]> jfr.im git - yt-dlp.git/blob - yt_dlp/YoutubeDL.py
[ie/loom] Add extractors (#8686)
[yt-dlp.git] / yt_dlp / YoutubeDL.py
1 import collections
2 import contextlib
3 import copy
4 import datetime
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, compat_shlex_quote, 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 args_to_str,
106 bug_reports_message,
107 date_from_str,
108 deprecation_warning,
109 determine_ext,
110 determine_protocol,
111 encode_compat_str,
112 encodeFilename,
113 error_to_compat_str,
114 escapeHTML,
115 expand_path,
116 extract_basic_auth,
117 filter_dict,
118 float_or_none,
119 format_bytes,
120 format_decimal_suffix,
121 format_field,
122 formatSeconds,
123 get_compatible_ext,
124 get_domain,
125 int_or_none,
126 iri_to_uri,
127 is_path_like,
128 join_nonempty,
129 locked_file,
130 make_archive_id,
131 make_dir,
132 number_of_digits,
133 orderedSet,
134 orderedSet_from_options,
135 parse_filesize,
136 preferredencoding,
137 prepend_extension,
138 remove_terminal_sequences,
139 render_table,
140 replace_extension,
141 sanitize_filename,
142 sanitize_path,
143 sanitize_url,
144 str_or_none,
145 strftime_or_none,
146 subtitles_filename,
147 supports_terminal_sequences,
148 system_identifier,
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.
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', 'hls_aes', 'downloader_options',
585 'page_url', 'app', 'play_path', 'tc_url', 'flash_version', 'rtmp_live', 'rtmp_conn', 'rtmp_protocol', 'rtmp_real_time'
586 }
587 _deprecated_multivalue_fields = {
588 'album_artist': 'album_artists',
589 'artist': 'artists',
590 'composer': 'composers',
591 'creator': 'creators',
592 'genre': 'genres',
593 }
594 _format_selection_exts = {
595 'audio': set(MEDIA_EXTENSIONS.common_audio),
596 'video': set(MEDIA_EXTENSIONS.common_video + ('3gp', )),
597 'storyboards': set(MEDIA_EXTENSIONS.storyboards),
598 }
599
600 def __init__(self, params=None, auto_init=True):
601 """Create a FileDownloader object with the given options.
602 @param auto_init Whether to load the default extractors and print header (if verbose).
603 Set to 'no_verbose_header' to not print the header
604 """
605 if params is None:
606 params = {}
607 self.params = params
608 self._ies = {}
609 self._ies_instances = {}
610 self._pps = {k: [] for k in POSTPROCESS_WHEN}
611 self._printed_messages = set()
612 self._first_webpage_request = True
613 self._post_hooks = []
614 self._progress_hooks = []
615 self._postprocessor_hooks = []
616 self._download_retcode = 0
617 self._num_downloads = 0
618 self._num_videos = 0
619 self._playlist_level = 0
620 self._playlist_urls = set()
621 self.cache = Cache(self)
622 self.__header_cookies = []
623
624 stdout = sys.stderr if self.params.get('logtostderr') else sys.stdout
625 self._out_files = Namespace(
626 out=stdout,
627 error=sys.stderr,
628 screen=sys.stderr if self.params.get('quiet') else stdout,
629 console=None if compat_os_name == 'nt' else next(
630 filter(supports_terminal_sequences, (sys.stderr, sys.stdout)), None)
631 )
632
633 try:
634 windows_enable_vt_mode()
635 except Exception as e:
636 self.write_debug(f'Failed to enable VT mode: {e}')
637
638 if self.params.get('no_color'):
639 if self.params.get('color') is not None:
640 self.params.setdefault('_warnings', []).append(
641 'Overwriting params from "color" with "no_color"')
642 self.params['color'] = 'no_color'
643
644 term_allow_color = os.getenv('TERM', '').lower() != 'dumb'
645 no_color = bool(os.getenv('NO_COLOR'))
646
647 def process_color_policy(stream):
648 stream_name = {sys.stdout: 'stdout', sys.stderr: 'stderr'}[stream]
649 policy = traverse_obj(self.params, ('color', (stream_name, None), {str}), get_all=False)
650 if policy in ('auto', None):
651 if term_allow_color and supports_terminal_sequences(stream):
652 return 'no_color' if no_color else True
653 return False
654 assert policy in ('always', 'never', 'no_color'), policy
655 return {'always': True, 'never': False}.get(policy, policy)
656
657 self._allow_colors = Namespace(**{
658 name: process_color_policy(stream)
659 for name, stream in self._out_files.items_ if name != 'console'
660 })
661
662 system_deprecation = _get_system_deprecation()
663 if system_deprecation:
664 self.deprecated_feature(system_deprecation.replace('\n', '\n '))
665
666 if self.params.get('allow_unplayable_formats'):
667 self.report_warning(
668 f'You have asked for {self._format_err("UNPLAYABLE", self.Styles.EMPHASIS)} formats to be listed/downloaded. '
669 'This is a developer option intended for debugging. \n'
670 ' If you experience any issues while using this option, '
671 f'{self._format_err("DO NOT", self.Styles.ERROR)} open a bug report')
672
673 if self.params.get('bidi_workaround', False):
674 try:
675 import pty
676 master, slave = pty.openpty()
677 width = shutil.get_terminal_size().columns
678 width_args = [] if width is None else ['-w', str(width)]
679 sp_kwargs = {'stdin': subprocess.PIPE, 'stdout': slave, 'stderr': self._out_files.error}
680 try:
681 self._output_process = Popen(['bidiv'] + width_args, **sp_kwargs)
682 except OSError:
683 self._output_process = Popen(['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
684 self._output_channel = os.fdopen(master, 'rb')
685 except OSError as ose:
686 if ose.errno == errno.ENOENT:
687 self.report_warning(
688 'Could not find fribidi executable, ignoring --bidi-workaround. '
689 'Make sure that fribidi is an executable file in one of the directories in your $PATH.')
690 else:
691 raise
692
693 self.params['compat_opts'] = set(self.params.get('compat_opts', ()))
694 self.params['http_headers'] = HTTPHeaderDict(std_headers, self.params.get('http_headers'))
695 self._load_cookies(self.params['http_headers'].get('Cookie')) # compat
696 self.params['http_headers'].pop('Cookie', None)
697
698 if auto_init and auto_init != 'no_verbose_header':
699 self.print_debug_header()
700
701 def check_deprecated(param, option, suggestion):
702 if self.params.get(param) is not None:
703 self.report_warning(f'{option} is deprecated. Use {suggestion} instead')
704 return True
705 return False
706
707 if check_deprecated('cn_verification_proxy', '--cn-verification-proxy', '--geo-verification-proxy'):
708 if self.params.get('geo_verification_proxy') is None:
709 self.params['geo_verification_proxy'] = self.params['cn_verification_proxy']
710
711 check_deprecated('autonumber', '--auto-number', '-o "%(autonumber)s-%(title)s.%(ext)s"')
712 check_deprecated('usetitle', '--title', '-o "%(title)s-%(id)s.%(ext)s"')
713 check_deprecated('useid', '--id', '-o "%(id)s.%(ext)s"')
714
715 for msg in self.params.get('_warnings', []):
716 self.report_warning(msg)
717 for msg in self.params.get('_deprecation_warnings', []):
718 self.deprecated_feature(msg)
719
720 if impersonate_target := self.params.get('impersonate'):
721 if not self._impersonate_target_available(impersonate_target):
722 raise YoutubeDLError(
723 f'Impersonate target "{impersonate_target}" is not available. '
724 f'Use --list-impersonate-targets to see available targets. '
725 f'You may be missing dependencies required to support this target.')
726
727 if 'list-formats' in self.params['compat_opts']:
728 self.params['listformats_table'] = False
729
730 if 'overwrites' not in self.params and self.params.get('nooverwrites') is not None:
731 # nooverwrites was unnecessarily changed to overwrites
732 # in 0c3d0f51778b153f65c21906031c2e091fcfb641
733 # This ensures compatibility with both keys
734 self.params['overwrites'] = not self.params['nooverwrites']
735 elif self.params.get('overwrites') is None:
736 self.params.pop('overwrites', None)
737 else:
738 self.params['nooverwrites'] = not self.params['overwrites']
739
740 if self.params.get('simulate') is None and any((
741 self.params.get('list_thumbnails'),
742 self.params.get('listformats'),
743 self.params.get('listsubtitles'),
744 )):
745 self.params['simulate'] = 'list_only'
746
747 self.params.setdefault('forceprint', {})
748 self.params.setdefault('print_to_file', {})
749
750 # Compatibility with older syntax
751 if not isinstance(params['forceprint'], dict):
752 self.params['forceprint'] = {'video': params['forceprint']}
753
754 if auto_init:
755 self.add_default_info_extractors()
756
757 if (sys.platform != 'win32'
758 and sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968']
759 and not self.params.get('restrictfilenames', False)):
760 # Unicode filesystem API will throw errors (#1474, #13027)
761 self.report_warning(
762 'Assuming --restrict-filenames since file system encoding '
763 'cannot encode all characters. '
764 'Set the LC_ALL environment variable to fix this.')
765 self.params['restrictfilenames'] = True
766
767 self._parse_outtmpl()
768
769 # Creating format selector here allows us to catch syntax errors before the extraction
770 self.format_selector = (
771 self.params.get('format') if self.params.get('format') in (None, '-')
772 else self.params['format'] if callable(self.params['format'])
773 else self.build_format_selector(self.params['format']))
774
775 hooks = {
776 'post_hooks': self.add_post_hook,
777 'progress_hooks': self.add_progress_hook,
778 'postprocessor_hooks': self.add_postprocessor_hook,
779 }
780 for opt, fn in hooks.items():
781 for ph in self.params.get(opt, []):
782 fn(ph)
783
784 for pp_def_raw in self.params.get('postprocessors', []):
785 pp_def = dict(pp_def_raw)
786 when = pp_def.pop('when', 'post_process')
787 self.add_post_processor(
788 get_postprocessor(pp_def.pop('key'))(self, **pp_def),
789 when=when)
790
791 def preload_download_archive(fn):
792 """Preload the archive, if any is specified"""
793 archive = set()
794 if fn is None:
795 return archive
796 elif not is_path_like(fn):
797 return fn
798
799 self.write_debug(f'Loading archive file {fn!r}')
800 try:
801 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
802 for line in archive_file:
803 archive.add(line.strip())
804 except OSError as ioe:
805 if ioe.errno != errno.ENOENT:
806 raise
807 return archive
808
809 self.archive = preload_download_archive(self.params.get('download_archive'))
810
811 def warn_if_short_id(self, argv):
812 # short YouTube ID starting with dash?
813 idxs = [
814 i for i, a in enumerate(argv)
815 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
816 if idxs:
817 correct_argv = (
818 ['yt-dlp']
819 + [a for i, a in enumerate(argv) if i not in idxs]
820 + ['--'] + [argv[i] for i in idxs]
821 )
822 self.report_warning(
823 'Long argument string detected. '
824 'Use -- to separate parameters and URLs, like this:\n%s' %
825 args_to_str(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 '%s%s' % (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('[download] %s has already been downloaded' % file_name)
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('Deleting existing file %s' % 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+)'}
1211 FIELD_RE = r'\w*(?:\.(?:%(inner)s|{%(field)s(?:,%(field)s)*}))*' % {
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'(?:%s)' % '|'.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.keys() 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 = ' '.join(map(compat_shlex_quote, value)), 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%s%s' % ('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 = '%s\0%s' % (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 dateRange = self.params.get('daterange', DateRange())
1482 if date not in dateRange:
1483 return f'{date_from_str(date).isoformat()} upload date is not in range {dateRange}'
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 'Skipping "%s" because it is age restricted' % video_title
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 %s.' % ', '.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] %s: %d additional URL(s) requested' % (ie_result['id'], len(additional_urls)))
1829 self.write_debug('Additional URLs: "%s"' % '", "'.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: %s'
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 %s returned a compat_list result. '
1898 'It needs to be updated.' % 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('Invalid result type: %s' % 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('[download] Downloading item %s of %s' % (
2032 self._format_screen(i + 1, self.Styles.ID), self._format_screen(n_entries, self.Styles.EMPHASIS)))
2033
2034 entry_result = self.__process_iterable_entry(entry, download, collections.ChainMap({
2035 'playlist_index': playlist_index,
2036 'playlist_autonumber': i + 1,
2037 }, extra))
2038 if not entry_result:
2039 failures += 1
2040 if failures >= max_failures:
2041 self.report_error(
2042 f'Skipping the remaining entries in playlist "{title}" since {failures} items failed extraction')
2043 break
2044 if keep_resolved_entries:
2045 resolved_entries[i] = (playlist_index, entry_result)
2046
2047 # Update with processed data
2048 ie_result['entries'] = [e for _, e in resolved_entries if e is not NO_DEFAULT]
2049 ie_result['requested_entries'] = [i for i, e in resolved_entries if e is not NO_DEFAULT]
2050 if ie_result['requested_entries'] == try_call(lambda: list(range(1, ie_result['playlist_count'] + 1))):
2051 # Do not set for full playlist
2052 ie_result.pop('requested_entries')
2053
2054 # Write the updated info to json
2055 if _infojson_written is True and self._write_info_json(
2056 'updated playlist', ie_result,
2057 self.prepare_filename(ie_copy, 'pl_infojson'), overwrite=True) is None:
2058 return
2059
2060 ie_result = self.run_all_pps('playlist', ie_result)
2061 self.to_screen(f'[download] Finished downloading playlist: {title}')
2062 return ie_result
2063
2064 @_handle_extraction_exceptions
2065 def __process_iterable_entry(self, entry, download, extra_info):
2066 return self.process_ie_result(
2067 entry, download=download, extra_info=extra_info)
2068
2069 def _build_format_filter(self, filter_spec):
2070 " Returns a function to filter the formats according to the filter_spec "
2071
2072 OPERATORS = {
2073 '<': operator.lt,
2074 '<=': operator.le,
2075 '>': operator.gt,
2076 '>=': operator.ge,
2077 '=': operator.eq,
2078 '!=': operator.ne,
2079 }
2080 operator_rex = re.compile(r'''(?x)\s*
2081 (?P<key>[\w.-]+)\s*
2082 (?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
2083 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)\s*
2084 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
2085 m = operator_rex.fullmatch(filter_spec)
2086 if m:
2087 try:
2088 comparison_value = int(m.group('value'))
2089 except ValueError:
2090 comparison_value = parse_filesize(m.group('value'))
2091 if comparison_value is None:
2092 comparison_value = parse_filesize(m.group('value') + 'B')
2093 if comparison_value is None:
2094 raise ValueError(
2095 'Invalid value %r in format specification %r' % (
2096 m.group('value'), filter_spec))
2097 op = OPERATORS[m.group('op')]
2098
2099 if not m:
2100 STR_OPERATORS = {
2101 '=': operator.eq,
2102 '^=': lambda attr, value: attr.startswith(value),
2103 '$=': lambda attr, value: attr.endswith(value),
2104 '*=': lambda attr, value: value in attr,
2105 '~=': lambda attr, value: value.search(attr) is not None
2106 }
2107 str_operator_rex = re.compile(r'''(?x)\s*
2108 (?P<key>[a-zA-Z0-9._-]+)\s*
2109 (?P<negation>!\s*)?(?P<op>%s)\s*(?P<none_inclusive>\?\s*)?
