]> jfr.im git - yt-dlp.git/blob - yt_dlp/YoutubeDL.py
8d96498a6780c35210c0ae22d006968abb73ab6a
[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 _CompatHTTPError,
44 network_exceptions,
45 )
46 from .plugins import directories as plugin_directories
47 from .postprocessor import _PLUGIN_CLASSES as plugin_pps
48 from .postprocessor import (
49 EmbedThumbnailPP,
50 FFmpegFixupDuplicateMoovPP,
51 FFmpegFixupDurationPP,
52 FFmpegFixupM3u8PP,
53 FFmpegFixupM4aPP,
54 FFmpegFixupStretchedPP,
55 FFmpegFixupTimestampPP,
56 FFmpegMergerPP,
57 FFmpegPostProcessor,
58 FFmpegVideoConvertorPP,
59 MoveFilesAfterDownloadPP,
60 get_postprocessor,
61 )
62 from .postprocessor.ffmpeg import resolve_mapping as resolve_recode_mapping
63 from .update import (
64 REPOSITORY,
65 _get_system_deprecation,
66 _make_label,
67 current_git_head,
68 detect_variant,
69 )
70 from .utils import (
71 DEFAULT_OUTTMPL,
72 IDENTITY,
73 LINK_TEMPLATES,
74 MEDIA_EXTENSIONS,
75 NO_DEFAULT,
76 NUMBER_RE,
77 OUTTMPL_TYPES,
78 POSTPROCESS_WHEN,
79 STR_FORMAT_RE_TMPL,
80 STR_FORMAT_TYPES,
81 ContentTooShortError,
82 DateRange,
83 DownloadCancelled,
84 DownloadError,
85 EntryNotInPlaylist,
86 ExistingVideoReached,
87 ExtractorError,
88 FormatSorter,
89 GeoRestrictedError,
90 ISO3166Utils,
91 LazyList,
92 MaxDownloadsReached,
93 Namespace,
94 PagedList,
95 PlaylistEntries,
96 Popen,
97 PostProcessingError,
98 ReExtractInfo,
99 RejectedVideoReached,
100 SameFileError,
101 UnavailableVideoError,
102 UserNotLive,
103 age_restricted,
104 args_to_str,
105 bug_reports_message,
106 date_from_str,
107 deprecation_warning,
108 determine_ext,
109 determine_protocol,
110 encode_compat_str,
111 encodeFilename,
112 error_to_compat_str,
113 escapeHTML,
114 expand_path,
115 extract_basic_auth,
116 filter_dict,
117 float_or_none,
118 format_bytes,
119 format_decimal_suffix,
120 format_field,
121 formatSeconds,
122 get_compatible_ext,
123 get_domain,
124 int_or_none,
125 iri_to_uri,
126 is_path_like,
127 join_nonempty,
128 locked_file,
129 make_archive_id,
130 make_dir,
131 number_of_digits,
132 orderedSet,
133 orderedSet_from_options,
134 parse_filesize,
135 preferredencoding,
136 prepend_extension,
137 remove_terminal_sequences,
138 render_table,
139 replace_extension,
140 sanitize_filename,
141 sanitize_path,
142 sanitize_url,
143 str_or_none,
144 strftime_or_none,
145 subtitles_filename,
146 supports_terminal_sequences,
147 system_identifier,
148 timetuple_from_msec,
149 to_high_limit_path,
150 traverse_obj,
151 try_call,
152 try_get,
153 url_basename,
154 variadic,
155 version_tuple,
156 windows_enable_vt_mode,
157 write_json_file,
158 write_string,
159 )
160 from .utils._utils import _YDLLogger
161 from .utils.networking import (
162 HTTPHeaderDict,
163 clean_headers,
164 clean_proxies,
165 std_headers,
166 )
167 from .version import CHANNEL, ORIGIN, RELEASE_GIT_HEAD, VARIANT, __version__
168
169 if compat_os_name == 'nt':
170 import ctypes
171
172
173 class YoutubeDL:
174 """YoutubeDL class.
175
176 YoutubeDL objects are the ones responsible of downloading the
177 actual video file and writing it to disk if the user has requested
178 it, among some other tasks. In most cases there should be one per
179 program. As, given a video URL, the downloader doesn't know how to
180 extract all the needed information, task that InfoExtractors do, it
181 has to pass the URL to one of them.
182
183 For this, YoutubeDL objects have a method that allows
184 InfoExtractors to be registered in a given order. When it is passed
185 a URL, the YoutubeDL object handles it to the first InfoExtractor it
186 finds that reports being able to handle it. The InfoExtractor extracts
187 all the information about the video or videos the URL refers to, and
188 YoutubeDL process the extracted information, possibly using a File
189 Downloader to download the video.
190
191 YoutubeDL objects accept a lot of parameters. In order not to saturate
192 the object constructor with arguments, it receives a dictionary of
193 options instead. These options are available through the params
194 attribute for the InfoExtractors to use. The YoutubeDL also
195 registers itself as the downloader in charge for the InfoExtractors
196 that are added to it, so this is a "mutual registration".
197
198 Available options:
199
200 username: Username for authentication purposes.
201 password: Password for authentication purposes.
202 videopassword: Password for accessing a video.
203 ap_mso: Adobe Pass multiple-system operator identifier.
204 ap_username: Multiple-system operator account username.
205 ap_password: Multiple-system operator account password.
206 usenetrc: Use netrc for authentication instead.
207 netrc_location: Location of the netrc file. Defaults to ~/.netrc.
208 netrc_cmd: Use a shell command to get credentials
209 verbose: Print additional info to stdout.
210 quiet: Do not print messages to stdout.
211 no_warnings: Do not print out anything for warnings.
212 forceprint: A dict with keys WHEN mapped to a list of templates to
213 print to stdout. The allowed keys are video or any of the
214 items in utils.POSTPROCESS_WHEN.
215 For compatibility, a single list is also accepted
216 print_to_file: A dict with keys WHEN (same as forceprint) mapped to
217 a list of tuples with (template, filename)
218 forcejson: Force printing info_dict as JSON.
219 dump_single_json: Force printing the info_dict of the whole playlist
220 (or video) as a single JSON line.
221 force_write_download_archive: Force writing download archive regardless
222 of 'skip_download' or 'simulate'.
