641 ('release_timestamp', 'release_date'),
2642 ('modified_timestamp', 'modified_date'),
2643 ):
2644 if info_dict.get(date_key) is None and info_dict.get(ts_key) is not None:
2645 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
2646 # see http://bugs.python.org/issue1646728)
2647 with contextlib.suppress(ValueError, OverflowError, OSError):
2648 upload_date = dt.datetime.fromtimestamp(info_dict[ts_key], dt.timezone.utc)
2649 info_dict[date_key] = upload_date.strftime('%Y%m%d')
2650
2651 if not info_dict.get('release_year'):
2652 info_dict['release_year'] = traverse_obj(info_dict, ('release_date', {lambda x: int(x[:4])}))
2653
2654 live_keys = ('is_live', 'was_live')
2655 live_status = info_dict.get('live_status')
2656 if live_status is None:
2657 for key in live_keys:
2658 if info_dict.get(key) is False:
2659 continue
2660 if info_dict.get(key):
2661 live_status = key
2662 break
2663 if all(info_dict.get(key) is False for key in live_keys):
2664 live_status = 'not_live'
2665 if live_status:
2666 info_dict['live_status'] = live_status
2667 for key in live_keys:
2668 if info_dict.get(key) is None:
2669 info_dict[key] = (live_status == key)
2670 if live_status == 'post_live':
2671 info_dict['was_live'] = True
2672
2673 # Auto generate title fields corresponding to the *_number fields when missing
2674 # in order to always have clean titles. This is very common for TV series.
2675 for field in ('chapter', 'season', 'episode'):
2676 if final and info_dict.get(f'{field}_number') is not None and not info_dict.get(field):
2677 info_dict[field] = '%s %d' % (field.capitalize(), info_dict[f'{field}_number'])
2678
2679 for old_key, new_key in self._deprecated_multivalue_fields.items():
2680 if new_key in info_dict and old_key in info_dict:
2681 if '_version' not in info_dict: # HACK: Do not warn when using --load-info-json
2682 self.deprecation_warning(f'Do not return {old_key!r} when {new_key!r} is present')
2683 elif old_value := info_dict.get(old_key):
2684 info_dict[new_key] = old_value.split(', ')
2685 elif new_value := info_dict.get(new_key):
2686 info_dict[old_key] = ', '.join(v.replace(',', '\N{FULLWIDTH COMMA}') for v in new_value)
2687
2688 def _raise_pending_errors(self, info):
2689 err = info.pop('__pending_error', None)
2690 if err:
2691 self.report_error(err, tb=False)
2692
2693 def sort_formats(self, info_dict):
2694 formats = self._get_formats(info_dict)
2695 formats.sort(key=FormatSorter(
2696 self, info_dict.get('_format_sort_fields') or []).calculate_preference)
2697
2698 def process_video_result(self, info_dict, download=True):
2699 assert info_dict.get('_type', 'video') == 'video'
2700 self._num_videos += 1
2701
2702 if 'id' not in info_dict:
2703 raise ExtractorError('Missing "id" field in extractor result', ie=info_dict['extractor'])
2704 elif not info_dict.get('id'):
2705 raise ExtractorError('Extractor failed to obtain "id"', ie=info_dict['extractor'])
2706
2707 def report_force_conversion(field, field_not, conversion):
2708 self.report_warning(
2709 f'"{field}" field is not {field_not} - forcing {conversion} conversion, '
2710 'there is an error in extractor')
2711
2712 def sanitize_string_field(info, string_field):
2713 field = info.get(string_field)
2714 if field is None or isinstance(field, str):
2715 return
2716 report_force_conversion(string_field, 'a string', 'string')
2717 info[string_field] = str(field)
2718
2719 def sanitize_numeric_fields(info):
2720 for numeric_field in self._NUMERIC_FIELDS:
2721 field = info.get(numeric_field)
2722 if field is None or isinstance(field, (int, float)):
2723 continue
2724 report_force_conversion(numeric_field, 'numeric', 'int')
2725 info[numeric_field] = int_or_none(field)
2726
2727 sanitize_string_field(info_dict, 'id')
2728 sanitize_numeric_fields(info_dict)
2729 if info_dict.get('section_end') and info_dict.get('section_start') is not None:
2730 info_dict['duration'] = round(info_dict['section_end'] - info_dict['section_start'], 3)
2731 if (info_dict.get('duration') or 0) <= 0 and info_dict.pop('duration', None):
2732 self.report_warning('"duration" field is negative, there is an error in extractor')
2733
2734 chapters = info_dict.get('chapters') or []
2735 if chapters and chapters[0].get('start_time'):
2736 chapters.insert(0, {'start_time': 0})
2737
2738 dummy_chapter = {'end_time': 0, 'start_time': info_dict.get('duration')}
2739 for idx, (prev, current, next_) in enumerate(zip(
2740 (dummy_chapter, *chapters), chapters, (*chapters[1:], dummy_chapter)), 1):
2741 if current.get('start_time') is None:
2742 current['start_time'] = prev.get('end_time')
2743 if not current.get('end_time'):
2744 current['end_time'] = next_.get('start_time')
2745 if not current.get('title'):
2746 current['title'] = f'<Untitled Chapter {idx}>'
2747
2748 if 'playlist' not in info_dict:
2749 # It isn't part of a playlist
2750 info_dict['playlist'] = None
2751 info_dict['playlist_index'] = None
2752
2753 self._sanitize_thumbnails(info_dict)
2754
2755 thumbnail = info_dict.get('thumbnail')
2756 thumbnails = info_dict.get('thumbnails')
2757 if thumbnail:
2758 info_dict['thumbnail'] = sanitize_url(thumbnail)
2759 elif thumbnails:
2760 info_dict['thumbnail'] = thumbnails[-1]['url']
2761
2762 if info_dict.get('display_id') is None and 'id' in info_dict:
2763 info_dict['display_id'] = info_dict['id']
2764
2765 self._fill_common_fields(info_dict)
2766
2767 for cc_kind in ('subtitles', 'automatic_captions'):
2768 cc = info_dict.get(cc_kind)
2769 if cc:
2770 for _, subtitle in cc.items():
2771 for subtitle_format in subtitle:
2772 if subtitle_format.get('url'):
2773 subtitle_format['url'] = sanitize_url(subtitle_format['url'])
2774 if subtitle_format.get('ext') is None:
2775 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
2776
2777 automatic_captions = info_dict.get('automatic_captions')
2778 subtitles = info_dict.get('subtitles')
2779
2780 info_dict['requested_subtitles'] = self.process_subtitles(
2781 info_dict['id'], subtitles, automatic_captions)
2782
2783 formats = self._get_formats(info_dict)
2784
2785 # Backward compatibility with InfoExtractor._sort_formats
2786 field_preference = (formats or [{}])[0].pop('__sort_fields', None)
2787 if field_preference:
2788 info_dict['_format_sort_fields'] = field_preference
2789
2790 info_dict['_has_drm'] = any( # or None ensures --clean-infojson removes it
2791 f.get('has_drm') and f['has_drm'] != 'maybe' for f in formats) or None
2792 if not self.params.get('allow_unplayable_formats'):
2793 formats = [f for f in formats if not f.get('has_drm') or f['has_drm'] == 'maybe']
2794
2795 if formats and all(f.get('acodec') == f.get('vcodec') == 'none' for f in formats):
2796 self.report_warning(
2797 f'{"This video is DRM protected and " if info_dict["_has_drm"] else ""}'
2798 'only images are available for download. Use --list-formats to see them'.capitalize())
2799
2800 get_from_start = not info_dict.get('is_live') or bool(self.params.get('live_from_start'))
2801 if not get_from_start:
2802 info_dict['title'] += ' ' + dt.datetime.now().strftime('%Y-%m-%d %H:%M')
2803 if info_dict.get('is_live') and formats:
2804 formats = [f for f in formats if bool(f.get('is_from_start')) == get_from_start]
2805 if get_from_start and not formats:
2806 self.raise_no_formats(info_dict, msg=(
2807 '--live-from-start is passed, but there are no formats that can be downloaded from the start. '
2808 'If you want to download from the current time, use --no-live-from-start'))
2809
2810 def is_wellformed(f):
2811 url = f.get('url')
2812 if not url:
2813 self.report_warning(
2814 '"url" field is missing or empty - skipping format, '
2815 'there is an error in extractor')
2816 return False
2817 if isinstance(url, bytes):
2818 sanitize_string_field(f, 'url')
2819 return True
2820
2821 # Filter out malformed formats for better extraction robustness
2822 formats = list(filter(is_wellformed, formats or []))
2823
2824 if not formats:
2825 self.raise_no_formats(info_dict)
2826
2827 for fmt in formats:
2828 sanitize_string_field(fmt, 'format_id')
2829 sanitize_numeric_fields(fmt)
2830 fmt['url'] = sanitize_url(fmt['url'])
2831 if fmt.get('ext') is None:
2832 fmt['ext'] = determine_ext(fmt['url']).lower()
2833 if fmt['ext'] in ('aac', 'opus', 'mp3', 'flac', 'vorbis'):
2834 if fmt.get('acodec') is None:
2835 fmt['acodec'] = fmt['ext']
2836 if fmt.get('protocol') is None:
2837 fmt['protocol'] = determine_protocol(fmt)
2838 if fmt.get('resolution') is None:
2839 fmt['resolution'] = self.format_resolution(fmt, default=None)
2840 if fmt.get('dynamic_range') is None and fmt.get('vcodec') != 'none':
2841 fmt['dynamic_range'] = 'SDR'
2842 if fmt.get('aspect_ratio') is None:
2843 fmt['aspect_ratio'] = try_call(lambda: round(fmt['width'] / fmt['height'], 2))
2844 # For fragmented formats, "tbr" is often max bitrate and not average
2845 if (('manifest-filesize-approx' in self.params['compat_opts'] or not fmt.get('manifest_url'))
