]> jfr.im git - yt-dlp.git/blob - yt_dlp/YoutubeDL.py
[core] Handle `--load-info-json` format selection errors (#9392)
[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 .plugins import directories as plugin_directories
46 from .postprocessor import _PLUGIN_CLASSES as plugin_pps
47 from .postprocessor import (
48 EmbedThumbnailPP,
49 FFmpegFixupDuplicateMoovPP,
50 FFmpegFixupDurationPP,
51 FFmpegFixupM3u8PP,
52 FFmpegFixupM4aPP,
53 FFmpegFixupStretchedPP,
54 FFmpegFixupTimestampPP,
55 FFmpegMergerPP,
56 FFmpegPostProcessor,
57 FFmpegVideoConvertorPP,
58 MoveFilesAfterDownloadPP,
59 get_postprocessor,
60 )
61 from .postprocessor.ffmpeg import resolve_mapping as resolve_recode_mapping
62 from .update import (
63 REPOSITORY,
64 _get_system_deprecation,
65 _make_label,
66 current_git_head,
67 detect_variant,
68 )
69 from .utils import (
70 DEFAULT_OUTTMPL,
71 IDENTITY,
72 LINK_TEMPLATES,
73 MEDIA_EXTENSIONS,
74 NO_DEFAULT,
75 NUMBER_RE,
76 OUTTMPL_TYPES,
77 POSTPROCESS_WHEN,
78 STR_FORMAT_RE_TMPL,
79 STR_FORMAT_TYPES,
80 ContentTooShortError,
81 DateRange,
82 DownloadCancelled,
83 DownloadError,
84 EntryNotInPlaylist,
85 ExistingVideoReached,
86 ExtractorError,
87 FormatSorter,
88 GeoRestrictedError,
89 ISO3166Utils,
90 LazyList,
91 MaxDownloadsReached,
92 Namespace,
93 PagedList,
94 PlaylistEntries,
95 Popen,
96 PostProcessingError,
97 ReExtractInfo,
98 RejectedVideoReached,
99 SameFileError,
100 UnavailableVideoError,
101 UserNotLive,
102 age_restricted,
103 args_to_str,
104 bug_reports_message,
105 date_from_str,
106 deprecation_warning,
107 determine_ext,
108 determine_protocol,
109 encode_compat_str,
110 encodeFilename,
111 error_to_compat_str,
112 escapeHTML,
113 expand_path,
114 extract_basic_auth,
115 filter_dict,
116 float_or_none,
117 format_bytes,
118 format_decimal_suffix,
119 format_field,
120 formatSeconds,
121 get_compatible_ext,
122 get_domain,
123 int_or_none,
124 iri_to_uri,
125 is_path_like,
126 join_nonempty,
127 locked_file,
128 make_archive_id,
129 make_dir,
130 number_of_digits,
131 orderedSet,
132 orderedSet_from_options,
133 parse_filesize,
134 preferredencoding,
135 prepend_extension,
136 remove_terminal_sequences,
137 render_table,
138 replace_extension,
139 sanitize_filename,
140 sanitize_path,
141 sanitize_url,
142 str_or_none,
143 strftime_or_none,
144 subtitles_filename,
145 supports_terminal_sequences,
146 system_identifier,
147 timetuple_from_msec,
148 to_high_limit_path,
149 traverse_obj,
150 try_call,
151 try_get,
152 url_basename,
153 variadic,
154 version_tuple,
155 windows_enable_vt_mode,
156 write_json_file,
157 write_string,
158 )
159 from .utils._utils import _YDLLogger
160 from .utils.networking import (
161 HTTPHeaderDict,
162 clean_headers,
163 clean_proxies,
164 std_headers,
165 )
166 from .version import CHANNEL, ORIGIN, RELEASE_GIT_HEAD, VARIANT, __version__
167
168 if compat_os_name == 'nt':
169 import ctypes
170
171
172 class YoutubeDL:
173 """YoutubeDL class.
174
175 YoutubeDL objects are the ones responsible of downloading the
176 actual video file and writing it to disk if the user has requested
177 it, among some other tasks. In most cases there should be one per
178 program. As, given a video URL, the downloader doesn't know how to
179 extract all the needed information, task that InfoExtractors do, it
180 has to pass the URL to one of them.
181
182 For this, YoutubeDL objects have a method that allows
183 InfoExtractors to be registered in a given order. When it is passed
184 a URL, the YoutubeDL object handles it to the first InfoExtractor it
185 finds that reports being able to handle it. The InfoExtractor extracts
186 all the information about the video or videos the URL refers to, and
187 YoutubeDL process the extracted information, possibly using a File
188 Downloader to download the video.
189
190 YoutubeDL objects accept a lot of parameters. In order not to saturate
191 the object constructor with arguments, it receives a dictionary of
192 options instead. These options are available through the params
193 attribute for the InfoExtractors to use. The YoutubeDL also
194 registers itself as the downloader in charge for the InfoExtractors
195 that are added to it, so this is a "mutual registration".
196
197 Available options:
198
199 username: Username for authentication purposes.
200 password: Password for authentication purposes.
201 videopassword: Password for accessing a video.
202 ap_mso: Adobe Pass multiple-system operator identifier.
203 ap_username: Multiple-system operator account username.
204 ap_password: Multiple-system operator account password.
205 usenetrc: Use netrc for authentication instead.
206 netrc_location: Location of the netrc file. Defaults to ~/.netrc.
207 netrc_cmd: Use a shell command to get credentials
208 verbose: Print additional info to stdout.
209 quiet: Do not print messages to stdout.
210 no_warnings: Do not print out anything for warnings.
211 forceprint: A dict with keys WHEN mapped to a list of templates to
212 print to stdout. The allowed keys are video or any of the
213 items in utils.POSTPROCESS_WHEN.
214 For compatibility, a single list is also accepted
215 print_to_file: A dict with keys WHEN (same as forceprint) mapped to
216 a list of tuples with (template, filename)
217 forcejson: Force printing info_dict as JSON.
218 dump_single_json: Force printing the info_dict of the whole playlist
219 (or video) as a single JSON line.
220 force_write_download_archive: Force writing download archive regardless
221 of 'skip_download' or 'simulate'.
222 simulate: Do not download the video files. If unset (or None),
223 simulate only if listsubtitles, listformats or list_thumbnails is used
224 format: Video format code. see "FORMAT SELECTION" for more details.
225 You can also pass a function. The function takes 'ctx' as
226 argument and returns the formats to download.
227 See "build_format_selector" for an implementation
228 allow_unplayable_formats: Allow unplayable formats to be extracted and downloaded.
229 ignore_no_formats_error: Ignore "No video formats" error. Usefull for
230 extracting metadata even if the video is not actually
231 available for download (experimental)
232 format_sort: A list of fields by which to sort the video formats.
233 See "Sorting Formats" for more details.
234 format_sort_force: Force the given format_sort. see "Sorting Formats"
235 for more details.
236 prefer_free_formats: Whether to prefer video formats with free containers
237 over non-free ones of same quality.
238 allow_multiple_video_streams: Allow multiple video streams to be merged
239 into a single file
240 allow_multiple_audio_streams: Allow multiple audio streams to be merged
241 into a single file
242 check_formats Whether to test if the formats are downloadable.
243 Can be True (check all), False (check none),
244 'selected' (check selected formats),
245 or None (check only if requested by extractor)
246 paths: Dictionary of output paths. The allowed keys are 'home'
247 'temp' and the keys of OUTTMPL_TYPES (in utils/_utils.py)
248 outtmpl: Dictionary of templates for output names. Allowed keys
249 are 'default' and the keys of OUTTMPL_TYPES (in utils/_utils.py).
250 For compatibility with youtube-dl, a single string can also be used
251 outtmpl_na_placeholder: Placeholder for unavailable meta fields.
252 restrictfilenames: Do not allow "&" and spaces in file names
253 trim_file_name: Limit length of filename (extension excluded)
254 windowsfilenames: Force the filenames to be windows compatible
255 ignoreerrors: Do not stop on download/postprocessing errors.
256 Can be 'only_download' to ignore only download errors.
257 Default is 'only_download' for CLI, but False for API
258 skip_playlist_after_errors: Number of allowed failures until the rest of
259 the playlist is skipped
260 allowed_extractors: List of regexes to match against extractor names that are allowed
261 overwrites: Overwrite all video and metadata files if True,
262 overwrite only non-video files if None
263 and don't overwrite any file if False
264 playlist_items: Specific indices of playlist to download.
265 playlistrandom: Download playlist items in random order.
266 lazy_playlist: Process playlist entries as they are received.
267 matchtitle: Download only matching titles.
268 rejecttitle: Reject downloads for matching titles.
269 logger: Log messages to a logging.Logger instance.
270 logtostderr: Print everything to stderr instead of stdout.
271 consoletitle: Display progress in console window's titlebar.
272 writedescription: Write the video description to a .description file
273 writeinfojson: Write the video description to a .info.json file
274 clean_infojson: Remove internal metadata from the infojson
275 getcomments: Extract video comments. This will not be written to disk
276 unless writeinfojson is also given
277 writeannotations: Write the video annotations to a .annotations.xml file
278 writethumbnail: Write the thumbnail image to a file
279 allow_playlist_files: Whether to write playlists' description, infojson etc
280 also to disk when using the 'write*' options
281 write_all_thumbnails: Write all thumbnail formats to files
282 writelink: Write an internet shortcut file, depending on the
283 current platform (.url/.webloc/.desktop)
284 writeurllink: Write a Windows internet shortcut file (.url)
285 writewebloclink: Write a macOS internet shortcut file (.webloc)
286 writedesktoplink: Write a Linux internet shortcut file (.desktop)
287 writesubtitles: Write the video subtitles to a file
288 writeautomaticsub: Write the automatically generated subtitles to a file
289 listsubtitles: Lists all available subtitles for the video
290 subtitlesformat: The format code for subtitles
291 subtitleslangs: List of languages of the subtitles to download (can be regex).
292 The list may contain "all" to refer to all the available
293 subtitles. The language can be prefixed with a "-" to
294 exclude it from the requested languages, e.g. ['all', '-live_chat']
295 keepvideo: Keep the video file after post-processing
296 daterange: A utils.DateRange object, download only if the upload_date is in the range.
297 skip_download: Skip the actual download of the video file
298 cachedir: Location of the cache files in the filesystem.
299 False to disable filesystem cache.
300 noplaylist: Download single video instead of a playlist if in doubt.
301 age_limit: An integer representing the user's age in years.
302 Unsuitable videos for the given age are skipped.
303 min_views: An integer representing the minimum view count the video
304 must have in order to not be skipped.
305 Videos without view count information are always
306 downloaded. None for no limit.
307 max_views: An integer representing the maximum view count.
308 Videos that are more popular than that are not
309 downloaded.
310 Videos without view count information are always
311 downloaded. None for no limit.
312 download_archive: A set, or the name of a file where all downloads are recorded.
313 Videos already present in the file are not downloaded again.
314 break_on_existing: Stop the download process after attempting to download a
315 file that is in the archive.
316 break_per_url: Whether break_on_reject and break_on_existing
317 should act on each input URL as opposed to for the entire queue
318 cookiefile: File name or text stream from where cookies should be read and dumped to
319 cookiesfrombrowser: A tuple containing the name of the browser, the profile
320 name/path from where cookies are loaded, the name of the keyring,
321 and the container name, e.g. ('chrome', ) or
322 ('vivaldi', 'default', 'BASICTEXT') or ('firefox', 'default', None, 'Meta')
323 legacyserverconnect: Explicitly allow HTTPS connection to servers that do not
324 support RFC 5746 secure renegotiation
325 nocheckcertificate: Do not verify SSL certificates
326 client_certificate: Path to client certificate file in PEM format. May include the private key
327 client_certificate_key: Path to private key file for client certificate
328 client_certificate_password: Password for client certificate private key, if encrypted.
329 If not provided and the key is encrypted, yt-dlp will ask interactively
330 prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
331 (Only supported by some extractors)
332 enable_file_urls: Enable file:// URLs. This is disabled by default for security reasons.
333 http_headers: A dictionary of custom headers to be used for all requests
334 proxy: URL of the proxy server to use
335 geo_verification_proxy: URL of the proxy to use for IP address verification
336 on geo-restricted sites.
337 socket_timeout: Time to wait for unresponsive hosts, in seconds
338 bidi_workaround: Work around buggy terminals without bidirectional text
339 support, using fridibi
340 debug_printtraffic:Print out sent and received HTTP traffic
341 default_search: Prepend this string if an input url is not valid.
342 'auto' for elaborate guessing
343 encoding: Use this encoding instead of the system-specified.
344 extract_flat: Whether to resolve and process url_results further
345 * False: Always process. Default for API
346 * True: Never process
347 * 'in_playlist': Do not process inside playlist/multi_video
348 * 'discard': Always process, but don't return the result
349 from inside playlist/multi_video
350 * 'discard_in_playlist': Same as "discard", but only for
351 playlists (not multi_video). Default for CLI
352 wait_for_video: If given, wait for scheduled streams to become available.
353 The value should be a tuple containing the range
354 (min_secs, max_secs) to wait between retries
355 postprocessors: A list of dictionaries, each with an entry
356 * key: The name of the postprocessor. See
357 yt_dlp/postprocessor/__init__.py for a list.
358 * when: When to run the postprocessor. Allowed values are
359 the entries of utils.POSTPROCESS_WHEN
360 Assumed to be 'post_process' if not given
361 progress_hooks: A list of functions that get called on download
362 progress, with a dictionary with the entries
363 * status: One of "downloading", "error", or "finished".
364 Check this first and ignore unknown values.
365 * info_dict: The extracted info_dict
366
367 If status is one of "downloading", or "finished", the
368 following properties may also be present:
369 * filename: The final filename (always present)
370 * tmpfilename: The filename we're currently writing to
371 * downloaded_bytes: Bytes on disk
372 * total_bytes: Size of the whole file, None if unknown
373 * total_bytes_estimate: Guess of the eventual file size,
374 None if unavailable.
375 * elapsed: The number of seconds since download started.
376 * eta: The estimated time in seconds, None if unknown
377 * speed: The download speed in bytes/second, None if
378 unknown
379 * fragment_index: The counter of the currently
380 downloaded video fragment.
381 * fragment_count: The number of fragments (= individual
382 files that will be merged)
383
384 Progress hooks are guaranteed to be called at least once
385 (with status "finished") if the download is successful.
386 postprocessor_hooks: A list of functions that get called on postprocessing
387 progress, with a dictionary with the entries
388 * status: One of "started", "processing", or "finished".
389 Check this first and ignore unknown values.
390 * postprocessor: Name of the postprocessor
391 * info_dict: The extracted info_dict
392
393 Progress hooks are guaranteed to be called at least twice
394 (with status "started" and "finished") if the processing is successful.
395 merge_output_format: "/" separated list of extensions to use when merging formats.
396 final_ext: Expected final extension; used to detect when the file was
397 already downloaded and converted
398 fixup: Automatically correct known faults of the file.
399 One of:
400 - "never": do nothing
401 - "warn": only emit a warning
402 - "detect_or_warn": check whether we can do anything
403 about it, warn otherwise (default)
404 source_address: Client-side IP address to bind to.
405 sleep_interval_requests: Number of seconds to sleep between requests
406 during extraction
407 sleep_interval: Number of seconds to sleep before each download when
408 used alone or a lower bound of a range for randomized
409 sleep before each download (minimum possible number
410 of seconds to sleep) when used along with
411 max_sleep_interval.
412 max_sleep_interval:Upper bound of a range for randomized sleep before each
413 download (maximum possible number of seconds to sleep).
414 Must only be used along with sleep_interval.
415 Actual sleep time will be a random float from range
416 [sleep_interval; max_sleep_interval].
417 sleep_interval_subtitles: Number of seconds to sleep before each subtitle download
418 listformats: Print an overview of available video formats and exit.
419 list_thumbnails: Print a table of all thumbnails and exit.
420 match_filter: A function that gets called for every video with the signature
421 (info_dict, *, incomplete: bool) -> Optional[str]
422 For backward compatibility with youtube-dl, the signature
423 (info_dict) -> Optional[str] is also allowed.
424 - If it returns a message, the video is ignored.
425 - If it returns None, the video is downloaded.
426 - If it returns utils.NO_DEFAULT, the user is interactively
427 asked whether to download the video.
428 - Raise utils.DownloadCancelled(msg) to abort remaining
429 downloads when a video is rejected.
430 match_filter_func in utils/_utils.py is one example for this.
431 color: A Dictionary with output stream names as keys
432 and their respective color policy as values.
433 Can also just be a single color policy,
434 in which case it applies to all outputs.
435 Valid stream names are 'stdout' and 'stderr'.
436 Valid color policies are one of 'always', 'auto', 'no_color' or 'never'.
437 geo_bypass: Bypass geographic restriction via faking X-Forwarded-For
438 HTTP header
439 geo_bypass_country:
440 Two-letter ISO 3166-2 country code that will be used for
441 explicit geographic restriction bypassing via faking
442 X-Forwarded-For HTTP header
443 geo_bypass_ip_block:
444 IP range in CIDR notation that will be used similarly to
445 geo_bypass_country
446 external_downloader: A dictionary of protocol keys and the executable of the
447 external downloader to use for it. The allowed protocols
448 are default|http|ftp|m3u8|dash|rtsp|rtmp|mms.
449 Set the value to 'native' to use the native downloader
450 compat_opts: Compatibility options. See "Differences in default behavior".
451 The following options do not work when used through the API:
452 filename, abort-on-error, multistreams, no-live-chat, format-sort
453 no-clean-infojson, no-playlist-metafiles, no-keep-subs, no-attach-info-json.
454 Refer __init__.py for their implementation
455 progress_template: Dictionary of templates for progress outputs.
456 Allowed keys are 'download', 'postprocess',
457 'download-title' (console title) and 'postprocess-title'.
458 The template is mapped on a dictionary with keys 'progress' and 'info'
459 retry_sleep_functions: Dictionary of functions that takes the number of attempts
460 as argument and returns the time to sleep in seconds.
461 Allowed keys are 'http', 'fragment', 'file_access'
462 download_ranges: A callback function that gets called for every video with
463 the signature (info_dict, ydl) -> Iterable[Section].
464 Only the returned sections will be downloaded.
465 Each Section is a dict with the following keys:
466 * start_time: Start time of the section in seconds
467 * end_time: End time of the section in seconds
468 * title: Section title (Optional)
469 * index: Section number (Optional)
470 force_keyframes_at_cuts: Re-encode the video when downloading ranges to get precise cuts
471 noprogress: Do not print the progress bar
472 live_from_start: Whether to download livestreams videos from the start
473
474 The following parameters are not used by YoutubeDL itself, they are used by
475 the downloader (see yt_dlp/downloader/common.py):
476 nopart, updatetime, buffersize, ratelimit, throttledratelimit, min_filesize,
477 max_filesize, test, noresizebuffer, retries, file_access_retries, fragment_retries,
478 continuedl, xattr_set_filesize, hls_use_mpegts, http_chunk_size,
479 external_downloader_args, concurrent_fragment_downloads.
480
481 The following options are used by the post processors:
482 ffmpeg_location: Location of the ffmpeg/avconv binary; either the path
483 to the binary or its containing directory.
484 postprocessor_args: A dictionary of postprocessor/executable keys (in lower case)
485 and a list of additional command-line arguments for the
486 postprocessor/executable. The dict can also have "PP+EXE" keys
487 which are used when the given exe is used by the given PP.
488 Use 'default' as the name for arguments to passed to all PP
489 For compatibility with youtube-dl, a single list of args
490 can also be used
491
492 The following options are used by the extractors:
493 extractor_retries: Number of times to retry for known errors (default: 3)
494 dynamic_mpd: Whether to process dynamic DASH manifests (default: True)
495 hls_split_discontinuity: Split HLS playlists to different formats at
496 discontinuities such as ad breaks (default: False)
497 extractor_args: A dictionary of arguments to be passed to the extractors.
498 See "EXTRACTOR ARGUMENTS" for details.
499 E.g. {'youtube': {'skip': ['dash', 'hls']}}
500 mark_watched: Mark videos watched (even with --simulate). Only for YouTube
501
502 The following options are deprecated and may be removed in the future:
503
504 break_on_reject: Stop the download process when encountering a video that
505 has been filtered out.
506 - `raise DownloadCancelled(msg)` in match_filter instead
507 force_generic_extractor: Force downloader to use the generic extractor
508 - Use allowed_extractors = ['generic', 'default']
509 playliststart: - Use playlist_items
510 Playlist item to start at.
511 playlistend: - Use playlist_items
512 Playlist item to end at.
513 playlistreverse: - Use playlist_items
514 Download playlist items in reverse order.
515 forceurl: - Use forceprint
516 Force printing final URL.
517 forcetitle: - Use forceprint
518 Force printing title.
519 forceid: - Use forceprint
520 Force printing ID.
521 forcethumbnail: - Use forceprint
522 Force printing thumbnail URL.
523 forcedescription: - Use forceprint
524 Force printing description.
525 forcefilename: - Use forceprint
526 Force printing final filename.
527 forceduration: - Use forceprint
528 Force printing duration.
529 allsubtitles: - Use subtitleslangs = ['all']
530 Downloads all the subtitles of the video
531 (requires writesubtitles or writeautomaticsub)
532 include_ads: - Doesn't work
533 Download ads as well
534 call_home: - Not implemented
535 Boolean, true iff we are allowed to contact the
536 yt-dlp servers for debugging.
537 post_hooks: - Register a custom postprocessor
538 A list of functions that get called as the final step
539 for each video file, after all postprocessors have been
540 called. The filename will be passed as the only argument.
541 hls_prefer_native: - Use external_downloader = {'m3u8': 'native'} or {'m3u8': 'ffmpeg'}.
542 Use the native HLS downloader instead of ffmpeg/avconv
543 if True, otherwise use ffmpeg/avconv if False, otherwise
544 use downloader suggested by extractor if None.
545 prefer_ffmpeg: - avconv support is deprecated
546 If False, use avconv instead of ffmpeg if both are available,
547 otherwise prefer ffmpeg.
548 youtube_include_dash_manifest: - Use extractor_args
549 If True (default), DASH manifests and related
550 data will be downloaded and processed by extractor.
551 You can reduce network I/O by disabling it if you don't
552 care about DASH. (only for youtube)
553 youtube_include_hls_manifest: - Use extractor_args
554 If True (default), HLS manifests and related
555 data will be downloaded and processed by extractor.