223 simulate: Do not download the video files. If unset (or None),
224 simulate only if listsubtitles, listformats or list_thumbnails is used
225 format: Video format code. see "FORMAT SELECTION" for more details.
226 You can also pass a function. The function takes 'ctx' as
227 argument and returns the formats to download.
228 See "build_format_selector" for an implementation
229 allow_unplayable_formats: Allow unplayable formats to be extracted and downloaded.
230 ignore_no_formats_error: Ignore "No video formats" error. Usefull for
231 extracting metadata even if the video is not actually
232 available for download (experimental)
233 format_sort: A list of fields by which to sort the video formats.
234 See "Sorting Formats" for more details.
235 format_sort_force: Force the given format_sort. see "Sorting Formats"
236 for more details.
237 prefer_free_formats: Whether to prefer video formats with free containers
238 over non-free ones of same quality.
239 allow_multiple_video_streams: Allow multiple video streams to be merged
240 into a single file
241 allow_multiple_audio_streams: Allow multiple audio streams to be merged
242 into a single file
243 check_formats Whether to test if the formats are downloadable.
244 Can be True (check all), False (check none),
245 'selected' (check selected formats),
246 or None (check only if requested by extractor)
247 paths: Dictionary of output paths. The allowed keys are 'home'
248 'temp' and the keys of OUTTMPL_TYPES (in utils/_utils.py)
249 outtmpl: Dictionary of templates for output names. Allowed keys
250 are 'default' and the keys of OUTTMPL_TYPES (in utils/_utils.py).
251 For compatibility with youtube-dl, a single string can also be used
252 outtmpl_na_placeholder: Placeholder for unavailable meta fields.
253 restrictfilenames: Do not allow "&" and spaces in file names
254 trim_file_name: Limit length of filename (extension excluded)
255 windowsfilenames: Force the filenames to be windows compatible
256 ignoreerrors: Do not stop on download/postprocessing errors.
257 Can be 'only_download' to ignore only download errors.
258 Default is 'only_download' for CLI, but False for API
259 skip_playlist_after_errors: Number of allowed failures until the rest of
260 the playlist is skipped
261 allowed_extractors: List of regexes to match against extractor names that are allowed
262 overwrites: Overwrite all video and metadata files if True,
263 overwrite only non-video files if None
264 and don't overwrite any file if False
265 playlist_items: Specific indices of playlist to download.
266 playlistrandom: Download playlist items in random order.
267 lazy_playlist: Process playlist entries as they are received.
268 matchtitle: Download only matching titles.
269 rejecttitle: Reject downloads for matching titles.
270 logger: Log messages to a logging.Logger instance.
271 logtostderr: Print everything to stderr instead of stdout.
272 consoletitle: Display progress in console window's titlebar.
273 writedescription: Write the video description to a .description file
274 writeinfojson: Write the video description to a .info.json file
275 clean_infojson: Remove internal metadata from the infojson
276 getcomments: Extract video comments. This will not be written to disk
277 unless writeinfojson is also given
278 writeannotations: Write the video annotations to a .annotations.xml file
279 writethumbnail: Write the thumbnail image to a file
280 allow_playlist_files: Whether to write playlists' description, infojson etc
281 also to disk when using the 'write*' options
282 write_all_thumbnails: Write all thumbnail formats to files
283 writelink: Write an internet shortcut file, depending on the
284 current platform (.url/.webloc/.desktop)
285 writeurllink: Write a Windows internet shortcut file (.url)
286 writewebloclink: Write a macOS internet shortcut file (.webloc)
287 writedesktoplink: Write a Linux internet shortcut file (.desktop)
288 writesubtitles: Write the video subtitles to a file
289 writeautomaticsub: Write the automatically generated subtitles to a file
290 listsubtitles: Lists all available subtitles for the video
291 subtitlesformat: The format code for subtitles
292 subtitleslangs: List of languages of the subtitles to download (can be regex).
293 The list may contain "all" to refer to all the available
294 subtitles. The language can be prefixed with a "-" to
295 exclude it from the requested languages, e.g. ['all', '-live_chat']
296 keepvideo: Keep the video file after post-processing
297 daterange: A utils.DateRange object, download only if the upload_date is in the range.
298 skip_download: Skip the actual download of the video file
299 cachedir: Location of the cache files in the filesystem.
300 False to disable filesystem cache.
301 noplaylist: Download single video instead of a playlist if in doubt.
302 age_limit: An integer representing the user's age in years.
303 Unsuitable videos for the given age are skipped.
304 min_views: An integer representing the minimum view count the video
305 must have in order to not be skipped.
306 Videos without view count information are always
307 downloaded. None for no limit.
308 max_views: An integer representing the maximum view count.
309 Videos that are more popular than that are not
310 downloaded.
311 Videos without view count information are always
312 downloaded. None for no limit.
313 download_archive: A set, or the name of a file where all downloads are recorded.
314 Videos already present in the file are not downloaded again.
315 break_on_existing: Stop the download process after attempting to download a
316 file that is in the archive.
317 break_per_url: Whether break_on_reject and break_on_existing
318 should act on each input URL as opposed to for the entire queue
319 cookiefile: File name or text stream from where cookies should be read and dumped to
320 cookiesfrombrowser: A tuple containing the name of the browser, the profile
321 name/path from where cookies are loaded, the name of the keyring,
322 and the container name, e.g. ('chrome', ) or
323 ('vivaldi', 'default', 'BASICTEXT') or ('firefox', 'default', None, 'Meta')
324 legacyserverconnect: Explicitly allow HTTPS connection to servers that do not
325 support RFC 5746 secure renegotiation
326 nocheckcertificate: Do not verify SSL certificates
327 client_certificate: Path to client certificate file in PEM format. May include the private key
328 client_certificate_key: Path to private key file for client certificate
329 client_certificate_password: Password for client certificate private key, if encrypted.
330 If not provided and the key is encrypted, yt-dlp will ask interactively
331 prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
332 (Only supported by some extractors)
333 enable_file_urls: Enable file:// URLs. This is disabled by default for security reasons.
334 http_headers: A dictionary of custom headers to be used for all requests
335 proxy: URL of the proxy server to use
336 geo_verification_proxy: URL of the proxy to use for IP address verification
337 on geo-restricted sites.