2846 and not fmt.get('filesize') and not fmt.get('filesize_approx')):
2847 fmt['filesize_approx'] = filesize_from_tbr(fmt.get('tbr'), info_dict.get('duration'))
2848 fmt['http_headers'] = self._calc_headers(collections.ChainMap(fmt, info_dict), load_cookies=True)
2849
2850 # Safeguard against old/insecure infojson when using --load-info-json
2851 if info_dict.get('http_headers'):
2852 info_dict['http_headers'] = HTTPHeaderDict(info_dict['http_headers'])
2853 info_dict['http_headers'].pop('Cookie', None)
2854
2855 # This is copied to http_headers by the above _calc_headers and can now be removed
2856 if '__x_forwarded_for_ip' in info_dict:
2857 del info_dict['__x_forwarded_for_ip']
2858
2859 self.sort_formats({
2860 'formats': formats,
2861 '_format_sort_fields': info_dict.get('_format_sort_fields'),
2862 })
2863
2864 # Sanitize and group by format_id
2865 formats_dict = {}
2866 for i, fmt in enumerate(formats):
2867 if not fmt.get('format_id'):
2868 fmt['format_id'] = str(i)
2869 else:
2870 # Sanitize format_id from characters used in format selector expression
2871 fmt['format_id'] = re.sub(r'[\s,/+\[\]()]', '_', fmt['format_id'])
2872 formats_dict.setdefault(fmt['format_id'], []).append(fmt)
2873
2874 # Make sure all formats have unique format_id
2875 common_exts = set(itertools.chain(*self._format_selection_exts.values()))
2876 for format_id, ambiguous_formats in formats_dict.items():
2877 ambigious_id = len(ambiguous_formats) > 1
2878 for i, fmt in enumerate(ambiguous_formats):
2879 if ambigious_id:
2880 fmt['format_id'] = f'{format_id}-{i}'
2881 # Ensure there is no conflict between id and ext in format selection
2882 # See https://github.com/yt-dlp/yt-dlp/issues/1282
2883 if fmt['format_id'] != fmt['ext'] and fmt['format_id'] in common_exts:
2884 fmt['format_id'] = 'f{}'.format(fmt['format_id'])
2885
2886 if fmt.get('format') is None:
2887 fmt['format'] = '{id} - {res}{note}'.format(
2888 id=fmt['format_id'],
2889 res=self.format_resolution(fmt),
2890 note=format_field(fmt, 'format_note', ' (%s)'),
2891 )
2892
2893 if self.params.get('check_formats') is True:
2894 formats = LazyList(self._check_formats(formats[::-1]), reverse=True)
2895
2896 if not formats or formats[0] is not info_dict:
2897 # only set the 'formats' fields if the original info_dict list them
2898 # otherwise we end up with a circular reference, the first (and unique)
2899 # element in the 'formats' field in info_dict is info_dict itself,
2900 # which can't be exported to json
2901 info_dict['formats'] = formats
2902
2903 info_dict, _ = self.pre_process(info_dict)
2904
2905 if self._match_entry(info_dict, incomplete=self._format_fields) is not None:
2906 return info_dict
2907
2908 self.post_extract(info_dict)
2909 info_dict, _ = self.pre_process(info_dict, 'after_filter')
2910
2911 # The pre-processors may have modified the formats
2912 formats = self._get_formats(info_dict)
2913
2914 list_only = self.params.get('simulate') == 'list_only'
2915 interactive_format_selection = not list_only and self.format_selector == '-'
2916 if self.params.get('list_thumbnails'):
2917 self.list_thumbnails(info_dict)
2918 if self.params.get('listsubtitles'):
2919 if 'automatic_captions' in info_dict:
2920 self.list_subtitles(
2921 info_dict['id'], automatic_captions, 'automatic captions')
2922 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
2923 if self.params.get('listformats') or interactive_format_selection:
2924 self.list_formats(info_dict)
2925 if list_only:
2926 # Without this printing, -F --print-json will not work
2927 self.__forced_printings(info_dict)
2928 return info_dict
2929
2930 format_selector = self.format_selector
2931 while True:
2932 if interactive_format_selection:
2933 req_format = input(self._format_screen('\nEnter format selector ', self.Styles.EMPHASIS)
2934 + '(Press ENTER for default, or Ctrl+C to quit)'
2935 + self._format_screen(': ', self.Styles.EMPHASIS))
2936 try:
2937 format_selector = self.build_format_selector(req_format) if req_format else None
2938 except SyntaxError as err:
2939 self.report_error(err, tb=False, is_error=False)
2940 continue
2941
2942 if format_selector is None:
2943 req_format = self._default_format_spec(info_dict, download=download)
2944 self.write_debug(f'Default format spec: {req_format}')
2945 format_selector = self.build_format_selector(req_format)
2946
2947 formats_to_download = self._select_formats(formats, format_selector)
2948 if interactive_format_selection and not formats_to_download:
2949 self.report_error('Requested format is not available', tb=False, is_error=False)
2950 continue
2951 break
2952
2953 if not formats_to_download:
2954 if not self.params.get('ignore_no_formats_error'):
2955 raise ExtractorError(
2956 'Requested format is not available. Use --list-formats for a list of available formats',
2957 expected=True, video_id=info_dict['id'], ie=info_dict['extractor'])
2958 self.report_warning('Requested format is not available')
2959 # Process what we can, even without any available formats.
2960 formats_to_download = [{}]
2961
2962 requested_ranges = tuple(self.params.get('download_ranges', lambda *_: [{}])(info_dict, self))
2963 best_format, downloaded_formats = formats_to_download[-1], []
2964 if download:
2965 if best_format and requested_ranges:
2966 def to_screen(*msg):
2967 self.to_screen(f'[info] {info_dict["id"]}: {" ".join(", ".join(variadic(m)) for m in msg)}')
2968
2969 to_screen(f'Downloading {len(formats_to_download)} format(s):',
2970 (f['format_id'] for f in formats_to_download))
2971 if requested_ranges != ({}, ):
2972 to_screen(f'Downloading {len(requested_ranges)} time ranges:',
2973 (f'{c["start_time"]:.1f}-{c["end_time"]:.1f}' for c in requested_ranges))
2974 max_downloads_reached = False
2975
2976 for fmt, chapter in itertools.product(formats_to_download, requested_ranges):
2977 new_info = self._copy_infodict(info_dict)
2978 new_info.update(fmt)
2979 offset, duration = info_dict.get('section_start') or 0, info_dict.get('duration') or float('inf')
2980 end_time = offset + min(chapter.get('end_time', duration), duration)
2981 # duration may not be accurate. So allow deviations <1sec
2982 if end_time == float('inf') or end_time > offset + duration + 1:
2983 end_time = None
2984 if chapter or offset:
2985 new_info.update({
2986 'section_start': offset + chapter.get('start_time', 0),
2987 'section_end': end_time,
2988 'section_title': chapter.get('title'),
2989 'section_number': chapter.get('index'),
2990 })
2991 downloaded_formats.append(new_info)
2992 try:
2993 self.process_info(new_info)
2994 except MaxDownloadsReached:
2995 max_downloads_reached = True
2996 self._raise_pending_errors(new_info)
2997 # Remove copied info
2998 for key, val in tuple(new_info.items()):
2999 if info_dict.get(key) == val:
3000 new_info.pop(key)
3001 if max_downloads_reached:
3002 break
3003
3004 write_archive = {f.get('__write_download_archive', False) for f in downloaded_formats}
3005 assert write_archive.issubset({True, False, 'ignore'})
3006 if True in write_archive and False not in write_archive:
3007 self.record_download_archive(info_dict)
3008
3009 info_dict['requested_downloads'] = downloaded_formats
3010 info_dict = self.run_all_pps('after_video', info_dict)
3011 if max_downloads_reached:
3012 raise MaxDownloadsReached
3013
3014 # We update the info dict with the selected best quality format (backwards compatibility)
3015 info_dict.update(best_format)
3016 return info_dict
3017
3018 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
3019 """Select the requested subtitles and their format"""
3020 available_subs, normal_sub_langs = {}, []
3021 if normal_subtitles and self.params.get('writesubtitles'):
3022 available_subs.update(normal_subtitles)
3023 normal_sub_langs = tuple(normal_subtitles.keys())
3024 if automatic_captions and self.params.get('writeautomaticsub'):
3025 for lang, cap_info in automatic_captions.items():
3026 if lang not in available_subs:
3027 available_subs[lang] = cap_info
3028
3029 if not available_subs or (
3030 not self.params.get('writesubtitles')
3031 and not self.params.get('writeautomaticsub')):
3032 return None
3033
3034 all_sub_langs = tuple(available_subs.keys())
3035 if self.params.get('allsubtitles', False):
3036 requested_langs = all_sub_langs
3037 elif self.params.get('subtitleslangs', False):
3038 try:
3039 requested_langs = orderedSet_from_options(
3040 self.params.get('subtitleslangs'), {'all': all_sub_langs}, use_regex=True)
3041 except re.error as e:
3042 raise ValueError(f'Wrong regex for subtitlelangs: {e.pattern}')
3043 else:
3044 requested_langs = LazyList(itertools.chain(
3045 ['en'] if 'en' in normal_sub_langs else [],
3046 filter(lambda f: f.startswith('en'), normal_sub_langs),
3047 ['en'] if 'en' in all_sub_langs else [],
3048 filter(lambda f: f.startswith('en'), all_sub_langs),
3049 normal_sub_langs, all_sub_langs,
3050 ))[:1]
3051 if requested_langs:
3052 self.to_screen(f'[info] {video_id}: Downloading subtitles: {", ".join(requested_langs)}')
3053
3054 formats_query = self.params.get('subtitlesformat', 'best')
3055 formats_preference = formats_query.split('/') if formats_query else []
3056 subs = {}
3057 for lang in requested_langs:
3058 formats = available_subs.get(lang)
3059 if formats is None:
3060 self.report_warning(f'{lang} subtitles not available for {video_id}')
3061 continue
3062 for ext in formats_preference:
3063 if ext == 'best':
3064 f = formats[-1]
3065 break
3066 matches = list(filter(lambda f: f['ext'] == ext, formats))
3067 if matches:
3068 f = matches[-1]
3069 break
3070 else:
3071 f = formats[-1]
3072 self.report_warning(
3073 'No subtitle format found matching "{}" for language {}, '
3074 'using {}. Use --list-subs for a list of available subtitles'.format(formats_query, lang, f['ext']))
3075 subs[lang] = f
3076 return subs
3077
3078 def _forceprint(self, key, info_dict):
3079 if info_dict is None:
3080 return
3081 info_copy = info_dict.copy()
3082 info_copy.setdefault('filename', self.prepare_filename(info_dict))
3083 if info_dict.get('requested_formats') is not None:
3084 # For RTMP URLs, also include the playpath
3085 info_copy['urls'] = '\n'.join(f['url'] + f.get('play_path', '') for f in info_dict['requested_formats'])
3086 elif info_dict.get('url'):
3087 info_copy['urls'] = info_dict['url'] + info_dict.get('play_path', '')
3088 info_copy['formats_table'] = self.render_formats_table(info_dict)
3089 info_copy['thumbnails_table'] = self.render_thumbnails_table(info_dict)
3090 info_copy['subtitles_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('subtitles'))
3091 info_copy['automatic_captions_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('automatic_captions'))
3092
3093 def format_tmpl(tmpl):
3094 mobj = re.fullmatch(r'([\w.:,]|-\d|(?P<dict>{([\w.:,]|-\d)+}))+=?', tmpl)
3095 if not mobj:
3096 return tmpl
3097
3098 fmt = '%({})s'
3099 if tmpl.startswith('{'):
3100 tmpl, fmt = f'.{tmpl}', '%({})j'
3101 if tmpl.endswith('='):
3102 tmpl, fmt = tmpl[:-1], '{0} = %({0})#j'
3103 return '\n'.join(map(fmt.format, [tmpl] if mobj.group('dict') else tmpl.split(',')))
3104
3105 for tmpl in self.params['forceprint'].get(key, []):
3106 self.to_stdout(self.evaluate_outtmpl(format_tmpl(tmpl), info_copy))
3107
3108 for tmpl, file_tmpl in self.params['print_to_file'].get(key, []):
3109 filename = self.prepare_filename(info_dict, outtmpl=file_tmpl)
3110 tmpl = format_tmpl(tmpl)
3111 self.to_screen(f'[info] Writing {tmpl!r} to: {filename}')
3112 if self._ensure_dir_exists(filename):
3113 with open(filename, 'a', encoding='utf-8', newline='') as f:
3114 f.write(self.evaluate_outtmpl(tmpl, info_copy) + os.linesep)
3115
3116 return info_copy
3117
3118 def __forced_printings(self, info_dict, filename=None, incomplete=True):
3119 if (self.params.get('forcejson')
3120 or self.params['forceprint'].get('video')
3121 or self.params['print_to_file'].get('video')):
3122 self.post_extract(info_dict)
3123 if filename:
3124 info_dict['filename'] = filename
3125 info_copy = self._forceprint('video', info_dict)
3126
3127 def print_field(field, actual_field=None, optional=False):
3128 if actual_field is None:
3129 actual_field = field
3130 if self.params.get(f'force{field}') and (
3131 info_copy.get(field) is not None or (not optional and not incomplete)):
3132 self.to_stdout(info_copy[actual_field])
3133
3134 print_field('title')
3135 print_field('id')
3136 print_field('url', 'urls')
3137 print_field('thumbnail', optional=True)
3138 print_field('description', optional=True)
3139 print_field('filename')
3140 if self.params.get('forceduration') and info_copy.get('duration') is not None:
3141 self.to_stdout(formatSeconds(info_copy['duration']))
3142 print_field('format')
3143
3144 if self.params.get('forcejson'):
3145 self.to_stdout(json.dumps(self.sanitize_info(info_dict)))
3146
3147 def dl(self, name, info, subtitle=False, test=False):
3148 if not info.get('url'):
3149 self.raise_no_formats(info, True)
3150
3151 if test:
3152 verbose = self.params.get('verbose')
3153 params = {
3154 'test': True,
3155 'quiet': self.params.get('quiet') or not verbose,
3156 'verbose': verbose,
3157 'noprogress': not verbose,
3158 'nopart': True,
3159 'skip_unavailable_fragments': False,
3160 'keep_fragments': False,
3161 'overwrites': True,
3162 '_no_ytdl_file': True,
3163 }
3164 else:
3165 params = self.params
3166 fd = get_suitable_downloader(info, params, to_stdout=(name == '-'))(self, params)
3167 if not test:
3168 for ph in self._progress_hooks:
3169 fd.add_progress_hook(ph)
3170 urls = '", "'.join(
3171 (f['url'].split(',')[0] + ',<data>' if f['url'].startswith('data:') else f['url'])
3172 for f in info.get('requested_formats', []) or [info])
3173 self.write_debug(f'Invoking {fd.FD_NAME} downloader on "{urls}"')
3174
3175 # Note: Ideally info should be a deep-copied so that hooks cannot modify it.
3176 # But it may contain objects that are not deep-copyable
3177 new_info = self._copy_infodict(info)
3178 if new_info.get('http_headers') is None:
3179 new_info['http_headers'] = self._calc_headers(new_info)
3180 return fd.download(name, new_info, subtitle)
3181
3182 def existing_file(self, filepaths, *, default_overwrite=True):
3183 existing_files = list(filter(os.path.exists, orderedSet(filepaths)))
3184 if existing_files and not self.params.get('overwrites', default_overwrite):
3185 return existing_files[0]
3186
3187 for file in existing_files:
3188 self.report_file_delete(file)
3189 os.remove(file)
3190 return None
3191
3192 def process_info(self, info_dict):
3193 """Process a single resolved IE result. (Modifies it in-place)"""
3194
3195 assert info_dict.get('_type', 'video') == 'video'
3196 original_infodict = info_dict
3197
3198 if 'format' not in info_dict and 'ext' in info_dict:
3199 info_dict['format'] = info_dict['ext']
3200
3201 if self._match_entry(info_dict) is not None:
3202 info_dict['__write_download_archive'] = 'ignore'
3203 return
3204
3205 # Does nothing under normal operation - for backward compatibility of process_info
3206 self.post_extract(info_dict)
3207
3208 def replace_info_dict(new_info):
3209 nonlocal info_dict
3210 if new_info == info_dict:
3211 return
3212 info_dict.clear()
3213 info_dict.update(new_info)
3214
3215 new_info, _ = self.pre_process(info_dict, 'video')
3216 replace_info_dict(new_info)
3217 self._num_downloads += 1
3218
3219 # info_dict['_filename'] needs to be set for backward compatibility
3220 info_dict['_filename'] = full_filename = self.prepare_filename(info_dict, warn=True)
3221 temp_filename = self.prepare_filename(info_dict, 'temp')
3222 files_to_move = {}
3223
3224 # Forced printings
3225 self.__forced_printings(info_dict, full_filename, incomplete=('format' not in info_dict))
3226
3227 def check_max_downloads():
3228 if self._num_downloads >= float(self.params.get('max_downloads') or 'inf'):
3229 raise MaxDownloadsReached
3230
3231 if self.params.get('simulate'):
3232 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
3233 check_max_downloads()
3234 return
3235
3236 if full_filename is None:
3237 return
3238 if not self._ensure_dir_exists(encodeFilename(full_filename)):
3239 return
3240 if not self._ensure_dir_exists(encodeFilename(temp_filename)):
3241 return
3242
3243 if self._write_description('video', info_dict,
3244 self.prepare_filename(info_dict, 'description')) is None:
3245 return
3246
3247 sub_files = self._write_subtitles(info_dict, temp_filename)
3248 if sub_files is None:
3249 return
3250 files_to_move.update(dict(sub_files))
3251
3252 thumb_files = self._write_thumbnails(
3253 'video', info_dict, temp_filename, self.prepare_filename(info_dict, 'thumbnail'))
3254 if thumb_files is None:
3255 return
3256 files_to_move.update(dict(thumb_files))
3257
3258 infofn = self.prepare_filename(info_dict, 'infojson')
3259 _infojson_written = self._write_info_json('video', info_dict, infofn)
3260 if _infojson_written:
3261 info_dict['infojson_filename'] = infofn
3262 # For backward compatibility, even though it was a private field
3263 info_dict['__infojson_filename'] = infofn
3264 elif _infojson_written is None:
3265 return
3266
3267 # Note: Annotations are deprecated
3268 annofn = None
3269 if self.params.get('writeannotations', False):
3270 annofn = self.prepare_filename(info_dict, 'annotation')
3271 if annofn:
3272 if not self._ensure_dir_exists(encodeFilename(annofn)):
3273 return
3274 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(annofn)):
3275 self.to_screen('[info] Video annotations are already present')
3276 elif not info_dict.get('annotations'):
3277 self.report_warning('There are no annotations to write.')
3278 else:
3279 try:
3280 self.to_screen('[info] Writing video annotations to: ' + annofn)
3281 with open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
3282 annofile.write(info_dict['annotations'])
3283 except (KeyError, TypeError):
3284 self.report_warning('There are no annotations to write.')