556 You can reduce network I/O by disabling it if you don't
557 care about HLS. (only for youtube)
558 no_color: Same as `color='no_color'`
559 no_overwrites: Same as `overwrites=False`
560 """
561
562 _NUMERIC_FIELDS = {
563 'width', 'height', 'asr', 'audio_channels', 'fps',
564 'tbr', 'abr', 'vbr', 'filesize', 'filesize_approx',
565 'timestamp', 'release_timestamp',
566 'duration', 'view_count', 'like_count', 'dislike_count', 'repost_count',
567 'average_rating', 'comment_count', 'age_limit',
568 'start_time', 'end_time',
569 'chapter_number', 'season_number', 'episode_number',
570 'track_number', 'disc_number', 'release_year',
571 }
572
573 _format_fields = {
574 # NB: Keep in sync with the docstring of extractor/common.py
575 'url', 'manifest_url', 'manifest_stream_number', 'ext', 'format', 'format_id', 'format_note',
576 'width', 'height', 'aspect_ratio', 'resolution', 'dynamic_range', 'tbr', 'abr', 'acodec', 'asr', 'audio_channels',
577 'vbr', 'fps', 'vcodec', 'container', 'filesize', 'filesize_approx', 'rows', 'columns',
578 'player_url', 'protocol', 'fragment_base_url', 'fragments', 'is_from_start',
579 'preference', 'language', 'language_preference', 'quality', 'source_preference', 'cookies',
580 'http_headers', 'stretched_ratio', 'no_resume', 'has_drm', 'extra_param_to_segment_url', 'hls_aes', 'downloader_options',
581 'page_url', 'app', 'play_path', 'tc_url', 'flash_version', 'rtmp_live', 'rtmp_conn', 'rtmp_protocol', 'rtmp_real_time'
582 }
583 _deprecated_multivalue_fields = {
584 'album_artist': 'album_artists',
585 'artist': 'artists',
586 'composer': 'composers',
587 'creator': 'creators',
588 'genre': 'genres',
589 }
590 _format_selection_exts = {
591 'audio': set(MEDIA_EXTENSIONS.common_audio),
592 'video': set(MEDIA_EXTENSIONS.common_video + ('3gp', )),
593 'storyboards': set(MEDIA_EXTENSIONS.storyboards),
594 }
595
596 def __init__(self, params=None, auto_init=True):
597 """Create a FileDownloader object with the given options.
598 @param auto_init Whether to load the default extractors and print header (if verbose).
599 Set to 'no_verbose_header' to not print the header
600 """
601 if params is None:
602 params = {}
603 self.params = params
604 self._ies = {}
605 self._ies_instances = {}
606 self._pps = {k: [] for k in POSTPROCESS_WHEN}
607 self._printed_messages = set()
608 self._first_webpage_request = True
609 self._post_hooks = []
610 self._progress_hooks = []
611 self._postprocessor_hooks = []
612 self._download_retcode = 0
613 self._num_downloads = 0
614 self._num_videos = 0
615 self._playlist_level = 0
616 self._playlist_urls = set()
617 self.cache = Cache(self)
618 self.__header_cookies = []
619
620 stdout = sys.stderr if self.params.get('logtostderr') else sys.stdout
621 self._out_files = Namespace(
622 out=stdout,
623 error=sys.stderr,
624 screen=sys.stderr if self.params.get('quiet') else stdout,
625 console=None if compat_os_name == 'nt' else next(
626 filter(supports_terminal_sequences, (sys.stderr, sys.stdout)), None)
627 )
628
629 try:
630 windows_enable_vt_mode()
631 except Exception as e:
632 self.write_debug(f'Failed to enable VT mode: {e}')
633
634 if self.params.get('no_color'):
635 if self.params.get('color') is not None:
636 self.params.setdefault('_warnings', []).append(
637 'Overwriting params from "color" with "no_color"')
638 self.params['color'] = 'no_color'
639
640 term_allow_color = os.getenv('TERM', '').lower() != 'dumb'
641 no_color = bool(os.getenv('NO_COLOR'))
642
643 def process_color_policy(stream):
644 stream_name = {sys.stdout: 'stdout', sys.stderr: 'stderr'}[stream]
645 policy = traverse_obj(self.params, ('color', (stream_name, None), {str}), get_all=False)
646 if policy in ('auto', None):
647 if term_allow_color and supports_terminal_sequences(stream):
648 return 'no_color' if no_color else True
649 return False
650 assert policy in ('always', 'never', 'no_color'), policy
651 return {'always': True, 'never': False}.get(policy, policy)
652
653 self._allow_colors = Namespace(**{
654 name: process_color_policy(stream)
655 for name, stream in self._out_files.items_ if name != 'console'
656 })
657
658 system_deprecation = _get_system_deprecation()
659 if system_deprecation:
660 self.deprecated_feature(system_deprecation.replace('\n', '\n '))
661
662 if self.params.get('allow_unplayable_formats'):
663 self.report_warning(
664 f'You have asked for {self._format_err("UNPLAYABLE", self.Styles.EMPHASIS)} formats to be listed/downloaded. '
665 'This is a developer option intended for debugging. \n'
666 ' If you experience any issues while using this option, '
667 f'{self._format_err("DO NOT", self.Styles.ERROR)} open a bug report')
668
669 if self.params.get('bidi_workaround', False):
670 try:
671 import pty
672 master, slave = pty.openpty()
673 width = shutil.get_terminal_size().columns
674 width_args = [] if width is None else ['-w', str(width)]
675 sp_kwargs = {'stdin': subprocess.PIPE, 'stdout': slave, 'stderr': self._out_files.error}
676 try:
677 self._output_process = Popen(['bidiv'] + width_args, **sp_kwargs)
678 except OSError:
679 self._output_process = Popen(['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
680 self._output_channel = os.fdopen(master, 'rb')
681 except OSError as ose:
682 if ose.errno == errno.ENOENT:
683 self.report_warning(
684 'Could not find fribidi executable, ignoring --bidi-workaround. '
685 'Make sure that fribidi is an executable file in one of the directories in your $PATH.')
686 else:
687 raise
688
689 self.params['compat_opts'] = set(self.params.get('compat_opts', ()))
690 self.params['http_headers'] = HTTPHeaderDict(std_headers, self.params.get('http_headers'))
691 self._load_cookies(self.params['http_headers'].get('Cookie')) # compat
692 self.params['http_headers'].pop('Cookie', None)
693
694 if auto_init and auto_init != 'no_verbose_header':
695 self.print_debug_header()
696
697 def check_deprecated(param, option, suggestion):
698 if self.params.get(param) is not None:
699 self.report_warning(f'{option} is deprecated. Use {suggestion} instead')
700 return True
701 return False
702
703 if check_deprecated('cn_verification_proxy', '--cn-verification-proxy', '--geo-verification-proxy'):
704 if self.params.get('geo_verification_proxy') is None:
705 self.params['geo_verification_proxy'] = self.params['cn_verification_proxy']
706
707 check_deprecated('autonumber', '--auto-number', '-o "%(autonumber)s-%(title)s.%(ext)s"')
708 check_deprecated('usetitle', '--title', '-o "%(title)s-%(id)s.%(ext)s"')
709 check_deprecated('useid', '--id', '-o "%(id)s.%(ext)s"')
710
711 for msg in self.params.get('_warnings', []):
712 self.report_warning(msg)
713 for msg in self.params.get('_deprecation_warnings', []):
714 self.deprecated_feature(msg)
715
716 if 'list-formats' in self.params['compat_opts']:
717 self.params['listformats_table'] = False
718
719 if 'overwrites' not in self.params and self.params.get('nooverwrites') is not None:
720 # nooverwrites was unnecessarily changed to overwrites
721 # in 0c3d0f51778b153f65c21906031c2e091fcfb641
722 # This ensures compatibility with both keys
723 self.params['overwrites'] = not self.params['nooverwrites']
724 elif self.params.get('overwrites') is None:
725 self.params.pop('overwrites', None)
726 else:
727 self.params['nooverwrites'] = not self.params['overwrites']
728
729 if self.params.get('simulate') is None and any((
730 self.params.get('list_thumbnails'),
731 self.params.get('listformats'),
732 self.params.get('listsubtitles'),
733 )):
734 self.params['simulate'] = 'list_only'
735
736 self.params.setdefault('forceprint', {})
737 self.params.setdefault('print_to_file', {})
738
739 # Compatibility with older syntax
740 if not isinstance(params['forceprint'], dict):
741 self.params['forceprint'] = {'video': params['forceprint']}
742
743 if auto_init:
744 self.add_default_info_extractors()
745
746 if (sys.platform != 'win32'
747 and sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968']
748 and not self.params.get('restrictfilenames', False)):
749 # Unicode filesystem API will throw errors (#1474, #13027)
750 self.report_warning(
751 'Assuming --restrict-filenames since file system encoding '
752 'cannot encode all characters. '
753 'Set the LC_ALL environment variable to fix this.')
754 self.params['restrictfilenames'] = True
755
756 self._parse_outtmpl()
757
758 # Creating format selector here allows us to catch syntax errors before the extraction
759 self.format_selector = (
760 self.params.get('format') if self.params.get('format') in (None, '-')
761 else self.params['format'] if callable(self.params['format'])
762 else self.build_format_selector(self.params['format']))
763
764 hooks = {
765 'post_hooks': self.add_post_hook,
766 'progress_hooks': self.add_progress_hook,
767 'postprocessor_hooks': self.add_postprocessor_hook,
768 }
769 for opt, fn in hooks.items():
770 for ph in self.params.get(opt, []):
771 fn(ph)
772
773 for pp_def_raw in self.params.get('postprocessors', []):
774 pp_def = dict(pp_def_raw)
775 when = pp_def.pop('when', 'post_process')
776 self.add_post_processor(
777 get_postprocessor(pp_def.pop('key'))(self, **pp_def),
778 when=when)
779
780 def preload_download_archive(fn):
781 """Preload the archive, if any is specified"""
782 archive = set()
783 if fn is None:
784 return archive
785 elif not is_path_like(fn):
786 return fn
787
788 self.write_debug(f'Loading archive file {fn!r}')
789 try:
790 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
791 for line in archive_file:
792 archive.add(line.strip())
793 except OSError as ioe:
794 if ioe.errno != errno.ENOENT:
795 raise
796 return archive
797
798 self.archive = preload_download_archive(self.params.get('download_archive'))
799
800 def warn_if_short_id(self, argv):
801 # short YouTube ID starting with dash?
802 idxs = [
803 i for i, a in enumerate(argv)
804 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
805 if idxs:
806 correct_argv = (
807 ['yt-dlp']
808 + [a for i, a in enumerate(argv) if i not in idxs]
809 + ['--'] + [argv[i] for i in idxs]
810 )
811 self.report_warning(
812 'Long argument string detected. '
813 'Use -- to separate parameters and URLs, like this:\n%s' %
814 args_to_str(correct_argv))
815
816 def add_info_extractor(self, ie):
817 """Add an InfoExtractor object to the end of the list."""
818 ie_key = ie.ie_key()
819 self._ies[ie_key] = ie
820 if not isinstance(ie, type):
821 self._ies_instances[ie_key] = ie
822 ie.set_downloader(self)
823
824 def get_info_extractor(self, ie_key):
825 """
826 Get an instance of an IE with name ie_key, it will try to get one from
827 the _ies list, if there's no instance it will create a new one and add
828 it to the extractor list.
829 """
830 ie = self._ies_instances.get(ie_key)
831 if ie is None:
832 ie = get_info_extractor(ie_key)()
833 self.add_info_extractor(ie)
834 return ie
835
836 def add_default_info_extractors(self):
837 """
838 Add the InfoExtractors returned by gen_extractors to the end of the list
839 """
840 all_ies = {ie.IE_NAME.lower(): ie for ie in gen_extractor_classes()}
841 all_ies['end'] = UnsupportedURLIE()
842 try:
843 ie_names = orderedSet_from_options(
844 self.params.get('allowed_extractors', ['default']), {
845 'all': list(all_ies),
846 'default': [name for name, ie in all_ies.items() if ie._ENABLED],
847 }, use_regex=True)
848 except re.error as e:
849 raise ValueError(f'Wrong regex for allowed_extractors: {e.pattern}')
850 for name in ie_names:
851 self.add_info_extractor(all_ies[name])
852 self.write_debug(f'Loaded {len(ie_names)} extractors')
853
854 def add_post_processor(self, pp, when='post_process'):
855 """Add a PostProcessor object to the end of the chain."""
856 assert when in POSTPROCESS_WHEN, f'Invalid when={when}'
857 self._pps[when].append(pp)
858 pp.set_downloader(self)
859
860 def add_post_hook(self, ph):
861 """Add the post hook"""
862 self._post_hooks.append(ph)
863
864 def add_progress_hook(self, ph):
865 """Add the download progress hook"""
866 self._progress_hooks.append(ph)
867
868 def add_postprocessor_hook(self, ph):
869 """Add the postprocessing progress hook"""
870 self._postprocessor_hooks.append(ph)
871 for pps in self._pps.values():
872 for pp in pps:
873 pp.add_progress_hook(ph)
874
875 def _bidi_workaround(self, message):
876 if not hasattr(self, '_output_channel'):
877 return message
878
879 assert hasattr(self, '_output_process')
880 assert isinstance(message, str)
881 line_count = message.count('\n') + 1
882 self._output_process.stdin.write((message + '\n').encode())
883 self._output_process.stdin.flush()
884 res = ''.join(self._output_channel.readline().decode()
885 for _ in range(line_count))
886 return res[:-len('\n')]
887
888 def _write_string(self, message, out=None, only_once=False):
889 if only_once:
890 if message in self._printed_messages:
891 return
892 self._printed_messages.add(message)
893 write_string(message, out=out, encoding=self.params.get('encoding'))
894
895 def to_stdout(self, message, skip_eol=False, quiet=None):
896 """Print message to stdout"""
897 if quiet is not None:
898 self.deprecation_warning('"YoutubeDL.to_stdout" no longer accepts the argument quiet. '
899 'Use "YoutubeDL.to_screen" instead')
900 if skip_eol is not False:
901 self.deprecation_warning('"YoutubeDL.to_stdout" no longer accepts the argument skip_eol. '
902 'Use "YoutubeDL.to_screen" instead')
903 self._write_string(f'{self._bidi_workaround(message)}\n', self._out_files.out)
904
905 def to_screen(self, message, skip_eol=False, quiet=None, only_once=False):
906 """Print message to screen if not in quiet mode"""
907 if self.params.get('logger'):
908 self.params['logger'].debug(message)
909 return
910 if (self.params.get('quiet') if quiet is None else quiet) and not self.params.get('verbose'):
911 return
912 self._write_string(
913 '%s%s' % (self._bidi_workaround(message), ('' if skip_eol else '\n')),
914 self._out_files.screen, only_once=only_once)
915
916 def to_stderr(self, message, only_once=False):
917 """Print message to stderr"""
918 assert isinstance(message, str)
919 if self.params.get('logger'):
920 self.params['logger'].error(message)
921 else:
922 self._write_string(f'{self._bidi_workaround(message)}\n', self._out_files.error, only_once=only_once)
923
924 def _send_console_code(self, code):
925 if compat_os_name == 'nt' or not self._out_files.console:
926 return
927 self._write_string(code, self._out_files.console)
928
929 def to_console_title(self, message):
930 if not self.params.get('consoletitle', False):
931 return
932 message = remove_terminal_sequences(message)
933 if compat_os_name == 'nt':
934 if ctypes.windll.kernel32.GetConsoleWindow():
935 # c_wchar_p() might not be necessary if `message` is
936 # already of type unicode()
937 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
938 else:
939 self._send_console_code(f'\033]0;{message}\007')
940
941 def save_console_title(self):
942 if not self.params.get('consoletitle') or self.params.get('simulate'):
943 return
944 self._send_console_code('\033[22;0t') # Save the title on stack
945
946 def restore_console_title(self):
947 if not self.params.get('consoletitle') or self.params.get('simulate'):
948 return
949 self._send_console_code('\033[23;0t') # Restore the title from stack
950
951 def __enter__(self):
952 self.save_console_title()
953 return self
954
955 def save_cookies(self):
956 if self.params.get('cookiefile') is not None:
957 self.cookiejar.save()
958
959 def __exit__(self, *args):
960 self.restore_console_title()
961 self.close()
962
963 def close(self):
964 self.save_cookies()
965 self._request_director.close()
966 del self._request_director
967
968 def trouble(self, message=None, tb=None, is_error=True):
969 """Determine action to take when a download problem appears.
970
971 Depending on if the downloader has been configured to ignore
972 download errors or not, this method may throw an exception or
973 not when errors are found, after printing the message.
974
975 @param tb If given, is additional traceback information
976 @param is_error Whether to raise error according to ignorerrors
977 """
978 if message is not None:
979 self.to_stderr(message)
980 if self.params.get('verbose'):
981 if tb is None:
982 if sys.exc_info()[0]: # if .trouble has been called from an except block
983 tb = ''
984 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
985 tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
986 tb += encode_compat_str(traceback.format_exc())
987 else:
988 tb_data = traceback.format_list(traceback.extract_stack())
989 tb = ''.join(tb_data)
990 if tb:
991 self.to_stderr(tb)
992 if not is_error:
993 return
994 if not self.params.get('ignoreerrors'):
995 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
996 exc_info = sys.exc_info()[1].exc_info
997 else:
998 exc_info = sys.exc_info()
999 raise DownloadError(message, exc_info)
1000 self._download_retcode = 1
1001
1002 Styles = Namespace(
1003 HEADERS='yellow',
1004 EMPHASIS='light blue',
1005 FILENAME='green',
1006 ID='green',
1007 DELIM='blue',
1008 ERROR='red',
1009 BAD_FORMAT='light red',
1010 WARNING='yellow',
1011 SUPPRESS='light black',
1012 )
1013
1014 def _format_text(self, handle, allow_colors, text, f, fallback=None, *, test_encoding=False):
1015 text = str(text)
1016 if test_encoding:
1017 original_text = text
1018 # handle.encoding can be None. See https://github.com/yt-dlp/yt-dlp/issues/2711
1019 encoding = self.params.get('encoding') or getattr(handle, 'encoding', None) or 'ascii'
1020 text = text.encode(encoding, 'ignore').decode(encoding)
1021 if fallback is not None and text != original_text:
1022 text = fallback
1023 return format_text(text, f) if allow_colors is True else text if fallback is None else fallback
1024
1025 def _format_out(self, *args, **kwargs):
1026 return self._format_text(self._out_files.out, self._allow_colors.out, *args, **kwargs)
1027
1028 def _format_screen(self, *args, **kwargs):
1029 return self._format_text(self._out_files.screen, self._allow_colors.screen, *args, **kwargs)
1030
1031 def _format_err(self, *args, **kwargs):
1032 return self._format_text(self._out_files.error, self._allow_colors.error, *args, **kwargs)
1033
1034 def report_warning(self, message, only_once=False):
1035 '''
1036 Print the message to stderr, it will be prefixed with 'WARNING:'
1037 If stderr is a tty file the 'WARNING:' will be colored
1038 '''
1039 if self.params.get('logger') is not None:
1040 self.params['logger'].warning(message)
1041 else:
1042 if self.params.get('no_warnings'):
1043 return
1044 self.to_stderr(f'{self._format_err("WARNING:", self.Styles.WARNING)} {message}', only_once)
1045
1046 def deprecation_warning(self, message, *, stacklevel=0):
1047 deprecation_warning(
1048 message, stacklevel=stacklevel + 1, printer=self.report_error, is_error=False)
1049
1050 def deprecated_feature(self, message):
1051 if self.params.get('logger') is not None:
1052 self.params['logger'].warning(f'Deprecated Feature: {message}')
1053 self.to_stderr(f'{self._format_err("Deprecated Feature:", self.Styles.ERROR)} {message}', True)
1054
1055 def report_error(self, message, *args, **kwargs):
1056 '''
1057 Do the same as trouble, but prefixes the message with 'ERROR:', colored
1058 in red if stderr is a tty file.
1059 '''
1060 self.trouble(f'{self._format_err("ERROR:", self.Styles.ERROR)} {message}', *args, **kwargs)
1061
1062 def write_debug(self, message, only_once=False):
1063 '''Log debug message or Print message to stderr'''
1064 if not self.params.get('verbose', False):
1065 return
1066 message = f'[debug] {message}'
1067 if self.params.get('logger'):
1068 self.params['logger'].debug(message)
1069 else:
1070 self.to_stderr(message, only_once)
1071
1072 def report_file_already_downloaded(self, file_name):
1073 """Report file has already been fully downloaded."""
1074 try:
1075 self.to_screen('[download] %s has already been downloaded' % file_name)
1076 except UnicodeEncodeError:
1077 self.to_screen('[download] The file has already been downloaded')
1078
1079 def report_file_delete(self, file_name):
1080 """Report that existing file will be deleted."""
1081 try:
1082 self.to_screen('Deleting existing file %s' % file_name)
1083 except UnicodeEncodeError:
1084 self.to_screen('Deleting existing file')
1085
1086 def raise_no_formats(self, info, forced=False, *, msg=None):
1087 has_drm = info.get('_has_drm')
1088 ignored, expected = self.params.get('ignore_no_formats_error'), bool(msg)
1089 msg = msg or has_drm and 'This video is DRM protected' or 'No video formats found!'
1090 if forced or not ignored:
1091 raise ExtractorError(msg, video_id=info['id'], ie=info['extractor'],
1092 expected=has_drm or ignored or expected)
1093 else:
1094 self.report_warning(msg)
1095
1096 def parse_outtmpl(self):
1097 self.deprecation_warning('"YoutubeDL.parse_outtmpl" is deprecated and may be removed in a future version')
1098 self._parse_outtmpl()
1099 return self.params['outtmpl']
1100
1101 def _parse_outtmpl(self):
1102 sanitize = IDENTITY
1103 if self.params.get('restrictfilenames'): # Remove spaces in the default template
1104 sanitize = lambda x: x.replace(' - ', ' ').replace(' ', '-')
1105
1106 outtmpl = self.params.setdefault('outtmpl', {})
1107 if not isinstance(outtmpl, dict):
1108 self.params['outtmpl'] = outtmpl = {'default': outtmpl}
1109 outtmpl.update({k: sanitize(v) for k, v in DEFAULT_OUTTMPL.items() if outtmpl.get(k) is None})
1110
1111 def get_output_path(self, dir_type='', filename=None):
1112 paths = self.params.get('paths', {})
1113 assert isinstance(paths, dict), '"paths" parameter must be a dictionary'
1114 path = os.path.join(
1115 expand_path(paths.get('home', '').strip()),
1116 expand_path(paths.get(dir_type, '').strip()) if dir_type else '',
1117 filename or '')
1118 return sanitize_path(path, force=self.params.get('windowsfilenames'))
1119
1120 @staticmethod
1121 def _outtmpl_expandpath(outtmpl):
1122 # expand_path translates '%%' into '%' and '$$' into '$'
1123 # correspondingly that is not what we want since we need to keep
1124 # '%%' intact for template dict substitution step. Working around
1125 # with boundary-alike separator hack.
1126 sep = ''.join(random.choices(string.ascii_letters, k=32))
1127 outtmpl = outtmpl.replace('%%', f'%{sep}%').replace('$$', f'${sep}$')
1128
1129 # outtmpl should be expand_path'ed before template dict substitution
1130 # because meta fields may contain env variables we don't want to
1131 # be expanded. E.g. for outtmpl "%(title)s.%(ext)s" and
1132 # title "Hello $PATH", we don't want `$PATH` to be expanded.