338 socket_timeout: Time to wait for unresponsive hosts, in seconds
339 bidi_workaround: Work around buggy terminals without bidirectional text
340 support, using fridibi
341 debug_printtraffic:Print out sent and received HTTP traffic
342 default_search: Prepend this string if an input url is not valid.
343 'auto' for elaborate guessing
344 encoding: Use this encoding instead of the system-specified.
345 extract_flat: Whether to resolve and process url_results further
346 * False: Always process. Default for API
347 * True: Never process
348 * 'in_playlist': Do not process inside playlist/multi_video
349 * 'discard': Always process, but don't return the result
350 from inside playlist/multi_video
351 * 'discard_in_playlist': Same as "discard", but only for
352 playlists (not multi_video). Default for CLI
353 wait_for_video: If given, wait for scheduled streams to become available.
354 The value should be a tuple containing the range
355 (min_secs, max_secs) to wait between retries
356 postprocessors: A list of dictionaries, each with an entry
357 * key: The name of the postprocessor. See
358 yt_dlp/postprocessor/__init__.py for a list.
359 * when: When to run the postprocessor. Allowed values are
360 the entries of utils.POSTPROCESS_WHEN
361 Assumed to be 'post_process' if not given
362 progress_hooks: A list of functions that get called on download
363 progress, with a dictionary with the entries
364 * status: One of "downloading", "error", or "finished".
365 Check this first and ignore unknown values.
366 * info_dict: The extracted info_dict
367
368 If status is one of "downloading", or "finished", the
369 following properties may also be present:
370 * filename: The final filename (always present)
371 * tmpfilename: The filename we're currently writing to
372 * downloaded_bytes: Bytes on disk
373 * total_bytes: Size of the whole file, None if unknown
374 * total_bytes_estimate: Guess of the eventual file size,
375 None if unavailable.
376 * elapsed: The number of seconds since download started.
377 * eta: The estimated time in seconds, None if unknown
378 * speed: The download speed in bytes/second, None if
379 unknown
380 * fragment_index: The counter of the currently
381 downloaded video fragment.
382 * fragment_count: The number of fragments (= individual
383 files that will be merged)
384
385 Progress hooks are guaranteed to be called at least once
386 (with status "finished") if the download is successful.
387 postprocessor_hooks: A list of functions that get called on postprocessing
388 progress, with a dictionary with the entries
389 * status: One of "started", "processing", or "finished".
390 Check this first and ignore unknown values.
391 * postprocessor: Name of the postprocessor
392 * info_dict: The extracted info_dict
393
394 Progress hooks are guaranteed to be called at least twice
395 (with status "started" and "finished") if the processing is successful.
396 merge_output_format: "/" separated list of extensions to use when merging formats.
397 final_ext: Expected final extension; used to detect when the file was
398 already downloaded and converted
399 fixup: Automatically correct known faults of the file.
400 One of:
401 - "never": do nothing
402 - "warn": only emit a warning
403 - "detect_or_warn": check whether we can do anything
404 about it, warn otherwise (default)
405 source_address: Client-side IP address to bind to.
406 sleep_interval_requests: Number of seconds to sleep between requests
407 during extraction
408 sleep_interval: Number of seconds to sleep before each download when
409 used alone or a lower bound of a range for randomized
410 sleep before each download (minimum possible number
411 of seconds to sleep) when used along with
412 max_sleep_interval.
413 max_sleep_interval:Upper bound of a range for randomized sleep before each
414 download (maximum possible number of seconds to sleep).
415 Must only be used along with sleep_interval.
416 Actual sleep time will be a random float from range
417 [sleep_interval; max_sleep_interval].
418 sleep_interval_subtitles: Number of seconds to sleep before each subtitle download
419 listformats: Print an overview of available video formats and exit.
420 list_thumbnails: Print a table of all thumbnails and exit.
421 match_filter: A function that gets called for every video with the signature
422 (info_dict, *, incomplete: bool) -> Optional[str]
423 For backward compatibility with youtube-dl, the signature
424 (info_dict) -> Optional[str] is also allowed.
425 - If it returns a message, the video is ignored.
426 - If it returns None, the video is downloaded.
427 - If it returns utils.NO_DEFAULT, the user is interactively
428 asked whether to download the video.
429 - Raise utils.DownloadCancelled(msg) to abort remaining
430 downloads when a video is rejected.
431 match_filter_func in utils/_utils.py is one example for this.
432 color: A Dictionary with output stream names as keys
433 and their respective color policy as values.
434 Can also just be a single color policy,
435 in which case it applies to all outputs.
436 Valid stream names are 'stdout' and 'stderr'.
437 Valid color policies are one of 'always', 'auto', 'no_color' or 'never'.
438 geo_bypass: Bypass geographic restriction via faking X-Forwarded-For
439 HTTP header
440 geo_bypass_country:
441 Two-letter ISO 3166-2 country code that will be used for
442 explicit geographic restriction bypassing via faking
443 X-Forwarded-For HTTP header
444 geo_bypass_ip_block:
445 IP range in CIDR notation that will be used similarly to
446 geo_bypass_country
447 external_downloader: A dictionary of protocol keys and the executable of the
448 external downloader to use for it. The allowed protocols
449 are default|http|ftp|m3u8|dash|rtsp|rtmp|mms.
450 Set the value to 'native' to use the native downloader
451 compat_opts: Compatibility options. See "Differences in default behavior".
452 The following options do not work when used through the API:
453 filename, abort-on-error, multistreams, no-live-chat, format-sort
454 no-clean-infojson, no-playlist-metafiles, no-keep-subs, no-attach-info-json.
455 Refer __init__.py for their implementation
456 progress_template: Dictionary of templates for progress outputs.
457 Allowed keys are 'download', 'postprocess',
458 'download-title' (console title) and 'postprocess-title'.
459 The template is mapped on a dictionary with keys 'progress' and 'info'
460 retry_sleep_functions: Dictionary of functions that takes the number of attempts
461 as argument and returns the time to sleep in seconds.
462 Allowed keys are 'http', 'fragment', 'file_access'
463 download_ranges: A callback function that gets called for every video with
464 the signature (info_dict, ydl) -> Iterable[Section].