3285 except OSError:
3286 self.report_error('Cannot write annotations file: ' + annofn)
3287 return
3288
3289 # Write internet shortcut files
3290 def _write_link_file(link_type):
3291 url = try_get(info_dict['webpage_url'], iri_to_uri)
3292 if not url:
3293 self.report_warning(
3294 f'Cannot write internet shortcut file because the actual URL of "{info_dict["webpage_url"]}" is unknown')
3295 return True
3296 linkfn = replace_extension(self.prepare_filename(info_dict, 'link'), link_type, info_dict.get('ext'))
3297 if not self._ensure_dir_exists(encodeFilename(linkfn)):
3298 return False
3299 if self.params.get('overwrites', True) and os.path.exists(encodeFilename(linkfn)):
3300 self.to_screen(f'[info] Internet shortcut (.{link_type}) is already present')
3301 return True
3302 try:
3303 self.to_screen(f'[info] Writing internet shortcut (.{link_type}) to: {linkfn}')
3304 with open(encodeFilename(to_high_limit_path(linkfn)), 'w', encoding='utf-8',
3305 newline='\r\n' if link_type == 'url' else '\n') as linkfile:
3306 template_vars = {'url': url}
3307 if link_type == 'desktop':
3308 template_vars['filename'] = linkfn[:-(len(link_type) + 1)]
3309 linkfile.write(LINK_TEMPLATES[link_type] % template_vars)
3310 except OSError:
3311 self.report_error(f'Cannot write internet shortcut {linkfn}')
3312 return False
3313 return True
3314
3315 write_links = {
3316 'url': self.params.get('writeurllink'),
3317 'webloc': self.params.get('writewebloclink'),
3318 'desktop': self.params.get('writedesktoplink'),
3319 }
3320 if self.params.get('writelink'):
3321 link_type = ('webloc' if sys.platform == 'darwin'
3322 else 'desktop' if sys.platform.startswith('linux')
3323 else 'url')
3324 write_links[link_type] = True
3325
3326 if any(should_write and not _write_link_file(link_type)
3327 for link_type, should_write in write_links.items()):
3328 return
3329
3330 new_info, files_to_move = self.pre_process(info_dict, 'before_dl', files_to_move)
3331 replace_info_dict(new_info)
3332
3333 if self.params.get('skip_download'):
3334 info_dict['filepath'] = temp_filename
3335 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
3336 info_dict['__files_to_move'] = files_to_move
3337 replace_info_dict(self.run_pp(MoveFilesAfterDownloadPP(self, False), info_dict))
3338 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
3339 else:
3340 # Download
3341 info_dict.setdefault('__postprocessors', [])
3342 try:
3343
3344 def existing_video_file(*filepaths):
3345 ext = info_dict.get('ext')
3346 converted = lambda file: replace_extension(file, self.params.get('final_ext') or ext, ext)
3347 file = self.existing_file(itertools.chain(*zip(map(converted, filepaths), filepaths)),
3348 default_overwrite=False)
3349 if file:
3350 info_dict['ext'] = os.path.splitext(file)[1][1:]
3351 return file
3352
3353 fd, success = None, True
3354 if info_dict.get('protocol') or info_dict.get('url'):
3355 fd = get_suitable_downloader(info_dict, self.params, to_stdout=temp_filename == '-')
3356 if fd != FFmpegFD and 'no-direct-merge' not in self.params['compat_opts'] and (
3357 info_dict.get('section_start') or info_dict.get('section_end')):
3358 msg = ('This format cannot be partially downloaded' if FFmpegFD.available()
3359 else 'You have requested downloading the video partially, but ffmpeg is not installed')
3360 self.report_error(f'{msg}. Aborting')
3361 return
3362
3363 if info_dict.get('requested_formats') is not None:
3364 old_ext = info_dict['ext']
3365 if self.params.get('merge_output_format') is None:
3366 if (info_dict['ext'] == 'webm'
3367 and info_dict.get('thumbnails')
3368 # check with type instead of pp_key, __name__, or isinstance
3369 # since we dont want any custom PPs to trigger this
3370 and any(type(pp) == EmbedThumbnailPP for pp in self._pps['post_process'])): # noqa: E721
3371 info_dict['ext'] = 'mkv'
3372 self.report_warning(
3373 'webm doesn\'t support embedding a thumbnail, mkv will be used')
3374 new_ext = info_dict['ext']
3375
3376 def correct_ext(filename, ext=new_ext):
3377 if filename == '-':
3378 return filename
3379 filename_real_ext = os.path.splitext(filename)[1][1:]
3380 filename_wo_ext = (
3381 os.path.splitext(filename)[0]
3382 if filename_real_ext in (old_ext, new_ext)
3383 else filename)
3384 return f'{filename_wo_ext}.{ext}'
3385
3386 # Ensure filename always has a correct extension for successful merge
3387 full_filename = correct_ext(full_filename)
3388 temp_filename = correct_ext(temp_filename)
3389 dl_filename = existing_video_file(full_filename, temp_filename)
3390
3391 info_dict['__real_download'] = False
3392 # NOTE: Copy so that original format dicts are not modified
3393 info_dict['requested_formats'] = list(map(dict, info_dict['requested_formats']))
3394
3395 merger = FFmpegMergerPP(self)
3396 downloaded = []
3397 if dl_filename is not None:
3398 self.report_file_already_downloaded(dl_filename)
3399 elif fd:
3400 for f in info_dict['requested_formats'] if fd != FFmpegFD else []:
3401 f['filepath'] = fname = prepend_extension(
3402 correct_ext(temp_filename, info_dict['ext']),
3403 'f{}'.format(f['format_id']), info_dict['ext'])
3404 downloaded.append(fname)
3405 info_dict['url'] = '\n'.join(f['url'] for f in info_dict['requested_formats'])
3406 success, real_download = self.dl(temp_filename, info_dict)
3407 info_dict['__real_download'] = real_download
3408 else:
3409 if self.params.get('allow_unplayable_formats'):
3410 self.report_warning(
3411 'You have requested merging of multiple formats '
3412 'while also allowing unplayable formats to be downloaded. '
3413 'The formats won\'t be merged to prevent data corruption.')
3414 elif not merger.available:
3415 msg = 'You have requested merging of multiple formats but ffmpeg is not installed'
3416 if not self.params.get('ignoreerrors'):
3417 self.report_error(f'{msg}. Aborting due to --abort-on-error')
3418 return
3419 self.report_warning(f'{msg}. The formats won\'t be merged')
3420
3421 if temp_filename == '-':
3422 reason = ('using a downloader other than ffmpeg' if FFmpegFD.can_merge_formats(info_dict, self.params)
3423 else 'but the formats are incompatible for simultaneous download' if merger.available
3424 else 'but ffmpeg is not installed')
3425 self.report_warning(
3426 f'You have requested downloading multiple formats to stdout {reason}. '
3427 'The formats will be streamed one after the other')
3428 fname = temp_filename
3429 for f in info_dict['requested_formats']:
3430 new_info = dict(info_dict)
3431 del new_info['requested_formats']
3432 new_info.update(f)
3433 if temp_filename != '-':
3434 fname = prepend_extension(
3435 correct_ext(temp_filename, new_info['ext']),
3436 'f{}'.format(f['format_id']), new_info['ext'])
3437 if not self._ensure_dir_exists(fname):
3438 return
3439 f['filepath'] = fname
3440 downloaded.append(fname)
3441 partial_success, real_download = self.dl(fname, new_info)
3442 info_dict['__real_download'] = info_dict['__real_download'] or real_download
3443 success = success and partial_success
3444
3445 if downloaded and merger.available and not self.params.get('allow_unplayable_formats'):
3446 info_dict['__postprocessors'].append(merger)
3447 info_dict['__files_to_merge'] = downloaded
3448 # Even if there were no downloads, it is being merged only now
3449 info_dict['__real_download'] = True
3450 else:
3451 for file in downloaded:
3452 files_to_move[file] = None
3453 else:
3454 # Just a single file
3455 dl_filename = existing_video_file(full_filename, temp_filename)
3456 if dl_filename is None or dl_filename == temp_filename:
3457 # dl_filename == temp_filename could mean that the file was partially downloaded with --no-part.