1133 return expand_path(outtmpl).replace(sep, '')
1134
1135 @staticmethod
1136 def escape_outtmpl(outtmpl):
1137 ''' Escape any remaining strings like %s, %abc% etc. '''
1138 return re.sub(
1139 STR_FORMAT_RE_TMPL.format('', '(?![%(\0])'),
1140 lambda mobj: ('' if mobj.group('has_key') else '%') + mobj.group(0),
1141 outtmpl)
1142
1143 @classmethod
1144 def validate_outtmpl(cls, outtmpl):
1145 ''' @return None or Exception object '''
1146 outtmpl = re.sub(
1147 STR_FORMAT_RE_TMPL.format('[^)]*', '[ljhqBUDS]'),
1148 lambda mobj: f'{mobj.group(0)[:-1]}s',
1149 cls._outtmpl_expandpath(outtmpl))
1150 try:
1151 cls.escape_outtmpl(outtmpl) % collections.defaultdict(int)
1152 return None
1153 except ValueError as err:
1154 return err
1155
1156 @staticmethod
1157 def _copy_infodict(info_dict):
1158 info_dict = dict(info_dict)
1159 info_dict.pop('__postprocessors', None)
1160 info_dict.pop('__pending_error', None)
1161 return info_dict
1162
1163 def prepare_outtmpl(self, outtmpl, info_dict, sanitize=False):
1164 """ Make the outtmpl and info_dict suitable for substitution: ydl.escape_outtmpl(outtmpl) % info_dict
1165 @param sanitize Whether to sanitize the output as a filename.
1166 For backward compatibility, a function can also be passed
1167 """
1168
1169 info_dict.setdefault('epoch', int(time.time())) # keep epoch consistent once set
1170
1171 info_dict = self._copy_infodict(info_dict)
1172 info_dict['duration_string'] = ( # %(duration>%H-%M-%S)s is wrong if duration > 24hrs
1173 formatSeconds(info_dict['duration'], '-' if sanitize else ':')
1174 if info_dict.get('duration', None) is not None
1175 else None)
1176 info_dict['autonumber'] = int(self.params.get('autonumber_start', 1) - 1 + self._num_downloads)
1177 info_dict['video_autonumber'] = self._num_videos
1178 if info_dict.get('resolution') is None:
1179 info_dict['resolution'] = self.format_resolution(info_dict, default=None)
1180
1181 # For fields playlist_index, playlist_autonumber and autonumber convert all occurrences
1182 # of %(field)s to %(field)0Nd for backward compatibility
1183 field_size_compat_map = {
1184 'playlist_index': number_of_digits(info_dict.get('__last_playlist_index') or 0),
1185 'playlist_autonumber': number_of_digits(info_dict.get('n_entries') or 0),
1186 'autonumber': self.params.get('autonumber_size') or 5,
1187 }
1188
1189 TMPL_DICT = {}
1190 EXTERNAL_FORMAT_RE = re.compile(STR_FORMAT_RE_TMPL.format('[^)]*', f'[{STR_FORMAT_TYPES}ljhqBUDS]'))
1191 MATH_FUNCTIONS = {
1192 '+': float.__add__,
1193 '-': float.__sub__,
1194 '*': float.__mul__,
1195 }
1196 # Field is of the form key1.key2...
1197 # where keys (except first) can be string, int, slice or "{field, ...}"
1198 FIELD_INNER_RE = r'(?:\w+|%(num)s|%(num)s?(?::%(num)s?){1,2})' % {'num': r'(?:-?\d+)'}
1199 FIELD_RE = r'\w*(?:\.(?:%(inner)s|{%(field)s(?:,%(field)s)*}))*' % {
1200 'inner': FIELD_INNER_RE,
1201 'field': rf'\w*(?:\.{FIELD_INNER_RE})*'
1202 }
1203 MATH_FIELD_RE = rf'(?:{FIELD_RE}|-?{NUMBER_RE})'
1204 MATH_OPERATORS_RE = r'(?:%s)' % '|'.join(map(re.escape, MATH_FUNCTIONS.keys()))
1205 INTERNAL_FORMAT_RE = re.compile(rf'''(?xs)
1206 (?P<negate>-)?
1207 (?P<fields>{FIELD_RE})
1208 (?P<maths>(?:{MATH_OPERATORS_RE}{MATH_FIELD_RE})*)
1209 (?:>(?P<strf_format>.+?))?
1210 (?P<remaining>
1211 (?P<alternate>(?<!\\),[^|&)]+)?
1212 (?:&(?P<replacement>.*?))?
1213 (?:\|(?P<default>.*?))?
1214 )$''')
1215
1216 def _from_user_input(field):
1217 if field == ':':
1218 return ...
1219 elif ':' in field:
1220 return slice(*map(int_or_none, field.split(':')))
1221 elif int_or_none(field) is not None:
1222 return int(field)
1223 return field
1224
1225 def _traverse_infodict(fields):
1226 fields = [f for x in re.split(r'\.({.+?})\.?', fields)
1227 for f in ([x] if x.startswith('{') else x.split('.'))]
1228 for i in (0, -1):
1229 if fields and not fields[i]:
1230 fields.pop(i)
1231
1232 for i, f in enumerate(fields):
1233 if not f.startswith('{'):
1234 fields[i] = _from_user_input(f)
1235 continue
1236 assert f.endswith('}'), f'No closing brace for {f} in {fields}'
1237 fields[i] = {k: list(map(_from_user_input, k.split('.'))) for k in f[1:-1].split(',')}
1238
1239 return traverse_obj(info_dict, fields, traverse_string=True)
1240
1241 def get_value(mdict):
1242 # Object traversal
1243 value = _traverse_infodict(mdict['fields'])
1244 # Negative
1245 if mdict['negate']:
1246 value = float_or_none(value)
1247 if value is not None:
1248 value *= -1
1249 # Do maths
1250 offset_key = mdict['maths']
1251 if offset_key:
1252 value = float_or_none(value)
1253 operator = None
1254 while offset_key:
1255 item = re.match(
1256 MATH_FIELD_RE if operator else MATH_OPERATORS_RE,
1257 offset_key).group(0)
1258 offset_key = offset_key[len(item):]
1259 if operator is None:
1260 operator = MATH_FUNCTIONS[item]
1261 continue
1262 item, multiplier = (item[1:], -1) if item[0] == '-' else (item, 1)
1263 offset = float_or_none(item)
1264 if offset is None:
1265 offset = float_or_none(_traverse_infodict(item))
1266 try:
1267 value = operator(value, multiplier * offset)
1268 except (TypeError, ZeroDivisionError):
1269 return None
1270 operator = None
1271 # Datetime formatting
1272 if mdict['strf_format']:
1273 value = strftime_or_none(value, mdict['strf_format'].replace('\\,', ','))
1274
1275 # XXX: Workaround for https://github.com/yt-dlp/yt-dlp/issues/4485
1276 if sanitize and value == '':
1277 value = None
1278 return value
1279
1280 na = self.params.get('outtmpl_na_placeholder', 'NA')
1281
1282 def filename_sanitizer(key, value, restricted=self.params.get('restrictfilenames')):
1283 return sanitize_filename(str(value), restricted=restricted, is_id=(
1284 bool(re.search(r'(^|[_.])id(\.|$)', key))
1285 if 'filename-sanitization' in self.params['compat_opts']
1286 else NO_DEFAULT))
1287
1288 sanitizer = sanitize if callable(sanitize) else filename_sanitizer
1289 sanitize = bool(sanitize)
1290
1291 def _dumpjson_default(obj):
1292 if isinstance(obj, (set, LazyList)):
1293 return list(obj)
1294 return repr(obj)
1295
1296 class _ReplacementFormatter(string.Formatter):
1297 def get_field(self, field_name, args, kwargs):
1298 if field_name.isdigit():
1299 return args[0], -1
1300 raise ValueError('Unsupported field')
1301
1302 replacement_formatter = _ReplacementFormatter()
1303
1304 def create_key(outer_mobj):
1305 if not outer_mobj.group('has_key'):
1306 return outer_mobj.group(0)
1307 key = outer_mobj.group('key')
1308 mobj = re.match(INTERNAL_FORMAT_RE, key)
1309 value, replacement, default, last_field = None, None, na, ''
1310 while mobj:
1311 mobj = mobj.groupdict()
1312 default = mobj['default'] if mobj['default'] is not None else default
1313 value = get_value(mobj)
1314 last_field, replacement = mobj['fields'], mobj['replacement']
1315 if value is None and mobj['alternate']:
1316 mobj = re.match(INTERNAL_FORMAT_RE, mobj['remaining'][1:])
1317 else:
1318 break
1319
1320 if None not in (value, replacement):
1321 try:
1322 value = replacement_formatter.format(replacement, value)
1323 except ValueError:
1324 value, default = None, na
1325
1326 fmt = outer_mobj.group('format')
1327 if fmt == 's' and last_field in field_size_compat_map.keys() and isinstance(value, int):
1328 fmt = f'0{field_size_compat_map[last_field]:d}d'
1329
1330 flags = outer_mobj.group('conversion') or ''
1331 str_fmt = f'{fmt[:-1]}s'
1332 if value is None:
1333 value, fmt = default, 's'
1334 elif fmt[-1] == 'l': # list
1335 delim = '\n' if '#' in flags else ', '
1336 value, fmt = delim.join(map(str, variadic(value, allowed_types=(str, bytes)))), str_fmt
1337 elif fmt[-1] == 'j': # json
1338 value, fmt = json.dumps(
1339 value, default=_dumpjson_default,
1340 indent=4 if '#' in flags else None, ensure_ascii='+' not in flags), str_fmt
1341 elif fmt[-1] == 'h': # html
1342 value, fmt = escapeHTML(str(value)), str_fmt
1343 elif fmt[-1] == 'q': # quoted
1344 value = map(str, variadic(value) if '#' in flags else [value])
1345 value, fmt = ' '.join(map(compat_shlex_quote, value)), str_fmt
1346 elif fmt[-1] == 'B': # bytes
1347 value = f'%{str_fmt}'.encode() % str(value).encode()
1348 value, fmt = value.decode('utf-8', 'ignore'), 's'
1349 elif fmt[-1] == 'U': # unicode normalized
1350 value, fmt = unicodedata.normalize(
1351 # "+" = compatibility equivalence, "#" = NFD
1352 'NF%s%s' % ('K' if '+' in flags else '', 'D' if '#' in flags else 'C'),
1353 value), str_fmt
1354 elif fmt[-1] == 'D': # decimal suffix
1355 num_fmt, fmt = fmt[:-1].replace('#', ''), 's'
1356 value = format_decimal_suffix(value, f'%{num_fmt}f%s' if num_fmt else '%d%s',
1357 factor=1024 if '#' in flags else 1000)
1358 elif fmt[-1] == 'S': # filename sanitization
1359 value, fmt = filename_sanitizer(last_field, value, restricted='#' in flags), str_fmt
1360 elif fmt[-1] == 'c':
1361 if value:
1362 value = str(value)[0]
1363 else:
1364 fmt = str_fmt
1365 elif fmt[-1] not in 'rsa': # numeric
1366 value = float_or_none(value)
1367 if value is None:
1368 value, fmt = default, 's'
1369
1370 if sanitize:
1371 # If value is an object, sanitize might convert it to a string
1372 # So we convert it to repr first
1373 if fmt[-1] == 'r':
1374 value, fmt = repr(value), str_fmt
1375 elif fmt[-1] == 'a':
1376 value, fmt = ascii(value), str_fmt
1377 if fmt[-1] in 'csra':
1378 value = sanitizer(last_field, value)
1379
1380 key = '%s\0%s' % (key.replace('%', '%\0'), outer_mobj.group('format'))
1381 TMPL_DICT[key] = value
1382 return '{prefix}%({key}){fmt}'.format(key=key, fmt=fmt, prefix=outer_mobj.group('prefix'))
1383
1384 return EXTERNAL_FORMAT_RE.sub(create_key, outtmpl), TMPL_DICT
1385
1386 def evaluate_outtmpl(self, outtmpl, info_dict, *args, **kwargs):
1387 outtmpl, info_dict = self.prepare_outtmpl(outtmpl, info_dict, *args, **kwargs)
1388 return self.escape_outtmpl(outtmpl) % info_dict
1389
1390 def _prepare_filename(self, info_dict, *, outtmpl=None, tmpl_type=None):
1391 assert None in (outtmpl, tmpl_type), 'outtmpl and tmpl_type are mutually exclusive'
1392 if outtmpl is None:
1393 outtmpl = self.params['outtmpl'].get(tmpl_type or 'default', self.params['outtmpl']['default'])
1394 try:
1395 outtmpl = self._outtmpl_expandpath(outtmpl)
1396 filename = self.evaluate_outtmpl(outtmpl, info_dict, True)
1397 if not filename:
1398 return None
1399
1400 if tmpl_type in ('', 'temp'):
1401 final_ext, ext = self.params.get('final_ext'), info_dict.get('ext')
1402 if final_ext and ext and final_ext != ext and filename.endswith(f'.{final_ext}'):
1403 filename = replace_extension(filename, ext, final_ext)
1404 elif tmpl_type:
1405 force_ext = OUTTMPL_TYPES[tmpl_type]
1406 if force_ext:
1407 filename = replace_extension(filename, force_ext, info_dict.get('ext'))
1408
1409 # https://github.com/blackjack4494/youtube-dlc/issues/85
1410 trim_file_name = self.params.get('trim_file_name', False)
1411 if trim_file_name:
1412 no_ext, *ext = filename.rsplit('.', 2)
1413 filename = join_nonempty(no_ext[:trim_file_name], *ext, delim='.')
1414
1415 return filename
1416 except ValueError as err:
1417 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
1418 return None
1419
1420 def prepare_filename(self, info_dict, dir_type='', *, outtmpl=None, warn=False):
1421 """Generate the output filename"""
1422 if outtmpl:
1423 assert not dir_type, 'outtmpl and dir_type are mutually exclusive'
1424 dir_type = None
1425 filename = self._prepare_filename(info_dict, tmpl_type=dir_type, outtmpl=outtmpl)
1426 if not filename and dir_type not in ('', 'temp'):
1427 return ''
1428
1429 if warn:
1430 if not self.params.get('paths'):
1431 pass
1432 elif filename == '-':
1433 self.report_warning('--paths is ignored when an outputting to stdout', only_once=True)
1434 elif os.path.isabs(filename):
1435 self.report_warning('--paths is ignored since an absolute path is given in output template', only_once=True)
1436 if filename == '-' or not filename:
1437 return filename
1438
1439 return self.get_output_path(dir_type, filename)
1440
1441 def _match_entry(self, info_dict, incomplete=False, silent=False):
1442 """Returns None if the file should be downloaded"""
1443 _type = 'video' if 'playlist-match-filter' in self.params['compat_opts'] else info_dict.get('_type', 'video')
1444 assert incomplete or _type == 'video', 'Only video result can be considered complete'
1445
1446 video_title = info_dict.get('title', info_dict.get('id', 'entry'))
1447
1448 def check_filter():
1449 if _type in ('playlist', 'multi_video'):
1450 return
1451 elif _type in ('url', 'url_transparent') and not try_call(
1452 lambda: self.get_info_extractor(info_dict['ie_key']).is_single_video(info_dict['url'])):
1453 return
1454
1455 if 'title' in info_dict:
1456 # This can happen when we're just evaluating the playlist
1457 title = info_dict['title']
1458 matchtitle = self.params.get('matchtitle', False)
1459 if matchtitle:
1460 if not re.search(matchtitle, title, re.IGNORECASE):
1461 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
1462 rejecttitle = self.params.get('rejecttitle', False)
1463 if rejecttitle:
1464 if re.search(rejecttitle, title, re.IGNORECASE):
1465 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
1466
1467 date = info_dict.get('upload_date')
1468 if date is not None:
1469 dateRange = self.params.get('daterange', DateRange())
1470 if date not in dateRange:
1471 return f'{date_from_str(date).isoformat()} upload date is not in range {dateRange}'
1472 view_count = info_dict.get('view_count')
1473 if view_count is not None:
1474 min_views = self.params.get('min_views')
1475 if min_views is not None and view_count < min_views:
1476 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
1477 max_views = self.params.get('max_views')
1478 if max_views is not None and view_count > max_views:
1479 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
1480 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
1481 return 'Skipping "%s" because it is age restricted' % video_title
1482
1483 match_filter = self.params.get('match_filter')
1484 if match_filter is None:
1485 return None
1486
1487 cancelled = None
1488 try:
1489 try:
1490 ret = match_filter(info_dict, incomplete=incomplete)
1491 except TypeError:
1492 # For backward compatibility
1493 ret = None if incomplete else match_filter(info_dict)
1494 except DownloadCancelled as err:
1495 if err.msg is not NO_DEFAULT:
1496 raise
1497 ret, cancelled = err.msg, err
1498
1499 if ret is NO_DEFAULT:
1500 while True:
1501 filename = self._format_screen(self.prepare_filename(info_dict), self.Styles.FILENAME)
1502 reply = input(self._format_screen(
1503 f'Download "{filename}"? (Y/n): ', self.Styles.EMPHASIS)).lower().strip()
1504 if reply in {'y', ''}:
1505 return None
1506 elif reply == 'n':
1507 if cancelled:
1508 raise type(cancelled)(f'Skipping {video_title}')
1509 return f'Skipping {video_title}'
1510 return ret
1511
1512 if self.in_download_archive(info_dict):
1513 reason = ''.join((
1514 format_field(info_dict, 'id', f'{self._format_screen("%s", self.Styles.ID)}: '),
1515 format_field(info_dict, 'title', f'{self._format_screen("%s", self.Styles.EMPHASIS)} '),
1516 'has already been recorded in the archive'))
1517 break_opt, break_err = 'break_on_existing', ExistingVideoReached
1518 else:
1519 try:
1520 reason = check_filter()
1521 except DownloadCancelled as e:
1522 reason, break_opt, break_err = e.msg, 'match_filter', type(e)
1523 else:
1524 break_opt, break_err = 'break_on_reject', RejectedVideoReached
1525 if reason is not None:
1526 if not silent:
1527 self.to_screen('[download] ' + reason)
1528 if self.params.get(break_opt, False):
1529 raise break_err()
1530 return reason
1531
1532 @staticmethod
1533 def add_extra_info(info_dict, extra_info):
1534 '''Set the keys from extra_info in info dict if they are missing'''
1535 for key, value in extra_info.items():
1536 info_dict.setdefault(key, value)
1537
1538 def extract_info(self, url, download=True, ie_key=None, extra_info=None,
1539 process=True, force_generic_extractor=False):
1540 """
1541 Extract and return the information dictionary of the URL
1542
1543 Arguments:
1544 @param url URL to extract
1545
1546 Keyword arguments:
1547 @param download Whether to download videos
1548 @param process Whether to resolve all unresolved references (URLs, playlist items).
1549 Must be True for download to work
1550 @param ie_key Use only the extractor with this key
1551
1552 @param extra_info Dictionary containing the extra values to add to the info (For internal use only)
1553 @force_generic_extractor Force using the generic extractor (Deprecated; use ie_key='Generic')
1554 """
1555
1556 if extra_info is None:
1557 extra_info = {}
1558
1559 if not ie_key and force_generic_extractor:
1560 ie_key = 'Generic'
1561
1562 if ie_key:
1563 ies = {ie_key: self._ies[ie_key]} if ie_key in self._ies else {}
1564 else:
1565 ies = self._ies
1566
1567 for key, ie in ies.items():
1568 if not ie.suitable(url):
1569 continue
1570
1571 if not ie.working():
1572 self.report_warning('The program functionality for this site has been marked as broken, '
1573 'and will probably not work.')
1574
1575 temp_id = ie.get_temp_id(url)
1576 if temp_id is not None and self.in_download_archive({'id': temp_id, 'ie_key': key}):
1577 self.to_screen(f'[download] {self._format_screen(temp_id, self.Styles.ID)}: '
1578 'has already been recorded in the archive')
1579 if self.params.get('break_on_existing', False):
1580 raise ExistingVideoReached()
1581 break
1582 return self.__extract_info(url, self.get_info_extractor(key), download, extra_info, process)
1583 else:
1584 extractors_restricted = self.params.get('allowed_extractors') not in (None, ['default'])
1585 self.report_error(f'No suitable extractor{format_field(ie_key, None, " (%s)")} found for URL {url}',
1586 tb=False if extractors_restricted else None)
1587
1588 def _handle_extraction_exceptions(func):
1589 @functools.wraps(func)
1590 def wrapper(self, *args, **kwargs):
1591 while True:
1592 try:
1593 return func(self, *args, **kwargs)
1594 except (DownloadCancelled, LazyList.IndexError, PagedList.IndexError):
1595 raise
1596 except ReExtractInfo as e:
1597 if e.expected:
1598 self.to_screen(f'{e}; Re-extracting data')
1599 else:
1600 self.to_stderr('\r')
1601 self.report_warning(f'{e}; Re-extracting data')
1602 continue
1603 except GeoRestrictedError as e:
1604 msg = e.msg
1605 if e.countries:
1606 msg += '\nThis video is available in %s.' % ', '.join(
1607 map(ISO3166Utils.short2full, e.countries))
1608 msg += '\nYou might want to use a VPN or a proxy server (with --proxy) to workaround.'
1609 self.report_error(msg)
1610 except ExtractorError as e: # An error we somewhat expected
1611 self.report_error(str(e), e.format_traceback())
1612 except Exception as e:
1613 if self.params.get('ignoreerrors'):
1614 self.report_error(str(e), tb=encode_compat_str(traceback.format_exc()))
1615 else:
1616 raise
1617 break
1618 return wrapper
1619
1620 def _wait_for_video(self, ie_result={}):
1621 if (not self.params.get('wait_for_video')
1622 or ie_result.get('_type', 'video') != 'video'
1623 or ie_result.get('formats') or ie_result.get('url')):
1624 return
1625
1626 format_dur = lambda dur: '%02d:%02d:%02d' % timetuple_from_msec(dur * 1000)[:-1]
1627 last_msg = ''
1628
1629 def progress(msg):
1630 nonlocal last_msg
1631 full_msg = f'{msg}\n'
1632 if not self.params.get('noprogress'):
1633 full_msg = msg + ' ' * (len(last_msg) - len(msg)) + '\r'
1634 elif last_msg:
1635 return
1636 self.to_screen(full_msg, skip_eol=True)
1637 last_msg = msg
1638
1639 min_wait, max_wait = self.params.get('wait_for_video')
1640 diff = try_get(ie_result, lambda x: x['release_timestamp'] - time.time())
1641 if diff is None and ie_result.get('live_status') == 'is_upcoming':
1642 diff = round(random.uniform(min_wait, max_wait) if (max_wait and min_wait) else (max_wait or min_wait), 0)
1643 self.report_warning('Release time of video is not known')
1644 elif ie_result and (diff or 0) <= 0:
1645 self.report_warning('Video should already be available according to extracted info')
1646 diff = min(max(diff or 0, min_wait or 0), max_wait or float('inf'))
1647 self.to_screen(f'[wait] Waiting for {format_dur(diff)} - Press Ctrl+C to try now')
1648
1649 wait_till = time.time() + diff
1650 try:
1651 while True:
1652 diff = wait_till - time.time()
1653 if diff <= 0:
1654 progress('')
1655 raise ReExtractInfo('[wait] Wait period ended', expected=True)
1656 progress(f'[wait] Remaining time until next attempt: {self._format_screen(format_dur(diff), self.Styles.EMPHASIS)}')
1657 time.sleep(1)
1658 except KeyboardInterrupt:
1659 progress('')
1660 raise ReExtractInfo('[wait] Interrupted by user', expected=True)
1661 except BaseException as e:
1662 if not isinstance(e, ReExtractInfo):
1663 self.to_screen('')
1664 raise
1665
1666 def _load_cookies(self, data, *, autoscope=True):
1667 """Loads cookies from a `Cookie` header
1668
1669 This tries to work around the security vulnerability of passing cookies to every domain.
1670 See: https://github.com/yt-dlp/yt-dlp/security/advisories/GHSA-v8mc-9377-rwjj
1671
1672 @param data The Cookie header as string to load the cookies from
1673 @param autoscope If `False`, scope cookies using Set-Cookie syntax and error for cookie without domains
1674 If `True`, save cookies for later to be stored in the jar with a limited scope
1675 If a URL, save cookies in the jar with the domain of the URL
1676 """
1677 for cookie in LenientSimpleCookie(data).values():
1678 if autoscope and any(cookie.values()):
1679 raise ValueError('Invalid syntax in Cookie Header')
1680
1681 domain = cookie.get('domain') or ''
1682 expiry = cookie.get('expires')
1683 if expiry == '': # 0 is valid
1684 expiry = None
1685 prepared_cookie = http.cookiejar.Cookie(
1686 cookie.get('version') or 0, cookie.key, cookie.value, None, False,
1687 domain, True, True, cookie.get('path') or '', bool(cookie.get('path')),
1688 cookie.get('secure') or False, expiry, False, None, None, {})
1689
1690 if domain:
1691 self.cookiejar.set_cookie(prepared_cookie)
1692 elif autoscope is True:
1693 self.deprecated_feature(
1694 'Passing cookies as a header is a potential security risk; '
1695 'they will be scoped to the domain of the downloaded urls. '
1696 'Please consider loading cookies from a file or browser instead.')