465 Only the returned sections will be downloaded.
466 Each Section is a dict with the following keys:
467 * start_time: Start time of the section in seconds
468 * end_time: End time of the section in seconds
469 * title: Section title (Optional)
470 * index: Section number (Optional)
471 force_keyframes_at_cuts: Re-encode the video when downloading ranges to get precise cuts
472 noprogress: Do not print the progress bar
473 live_from_start: Whether to download livestreams videos from the start
474
475 The following parameters are not used by YoutubeDL itself, they are used by
476 the downloader (see yt_dlp/downloader/common.py):
477 nopart, updatetime, buffersize, ratelimit, throttledratelimit, min_filesize,
478 max_filesize, test, noresizebuffer, retries, file_access_retries, fragment_retries,
479 continuedl, xattr_set_filesize, hls_use_mpegts, http_chunk_size,
480 external_downloader_args, concurrent_fragment_downloads.
481
482 The following options are used by the post processors:
483 ffmpeg_location: Location of the ffmpeg/avconv binary; either the path
484 to the binary or its containing directory.
485 postprocessor_args: A dictionary of postprocessor/executable keys (in lower case)
486 and a list of additional command-line arguments for the
487 postprocessor/executable. The dict can also have "PP+EXE" keys
488 which are used when the given exe is used by the given PP.
489 Use 'default' as the name for arguments to passed to all PP
490 For compatibility with youtube-dl, a single list of args
491 can also be used
492
493 The following options are used by the extractors:
494 extractor_retries: Number of times to retry for known errors (default: 3)
495 dynamic_mpd: Whether to process dynamic DASH manifests (default: True)
496 hls_split_discontinuity: Split HLS playlists to different formats at
497 discontinuities such as ad breaks (default: False)
498 extractor_args: A dictionary of arguments to be passed to the extractors.
499 See "EXTRACTOR ARGUMENTS" for details.
500 E.g. {'youtube': {'skip': ['dash', 'hls']}}
501 mark_watched: Mark videos watched (even with --simulate). Only for YouTube
502
503 The following options are deprecated and may be removed in the future:
504
505 break_on_reject: Stop the download process when encountering a video that
506 has been filtered out.
507 - `raise DownloadCancelled(msg)` in match_filter instead
508 force_generic_extractor: Force downloader to use the generic extractor
509 - Use allowed_extractors = ['generic', 'default']
510 playliststart: - Use playlist_items
511 Playlist item to start at.
512 playlistend: - Use playlist_items
513 Playlist item to end at.
514 playlistreverse: - Use playlist_items
515 Download playlist items in reverse order.
516 forceurl: - Use forceprint
517 Force printing final URL.
518 forcetitle: - Use forceprint
519 Force printing title.
520 forceid: - Use forceprint
521 Force printing ID.
522 forcethumbnail: - Use forceprint
523 Force printing thumbnail URL.
524 forcedescription: - Use forceprint
525 Force printing description.
526 forcefilename: - Use forceprint
527 Force printing final filename.
528 forceduration: - Use forceprint
529 Force printing duration.
530 allsubtitles: - Use subtitleslangs = ['all']
531 Downloads all the subtitles of the video
532 (requires writesubtitles or writeautomaticsub)
533 include_ads: - Doesn't work
534 Download ads as well
535 call_home: - Not implemented
536 Boolean, true iff we are allowed to contact the
537 yt-dlp servers for debugging.
538 post_hooks: - Register a custom postprocessor
539 A list of functions that get called as the final step
540 for each video file, after all postprocessors have been
541 called. The filename will be passed as the only argument.
542 hls_prefer_native: - Use external_downloader = {'m3u8': 'native'} or {'m3u8': 'ffmpeg'}.
543 Use the native HLS downloader instead of ffmpeg/avconv
544 if True, otherwise use ffmpeg/avconv if False, otherwise
545 use downloader suggested by extractor if None.
546 prefer_ffmpeg: - avconv support is deprecated
547 If False, use avconv instead of ffmpeg if both are available,
548 otherwise prefer ffmpeg.
549 youtube_include_dash_manifest: - Use extractor_args
550 If True (default), DASH manifests and related
551 data will be downloaded and processed by extractor.
552 You can reduce network I/O by disabling it if you don't
553 care about DASH. (only for youtube)
554 youtube_include_hls_manifest: - Use extractor_args
555 If True (default), HLS manifests and related
556 data will be downloaded and processed by extractor.
557 You can reduce network I/O by disabling it if you don't
558 care about HLS. (only for youtube)
559 no_color: Same as `color='no_color'`
560 no_overwrites: Same as `overwrites=False`
561 """
562
563 _NUMERIC_FIELDS = {
564 'width', 'height', 'asr', 'audio_channels', 'fps',
565 'tbr', 'abr', 'vbr', 'filesize', 'filesize_approx',
566 'timestamp', 'release_timestamp',
567 'duration', 'view_count', 'like_count', 'dislike_count', 'repost_count',
568 'average_rating', 'comment_count', 'age_limit',
569 'start_time', 'end_time',
570 'chapter_number', 'season_number', 'episode_number',
571 'track_number', 'disc_number', 'release_year',
572 }
573
574 _format_fields = {
575 # NB: Keep in sync with the docstring of extractor/common.py
576 'url', 'manifest_url', 'manifest_stream_number', 'ext', 'format', 'format_id', 'format_note',
577 'width', 'height', 'aspect_ratio', 'resolution', 'dynamic_range', 'tbr', 'abr', 'acodec', 'asr', 'audio_channels',
578 'vbr', 'fps', 'vcodec', 'container', 'filesize', 'filesize_approx', 'rows', 'columns',
579 'player_url', 'protocol', 'fragment_base_url', 'fragments', 'is_from_start',
580 'preference', 'language', 'language_preference', 'quality', 'source_preference', 'cookies',
581 'http_headers', 'stretched_ratio', 'no_resume', 'has_drm', 'extra_param_to_segment_url', 'hls_aes', 'downloader_options',
582 'page_url', 'app', 'play_path', 'tc_url', 'flash_version', 'rtmp_live', 'rtmp_conn', 'rtmp_protocol', 'rtmp_real_time'
583 }
584 _format_selection_exts = {
585 'audio': set(MEDIA_EXTENSIONS.common_audio),
586 'video': set(MEDIA_EXTENSIONS.common_video + ('3gp', )),
587 'storyboards': set(MEDIA_EXTENSIONS.storyboards),
588 }
589
590 def __init__(self, params=None, auto_init=True):
591 """Create a FileDownloader object with the given options.