3458 # So we should try to resume the download
3459 success, real_download = self.dl(temp_filename, info_dict)
3460 info_dict['__real_download'] = real_download
3461 else:
3462 self.report_file_already_downloaded(dl_filename)
3463
3464 dl_filename = dl_filename or temp_filename
3465 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
3466
3467 except network_exceptions as err:
3468 self.report_error(f'unable to download video data: {err}')
3469 return
3470 except OSError as err:
3471 raise UnavailableVideoError(err)
3472 except ContentTooShortError as err:
3473 self.report_error(f'content too short (expected {err.expected} bytes and served {err.downloaded})')
3474 return
3475
3476 self._raise_pending_errors(info_dict)
3477 if success and full_filename != '-':
3478
3479 def fixup():
3480 do_fixup = True
3481 fixup_policy = self.params.get('fixup')
3482 vid = info_dict['id']
3483
3484 if fixup_policy in ('ignore', 'never'):
3485 return
3486 elif fixup_policy == 'warn':
3487 do_fixup = 'warn'
3488 elif fixup_policy != 'force':
3489 assert fixup_policy in ('detect_or_warn', None)
3490 if not info_dict.get('__real_download'):
3491 do_fixup = False
3492
3493 def ffmpeg_fixup(cndn, msg, cls):
3494 if not (do_fixup and cndn):
3495 return
3496 elif do_fixup == 'warn':
3497 self.report_warning(f'{vid}: {msg}')
3498 return
3499 pp = cls(self)
3500 if pp.available:
3501 info_dict['__postprocessors'].append(pp)
3502 else:
3503 self.report_warning(f'{vid}: {msg}. Install ffmpeg to fix this automatically')
3504
3505 stretched_ratio = info_dict.get('stretched_ratio')
3506 ffmpeg_fixup(stretched_ratio not in (1, None),
3507 f'Non-uniform pixel ratio {stretched_ratio}',
3508 FFmpegFixupStretchedPP)
3509
3510 downloader = get_suitable_downloader(info_dict, self.params) if 'protocol' in info_dict else None
3511 downloader = downloader.FD_NAME if downloader else None
3512
3513 ext = info_dict.get('ext')
3514 postprocessed_by_ffmpeg = info_dict.get('requested_formats') or any((
3515 isinstance(pp, FFmpegVideoConvertorPP)
3516 and resolve_recode_mapping(ext, pp.mapping)[0] not in (ext, None)
3517 ) for pp in self._pps['post_process'])
3518
3519 if not postprocessed_by_ffmpeg:
3520 ffmpeg_fixup(fd != FFmpegFD and ext == 'm4a'
3521 and info_dict.get('container') == 'm4a_dash',
3522 'writing DASH m4a. Only some players support this container',
3523 FFmpegFixupM4aPP)
3524 ffmpeg_fixup(downloader == 'hlsnative' and not self.params.get('hls_use_mpegts')
3525 or info_dict.get('is_live') and self.params.get('hls_use_mpegts') is None,
3526 'Possible MPEG-TS in MP4 container or malformed AAC timestamps',
3527 FFmpegFixupM3u8PP)
3528 ffmpeg_fixup(downloader == 'dashsegments'
3529 and (info_dict.get('is_live') or info_dict.get('is_dash_periods')),
3530 'Possible duplicate MOOV atoms', FFmpegFixupDuplicateMoovPP)
3531
3532 ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed timestamps detected', FFmpegFixupTimestampPP)
3533 ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed duration detected', FFmpegFixupDurationPP)
3534
3535 fixup()
3536 try:
3537 replace_info_dict(self.post_process(dl_filename, info_dict, files_to_move))
3538 except PostProcessingError as err:
3539 self.report_error(f'Postprocessing: {err}')
3540 return
3541 try:
3542 for ph in self._post_hooks:
3543 ph(info_dict['filepath'])
3544 except Exception as err:
3545 self.report_error(f'post hooks: {err}')
3546 return
3547 info_dict['__write_download_archive'] = True
3548
3549 assert info_dict is original_infodict # Make sure the info_dict was modified in-place
3550 if self.params.get('force_write_download_archive'):
3551 info_dict['__write_download_archive'] = True
3552 check_max_downloads()
3553
3554 def __download_wrapper(self, func):
3555 @functools.wraps(func)
3556 def wrapper(*args, **kwargs):
3557 try:
3558 res = func(*args, **kwargs)
3559 except UnavailableVideoError as e:
3560 self.report_error(e)
3561 except DownloadCancelled as e:
3562 self.to_screen(f'[info] {e}')
3563 if not self.params.get('break_per_url'):
3564 raise
3565 self._num_downloads = 0
3566 else:
3567 if self.params.get('dump_single_json', False):
3568 self.post_extract(res)
3569 self.to_stdout(json.dumps(self.sanitize_info(res)))
3570 return wrapper
3571
3572 def download(self, url_list):
3573 """Download a given list of URLs."""
3574 url_list = variadic(url_list) # Passing a single URL is a common mistake
3575 outtmpl = self.params['outtmpl']['default']
3576 if (len(url_list) > 1
3577 and outtmpl != '-'
3578 and '%' not in outtmpl
3579 and self.params.get('max_downloads') != 1):
3580 raise SameFileError(outtmpl)
3581
3582 for url in url_list:
3583 self.__download_wrapper(self.extract_info)(
3584 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
3585
3586 return self._download_retcode
3587
3588 def download_with_info_file(self, info_filename):
3589 with contextlib.closing(fileinput.FileInput(
3590 [info_filename], mode='r',
3591 openhook=fileinput.hook_encoded('utf-8'))) as f:
3592 # FileInput doesn't have a read method, we can't call json.load
3593 infos = [self.sanitize_info(info, self.params.get('clean_infojson', True))
3594 for info in variadic(json.loads('\n'.join(f)))]
3595 for info in infos:
3596 try:
3597 self.__download_wrapper(self.process_ie_result)(info, download=True)
3598 except (DownloadError, EntryNotInPlaylist, ReExtractInfo) as e:
3599 if not isinstance(e, EntryNotInPlaylist):
3600 self.to_stderr('\r')
3601 webpage_url = info.get('webpage_url')
3602 if webpage_url is None:
3603 raise
3604 self.report_warning(f'The info failed to download: {e}; trying with URL {webpage_url}')
3605 self.download([webpage_url])
3606 except ExtractorError as e:
3607 self.report_error(e)
3608 return self._download_retcode
3609
3610 @staticmethod
3611 def sanitize_info(info_dict, remove_private_keys=False):
3612 """ Sanitize the infodict for converting to json """
3613 if info_dict is None:
3614 return info_dict
3615 info_dict.setdefault('epoch', int(time.time()))
3616 info_dict.setdefault('_type', 'video')
3617 info_dict.setdefault('_version', {
3618 'version': __version__,
3619 'current_git_head': current_git_head(),
3620 'release_git_head': RELEASE_GIT_HEAD,
3621 'repository': ORIGIN,
3622 })
3623
3624 if remove_private_keys:
3625 reject = lambda k, v: v is None or k.startswith('__') or k in {
3626 'requested_downloads', 'requested_formats', 'requested_subtitles', 'requested_entries',
3627 'entries', 'filepath', '_filename', 'filename', 'infojson_filename', 'original_url',
3628 'playlist_autonumber',
3629 }
3630 else:
3631 reject = lambda k, v: False
3632
3633 def filter_fn(obj):
3634 if isinstance(obj, dict):
3635 return {k: filter_fn(v) for k, v in obj.items() if not reject(k, v)}
3636 elif isinstance(obj, (list, tuple, set, LazyList)):
3637 return list(map(filter_fn, obj))
3638 elif obj is None or isinstance(obj, (str, int, float, bool)):
3639 return obj
3640 else:
3641 return repr(obj)
3642
3643 return filter_fn(info_dict)
3644
3645 @staticmethod
3646 def filter_requested_info(info_dict, actually_filter=True):
3647 """ Alias of sanitize_info for backward compatibility """
3648 return YoutubeDL.sanitize_info(info_dict, actually_filter)
3649
3650 def _delete_downloaded_files(self, *files_to_delete, info={}, msg=None):
3651 for filename in set(filter(None, files_to_delete)):
3652 if msg:
3653 self.to_screen(msg % filename)
3654 try:
3655 os.remove(filename)
3656 except OSError:
3657 self.report_warning(f'Unable to delete file {filename}')
3658 if filename in info.get('__files_to_move', []): # NB: Delete even if None
3659 del info['__files_to_move'][filename]
3660
3661 @staticmethod
3662 def post_extract(info_dict):
3663 def actual_post_extract(info_dict):
3664 if info_dict.get('_type') in ('playlist', 'multi_video'):
3665 for video_dict in info_dict.get('entries', {}):
3666 actual_post_extract(video_dict or {})
3667 return
3668
3669 post_extractor = info_dict.pop('__post_extractor', None) or dict
3670 info_dict.update(post_extractor())
3671
3672 actual_post_extract(info_dict or {})
3673
3674 def run_pp(self, pp, infodict):
3675 files_to_delete = []
3676 if '__files_to_move' not in infodict:
3677 infodict['__files_to_move'] = {}
3678 try:
3679 files_to_delete, infodict = pp.run(infodict)
3680 except PostProcessingError as e:
3681 # Must be True and not 'only_download'
3682 if self.params.get('ignoreerrors') is True:
3683 self.report_error(e)
3684 return infodict
3685 raise
3686
3687 if not files_to_delete:
3688 return infodict
3689 if self.params.get('keepvideo', False):
3690 for f in files_to_delete:
3691 infodict['__files_to_move'].setdefault(f, '')
3692 else:
3693 self._delete_downloaded_files(
3694 *files_to_delete, info=infodict, msg='Deleting original file %s (pass -k to keep)')
3695 return infodict
3696
3697 def run_all_pps(self, key, info, *, additional_pps=None):
3698 if key != 'video':
3699 self._forceprint(key, info)
3700 for pp in (additional_pps or []) + self._pps[key]:
3701 info = self.run_pp(pp, info)
3702 return info
3703
3704 def pre_process(self, ie_info, key='pre_process', files_to_move=None):
3705 info = dict(ie_info)
3706 info['__files_to_move'] = files_to_move or {}
3707 try:
3708 info = self.run_all_pps(key, info)
3709 except PostProcessingError as err:
3710 msg = f'Preprocessing: {err}'
3711 info.setdefault('__pending_error', msg)
3712 self.report_error(msg, is_error=False)
3713 return info, info.pop('__files_to_move', None)
3714
3715 def post_process(self, filename, info, files_to_move=None):
3716 """Run all the postprocessors on the given file."""