1697 self.__header_cookies.append(prepared_cookie)
1698 elif autoscope:
1699 self.report_warning(
1700 'The extractor result contains an unscoped cookie as an HTTP header. '
1701 f'If you are using yt-dlp with an input URL{bug_reports_message(before=",")}',
1702 only_once=True)
1703 self._apply_header_cookies(autoscope, [prepared_cookie])
1704 else:
1705 self.report_error('Unscoped cookies are not allowed; please specify some sort of scoping',
1706 tb=False, is_error=False)
1707
1708 def _apply_header_cookies(self, url, cookies=None):
1709 """Applies stray header cookies to the provided url
1710
1711 This loads header cookies and scopes them to the domain provided in `url`.
1712 While this is not ideal, it helps reduce the risk of them being sent
1713 to an unintended destination while mostly maintaining compatibility.
1714 """
1715 parsed = urllib.parse.urlparse(url)
1716 if not parsed.hostname:
1717 return
1718
1719 for cookie in map(copy.copy, cookies or self.__header_cookies):
1720 cookie.domain = f'.{parsed.hostname}'
1721 self.cookiejar.set_cookie(cookie)
1722
1723 @_handle_extraction_exceptions
1724 def __extract_info(self, url, ie, download, extra_info, process):
1725 self._apply_header_cookies(url)
1726
1727 try:
1728 ie_result = ie.extract(url)
1729 except UserNotLive as e:
1730 if process:
1731 if self.params.get('wait_for_video'):
1732 self.report_warning(e)
1733 self._wait_for_video()
1734 raise
1735 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
1736 self.report_warning(f'Extractor {ie.IE_NAME} returned nothing{bug_reports_message()}')
1737 return
1738 if isinstance(ie_result, list):
1739 # Backwards compatibility: old IE result format
1740 ie_result = {
1741 '_type': 'compat_list',
1742 'entries': ie_result,
1743 }
1744 if extra_info.get('original_url'):
1745 ie_result.setdefault('original_url', extra_info['original_url'])
1746 self.add_default_extra_info(ie_result, ie, url)
1747 if process:
1748 self._wait_for_video(ie_result)
1749 return self.process_ie_result(ie_result, download, extra_info)
1750 else:
1751 return ie_result
1752
1753 def add_default_extra_info(self, ie_result, ie, url):
1754 if url is not None:
1755 self.add_extra_info(ie_result, {
1756 'webpage_url': url,
1757 'original_url': url,
1758 })
1759 webpage_url = ie_result.get('webpage_url')
1760 if webpage_url:
1761 self.add_extra_info(ie_result, {
1762 'webpage_url_basename': url_basename(webpage_url),
1763 'webpage_url_domain': get_domain(webpage_url),
1764 })
1765 if ie is not None:
1766 self.add_extra_info(ie_result, {
1767 'extractor': ie.IE_NAME,
1768 'extractor_key': ie.ie_key(),
1769 })
1770
1771 def process_ie_result(self, ie_result, download=True, extra_info=None):
1772 """
1773 Take the result of the ie(may be modified) and resolve all unresolved
1774 references (URLs, playlist items).
1775
1776 It will also download the videos if 'download'.
1777 Returns the resolved ie_result.
1778 """
1779 if extra_info is None:
1780 extra_info = {}
1781 result_type = ie_result.get('_type', 'video')
1782
1783 if result_type in ('url', 'url_transparent'):
1784 ie_result['url'] = sanitize_url(
1785 ie_result['url'], scheme='http' if self.params.get('prefer_insecure') else 'https')
1786 if ie_result.get('original_url') and not extra_info.get('original_url'):
1787 extra_info = {'original_url': ie_result['original_url'], **extra_info}
1788
1789 extract_flat = self.params.get('extract_flat', False)
1790 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info)
1791 or extract_flat is True):
1792 info_copy = ie_result.copy()
1793 ie = try_get(ie_result.get('ie_key'), self.get_info_extractor)
1794 if ie and not ie_result.get('id'):
1795 info_copy['id'] = ie.get_temp_id(ie_result['url'])
1796 self.add_default_extra_info(info_copy, ie, ie_result['url'])
1797 self.add_extra_info(info_copy, extra_info)
1798 info_copy, _ = self.pre_process(info_copy)
1799 self._fill_common_fields(info_copy, False)
1800 self.__forced_printings(info_copy)
1801 self._raise_pending_errors(info_copy)
1802 if self.params.get('force_write_download_archive', False):
1803 self.record_download_archive(info_copy)
1804 return ie_result
1805
1806 if result_type == 'video':
1807 self.add_extra_info(ie_result, extra_info)
1808 ie_result = self.process_video_result(ie_result, download=download)
1809 self._raise_pending_errors(ie_result)
1810 additional_urls = (ie_result or {}).get('additional_urls')
1811 if additional_urls:
1812 # TODO: Improve MetadataParserPP to allow setting a list
1813 if isinstance(additional_urls, str):
1814 additional_urls = [additional_urls]
1815 self.to_screen(
1816 '[info] %s: %d additional URL(s) requested' % (ie_result['id'], len(additional_urls)))
1817 self.write_debug('Additional URLs: "%s"' % '", "'.join(additional_urls))
1818 ie_result['additional_entries'] = [
1819 self.extract_info(
1820 url, download, extra_info=extra_info,
1821 force_generic_extractor=self.params.get('force_generic_extractor'))
1822 for url in additional_urls
1823 ]
1824 return ie_result
1825 elif result_type == 'url':
1826 # We have to add extra_info to the results because it may be
1827 # contained in a playlist
1828 return self.extract_info(
1829 ie_result['url'], download,
1830 ie_key=ie_result.get('ie_key'),
1831 extra_info=extra_info)
1832 elif result_type == 'url_transparent':
1833 # Use the information from the embedding page
1834 info = self.extract_info(
1835 ie_result['url'], ie_key=ie_result.get('ie_key'),
1836 extra_info=extra_info, download=False, process=False)
1837
1838 # extract_info may return None when ignoreerrors is enabled and
1839 # extraction failed with an error, don't crash and return early
1840 # in this case
1841 if not info:
1842 return info
1843
1844 exempted_fields = {'_type', 'url', 'ie_key'}
1845 if not ie_result.get('section_end') and ie_result.get('section_start') is None:
1846 # For video clips, the id etc of the clip extractor should be used
1847 exempted_fields |= {'id', 'extractor', 'extractor_key'}
1848
1849 new_result = info.copy()
1850 new_result.update(filter_dict(ie_result, lambda k, v: v is not None and k not in exempted_fields))
1851
1852 # Extracted info may not be a video result (i.e.
1853 # info.get('_type', 'video') != video) but rather an url or
1854 # url_transparent. In such cases outer metadata (from ie_result)
1855 # should be propagated to inner one (info). For this to happen
1856 # _type of info should be overridden with url_transparent. This
1857 # fixes issue from https://github.com/ytdl-org/youtube-dl/pull/11163.
1858 if new_result.get('_type') == 'url':
1859 new_result['_type'] = 'url_transparent'
1860
1861 return self.process_ie_result(
1862 new_result, download=download, extra_info=extra_info)
1863 elif result_type in ('playlist', 'multi_video'):
1864 # Protect from infinite recursion due to recursively nested playlists
1865 # (see https://github.com/ytdl-org/youtube-dl/issues/27833)
1866 webpage_url = ie_result.get('webpage_url') # Playlists maynot have webpage_url
1867 if webpage_url and webpage_url in self._playlist_urls:
1868 self.to_screen(
1869 '[download] Skipping already downloaded playlist: %s'
1870 % ie_result.get('title') or ie_result.get('id'))
1871 return
1872
1873 self._playlist_level += 1
1874 self._playlist_urls.add(webpage_url)
1875 self._fill_common_fields(ie_result, False)
1876 self._sanitize_thumbnails(ie_result)
1877 try:
1878 return self.__process_playlist(ie_result, download)
1879 finally:
1880 self._playlist_level -= 1
1881 if not self._playlist_level:
1882 self._playlist_urls.clear()
1883 elif result_type == 'compat_list':
1884 self.report_warning(
1885 'Extractor %s returned a compat_list result. '
1886 'It needs to be updated.' % ie_result.get('extractor'))
1887
1888 def _fixup(r):
1889 self.add_extra_info(r, {
1890 'extractor': ie_result['extractor'],
1891 'webpage_url': ie_result['webpage_url'],
1892 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1893 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1894 'extractor_key': ie_result['extractor_key'],
1895 })
1896 return r
1897 ie_result['entries'] = [
1898 self.process_ie_result(_fixup(r), download, extra_info)
1899 for r in ie_result['entries']
1900 ]
1901 return ie_result
1902 else:
1903 raise Exception('Invalid result type: %s' % result_type)
1904
1905 def _ensure_dir_exists(self, path):
1906 return make_dir(path, self.report_error)
1907
1908 @staticmethod
1909 def _playlist_infodict(ie_result, strict=False, **kwargs):
1910 info = {
1911 'playlist_count': ie_result.get('playlist_count'),
1912 'playlist': ie_result.get('title') or ie_result.get('id'),
1913 'playlist_id': ie_result.get('id'),
1914 'playlist_title': ie_result.get('title'),
1915 'playlist_uploader': ie_result.get('uploader'),
1916 'playlist_uploader_id': ie_result.get('uploader_id'),
1917 **kwargs,
1918 }
1919 if strict:
1920 return info
1921 if ie_result.get('webpage_url'):
1922 info.update({
1923 'webpage_url': ie_result['webpage_url'],
1924 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1925 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1926 })
1927 return {
1928 **info,
1929 'playlist_index': 0,
1930 '__last_playlist_index': max(ie_result.get('requested_entries') or (0, 0)),
1931 'extractor': ie_result['extractor'],
1932 'extractor_key': ie_result['extractor_key'],
1933 }
1934
1935 def __process_playlist(self, ie_result, download):
1936 """Process each entry in the playlist"""
1937 assert ie_result['_type'] in ('playlist', 'multi_video')
1938
1939 common_info = self._playlist_infodict(ie_result, strict=True)
1940 title = common_info.get('playlist') or '<Untitled>'
1941 if self._match_entry(common_info, incomplete=True) is not None:
1942 return
1943 self.to_screen(f'[download] Downloading {ie_result["_type"]}: {title}')
1944
1945 all_entries = PlaylistEntries(self, ie_result)
1946 entries = orderedSet(all_entries.get_requested_items(), lazy=True)
1947
1948 lazy = self.params.get('lazy_playlist')
1949 if lazy:
1950 resolved_entries, n_entries = [], 'N/A'
1951 ie_result['requested_entries'], ie_result['entries'] = None, None
1952 else:
1953 entries = resolved_entries = list(entries)
1954 n_entries = len(resolved_entries)
1955 ie_result['requested_entries'], ie_result['entries'] = tuple(zip(*resolved_entries)) or ([], [])
1956 if not ie_result.get('playlist_count'):
1957 # Better to do this after potentially exhausting entries
1958 ie_result['playlist_count'] = all_entries.get_full_count()
1959
1960 extra = self._playlist_infodict(ie_result, n_entries=int_or_none(n_entries))
1961 ie_copy = collections.ChainMap(ie_result, extra)
1962
1963 _infojson_written = False
1964 write_playlist_files = self.params.get('allow_playlist_files', True)
1965 if write_playlist_files and self.params.get('list_thumbnails'):
1966 self.list_thumbnails(ie_result)
1967 if write_playlist_files and not self.params.get('simulate'):
1968 _infojson_written = self._write_info_json(
1969 'playlist', ie_result, self.prepare_filename(ie_copy, 'pl_infojson'))
1970 if _infojson_written is None:
1971 return
1972 if self._write_description('playlist', ie_result,
1973 self.prepare_filename(ie_copy, 'pl_description')) is None:
1974 return
1975 # TODO: This should be passed to ThumbnailsConvertor if necessary
1976 self._write_thumbnails('playlist', ie_result, self.prepare_filename(ie_copy, 'pl_thumbnail'))
1977
1978 if lazy:
1979 if self.params.get('playlistreverse') or self.params.get('playlistrandom'):
1980 self.report_warning('playlistreverse and playlistrandom are not supported with lazy_playlist', only_once=True)
1981 elif self.params.get('playlistreverse'):
1982 entries.reverse()
1983 elif self.params.get('playlistrandom'):
1984 random.shuffle(entries)
1985
1986 self.to_screen(f'[{ie_result["extractor"]}] Playlist {title}: Downloading {n_entries} items'
1987 f'{format_field(ie_result, "playlist_count", " of %s")}')
1988
1989 keep_resolved_entries = self.params.get('extract_flat') != 'discard'
1990 if self.params.get('extract_flat') == 'discard_in_playlist':
1991 keep_resolved_entries = ie_result['_type'] != 'playlist'
1992 if keep_resolved_entries:
1993 self.write_debug('The information of all playlist entries will be held in memory')
1994
1995 failures = 0
1996 max_failures = self.params.get('skip_playlist_after_errors') or float('inf')
1997 for i, (playlist_index, entry) in enumerate(entries):
1998 if lazy:
1999 resolved_entries.append((playlist_index, entry))
2000 if not entry:
2001 continue
2002
2003 entry['__x_forwarded_for_ip'] = ie_result.get('__x_forwarded_for_ip')
2004 if not lazy and 'playlist-index' in self.params['compat_opts']:
2005 playlist_index = ie_result['requested_entries'][i]
2006
2007 entry_copy = collections.ChainMap(entry, {
2008 **common_info,
2009 'n_entries': int_or_none(n_entries),
2010 'playlist_index': playlist_index,
2011 'playlist_autonumber': i + 1,
2012 })
2013
2014 if self._match_entry(entry_copy, incomplete=True) is not None:
2015 # For compatabilty with youtube-dl. See https://github.com/yt-dlp/yt-dlp/issues/4369
2016 resolved_entries[i] = (playlist_index, NO_DEFAULT)
2017 continue
2018
2019 self.to_screen('[download] Downloading item %s of %s' % (
2020 self._format_screen(i + 1, self.Styles.ID), self._format_screen(n_entries, self.Styles.EMPHASIS)))
2021
2022 entry_result = self.__process_iterable_entry(entry, download, collections.ChainMap({
2023 'playlist_index': playlist_index,
2024 'playlist_autonumber': i + 1,
2025 }, extra))
2026 if not entry_result:
2027 failures += 1
2028 if failures >= max_failures:
2029 self.report_error(
2030 f'Skipping the remaining entries in playlist "{title}" since {failures} items failed extraction')
2031 break
2032 if keep_resolved_entries:
2033 resolved_entries[i] = (playlist_index, entry_result)
2034
2035 # Update with processed data
2036 ie_result['entries'] = [e for _, e in resolved_entries if e is not NO_DEFAULT]
2037 ie_result['requested_entries'] = [i for i, e in resolved_entries if e is not NO_DEFAULT]
2038 if ie_result['requested_entries'] == try_call(lambda: list(range(1, ie_result['playlist_count'] + 1))):
2039 # Do not set for full playlist
2040 ie_result.pop('requested_entries')
2041
2042 # Write the updated info to json
2043 if _infojson_written is True and self._write_info_json(
2044 'updated playlist', ie_result,
2045 self.prepare_filename(ie_copy, 'pl_infojson'), overwrite=True) is None:
2046 return
2047
2048 ie_result = self.run_all_pps('playlist', ie_result)
2049 self.to_screen(f'[download] Finished downloading playlist: {title}')
2050 return ie_result
2051
2052 @_handle_extraction_exceptions
2053 def __process_iterable_entry(self, entry, download, extra_info):
2054 return self.process_ie_result(
2055 entry, download=download, extra_info=extra_info)
2056
2057 def _build_format_filter(self, filter_spec):
2058 " Returns a function to filter the formats according to the filter_spec "
2059
2060 OPERATORS = {
2061 '<': operator.lt,
2062 '<=': operator.le,
2063 '>': operator.gt,
2064 '>=': operator.ge,
2065 '=': operator.eq,
2066 '!=': operator.ne,
2067 }
2068 operator_rex = re.compile(r'''(?x)\s*
2069 (?P<key>[\w.-]+)\s*
2070 (?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
2071 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)\s*
2072 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
2073 m = operator_rex.fullmatch(filter_spec)
2074 if m:
2075 try:
2076 comparison_value = int(m.group('value'))
2077 except ValueError:
2078 comparison_value = parse_filesize(m.group('value'))
2079 if comparison_value is None:
2080 comparison_value = parse_filesize(m.group('value') + 'B')
2081 if comparison_value is None:
2082 raise ValueError(
2083 'Invalid value %r in format specification %r' % (
2084 m.group('value'), filter_spec))
2085 op = OPERATORS[m.group('op')]
2086
2087 if not m:
2088 STR_OPERATORS = {
2089 '=': operator.eq,
2090 '^=': lambda attr, value: attr.startswith(value),
2091 '$=': lambda attr, value: attr.endswith(value),
2092 '*=': lambda attr, value: value in attr,
2093 '~=': lambda attr, value: value.search(attr) is not None
2094 }
2095 str_operator_rex = re.compile(r'''(?x)\s*
2096 (?P<key>[a-zA-Z0-9._-]+)\s*
2097 (?P<negation>!\s*)?(?P<op>%s)\s*(?P<none_inclusive>\?\s*)?
2098 (?P<quote>["'])?
2099 (?P<value>(?(quote)(?:(?!(?P=quote))[^\\]|\\.)+|[\w.-]+))
2100 (?(quote)(?P=quote))\s*
2101 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
2102 m = str_operator_rex.fullmatch(filter_spec)
2103 if m:
2104 if m.group('op') == '~=':
2105 comparison_value = re.compile(m.group('value'))
2106 else:
2107 comparison_value = re.sub(r'''\\([\\"'])''', r'\1', m.group('value'))
2108 str_op = STR_OPERATORS[m.group('op')]
2109 if m.group('negation'):
2110 op = lambda attr, value: not str_op(attr, value)
2111 else:
2112 op = str_op
2113
2114 if not m:
2115 raise SyntaxError('Invalid filter specification %r' % filter_spec)
2116
2117 def _filter(f):
2118 actual_value = f.get(m.group('key'))
2119 if actual_value is None:
2120 return m.group('none_inclusive')
2121 return op(actual_value, comparison_value)
2122 return _filter
2123
2124 def _check_formats(self, formats):
2125 for f in formats:
2126 self.to_screen('[info] Testing format %s' % f['format_id'])
2127 path = self.get_output_path('temp')
2128 if not self._ensure_dir_exists(f'{path}/'):
2129 continue
2130 temp_file = tempfile.NamedTemporaryFile(suffix='.tmp', delete=False, dir=path or None)
2131 temp_file.close()
2132 try:
2133 success, _ = self.dl(temp_file.name, f, test=True)
2134 except (DownloadError, OSError, ValueError) + network_exceptions:
2135 success = False
2136 finally:
2137 if os.path.exists(temp_file.name):
2138 try:
2139 os.remove(temp_file.name)
2140 except OSError:
2141 self.report_warning('Unable to delete temporary file "%s"' % temp_file.name)
2142 if success:
2143 yield f
2144 else:
2145 self.to_screen('[info] Unable to download format %s. Skipping...' % f['format_id'])
2146
2147 def _default_format_spec(self, info_dict, download=True):
2148
2149 def can_merge():
2150 merger = FFmpegMergerPP(self)
2151 return merger.available and merger.can_merge()
2152
2153 prefer_best = (
2154 not self.params.get('simulate')
2155 and download
2156 and (
2157 not can_merge()
2158 or info_dict.get('is_live') and not self.params.get('live_from_start')
2159 or self.params['outtmpl']['default'] == '-'))
2160 compat = (
2161 prefer_best
2162 or self.params.get('allow_multiple_audio_streams', False)
2163 or 'format-spec' in self.params['compat_opts'])
2164
2165 return (
2166 'best/bestvideo+bestaudio' if prefer_best
2167 else 'bestvideo*+bestaudio/best' if not compat
2168 else 'bestvideo+bestaudio/best')
2169
2170 def build_format_selector(self, format_spec):
2171 def syntax_error(note, start):
2172 message = (
2173 'Invalid format specification: '
2174 '{}\n\t{}\n\t{}^'.format(note, format_spec, ' ' * start[1]))
2175 return SyntaxError(message)
2176
2177 PICKFIRST = 'PICKFIRST'
2178 MERGE = 'MERGE'
2179 SINGLE = 'SINGLE'
2180 GROUP = 'GROUP'
2181 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
2182
2183 allow_multiple_streams = {'audio': self.params.get('allow_multiple_audio_streams', False),
2184 'video': self.params.get('allow_multiple_video_streams', False)}
2185
2186 def _parse_filter(tokens):
2187 filter_parts = []
2188 for type, string_, start, _, _ in tokens:
2189 if type == tokenize.OP and string_ == ']':
2190 return ''.join(filter_parts)
2191 else:
2192 filter_parts.append(string_)
2193
2194 def _remove_unused_ops(tokens):
2195 # Remove operators that we don't use and join them with the surrounding strings.