2110 (?P<quote>["'])?
2111 (?P<value>(?(quote)(?:(?!(?P=quote))[^\\]|\\.)+|[\w.-]+))
2112 (?(quote)(?P=quote))\s*
2113 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
2114 m = str_operator_rex.fullmatch(filter_spec)
2115 if m:
2116 if m.group('op') == '~=':
2117 comparison_value = re.compile(m.group('value'))
2118 else:
2119 comparison_value = re.sub(r'''\\([\\"'])''', r'\1', m.group('value'))
2120 str_op = STR_OPERATORS[m.group('op')]
2121 if m.group('negation'):
2122 op = lambda attr, value: not str_op(attr, value)
2123 else:
2124 op = str_op
2125
2126 if not m:
2127 raise SyntaxError('Invalid filter specification %r' % filter_spec)
2128
2129 def _filter(f):
2130 actual_value = f.get(m.group('key'))
2131 if actual_value is None:
2132 return m.group('none_inclusive')
2133 return op(actual_value, comparison_value)
2134 return _filter
2135
2136 def _check_formats(self, formats):
2137 for f in formats:
2138 self.to_screen('[info] Testing format %s' % f['format_id'])
2139 path = self.get_output_path('temp')
2140 if not self._ensure_dir_exists(f'{path}/'):
2141 continue
2142 temp_file = tempfile.NamedTemporaryFile(suffix='.tmp', delete=False, dir=path or None)
2143 temp_file.close()
2144 try:
2145 success, _ = self.dl(temp_file.name, f, test=True)
2146 except (DownloadError, OSError, ValueError) + network_exceptions:
2147 success = False
2148 finally:
2149 if os.path.exists(temp_file.name):
2150 try:
2151 os.remove(temp_file.name)
2152 except OSError:
2153 self.report_warning('Unable to delete temporary file "%s"' % temp_file.name)
2154 if success:
2155 yield f
2156 else:
2157 self.to_screen('[info] Unable to download format %s. Skipping...' % f['format_id'])
2158
2159 def _default_format_spec(self, info_dict, download=True):
2160
2161 def can_merge():
2162 merger = FFmpegMergerPP(self)
2163 return merger.available and merger.can_merge()
2164
2165 prefer_best = (
2166 not self.params.get('simulate')
2167 and download
2168 and (
2169 not can_merge()
2170 or info_dict.get('is_live') and not self.params.get('live_from_start')
2171 or self.params['outtmpl']['default'] == '-'))
2172 compat = (
2173 prefer_best
2174 or self.params.get('allow_multiple_audio_streams', False)
2175 or 'format-spec' in self.params['compat_opts'])
2176
2177 return (
2178 'best/bestvideo+bestaudio' if prefer_best
2179 else 'bestvideo*+bestaudio/best' if not compat
2180 else 'bestvideo+bestaudio/best')
2181
2182 def build_format_selector(self, format_spec):
2183 def syntax_error(note, start):
2184 message = (
2185 'Invalid format specification: '
2186 '{}\n\t{}\n\t{}^'.format(note, format_spec, ' ' * start[1]))
2187 return SyntaxError(message)
2188
2189 PICKFIRST = 'PICKFIRST'
2190 MERGE = 'MERGE'
2191 SINGLE = 'SINGLE'
2192 GROUP = 'GROUP'
2193 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
2194
2195 allow_multiple_streams = {'audio': self.params.get('allow_multiple_audio_streams', False),
2196 'video': self.params.get('allow_multiple_video_streams', False)}
2197
2198 def _parse_filter(tokens):
2199 filter_parts = []
2200 for type, string_, start, _, _ in tokens:
2201 if type == tokenize.OP and string_ == ']':
2202 return ''.join(filter_parts)
2203 else:
2204 filter_parts.append(string_)
2205
2206 def _remove_unused_ops(tokens):
2207 # Remove operators that we don't use and join them with the surrounding strings.
2208 # E.g. 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
2209 ALLOWED_OPS = ('/', '+', ',', '(', ')')
2210 last_string, last_start, last_end, last_line = None, None, None, None
2211 for type, string_, start, end, line in tokens:
2212 if type == tokenize.OP and string_ == '[':
2213 if last_string:
2214 yield tokenize.NAME, last_string, last_start, last_end, last_line
2215 last_string = None
2216 yield type, string_, start, end, line
2217 # everything inside brackets will be handled by _parse_filter
2218 for type, string_, start, end, line in tokens:
2219 yield type, string_, start, end, line
2220 if type == tokenize.OP and string_ == ']':
2221 break
2222 elif type == tokenize.OP and string_ in ALLOWED_OPS:
2223 if last_string:
2224 yield tokenize.NAME, last_string, last_start, last_end, last_line
2225 last_string = None
2226 yield type, string_, start, end, line
2227 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
2228 if not last_string:
2229 last_string = string_
2230 last_start = start
2231 last_end = end
2232 else:
2233 last_string += string_
2234 if last_string:
2235 yield tokenize.NAME, last_string, last_start, last_end, last_line
2236
2237 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
2238 selectors = []
2239 current_selector = None
2240 for type, string_, start, _, _ in tokens:
2241 # ENCODING is only defined in Python 3.x
2242 if type == getattr(tokenize, 'ENCODING', None):
2243 continue
2244 elif type in [tokenize.NAME, tokenize.NUMBER]:
2245 current_selector = FormatSelector(SINGLE, string_, [])
2246 elif type == tokenize.OP:
2247 if string_ == ')':
2248 if not inside_group:
2249 # ')' will be handled by the parentheses group
2250 tokens.restore_last_token()
2251 break
2252 elif inside_merge and string_ in ['/', ',']:
2253 tokens.restore_last_token()
2254 break
2255 elif inside_choice and string_ == ',':
2256 tokens.restore_last_token()
2257 break
2258 elif string_ == ',':
2259 if not current_selector:
2260 raise syntax_error('"," must follow a format selector', start)
2261 selectors.append(current_selector)
2262 current_selector = None
2263 elif string_ == '/':
2264 if not current_selector:
2265 raise syntax_error('"/" must follow a format selector', start)
2266 first_choice = current_selector
2267 second_choice = _parse_format_selection(tokens, inside_choice=True)
2268 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
2269 elif string_ == '[':
2270 if not current_selector:
2271 current_selector = FormatSelector(SINGLE, 'best', [])
2272 format_filter = _parse_filter(tokens)
2273 current_selector.filters.append(format_filter)
2274 elif string_ == '(':
2275 if current_selector:
2276 raise syntax_error('Unexpected "("', start)
2277 group = _parse_format_selection(tokens, inside_group=True)
2278 current_selector = FormatSelector(GROUP, group, [])
2279 elif string_ == '+':
2280 if not current_selector:
2281 raise syntax_error('Unexpected "+"', start)
2282 selector_1 = current_selector
2283 selector_2 = _parse_format_selection(tokens, inside_merge=True)
2284 if not selector_2:
2285 raise syntax_error('Expected a selector', start)
2286 current_selector = FormatSelector(MERGE, (selector_1, selector_2), [])
2287 else:
2288 raise syntax_error(f'Operator not recognized: "{string_}"', start)
2289 elif type == tokenize.ENDMARKER:
2290 break
2291 if current_selector:
2292 selectors.append(current_selector)
2293 return selectors
2294
2295 def _merge(formats_pair):
2296 format_1, format_2 = formats_pair
2297
2298 formats_info = []
2299 formats_info.extend(format_1.get('requested_formats', (format_1,)))
2300 formats_info.extend(format_2.get('requested_formats', (format_2,)))
2301
2302 if not allow_multiple_streams['video'] or not allow_multiple_streams['audio']:
2303 get_no_more = {'video': False, 'audio': False}
2304 for (i, fmt_info) in enumerate(formats_info):
2305 if fmt_info.get('acodec') == fmt_info.get('vcodec') == 'none':
2306 formats_info.pop(i)
2307 continue
2308 for aud_vid in ['audio', 'video']:
2309 if not allow_multiple_streams[aud_vid] and fmt_info.get(aud_vid[0] + 'codec') != 'none':
2310 if get_no_more[aud_vid]:
2311 formats_info.pop(i)
2312 break
2313 get_no_more[aud_vid] = True
2314
2315 if len(formats_info) == 1:
2316 return formats_info[0]
2317
2318 video_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('vcodec') != 'none']
2319 audio_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('acodec') != 'none']
2320
2321 the_only_video = video_fmts[0] if len(video_fmts) == 1 else None
2322 the_only_audio = audio_fmts[0] if len(audio_fmts) == 1 else None
2323
2324 output_ext = get_compatible_ext(
2325 vcodecs=[f.get('vcodec') for f in video_fmts],
2326 acodecs=[f.get('acodec') for f in audio_fmts],
2327 vexts=[f['ext'] for f in video_fmts],
2328 aexts=[f['ext'] for f in audio_fmts],
2329 preferences=(try_call(lambda: self.params['merge_output_format'].split('/'))
2330 or self.params.get('prefer_free_formats') and ('webm', 'mkv')))
2331
2332 filtered = lambda *keys: filter(None, (traverse_obj(fmt, *keys) for fmt in formats_info))
2333
2334 new_dict = {
2335 'requested_formats': formats_info,
2336 'format': '+'.join(filtered('format')),
2337 'format_id': '+'.join(filtered('format_id')),
2338 'ext': output_ext,
2339 'protocol': '+'.join(map(determine_protocol, formats_info)),
2340 'language': '+'.join(orderedSet(filtered('language'))) or None,
2341 'format_note': '+'.join(orderedSet(filtered('format_note'))) or None,
2342 'filesize_approx': sum(filtered('filesize', 'filesize_approx')) or None,
2343 'tbr': sum(filtered('tbr', 'vbr', 'abr')),
2344 }
2345
2346 if the_only_video:
2347 new_dict.update({
2348 'width': the_only_video.get('width'),
2349 'height': the_only_video.get('height'),
2350 'resolution': the_only_video.get('resolution') or self.format_resolution(the_only_video),
2351 'fps': the_only_video.get('fps'),
2352 'dynamic_range': the_only_video.get('dynamic_range'),
2353 'vcodec': the_only_video.get('vcodec'),
2354 'vbr': the_only_video.get('vbr'),
2355 'stretched_ratio': the_only_video.get('stretched_ratio'),
2356 'aspect_ratio': the_only_video.get('aspect_ratio'),
2357 })
2358
2359 if the_only_audio:
2360 new_dict.update({
2361 'acodec': the_only_audio.get('acodec'),
2362 'abr': the_only_audio.get('abr'),
2363 'asr': the_only_audio.get('asr'),
2364 'audio_channels': the_only_audio.get('audio_channels')
2365 })
2366
2367 return new_dict
2368
2369 def _check_formats(formats):
2370 if self.params.get('check_formats') == 'selected':
2371 yield from self._check_formats(formats)
2372 return
2373 elif (self.params.get('check_formats') is not None
2374 or self.params.get('allow_unplayable_formats')):
2375 yield from formats
2376 return
2377
2378 for f in formats:
2379 if f.get('has_drm') or f.get('__needs_testing'):
2380 yield from self._check_formats([f])
2381 else:
2382 yield f
2383
2384 def _build_selector_function(selector):
2385 if isinstance(selector, list): # ,
2386 fs = [_build_selector_function(s) for s in selector]
2387
2388 def selector_function(ctx):
2389 for f in fs:
2390 yield from f(ctx)
2391 return selector_function
2392
2393 elif selector.type == GROUP: # ()
2394 selector_function = _build_selector_function(selector.selector)
2395
2396 elif selector.type == PICKFIRST: # /
2397 fs = [_build_selector_function(s) for s in selector.selector]
2398
2399 def selector_function(ctx):
2400 for f in fs:
2401 picked_formats = list(f(ctx))
2402 if picked_formats:
2403 return picked_formats
2404 return []
2405
2406 elif selector.type == MERGE: # +
2407 selector_1, selector_2 = map(_build_selector_function, selector.selector)
2408
2409 def selector_function(ctx):
2410 for pair in itertools.product(selector_1(ctx), selector_2(ctx)):
2411 yield _merge(pair)
2412
2413 elif selector.type == SINGLE: # atom
2414 format_spec = selector.selector or 'best'
2415
2416 # TODO: Add allvideo, allaudio etc by generalizing the code with best/worst selector
2417 if format_spec == 'all':
2418 def selector_function(ctx):
2419 yield from _check_formats(ctx['formats'][::-1])
2420 elif format_spec == 'mergeall':
2421 def selector_function(ctx):
2422 formats = list(_check_formats(
2423 f for f in ctx['formats'] if f.get('vcodec') != 'none' or f.get('acodec') != 'none'))
2424 if not formats:
2425 return
2426 merged_format = formats[-1]
2427 for f in formats[-2::-1]:
2428 merged_format = _merge((merged_format, f))
2429 yield merged_format
2430
2431 else:
2432 format_fallback, seperate_fallback, format_reverse, format_idx = False, None, True, 1
2433 mobj = re.match(
2434 r'(?P<bw>best|worst|b|w)(?P<type>video|audio|v|a)?(?P<mod>\*)?(?:\.(?P<n>[1-9]\d*))?$',
2435 format_spec)
2436 if mobj is not None:
2437 format_idx = int_or_none(mobj.group('n'), default=1)
2438 format_reverse = mobj.group('bw')[0] == 'b'
2439 format_type = (mobj.group('type') or [None])[0]
2440 not_format_type = {'v': 'a', 'a': 'v'}.get(format_type)
2441 format_modified = mobj.group('mod') is not None
2442
2443 format_fallback = not format_type and not format_modified # for b, w
2444 _filter_f = (
2445 (lambda f: f.get('%scodec' % format_type) != 'none')
2446 if format_type and format_modified # bv*, ba*, wv*, wa*
2447 else (lambda f: f.get('%scodec' % not_format_type) == 'none')
2448 if format_type # bv, ba, wv, wa
2449 else (lambda f: f.get('vcodec') != 'none' and f.get('acodec') != 'none')
2450 if not format_modified # b, w
2451 else lambda f: True) # b*, w*
2452 filter_f = lambda f: _filter_f(f) and (
2453 f.get('vcodec') != 'none' or f.get('acodec') != 'none')
2454 else:
2455 if format_spec in self._format_selection_exts['audio']:
2456 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none'
2457 elif format_spec in self._format_selection_exts['video']:
2458 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none' and f.get('vcodec') != 'none'
2459 seperate_fallback = lambda f: f.get('ext') == format_spec and f.get('vcodec') != 'none'
2460 elif format_spec in self._format_selection_exts['storyboards']:
2461 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') == 'none' and f.get('vcodec') == 'none'
2462 else:
2463 filter_f = lambda f: f.get('format_id') == format_spec # id
2464
2465 def selector_function(ctx):
2466 formats = list(ctx['formats'])
2467 matches = list(filter(filter_f, formats)) if filter_f is not None else formats
2468 if not matches:
2469 if format_fallback and ctx['incomplete_formats']:
2470 # for extractors with incomplete formats (audio only (soundcloud)
2471 # or video only (imgur)) best/worst will fallback to
2472 # best/worst {video,audio}-only format
2473 matches = list(filter(lambda f: f.get('vcodec') != 'none' or f.get('acodec') != 'none', formats))
2474 elif seperate_fallback and not ctx['has_merged_format']:
2475 # for compatibility with youtube-dl when there is no pre-merged format
2476 matches = list(filter(seperate_fallback, formats))