592 @param auto_init Whether to load the default extractors and print header (if verbose).
593 Set to 'no_verbose_header' to not print the header
594 """
595 if params is None:
596 params = {}
597 self.params = params
598 self._ies = {}
599 self._ies_instances = {}
600 self._pps = {k: [] for k in POSTPROCESS_WHEN}
601 self._printed_messages = set()
602 self._first_webpage_request = True
603 self._post_hooks = []
604 self._progress_hooks = []
605 self._postprocessor_hooks = []
606 self._download_retcode = 0
607 self._num_downloads = 0
608 self._num_videos = 0
609 self._playlist_level = 0
610 self._playlist_urls = set()
611 self.cache = Cache(self)
612 self.__header_cookies = []
613
614 stdout = sys.stderr if self.params.get('logtostderr') else sys.stdout
615 self._out_files = Namespace(
616 out=stdout,
617 error=sys.stderr,
618 screen=sys.stderr if self.params.get('quiet') else stdout,
619 console=None if compat_os_name == 'nt' else next(
620 filter(supports_terminal_sequences, (sys.stderr, sys.stdout)), None)
621 )
622
623 try:
624 windows_enable_vt_mode()
625 except Exception as e:
626 self.write_debug(f'Failed to enable VT mode: {e}')
627
628 if self.params.get('no_color'):
629 if self.params.get('color') is not None:
630 self.params.setdefault('_warnings', []).append(
631 'Overwriting params from "color" with "no_color"')
632 self.params['color'] = 'no_color'
633
634 term_allow_color = os.getenv('TERM', '').lower() != 'dumb'
635 no_color = bool(os.getenv('NO_COLOR'))
636
637 def process_color_policy(stream):
638 stream_name = {sys.stdout: 'stdout', sys.stderr: 'stderr'}[stream]
639 policy = traverse_obj(self.params, ('color', (stream_name, None), {str}), get_all=False)
640 if policy in ('auto', None):
641 if term_allow_color and supports_terminal_sequences(stream):
642 return 'no_color' if no_color else True
643 return False
644 assert policy in ('always', 'never', 'no_color'), policy
645 return {'always': True, 'never': False}.get(policy, policy)
646
647 self._allow_colors = Namespace(**{
648 name: process_color_policy(stream)
649 for name, stream in self._out_files.items_ if name != 'console'
650 })
651
652 system_deprecation = _get_system_deprecation()
653 if system_deprecation:
654 self.deprecated_feature(system_deprecation.replace('\n', '\n '))
655
656 if self.params.get('allow_unplayable_formats'):
657 self.report_warning(
658 f'You have asked for {self._format_err("UNPLAYABLE", self.Styles.EMPHASIS)} formats to be listed/downloaded. '
659 'This is a developer option intended for debugging. \n'
660 ' If you experience any issues while using this option, '
661 f'{self._format_err("DO NOT", self.Styles.ERROR)} open a bug report')
662
663 if self.params.get('bidi_workaround', False):
664 try:
665 import pty
666 master, slave = pty.openpty()
667 width = shutil.get_terminal_size().columns
668 width_args = [] if width is None else ['-w', str(width)]
669 sp_kwargs = {'stdin': subprocess.PIPE, 'stdout': slave, 'stderr': self._out_files.error}
670 try:
671 self._output_process = Popen(['bidiv'] + width_args, **sp_kwargs)
672 except OSError:
673 self._output_process = Popen(['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
674 self._output_channel = os.fdopen(master, 'rb')
675 except OSError as ose:
676 if ose.errno == errno.ENOENT:
677 self.report_warning(
678 'Could not find fribidi executable, ignoring --bidi-workaround. '
679 'Make sure that fribidi is an executable file in one of the directories in your $PATH.')
680 else:
681 raise
682
683 self.params['compat_opts'] = set(self.params.get('compat_opts', ()))
684 self.params['http_headers'] = HTTPHeaderDict(std_headers, self.params.get('http_headers'))
685 self._load_cookies(self.params['http_headers'].get('Cookie')) # compat
686 self.params['http_headers'].pop('Cookie', None)
687 self._request_director = self.build_request_director(_REQUEST_HANDLERS.values(), _RH_PREFERENCES)
688
689 if auto_init and auto_init != 'no_verbose_header':
690 self.print_debug_header()
691
692 def check_deprecated(param, option, suggestion):
693 if self.params.get(param) is not None:
694 self.report_warning(f'{option} is deprecated. Use {suggestion} instead')
695 return True
696 return False
697
698 if check_deprecated('cn_verification_proxy', '--cn-verification-proxy', '--geo-verification-proxy'):
699 if self.params.get('geo_verification_proxy') is None:
700 self.params['geo_verification_proxy'] = self.params['cn_verification_proxy']
701
702 check_deprecated('autonumber', '--auto-number', '-o "%(autonumber)s-%(title)s.%(ext)s"')
703 check_deprecated('usetitle', '--title', '-o "%(title)s-%(id)s.%(ext)s"')
704 check_deprecated('useid', '--id', '-o "%(id)s.%(ext)s"')
705
706 for msg in self.params.get('_warnings', []):
707 self.report_warning(msg)
708 for msg in self.params.get('_deprecation_warnings', []):
709 self.deprecated_feature(msg)
710
711 if 'list-formats' in self.params['compat_opts']:
712 self.params['listformats_table'] = False
713
714 if 'overwrites' not in self.params and self.params.get('nooverwrites') is not None:
715 # nooverwrites was unnecessarily changed to overwrites
716 # in 0c3d0f51778b153f65c21906031c2e091fcfb641
717 # This ensures compatibility with both keys
718 self.params['overwrites'] = not self.params['nooverwrites']
719 elif self.params.get('overwrites') is None:
720 self.params.pop('overwrites', None)
721 else:
722 self.params['nooverwrites'] = not self.params['overwrites']
723
724 if self.params.get('simulate') is None and any((
725 self.params.get('list_thumbnails'),
726 self.params.get('listformats'),
727 self.params.get('listsubtitles'),
728 )):
729 self.params['simulate'] = 'list_only'
730
731 self.params.setdefault('forceprint', {})
732 self.params.setdefault('print_to_file', {})
733
734 # Compatibility with older syntax
735 if not isinstance(params['forceprint'], dict):
736 self.params['forceprint'] = {'video': params['forceprint']}
737
738 if auto_init:
739 self.add_default_info_extractors()
740
741 if (sys.platform != 'win32'
742 and sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968']
743 and not self.params.get('restrictfilenames', False)):
744 # Unicode filesystem API will throw errors (#1474, #13027)
745 self.report_warning(
746 'Assuming --restrict-filenames since file system encoding '
747 'cannot encode all characters. '
748 'Set the LC_ALL environment variable to fix this.')