3717 info['filepath'] = filename
3718 info['__files_to_move'] = files_to_move or {}
3719 info = self.run_all_pps('post_process', info, additional_pps=info.get('__postprocessors'))
3720 info = self.run_pp(MoveFilesAfterDownloadPP(self), info)
3721 del info['__files_to_move']
3722 return self.run_all_pps('after_move', info)
3723
3724 def _make_archive_id(self, info_dict):
3725 video_id = info_dict.get('id')
3726 if not video_id:
3727 return
3728 # Future-proof against any change in case
3729 # and backwards compatibility with prior versions
3730 extractor = info_dict.get('extractor_key') or info_dict.get('ie_key') # key in a playlist
3731 if extractor is None:
3732 url = str_or_none(info_dict.get('url'))
3733 if not url:
3734 return
3735 # Try to find matching extractor for the URL and take its ie_key
3736 for ie_key, ie in self._ies.items():
3737 if ie.suitable(url):
3738 extractor = ie_key
3739 break
3740 else:
3741 return
3742 return make_archive_id(extractor, video_id)
3743
3744 def in_download_archive(self, info_dict):
3745 if not self.archive:
3746 return False
3747
3748 vid_ids = [self._make_archive_id(info_dict)]
3749 vid_ids.extend(info_dict.get('_old_archive_ids') or [])
3750 return any(id_ in self.archive for id_ in vid_ids)
3751
3752 def record_download_archive(self, info_dict):
3753 fn = self.params.get('download_archive')
3754 if fn is None:
3755 return
3756 vid_id = self._make_archive_id(info_dict)
3757 assert vid_id
3758
3759 self.write_debug(f'Adding to archive: {vid_id}')
3760 if is_path_like(fn):
3761 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
3762 archive_file.write(vid_id + '\n')
3763 self.archive.add(vid_id)
3764
3765 @staticmethod
3766 def format_resolution(format, default='unknown'):
3767 if format.get('vcodec') == 'none' and format.get('acodec') != 'none':
3768 return 'audio only'
3769 if format.get('resolution') is not None:
3770 return format['resolution']
3771 if format.get('width') and format.get('height'):
3772 return '%dx%d' % (format['width'], format['height'])
3773 elif format.get('height'):
3774 return '{}p'.format(format['height'])
3775 elif format.get('width'):
3776 return '%dx?' % format['width']
3777 return default
3778
3779 def _list_format_headers(self, *headers):
3780 if self.params.get('listformats_table', True) is not False:
3781 return [self._format_out(header, self.Styles.HEADERS) for header in headers]
3782 return headers
3783
3784 def _format_note(self, fdict):
3785 res = ''
3786 if fdict.get('ext') in ['f4f', 'f4m']:
3787 res += '(unsupported)'
3788 if fdict.get('language'):
3789 if res:
3790 res += ' '
3791 res += '[{}]'.format(fdict['language'])
3792 if fdict.get('format_note') is not None:
3793 if res:
3794 res += ' '
3795 res += fdict['format_note']
3796 if fdict.get('tbr') is not None:
3797 if res:
3798 res += ', '
3799 res += '%4dk' % fdict['tbr']
3800 if fdict.get('container') is not None:
3801 if res:
3802 res += ', '
3803 res += '{} container'.format(fdict['container'])
3804 if (fdict.get('vcodec') is not None
3805 and fdict.get('vcodec') != 'none'):
3806 if res:
3807 res += ', '
3808 res += fdict['vcodec']
3809 if fdict.get('vbr') is not None:
3810 res += '@'
3811 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
3812 res += 'video@'
3813 if fdict.get('vbr') is not None:
3814 res += '%4dk' % fdict['vbr']
3815 if fdict.get('fps') is not None:
3816 if res:
3817 res += ', '
3818 res += '{}fps'.format(fdict['fps'])
3819 if fdict.get('acodec') is not None:
3820 if res:
3821 res += ', '
3822 if fdict['acodec'] == 'none':
3823 res += 'video only'
3824 else:
3825 res += '%-5s' % fdict['acodec']
3826 elif fdict.get('abr') is not None:
3827 if res:
3828 res += ', '
3829 res += 'audio'
3830 if fdict.get('abr') is not None:
3831 res += '@%3dk' % fdict['abr']
3832 if fdict.get('asr') is not None:
3833 res += ' (%5dHz)' % fdict['asr']
3834 if fdict.get('filesize') is not None:
3835 if res:
3836 res += ', '
3837 res += format_bytes(fdict['filesize'])
3838 elif fdict.get('filesize_approx') is not None:
3839 if res:
3840 res += ', '
3841 res += '~' + format_bytes(fdict['filesize_approx'])
3842 return res
3843
3844 def _get_formats(self, info_dict):
3845 if info_dict.get('formats') is None:
3846 if info_dict.get('url') and info_dict.get('_type', 'video') == 'video':
3847 return [info_dict]
3848 return []
3849 return info_dict['formats']
3850
3851 def render_formats_table(self, info_dict):
3852 formats = self._get_formats(info_dict)
3853 if not formats:
3854 return
3855 if not self.params.get('listformats_table', True) is not False:
3856 table = [
3857 [
3858 format_field(f, 'format_id'),
3859 format_field(f, 'ext'),
3860 self.format_resolution(f),
3861 self._format_note(f),
3862 ] for f in formats if (f.get('preference') or 0) >= -1000]
3863 return render_table(['format code', 'extension', 'resolution', 'note'], table, extra_gap=1)
3864
3865 def simplified_codec(f, field):
3866 assert field in ('acodec', 'vcodec')
3867 codec = f.get(field)
3868 if not codec:
3869 return 'unknown'
3870 elif codec != 'none':
3871 return '.'.join(codec.split('.')[:4])
3872
3873 if field == 'vcodec' and f.get('acodec') == 'none':
3874 return 'images'
3875 elif field == 'acodec' and f.get('vcodec') == 'none':
3876 return ''
3877 return self._format_out('audio only' if field == 'vcodec' else 'video only',
3878 self.Styles.SUPPRESS)
3879
3880 delim = self._format_out('\u2502', self.Styles.DELIM, '|', test_encoding=True)
3881 table = [
3882 [
3883 self._format_out(format_field(f, 'format_id'), self.Styles.ID),
3884 format_field(f, 'ext'),
3885 format_field(f, func=self.format_resolution, ignore=('audio only', 'images')),
3886 format_field(f, 'fps', '\t%d', func=round),
3887 format_field(f, 'dynamic_range', '%s', ignore=(None, 'SDR')).replace('HDR', ''),
3888 format_field(f, 'audio_channels', '\t%s'),
3889 delim, (
3890 format_field(f, 'filesize', ' \t%s', func=format_bytes)
3891 or format_field(f, 'filesize_approx', '≈\t%s', func=format_bytes)
3892 or format_field(filesize_from_tbr(f.get('tbr'), info_dict.get('duration')), None,
3893 self._format_out('~\t%s', self.Styles.SUPPRESS), func=format_bytes)),
3894 format_field(f, 'tbr', '\t%dk', func=round),
3895 shorten_protocol_name(f.get('protocol', '')),
3896 delim,
3897 simplified_codec(f, 'vcodec'),
3898 format_field(f, 'vbr', '\t%dk', func=round),
3899 simplified_codec(f, 'acodec'),
3900 format_field(f, 'abr', '\t%dk', func=round),
3901 format_field(f, 'asr', '\t%s', func=format_decimal_suffix),
3902 join_nonempty(format_field(f, 'language', '[%s]'), join_nonempty(
3903 self._format_out('UNSUPPORTED', self.Styles.BAD_FORMAT) if f.get('ext') in ('f4f', 'f4m') else None,
3904 (self._format_out('Maybe DRM', self.Styles.WARNING) if f.get('has_drm') == 'maybe'
3905 else self._format_out('DRM', self.Styles.BAD_FORMAT) if f.get('has_drm') else None),
3906 format_field(f, 'format_note'),
3907 format_field(f, 'container', ignore=(None, f.get('ext'))),
3908 delim=', '), delim=' '),
3909 ] for f in formats if f.get('preference') is None or f['preference'] >= -1000]
3910 header_line = self._list_format_headers(
3911 'ID', 'EXT', 'RESOLUTION', '\tFPS', 'HDR', 'CH', delim, '\tFILESIZE', '\tTBR', 'PROTO',
3912 delim, 'VCODEC', '\tVBR', 'ACODEC', '\tABR', '\tASR', 'MORE INFO')
3913
3914 return render_table(
3915 header_line, table, hide_empty=True,
3916 delim=self._format_out('\u2500', self.Styles.DELIM, '-', test_encoding=True))
3917
3918 def render_thumbnails_table(self, info_dict):
3919 thumbnails = list(info_dict.get('thumbnails') or [])
3920 if not thumbnails:
3921 return None
3922 return render_table(
3923 self._list_format_headers('ID', 'Width', 'Height', 'URL'),
3924 [[t.get('id'), t.get('width') or 'unknown', t.get('height') or 'unknown', t['url']] for t in thumbnails])
3925
3926 def render_subtitles_table(self, video_id, subtitles):
3927 def _row(lang, formats):
3928 exts, names = zip(*((f['ext'], f.get('name') or 'unknown') for f in reversed(formats)))
3929 if len(set(names)) == 1:
3930 names = [] if names[0] == 'unknown' else names[:1]
3931 return [lang, ', '.join(names), ', '.join(exts)]
3932
3933 if not subtitles:
3934 return None
3935 return render_table(
3936 self._list_format_headers('Language', 'Name', 'Formats'),
3937 [_row(lang, formats) for lang, formats in subtitles.items()],
3938 hide_empty=True)
3939
3940 def __list_table(self, video_id, name, func, *args):
3941 table = func(*args)
3942 if not table:
3943 self.to_screen(f'{video_id} has no {name}')
3944 return
3945 self.to_screen(f'[info] Available {name} for {video_id}:')
3946 self.to_stdout(table)
3947
3948 def list_formats(self, info_dict):
3949 self.__list_table(info_dict['id'], 'formats', self.render_formats_table, info_dict)
3950
3951 def list_thumbnails(self, info_dict):
3952 self.__list_table(info_dict['id'], 'thumbnails', self.render_thumbnails_table, info_dict)
3953