2477 matches = LazyList(_check_formats(matches[::-1 if format_reverse else 1]))
2478 try:
2479 yield matches[format_idx - 1]
2480 except LazyList.IndexError:
2481 return
2482
2483 filters = [self._build_format_filter(f) for f in selector.filters]
2484
2485 def final_selector(ctx):
2486 ctx_copy = dict(ctx)
2487 for _filter in filters:
2488 ctx_copy['formats'] = list(filter(_filter, ctx_copy['formats']))
2489 return selector_function(ctx_copy)
2490 return final_selector
2491
2492 # HACK: Python 3.12 changed the underlying parser, rendering '7_a' invalid
2493 # Prefix numbers with random letters to avoid it being classified as a number
2494 # See: https://github.com/yt-dlp/yt-dlp/pulls/8797
2495 # TODO: Implement parser not reliant on tokenize.tokenize
2496 prefix = ''.join(random.choices(string.ascii_letters, k=32))
2497 stream = io.BytesIO(re.sub(r'\d[_\d]*', rf'{prefix}\g<0>', format_spec).encode())
2498 try:
2499 tokens = list(_remove_unused_ops(
2500 token._replace(string=token.string.replace(prefix, ''))
2501 for token in tokenize.tokenize(stream.readline)))
2502 except tokenize.TokenError:
2503 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
2504
2505 class TokenIterator:
2506 def __init__(self, tokens):
2507 self.tokens = tokens
2508 self.counter = 0
2509
2510 def __iter__(self):
2511 return self
2512
2513 def __next__(self):
2514 if self.counter >= len(self.tokens):
2515 raise StopIteration()
2516 value = self.tokens[self.counter]
2517 self.counter += 1
2518 return value
2519
2520 next = __next__
2521
2522 def restore_last_token(self):
2523 self.counter -= 1
2524
2525 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
2526 return _build_selector_function(parsed_selector)
2527
2528 def _calc_headers(self, info_dict, load_cookies=False):
2529 res = HTTPHeaderDict(self.params['http_headers'], info_dict.get('http_headers'))
2530 clean_headers(res)
2531
2532 if load_cookies: # For --load-info-json
2533 self._load_cookies(res.get('Cookie'), autoscope=info_dict['url']) # compat
2534 self._load_cookies(info_dict.get('cookies'), autoscope=False)
2535 # The `Cookie` header is removed to prevent leaks and unscoped cookies.
2536 # See: https://github.com/yt-dlp/yt-dlp/security/advisories/GHSA-v8mc-9377-rwjj
2537 res.pop('Cookie', None)
2538 cookies = self.cookiejar.get_cookies_for_url(info_dict['url'])
2539 if cookies:
2540 encoder = LenientSimpleCookie()
2541 values = []
2542 for cookie in cookies:
2543 _, value = encoder.value_encode(cookie.value)
2544 values.append(f'{cookie.name}={value}')
2545 if cookie.domain:
2546 values.append(f'Domain={cookie.domain}')
2547 if cookie.path:
2548 values.append(f'Path={cookie.path}')
2549 if cookie.secure:
2550 values.append('Secure')
2551 if cookie.expires:
2552 values.append(f'Expires={cookie.expires}')
2553 if cookie.version:
2554 values.append(f'Version={cookie.version}')
2555 info_dict['cookies'] = '; '.join(values)
2556
2557 if 'X-Forwarded-For' not in res:
2558 x_forwarded_for_ip = info_dict.get('__x_forwarded_for_ip')
2559 if x_forwarded_for_ip:
2560 res['X-Forwarded-For'] = x_forwarded_for_ip
2561
2562 return res
2563
2564 def _calc_cookies(self, url):
2565 self.deprecation_warning('"YoutubeDL._calc_cookies" is deprecated and may be removed in a future version')
2566 return self.cookiejar.get_cookie_header(url)
2567
2568 def _sort_thumbnails(self, thumbnails):
2569 thumbnails.sort(key=lambda t: (
2570 t.get('preference') if t.get('preference') is not None else -1,
2571 t.get('width') if t.get('width') is not None else -1,
2572 t.get('height') if t.get('height') is not None else -1,
2573 t.get('id') if t.get('id') is not None else '',
2574 t.get('url')))
2575
2576 def _sanitize_thumbnails(self, info_dict):
2577 thumbnails = info_dict.get('thumbnails')
2578 if thumbnails is None:
2579 thumbnail = info_dict.get('thumbnail')
2580 if thumbnail:
2581 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
2582 if not thumbnails:
2583 return
2584
2585 def check_thumbnails(thumbnails):
2586 for t in thumbnails:
2587 self.to_screen(f'[info] Testing thumbnail {t["id"]}')
2588 try:
2589 self.urlopen(HEADRequest(t['url']))
2590 except network_exceptions as err:
2591 self.to_screen(f'[info] Unable to connect to thumbnail {t["id"]} URL {t["url"]!r} - {err}. Skipping...')
2592 continue
2593 yield t
2594
2595 self._sort_thumbnails(thumbnails)
2596 for i, t in enumerate(thumbnails):
2597 if t.get('id') is None:
2598 t['id'] = '%d' % i
2599 if t.get('width') and t.get('height'):
2600 t['resolution'] = '%dx%d' % (t['width'], t['height'])
2601 t['url'] = sanitize_url(t['url'])
2602
2603 if self.params.get('check_formats') is True:
2604 info_dict['thumbnails'] = LazyList(check_thumbnails(thumbnails[::-1]), reverse=True)
2605 else:
2606 info_dict['thumbnails'] = thumbnails
2607
2608 def _fill_common_fields(self, info_dict, final=True):
2609 # TODO: move sanitization here
2610 if final:
2611 title = info_dict['fulltitle'] = info_dict.get('title')
2612 if not title:
2613 if title == '':
2614 self.write_debug('Extractor gave empty title. Creating a generic title')
2615 else:
2616 self.report_warning('Extractor failed to obtain "title". Creating a generic title instead')
2617 info_dict['title'] = f'{info_dict["extractor"].replace(":", "-")} video #{info_dict["id"]}'
2618
2619 if info_dict.get('duration') is not None:
2620 info_dict['duration_string'] = formatSeconds(info_dict['duration'])
2621
2622 for ts_key, date_key in (
2623 ('timestamp', 'upload_date'),
2624 ('release_timestamp', 'release_date'),
2625 ('modified_timestamp', 'modified_date'),
2626 ):
2627 if info_dict.get(date_key) is None and info_dict.get(ts_key) is not None:
2628 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
2629 # see http://bugs.python.org/issue1646728)
2630 with contextlib.suppress(ValueError, OverflowError, OSError):
2631 upload_date = datetime.datetime.fromtimestamp(info_dict[ts_key], datetime.timezone.utc)
2632 info_dict[date_key] = upload_date.strftime('%Y%m%d')
2633
2634 if not info_dict.get('release_year'):
2635 info_dict['release_year'] = traverse_obj(info_dict, ('release_date', {lambda x: int(x[:4])}))
2636
2637 live_keys = ('is_live', 'was_live')
2638 live_status = info_dict.get('live_status')
2639 if live_status is None:
2640 for key in live_keys:
2641 if info_dict.get(key) is False:
2642 continue
2643 if info_dict.get(key):
2644 live_status = key
2645 break
2646 if all(info_dict.get(key) is False for key in live_keys):
2647 live_status = 'not_live'
2648 if live_status:
2649 info_dict['live_status'] = live_status
2650 for key in live_keys:
2651 if info_dict.get(key) is None:
2652 info_dict[key] = (live_status == key)
2653 if live_status == 'post_live':
2654 info_dict['was_live'] = True
2655
2656 # Auto generate title fields corresponding to the *_number fields when missing
2657 # in order to always have clean titles. This is very common for TV series.
2658 for field in ('chapter', 'season', 'episode'):
2659 if final and info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
2660 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
2661
2662 for old_key, new_key in self._deprecated_multivalue_fields.items():
2663 if new_key in info_dict and old_key in info_dict:
2664 if '_version' not in info_dict: # HACK: Do not warn when using --load-info-json
2665 self.deprecation_warning(f'Do not return {old_key!r} when {new_key!r} is present')
2666 elif old_value := info_dict.get(old_key):
2667 info_dict[new_key] = old_value.split(', ')
2668 elif new_value := info_dict.get(new_key):
2669 info_dict[old_key] = ', '.join(v.replace(',', '\N{FULLWIDTH COMMA}') for v in new_value)
2670
2671 def _raise_pending_errors(self, info):
2672 err = info.pop('__pending_error', None)
2673 if err:
2674 self.report_error(err, tb=False)
2675
2676 def sort_formats(self, info_dict):
2677 formats = self._get_formats(info_dict)
2678 formats.sort(key=FormatSorter(
2679 self, info_dict.get('_format_sort_fields') or []).calculate_preference)
2680
2681 def process_video_result(self, info_dict, download=True):
2682 assert info_dict.get('_type', 'video') == 'video'
2683 self._num_videos += 1
2684
2685 if 'id' not in info_dict:
2686 raise ExtractorError('Missing "id" field in extractor result', ie=info_dict['extractor'])
2687 elif not info_dict.get('id'):
2688 raise ExtractorError('Extractor failed to obtain "id"', ie=info_dict['extractor'])
2689
2690 def report_force_conversion(field, field_not, conversion):
2691 self.report_warning(
2692 '"%s" field is not %s - forcing %s conversion, there is an error in extractor'
2693 % (field, field_not, conversion))
2694
2695 def sanitize_string_field(info, string_field):
2696 field = info.get(string_field)
2697 if field is None or isinstance(field, str):
2698 return
2699 report_force_conversion(string_field, 'a string', 'string')
2700 info[string_field] = str(field)
2701
2702 def sanitize_numeric_fields(info):
2703 for numeric_field in self._NUMERIC_FIELDS:
2704 field = info.get(numeric_field)
2705 if field is None or isinstance(field, (int, float)):
2706 continue
2707 report_force_conversion(numeric_field, 'numeric', 'int')
2708 info[numeric_field] = int_or_none(field)
2709
2710 sanitize_string_field(info_dict, 'id')
2711 sanitize_numeric_fields(info_dict)
2712 if info_dict.get('section_end') and info_dict.get('section_start') is not None:
2713 info_dict['duration'] = round(info_dict['section_end'] - info_dict['section_start'], 3)
2714 if (info_dict.get('duration') or 0) <= 0 and info_dict.pop('duration', None):
2715 self.report_warning('"duration" field is negative, there is an error in extractor')
2716
2717 chapters = info_dict.get('chapters') or []
2718 if chapters and chapters[0].get('start_time'):
2719 chapters.insert(0, {'start_time': 0})
2720
2721 dummy_chapter = {'end_time': 0, 'start_time': info_dict.get('duration')}
2722 for idx, (prev, current, next_) in enumerate(zip(
2723 (dummy_chapter, *chapters), chapters, (*chapters[1:], dummy_chapter)), 1):
2724 if current.get('start_time') is None:
2725 current['start_time'] = prev.get('end_time')
2726 if not current.get('end_time'):
2727 current['end_time'] = next_.get('start_time')
2728 if not current.get('title'):
2729 current['title'] = f'<Untitled Chapter {idx}>'
2730
2731 if 'playlist' not in info_dict:
2732 # It isn't part of a playlist
2733 info_dict['playlist'] = None
2734 info_dict['playlist_index'] = None
2735
2736 self._sanitize_thumbnails(info_dict)
2737
2738 thumbnail = info_dict.get('thumbnail')
2739 thumbnails = info_dict.get('thumbnails')
2740 if thumbnail:
2741 info_dict['thumbnail'] = sanitize_url(thumbnail)
2742 elif thumbnails:
2743 info_dict['thumbnail'] = thumbnails[-1]['url']
2744
2745 if info_dict.get('display_id') is None and 'id' in info_dict:
2746 info_dict['display_id'] = info_dict['id']
2747
2748 self._fill_common_fields(info_dict)
2749
2750 for cc_kind in ('subtitles', 'automatic_captions'):
2751 cc = info_dict.get(cc_kind)
2752 if cc:
2753 for _, subtitle in cc.items():
2754 for subtitle_format in subtitle:
2755 if subtitle_format.get('url'):
2756 subtitle_format['url'] = sanitize_url(subtitle_format['url'])
2757 if subtitle_format.get('ext') is None:
2758 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
2759
2760 automatic_captions = info_dict.get('automatic_captions')
2761 subtitles = info_dict.get('subtitles')
2762
2763 info_dict['requested_subtitles'] = self.process_subtitles(
2764 info_dict['id'], subtitles, automatic_captions)
2765
2766 formats = self._get_formats(info_dict)
2767
2768 # Backward compatibility with InfoExtractor._sort_formats
2769 field_preference = (formats or [{}])[0].pop('__sort_fields', None)
2770 if field_preference:
2771 info_dict['_format_sort_fields'] = field_preference
2772
2773 info_dict['_has_drm'] = any( # or None ensures --clean-infojson removes it
2774 f.get('has_drm') and f['has_drm'] != 'maybe' for f in formats) or None
2775 if not self.params.get('allow_unplayable_formats'):
2776 formats = [f for f in formats if not f.get('has_drm') or f['has_drm'] == 'maybe']
2777
2778 if formats and all(f.get('acodec') == f.get('vcodec') == 'none' for f in formats):
2779 self.report_warning(
2780 f'{"This video is DRM protected and " if info_dict["_has_drm"] else ""}'
2781 'only images are available for download. Use --list-formats to see them'.capitalize())
2782
2783 get_from_start = not info_dict.get('is_live') or bool(self.params.get('live_from_start'))
2784 if not get_from_start:
2785 info_dict['title'] += ' ' + datetime.datetime.now().strftime('%Y-%m-%d %H:%M')
2786 if info_dict.get('is_live') and formats:
2787 formats = [f for f in formats if bool(f.get('is_from_start')) == get_from_start]
2788 if get_from_start and not formats:
2789 self.raise_no_formats(info_dict, msg=(
2790 '--live-from-start is passed, but there are no formats that can be downloaded from the start. '
2791 'If you want to download from the current time, use --no-live-from-start'))
2792
2793 def is_wellformed(f):
2794 url = f.get('url')
2795 if not url:
2796 self.report_warning(
2797 '"url" field is missing or empty - skipping format, '
2798 'there is an error in extractor')
2799 return False
2800 if isinstance(url, bytes):
2801 sanitize_string_field(f, 'url')
2802 return True
2803
2804 # Filter out malformed formats for better extraction robustness
2805 formats = list(filter(is_wellformed, formats or []))
2806
2807 if not formats:
2808 self.raise_no_formats(info_dict)
2809
2810 for format in formats:
2811 sanitize_string_field(format, 'format_id')
2812 sanitize_numeric_fields(format)
2813 format['url'] = sanitize_url(format['url'])
2814 if format.get('ext') is None:
2815 format['ext'] = determine_ext(format['url']).lower()
2816 if format.get('protocol') is None:
2817 format['protocol'] = determine_protocol(format)
2818 if format.get('resolution') is None:
2819 format['resolution'] = self.format_resolution(format, default=None)
2820 if format.get('dynamic_range') is None and format.get('vcodec') != 'none':
2821 format['dynamic_range'] = 'SDR'
2822 if format.get('aspect_ratio') is None:
2823 format['aspect_ratio'] = try_call(lambda: round(format['width'] / format['height'], 2))
2824 # For fragmented formats, "tbr" is often max bitrate and not average