749 self.params['restrictfilenames'] = True
750
751 self._parse_outtmpl()
752
753 # Creating format selector here allows us to catch syntax errors before the extraction
754 self.format_selector = (
755 self.params.get('format') if self.params.get('format') in (None, '-')
756 else self.params['format'] if callable(self.params['format'])
757 else self.build_format_selector(self.params['format']))
758
759 hooks = {
760 'post_hooks': self.add_post_hook,
761 'progress_hooks': self.add_progress_hook,
762 'postprocessor_hooks': self.add_postprocessor_hook,
763 }
764 for opt, fn in hooks.items():
765 for ph in self.params.get(opt, []):
766 fn(ph)
767
768 for pp_def_raw in self.params.get('postprocessors', []):
769 pp_def = dict(pp_def_raw)
770 when = pp_def.pop('when', 'post_process')
771 self.add_post_processor(
772 get_postprocessor(pp_def.pop('key'))(self, **pp_def),
773 when=when)
774
775 def preload_download_archive(fn):
776 """Preload the archive, if any is specified"""
777 archive = set()
778 if fn is None:
779 return archive
780 elif not is_path_like(fn):
781 return fn
782
783 self.write_debug(f'Loading archive file {fn!r}')
784 try:
785 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
786 for line in archive_file:
787 archive.add(line.strip())
788 except OSError as ioe:
789 if ioe.errno != errno.ENOENT:
790 raise
791 return archive
792
793 self.archive = preload_download_archive(self.params.get('download_archive'))
794
795 def warn_if_short_id(self, argv):
796 # short YouTube ID starting with dash?
797 idxs = [
798 i for i, a in enumerate(argv)
799 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
800 if idxs:
801 correct_argv = (
802 ['yt-dlp']
803 + [a for i, a in enumerate(argv) if i not in idxs]
804 + ['--'] + [argv[i] for i in idxs]
805 )
806 self.report_warning(
807 'Long argument string detected. '
808 'Use -- to separate parameters and URLs, like this:\n%s' %
809 args_to_str(correct_argv))
810
811 def add_info_extractor(self, ie):
812 """Add an InfoExtractor object to the end of the list."""
813 ie_key = ie.ie_key()
814 self._ies[ie_key] = ie
815 if not isinstance(ie, type):
816 self._ies_instances[ie_key] = ie
817 ie.set_downloader(self)
818
819 def get_info_extractor(self, ie_key):
820 """
821 Get an instance of an IE with name ie_key, it will try to get one from
822 the _ies list, if there's no instance it will create a new one and add
823 it to the extractor list.
824 """
825 ie = self._ies_instances.get(ie_key)
826 if ie is None:
827 ie = get_info_extractor(ie_key)()
828 self.add_info_extractor(ie)
829 return ie
830
831 def add_default_info_extractors(self):
832 """
833 Add the InfoExtractors returned by gen_extractors to the end of the list
834 """
835 all_ies = {ie.IE_NAME.lower(): ie for ie in gen_extractor_classes()}
836 all_ies['end'] = UnsupportedURLIE()
837 try:
838 ie_names = orderedSet_from_options(
839 self.params.get('allowed_extractors', ['default']), {
840 'all': list(all_ies),
841 'default': [name for name, ie in all_ies.items() if ie._ENABLED],
842 }, use_regex=True)
843 except re.error as e:
844 raise ValueError(f'Wrong regex for allowed_extractors: {e.pattern}')
845 for name in ie_names:
846 self.add_info_extractor(all_ies[name])
847 self.write_debug(f'Loaded {len(ie_names)} extractors')
848
849 def add_post_processor(self, pp, when='post_process'):
850 """Add a PostProcessor object to the end of the chain."""