3954 def list_subtitles(self, video_id, subtitles, name='subtitles'):
3955 self.__list_table(video_id, name, self.render_subtitles_table, video_id, subtitles)
3956
3957 def print_debug_header(self):
3958 if not self.params.get('verbose'):
3959 return
3960
3961 from . import _IN_CLI # Must be delayed import
3962
3963 # These imports can be slow. So import them only as needed
3964 from .extractor.extractors import _LAZY_LOADER
3965 from .extractor.extractors import (
3966 _PLUGIN_CLASSES as plugin_ies,
3967 _PLUGIN_OVERRIDES as plugin_ie_overrides,
3968 )
3969
3970 def get_encoding(stream):
3971 ret = str(getattr(stream, 'encoding', f'missing ({type(stream).__name__})'))
3972 additional_info = []
3973 if os.environ.get('TERM', '').lower() == 'dumb':
3974 additional_info.append('dumb')
3975 if not supports_terminal_sequences(stream):
3976 from .utils import WINDOWS_VT_MODE # Must be imported locally
3977 additional_info.append('No VT' if WINDOWS_VT_MODE is False else 'No ANSI')
3978 if additional_info:
3979 ret = f'{ret} ({",".join(additional_info)})'
3980 return ret
3981
3982 encoding_str = 'Encodings: locale {}, fs {}, pref {}, {}'.format(
3983 locale.getpreferredencoding(),
3984 sys.getfilesystemencoding(),
3985 self.get_encoding(),
3986 ', '.join(
3987 f'{key} {get_encoding(stream)}' for key, stream in self._out_files.items_
3988 if stream is not None and key != 'console'),
3989 )
3990
3991 logger = self.params.get('logger')
3992 if logger:
3993 write_debug = lambda msg: logger.debug(f'[debug] {msg}')
3994 write_debug(encoding_str)
3995 else:
3996 write_string(f'[debug] {encoding_str}\n', encoding=None)
3997 write_debug = lambda msg: self._write_string(f'[debug] {msg}\n')
3998
3999 source = detect_variant()
4000 if VARIANT not in (None, 'pip'):
4001 source += '*'
4002 klass = type(self)
4003 write_debug(join_nonempty(
4004 f'{REPOSITORY.rpartition("/")[2]} version',
4005 _make_label(ORIGIN, CHANNEL.partition('@')[2] or __version__, __version__),
4006 f'[{RELEASE_GIT_HEAD[:9]}]' if RELEASE_GIT_HEAD else '',
4007 '' if source == 'unknown' else f'({source})',
4008 '' if _IN_CLI else 'API' if klass == YoutubeDL else f'API:{self.__module__}.{klass.__qualname__}',
4009 delim=' '))
4010
4011 if not _IN_CLI:
4012 write_debug(f'params: {self.params}')
4013
4014 if not _LAZY_LOADER:
4015 if os.environ.get('YTDLP_NO_LAZY_EXTRACTORS'):
4016 write_debug('Lazy loading extractors is forcibly disabled')
4017 else:
4018 write_debug('Lazy loading extractors is disabled')
4019 if self.params['compat_opts']:
4020 write_debug('Compatibility options: {}'.format(', '.join(self.params['compat_opts'])))
4021
4022 if current_git_head():
4023 write_debug(f'Git HEAD: {current_git_head()}')
4024 write_debug(system_identifier())
4025
4026 exe_versions, ffmpeg_features = FFmpegPostProcessor.get_versions_and_features(self)
4027 ffmpeg_features = {key for key, val in ffmpeg_features.items() if val}
4028 if ffmpeg_features:
4029 exe_versions['ffmpeg'] += ' ({})'.format(','.join(sorted(ffmpeg_features)))
4030
4031 exe_versions['rtmpdump'] = rtmpdump_version()
4032 exe_versions['phantomjs'] = PhantomJSwrapper._version()
4033 exe_str = ', '.join(
4034 f'{exe} {v}' for exe, v in sorted(exe_versions.items()) if v
4035 ) or 'none'
4036 write_debug(f'exe versions: {exe_str}')
4037
4038 from .compat.compat_utils import get_package_info
4039 from .dependencies import available_dependencies
4040
4041 write_debug('Optional libraries: %s' % (', '.join(sorted({
4042 join_nonempty(*get_package_info(m)) for m in available_dependencies.values()
4043 })) or 'none'))
4044
4045 write_debug(f'Proxy map: {self.proxies}')
4046 write_debug(f'Request Handlers: {", ".join(rh.RH_NAME for rh in self._request_director.handlers.values())}')
4047 for plugin_type, plugins in {'Extractor': plugin_ies, 'Post-Processor': plugin_pps}.items():
4048 display_list = ['{}{}'.format(
4049 klass.__name__, '' if klass.__name__ == name else f' as {name}')
4050 for name, klass in plugins.items()]
4051 if plugin_type == 'Extractor':
4052 display_list.extend(f'{plugins[-1].IE_NAME.partition("+")[2]} ({parent.__name__})'
4053 for parent, plugins in plugin_ie_overrides.items())
4054 if not display_list:
4055 continue
4056 write_debug(f'{plugin_type} Plugins: {", ".join(sorted(display_list))}')
4057
4058 plugin_dirs = plugin_directories()
4059 if plugin_dirs:
4060 write_debug(f'Plugin directories: {plugin_dirs}')
4061
4062 # Not implemented
4063 if False and self.params.get('call_home'):
4064 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode()
4065 write_debug(f'Public IP address: {ipaddr}')
4066 latest_version = self.urlopen(
4067 'https://yt-dl.org/latest/version').read().decode()
4068 if version_tuple(latest_version) > version_tuple(__version__):
4069 self.report_warning(
4070 f'You are using an outdated version (newest version: {latest_version})! '
4071 'See https://yt-dl.org/update if you need help updating.')
4072
4073 @functools.cached_property
4074 def proxies(self):
4075 """Global proxy configuration"""
4076 opts_proxy = self.params.get('proxy')
4077 if opts_proxy is not None:
4078 if opts_proxy == '':
4079 opts_proxy = '__noproxy__'
4080 proxies = {'all': opts_proxy}
4081 else:
4082 proxies = urllib.request.getproxies()
4083 # compat. Set HTTPS_PROXY to __noproxy__ to revert
4084 if 'http' in proxies and 'https' not in proxies:
4085 proxies['https'] = proxies['http']
4086
4087 return proxies
4088
4089 @functools.cached_property
4090 def cookiejar(self):
4091 """Global cookiejar instance"""
4092 return load_cookies(
4093 self.params.get('cookiefile'), self.params.get('cookiesfrombrowser'), self)
4094
4095 @property
4096 def _opener(self):
4097 """
4098 Get a urllib OpenerDirector from the Urllib handler (deprecated).
4099 """
4100 self.deprecation_warning('YoutubeDL._opener is deprecated, use YoutubeDL.urlopen()')
4101 handler = self._request_director.handlers['Urllib']
4102 return handler._get_instance(cookiejar=self.cookiejar, proxies=self.proxies)
4103
4104 def _get_available_impersonate_targets(self):
4105 # TODO(future): make available as public API
4106 return [
4107 (target, rh.RH_NAME)
4108 for rh in self._request_director.handlers.values()
4109 if isinstance(rh, ImpersonateRequestHandler)
4110 for target in rh.supported_targets
4111 ]
4112
4113 def _impersonate_target_available(self, target):
4114 # TODO(future): make available as public API
4115 return any(
4116 rh.is_supported_target(target)
4117 for rh in self._request_director.handlers.values()
4118 if isinstance(rh, ImpersonateRequestHandler))
4119
4120 def urlopen(self, req):
4121 """ Start an HTTP download """
4122 if isinstance(req, str):
4123 req = Request(req)
4124 elif isinstance(req, urllib.request.Request):
4125 self.deprecation_warning(
4126 'Passing a urllib.request.Request object to YoutubeDL.urlopen() is deprecated. '
4127 'Use yt_dlp.networking.common.Request instead.')
4128 req = urllib_req_to_req(req)
4129 assert isinstance(req, Request)
4130
4131 # compat: Assume user:pass url params are basic auth
4132 url, basic_auth_header = extract_basic_auth(req.url)
4133 if basic_auth_header:
4134 req.headers['Authorization'] = basic_auth_header
4135 req.url = sanitize_url(url)
4136
4137 clean_proxies(proxies=req.proxies, headers=req.headers)
4138 clean_headers(req.headers)
4139
4140 try:
4141 return self._request_director.send(req)
4142 except NoSupportingHandlers as e:
4143 for ue in e.unsupported_errors:
4144 # FIXME: This depends on the order of errors.
4145 if not (ue.handler and ue.msg):
4146 continue
4147 if ue.handler.RH_KEY == 'Urllib' and 'unsupported url scheme: "file"' in ue.msg.lower():
4148 raise RequestError(
4149 'file:// URLs are disabled by default in yt-dlp for security reasons. '
4150 'Use --enable-file-urls to enable at your own risk.', cause=ue) from ue
4151 if (
4152 'unsupported proxy type: "https"' in ue.msg.lower()
4153 and 'requests' not in self._request_director.handlers
4154 and 'curl_cffi' not in self._request_director.handlers
4155 ):
4156 raise RequestError(
4157 'To use an HTTPS proxy for this request, one of the following dependencies needs to be installed: requests, curl_cffi')
4158
4159 elif (
4160 re.match(r'unsupported url scheme: "wss?"', ue.msg.lower())
4161 and 'websockets' not in self._request_director.handlers
4162 ):
4163 raise RequestError(
4164 'This request requires WebSocket support. '
4165 'Ensure one of the following dependencies are installed: websockets',
4166 cause=ue) from ue
4167
4168 elif re.match(r'unsupported (?:extensions: impersonate|impersonate target)', ue.msg.lower()):
4169 raise RequestError(
4170 f'Impersonate target "{req.extensions["impersonate"]}" is not available.'
4171 f' See --list-impersonate-targets for available targets.'
4172 f' This request requires browser impersonation, however you may be missing dependencies'
4173 f' required to support this target.')