2825 if (('manifest-filesize-approx' in self.params['compat_opts'] or not format.get('manifest_url'))
2826 and info_dict.get('duration') and format.get('tbr')
2827 and not format.get('filesize') and not format.get('filesize_approx')):
2828 format['filesize_approx'] = int(info_dict['duration'] * format['tbr'] * (1024 / 8))
2829 format['http_headers'] = self._calc_headers(collections.ChainMap(format, info_dict), load_cookies=True)
2830
2831 # Safeguard against old/insecure infojson when using --load-info-json
2832 if info_dict.get('http_headers'):
2833 info_dict['http_headers'] = HTTPHeaderDict(info_dict['http_headers'])
2834 info_dict['http_headers'].pop('Cookie', None)
2835
2836 # This is copied to http_headers by the above _calc_headers and can now be removed
2837 if '__x_forwarded_for_ip' in info_dict:
2838 del info_dict['__x_forwarded_for_ip']
2839
2840 self.sort_formats({
2841 'formats': formats,
2842 '_format_sort_fields': info_dict.get('_format_sort_fields')
2843 })
2844
2845 # Sanitize and group by format_id
2846 formats_dict = {}
2847 for i, format in enumerate(formats):
2848 if not format.get('format_id'):
2849 format['format_id'] = str(i)
2850 else:
2851 # Sanitize format_id from characters used in format selector expression
2852 format['format_id'] = re.sub(r'[\s,/+\[\]()]', '_', format['format_id'])
2853 formats_dict.setdefault(format['format_id'], []).append(format)
2854
2855 # Make sure all formats have unique format_id
2856 common_exts = set(itertools.chain(*self._format_selection_exts.values()))
2857 for format_id, ambiguous_formats in formats_dict.items():
2858 ambigious_id = len(ambiguous_formats) > 1
2859 for i, format in enumerate(ambiguous_formats):
2860 if ambigious_id:
2861 format['format_id'] = '%s-%d' % (format_id, i)
2862 # Ensure there is no conflict between id and ext in format selection
2863 # See https://github.com/yt-dlp/yt-dlp/issues/1282
2864 if format['format_id'] != format['ext'] and format['format_id'] in common_exts:
2865 format['format_id'] = 'f%s' % format['format_id']
2866
2867 if format.get('format') is None:
2868 format['format'] = '{id} - {res}{note}'.format(
2869 id=format['format_id'],
2870 res=self.format_resolution(format),
2871 note=format_field(format, 'format_note', ' (%s)'),
2872 )
2873
2874 if self.params.get('check_formats') is True:
2875 formats = LazyList(self._check_formats(formats[::-1]), reverse=True)
2876
2877 if not formats or formats[0] is not info_dict:
2878 # only set the 'formats' fields if the original info_dict list them
2879 # otherwise we end up with a circular reference, the first (and unique)
2880 # element in the 'formats' field in info_dict is info_dict itself,
2881 # which can't be exported to json
2882 info_dict['formats'] = formats
2883
2884 info_dict, _ = self.pre_process(info_dict)
2885
2886 if self._match_entry(info_dict, incomplete=self._format_fields) is not None:
2887 return info_dict
2888
2889 self.post_extract(info_dict)
2890 info_dict, _ = self.pre_process(info_dict, 'after_filter')
2891
2892 # The pre-processors may have modified the formats
2893 formats = self._get_formats(info_dict)
2894
2895 list_only = self.params.get('simulate') == 'list_only'
2896 interactive_format_selection = not list_only and self.format_selector == '-'
2897 if self.params.get('list_thumbnails'):
2898 self.list_thumbnails(info_dict)
2899 if self.params.get('listsubtitles'):
2900 if 'automatic_captions' in info_dict:
2901 self.list_subtitles(
2902 info_dict['id'], automatic_captions, 'automatic captions')
2903 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
2904 if self.params.get('listformats') or interactive_format_selection:
2905 self.list_formats(info_dict)
2906 if list_only:
2907 # Without this printing, -F --print-json will not work
2908 self.__forced_printings(info_dict)
2909 return info_dict
2910
2911 format_selector = self.format_selector
2912 while True:
2913 if interactive_format_selection:
2914 req_format = input(self._format_screen('\nEnter format selector ', self.Styles.EMPHASIS)
2915 + '(Press ENTER for default, or Ctrl+C to quit)'
2916 + self._format_screen(': ', self.Styles.EMPHASIS))
2917 try:
2918 format_selector = self.build_format_selector(req_format) if req_format else None
2919 except SyntaxError as err:
2920 self.report_error(err, tb=False, is_error=False)
2921 continue
2922
2923 if format_selector is None:
2924 req_format = self._default_format_spec(info_dict, download=download)
2925 self.write_debug(f'Default format spec: {req_format}')
2926 format_selector = self.build_format_selector(req_format)
2927
2928 formats_to_download = list(format_selector({
2929 'formats': formats,
2930 'has_merged_format': any('none' not in (f.get('acodec'), f.get('vcodec')) for f in formats),
2931 'incomplete_formats': (all(f.get('vcodec') == 'none' for f in formats) # No formats with video
2932 or all(f.get('acodec') == 'none' for f in formats)), # OR, No formats with audio
2933 }))
2934 if interactive_format_selection and not formats_to_download:
2935 self.report_error('Requested format is not available', tb=False, is_error=False)
2936 continue
2937 break
2938
2939 if not formats_to_download:
2940 if not self.params.get('ignore_no_formats_error'):
2941 raise ExtractorError(
2942 'Requested format is not available. Use --list-formats for a list of available formats',
2943 expected=True, video_id=info_dict['id'], ie=info_dict['extractor'])
2944 self.report_warning('Requested format is not available')
2945 # Process what we can, even without any available formats.
2946 formats_to_download = [{}]
2947
2948 requested_ranges = tuple(self.params.get('download_ranges', lambda *_: [{}])(info_dict, self))
2949 best_format, downloaded_formats = formats_to_download[-1], []
2950 if download:
2951 if best_format and requested_ranges:
2952 def to_screen(*msg):
2953 self.to_screen(f'[info] {info_dict["id"]}: {" ".join(", ".join(variadic(m)) for m in msg)}')
2954
2955 to_screen(f'Downloading {len(formats_to_download)} format(s):',
2956 (f['format_id'] for f in formats_to_download))
2957 if requested_ranges != ({}, ):
2958 to_screen(f'Downloading {len(requested_ranges)} time ranges:',
2959 (f'{c["start_time"]:.1f}-{c["end_time"]:.1f}' for c in requested_ranges))
2960 max_downloads_reached = False
2961
2962 for fmt, chapter in itertools.product(formats_to_download, requested_ranges):
2963 new_info = self._copy_infodict(info_dict)
2964 new_info.update(fmt)
2965 offset, duration = info_dict.get('section_start') or 0, info_dict.get('duration') or float('inf')
2966 end_time = offset + min(chapter.get('end_time', duration), duration)
2967 # duration may not be accurate. So allow deviations <1sec
2968 if end_time == float('inf') or end_time > offset + duration + 1:
2969 end_time = None
2970 if chapter or offset:
2971 new_info.update({
2972 'section_start': offset + chapter.get('start_time', 0),
2973 'section_end': end_time,
2974 'section_title': chapter.get('title'),
2975 'section_number': chapter.get('index'),
2976 })
2977 downloaded_formats.append(new_info)
2978 try:
2979 self.process_info(new_info)
2980 except MaxDownloadsReached:
2981 max_downloads_reached = True
2982 self._raise_pending_errors(new_info)
2983 # Remove copied info
2984 for key, val in tuple(new_info.items()):
2985 if info_dict.get(key) == val:
2986 new_info.pop(key)
2987 if max_downloads_reached:
2988 break
2989
2990 write_archive = {f.get('__write_download_archive', False) for f in downloaded_formats}
2991 assert write_archive.issubset({True, False, 'ignore'})
2992 if True in write_archive and False not in write_archive:
2993 self.record_download_archive(info_dict)
2994
2995 info_dict['requested_downloads'] = downloaded_formats
2996 info_dict = self.run_all_pps('after_video', info_dict)
2997 if max_downloads_reached:
2998 raise MaxDownloadsReached()
2999
3000 # We update the info dict with the selected best quality format (backwards compatibility)
3001 info_dict.update(best_format)
3002 return info_dict
3003
3004 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
3005 """Select the requested subtitles and their format"""
3006 available_subs, normal_sub_langs = {}, []
3007 if normal_subtitles and self.params.get('writesubtitles'):
3008 available_subs.update(normal_subtitles)
3009 normal_sub_langs = tuple(normal_subtitles.keys())
3010 if automatic_captions and self.params.get('writeautomaticsub'):
3011 for lang, cap_info in automatic_captions.items():
3012 if lang not in available_subs:
3013 available_subs[lang] = cap_info
3014
3015 if not available_subs or (
3016 not self.params.get('writesubtitles')
3017 and not self.params.get('writeautomaticsub')):
3018 return None
3019
3020 all_sub_langs = tuple(available_subs.keys())
3021 if self.params.get('allsubtitles', False):
3022 requested_langs = all_sub_langs
3023 elif self.params.get('subtitleslangs', False):
3024 try:
3025 requested_langs = orderedSet_from_options(
3026 self.params.get('subtitleslangs'), {'all': all_sub_langs}, use_regex=True)
3027 except re.error as e:
3028 raise ValueError(f'Wrong regex for subtitlelangs: {e.pattern}')
3029 else:
3030 requested_langs = LazyList(itertools.chain(
3031 ['en'] if 'en' in normal_sub_langs else [],
3032 filter(lambda f: f.startswith('en'), normal_sub_langs),
3033 ['en'] if 'en' in all_sub_langs else [],
3034 filter(lambda f: f.startswith('en'), all_sub_langs),
3035 normal_sub_langs, all_sub_langs,
3036 ))[:1]
3037 if requested_langs:
3038 self.to_screen(f'[info] {video_id}: Downloading subtitles: {", ".join(requested_langs)}')
3039
3040 formats_query = self.params.get('subtitlesformat', 'best')
3041 formats_preference = formats_query.split('/') if formats_query else []
3042 subs = {}
3043 for lang in requested_langs:
3044 formats = available_subs.get(lang)
3045 if formats is None:
3046 self.report_warning(f'{lang} subtitles not available for {video_id}')
3047 continue
3048 for ext in formats_preference:
3049 if ext == 'best':
3050 f = formats[-1]
3051 break
3052 matches = list(filter(lambda f: f['ext'] == ext, formats))
3053 if matches:
3054 f = matches[-1]
3055 break
3056 else:
3057 f = formats[-1]
3058 self.report_warning(
3059 'No subtitle format found matching "%s" for language %s, '
3060 'using %s' % (formats_query, lang, f['ext']))
3061 subs[lang] = f
3062 return subs
3063
3064 def _forceprint(self, key, info_dict):
3065 if info_dict is None:
3066 return
3067 info_copy = info_dict.copy()
3068 info_copy.setdefault('filename', self.prepare_filename(info_dict))
3069 if info_dict.get('requested_formats') is not None:
3070 # For RTMP URLs, also include the playpath
3071 info_copy['urls'] = '\n'.join(f['url'] + f.get('play_path', '') for f in info_dict['requested_formats'])
3072 elif info_dict.get('url'):
3073 info_copy['urls'] = info_dict['url'] + info_dict.get('play_path', '')
3074 info_copy['formats_table'] = self.render_formats_table(info_dict)
3075 info_copy['thumbnails_table'] = self.render_thumbnails_table(info_dict)
3076 info_copy['subtitles_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('subtitles'))
3077 info_copy['automatic_captions_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('automatic_captions'))
3078
3079 def format_tmpl(tmpl):
3080 mobj = re.fullmatch(r'([\w.:,]|-\d|(?P<dict>{([\w.:,]|-\d)+}))+=?', tmpl)
3081 if not mobj:
3082 return tmpl
3083
3084 fmt = '%({})s'
3085 if tmpl.startswith('{'):
3086 tmpl, fmt = f'.{tmpl}', '%({})j'
3087 if tmpl.endswith('='):
3088 tmpl, fmt = tmpl[:-1], '{0} = %({0})#j'
3089 return '\n'.join(map(fmt.format, [tmpl] if mobj.group('dict') else tmpl.split(',')))
3090
3091 for tmpl in self.params['forceprint'].get(key, []):
3092 self.to_stdout(self.evaluate_outtmpl(format_tmpl(tmpl), info_copy))
3093
3094 for tmpl, file_tmpl in self.params['print_to_file'].get(key, []):
3095 filename = self.prepare_filename(info_dict, outtmpl=file_tmpl)
3096 tmpl = format_tmpl(tmpl)
3097 self.to_screen(f'[info] Writing {tmpl!r} to: {filename}')
3098 if self._ensure_dir_exists(filename):
3099 with open(filename, 'a', encoding='utf-8', newline='') as f:
3100 f.write(self.evaluate_outtmpl(tmpl, info_copy) + os.linesep)
3101
3102 return info_copy
3103
3104 def __forced_printings(self, info_dict, filename=None, incomplete=True):
3105 if (self.params.get('forcejson')
3106 or self.params['forceprint'].get('video')
3107 or self.params['print_to_file'].get('video')):
3108 self.post_extract(info_dict)
3109 if filename:
3110 info_dict['filename'] = filename
3111 info_copy = self._forceprint('video', info_dict)
3112
3113 def print_field(field, actual_field=None, optional=False):
3114 if actual_field is None:
3115 actual_field = field
3116 if self.params.get(f'force{field}') and (
3117 info_copy.get(field) is not None or (not optional and not incomplete)):
3118 self.to_stdout(info_copy[actual_field])
3119
3120 print_field('title')
3121 print_field('id')
3122 print_field('url', 'urls')
3123 print_field('thumbnail', optional=True)
3124 print_field('description', optional=True)
3125 print_field('filename')
3126 if self.params.get('forceduration') and info_copy.get('duration') is not None:
3127 self.to_stdout(formatSeconds(info_copy['duration']))
3128 print_field('format')
3129
3130 if self.params.get('forcejson'):
3131 self.to_stdout(json.dumps(self.sanitize_info(info_dict)))
3132
3133 def dl(self, name, info, subtitle=False, test=False):
3134 if not info.get('url'):
3135 self.raise_no_formats(info, True)
3136
3137 if test:
3138 verbose = self.params.get('verbose')
3139 params = {
3140 'test': True,
3141 'quiet': self.params.get('quiet') or not verbose,
3142 'verbose': verbose,
3143 'noprogress': not verbose,
3144 'nopart': True,
3145 'skip_unavailable_fragments': False,
3146 'keep_fragments': False,
3147 'overwrites': True,
3148 '_no_ytdl_file': True,
3149 }
3150 else:
3151 params = self.params
3152 fd = get_suitable_downloader(info, params, to_stdout=(name == '-'))(self, params)
3153 if not test:
3154 for ph in self._progress_hooks:
3155 fd.add_progress_hook(ph)
3156 urls = '", "'.join(
3157 (f['url'].split(',')[0] + ',<data>' if f['url'].startswith('data:') else f['url'])
3158 for f in info.get('requested_formats', []) or [info])
3159 self.write_debug(f'Invoking {fd.FD_NAME} downloader on "{urls}"')
3160
3161 # Note: Ideally info should be a deep-copied so that hooks cannot modify it.