2196 # E.g. 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
2197 ALLOWED_OPS = ('/', '+', ',', '(', ')')
2198 last_string, last_start, last_end, last_line = None, None, None, None
2199 for type, string_, start, end, line in tokens:
2200 if type == tokenize.OP and string_ == '[':
2201 if last_string:
2202 yield tokenize.NAME, last_string, last_start, last_end, last_line
2203 last_string = None
2204 yield type, string_, start, end, line
2205 # everything inside brackets will be handled by _parse_filter
2206 for type, string_, start, end, line in tokens:
2207 yield type, string_, start, end, line
2208 if type == tokenize.OP and string_ == ']':
2209 break
2210 elif type == tokenize.OP and string_ in ALLOWED_OPS:
2211 if last_string:
2212 yield tokenize.NAME, last_string, last_start, last_end, last_line
2213 last_string = None
2214 yield type, string_, start, end, line
2215 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
2216 if not last_string:
2217 last_string = string_
2218 last_start = start
2219 last_end = end
2220 else:
2221 last_string += string_
2222 if last_string:
2223 yield tokenize.NAME, last_string, last_start, last_end, last_line
2224
2225 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
2226 selectors = []
2227 current_selector = None
2228 for type, string_, start, _, _ in tokens:
2229 # ENCODING is only defined in python 3.x
2230 if type == getattr(tokenize, 'ENCODING', None):
2231 continue
2232 elif type in [tokenize.NAME, tokenize.NUMBER]:
2233 current_selector = FormatSelector(SINGLE, string_, [])
2234 elif type == tokenize.OP:
2235 if string_ == ')':
2236 if not inside_group:
2237 # ')' will be handled by the parentheses group
2238 tokens.restore_last_token()
2239 break
2240 elif inside_merge and string_ in ['/', ',']:
2241 tokens.restore_last_token()
2242 break
2243 elif inside_choice and string_ == ',':
2244 tokens.restore_last_token()
2245 break
2246 elif string_ == ',':
2247 if not current_selector:
2248 raise syntax_error('"," must follow a format selector', start)
2249 selectors.append(current_selector)
2250 current_selector = None
2251 elif string_ == '/':
2252 if not current_selector:
2253 raise syntax_error('"/" must follow a format selector', start)
2254 first_choice = current_selector
2255 second_choice = _parse_format_selection(tokens, inside_choice=True)
2256 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
2257 elif string_ == '[':
2258 if not current_selector:
2259 current_selector = FormatSelector(SINGLE, 'best', [])
2260 format_filter = _parse_filter(tokens)
2261 current_selector.filters.append(format_filter)
2262 elif string_ == '(':
2263 if current_selector:
2264 raise syntax_error('Unexpected "("', start)
2265 group = _parse_format_selection(tokens, inside_group=True)
2266 current_selector = FormatSelector(GROUP, group, [])
2267 elif string_ == '+':
2268 if not current_selector:
2269 raise syntax_error('Unexpected "+"', start)
2270 selector_1 = current_selector
2271 selector_2 = _parse_format_selection(tokens, inside_merge=True)
2272 if not selector_2:
2273 raise syntax_error('Expected a selector', start)
2274 current_selector = FormatSelector(MERGE, (selector_1, selector_2), [])
2275 else:
2276 raise syntax_error(f'Operator not recognized: "{string_}"', start)
2277 elif type == tokenize.ENDMARKER:
2278 break
2279 if current_selector:
2280 selectors.append(current_selector)
2281 return selectors
2282
2283 def _merge(formats_pair):
2284 format_1, format_2 = formats_pair
2285
2286 formats_info = []
2287 formats_info.extend(format_1.get('requested_formats', (format_1,)))
2288 formats_info.extend(format_2.get('requested_formats', (format_2,)))
2289
2290 if not allow_multiple_streams['video'] or not allow_multiple_streams['audio']:
2291 get_no_more = {'video': False, 'audio': False}
2292 for (i, fmt_info) in enumerate(formats_info):
2293 if fmt_info.get('acodec') == fmt_info.get('vcodec') == 'none':
2294 formats_info.pop(i)
2295 continue
2296 for aud_vid in ['audio', 'video']:
2297 if not allow_multiple_streams[aud_vid] and fmt_info.get(aud_vid[0] + 'codec') != 'none':
2298 if get_no_more[aud_vid]:
2299 formats_info.pop(i)
2300 break
2301 get_no_more[aud_vid] = True
2302
2303 if len(formats_info) == 1:
2304 return formats_info[0]
2305
2306 video_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('vcodec') != 'none']
2307 audio_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('acodec') != 'none']
2308
2309 the_only_video = video_fmts[0] if len(video_fmts) == 1 else None
2310 the_only_audio = audio_fmts[0] if len(audio_fmts) == 1 else None
2311
2312 output_ext = get_compatible_ext(
2313 vcodecs=[f.get('vcodec') for f in video_fmts],
2314 acodecs=[f.get('acodec') for f in audio_fmts],
2315 vexts=[f['ext'] for f in video_fmts],
2316 aexts=[f['ext'] for f in audio_fmts],
2317 preferences=(try_call(lambda: self.params['merge_output_format'].split('/'))
2318 or self.params.get('prefer_free_formats') and ('webm', 'mkv')))
2319
2320 filtered = lambda *keys: filter(None, (traverse_obj(fmt, *keys) for fmt in formats_info))
2321
2322 new_dict = {
2323 'requested_formats': formats_info,
2324 'format': '+'.join(filtered('format')),
2325 'format_id': '+'.join(filtered('format_id')),
2326 'ext': output_ext,
2327 'protocol': '+'.join(map(determine_protocol, formats_info)),
2328 'language': '+'.join(orderedSet(filtered('language'))) or None,
2329 'format_note': '+'.join(orderedSet(filtered('format_note'))) or None,
2330 'filesize_approx': sum(filtered('filesize', 'filesize_approx')) or None,
2331 'tbr': sum(filtered('tbr', 'vbr', 'abr')),
2332 }
2333
2334 if the_only_video:
2335 new_dict.update({
2336 'width': the_only_video.get('width'),
2337 'height': the_only_video.get('height'),
2338 'resolution': the_only_video.get('resolution') or self.format_resolution(the_only_video),
2339 'fps': the_only_video.get('fps'),
2340 'dynamic_range': the_only_video.get('dynamic_range'),
2341 'vcodec': the_only_video.get('vcodec'),
2342 'vbr': the_only_video.get('vbr'),
2343 'stretched_ratio': the_only_video.get('stretched_ratio'),
2344 'aspect_ratio': the_only_video.get('aspect_ratio'),
2345 })
2346
2347 if the_only_audio:
2348 new_dict.update({
2349 'acodec': the_only_audio.get('acodec'),
2350 'abr': the_only_audio.get('abr'),
2351 'asr': the_only_audio.get('asr'),
2352 'audio_channels': the_only_audio.get('audio_channels')
2353 })
2354
2355 return new_dict
2356
2357 def _check_formats(formats):
2358 if self.params.get('check_formats') == 'selected':
2359 yield from self._check_formats(formats)
2360 return
2361 elif (self.params.get('check_formats') is not None
2362 or self.params.get('allow_unplayable_formats')):
2363 yield from formats
2364 return
2365
2366 for f in formats:
2367 if f.get('has_drm') or f.get('__needs_testing'):
2368 yield from self._check_formats([f])
2369 else:
2370 yield f
2371
2372 def _build_selector_function(selector):
2373 if isinstance(selector, list): # ,
2374 fs = [_build_selector_function(s) for s in selector]
2375
2376 def selector_function(ctx):
2377 for f in fs:
2378 yield from f(ctx)
2379 return selector_function
2380
2381 elif selector.type == GROUP: # ()
2382 selector_function = _build_selector_function(selector.selector)
2383
2384 elif selector.type == PICKFIRST: # /
2385 fs = [_build_selector_function(s) for s in selector.selector]
2386
2387 def selector_function(ctx):
2388 for f in fs:
2389 picked_formats = list(f(ctx))
2390 if picked_formats:
2391 return picked_formats
2392 return []
2393
2394 elif selector.type == MERGE: # +
2395 selector_1, selector_2 = map(_build_selector_function, selector.selector)
2396
2397 def selector_function(ctx):
2398 for pair in itertools.product(selector_1(ctx), selector_2(ctx)):
2399 yield _merge(pair)
2400
2401 elif selector.type == SINGLE: # atom
2402 format_spec = selector.selector or 'best'
2403
2404 # TODO: Add allvideo, allaudio etc by generalizing the code with best/worst selector
2405 if format_spec == 'all':
2406 def selector_function(ctx):
2407 yield from _check_formats(ctx['formats'][::-1])
2408 elif format_spec == 'mergeall':
2409 def selector_function(ctx):
2410 formats = list(_check_formats(
2411 f for f in ctx['formats'] if f.get('vcodec') != 'none' or f.get('acodec') != 'none'))
2412 if not formats:
2413 return
2414 merged_format = formats[-1]
2415 for f in formats[-2::-1]:
2416 merged_format = _merge((merged_format, f))
2417 yield merged_format
2418
2419 else:
2420 format_fallback, seperate_fallback, format_reverse, format_idx = False, None, True, 1
2421 mobj = re.match(
2422 r'(?P<bw>best|worst|b|w)(?P<type>video|audio|v|a)?(?P<mod>\*)?(?:\.(?P<n>[1-9]\d*))?$',
2423 format_spec)
2424 if mobj is not None:
2425 format_idx = int_or_none(mobj.group('n'), default=1)
2426 format_reverse = mobj.group('bw')[0] == 'b'
2427 format_type = (mobj.group('type') or [None])[0]
2428 not_format_type = {'v': 'a', 'a': 'v'}.get(format_type)
2429 format_modified = mobj.group('mod') is not None
2430
2431 format_fallback = not format_type and not format_modified # for b, w
2432 _filter_f = (
2433 (lambda f: f.get('%scodec' % format_type) != 'none')
2434 if format_type and format_modified # bv*, ba*, wv*, wa*
2435 else (lambda f: f.get('%scodec' % not_format_type) == 'none')
2436 if format_type # bv, ba, wv, wa
2437 else (lambda f: f.get('vcodec') != 'none' and f.get('acodec') != 'none')
2438 if not format_modified # b, w
2439 else lambda f: True) # b*, w*
2440 filter_f = lambda f: _filter_f(f) and (
2441 f.get('vcodec') != 'none' or f.get('acodec') != 'none')
2442 else:
2443 if format_spec in self._format_selection_exts['audio']:
2444 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none'
2445 elif format_spec in self._format_selection_exts['video']:
2446 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none' and f.get('vcodec') != 'none'
2447 seperate_fallback = lambda f: f.get('ext') == format_spec and f.get('vcodec') != 'none'
2448 elif format_spec in self._format_selection_exts['storyboards']:
2449 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') == 'none' and f.get('vcodec') == 'none'
2450 else:
2451 filter_f = lambda f: f.get('format_id') == format_spec # id
2452
2453 def selector_function(ctx):
2454 formats = list(ctx['formats'])
2455 matches = list(filter(filter_f, formats)) if filter_f is not None else formats
2456 if not matches:
2457 if format_fallback and ctx['incomplete_formats']:
2458 # for extractors with incomplete formats (audio only (soundcloud)
2459 # or video only (imgur)) best/worst will fallback to
2460 # best/worst {video,audio}-only format
2461 matches = list(filter(lambda f: f.get('vcodec') != 'none' or f.get('acodec') != 'none', formats))
2462 elif seperate_fallback and not ctx['has_merged_format']:
2463 # for compatibility with youtube-dl when there is no pre-merged format
2464 matches = list(filter(seperate_fallback, formats))
2465 matches = LazyList(_check_formats(matches[::-1 if format_reverse else 1]))
2466 try:
2467 yield matches[format_idx - 1]
2468 except LazyList.IndexError:
2469 return
2470
2471 filters = [self._build_format_filter(f) for f in selector.filters]
2472
2473 def final_selector(ctx):
2474 ctx_copy = dict(ctx)
2475 for _filter in filters:
2476 ctx_copy['formats'] = list(filter(_filter, ctx_copy['formats']))
2477 return selector_function(ctx_copy)
2478 return final_selector
2479
2480 # HACK: Python 3.12 changed the underlying parser, rendering '7_a' invalid
2481 # Prefix numbers with random letters to avoid it being classified as a number
2482 # See: https://github.com/yt-dlp/yt-dlp/pulls/8797
2483 # TODO: Implement parser not reliant on tokenize.tokenize
2484 prefix = ''.join(random.choices(string.ascii_letters, k=32))
2485 stream = io.BytesIO(re.sub(r'\d[_\d]*', rf'{prefix}\g<0>', format_spec).encode())
2486 try:
2487 tokens = list(_remove_unused_ops(
2488 token._replace(string=token.string.replace(prefix, ''))
2489 for token in tokenize.tokenize(stream.readline)))
2490 except tokenize.TokenError:
2491 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
2492
2493 class TokenIterator:
2494 def __init__(self, tokens):
2495 self.tokens = tokens
2496 self.counter = 0
2497
2498 def __iter__(self):
2499 return self
2500
2501 def __next__(self):
2502 if self.counter >= len(self.tokens):
2503 raise StopIteration()
2504 value = self.tokens[self.counter]
2505 self.counter += 1
2506 return value
2507
2508 next = __next__
2509
2510 def restore_last_token(self):
2511 self.counter -= 1
2512
2513 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
2514 return _build_selector_function(parsed_selector)
2515
2516 def _calc_headers(self, info_dict, load_cookies=False):
2517 res = HTTPHeaderDict(self.params['http_headers'], info_dict.get('http_headers'))
2518 clean_headers(res)
2519
2520 if load_cookies: # For --load-info-json
2521 self._load_cookies(res.get('Cookie'), autoscope=info_dict['url']) # compat
2522 self._load_cookies(info_dict.get('cookies'), autoscope=False)
2523 # The `Cookie` header is removed to prevent leaks and unscoped cookies.
2524 # See: https://github.com/yt-dlp/yt-dlp/security/advisories/GHSA-v8mc-9377-rwjj
2525 res.pop('Cookie', None)
2526 cookies = self.cookiejar.get_cookies_for_url(info_dict['url'])
2527 if cookies:
2528 encoder = LenientSimpleCookie()
2529 values = []
2530 for cookie in cookies:
2531 _, value = encoder.value_encode(cookie.value)
2532 values.append(f'{cookie.name}={value}')
2533 if cookie.domain:
2534 values.append(f'Domain={cookie.domain}')
2535 if cookie.path:
2536 values.append(f'Path={cookie.path}')
2537 if cookie.secure:
2538 values.append('Secure')
2539 if cookie.expires:
2540 values.append(f'Expires={cookie.expires}')
2541 if cookie.version:
2542 values.append(f'Version={cookie.version}')
2543 info_dict['cookies'] = '; '.join(values)
2544
2545 if 'X-Forwarded-For' not in res:
2546 x_forwarded_for_ip = info_dict.get('__x_forwarded_for_ip')
2547 if x_forwarded_for_ip:
2548 res['X-Forwarded-For'] = x_forwarded_for_ip
2549
2550 return res
2551
2552 def _calc_cookies(self, url):
2553 self.deprecation_warning('"YoutubeDL._calc_cookies" is deprecated and may be removed in a future version')
2554 return self.cookiejar.get_cookie_header(url)
2555
2556 def _sort_thumbnails(self, thumbnails):
2557 thumbnails.sort(key=lambda t: (
2558 t.get('preference') if t.get('preference') is not None else -1,
2559 t.get('width') if t.get('width') is not None else -1,
2560 t.get('height') if t.get('height') is not None else -1,
2561 t.get('id') if t.get('id') is not None else '',
2562 t.get('url')))
2563
2564 def _sanitize_thumbnails(self, info_dict):
2565 thumbnails = info_dict.get('thumbnails')
2566 if thumbnails is None:
2567 thumbnail = info_dict.get('thumbnail')
2568 if thumbnail:
2569 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
2570 if not thumbnails:
2571 return
2572
2573 def check_thumbnails(thumbnails):
2574 for t in thumbnails:
2575 self.to_screen(f'[info] Testing thumbnail {t["id"]}')
2576 try:
2577 self.urlopen(HEADRequest(t['url']))
2578 except network_exceptions as err:
2579 self.to_screen(f'[info] Unable to connect to thumbnail {t["id"]} URL {t["url"]!r} - {err}. Skipping...')
2580 continue
2581 yield t
2582
2583 self._sort_thumbnails(thumbnails)
2584 for i, t in enumerate(thumbnails):
2585 if t.get('id') is None:
2586 t['id'] = '%d' % i
2587 if t.get('width') and t.get('height'):
2588 t['resolution'] = '%dx%d' % (t['width'], t['height'])
2589 t['url'] = sanitize_url(t['url'])
2590
2591 if self.params.get('check_formats') is True:
2592 info_dict['thumbnails'] = LazyList(check_thumbnails(thumbnails[::-1]), reverse=True)
2593 else:
2594 info_dict['thumbnails'] = thumbnails
2595
2596 def _fill_common_fields(self, info_dict, final=True):
2597 # TODO: move sanitization here
2598 if final:
2599 title = info_dict['fulltitle'] = info_dict.get('title')
2600 if not title:
2601 if title == '':
2602 self.write_debug('Extractor gave empty title. Creating a generic title')
2603 else:
2604 self.report_warning('Extractor failed to obtain "title". Creating a generic title instead')
2605 info_dict['title'] = f'{info_dict["extractor"].replace(":", "-")} video #{info_dict["id"]}'
2606
2607 if info_dict.get('duration') is not None:
2608 info_dict['duration_string'] = formatSeconds(info_dict['duration'])
2609
2610 for ts_key, date_key in (
2611 ('timestamp', 'upload_date'),
2612 ('release_timestamp', 'release_date'),
2613 ('modified_timestamp', 'modified_date'),
2614 ):
2615 if info_dict.get(date_key) is None and info_dict.get(ts_key) is not None:
2616 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
2617 # see http://bugs.python.org/issue1646728)
2618 with contextlib.suppress(ValueError, OverflowError, OSError):
2619 upload_date = datetime.datetime.fromtimestamp(info_dict[ts_key], datetime.timezone.utc)
2620 info_dict[date_key] = upload_date.strftime('%Y%m%d')
2621
2622 if not info_dict.get('release_year'):
2623 info_dict['release_year'] = traverse_obj(info_dict, ('release_date', {lambda x: int(x[:4])}))
2624
2625 live_keys = ('is_live', 'was_live')
2626 live_status = info_dict.get('live_status')
2627 if live_status is None:
2628 for key in live_keys:
2629 if info_dict.get(key) is False:
2630 continue
2631 if info_dict.get(key):
2632 live_status = key
2633 break
2634 if all(info_dict.get(key) is False for key in live_keys):
2635 live_status = 'not_live'
2636 if live_status:
2637 info_dict['live_status'] = live_status
2638 for key in live_keys:
2639 if info_dict.get(key) is None:
2640 info_dict[key] = (live_status == key)
2641 if live_status == 'post_live':
2642 info_dict['was_live'] = True
2643
2644 # Auto generate title fields corresponding to the *_number fields when missing
2645 # in order to always have clean titles. This is very common for TV series.