4174 raise
4175 except SSLError as e:
4176 if 'UNSAFE_LEGACY_RENEGOTIATION_DISABLED' in str(e):
4177 raise RequestError('UNSAFE_LEGACY_RENEGOTIATION_DISABLED: Try using --legacy-server-connect', cause=e) from e
4178 elif 'SSLV3_ALERT_HANDSHAKE_FAILURE' in str(e):
4179 raise RequestError(
4180 'SSLV3_ALERT_HANDSHAKE_FAILURE: The server may not support the current cipher list. '
4181 'Try using --legacy-server-connect', cause=e) from e
4182 raise
4183
4184 def build_request_director(self, handlers, preferences=None):
4185 logger = _YDLLogger(self)
4186 headers = self.params['http_headers'].copy()
4187 proxies = self.proxies.copy()
4188 clean_headers(headers)
4189 clean_proxies(proxies, headers)
4190
4191 director = RequestDirector(logger=logger, verbose=self.params.get('debug_printtraffic'))
4192 for handler in handlers:
4193 director.add_handler(handler(
4194 logger=logger,
4195 headers=headers,
4196 cookiejar=self.cookiejar,
4197 proxies=proxies,
4198 prefer_system_certs='no-certifi' in self.params['compat_opts'],
4199 verify=not self.params.get('nocheckcertificate'),
4200 **traverse_obj(self.params, {
4201 'verbose': 'debug_printtraffic',
4202 'source_address': 'source_address',
4203 'timeout': 'socket_timeout',
4204 'legacy_ssl_support': 'legacyserverconnect',
4205 'enable_file_urls': 'enable_file_urls',
4206 'impersonate': 'impersonate',
4207 'client_cert': {
4208 'client_certificate': 'client_certificate',
4209 'client_certificate_key': 'client_certificate_key',
4210 'client_certificate_password': 'client_certificate_password',
4211 },
4212 }),
4213 ))
4214 director.preferences.update(preferences or [])
4215 if 'prefer-legacy-http-handler' in self.params['compat_opts']:
4216 director.preferences.add(lambda rh, _: 500 if rh.RH_KEY == 'Urllib' else 0)
4217 return director
4218
4219 @functools.cached_property
4220 def _request_director(self):
4221 return self.build_request_director(_REQUEST_HANDLERS.values(), _RH_PREFERENCES)
4222
4223 def encode(self, s):
4224 if isinstance(s, bytes):
4225 return s # Already encoded
4226
4227 try:
4228 return s.encode(self.get_encoding())
4229 except UnicodeEncodeError as err:
4230 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
4231 raise
4232
4233 def get_encoding(self):
4234 encoding = self.params.get('encoding')
4235 if encoding is None:
4236 encoding = preferredencoding()
4237 return encoding
4238
4239 def _write_info_json(self, label, ie_result, infofn, overwrite=None):
4240 """ Write infojson and returns True = written, 'exists' = Already exists, False = skip, None = error """
4241 if overwrite is None:
4242 overwrite = self.params.get('overwrites', True)
4243 if not self.params.get('writeinfojson'):
4244 return False
4245 elif not infofn:
4246 self.write_debug(f'Skipping writing {label} infojson')
4247 return False
4248 elif not self._ensure_dir_exists(infofn):
4249 return None
4250 elif not overwrite and os.path.exists(infofn):
4251 self.to_screen(f'[info] {label.title()} metadata is already present')
4252 return 'exists'
4253
4254 self.to_screen(f'[info] Writing {label} metadata as JSON to: {infofn}')
4255 try:
4256 write_json_file(self.sanitize_info(ie_result, self.params.get('clean_infojson', True)), infofn)
4257 return True
4258 except OSError:
4259 self.report_error(f'Cannot write {label} metadata to JSON file {infofn}')
4260 return None
4261
4262 def _write_description(self, label, ie_result, descfn):
4263 """ Write description and returns True = written, False = skip, None = error """
4264 if not self.params.get('writedescription'):
4265 return False
4266 elif not descfn:
4267 self.write_debug(f'Skipping writing {label} description')
4268 return False
4269 elif not self._ensure_dir_exists(descfn):
4270 return None
4271 elif not self.params.get('overwrites', True) and os.path.exists(descfn):
4272 self.to_screen(f'[info] {label.title()} description is already present')
4273 elif ie_result.get('description') is None:
4274 self.to_screen(f'[info] There\'s no {label} description to write')
4275 return False
4276 else:
4277 try:
4278 self.to_screen(f'[info] Writing {label} description to: {descfn}')
4279 with open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
4280 descfile.write(ie_result['description'])
4281 except OSError:
4282 self.report_error(f'Cannot write {label} description file {descfn}')
4283 return None
4284 return True
4285
4286 def _write_subtitles(self, info_dict, filename):
4287 """ Write subtitles to file and return list of (sub_filename, final_sub_filename); or None if error"""
4288 ret = []
4289 subtitles = info_dict.get('requested_subtitles')
4290 if not (self.params.get('writesubtitles') or self.params.get('writeautomaticsub')):
4291 # subtitles download errors are already managed as troubles in relevant IE
4292 # that way it will silently go on when used with unsupporting IE
4293 return ret
4294 elif not subtitles:
4295 self.to_screen('[info] There are no subtitles for the requested languages')
4296 return ret
4297 sub_filename_base = self.prepare_filename(info_dict, 'subtitle')
4298 if not sub_filename_base:
4299 self.to_screen('[info] Skipping writing video subtitles')
4300 return ret
4301
4302 for sub_lang, sub_info in subtitles.items():
4303 sub_format = sub_info['ext']
4304 sub_filename = subtitles_filename(filename, sub_lang, sub_format, info_dict.get('ext'))
4305 sub_filename_final = subtitles_filename(sub_filename_base, sub_lang, sub_format, info_dict.get('ext'))
4306 existing_sub = self.existing_file((sub_filename_final, sub_filename))
4307 if existing_sub:
4308 self.to_screen(f'[info] Video subtitle {sub_lang}.{sub_format} is already present')
4309 sub_info['filepath'] = existing_sub
4310 ret.append((existing_sub, sub_filename_final))
4311 continue
4312
4313 self.to_screen(f'[info] Writing video subtitles to: {sub_filename}')
4314 if sub_info.get('data') is not None:
4315 try:
4316 # Use newline='' to prevent conversion of newline characters
4317 # See https://github.com/ytdl-org/youtube-dl/issues/10268
4318 with open(sub_filename, 'w', encoding='utf-8', newline='') as subfile:
4319 subfile.write(sub_info['data'])
4320 sub_info['filepath'] = sub_filename
4321 ret.append((sub_filename, sub_filename_final))
4322 continue
4323 except OSError:
4324 self.report_error(f'Cannot write video subtitles file {sub_filename}')
4325 return None
4326
4327 try:
4328 sub_copy = sub_info.copy()
4329 sub_copy.setdefault('http_headers', info_dict.get('http_headers'))
4330 self.dl(sub_filename, sub_copy, subtitle=True)
4331 sub_info['filepath'] = sub_filename
4332 ret.append((sub_filename, sub_filename_final))
4333 except (DownloadError, ExtractorError, OSError, ValueError, *network_exceptions) as err:
4334 msg = f'Unable to download video subtitles for {sub_lang!r}: {err}'
4335 if self.params.get('ignoreerrors') is not True: # False or 'only_download'
4336 if not self.params.get('ignoreerrors'):
4337 self.report_error(msg)
4338 raise DownloadError(msg)
4339 self.report_warning(msg)
4340 return ret
4341
4342 def _write_thumbnails(self, label, info_dict, filename, thumb_filename_base=None):
4343 """ Write thumbnails to file and return list of (thumb_filename, final_thumb_filename); or None if error """
4344 write_all = self.params.get('write_all_thumbnails', False)
4345 thumbnails, ret = [], []
4346 if write_all or self.params.get('writethumbnail', False):
4347 thumbnails = info_dict.get('thumbnails') or []
4348 if not thumbnails:
4349 self.to_screen(f'[info] There are no {label} thumbnails to download')
4350 return ret
4351 multiple = write_all and len(thumbnails) > 1
4352
4353 if thumb_filename_base is None:
4354 thumb_filename_base = filename
4355 if thumbnails and not thumb_filename_base:
4356 self.write_debug(f'Skipping writing {label} thumbnail')
4357 return ret
4358
4359 if thumbnails and not self._ensure_dir_exists(filename):
4360 return None
4361
4362 for idx, t in list(enumerate(thumbnails))[::-1]:
4363 thumb_ext = (f'{t["id"]}.' if multiple else '') + determine_ext(t['url'], 'jpg')
4364 thumb_display_id = f'{label} thumbnail {t["id"]}'
4365 thumb_filename = replace_extension(filename, thumb_ext, info_dict.get('ext'))
4366 thumb_filename_final = replace_extension(thumb_filename_base, thumb_ext, info_dict.get('ext'))
4367
4368 existing_thumb = self.existing_file((thumb_filename_final, thumb_filename))
4369 if existing_thumb:
4370 self.to_screen('[info] {} is already present'.format((
4371 thumb_display_id if multiple else f'{label} thumbnail').capitalize()))
4372 t['filepath'] = existing_thumb
4373 ret.append((existing_thumb, thumb_filename_final))
4374 else:
4375 self.to_screen(f'[info] Downloading {thumb_display_id} ...')
4376 try:
4377 uf = self.urlopen(Request(t['url'], headers=t.get('http_headers', {})))
4378 self.to_screen(f'[info] Writing {thumb_display_id} to: {thumb_filename}')
4379 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
4380 shutil.copyfileobj(uf, thumbf)
4381 ret.append((thumb_filename, thumb_filename_final))
4382 t['filepath'] = thumb_filename
4383 except network_exceptions as err:
4384 if isinstance(err, HTTPError) and err.status == 404:
4385 self.to_screen(f'[info] {thumb_display_id.title()} does not exist')
4386 else:
4387 self.report_warning(f'Unable to download {thumb_display_id}: {err}')
4388 thumbnails.pop(idx)
4389 if ret and not write_all:
4390 break
4391 return ret