3162 # But it may contain objects that are not deep-copyable
3163 new_info = self._copy_infodict(info)
3164 if new_info.get('http_headers') is None:
3165 new_info['http_headers'] = self._calc_headers(new_info)
3166 return fd.download(name, new_info, subtitle)
3167
3168 def existing_file(self, filepaths, *, default_overwrite=True):
3169 existing_files = list(filter(os.path.exists, orderedSet(filepaths)))
3170 if existing_files and not self.params.get('overwrites', default_overwrite):
3171 return existing_files[0]
3172
3173 for file in existing_files:
3174 self.report_file_delete(file)
3175 os.remove(file)
3176 return None
3177
3178 def process_info(self, info_dict):
3179 """Process a single resolved IE result. (Modifies it in-place)"""
3180
3181 assert info_dict.get('_type', 'video') == 'video'
3182 original_infodict = info_dict
3183
3184 if 'format' not in info_dict and 'ext' in info_dict:
3185 info_dict['format'] = info_dict['ext']
3186
3187 if self._match_entry(info_dict) is not None:
3188 info_dict['__write_download_archive'] = 'ignore'
3189 return
3190
3191 # Does nothing under normal operation - for backward compatibility of process_info
3192 self.post_extract(info_dict)
3193
3194 def replace_info_dict(new_info):
3195 nonlocal info_dict
3196 if new_info == info_dict:
3197 return
3198 info_dict.clear()
3199 info_dict.update(new_info)
3200
3201 new_info, _ = self.pre_process(info_dict, 'video')
3202 replace_info_dict(new_info)
3203 self._num_downloads += 1
3204
3205 # info_dict['_filename'] needs to be set for backward compatibility
3206 info_dict['_filename'] = full_filename = self.prepare_filename(info_dict, warn=True)
3207 temp_filename = self.prepare_filename(info_dict, 'temp')
3208 files_to_move = {}
3209
3210 # Forced printings
3211 self.__forced_printings(info_dict, full_filename, incomplete=('format' not in info_dict))
3212
3213 def check_max_downloads():
3214 if self._num_downloads >= float(self.params.get('max_downloads') or 'inf'):
3215 raise MaxDownloadsReached()
3216
3217 if self.params.get('simulate'):
3218 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
3219 check_max_downloads()
3220 return
3221
3222 if full_filename is None:
3223 return
3224 if not self._ensure_dir_exists(encodeFilename(full_filename)):
3225 return
3226 if not self._ensure_dir_exists(encodeFilename(temp_filename)):
3227 return
3228
3229 if self._write_description('video', info_dict,
3230 self.prepare_filename(info_dict, 'description')) is None:
3231 return
3232
3233 sub_files = self._write_subtitles(info_dict, temp_filename)
3234 if sub_files is None:
3235 return
3236 files_to_move.update(dict(sub_files))
3237
3238 thumb_files = self._write_thumbnails(
3239 'video', info_dict, temp_filename, self.prepare_filename(info_dict, 'thumbnail'))
3240 if thumb_files is None:
3241 return
3242 files_to_move.update(dict(thumb_files))
3243
3244 infofn = self.prepare_filename(info_dict, 'infojson')
3245 _infojson_written = self._write_info_json('video', info_dict, infofn)
3246 if _infojson_written:
3247 info_dict['infojson_filename'] = infofn
3248 # For backward compatibility, even though it was a private field
3249 info_dict['__infojson_filename'] = infofn
3250 elif _infojson_written is None:
3251 return
3252
3253 # Note: Annotations are deprecated
3254 annofn = None
3255 if self.params.get('writeannotations', False):
3256 annofn = self.prepare_filename(info_dict, 'annotation')
3257 if annofn:
3258 if not self._ensure_dir_exists(encodeFilename(annofn)):
3259 return
3260 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(annofn)):
3261 self.to_screen('[info] Video annotations are already present')
3262 elif not info_dict.get('annotations'):
3263 self.report_warning('There are no annotations to write.')
3264 else:
3265 try:
3266 self.to_screen('[info] Writing video annotations to: ' + annofn)
3267 with open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
3268 annofile.write(info_dict['annotations'])
3269 except (KeyError, TypeError):
3270 self.report_warning('There are no annotations to write.')
3271 except OSError:
3272 self.report_error('Cannot write annotations file: ' + annofn)
3273 return
3274
3275 # Write internet shortcut files
3276 def _write_link_file(link_type):
3277 url = try_get(info_dict['webpage_url'], iri_to_uri)
3278 if not url:
3279 self.report_warning(
3280 f'Cannot write internet shortcut file because the actual URL of "{info_dict["webpage_url"]}" is unknown')
3281 return True
3282 linkfn = replace_extension(self.prepare_filename(info_dict, 'link'), link_type, info_dict.get('ext'))
3283 if not self._ensure_dir_exists(encodeFilename(linkfn)):
3284 return False
3285 if self.params.get('overwrites', True) and os.path.exists(encodeFilename(linkfn)):
3286 self.to_screen(f'[info] Internet shortcut (.{link_type}) is already present')
3287 return True
3288 try:
3289 self.to_screen(f'[info] Writing internet shortcut (.{link_type}) to: {linkfn}')
3290 with open(encodeFilename(to_high_limit_path(linkfn)), 'w', encoding='utf-8',
3291 newline='\r\n' if link_type == 'url' else '\n') as linkfile:
3292 template_vars = {'url': url}
3293 if link_type == 'desktop':
3294 template_vars['filename'] = linkfn[:-(len(link_type) + 1)]
3295 linkfile.write(LINK_TEMPLATES[link_type] % template_vars)
3296 except OSError:
3297 self.report_error(f'Cannot write internet shortcut {linkfn}')
3298 return False
3299 return True
3300
3301 write_links = {
3302 'url': self.params.get('writeurllink'),
3303 'webloc': self.params.get('writewebloclink'),
3304 'desktop': self.params.get('writedesktoplink'),
3305 }
3306 if self.params.get('writelink'):
3307 link_type = ('webloc' if sys.platform == 'darwin'
3308 else 'desktop' if sys.platform.startswith('linux')
3309 else 'url')
3310 write_links[link_type] = True
3311
3312 if any(should_write and not _write_link_file(link_type)
3313 for link_type, should_write in write_links.items()):
3314 return
3315
3316 new_info, files_to_move = self.pre_process(info_dict, 'before_dl', files_to_move)
3317 replace_info_dict(new_info)
3318
3319 if self.params.get('skip_download'):
3320 info_dict['filepath'] = temp_filename
3321 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
3322 info_dict['__files_to_move'] = files_to_move
3323 replace_info_dict(self.run_pp(MoveFilesAfterDownloadPP(self, False), info_dict))
3324 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
3325 else:
3326 # Download
3327 info_dict.setdefault('__postprocessors', [])
3328 try:
3329
3330 def existing_video_file(*filepaths):
3331 ext = info_dict.get('ext')
3332 converted = lambda file: replace_extension(file, self.params.get('final_ext') or ext, ext)
3333 file = self.existing_file(itertools.chain(*zip(map(converted, filepaths), filepaths)),
3334 default_overwrite=False)
3335 if file:
3336 info_dict['ext'] = os.path.splitext(file)[1][1:]
3337 return file
3338
3339 fd, success = None, True
3340 if info_dict.get('protocol') or info_dict.get('url'):
3341 fd = get_suitable_downloader(info_dict, self.params, to_stdout=temp_filename == '-')
3342 if fd != FFmpegFD and 'no-direct-merge' not in self.params['compat_opts'] and (
3343 info_dict.get('section_start') or info_dict.get('section_end')):
3344 msg = ('This format cannot be partially downloaded' if FFmpegFD.available()
3345 else 'You have requested downloading the video partially, but ffmpeg is not installed')
3346 self.report_error(f'{msg}. Aborting')
3347 return
3348
3349 if info_dict.get('requested_formats') is not None:
3350 old_ext = info_dict['ext']
3351 if self.params.get('merge_output_format') is None:
3352 if (info_dict['ext'] == 'webm'
3353 and info_dict.get('thumbnails')
3354 # check with type instead of pp_key, __name__, or isinstance
3355 # since we dont want any custom PPs to trigger this
3356 and any(type(pp) == EmbedThumbnailPP for pp in self._pps['post_process'])): # noqa: E721
3357 info_dict['ext'] = 'mkv'
3358 self.report_warning(
3359 'webm doesn\'t support embedding a thumbnail, mkv will be used')
3360 new_ext = info_dict['ext']
3361
3362 def correct_ext(filename, ext=new_ext):
3363 if filename == '-':
3364 return filename
3365 filename_real_ext = os.path.splitext(filename)[1][1:]
3366 filename_wo_ext = (
3367 os.path.splitext(filename)[0]
3368 if filename_real_ext in (old_ext, new_ext)
3369 else filename)
3370 return f'{filename_wo_ext}.{ext}'
3371
3372 # Ensure filename always has a correct extension for successful merge
3373 full_filename = correct_ext(full_filename)
3374 temp_filename = correct_ext(temp_filename)
3375 dl_filename = existing_video_file(full_filename, temp_filename)
3376
3377 info_dict['__real_download'] = False
3378 # NOTE: Copy so that original format dicts are not modified
3379 info_dict['requested_formats'] = list(map(dict, info_dict['requested_formats']))
3380
3381 merger = FFmpegMergerPP(self)
3382 downloaded = []
3383 if dl_filename is not None:
3384 self.report_file_already_downloaded(dl_filename)
3385 elif fd:
3386 for f in info_dict['requested_formats'] if fd != FFmpegFD else []:
3387 f['filepath'] = fname = prepend_extension(
3388 correct_ext(temp_filename, info_dict['ext']),
3389 'f%s' % f['format_id'], info_dict['ext'])
3390 downloaded.append(fname)
3391 info_dict['url'] = '\n'.join(f['url'] for f in info_dict['requested_formats'])
3392 success, real_download = self.dl(temp_filename, info_dict)
3393 info_dict['__real_download'] = real_download
3394 else:
3395 if self.params.get('allow_unplayable_formats'):
3396 self.report_warning(
3397 'You have requested merging of multiple formats '
3398 'while also allowing unplayable formats to be downloaded. '
3399 'The formats won\'t be merged to prevent data corruption.')