2646 for field in ('chapter', 'season', 'episode'):
2647 if final and info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
2648 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
2649
2650 for old_key, new_key in self._deprecated_multivalue_fields.items():
2651 if new_key in info_dict and old_key in info_dict:
2652 if '_version' not in info_dict: # HACK: Do not warn when using --load-info-json
2653 self.deprecation_warning(f'Do not return {old_key!r} when {new_key!r} is present')
2654 elif old_value := info_dict.get(old_key):
2655 info_dict[new_key] = old_value.split(', ')
2656 elif new_value := info_dict.get(new_key):
2657 info_dict[old_key] = ', '.join(v.replace(',', '\N{FULLWIDTH COMMA}') for v in new_value)
2658
2659 def _raise_pending_errors(self, info):
2660 err = info.pop('__pending_error', None)
2661 if err:
2662 self.report_error(err, tb=False)
2663
2664 def sort_formats(self, info_dict):
2665 formats = self._get_formats(info_dict)
2666 formats.sort(key=FormatSorter(
2667 self, info_dict.get('_format_sort_fields') or []).calculate_preference)
2668
2669 def process_video_result(self, info_dict, download=True):
2670 assert info_dict.get('_type', 'video') == 'video'
2671 self._num_videos += 1
2672
2673 if 'id' not in info_dict:
2674 raise ExtractorError('Missing "id" field in extractor result', ie=info_dict['extractor'])
2675 elif not info_dict.get('id'):
2676 raise ExtractorError('Extractor failed to obtain "id"', ie=info_dict['extractor'])
2677
2678 def report_force_conversion(field, field_not, conversion):
2679 self.report_warning(
2680 '"%s" field is not %s - forcing %s conversion, there is an error in extractor'
2681 % (field, field_not, conversion))
2682
2683 def sanitize_string_field(info, string_field):
2684 field = info.get(string_field)
2685 if field is None or isinstance(field, str):
2686 return
2687 report_force_conversion(string_field, 'a string', 'string')
2688 info[string_field] = str(field)
2689
2690 def sanitize_numeric_fields(info):
2691 for numeric_field in self._NUMERIC_FIELDS:
2692 field = info.get(numeric_field)
2693 if field is None or isinstance(field, (int, float)):
2694 continue
2695 report_force_conversion(numeric_field, 'numeric', 'int')
2696 info[numeric_field] = int_or_none(field)
2697
2698 sanitize_string_field(info_dict, 'id')
2699 sanitize_numeric_fields(info_dict)
2700 if info_dict.get('section_end') and info_dict.get('section_start') is not None:
2701 info_dict['duration'] = round(info_dict['section_end'] - info_dict['section_start'], 3)
2702 if (info_dict.get('duration') or 0) <= 0 and info_dict.pop('duration', None):
2703 self.report_warning('"duration" field is negative, there is an error in extractor')
2704
2705 chapters = info_dict.get('chapters') or []
2706 if chapters and chapters[0].get('start_time'):
2707 chapters.insert(0, {'start_time': 0})
2708
2709 dummy_chapter = {'end_time': 0, 'start_time': info_dict.get('duration')}
2710 for idx, (prev, current, next_) in enumerate(zip(
2711 (dummy_chapter, *chapters), chapters, (*chapters[1:], dummy_chapter)), 1):
2712 if current.get('start_time') is None:
2713 current['start_time'] = prev.get('end_time')
2714 if not current.get('end_time'):
2715 current['end_time'] = next_.get('start_time')
2716 if not current.get('title'):
2717 current['title'] = f'<Untitled Chapter {idx}>'
2718
2719 if 'playlist' not in info_dict:
2720 # It isn't part of a playlist
2721 info_dict['playlist'] = None
2722 info_dict['playlist_index'] = None
2723
2724 self._sanitize_thumbnails(info_dict)
2725
2726 thumbnail = info_dict.get('thumbnail')
2727 thumbnails = info_dict.get('thumbnails')
2728 if thumbnail:
2729 info_dict['thumbnail'] = sanitize_url(thumbnail)
2730 elif thumbnails:
2731 info_dict['thumbnail'] = thumbnails[-1]['url']
2732
2733 if info_dict.get('display_id') is None and 'id' in info_dict:
2734 info_dict['display_id'] = info_dict['id']
2735
2736 self._fill_common_fields(info_dict)
2737
2738 for cc_kind in ('subtitles', 'automatic_captions'):
2739 cc = info_dict.get(cc_kind)
2740 if cc:
2741 for _, subtitle in cc.items():
2742 for subtitle_format in subtitle:
2743 if subtitle_format.get('url'):
2744 subtitle_format['url'] = sanitize_url(subtitle_format['url'])
2745 if subtitle_format.get('ext') is None:
2746 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
2747
2748 automatic_captions = info_dict.get('automatic_captions')
2749 subtitles = info_dict.get('subtitles')
2750
2751 info_dict['requested_subtitles'] = self.process_subtitles(
2752 info_dict['id'], subtitles, automatic_captions)
2753
2754 formats = self._get_formats(info_dict)
2755
2756 # Backward compatibility with InfoExtractor._sort_formats
2757 field_preference = (formats or [{}])[0].pop('__sort_fields', None)
2758 if field_preference:
2759 info_dict['_format_sort_fields'] = field_preference
2760
2761 info_dict['_has_drm'] = any( # or None ensures --clean-infojson removes it
2762 f.get('has_drm') and f['has_drm'] != 'maybe' for f in formats) or None
2763 if not self.params.get('allow_unplayable_formats'):
2764 formats = [f for f in formats if not f.get('has_drm') or f['has_drm'] == 'maybe']
2765
2766 if formats and all(f.get('acodec') == f.get('vcodec') == 'none' for f in formats):
2767 self.report_warning(
2768 f'{"This video is DRM protected and " if info_dict["_has_drm"] else ""}'
2769 'only images are available for download. Use --list-formats to see them'.capitalize())
2770
2771 get_from_start = not info_dict.get('is_live') or bool(self.params.get('live_from_start'))
2772 if not get_from_start:
2773 info_dict['title'] += ' ' + datetime.datetime.now().strftime('%Y-%m-%d %H:%M')
2774 if info_dict.get('is_live') and formats:
2775 formats = [f for f in formats if bool(f.get('is_from_start')) == get_from_start]
2776 if get_from_start and not formats:
2777 self.raise_no_formats(info_dict, msg=(
2778 '--live-from-start is passed, but there are no formats that can be downloaded from the start. '
2779 'If you want to download from the current time, use --no-live-from-start'))
2780
2781 def is_wellformed(f):
2782 url = f.get('url')
2783 if not url:
2784 self.report_warning(
2785 '"url" field is missing or empty - skipping format, '
2786 'there is an error in extractor')
2787 return False
2788 if isinstance(url, bytes):
2789 sanitize_string_field(f, 'url')
2790 return True
2791
2792 # Filter out malformed formats for better extraction robustness
2793 formats = list(filter(is_wellformed, formats or []))
2794
2795 if not formats:
2796 self.raise_no_formats(info_dict)
2797
2798 for format in formats:
2799 sanitize_string_field(format, 'format_id')
2800 sanitize_numeric_fields(format)
2801 format['url'] = sanitize_url(format['url'])
2802 if format.get('ext') is None:
2803 format['ext'] = determine_ext(format['url']).lower()
2804 if format.get('protocol') is None:
2805 format['protocol'] = determine_protocol(format)
2806 if format.get('resolution') is None:
2807 format['resolution'] = self.format_resolution(format, default=None)
2808 if format.get('dynamic_range') is None and format.get('vcodec') != 'none':
2809 format['dynamic_range'] = 'SDR'
2810 if format.get('aspect_ratio') is None:
2811 format['aspect_ratio'] = try_call(lambda: round(format['width'] / format['height'], 2))
2812 # For fragmented formats, "tbr" is often max bitrate and not average
2813 if (('manifest-filesize-approx' in self.params['compat_opts'] or not format.get('manifest_url'))
2814 and info_dict.get('duration') and format.get('tbr')
2815 and not format.get('filesize') and not format.get('filesize_approx')):
2816 format['filesize_approx'] = int(info_dict['duration'] * format['tbr'] * (1024 / 8))
2817 format['http_headers'] = self._calc_headers(collections.ChainMap(format, info_dict), load_cookies=True)
2818
2819 # Safeguard against old/insecure infojson when using --load-info-json
2820 if info_dict.get('http_headers'):
2821 info_dict['http_headers'] = HTTPHeaderDict(info_dict['http_headers'])
2822 info_dict['http_headers'].pop('Cookie', None)
2823
2824 # This is copied to http_headers by the above _calc_headers and can now be removed
2825 if '__x_forwarded_for_ip' in info_dict:
2826 del info_dict['__x_forwarded_for_ip']
2827
2828 self.sort_formats({
2829 'formats': formats,
2830 '_format_sort_fields': info_dict.get('_format_sort_fields')
2831 })
2832
2833 # Sanitize and group by format_id
2834 formats_dict = {}
2835 for i, format in enumerate(formats):
2836 if not format.get('format_id'):
2837 format['format_id'] = str(i)
2838 else:
2839 # Sanitize format_id from characters used in format selector expression
2840 format['format_id'] = re.sub(r'[\s,/+\[\]()]', '_', format['format_id'])
2841 formats_dict.setdefault(format['format_id'], []).append(format)
2842
2843 # Make sure all formats have unique format_id
2844 common_exts = set(itertools.chain(*self._format_selection_exts.values()))
2845 for format_id, ambiguous_formats in formats_dict.items():
2846 ambigious_id = len(ambiguous_formats) > 1
2847 for i, format in enumerate(ambiguous_formats):
2848 if ambigious_id:
2849 format['format_id'] = '%s-%d' % (format_id, i)
2850 # Ensure there is no conflict between id and ext in format selection
2851 # See https://github.com/yt-dlp/yt-dlp/issues/1282
2852 if format['format_id'] != format['ext'] and format['format_id'] in common_exts:
2853 format['format_id'] = 'f%s' % format['format_id']
2854
2855 if format.get('format') is None:
2856 format['format'] = '{id} - {res}{note}'.format(
2857 id=format['format_id'],
2858 res=self.format_resolution(format),
2859 note=format_field(format, 'format_note', ' (%s)'),
2860 )
2861
2862 if self.params.get('check_formats') is True:
2863 formats = LazyList(self._check_formats(formats[::-1]), reverse=True)
2864
2865 if not formats or formats[0] is not info_dict:
2866 # only set the 'formats' fields if the original info_dict list them
2867 # otherwise we end up with a circular reference, the first (and unique)
2868 # element in the 'formats' field in info_dict is info_dict itself,
2869 # which can't be exported to json
2870 info_dict['formats'] = formats
2871
2872 info_dict, _ = self.pre_process(info_dict)
2873
2874 if self._match_entry(info_dict, incomplete=self._format_fields) is not None:
2875 return info_dict
2876
2877 self.post_extract(info_dict)
2878 info_dict, _ = self.pre_process(info_dict, 'after_filter')
2879
2880 # The pre-processors may have modified the formats
2881 formats = self._get_formats(info_dict)
2882
2883 list_only = self.params.get('simulate') == 'list_only'
2884 interactive_format_selection = not list_only and self.format_selector == '-'
2885 if self.params.get('list_thumbnails'):
2886 self.list_thumbnails(info_dict)
2887 if self.params.get('listsubtitles'):
2888 if 'automatic_captions' in info_dict:
2889 self.list_subtitles(
2890 info_dict['id'], automatic_captions, 'automatic captions')
2891 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
2892 if self.params.get('listformats') or interactive_format_selection:
2893 self.list_formats(info_dict)
2894 if list_only:
2895 # Without this printing, -F --print-json will not work
2896 self.__forced_printings(info_dict)
2897 return info_dict
2898
2899 format_selector = self.format_selector
2900 while True:
2901 if interactive_format_selection:
2902 req_format = input(self._format_screen('\nEnter format selector ', self.Styles.EMPHASIS)
2903 + '(Press ENTER for default, or Ctrl+C to quit)'
2904 + self._format_screen(': ', self.Styles.EMPHASIS))
2905 try:
2906 format_selector = self.build_format_selector(req_format) if req_format else None
2907 except SyntaxError as err:
2908 self.report_error(err, tb=False, is_error=False)
2909 continue
2910
2911 if format_selector is None:
2912 req_format = self._default_format_spec(info_dict, download=download)
2913 self.write_debug(f'Default format spec: {req_format}')
2914 format_selector = self.build_format_selector(req_format)
2915
2916 formats_to_download = list(format_selector({
2917 'formats': formats,
2918 'has_merged_format': any('none' not in (f.get('acodec'), f.get('vcodec')) for f in formats),
2919 'incomplete_formats': (all(f.get('vcodec') == 'none' for f in formats) # No formats with video
2920 or all(f.get('acodec') == 'none' for f in formats)), # OR, No formats with audio
2921 }))
2922 if interactive_format_selection and not formats_to_download:
2923 self.report_error('Requested format is not available', tb=False, is_error=False)
2924 continue
2925 break
2926
2927 if not formats_to_download:
2928 if not self.params.get('ignore_no_formats_error'):
2929 raise ExtractorError(
2930 'Requested format is not available. Use --list-formats for a list of available formats',
2931 expected=True, video_id=info_dict['id'], ie=info_dict['extractor'])
2932 self.report_warning('Requested format is not available')
2933 # Process what we can, even without any available formats.
2934 formats_to_download = [{}]
2935
2936 requested_ranges = tuple(self.params.get('download_ranges', lambda *_: [{}])(info_dict, self))
2937 best_format, downloaded_formats = formats_to_download[-1], []
2938 if download:
2939 if best_format and requested_ranges:
2940 def to_screen(*msg):
2941 self.to_screen(f'[info] {info_dict["id"]}: {" ".join(", ".join(variadic(m)) for m in msg)}')
2942
2943 to_screen(f'Downloading {len(formats_to_download)} format(s):',
2944 (f['format_id'] for f in formats_to_download))
2945 if requested_ranges != ({}, ):
2946 to_screen(f'Downloading {len(requested_ranges)} time ranges:',
2947 (f'{c["start_time"]:.1f}-{c["end_time"]:.1f}' for c in requested_ranges))
2948 max_downloads_reached = False
2949
2950 for fmt, chapter in itertools.product(formats_to_download, requested_ranges):
2951 new_info = self._copy_infodict(info_dict)
2952 new_info.update(fmt)
2953 offset, duration = info_dict.get('section_start') or 0, info_dict.get('duration') or float('inf')
2954 end_time = offset + min(chapter.get('end_time', duration), duration)
2955 # duration may not be accurate. So allow deviations <1sec
2956 if end_time == float('inf') or end_time > offset + duration + 1:
2957 end_time = None
2958 if chapter or offset:
2959 new_info.update({
2960 'section_start': offset + chapter.get('start_time', 0),
2961 'section_end': end_time,
2962 'section_title': chapter.get('title'),
2963 'section_number': chapter.get('index'),
2964 })
2965 downloaded_formats.append(new_info)
2966 try:
2967 self.process_info(new_info)
2968 except MaxDownloadsReached:
2969 max_downloads_reached = True
2970 self._raise_pending_errors(new_info)
2971 # Remove copied info
2972 for key, val in tuple(new_info.items()):
2973 if info_dict.get(key) == val:
2974 new_info.pop(key)
2975 if max_downloads_reached:
2976 break
2977
2978 write_archive = {f.get('__write_download_archive', False) for f in downloaded_formats}
2979 assert write_archive.issubset({True, False, 'ignore'})
2980 if True in write_archive and False not in write_archive:
2981 self.record_download_archive(info_dict)
2982
2983 info_dict['requested_downloads'] = downloaded_formats
2984 info_dict = self.run_all_pps('after_video', info_dict)
2985 if max_downloads_reached:
2986 raise MaxDownloadsReached()
2987
2988 # We update the info dict with the selected best quality format (backwards compatibility)
2989 info_dict.update(best_format)
2990 return info_dict
2991
2992 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
2993 """Select the requested subtitles and their format"""
2994 available_subs, normal_sub_langs = {}, []
2995 if normal_subtitles and self.params.get('writesubtitles'):
2996 available_subs.update(normal_subtitles)
2997 normal_sub_langs = tuple(normal_subtitles.keys())
2998 if automatic_captions and self.params.get('writeautomaticsub'):
2999 for lang, cap_info in automatic_captions.items():
3000 if lang not in available_subs:
3001 available_subs[lang] = cap_info
3002
3003 if not available_subs or (
3004 not self.params.get('writesubtitles')
3005 and not self.params.get('writeautomaticsub')):
3006 return None
3007
3008 all_sub_langs = tuple(available_subs.keys())
3009 if self.params.get('allsubtitles', False):
3010 requested_langs = all_sub_langs
3011 elif self.params.get('subtitleslangs', False):
3012 try:
3013 requested_langs = orderedSet_from_options(
3014 self.params.get('subtitleslangs'), {'all': all_sub_langs}, use_regex=True)
3015 except re.error as e:
3016 raise ValueError(f'Wrong regex for subtitlelangs: {e.pattern}')
3017 else:
3018 requested_langs = LazyList(itertools.chain(
3019 ['en'] if 'en' in normal_sub_langs else [],
3020 filter(lambda f: f.startswith('en'), normal_sub_langs),
3021 ['en'] if 'en' in all_sub_langs else [],
3022 filter(lambda f: f.startswith('en'), all_sub_langs),
3023 normal_sub_langs, all_sub_langs,
3024 ))[:1]
3025 if requested_langs:
3026 self.to_screen(f'[info] {video_id}: Downloading subtitles: {", ".join(requested_langs)}')
3027
3028 formats_query = self.params.get('subtitlesformat', 'best')
3029 formats_preference = formats_query.split('/') if formats_query else []
3030 subs = {}
3031 for lang in requested_langs:
3032 formats = available_subs.get(lang)
3033 if formats is None:
3034 self.report_warning(f'{lang} subtitles not available for {video_id}')
3035 continue
3036 for ext in formats_preference:
3037 if ext == 'best':
3038 f = formats[-1]
3039 break
3040 matches = list(filter(lambda f: f['ext'] == ext, formats))
3041 if matches:
3042 f = matches[-1]
3043 break
3044 else:
3045 f = formats[-1]
3046 self.report_warning(
3047 'No subtitle format found matching "%s" for language %s, '
3048 'using %s' % (formats_query, lang, f['ext']))
3049 subs[lang] = f
3050 return subs
3051
3052 def _forceprint(self, key, info_dict):
3053 if info_dict is None:
3054 return
3055 info_copy = info_dict.copy()
3056 info_copy.setdefault('filename', self.prepare_filename(info_dict))
3057 if info_dict.get('requested_formats') is not None:
3058 # For RTMP URLs, also include the playpath
3059 info_copy['urls'] = '\n'.join(f['url'] + f.get('play_path', '') for f in info_dict['requested_formats'])
3060 elif info_dict.get('url'):
3061 info_copy['urls'] = info_dict['url'] + info_dict.get('play_path', '')
3062 info_copy['formats_table'] = self.render_formats_table(info_dict)
3063 info_copy['thumbnails_table'] = self.render_thumbnails_table(info_dict)
3064 info_copy['subtitles_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('subtitles'))
3065 info_copy['automatic_captions_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('automatic_captions'))
3066
3067 def format_tmpl(tmpl):
3068 mobj = re.fullmatch(r'([\w.:,]|-\d|(?P<dict>{([\w.:,]|-\d)+}))+=?', tmpl)
3069 if not mobj:
3070 return tmpl
3071
3072 fmt = '%({})s'
3073 if tmpl.startswith('{'):
3074 tmpl, fmt = f'.{tmpl}', '%({})j'
3075 if tmpl.endswith('='):
3076 tmpl, fmt = tmpl[:-1], '{0} = %({0})#j'
3077 return '\n'.join(map(fmt.format, [tmpl] if mobj.group('dict') else tmpl.split(',')))
3078
3079 for tmpl in self.params['forceprint'].get(key, []):
3080 self.to_stdout(self.evaluate_outtmpl(format_tmpl(tmpl), info_copy))
3081
3082 for tmpl, file_tmpl in self.params['print_to_file'].get(key, []):
3083 filename = self.prepare_filename(info_dict, outtmpl=file_tmpl)
3084 tmpl = format_tmpl(tmpl)
3085 self.to_screen(f'[info] Writing {tmpl!r} to: {filename}')
3086 if self._ensure_dir_exists(filename):
3087 with open(filename, 'a', encoding='utf-8', newline='') as f:
3088 f.write(self.evaluate_outtmpl(tmpl, info_copy) + os.linesep)
3089
3090 return info_copy
3091
3092 def __forced_printings(self, info_dict, filename=None, incomplete=True):
3093 if (self.params.get('forcejson')
3094 or self.params['forceprint'].get('video')
3095 or self.params['print_to_file'].get('video')):
3096 self.post_extract(info_dict)
3097 if filename:
3098 info_dict['filename'] = filename
3099 info_copy = self._forceprint('video', info_dict)
3100
3101 def print_field(field, actual_field=None, optional=False):
3102 if actual_field is None:
3103 actual_field = field
3104 if self.params.get(f'force{field}') and (
3105 info_copy.get(field) is not None or (not optional and not incomplete)):
3106 self.to_stdout(info_copy[actual_field])
3107
3108 print_field('title')
3109 print_field('id')
3110 print_field('url', 'urls')
3111 print_field('thumbnail', optional=True)
3112 print_field('description', optional=True)
3113 print_field('filename')
3114 if self.params.get('forceduration') and info_copy.get('duration') is not None:
3115 self.to_stdout(formatSeconds(info_copy['duration']))
3116 print_field('format')
3117
3118 if self.params.get('forcejson'):
3119 self.to_stdout(json.dumps(self.sanitize_info(info_dict)))
3120
3121 def dl(self, name, info, subtitle=False, test=False):
3122 if not info.get('url'):
3123 self.raise_no_formats(info, True)
3124
3125 if test:
3126 verbose = self.params.get('verbose')
3127 params = {
3128 'test': True,
3129 'quiet': self.params.get('quiet') or not verbose,
3130 'verbose': verbose,
3131 'noprogress': not verbose,
3132 'nopart': True,
3133 'skip_unavailable_fragments': False,
3134 'keep_fragments': False,
3135 'overwrites': True,
3136 '_no_ytdl_file': True,
3137 }
3138 else:
3139 params = self.params
3140 fd = get_suitable_downloader(info, params, to_stdout=(name == '-'))(self, params)
3141 if not test:
3142 for ph in self._progress_hooks:
3143 fd.add_progress_hook(ph)
3144 urls = '", "'.join(
3145 (f['url'].split(',')[0] + ',<data>' if f['url'].startswith('data:') else f['url'])
3146 for f in info.get('requested_formats', []) or [info])
3147 self.write_debug(f'Invoking {fd.FD_NAME} downloader on "{urls}"')
3148
3149 # Note: Ideally info should be a deep-copied so that hooks cannot modify it.
3150 # But it may contain objects that are not deep-copyable
3151 new_info = self._copy_infodict(info)
3152 if new_info.get('http_headers') is None:
3153 new_info['http_headers'] = self._calc_headers(new_info)
3154 return fd.download(name, new_info, subtitle)
3155
3156 def existing_file(self, filepaths, *, default_overwrite=True):
3157 existing_files = list(filter(os.path.exists, orderedSet(filepaths)))
3158 if existing_files and not self.params.get('overwrites', default_overwrite):
3159 return existing_files[0]
3160
3161 for file in existing_files:
3162 self.report_file_delete(file)
3163 os.remove(file)
3164 return None
3165
3166 def process_info(self, info_dict):
3167 """Process a single resolved IE result. (Modifies it in-place)"""
3168
3169 assert info_dict.get('_type', 'video') == 'video'
3170 original_infodict = info_dict
3171
3172 if 'format' not in info_dict and 'ext' in info_dict:
3173 info_dict['format'] = info_dict['ext']
3174
3175 if self._match_entry(info_dict) is not None:
3176 info_dict['__write_download_archive'] = 'ignore'
3177 return
3178
3179 # Does nothing under normal operation - for backward compatibility of process_info
3180 self.post_extract(info_dict)
3181
3182 def replace_info_dict(new_info):
3183 nonlocal info_dict
3184 if new_info == info_dict:
3185 return
3186 info_dict.clear()
3187 info_dict.update(new_info)
3188
3189 new_info, _ = self.pre_process(info_dict, 'video')
3190 replace_info_dict(new_info)
3191 self._num_downloads += 1
3192
3193 # info_dict['_filename'] needs to be set for backward compatibility
3194 info_dict['_filename'] = full_filename = self.prepare_filename(info_dict, warn=True)
3195 temp_filename = self.prepare_filename(info_dict, 'temp')
3196 files_to_move = {}
3197
3198 # Forced printings
3199 self.__forced_printings(info_dict, full_filename, incomplete=('format' not in info_dict))
3200
3201 def check_max_downloads():
3202 if self._num_downloads >= float(self.params.get('max_downloads') or 'inf'):
3203 raise MaxDownloadsReached()
3204
3205 if self.params.get('simulate'):
3206 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
3207 check_max_downloads()
3208 return
3209
3210 if full_filename is None:
3211 return
3212 if not self._ensure_dir_exists(encodeFilename(full_filename)):
3213 return
3214 if not self._ensure_dir_exists(encodeFilename(temp_filename)):
3215 return
3216
3217 if self._write_description('video', info_dict,
3218 self.prepare_filename(info_dict, 'description')) is None:
3219 return
3220
3221 sub_files = self._write_subtitles(info_dict, temp_filename)
3222 if sub_files is None:
3223 return
3224 files_to_move.update(dict(sub_files))
3225
3226 thumb_files = self._write_thumbnails(
3227 'video', info_dict, temp_filename, self.prepare_filename(info_dict, 'thumbnail'))
3228 if thumb_files is None:
3229 return
3230 files_to_move.update(dict(thumb_files))
3231
3232 infofn = self.prepare_filename(info_dict, 'infojson')
3233 _infojson_written = self._write_info_json('video', info_dict, infofn)
3234 if _infojson_written:
3235 info_dict['infojson_filename'] = infofn
3236 # For backward compatibility, even though it was a private field
3237 info_dict['__infojson_filename'] = infofn
3238 elif _infojson_written is None:
3239 return
3240
3241 # Note: Annotations are deprecated
3242 annofn = None
3243 if self.params.get('writeannotations', False):
3244 annofn = self.prepare_filename(info_dict, 'annotation')
3245 if annofn:
3246 if not self._ensure_dir_exists(encodeFilename(annofn)):
3247 return
3248 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(annofn)):
3249 self.to_screen('[info] Video annotations are already present')
3250 elif not info_dict.get('annotations'):
3251 self.report_warning('There are no annotations to write.')
3252 else:
3253 try:
3254 self.to_screen('[info] Writing video annotations to: ' + annofn)
3255 with open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
3256 annofile.write(info_dict['annotations'])
3257 except (KeyError, TypeError):
3258 self.report_warning('There are no annotations to write.')