851 assert when in POSTPROCESS_WHEN, f'Invalid when={when}'
852 self._pps[when].append(pp)
853 pp.set_downloader(self)
854
855 def add_post_hook(self, ph):
856 """Add the post hook"""
857 self._post_hooks.append(ph)
858
859 def add_progress_hook(self, ph):
860 """Add the download progress hook"""
861 self._progress_hooks.append(ph)
862
863 def add_postprocessor_hook(self, ph):
864 """Add the postprocessing progress hook"""
865 self._postprocessor_hooks.append(ph)
866 for pps in self._pps.values():
867 for pp in pps:
868 pp.add_progress_hook(ph)
869
870 def _bidi_workaround(self, message):
871 if not hasattr(self, '_output_channel'):
872 return message
873
874 assert hasattr(self, '_output_process')
875 assert isinstance(message, str)
876 line_count = message.count('\n') + 1
877 self._output_process.stdin.write((message + '\n').encode())
878 self._output_process.stdin.flush()
879 res = ''.join(self._output_channel.readline().decode()
880 for _ in range(line_count))
881 return res[:-len('\n')]
882
883 def _write_string(self, message, out=None, only_once=False):
884 if only_once:
885 if message in self._printed_messages:
886 return
887 self._printed_messages.add(message)
888 write_string(message, out=out, encoding=self.params.get('encoding'))
889
890 def to_stdout(self, message, skip_eol=False, quiet=None):
891 """Print message to stdout"""
892 if quiet is not None:
893 self.deprecation_warning('"YoutubeDL.to_stdout" no longer accepts the argument quiet. '
894 'Use "YoutubeDL.to_screen" instead')
895 if skip_eol is not False:
896 self.deprecation_warning('"YoutubeDL.to_stdout" no longer accepts the argument skip_eol. '
897 'Use "YoutubeDL.to_screen" instead')
898 self._write_string(f'{self._bidi_workaround(message)}\n', self._out_files.out)
899
900 def to_screen(self, message, skip_eol=False, quiet=None, only_once=False):
901 """Print message to screen if not in quiet mode"""
902 if self.params.get('logger'):
903 self.params['logger'].debug(message)
904 return
905 if (self.params.get('quiet') if quiet is None else quiet) and not self.params.get('verbose'):
906 return
907 self._write_string(
908 '%s%s' % (self._bidi_workaround(message), ('' if skip_eol else '\n')),
909 self._out_files.screen, only_once=only_once)
910
911 def to_stderr(self, message, only_once=False):
912 """Print message to stderr"""
913 assert isinstance(message, str)
914 if self.params.get('logger'):
915 self.params['logger'].error(message)
916 else:
917 self._write_string(f'{self._bidi_workaround(message)}\n', self._out_files.error, only_once=only_once)
918
919 def _send_console_code(self, code):
920 if compat_os_name == 'nt' or not self._out_files.console:
921 return
922 self._write_string(code, self._out_files.console)
923
924 def to_console_title(self, message):
925 if not self.params.get('consoletitle', False):
926 return
927 message = remove_terminal_sequences(message)
928 if compat_os_name == 'nt':
929 if ctypes.windll.kernel32.GetConsoleWindow():
930 # c_wchar_p() might not be necessary if `message` is
931 # already of type unicode()
932 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
933 else:
934 self._send_console_code(f'\033]0;{message}\007')
935
936 def save_console_title(self):
937 if not self.params.get('consoletitle') or self.params.get('simulate'):
938 return
939 self._send_console_code('\033[22;0t') # Save the title on stack
940
941 def restore_console_title(self):
942 if not self.params.get('consoletitle') or self.params.get('simulate'):
943 return
944 self._send_console_code('\033[23;0t') # Restore the title from stack
945
946 def __enter__(self):
947 self.save_console_title()
948 return self
949
950 def save_cookies(self):
951 if self.params.get('cookiefile') is not None:
952 self.cookiejar.save()
953
954 def __exit__(self, *args):
955 self.restore_console_title()
956 self.close()
957
958 def close(self):
959 self.save_cookies()
960 self._request_director.close()
961
962 def trouble(self, message=None, tb=None, is_error=True):
963 """Determine action to take when a download problem appears.
964
965 Depending on if the downloader has been configured to ignore
966 download errors or not, this method may throw an exception or
967 not when errors are found, after printing the message.
968
969 @param tb If given, is additional traceback information
970 @param is_error Whether to raise error according to ignorerrors
971 """
972 if message is not None:
973 self.to_stderr(message)
974 if self.params.get('verbose'):
975 if tb is None:
976 if sys.exc_info()[0]: # if .trouble has been called from an except block
977 tb = ''
978 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
979 tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
980 tb += encode_compat_str(traceback.format_exc())
981 else:
982 tb_data = traceback.format_list(traceback.extract_stack())
983 tb = ''.join(tb_data)
984 if tb:
985 self.to_stderr(tb)
986 if not is_error:
987 return
988 if not self.params.get('ignoreerrors'):
989 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
990 exc_info = sys.exc_info()[1].exc_info
991 else:
992 exc_info = sys.exc_info()
993 raise DownloadError(message, exc_info)
994 self._download_retcode = 1
995
996 Styles = Namespace(
997 HEADERS='yellow',
998 EMPHASIS='light blue',
999 FILENAME='green',
1000 ID='green',
1001 DELIM='blue',
1002 ERROR='red',
1003 BAD_FORMAT='light red',
1004 WARNING='yellow',
1005 SUPPRESS='light black',
1006 )
1007
1008 def _format_text(self, handle, allow_colors, text, f, fallback=None, *, test_encoding=False):
1009 text = str(text)
1010 if test_encoding:
1011 original_text = text
1012 # handle.encoding can be None. See https://github.com/yt-dlp/yt-dlp/issues/2711
1013 encoding = self.params.get('encoding') or getattr(handle, 'encoding', None) or 'ascii'
1014 text = text.encode(encoding, 'ignore').decode(encoding)
1015 if fallback is not None and text != original_text:
1016 text = fallback
1017 return format_text(text, f) if allow_colors is True else text if fallback is None else fallback
1018
1019 def _format_out(self, *args, **kwargs):
1020 return self._format_text(self._out_files.out, self._allow_colors.out, *args, **kwargs)
1021
1022 def _format_screen(self, *args, **kwargs):
1023 return self._format_text(self._out_files.screen, self._allow_colors.screen, *args, **kwargs)
1024
1025 def _format_err(self, *args, **kwargs):
1026 return self._format_text(self._out_files.error, self._allow_colors.error, *args, **kwargs)
1027
1028 def report_warning(self, message, only_once=False):
1029 '''
1030 Print the message to stderr, it will be prefixed with 'WARNING:'
1031 If stderr is a tty file the 'WARNING:' will be colored
1032 '''
1033 if self.params.get('logger') is not None:
1034 self.params['logger'].warning(message)
1035 else:
1036 if self.params.get('no_warnings'):
1037 return
1038 self.to_stderr(f'{self._format_err("WARNING:", self.Styles.WARNING)} {message}', only_once)
1039
1040 def deprecation_warning(self, message, *, stacklevel=0):
1041 deprecation_warning(
1042 message, stacklevel=stacklevel + 1, printer=self.report_error, is_error=False)
1043
1044 def deprecated_feature(self, message):
1045 if self.params.get('logger') is not None:
1046 self.params['logger'].warning(f'Deprecated Feature: {message}')
1047 self.to_stderr(f'{self._format_err("Deprecated Feature:", self.Styles.ERROR)} {message}', True)
1048
1049 def report_error(self, message, *args, **kwargs):
1050 '''
1051 Do the same as trouble, but prefixes the message with 'ERROR:', colored
1052 in red if stderr is a tty file.
1053 '''
1054 self.trouble(f'{self._format_err("ERROR:", self.Styles.ERROR)} {message}', *args, **kwargs)
1055
1056 def write_debug(self, message, only_once=False):
1057 '''Log debug message or Print message to stderr'''
1058 if not self.params.get('verbose', False):
1059 return
1060 message = f'[debug] {message}'
1061 if self.params.get('logger'):
1062 self.params['logger'].debug(message)
1063 else:
1064 self.to_stderr(message, only_once)
1065
1066 def report_file_already_downloaded(self, file_name):
1067 """Report file has already been fully downloaded."""