3400 elif not merger.available:
3401 msg = 'You have requested merging of multiple formats but ffmpeg is not installed'
3402 if not self.params.get('ignoreerrors'):
3403 self.report_error(f'{msg}. Aborting due to --abort-on-error')
3404 return
3405 self.report_warning(f'{msg}. The formats won\'t be merged')
3406
3407 if temp_filename == '-':
3408 reason = ('using a downloader other than ffmpeg' if FFmpegFD.can_merge_formats(info_dict, self.params)
3409 else 'but the formats are incompatible for simultaneous download' if merger.available
3410 else 'but ffmpeg is not installed')
3411 self.report_warning(
3412 f'You have requested downloading multiple formats to stdout {reason}. '
3413 'The formats will be streamed one after the other')
3414 fname = temp_filename
3415 for f in info_dict['requested_formats']:
3416 new_info = dict(info_dict)
3417 del new_info['requested_formats']
3418 new_info.update(f)
3419 if temp_filename != '-':
3420 fname = prepend_extension(
3421 correct_ext(temp_filename, new_info['ext']),
3422 'f%s' % f['format_id'], new_info['ext'])
3423 if not self._ensure_dir_exists(fname):
3424 return
3425 f['filepath'] = fname
3426 downloaded.append(fname)
3427 partial_success, real_download = self.dl(fname, new_info)
3428 info_dict['__real_download'] = info_dict['__real_download'] or real_download
3429 success = success and partial_success
3430
3431 if downloaded and merger.available and not self.params.get('allow_unplayable_formats'):
3432 info_dict['__postprocessors'].append(merger)
3433 info_dict['__files_to_merge'] = downloaded
3434 # Even if there were no downloads, it is being merged only now
3435 info_dict['__real_download'] = True
3436 else:
3437 for file in downloaded:
3438 files_to_move[file] = None
3439 else:
3440 # Just a single file
3441 dl_filename = existing_video_file(full_filename, temp_filename)
3442 if dl_filename is None or dl_filename == temp_filename:
3443 # dl_filename == temp_filename could mean that the file was partially downloaded with --no-part.
3444 # So we should try to resume the download
3445 success, real_download = self.dl(temp_filename, info_dict)
3446 info_dict['__real_download'] = real_download
3447 else:
3448 self.report_file_already_downloaded(dl_filename)
3449
3450 dl_filename = dl_filename or temp_filename
3451 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
3452
3453 except network_exceptions as err:
3454 self.report_error('unable to download video data: %s' % error_to_compat_str(err))
3455 return
3456 except OSError as err:
3457 raise UnavailableVideoError(err)
3458 except (ContentTooShortError, ) as err:
3459 self.report_error(f'content too short (expected {err.expected} bytes and served {err.downloaded})')
3460 return
3461
3462 self._raise_pending_errors(info_dict)
3463 if success and full_filename != '-':
3464
3465 def fixup():
3466 do_fixup = True
3467 fixup_policy = self.params.get('fixup')
3468 vid = info_dict['id']
3469
3470 if fixup_policy in ('ignore', 'never'):
3471 return
3472 elif fixup_policy == 'warn':
3473 do_fixup = 'warn'
3474 elif fixup_policy != 'force':
3475 assert fixup_policy in ('detect_or_warn', None)
3476 if not info_dict.get('__real_download'):
3477 do_fixup = False
3478
3479 def ffmpeg_fixup(cndn, msg, cls):
3480 if not (do_fixup and cndn):
3481 return
3482 elif do_fixup == 'warn':
3483 self.report_warning(f'{vid}: {msg}')
3484 return
3485 pp = cls(self)
3486 if pp.available:
3487 info_dict['__postprocessors'].append(pp)
3488 else:
3489 self.report_warning(f'{vid}: {msg}. Install ffmpeg to fix this automatically')
3490
3491 stretched_ratio = info_dict.get('stretched_ratio')
3492 ffmpeg_fixup(stretched_ratio not in (1, None),
3493 f'Non-uniform pixel ratio {stretched_ratio}',
3494 FFmpegFixupStretchedPP)
3495
3496 downloader = get_suitable_downloader(info_dict, self.params) if 'protocol' in info_dict else None
3497 downloader = downloader.FD_NAME if downloader else None
3498
3499 ext = info_dict.get('ext')
3500 postprocessed_by_ffmpeg = info_dict.get('requested_formats') or any((
3501 isinstance(pp, FFmpegVideoConvertorPP)
3502 and resolve_recode_mapping(ext, pp.mapping)[0] not in (ext, None)
3503 ) for pp in self._pps['post_process'])
3504
3505 if not postprocessed_by_ffmpeg:
3506 ffmpeg_fixup(fd != FFmpegFD and ext == 'm4a'
3507 and info_dict.get('container') == 'm4a_dash',
3508 'writing DASH m4a. Only some players support this container',
3509 FFmpegFixupM4aPP)
3510 ffmpeg_fixup(downloader == 'hlsnative' and not self.params.get('hls_use_mpegts')
3511 or info_dict.get('is_live') and self.params.get('hls_use_mpegts') is None,
3512 'Possible MPEG-TS in MP4 container or malformed AAC timestamps',
3513 FFmpegFixupM3u8PP)
3514 ffmpeg_fixup(downloader == 'dashsegments'
3515 and (info_dict.get('is_live') or info_dict.get('is_dash_periods')),
3516 'Possible duplicate MOOV atoms', FFmpegFixupDuplicateMoovPP)
3517
3518 ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed timestamps detected', FFmpegFixupTimestampPP)
3519 ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed duration detected', FFmpegFixupDurationPP)
3520
3521 fixup()
3522 try:
3523 replace_info_dict(self.post_process(dl_filename, info_dict, files_to_move))
3524 except PostProcessingError as err:
3525 self.report_error('Postprocessing: %s' % str(err))
3526 return
3527 try:
3528 for ph in self._post_hooks:
3529 ph(info_dict['filepath'])
3530 except Exception as err:
3531 self.report_error('post hooks: %s' % str(err))
3532 return
3533 info_dict['__write_download_archive'] = True
3534
3535 assert info_dict is original_infodict # Make sure the info_dict was modified in-place
3536 if self.params.get('force_write_download_archive'):
3537 info_dict['__write_download_archive'] = True
3538 check_max_downloads()
3539
3540 def __download_wrapper(self, func):
3541 @functools.wraps(func)
3542 def wrapper(*args, **kwargs):
3543 try:
3544 res = func(*args, **kwargs)
3545 except UnavailableVideoError as e:
3546 self.report_error(e)
3547 except DownloadCancelled as e:
3548 self.to_screen(f'[info] {e}')
3549 if not self.params.get('break_per_url'):
3550 raise
3551 self._num_downloads = 0
3552 else:
3553 if self.params.get('dump_single_json', False):
3554 self.post_extract(res)
3555 self.to_stdout(json.dumps(self.sanitize_info(res)))
3556 return wrapper
3557
3558 def download(self, url_list):
3559 """Download a given list of URLs."""
3560 url_list = variadic(url_list) # Passing a single URL is a common mistake
3561 outtmpl = self.params['outtmpl']['default']
3562 if (len(url_list) > 1
3563 and outtmpl != '-'
3564 and '%' not in outtmpl
3565 and self.params.get('max_downloads') != 1):
3566 raise SameFileError(outtmpl)
3567
3568 for url in url_list:
3569 self.__download_wrapper(self.extract_info)(
3570 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
3571
3572 return self._download_retcode
3573
3574 def download_with_info_file(self, info_filename):
3575 with contextlib.closing(fileinput.FileInput(
3576 [info_filename], mode='r',
3577 openhook=fileinput.hook_encoded('utf-8'))) as f:
3578 # FileInput doesn't have a read method, we can't call json.load
3579 infos = [self.sanitize_info(info, self.params.get('clean_infojson', True))
3580 for info in variadic(json.loads('\n'.join(f)))]
3581 for info in infos:
3582 try:
3583 self.__download_wrapper(self.process_ie_result)(info, download=True)
3584 except (DownloadError, EntryNotInPlaylist, ReExtractInfo) as e:
3585 if not isinstance(e, EntryNotInPlaylist):
3586 self.to_stderr('\r')
3587 webpage_url = info.get('webpage_url')
3588 if webpage_url is None:
3589 raise
3590 self.report_warning(f'The info failed to download: {e}; trying with URL {webpage_url}')
3591 self.download([webpage_url])
3592 except ExtractorError as e:
3593 self.report_error(e)
3594 return self._download_retcode
3595
3596 @staticmethod
3597 def sanitize_info(info_dict, remove_private_keys=False):
3598 ''' Sanitize the infodict for converting to json '''
3599 if info_dict is None:
3600 return info_dict
3601 info_dict.setdefault('epoch', int(time.time()))
3602 info_dict.setdefault('_type', 'video')
3603 info_dict.setdefault('_version', {
3604 'version': __version__,
3605 'current_git_head': current_git_head(),
3606 'release_git_head': RELEASE_GIT_HEAD,
3607 'repository': ORIGIN,
3608 })
3609
3610 if remove_private_keys:
3611 reject = lambda k, v: v is None or k.startswith('__') or k in {
3612 'requested_downloads', 'requested_formats', 'requested_subtitles', 'requested_entries',
3613 'entries', 'filepath', '_filename', 'filename', 'infojson_filename', 'original_url',
3614 'playlist_autonumber',
3615 }
3616 else:
3617 reject = lambda k, v: False
3618
3619 def filter_fn(obj):
3620 if isinstance(obj, dict):
3621 return {k: filter_fn(v) for k, v in obj.items() if not reject(k, v)}
3622 elif isinstance(obj, (list, tuple, set, LazyList)):
3623 return list(map(filter_fn, obj))
3624 elif obj is None or isinstance(obj, (str, int, float, bool)):
3625 return obj
3626 else:
3627 return repr(obj)
3628
3629 return filter_fn(info_dict)
3630
3631 @staticmethod
3632 def filter_requested_info(info_dict, actually_filter=True):
3633 ''' Alias of sanitize_info for backward compatibility '''
3634 return YoutubeDL.sanitize_info(info_dict, actually_filter)
3635
3636 def _delete_downloaded_files(self, *files_to_delete, info={}, msg=None):
3637 for filename in set(filter(None, files_to_delete)):
3638 if msg:
3639 self.to_screen(msg % filename)
3640 try:
3641 os.remove(filename)
3642 except OSError:
3643 self.report_warning(f'Unable to delete file {filename}')
3644 if filename in info.get('__files_to_move', []): # NB: Delete even if None
3645 del info['__files_to_move'][filename]
3646
3647 @staticmethod
3648 def post_extract(info_dict):
3649 def actual_post_extract(info_dict):
3650 if info_dict.get('_type') in ('playlist', 'multi_video'):
3651 for video_dict in info_dict.get('entries', {}):
3652 actual_post_extract(video_dict or {})
3653 return
3654
3655 post_extractor = info_dict.pop('__post_extractor', None) or (lambda: {})
3656 info_dict.update(post_extractor())
3657
3658 actual_post_extract(info_dict or {})
3659
3660 def run_pp(self, pp, infodict):
3661 files_to_delete = []
3662 if '__files_to_move' not in infodict:
3663 infodict['__files_to_move'] = {}
3664 try:
3665 files_to_delete, infodict = pp.run(infodict)
3666 except PostProcessingError as e:
3667 # Must be True and not 'only_download'
3668 if self.params.get('ignoreerrors') is True:
3669 self.report_error(e)
3670 return infodict
3671 raise
3672
3673 if not files_to_delete:
3674 return infodict
3675 if self.params.get('keepvideo', False):
3676 for f in files_to_delete:
3677 infodict['__files_to_move'].setdefault(f, '')
3678 else:
3679 self._delete_downloaded_files(
3680 *files_to_delete, info=infodict, msg='Deleting original file %s (pass -k to keep)')
3681 return infodict
3682
3683 def run_all_pps(self, key, info, *, additional_pps=None):
3684 if key != 'video':
3685 self._forceprint(key, info)
3686 for pp in (additional_pps or []) + self._pps[key]:
3687 info = self.run_pp(pp, info)
3688 return info
3689
3690 def pre_process(self, ie_info, key='pre_process', files_to_move=None):
3691 info = dict(ie_info)
3692 info['__files_to_move'] = files_to_move or {}
3693 try:
3694 info = self.run_all_pps(key, info)
3695 except PostProcessingError as err:
3696 msg = f'Preprocessing: {err}'
3697 info.setdefault('__pending_error', msg)
3698 self.report_error(msg, is_error=False)
3699 return info, info.pop('__files_to_move', None)
3700
3701 def post_process(self, filename, info, files_to_move=None):
3702 """Run all the postprocessors on the given file."""