3259 except OSError:
3260 self.report_error('Cannot write annotations file: ' + annofn)
3261 return
3262
3263 # Write internet shortcut files
3264 def _write_link_file(link_type):
3265 url = try_get(info_dict['webpage_url'], iri_to_uri)
3266 if not url:
3267 self.report_warning(
3268 f'Cannot write internet shortcut file because the actual URL of "{info_dict["webpage_url"]}" is unknown')
3269 return True
3270 linkfn = replace_extension(self.prepare_filename(info_dict, 'link'), link_type, info_dict.get('ext'))
3271 if not self._ensure_dir_exists(encodeFilename(linkfn)):
3272 return False
3273 if self.params.get('overwrites', True) and os.path.exists(encodeFilename(linkfn)):
3274 self.to_screen(f'[info] Internet shortcut (.{link_type}) is already present')
3275 return True
3276 try:
3277 self.to_screen(f'[info] Writing internet shortcut (.{link_type}) to: {linkfn}')
3278 with open(encodeFilename(to_high_limit_path(linkfn)), 'w', encoding='utf-8',
3279 newline='\r\n' if link_type == 'url' else '\n') as linkfile:
3280 template_vars = {'url': url}
3281 if link_type == 'desktop':
3282 template_vars['filename'] = linkfn[:-(len(link_type) + 1)]
3283 linkfile.write(LINK_TEMPLATES[link_type] % template_vars)
3284 except OSError:
3285 self.report_error(f'Cannot write internet shortcut {linkfn}')
3286 return False
3287 return True
3288
3289 write_links = {
3290 'url': self.params.get('writeurllink'),
3291 'webloc': self.params.get('writewebloclink'),
3292 'desktop': self.params.get('writedesktoplink'),
3293 }
3294 if self.params.get('writelink'):
3295 link_type = ('webloc' if sys.platform == 'darwin'
3296 else 'desktop' if sys.platform.startswith('linux')
3297 else 'url')
3298 write_links[link_type] = True
3299
3300 if any(should_write and not _write_link_file(link_type)
3301 for link_type, should_write in write_links.items()):
3302 return
3303
3304 new_info, files_to_move = self.pre_process(info_dict, 'before_dl', files_to_move)
3305 replace_info_dict(new_info)
3306
3307 if self.params.get('skip_download'):
3308 info_dict['filepath'] = temp_filename
3309 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
3310 info_dict['__files_to_move'] = files_to_move
3311 replace_info_dict(self.run_pp(MoveFilesAfterDownloadPP(self, False), info_dict))
3312 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
3313 else:
3314 # Download
3315 info_dict.setdefault('__postprocessors', [])
3316 try:
3317
3318 def existing_video_file(*filepaths):
3319 ext = info_dict.get('ext')
3320 converted = lambda file: replace_extension(file, self.params.get('final_ext') or ext, ext)
3321 file = self.existing_file(itertools.chain(*zip(map(converted, filepaths), filepaths)),
3322 default_overwrite=False)
3323 if file:
3324 info_dict['ext'] = os.path.splitext(file)[1][1:]
3325 return file
3326
3327 fd, success = None, True
3328 if info_dict.get('protocol') or info_dict.get('url'):
3329 fd = get_suitable_downloader(info_dict, self.params, to_stdout=temp_filename == '-')
3330 if fd != FFmpegFD and 'no-direct-merge' not in self.params['compat_opts'] and (
3331 info_dict.get('section_start') or info_dict.get('section_end')):
3332 msg = ('This format cannot be partially downloaded' if FFmpegFD.available()
3333 else 'You have requested downloading the video partially, but ffmpeg is not installed')
3334 self.report_error(f'{msg}. Aborting')
3335 return
3336
3337 if info_dict.get('requested_formats') is not None:
3338 old_ext = info_dict['ext']
3339 if self.params.get('merge_output_format') is None:
3340 if (info_dict['ext'] == 'webm'
3341 and info_dict.get('thumbnails')
3342 # check with type instead of pp_key, __name__, or isinstance
3343 # since we dont want any custom PPs to trigger this
3344 and any(type(pp) == EmbedThumbnailPP for pp in self._pps['post_process'])): # noqa: E721
3345 info_dict['ext'] = 'mkv'
3346 self.report_warning(
3347 'webm doesn\'t support embedding a thumbnail, mkv will be used')
3348 new_ext = info_dict['ext']
3349
3350 def correct_ext(filename, ext=new_ext):
3351 if filename == '-':
3352 return filename
3353 filename_real_ext = os.path.splitext(filename)[1][1:]
3354 filename_wo_ext = (
3355 os.path.splitext(filename)[0]
3356 if filename_real_ext in (old_ext, new_ext)
3357 else filename)
3358 return f'{filename_wo_ext}.{ext}'
3359
3360 # Ensure filename always has a correct extension for successful merge
3361 full_filename = correct_ext(full_filename)
3362 temp_filename = correct_ext(temp_filename)
3363 dl_filename = existing_video_file(full_filename, temp_filename)
3364
3365 info_dict['__real_download'] = False
3366 # NOTE: Copy so that original format dicts are not modified
3367 info_dict['requested_formats'] = list(map(dict, info_dict['requested_formats']))
3368
3369 merger = FFmpegMergerPP(self)
3370 downloaded = []
3371 if dl_filename is not None:
3372 self.report_file_already_downloaded(dl_filename)
3373 elif fd:
3374 for f in info_dict['requested_formats'] if fd != FFmpegFD else []:
3375 f['filepath'] = fname = prepend_extension(
3376 correct_ext(temp_filename, info_dict['ext']),
3377 'f%s' % f['format_id'], info_dict['ext'])
3378 downloaded.append(fname)
3379 info_dict['url'] = '\n'.join(f['url'] for f in info_dict['requested_formats'])
3380 success, real_download = self.dl(temp_filename, info_dict)
3381 info_dict['__real_download'] = real_download
3382 else:
3383 if self.params.get('allow_unplayable_formats'):
3384 self.report_warning(
3385 'You have requested merging of multiple formats '
3386 'while also allowing unplayable formats to be downloaded. '
3387 'The formats won\'t be merged to prevent data corruption.')
3388 elif not merger.available:
3389 msg = 'You have requested merging of multiple formats but ffmpeg is not installed'
3390 if not self.params.get('ignoreerrors'):
3391 self.report_error(f'{msg}. Aborting due to --abort-on-error')
3392 return
3393 self.report_warning(f'{msg}. The formats won\'t be merged')
3394
3395 if temp_filename == '-':
3396 reason = ('using a downloader other than ffmpeg' if FFmpegFD.can_merge_formats(info_dict, self.params)
3397 else 'but the formats are incompatible for simultaneous download' if merger.available
3398 else 'but ffmpeg is not installed')
3399 self.report_warning(
3400 f'You have requested downloading multiple formats to stdout {reason}. '
3401 'The formats will be streamed one after the other')
3402 fname = temp_filename
3403 for f in info_dict['requested_formats']:
3404 new_info = dict(info_dict)
3405 del new_info['requested_formats']
3406 new_info.update(f)
3407 if temp_filename != '-':
3408 fname = prepend_extension(
3409 correct_ext(temp_filename, new_info['ext']),
3410 'f%s' % f['format_id'], new_info['ext'])
3411 if not self._ensure_dir_exists(fname):
3412 return
3413 f['filepath'] = fname
3414 downloaded.append(fname)
3415 partial_success, real_download = self.dl(fname, new_info)
3416 info_dict['__real_download'] = info_dict['__real_download'] or real_download
3417 success = success and partial_success
3418
3419 if downloaded and merger.available and not self.params.get('allow_unplayable_formats'):
3420 info_dict['__postprocessors'].append(merger)
3421 info_dict['__files_to_merge'] = downloaded
3422 # Even if there were no downloads, it is being merged only now
3423 info_dict['__real_download'] = True
3424 else:
3425 for file in downloaded:
3426 files_to_move[file] = None
3427 else:
3428 # Just a single file
3429 dl_filename = existing_video_file(full_filename, temp_filename)
3430 if dl_filename is None or dl_filename == temp_filename:
3431 # dl_filename == temp_filename could mean that the file was partially downloaded with --no-part.
3432 # So we should try to resume the download
3433 success, real_download = self.dl(temp_filename, info_dict)
3434 info_dict['__real_download'] = real_download
3435 else:
3436 self.report_file_already_downloaded(dl_filename)
3437
3438 dl_filename = dl_filename or temp_filename
3439 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
3440
3441 except network_exceptions as err:
3442 self.report_error('unable to download video data: %s' % error_to_compat_str(err))
3443 return
3444 except OSError as err:
3445 raise UnavailableVideoError(err)
3446 except (ContentTooShortError, ) as err:
3447 self.report_error(f'content too short (expected {err.expected} bytes and served {err.downloaded})')
3448 return
3449
3450 self._raise_pending_errors(info_dict)
3451 if success and full_filename != '-':
3452
3453 def fixup():
3454 do_fixup = True
3455 fixup_policy = self.params.get('fixup')
3456 vid = info_dict['id']
3457
3458 if fixup_policy in ('ignore', 'never'):
3459 return
3460 elif fixup_policy == 'warn':
3461 do_fixup = 'warn'
3462 elif fixup_policy != 'force':
3463 assert fixup_policy in ('detect_or_warn', None)
3464 if not info_dict.get('__real_download'):
3465 do_fixup = False
3466
3467 def ffmpeg_fixup(cndn, msg, cls):
3468 if not (do_fixup and cndn):
3469 return
3470 elif do_fixup == 'warn':
3471 self.report_warning(f'{vid}: {msg}')
3472 return
3473 pp = cls(self)
3474 if pp.available:
3475 info_dict['__postprocessors'].append(pp)
3476 else:
3477 self.report_warning(f'{vid}: {msg}. Install ffmpeg to fix this automatically')
3478
3479 stretched_ratio = info_dict.get('stretched_ratio')
3480 ffmpeg_fixup(stretched_ratio not in (1, None),
3481 f'Non-uniform pixel ratio {stretched_ratio}',
3482 FFmpegFixupStretchedPP)
3483
3484 downloader = get_suitable_downloader(info_dict, self.params) if 'protocol' in info_dict else None
3485 downloader = downloader.FD_NAME if downloader else None
3486
3487 ext = info_dict.get('ext')
3488 postprocessed_by_ffmpeg = info_dict.get('requested_formats') or any((
3489 isinstance(pp, FFmpegVideoConvertorPP)
3490 and resolve_recode_mapping(ext, pp.mapping)[0] not in (ext, None)
3491 ) for pp in self._pps['post_process'])
3492
3493 if not postprocessed_by_ffmpeg:
3494 ffmpeg_fixup(fd != FFmpegFD and ext == 'm4a'
3495 and info_dict.get('container') == 'm4a_dash',
3496 'writing DASH m4a. Only some players support this container',
3497 FFmpegFixupM4aPP)
3498 ffmpeg_fixup(downloader == 'hlsnative' and not self.params.get('hls_use_mpegts')
3499 or info_dict.get('is_live') and self.params.get('hls_use_mpegts') is None,
3500 'Possible MPEG-TS in MP4 container or malformed AAC timestamps',
3501 FFmpegFixupM3u8PP)
3502 ffmpeg_fixup(downloader == 'dashsegments'
3503 and (info_dict.get('is_live') or info_dict.get('is_dash_periods')),
3504 'Possible duplicate MOOV atoms', FFmpegFixupDuplicateMoovPP)
3505
3506 ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed timestamps detected', FFmpegFixupTimestampPP)
3507 ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed duration detected', FFmpegFixupDurationPP)
3508
3509 fixup()
3510 try:
3511 replace_info_dict(self.post_process(dl_filename, info_dict, files_to_move))
3512 except PostProcessingError as err:
3513 self.report_error('Postprocessing: %s' % str(err))
3514 return
3515 try:
3516 for ph in self._post_hooks:
3517 ph(info_dict['filepath'])
3518 except Exception as err:
3519 self.report_error('post hooks: %s' % str(err))
3520 return
3521 info_dict['__write_download_archive'] = True
3522
3523 assert info_dict is original_infodict # Make sure the info_dict was modified in-place
3524 if self.params.get('force_write_download_archive'):
3525 info_dict['__write_download_archive'] = True
3526 check_max_downloads()
3527
3528 def __download_wrapper(self, func):
3529 @functools.wraps(func)
3530 def wrapper(*args, **kwargs):
3531 try:
3532 res = func(*args, **kwargs)
3533 except UnavailableVideoError as e:
3534 self.report_error(e)
3535 except DownloadCancelled as e:
3536 self.to_screen(f'[info] {e}')
3537 if not self.params.get('break_per_url'):
3538 raise
3539 self._num_downloads = 0
3540 else:
3541 if self.params.get('dump_single_json', False):
3542 self.post_extract(res)
3543 self.to_stdout(json.dumps(self.sanitize_info(res)))
3544 return wrapper
3545
3546 def download(self, url_list):
3547 """Download a given list of URLs."""
3548 url_list = variadic(url_list) # Passing a single URL is a common mistake
3549 outtmpl = self.params['outtmpl']['default']
3550 if (len(url_list) > 1
3551 and outtmpl != '-'
3552 and '%' not in outtmpl
3553 and self.params.get('max_downloads') != 1):
3554 raise SameFileError(outtmpl)
3555
3556 for url in url_list:
3557 self.__download_wrapper(self.extract_info)(
3558 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
3559
3560 return self._download_retcode
3561
3562 def download_with_info_file(self, info_filename):
3563 with contextlib.closing(fileinput.FileInput(
3564 [info_filename], mode='r',
3565 openhook=fileinput.hook_encoded('utf-8'))) as f:
3566 # FileInput doesn't have a read method, we can't call json.load
3567 infos = [self.sanitize_info(info, self.params.get('clean_infojson', True))
3568 for info in variadic(json.loads('\n'.join(f)))]
3569 for info in infos:
3570 try:
3571 self.__download_wrapper(self.process_ie_result)(info, download=True)
3572 except (DownloadError, EntryNotInPlaylist, ReExtractInfo) as e:
3573 if not isinstance(e, EntryNotInPlaylist):
3574 self.to_stderr('\r')
3575 webpage_url = info.get('webpage_url')
3576 if webpage_url is None:
3577 raise
3578 self.report_warning(f'The info failed to download: {e}; trying with URL {webpage_url}')
3579 self.download([webpage_url])
3580 except ExtractorError as e:
3581 self.report_error(e)
3582 return self._download_retcode
3583
3584 @staticmethod
3585 def sanitize_info(info_dict, remove_private_keys=False):
3586 ''' Sanitize the infodict for converting to json '''
3587 if info_dict is None:
3588 return info_dict
3589 info_dict.setdefault('epoch', int(time.time()))
3590 info_dict.setdefault('_type', 'video')
3591 info_dict.setdefault('_version', {
3592 'version': __version__,
3593 'current_git_head': current_git_head(),
3594 'release_git_head': RELEASE_GIT_HEAD,
3595 'repository': ORIGIN,
3596 })
3597
3598 if remove_private_keys:
3599 reject = lambda k, v: v is None or k.startswith('__') or k in {
3600 'requested_downloads', 'requested_formats', 'requested_subtitles', 'requested_entries',
3601 'entries', 'filepath', '_filename', 'filename', 'infojson_filename', 'original_url',
3602 'playlist_autonumber',
3603 }
3604 else:
3605 reject = lambda k, v: False
3606
3607 def filter_fn(obj):
3608 if isinstance(obj, dict):
3609 return {k: filter_fn(v) for k, v in obj.items() if not reject(k, v)}
3610 elif isinstance(obj, (list, tuple, set, LazyList)):
3611 return list(map(filter_fn, obj))
3612 elif obj is None or isinstance(obj, (str, int, float, bool)):
3613 return obj
3614 else:
3615 return repr(obj)
3616
3617 return filter_fn(info_dict)
3618
3619 @staticmethod
3620 def filter_requested_info(info_dict, actually_filter=True):
3621 ''' Alias of sanitize_info for backward compatibility '''
3622 return YoutubeDL.sanitize_info(info_dict, actually_filter)
3623
3624 def _delete_downloaded_files(self, *files_to_delete, info={}, msg=None):
3625 for filename in set(filter(None, files_to_delete)):
3626 if msg:
3627 self.to_screen(msg % filename)
3628 try:
3629 os.remove(filename)
3630 except OSError:
3631 self.report_warning(f'Unable to delete file {filename}')
3632 if filename in info.get('__files_to_move', []): # NB: Delete even if None
3633 del info['__files_to_move'][filename]
3634
3635 @staticmethod
3636 def post_extract(info_dict):
3637 def actual_post_extract(info_dict):
3638 if info_dict.get('_type') in ('playlist', 'multi_video'):
3639 for video_dict in info_dict.get('entries', {}):
3640 actual_post_extract(video_dict or {})
3641 return
3642
3643 post_extractor = info_dict.pop('__post_extractor', None) or (lambda: {})
3644 info_dict.update(post_extractor())
3645
3646 actual_post_extract(info_dict or {})
3647
3648 def run_pp(self, pp, infodict):
3649 files_to_delete = []
3650 if '__files_to_move' not in infodict:
3651 infodict['__files_to_move'] = {}
3652 try:
3653 files_to_delete, infodict = pp.run(infodict)
3654 except PostProcessingError as e:
3655 # Must be True and not 'only_download'
3656 if self.params.get('ignoreerrors') is True:
3657 self.report_error(e)
3658 return infodict
3659 raise
3660
3661 if not files_to_delete:
3662 return infodict
3663 if self.params.get('keepvideo', False):
3664 for f in files_to_delete:
3665 infodict['__files_to_move'].setdefault(f, '')
3666 else:
3667 self._delete_downloaded_files(
3668 *files_to_delete, info=infodict, msg='Deleting original file %s (pass -k to keep)')
3669 return infodict
3670
3671 def run_all_pps(self, key, info, *, additional_pps=None):
3672 if key != 'video':
3673 self._forceprint(key, info)
3674 for pp in (additional_pps or []) + self._pps[key]:
3675 info = self.run_pp(pp, info)
3676 return info
3677
3678 def pre_process(self, ie_info, key='pre_process', files_to_move=None):
3679 info = dict(ie_info)
3680 info['__files_to_move'] = files_to_move or {}
3681 try:
3682 info = self.run_all_pps(key, info)
3683 except PostProcessingError as err:
3684 msg = f'Preprocessing: {err}'
3685 info.setdefault('__pending_error', msg)
3686 self.report_error(msg, is_error=False)
3687 return info, info.pop('__files_to_move', None)
3688
3689 def post_process(self, filename, info, files_to_move=None):
3690 """Run all the postprocessors on the given file."""
3691 info['filepath'] = filename
3692 info['__files_to_move'] = files_to_move or {}
3693 info = self.run_all_pps('post_process', info, additional_pps=info.get('__postprocessors'))
3694 info = self.run_pp(MoveFilesAfterDownloadPP(self), info)
3695 del info['__files_to_move']
3696 return self.run_all_pps('after_move', info)
3697
3698 def _make_archive_id(self, info_dict):
3699 video_id = info_dict.get('id')
3700 if not video_id:
3701 return
3702 # Future-proof against any change in case
3703 # and backwards compatibility with prior versions
3704 extractor = info_dict.get('extractor_key') or info_dict.get('ie_key') # key in a playlist
3705 if extractor is None:
3706 url = str_or_none(info_dict.get('url'))
3707 if not url:
3708 return
3709 # Try to find matching extractor for the URL and take its ie_key
3710 for ie_key, ie in self._ies.items():
3711 if ie.suitable(url):
3712 extractor = ie_key
3713 break
3714 else:
3715 return
3716 return make_archive_id(extractor, video_id)
3717
3718 def in_download_archive(self, info_dict):
3719 if not self.archive:
3720 return False
3721
3722 vid_ids = [self._make_archive_id(info_dict)]
3723 vid_ids.extend(info_dict.get('_old_archive_ids') or [])
3724 return any(id_ in self.archive for id_ in vid_ids)
3725
3726 def record_download_archive(self, info_dict):
3727 fn = self.params.get('download_archive')
3728 if fn is None:
3729 return
3730 vid_id = self._make_archive_id(info_dict)
3731 assert vid_id
3732
3733 self.write_debug(f'Adding to archive: {vid_id}')
3734 if is_path_like(fn):
3735 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
3736 archive_file.write(vid_id + '\n')
3737 self.archive.add(vid_id)
3738
3739 @staticmethod
3740 def format_resolution(format, default='unknown'):
3741 if format.get('vcodec') == 'none' and format.get('acodec') != 'none':
3742 return 'audio only'
3743 if format.get('resolution') is not None:
3744 return format['resolution']
3745 if format.get('width') and format.get('height'):
3746 return '%dx%d' % (format['width'], format['height'])
3747 elif format.get('height'):
3748 return '%sp' % format['height']
3749 elif format.get('width'):
3750 return '%dx?' % format['width']
3751 return default
3752
3753 def _list_format_headers(self, *headers):
3754 if self.params.get('listformats_table', True) is not False:
3755 return [self._format_out(header, self.Styles.HEADERS) for header in headers]
3756 return headers
3757
3758 def _format_note(self, fdict):
3759 res = ''
3760 if fdict.get('ext') in ['f4f', 'f4m']:
3761 res += '(unsupported)'
3762 if fdict.get('language'):
3763 if res:
3764 res += ' '
3765 res += '[%s]' % fdict['language']
3766 if fdict.get('format_note') is not None:
3767 if res:
3768 res += ' '
3769 res += fdict['format_note']
3770 if fdict.get('tbr') is not None:
3771 if res:
3772 res += ', '
3773 res += '%4dk' % fdict['tbr']
3774 if fdict.get('container') is not None:
3775 if res:
3776 res += ', '
3777 res += '%s container' % fdict['container']
3778 if (fdict.get('vcodec') is not None
3779 and fdict.get('vcodec') != 'none'):
3780 if res:
3781 res += ', '
3782 res += fdict['vcodec']
3783 if fdict.get('vbr') is not None:
3784 res += '@'
3785 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
3786 res += 'video@'
3787 if fdict.get('vbr') is not None:
3788 res += '%4dk' % fdict['vbr']
3789 if fdict.get('fps') is not None:
3790 if res:
3791 res += ', '
3792 res += '%sfps' % fdict['fps']
3793 if fdict.get('acodec') is not None:
3794 if res:
3795 res += ', '
3796 if fdict['acodec'] == 'none':
3797 res += 'video only'
3798 else:
3799 res += '%-5s' % fdict['acodec']
3800 elif fdict.get('abr') is not None:
3801 if res:
3802 res += ', '
3803 res += 'audio'
3804 if fdict.get('abr') is not None:
3805 res += '@%3dk' % fdict['abr']
3806 if fdict.get('asr') is not None:
3807 res += ' (%5dHz)' % fdict['asr']
3808 if fdict.get('filesize') is not None:
3809 if res:
3810 res += ', '
3811 res += format_bytes(fdict['filesize'])
3812 elif fdict.get('filesize_approx') is not None:
3813 if res:
3814 res += ', '
3815 res += '~' + format_bytes(fdict['filesize_approx'])
3816 return res
3817
3818 def _get_formats(self, info_dict):
3819 if info_dict.get('formats') is None:
3820 if info_dict.get('url') and info_dict.get('_type', 'video') == 'video':
3821 return [info_dict]
3822 return []
3823 return info_dict['formats']
3824
3825 def render_formats_table(self, info_dict):
3826 formats = self._get_formats(info_dict)
3827 if not formats:
3828 return
3829 if not self.params.get('listformats_table', True) is not False:
3830 table = [
3831 [
3832 format_field(f, 'format_id'),
3833 format_field(f, 'ext'),
3834 self.format_resolution(f),
3835 self._format_note(f)
3836 ] for f in formats if (f.get('preference') or 0) >= -1000]
3837 return render_table(['format code', 'extension', 'resolution', 'note'], table, extra_gap=1)
3838
3839 def simplified_codec(f, field):
3840 assert field in ('acodec', 'vcodec')
3841 codec = f.get(field)
3842 if not codec:
3843 return 'unknown'
3844 elif codec != 'none':
3845 return '.'.join(codec.split('.')[:4])
3846
3847 if field == 'vcodec' and f.get('acodec') == 'none':
3848 return 'images'
3849 elif field == 'acodec' and f.get('vcodec') == 'none':
3850 return ''
3851 return self._format_out('audio only' if field == 'vcodec' else 'video only',
3852 self.Styles.SUPPRESS)
3853
3854 delim = self._format_out('\u2502', self.Styles.DELIM, '|', test_encoding=True)
3855 table = [
3856 [
3857 self._format_out(format_field(f, 'format_id'), self.Styles.ID),
3858 format_field(f, 'ext'),
3859 format_field(f, func=self.format_resolution, ignore=('audio only', 'images')),
3860 format_field(f, 'fps', '\t%d', func=round),
3861 format_field(f, 'dynamic_range', '%s', ignore=(None, 'SDR')).replace('HDR', ''),
3862 format_field(f, 'audio_channels', '\t%s'),
3863 delim, (
3864 format_field(f, 'filesize', ' \t%s', func=format_bytes)
3865 or format_field(f, 'filesize_approx', '≈\t%s', func=format_bytes)
3866 or format_field(try_call(lambda: format_bytes(int(info_dict['duration'] * f['tbr'] * (1024 / 8)))),
3867 None, self._format_out('~\t%s', self.Styles.SUPPRESS))),
3868 format_field(f, 'tbr', '\t%dk', func=round),
3869 shorten_protocol_name(f.get('protocol', '')),
3870 delim,
3871 simplified_codec(f, 'vcodec'),
3872 format_field(f, 'vbr', '\t%dk', func=round),
3873 simplified_codec(f, 'acodec'),
3874 format_field(f, 'abr', '\t%dk', func=round),
3875 format_field(f, 'asr', '\t%s', func=format_decimal_suffix),
3876 join_nonempty(format_field(f, 'language', '[%s]'), join_nonempty(
3877 self._format_out('UNSUPPORTED', self.Styles.BAD_FORMAT) if f.get('ext') in ('f4f', 'f4m') else None,
3878 (self._format_out('Maybe DRM', self.Styles.WARNING) if f.get('has_drm') == 'maybe'
3879 else self._format_out('DRM', self.Styles.BAD_FORMAT) if f.get('has_drm') else None),
3880 format_field(f, 'format_note'),
3881 format_field(f, 'container', ignore=(None, f.get('ext'))),
3882 delim=', '), delim=' '),
3883 ] for f in formats if f.get('preference') is None or f['preference'] >= -1000]
3884 header_line = self._list_format_headers(
3885 'ID', 'EXT', 'RESOLUTION', '\tFPS', 'HDR', 'CH', delim, '\tFILESIZE', '\tTBR', 'PROTO',
3886 delim, 'VCODEC', '\tVBR', 'ACODEC', '\tABR', '\tASR', 'MORE INFO')
3887
3888 return render_table(
3889 header_line, table, hide_empty=True,
3890 delim=self._format_out('\u2500', self.Styles.DELIM, '-', test_encoding=True))
3891
3892 def render_thumbnails_table(self, info_dict):
3893 thumbnails = list(info_dict.get('thumbnails') or [])
3894 if not thumbnails:
3895 return None
3896 return render_table(
3897 self._list_format_headers('ID', 'Width', 'Height', 'URL'),
3898 [[t.get('id'), t.get('width') or 'unknown', t.get('height') or 'unknown', t['url']] for t in thumbnails])
3899
3900 def render_subtitles_table(self, video_id, subtitles):
3901 def _row(lang, formats):
3902 exts, names = zip(*((f['ext'], f.get('name') or 'unknown') for f in reversed(formats)))
3903 if len(set(names)) == 1:
3904 names = [] if names[0] == 'unknown' else names[:1]
3905 return [lang, ', '.join(names), ', '.join(exts)]
3906
3907 if not subtitles:
3908 return None
3909 return render_table(
3910 self._list_format_headers('Language', 'Name', 'Formats'),
3911 [_row(lang, formats) for lang, formats in subtitles.items()],
3912 hide_empty=True)
3913
3914 def __list_table(self, video_id, name, func, *args):
3915 table = func(*args)
3916 if not table:
3917 self.to_screen(f'{video_id} has no {name}')
3918 return
3919 self.to_screen(f'[info] Available {name} for {video_id}:')
3920 self.to_stdout(table)
3921
3922 def list_formats(self, info_dict):
3923 self.__list_table(info_dict['id'], 'formats', self.render_formats_table, info_dict)
3924
3925 def list_thumbnails(self, info_dict):
3926 self.__list_table(info_dict['id'], 'thumbnails', self.render_thumbnails_table, info_dict)
3927
3928 def list_subtitles(self, video_id, subtitles, name='subtitles'):
3929 self.__list_table(video_id, name, self.render_subtitles_table, video_id, subtitles)
3930
3931 def print_debug_header(self):
3932 if not self.params.get('verbose'):
3933 return
3934
3935 from . import _IN_CLI # Must be delayed import
3936
3937 # These imports can be slow. So import them only as needed
3938 from .extractor.extractors import _LAZY_LOADER
3939 from .extractor.extractors import (
3940 _PLUGIN_CLASSES as plugin_ies,
3941 _PLUGIN_OVERRIDES as plugin_ie_overrides
3942 )
3943
3944 def get_encoding(stream):
3945 ret = str(getattr(stream, 'encoding', 'missing (%s)' % type(stream).__name__))
3946 additional_info = []
3947 if os.environ.get('TERM', '').lower() == 'dumb':
3948 additional_info.append('dumb')
3949 if not supports_terminal_sequences(stream):
3950 from .utils import WINDOWS_VT_MODE # Must be imported locally
3951 additional_info.append('No VT' if WINDOWS_VT_MODE is False else 'No ANSI')
3952 if additional_info:
3953 ret = f'{ret} ({",".join(additional_info)})'
3954 return ret
3955
3956 encoding_str = 'Encodings: locale %s, fs %s, pref %s, %s' % (
3957 locale.getpreferredencoding(),
3958 sys.getfilesystemencoding(),
3959 self.get_encoding(),
3960 ', '.join(
3961 f'{key} {get_encoding(stream)}' for key, stream in self._out_files.items_
3962 if stream is not None and key != 'console')
3963 )
3964
3965 logger = self.params.get('logger')
3966 if logger:
3967 write_debug = lambda msg: logger.debug(f'[debug] {msg}')
3968 write_debug(encoding_str)
3969 else:
3970 write_string(f'[debug] {encoding_str}\n', encoding=None)
3971 write_debug = lambda msg: self._write_string(f'[debug] {msg}\n')
3972
3973 source = detect_variant()
3974 if VARIANT not in (None, 'pip'):
3975 source += '*'
3976 klass = type(self)
3977 write_debug(join_nonempty(
3978 f'{REPOSITORY.rpartition("/")[2]} version',
3979 _make_label(ORIGIN, CHANNEL.partition('@')[2] or __version__, __version__),
3980 f'[{RELEASE_GIT_HEAD[:9]}]' if RELEASE_GIT_HEAD else '',
3981 '' if source == 'unknown' else f'({source})',
3982 '' if _IN_CLI else 'API' if klass == YoutubeDL else f'API:{self.__module__}.{klass.__qualname__}',
3983 delim=' '))
3984
3985 if not _IN_CLI:
3986 write_debug(f'params: {self.params}')
3987
3988 if not _LAZY_LOADER:
3989 if os.environ.get('YTDLP_NO_LAZY_EXTRACTORS'):
3990 write_debug('Lazy loading extractors is forcibly disabled')
3991 else:
3992 write_debug('Lazy loading extractors is disabled')
3993 if self.params['compat_opts']:
3994 write_debug('Compatibility options: %s' % ', '.join(self.params['compat_opts']))
3995
3996 if current_git_head():
3997 write_debug(f'Git HEAD: {current_git_head()}')
3998 write_debug(system_identifier())
3999
4000 exe_versions, ffmpeg_features = FFmpegPostProcessor.get_versions_and_features(self)
4001 ffmpeg_features = {key for key, val in ffmpeg_features.items() if val}
4002 if ffmpeg_features:
4003 exe_versions['ffmpeg'] += ' (%s)' % ','.join(sorted(ffmpeg_features))
4004
4005 exe_versions['rtmpdump'] = rtmpdump_version()
4006 exe_versions['phantomjs'] = PhantomJSwrapper._version()
4007 exe_str = ', '.join(
4008 f'{exe} {v}' for exe, v in sorted(exe_versions.items()) if v
4009 ) or 'none'
4010 write_debug('exe versions: %s' % exe_str)
4011
4012 from .compat.compat_utils import get_package_info
4013 from .dependencies import available_dependencies
4014
4015 write_debug('Optional libraries: %s' % (', '.join(sorted({
4016 join_nonempty(*get_package_info(m)) for m in available_dependencies.values()
4017 })) or 'none'))
4018
4019 write_debug(f'Proxy map: {self.proxies}')
4020 write_debug(f'Request Handlers: {", ".join(rh.RH_NAME for rh in self._request_director.handlers.values())}')
4021 for plugin_type, plugins in {'Extractor': plugin_ies, 'Post-Processor': plugin_pps}.items():
4022 display_list = ['%s%s' % (
4023 klass.__name__, '' if klass.__name__ == name else f' as {name}')
4024 for name, klass in plugins.items()]
4025 if plugin_type == 'Extractor':
4026 display_list.extend(f'{plugins[-1].IE_NAME.partition("+")[2]} ({parent.__name__})'
4027 for parent, plugins in plugin_ie_overrides.items())
4028 if not display_list:
4029 continue
4030 write_debug(f'{plugin_type} Plugins: {", ".join(sorted(display_list))}')
4031
4032 plugin_dirs = plugin_directories()
4033 if plugin_dirs:
4034 write_debug(f'Plugin directories: {plugin_dirs}')
4035
4036 # Not implemented
4037 if False and self.params.get('call_home'):
4038 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode()
4039 write_debug('Public IP address: %s' % ipaddr)
4040 latest_version = self.urlopen(
4041 'https://yt-dl.org/latest/version').read().decode()
4042 if version_tuple(latest_version) > version_tuple(__version__):
4043 self.report_warning(
4044 'You are using an outdated version (newest version: %s)! '
4045 'See https://yt-dl.org/update if you need help updating.' %
4046 latest_version)
4047
4048 @functools.cached_property
4049 def proxies(self):
4050 """Global proxy configuration"""
4051 opts_proxy = self.params.get('proxy')
4052 if opts_proxy is not None:
4053 if opts_proxy == '':
4054 opts_proxy = '__noproxy__'
4055 proxies = {'all': opts_proxy}
4056 else:
4057 proxies = urllib.request.getproxies()
4058 # compat. Set HTTPS_PROXY to __noproxy__ to revert
4059 if 'http' in proxies and 'https' not in proxies:
4060 proxies['https'] = proxies['http']
4061
4062 return proxies
4063
4064 @functools.cached_property
4065 def cookiejar(self):
4066 """Global cookiejar instance"""
4067 return load_cookies(
4068 self.params.get('cookiefile'), self.params.get('cookiesfrombrowser'), self)
4069
4070 @property
4071 def _opener(self):
4072 """
4073 Get a urllib OpenerDirector from the Urllib handler (deprecated).
4074 """
4075 self.deprecation_warning('YoutubeDL._opener is deprecated, use YoutubeDL.urlopen()')
4076 handler = self._request_director.handlers['Urllib']
4077 return handler._get_instance(cookiejar=self.cookiejar, proxies=self.proxies)
4078
4079 def urlopen(self, req):
4080 """ Start an HTTP download """
4081 if isinstance(req, str):
4082 req = Request(req)
4083 elif isinstance(req, urllib.request.Request):
4084 self.deprecation_warning(
4085 'Passing a urllib.request.Request object to YoutubeDL.urlopen() is deprecated. '
4086 'Use yt_dlp.networking.common.Request instead.')
4087 req = urllib_req_to_req(req)
4088 assert isinstance(req, Request)
4089
4090 # compat: Assume user:pass url params are basic auth
4091 url, basic_auth_header = extract_basic_auth(req.url)
4092 if basic_auth_header:
4093 req.headers['Authorization'] = basic_auth_header
4094 req.url = sanitize_url(url)
4095
4096 clean_proxies(proxies=req.proxies, headers=req.headers)
4097 clean_headers(req.headers)
4098
4099 try:
4100 return self._request_director.send(req)
4101 except NoSupportingHandlers as e:
4102 for ue in e.unsupported_errors:
4103 # FIXME: This depends on the order of errors.
4104 if not (ue.handler and ue.msg):
4105 continue
4106 if ue.handler.RH_KEY == 'Urllib' and 'unsupported url scheme: "file"' in ue.msg.lower():
4107 raise RequestError(
4108 'file:// URLs are disabled by default in yt-dlp for security reasons. '
4109 'Use --enable-file-urls to enable at your own risk.', cause=ue) from ue
4110 if 'unsupported proxy type: "https"' in ue.msg.lower():
4111 raise RequestError(
4112 'To use an HTTPS proxy for this request, one of the following dependencies needs to be installed: requests')
4113
4114 elif (
4115 re.match(r'unsupported url scheme: "wss?"', ue.msg.lower())
4116 and 'websockets' not in self._request_director.handlers
4117 ):
4118 raise RequestError(
4119 'This request requires WebSocket support. '
4120 'Ensure one of the following dependencies are installed: websockets',
4121 cause=ue) from ue
4122 raise
4123 except SSLError as e:
4124 if 'UNSAFE_LEGACY_RENEGOTIATION_DISABLED' in str(e):
4125 raise RequestError('UNSAFE_LEGACY_RENEGOTIATION_DISABLED: Try using --legacy-server-connect', cause=e) from e
4126 elif 'SSLV3_ALERT_HANDSHAKE_FAILURE' in str(e):
4127 raise RequestError(
4128 'SSLV3_ALERT_HANDSHAKE_FAILURE: The server may not support the current cipher list. '
4129 'Try using --legacy-server-connect', cause=e) from e
4130 raise
4131
4132 def build_request_director(self, handlers, preferences=None):
4133 logger = _YDLLogger(self)
4134 headers = self.params['http_headers'].copy()
4135 proxies = self.proxies.copy()
4136 clean_headers(headers)
4137 clean_proxies(proxies, headers)
4138
4139 director = RequestDirector(logger=logger, verbose=self.params.get('debug_printtraffic'))
4140 for handler in handlers:
4141 director.add_handler(handler(
4142 logger=logger,
4143 headers=headers,
4144 cookiejar=self.cookiejar,
4145 proxies=proxies,
4146 prefer_system_certs='no-certifi' in self.params['compat_opts'],
4147 verify=not self.params.get('nocheckcertificate'),
4148 **traverse_obj(self.params, {
4149 'verbose': 'debug_printtraffic',
4150 'source_address': 'source_address',
4151 'timeout': 'socket_timeout',
4152 'legacy_ssl_support': 'legacyserverconnect',
4153 'enable_file_urls': 'enable_file_urls',
4154 'client_cert': {
4155 'client_certificate': 'client_certificate',
4156 'client_certificate_key': 'client_certificate_key',
4157 'client_certificate_password': 'client_certificate_password',
4158 },
4159 }),
4160 ))
4161 director.preferences.update(preferences or [])
4162 if 'prefer-legacy-http-handler' in self.params['compat_opts']:
4163 director.preferences.add(lambda rh, _: 500 if rh.RH_KEY == 'Urllib' else 0)
4164 return director
4165
4166 @functools.cached_property
4167 def _request_director(self):
4168 return self.build_request_director(_REQUEST_HANDLERS.values(), _RH_PREFERENCES)
4169
4170 def encode(self, s):
4171 if isinstance(s, bytes):
4172 return s # Already encoded
4173
4174 try:
4175 return s.encode(self.get_encoding())
4176 except UnicodeEncodeError as err:
4177 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
4178 raise
4179
4180 def get_encoding(self):
4181 encoding = self.params.get('encoding')
4182 if encoding is None:
4183 encoding = preferredencoding()
4184 return encoding
4185
4186 def _write_info_json(self, label, ie_result, infofn, overwrite=None):
4187 ''' Write infojson and returns True = written, 'exists' = Already exists, False = skip, None = error '''
4188 if overwrite is None:
4189 overwrite = self.params.get('overwrites', True)
4190 if not self.params.get('writeinfojson'):
4191 return False
4192 elif not infofn:
4193 self.write_debug(f'Skipping writing {label} infojson')
4194 return False
4195 elif not self._ensure_dir_exists(infofn):
4196 return None
4197 elif not overwrite and os.path.exists(infofn):
4198 self.to_screen(f'[info] {label.title()} metadata is already present')
4199 return 'exists'
4200
4201 self.to_screen(f'[info] Writing {label} metadata as JSON to: {infofn}')
4202 try:
4203 write_json_file(self.sanitize_info(ie_result, self.params.get('clean_infojson', True)), infofn)
4204 return True
4205 except OSError:
4206 self.report_error(f'Cannot write {label} metadata to JSON file {infofn}')
4207 return None
4208
4209 def _write_description(self, label, ie_result, descfn):
4210 ''' Write description and returns True = written, False = skip, None = error '''
4211 if not self.params.get('writedescription'):
4212 return False
4213 elif not descfn:
4214 self.write_debug(f'Skipping writing {label} description')
4215 return False
4216 elif not self._ensure_dir_exists(descfn):
4217 return None
4218 elif not self.params.get('overwrites', True) and os.path.exists(descfn):
4219 self.to_screen(f'[info] {label.title()} description is already present')
4220 elif ie_result.get('description') is None:
4221 self.to_screen(f'[info] There\'s no {label} description to write')
4222 return False
4223 else:
4224 try:
4225 self.to_screen(f'[info] Writing {label} description to: {descfn}')
4226 with open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
4227 descfile.write(ie_result['description'])
4228 except OSError:
4229 self.report_error(f'Cannot write {label} description file {descfn}')
4230 return None
4231 return True
4232
4233 def _write_subtitles(self, info_dict, filename):
4234 ''' Write subtitles to file and return list of (sub_filename, final_sub_filename); or None if error'''
4235 ret = []
4236 subtitles = info_dict.get('requested_subtitles')
4237 if not (self.params.get('writesubtitles') or self.params.get('writeautomaticsub')):
4238 # subtitles download errors are already managed as troubles in relevant IE
4239 # that way it will silently go on when used with unsupporting IE
4240 return ret
4241 elif not subtitles:
4242 self.to_screen('[info] There are no subtitles for the requested languages')
4243 return ret
4244 sub_filename_base = self.prepare_filename(info_dict, 'subtitle')
4245 if not sub_filename_base:
4246 self.to_screen('[info] Skipping writing video subtitles')
4247 return ret
4248
4249 for sub_lang, sub_info in subtitles.items():
4250 sub_format = sub_info['ext']
4251 sub_filename = subtitles_filename(filename, sub_lang, sub_format, info_dict.get('ext'))
4252 sub_filename_final = subtitles_filename(sub_filename_base, sub_lang, sub_format, info_dict.get('ext'))
4253 existing_sub = self.existing_file((sub_filename_final, sub_filename))
4254 if existing_sub:
4255 self.to_screen(f'[info] Video subtitle {sub_lang}.{sub_format} is already present')
4256 sub_info['filepath'] = existing_sub
4257 ret.append((existing_sub, sub_filename_final))
4258 continue
4259
4260 self.to_screen(f'[info] Writing video subtitles to: {sub_filename}')
4261 if sub_info.get('data') is not None:
4262 try:
4263 # Use newline='' to prevent conversion of newline characters
4264 # See https://github.com/ytdl-org/youtube-dl/issues/10268
4265 with open(sub_filename, 'w', encoding='utf-8', newline='') as subfile:
4266 subfile.write(sub_info['data'])
4267 sub_info['filepath'] = sub_filename
4268 ret.append((sub_filename, sub_filename_final))
4269 continue
4270 except OSError:
4271 self.report_error(f'Cannot write video subtitles file {sub_filename}')
4272 return None
4273
4274 try:
4275 sub_copy = sub_info.copy()
4276 sub_copy.setdefault('http_headers', info_dict.get('http_headers'))
4277 self.dl(sub_filename, sub_copy, subtitle=True)
4278 sub_info['filepath'] = sub_filename
4279 ret.append((sub_filename, sub_filename_final))
4280 except (DownloadError, ExtractorError, IOError, OSError, ValueError) + network_exceptions as err:
4281 msg = f'Unable to download video subtitles for {sub_lang!r}: {err}'
4282 if self.params.get('ignoreerrors') is not True: # False or 'only_download'
4283 if not self.params.get('ignoreerrors'):
4284 self.report_error(msg)
4285 raise DownloadError(msg)
4286 self.report_warning(msg)
4287 return ret
4288
4289 def _write_thumbnails(self, label, info_dict, filename, thumb_filename_base=None):
4290 ''' Write thumbnails to file and return list of (thumb_filename, final_thumb_filename); or None if error '''
4291 write_all = self.params.get('write_all_thumbnails', False)
4292 thumbnails, ret = [], []
4293 if write_all or self.params.get('writethumbnail', False):
4294 thumbnails = info_dict.get('thumbnails') or []
4295 if not thumbnails:
4296 self.to_screen(f'[info] There are no {label} thumbnails to download')
4297 return ret
4298 multiple = write_all and len(thumbnails) > 1
4299
4300 if thumb_filename_base is None:
4301 thumb_filename_base = filename
4302 if thumbnails and not thumb_filename_base:
4303 self.write_debug(f'Skipping writing {label} thumbnail')
4304 return ret
4305
4306 if thumbnails and not self._ensure_dir_exists(filename):
4307 return None
4308
4309 for idx, t in list(enumerate(thumbnails))[::-1]:
4310 thumb_ext = (f'{t["id"]}.' if multiple else '') + determine_ext(t['url'], 'jpg')
4311 thumb_display_id = f'{label} thumbnail {t["id"]}'
4312 thumb_filename = replace_extension(filename, thumb_ext, info_dict.get('ext'))
4313 thumb_filename_final = replace_extension(thumb_filename_base, thumb_ext, info_dict.get('ext'))
4314
4315 existing_thumb = self.existing_file((thumb_filename_final, thumb_filename))
4316 if existing_thumb:
4317 self.to_screen('[info] %s is already present' % (
4318 thumb_display_id if multiple else f'{label} thumbnail').capitalize())
4319 t['filepath'] = existing_thumb
4320 ret.append((existing_thumb, thumb_filename_final))
4321 else:
4322 self.to_screen(f'[info] Downloading {thumb_display_id} ...')
4323 try:
4324 uf = self.urlopen(Request(t['url'], headers=t.get('http_headers', {})))
4325 self.to_screen(f'[info] Writing {thumb_display_id} to: {thumb_filename}')
4326 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
4327 shutil.copyfileobj(uf, thumbf)
4328 ret.append((thumb_filename, thumb_filename_final))
4329 t['filepath'] = thumb_filename
4330 except network_exceptions as err:
4331 if isinstance(err, HTTPError) and err.status == 404:
4332 self.to_screen(f'[info] {thumb_display_id.title()} does not exist')
4333 else:
4334 self.report_warning(f'Unable to download {thumb_display_id}: {err}')
4335 thumbnails.pop(idx)
4336 if ret and not write_all:
4337 break
4338 return ret