3703 info['filepath'] = filename
3704 info['__files_to_move'] = files_to_move or {}
3705 info = self.run_all_pps('post_process', info, additional_pps=info.get('__postprocessors'))
3706 info = self.run_pp(MoveFilesAfterDownloadPP(self), info)
3707 del info['__files_to_move']
3708 return self.run_all_pps('after_move', info)
3709
3710 def _make_archive_id(self, info_dict):
3711 video_id = info_dict.get('id')
3712 if not video_id:
3713 return
3714 # Future-proof against any change in case
3715 # and backwards compatibility with prior versions
3716 extractor = info_dict.get('extractor_key') or info_dict.get('ie_key') # key in a playlist
3717 if extractor is None:
3718 url = str_or_none(info_dict.get('url'))
3719 if not url:
3720 return
3721 # Try to find matching extractor for the URL and take its ie_key
3722 for ie_key, ie in self._ies.items():
3723 if ie.suitable(url):
3724 extractor = ie_key
3725 break
3726 else:
3727 return
3728 return make_archive_id(extractor, video_id)
3729
3730 def in_download_archive(self, info_dict):
3731 if not self.archive:
3732 return False
3733
3734 vid_ids = [self._make_archive_id(info_dict)]
3735 vid_ids.extend(info_dict.get('_old_archive_ids') or [])
3736 return any(id_ in self.archive for id_ in vid_ids)
3737
3738 def record_download_archive(self, info_dict):
3739 fn = self.params.get('download_archive')
3740 if fn is None:
3741 return
3742 vid_id = self._make_archive_id(info_dict)
3743 assert vid_id
3744
3745 self.write_debug(f'Adding to archive: {vid_id}')
3746 if is_path_like(fn):
3747 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
3748 archive_file.write(vid_id + '\n')
3749 self.archive.add(vid_id)
3750
3751 @staticmethod
3752 def format_resolution(format, default='unknown'):
3753 if format.get('vcodec') == 'none' and format.get('acodec') != 'none':
3754 return 'audio only'
3755 if format.get('resolution') is not None:
3756 return format['resolution']
3757 if format.get('width') and format.get('height'):
3758 return '%dx%d' % (format['width'], format['height'])
3759 elif format.get('height'):
3760 return '%sp' % format['height']
3761 elif format.get('width'):
3762 return '%dx?' % format['width']
3763 return default
3764
3765 def _list_format_headers(self, *headers):
3766 if self.params.get('listformats_table', True) is not False:
3767 return [self._format_out(header, self.Styles.HEADERS) for header in headers]
3768 return headers
3769
3770 def _format_note(self, fdict):
3771 res = ''
3772 if fdict.get('ext') in ['f4f', 'f4m']:
3773 res += '(unsupported)'
3774 if fdict.get('language'):
3775 if res:
3776 res += ' '
3777 res += '[%s]' % fdict['language']
3778 if fdict.get('format_note') is not None:
3779 if res:
3780 res += ' '
3781 res += fdict['format_note']
3782 if fdict.get('tbr') is not None:
3783 if res:
3784 res += ', '
3785 res += '%4dk' % fdict['tbr']
3786 if fdict.get('container') is not None:
3787 if res:
3788 res += ', '
3789 res += '%s container' % fdict['container']
3790 if (fdict.get('vcodec') is not None
3791 and fdict.get('vcodec') != 'none'):
3792 if res:
3793 res += ', '
3794 res += fdict['vcodec']
3795 if fdict.get('vbr') is not None:
3796 res += '@'
3797 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
3798 res += 'video@'
3799 if fdict.get('vbr') is not None:
3800 res += '%4dk' % fdict['vbr']
3801 if fdict.get('fps') is not None:
3802 if res:
3803 res += ', '
3804 res += '%sfps' % fdict['fps']
3805 if fdict.get('acodec') is not None:
3806 if res:
3807 res += ', '
3808 if fdict['acodec'] == 'none':
3809 res += 'video only'
3810 else:
3811 res += '%-5s' % fdict['acodec']
3812 elif fdict.get('abr') is not None:
3813 if res:
3814 res += ', '
3815 res += 'audio'
3816 if fdict.get('abr') is not None:
3817 res += '@%3dk' % fdict['abr']
3818 if fdict.get('asr') is not None:
3819 res += ' (%5dHz)' % fdict['asr']
3820 if fdict.get('filesize') is not None:
3821 if res:
3822 res += ', '
3823 res += format_bytes(fdict['filesize'])
3824 elif fdict.get('filesize_approx') is not None:
3825 if res:
3826 res += ', '
3827 res += '~' + format_bytes(fdict['filesize_approx'])
3828 return res
3829
3830 def _get_formats(self, info_dict):
3831 if info_dict.get('formats') is None:
3832 if info_dict.get('url') and info_dict.get('_type', 'video') == 'video':
3833 return [info_dict]
3834 return []
3835 return info_dict['formats']
3836
3837 def render_formats_table(self, info_dict):
3838 formats = self._get_formats(info_dict)
3839 if not formats:
3840 return
3841 if not self.params.get('listformats_table', True) is not False:
3842 table = [
3843 [
3844 format_field(f, 'format_id'),
3845 format_field(f, 'ext'),
3846 self.format_resolution(f),
3847 self._format_note(f)
3848 ] for f in formats if (f.get('preference') or 0) >= -1000]
3849 return render_table(['format code', 'extension', 'resolution', 'note'], table, extra_gap=1)
3850
3851 def simplified_codec(f, field):
3852 assert field in ('acodec', 'vcodec')
3853 codec = f.get(field)
3854 if not codec:
3855 return 'unknown'
3856 elif codec != 'none':
3857 return '.'.join(codec.split('.')[:4])
3858
3859 if field == 'vcodec' and f.get('acodec') == 'none':
3860 return 'images'
3861 elif field == 'acodec' and f.get('vcodec') == 'none':
3862 return ''
3863 return self._format_out('audio only' if field == 'vcodec' else 'video only',
3864 self.Styles.SUPPRESS)
3865
3866 delim = self._format_out('\u2502', self.Styles.DELIM, '|', test_encoding=True)
3867 table = [
3868 [
3869 self._format_out(format_field(f, 'format_id'), self.Styles.ID),
3870 format_field(f, 'ext'),
3871 format_field(f, func=self.format_resolution, ignore=('audio only', 'images')),
3872 format_field(f, 'fps', '\t%d', func=round),
3873 format_field(f, 'dynamic_range', '%s', ignore=(None, 'SDR')).replace('HDR', ''),
3874 format_field(f, 'audio_channels', '\t%s'),
3875 delim, (
3876 format_field(f, 'filesize', ' \t%s', func=format_bytes)
3877 or format_field(f, 'filesize_approx', '≈\t%s', func=format_bytes)
3878 or format_field(try_call(lambda: format_bytes(int(info_dict['duration'] * f['tbr'] * (1024 / 8)))),
3879 None, self._format_out('~\t%s', self.Styles.SUPPRESS))),
3880 format_field(f, 'tbr', '\t%dk', func=round),
3881 shorten_protocol_name(f.get('protocol', '')),
3882 delim,
3883 simplified_codec(f, 'vcodec'),
3884 format_field(f, 'vbr', '\t%dk', func=round),
3885 simplified_codec(f, 'acodec'),
3886 format_field(f, 'abr', '\t%dk', func=round),
3887 format_field(f, 'asr', '\t%s', func=format_decimal_suffix),
3888 join_nonempty(format_field(f, 'language', '[%s]'), join_nonempty(
3889 self._format_out('UNSUPPORTED', self.Styles.BAD_FORMAT) if f.get('ext') in ('f4f', 'f4m') else None,
3890 (self._format_out('Maybe DRM', self.Styles.WARNING) if f.get('has_drm') == 'maybe'
3891 else self._format_out('DRM', self.Styles.BAD_FORMAT) if f.get('has_drm') else None),
3892 format_field(f, 'format_note'),
3893 format_field(f, 'container', ignore=(None, f.get('ext'))),
3894 delim=', '), delim=' '),
3895 ] for f in formats if f.get('preference') is None or f['preference'] >= -1000]
3896 header_line = self._list_format_headers(
3897 'ID', 'EXT', 'RESOLUTION', '\tFPS', 'HDR', 'CH', delim, '\tFILESIZE', '\tTBR', 'PROTO',
3898 delim, 'VCODEC', '\tVBR', 'ACODEC', '\tABR', '\tASR', 'MORE INFO')
3899
3900 return render_table(
3901 header_line, table, hide_empty=True,
3902 delim=self._format_out('\u2500', self.Styles.DELIM, '-', test_encoding=True))
3903
3904 def render_thumbnails_table(self, info_dict):
3905 thumbnails = list(info_dict.get('thumbnails') or [])
3906 if not thumbnails:
3907 return None
3908 return render_table(
3909 self._list_format_headers('ID', 'Width', 'Height', 'URL'),
3910 [[t.get('id'), t.get('width') or 'unknown', t.get('height') or 'unknown', t['url']] for t in thumbnails])
3911
3912 def render_subtitles_table(self, video_id, subtitles):
3913 def _row(lang, formats):
3914 exts, names = zip(*((f['ext'], f.get('name') or 'unknown') for f in reversed(formats)))
3915 if len(set(names)) == 1:
3916 names = [] if names[0] == 'unknown' else names[:1]
3917 return [lang, ', '.join(names), ', '.join(exts)]
3918
3919 if not subtitles:
3920 return None
3921 return render_table(
3922 self._list_format_headers('Language', 'Name', 'Formats'),
3923 [_row(lang, formats) for lang, formats in subtitles.items()],
3924 hide_empty=True)
3925
3926 def __list_table(self, video_id, name, func, *args):
3927 table = func(*args)
3928 if not table:
3929 self.to_screen(f'{video_id} has no {name}')
3930 return
3931 self.to_screen(f'[info] Available {name} for {video_id}:')
3932 self.to_stdout(table)
3933
3934 def list_formats(self, info_dict):
3935 self.__list_table(info_dict['id'], 'formats', self.render_formats_table, info_dict)
3936
3937 def list_thumbnails(self, info_dict):
3938 self.__list_table(info_dict['id'], 'thumbnails', self.render_thumbnails_table, info_dict)
3939
3940 def list_subtitles(self, video_id, subtitles, name='subtitles'):
3941 self.__list_table(video_id, name, self.render_subtitles_table, video_id, subtitles)
3942
3943 def print_debug_header(self):
3944 if not self.params.get('verbose'):
3945 return
3946
3947 from . import _IN_CLI # Must be delayed import
3948
3949 # These imports can be slow. So import them only as needed
3950 from .extractor.extractors import _LAZY_LOADER
3951 from .extractor.extractors import (
3952 _PLUGIN_CLASSES as plugin_ies,
3953 _PLUGIN_OVERRIDES as plugin_ie_overrides
3954 )
3955
3956 def get_encoding(stream):
3957 ret = str(getattr(stream, 'encoding', 'missing (%s)' % type(stream).__name__))
3958 additional_info = []
3959 if os.environ.get('TERM', '').lower() == 'dumb':
3960 additional_info.append('dumb')
3961 if not supports_terminal_sequences(stream):
3962 from .utils import WINDOWS_VT_MODE # Must be imported locally
3963 additional_info.append('No VT' if WINDOWS_VT_MODE is False else 'No ANSI')
3964 if additional_info:
3965 ret = f'{ret} ({",".join(additional_info)})'
3966 return ret
3967
3968 encoding_str = 'Encodings: locale %s, fs %s, pref %s, %s' % (
3969 locale.getpreferredencoding(),
3970 sys.getfilesystemencoding(),
3971 self.get_encoding(),
3972 ', '.join(
3973 f'{key} {get_encoding(stream)}' for key, stream in self._out_files.items_
3974 if stream is not None and key != 'console')
3975 )
3976
3977 logger = self.params.get('logger')
3978 if logger:
3979 write_debug = lambda msg: logger.debug(f'[debug] {msg}')
3980 write_debug(encoding_str)
3981 else:
3982 write_string(f'[debug] {encoding_str}\n', encoding=None)
3983 write_debug = lambda msg: self._write_string(f'[debug] {msg}\n')
3984
3985 source = detect_variant()
3986 if VARIANT not in (None, 'pip'):
3987 source += '*'
3988 klass = type(self)
3989 write_debug(join_nonempty(
3990 f'{REPOSITORY.rpartition("/")[2]} version',
3991 _make_label(ORIGIN, CHANNEL.partition('@')[2] or __version__, __version__),
3992 f'[{RELEASE_GIT_HEAD[:9]}]' if RELEASE_GIT_HEAD else '',
3993 '' if source == 'unknown' else f'({source})',
3994 '' if _IN_CLI else 'API' if klass == YoutubeDL else f'API:{self.__module__}.{klass.__qualname__}',
3995 delim=' '))
3996
3997 if not _IN_CLI:
3998 write_debug(f'params: {self.params}')
3999
4000 if not _LAZY_LOADER:
4001 if os.environ.get('YTDLP_NO_LAZY_EXTRACTORS'):
4002 write_debug('Lazy loading extractors is forcibly disabled')
4003 else:
4004 write_debug('Lazy loading extractors is disabled')
4005 if self.params['compat_opts']:
4006 write_debug('Compatibility options: %s' % ', '.join(self.params['compat_opts']))
4007
4008 if current_git_head():
4009 write_debug(f'Git HEAD: {current_git_head()}')
4010 write_debug(system_identifier())
4011
4012 exe_versions, ffmpeg_features = FFmpegPostProcessor.get_versions_and_features(self)
4013 ffmpeg_features = {key for key, val in ffmpeg_features.items() if val}
4014 if ffmpeg_features:
4015 exe_versions['ffmpeg'] += ' (%s)' % ','.join(sorted(ffmpeg_features))
4016
4017 exe_versions['rtmpdump'] = rtmpdump_version()
4018 exe_versions['phantomjs'] = PhantomJSwrapper._version()
4019 exe_str = ', '.join(
4020 f'{exe} {v}' for exe, v in sorted(exe_versions.items()) if v
4021 ) or 'none'
4022 write_debug('exe versions: %s' % exe_str)
4023
4024 from .compat.compat_utils import get_package_info
4025 from .dependencies import available_dependencies
4026
4027 write_debug('Optional libraries: %s' % (', '.join(sorted({
4028 join_nonempty(*get_package_info(m)) for m in available_dependencies.values()
4029 })) or 'none'))
4030
4031 write_debug(f'Proxy map: {self.proxies}')
4032 write_debug(f'Request Handlers: {", ".join(rh.RH_NAME for rh in self._request_director.handlers.values())}')
4033 for plugin_type, plugins in {'Extractor': plugin_ies, 'Post-Processor': plugin_pps}.items():
4034 display_list = ['%s%s' % (
4035 klass.__name__, '' if klass.__name__ == name else f' as {name}')
4036 for name, klass in plugins.items()]
4037 if plugin_type == 'Extractor':
4038 display_list.extend(f'{plugins[-1].IE_NAME.partition("+")[2]} ({parent.__name__})'
4039 for parent, plugins in plugin_ie_overrides.items())
4040 if not display_list:
4041 continue
4042 write_debug(f'{plugin_type} Plugins: {", ".join(sorted(display_list))}')
4043
4044 plugin_dirs = plugin_directories()
4045 if plugin_dirs:
4046 write_debug(f'Plugin directories: {plugin_dirs}')
4047
4048 # Not implemented
4049 if False and self.params.get('call_home'):
4050 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode()
4051 write_debug('Public IP address: %s' % ipaddr)
4052 latest_version = self.urlopen(
4053 'https://yt-dl.org/latest/version').read().decode()
4054 if version_tuple(latest_version) > version_tuple(__version__):
4055 self.report_warning(
4056 'You are using an outdated version (newest version: %s)! '
4057 'See https://yt-dl.org/update if you need help updating.' %
4058 latest_version)
4059
4060 @functools.cached_property
4061 def proxies(self):
4062 """Global proxy configuration"""
4063 opts_proxy = self.params.get('proxy')
4064 if opts_proxy is not None:
4065 if opts_proxy == '':
4066 opts_proxy = '__noproxy__'
4067 proxies = {'all': opts_proxy}
4068 else:
4069 proxies = urllib.request.getproxies()
4070 # compat. Set HTTPS_PROXY to __noproxy__ to revert
4071 if 'http' in proxies and 'https' not in proxies:
4072 proxies['https'] = proxies['http']
4073
4074 return proxies
4075
4076 @functools.cached_property
4077 def cookiejar(self):
4078 """Global cookiejar instance"""
4079 return load_cookies(
4080 self.params.get('cookiefile'), self.params.get('cookiesfrombrowser'), self)
4081
4082 @property
4083 def _opener(self):
4084 """
4085 Get a urllib OpenerDirector from the Urllib handler (deprecated).
4086 """
4087 self.deprecation_warning('YoutubeDL._opener is deprecated, use YoutubeDL.urlopen()')
4088 handler = self._request_director.handlers['Urllib']
4089 return handler._get_instance(cookiejar=self.cookiejar, proxies=self.proxies)
4090
4091 def _get_available_impersonate_targets(self):
4092 # todo(future): make available as public API
4093 return [
4094 (target, rh.RH_NAME)
4095 for rh in self._request_director.handlers.values()
4096 if isinstance(rh, ImpersonateRequestHandler)
4097 for target in rh.supported_targets
4098 ]
4099
4100 def _impersonate_target_available(self, target):
4101 # todo(future): make available as public API
4102 return any(
4103 rh.is_supported_target(target)
4104 for rh in self._request_director.handlers.values()
4105 if isinstance(rh, ImpersonateRequestHandler))
4106
4107 def urlopen(self, req):
4108 """ Start an HTTP download """
4109 if isinstance(req, str):
4110 req = Request(req)
4111 elif isinstance(req, urllib.request.Request):
4112 self.deprecation_warning(
4113 'Passing a urllib.request.Request object to YoutubeDL.urlopen() is deprecated. '
4114 'Use yt_dlp.networking.common.Request instead.')
4115 req = urllib_req_to_req(req)
4116 assert isinstance(req, Request)
4117
4118 # compat: Assume user:pass url params are basic auth
4119 url, basic_auth_header = extract_basic_auth(req.url)
4120 if basic_auth_header:
4121 req.headers['Authorization'] = basic_auth_header
4122 req.url = sanitize_url(url)
4123
4124 clean_proxies(proxies=req.proxies, headers=req.headers)
4125 clean_headers(req.headers)
4126
4127 try:
4128 return self._request_director.send(req)
4129 except NoSupportingHandlers as e:
4130 for ue in e.unsupported_errors:
4131 # FIXME: This depends on the order of errors.
4132 if not (ue.handler and ue.msg):
4133 continue
4134 if ue.handler.RH_KEY == 'Urllib' and 'unsupported url scheme: "file"' in ue.msg.lower():
4135 raise RequestError(
4136 'file:// URLs are disabled by default in yt-dlp for security reasons. '
4137 'Use --enable-file-urls to enable at your own risk.', cause=ue) from ue
4138 if (
4139 'unsupported proxy type: "https"' in ue.msg.lower()
4140 and 'requests' not in self._request_director.handlers
4141 and 'curl_cffi' not in self._request_director.handlers
4142 ):
4143 raise RequestError(
4144 'To use an HTTPS proxy for this request, one of the following dependencies needs to be installed: requests, curl_cffi')
4145
4146 elif (
4147 re.match(r'unsupported url scheme: "wss?"', ue.msg.lower())
4148 and 'websockets' not in self._request_director.handlers
4149 ):
4150 raise RequestError(
4151 'This request requires WebSocket support. '
4152 'Ensure one of the following dependencies are installed: websockets',
4153 cause=ue) from ue
4154
4155 elif re.match(r'unsupported (?:extensions: impersonate|impersonate target)', ue.msg.lower()):
4156 raise RequestError(
4157 f'Impersonate target "{req.extensions["impersonate"]}" is not available.'
4158 f' See --list-impersonate-targets for available targets.'
4159 f' This request requires browser impersonation, however you may be missing dependencies'
4160 f' required to support this target.')
4161 raise
4162 except SSLError as e:
4163 if 'UNSAFE_LEGACY_RENEGOTIATION_DISABLED' in str(e):
4164 raise RequestError('UNSAFE_LEGACY_RENEGOTIATION_DISABLED: Try using --legacy-server-connect', cause=e) from e
4165 elif 'SSLV3_ALERT_HANDSHAKE_FAILURE' in str(e):
4166 raise RequestError(
4167 'SSLV3_ALERT_HANDSHAKE_FAILURE: The server may not support the current cipher list. '
4168 'Try using --legacy-server-connect', cause=e) from e
4169 raise
4170
4171 def build_request_director(self, handlers, preferences=None):
4172 logger = _YDLLogger(self)
4173 headers = self.params['http_headers'].copy()
4174 proxies = self.proxies.copy()
4175 clean_headers(headers)
4176 clean_proxies(proxies, headers)
4177
4178 director = RequestDirector(logger=logger, verbose=self.params.get('debug_printtraffic'))
4179 for handler in handlers:
4180 director.add_handler(handler(
4181 logger=logger,
4182 headers=headers,
4183 cookiejar=self.cookiejar,
4184 proxies=proxies,
4185 prefer_system_certs='no-certifi' in self.params['compat_opts'],
4186 verify=not self.params.get('nocheckcertificate'),
4187 **traverse_obj(self.params, {
4188 'verbose': 'debug_printtraffic',
4189 'source_address': 'source_address',
4190 'timeout': 'socket_timeout',
4191 'legacy_ssl_support': 'legacyserverconnect',
4192 'enable_file_urls': 'enable_file_urls',
4193 'impersonate': 'impersonate',
4194 'client_cert': {
4195 'client_certificate': 'client_certificate',
4196 'client_certificate_key': 'client_certificate_key',
4197 'client_certificate_password': 'client_certificate_password',
4198 },
4199 }),
4200 ))
4201 director.preferences.update(preferences or [])
4202 if 'prefer-legacy-http-handler' in self.params['compat_opts']:
4203 director.preferences.add(lambda rh, _: 500 if rh.RH_KEY == 'Urllib' else 0)
4204 return director
4205
4206 @functools.cached_property
4207 def _request_director(self):
4208 return self.build_request_director(_REQUEST_HANDLERS.values(), _RH_PREFERENCES)
4209
4210 def encode(self, s):
4211 if isinstance(s, bytes):
4212 return s # Already encoded
4213
4214 try:
4215 return s.encode(self.get_encoding())
4216 except UnicodeEncodeError as err:
4217 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
4218 raise
4219
4220 def get_encoding(self):
4221 encoding = self.params.get('encoding')
4222 if encoding is None:
4223 encoding = preferredencoding()
4224 return encoding
4225
4226 def _write_info_json(self, label, ie_result, infofn, overwrite=None):
4227 ''' Write infojson and returns True = written, 'exists' = Already exists, False = skip, None = error '''
4228 if overwrite is None:
4229 overwrite = self.params.get('overwrites', True)
4230 if not self.params.get('writeinfojson'):
4231 return False
4232 elif not infofn:
4233 self.write_debug(f'Skipping writing {label} infojson')
4234 return False
4235 elif not self._ensure_dir_exists(infofn):
4236 return None
4237 elif not overwrite and os.path.exists(infofn):
4238 self.to_screen(f'[info] {label.title()} metadata is already present')
4239 return 'exists'
4240
4241 self.to_screen(f'[info] Writing {label} metadata as JSON to: {infofn}')
4242 try:
4243 write_json_file(self.sanitize_info(ie_result, self.params.get('clean_infojson', True)), infofn)
4244 return True
4245 except OSError:
4246 self.report_error(f'Cannot write {label} metadata to JSON file {infofn}')
4247 return None
4248
4249 def _write_description(self, label, ie_result, descfn):
4250 ''' Write description and returns True = written, False = skip, None = error '''
4251 if not self.params.get('writedescription'):
4252 return False
4253 elif not descfn:
4254 self.write_debug(f'Skipping writing {label} description')
4255 return False
4256 elif not self._ensure_dir_exists(descfn):
4257 return None
4258 elif not self.params.get('overwrites', True) and os.path.exists(descfn):
4259 self.to_screen(f'[info] {label.title()} description is already present')
4260 elif ie_result.get('description') is None:
4261 self.to_screen(f'[info] There\'s no {label} description to write')
4262 return False
4263 else:
4264 try:
4265 self.to_screen(f'[info] Writing {label} description to: {descfn}')
4266 with open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
4267 descfile.write(ie_result['description'])
4268 except OSError:
4269 self.report_error(f'Cannot write {label} description file {descfn}')
4270 return None
4271 return True
4272
4273 def _write_subtitles(self, info_dict, filename):
4274 ''' Write subtitles to file and return list of (sub_filename, final_sub_filename); or None if error'''
4275 ret = []
4276 subtitles = info_dict.get('requested_subtitles')
4277 if not (self.params.get('writesubtitles') or self.params.get('writeautomaticsub')):
4278 # subtitles download errors are already managed as troubles in relevant IE
4279 # that way it will silently go on when used with unsupporting IE
4280 return ret
4281 elif not subtitles:
4282 self.to_screen('[info] There are no subtitles for the requested languages')
4283 return ret
4284 sub_filename_base = self.prepare_filename(info_dict, 'subtitle')
4285 if not sub_filename_base:
4286 self.to_screen('[info] Skipping writing video subtitles')
4287 return ret
4288
4289 for sub_lang, sub_info in subtitles.items():
4290 sub_format = sub_info['ext']
4291 sub_filename = subtitles_filename(filename, sub_lang, sub_format, info_dict.get('ext'))
4292 sub_filename_final = subtitles_filename(sub_filename_base, sub_lang, sub_format, info_dict.get('ext'))
4293 existing_sub = self.existing_file((sub_filename_final, sub_filename))
4294 if existing_sub:
4295 self.to_screen(f'[info] Video subtitle {sub_lang}.{sub_format} is already present')
4296 sub_info['filepath'] = existing_sub
4297 ret.append((existing_sub, sub_filename_final))
4298 continue
4299
4300 self.to_screen(f'[info] Writing video subtitles to: {sub_filename}')
4301 if sub_info.get('data') is not None:
4302 try:
4303 # Use newline='' to prevent conversion of newline characters
4304 # See https://github.com/ytdl-org/youtube-dl/issues/10268
4305 with open(sub_filename, 'w', encoding='utf-8', newline='') as subfile:
4306 subfile.write(sub_info['data'])
4307 sub_info['filepath'] = sub_filename
4308 ret.append((sub_filename, sub_filename_final))
4309 continue
4310 except OSError:
4311 self.report_error(f'Cannot write video subtitles file {sub_filename}')
4312 return None
4313
4314 try:
4315 sub_copy = sub_info.copy()
4316 sub_copy.setdefault('http_headers', info_dict.get('http_headers'))
4317 self.dl(sub_filename, sub_copy, subtitle=True)
4318 sub_info['filepath'] = sub_filename
4319 ret.append((sub_filename, sub_filename_final))
4320 except (DownloadError, ExtractorError, IOError, OSError, ValueError) + network_exceptions as err:
4321 msg = f'Unable to download video subtitles for {sub_lang!r}: {err}'
4322 if self.params.get('ignoreerrors') is not True: # False or 'only_download'
4323 if not self.params.get('ignoreerrors'):
4324 self.report_error(msg)
4325 raise DownloadError(msg)
4326 self.report_warning(msg)
4327 return ret
4328
4329 def _write_thumbnails(self, label, info_dict, filename, thumb_filename_base=None):
4330 ''' Write thumbnails to file and return list of (thumb_filename, final_thumb_filename); or None if error '''
4331 write_all = self.params.get('write_all_thumbnails', False)
4332 thumbnails, ret = [], []
4333 if write_all or self.params.get('writethumbnail', False):
4334 thumbnails = info_dict.get('thumbnails') or []
4335 if not thumbnails:
4336 self.to_screen(f'[info] There are no {label} thumbnails to download')
4337 return ret
4338 multiple = write_all and len(thumbnails) > 1
4339
4340 if thumb_filename_base is None:
4341 thumb_filename_base = filename
4342 if thumbnails and not thumb_filename_base:
4343 self.write_debug(f'Skipping writing {label} thumbnail')
4344 return ret
4345
4346 if thumbnails and not self._ensure_dir_exists(filename):
4347 return None
4348
4349 for idx, t in list(enumerate(thumbnails))[::-1]:
4350 thumb_ext = (f'{t["id"]}.' if multiple else '') + determine_ext(t['url'], 'jpg')
4351 thumb_display_id = f'{label} thumbnail {t["id"]}'
4352 thumb_filename = replace_extension(filename, thumb_ext, info_dict.get('ext'))
4353 thumb_filename_final = replace_extension(thumb_filename_base, thumb_ext, info_dict.get('ext'))
4354
4355 existing_thumb = self.existing_file((thumb_filename_final, thumb_filename))
4356 if existing_thumb:
4357 self.to_screen('[info] %s is already present' % (
4358 thumb_display_id if multiple else f'{label} thumbnail').capitalize())
4359 t['filepath'] = existing_thumb
4360 ret.append((existing_thumb, thumb_filename_final))
4361 else:
4362 self.to_screen(f'[info] Downloading {thumb_display_id} ...')
4363 try:
4364 uf = self.urlopen(Request(t['url'], headers=t.get('http_headers', {})))
4365 self.to_screen(f'[info] Writing {thumb_display_id} to: {thumb_filename}')
4366 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
4367 shutil.copyfileobj(uf, thumbf)
4368 ret.append((thumb_filename, thumb_filename_final))
4369 t['filepath'] = thumb_filename
4370 except network_exceptions as err:
4371 if isinstance(err, HTTPError) and err.status == 404:
4372 self.to_screen(f'[info] {thumb_display_id.title()} does not exist')
4373 else:
4374 self.report_warning(f'Unable to download {thumb_display_id}: {err}')
4375 thumbnails.pop(idx)
4376 if ret and not write_all:
4377 break
4378 return ret