1068 try:
1069 self.to_screen('[download] %s has already been downloaded' % file_name)
1070 except UnicodeEncodeError:
1071 self.to_screen('[download] The file has already been downloaded')
1072
1073 def report_file_delete(self, file_name):
1074 """Report that existing file will be deleted."""
1075 try:
1076 self.to_screen('Deleting existing file %s' % file_name)
1077 except UnicodeEncodeError:
1078 self.to_screen('Deleting existing file')
1079
1080 def raise_no_formats(self, info, forced=False, *, msg=None):
1081 has_drm = info.get('_has_drm')
1082 ignored, expected = self.params.get('ignore_no_formats_error'), bool(msg)
1083 msg = msg or has_drm and 'This video is DRM protected' or 'No video formats found!'
1084 if forced or not ignored:
1085 raise ExtractorError(msg, video_id=info['id'], ie=info['extractor'],
1086 expected=has_drm or ignored or expected)
1087 else:
1088 self.report_warning(msg)
1089
1090 def parse_outtmpl(self):
1091 self.deprecation_warning('"YoutubeDL.parse_outtmpl" is deprecated and may be removed in a future version')
1092 self._parse_outtmpl()
1093 return self.params['outtmpl']
1094
1095 def _parse_outtmpl(self):
1096 sanitize = IDENTITY
1097 if self.params.get('restrictfilenames'): # Remove spaces in the default template
1098 sanitize = lambda x: x.replace(' - ', ' ').replace(' ', '-')
1099
1100 outtmpl = self.params.setdefault('outtmpl', {})
1101 if not isinstance(outtmpl, dict):
1102 self.params['outtmpl'] = outtmpl = {'default': outtmpl}
1103 outtmpl.update({k: sanitize(v) for k, v in DEFAULT_OUTTMPL.items() if outtmpl.get(k) is None})
1104
1105 def get_output_path(self, dir_type='', filename=None):
1106 paths = self.params.get('paths', {})
1107 assert isinstance(paths, dict), '"paths" parameter must be a dictionary'
1108 path = os.path.join(
1109 expand_path(paths.get('home', '').strip()),
1110 expand_path(paths.get(dir_type, '').strip()) if dir_type else '',
1111 filename or '')
1112 return sanitize_path(path, force=self.params.get('windowsfilenames'))
1113
1114 @staticmethod
1115 def _outtmpl_expandpath(outtmpl):
1116 # expand_path translates '%%' into '%' and '$$' into '$'
1117 # correspondingly that is not what we want since we need to keep
1118 # '%%' intact for template dict substitution step. Working around
1119 # with boundary-alike separator hack.
1120 sep = ''.join(random.choices(string.ascii_letters, k=32))
1121 outtmpl = outtmpl.replace('%%', f'%{sep}%').replace('$$', f'${sep}$')
1122
1123 # outtmpl should be expand_path'ed before template dict substitution
1124 # because meta fields may contain env variables we don't want to
1125 # be expanded. E.g. for outtmpl "%(title)s.%(ext)s" and
1126 # title "Hello $PATH", we don't want `$PATH` to be expanded.
1127 return expand_path(outtmpl).replace(sep, '')
1128
1129 @staticmethod
1130 def escape_outtmpl(outtmpl):
1131 ''' Escape any remaining strings like %s, %abc% etc. '''
1132 return re.sub(
1133 STR_FORMAT_RE_TMPL.format('', '(?![%(\0])'),
1134 lambda mobj: ('' if mobj.group('has_key') else '%') + mobj.group(0),
1135 outtmpl)
1136
1137 @classmethod
1138 def validate_outtmpl(cls, outtmpl):
1139 ''' @return None or Exception object '''
1140 outtmpl = re.sub(
1141 STR_FORMAT_RE_TMPL.format('[^)]*', '[ljhqBUDS]'),
1142 lambda mobj: f'{mobj.group(0)[:-1]}s',
1143 cls._outtmpl_expandpath(outtmpl))
1144 try:
1145 cls.escape_outtmpl(outtmpl) % collections.defaultdict(int)
1146 return None
1147 except ValueError as err:
1148 return err
1149
1150 @staticmethod
1151 def _copy_infodict(info_dict):
1152 info_dict = dict(info_dict)
1153 info_dict.pop('__postprocessors', None)
1154 info_dict.pop('__pending_error', None)
1155 return info_dict
1156
1157 def prepare_outtmpl(self, outtmpl, info_dict, sanitize=False):
1158 """ Make the outtmpl and info_dict suitable for substitution: ydl.escape_outtmpl(outtmpl) % info_dict
1159 @param sanitize Whether to sanitize the output as a filename.
1160 For backward compatibility, a function can also be passed
1161 """
1162
1163 info_dict.setdefault('epoch', int(time.time())) # keep epoch consistent once set
1164
1165 info_dict = self._copy_infodict(info_dict)
1166 info_dict['duration_string'] = ( # %(duration>%H-%M-%S)s is wrong if duration > 24hrs
1167 formatSeconds(info_dict['duration'], '-' if sanitize else ':')
1168 if info_dict.get('duration', None) is not None
1169 else None)
1170 info_dict['autonumber'] = int(self.params.get('autonumber_start', 1) - 1 + self._num_downloads)
1171 info_dict['video_autonumber'] = self._num_videos
1172 if info_dict.get('resolution') is None:
1173 info_dict['resolution'] = self.format_resolution(info_dict, default=None)
1174
1175 # For fields playlist_index, playlist_autonumber and autonumber convert all occurrences
1176 # of %(field)s to %(field)0Nd for backward compatibility
1177 field_size_compat_map = {
1178 'playlist_index': number_of_digits(info_dict.get('__last_playlist_index') or 0),
1179 'playlist_autonumber': number_of_digits(info_dict.get('n_entries') or 0),
1180 'autonumber': self.params.get('autonumber_size') or 5,
1181 }
1182
1183 TMPL_DICT = {}
1184 EXTERNAL_FORMAT_RE = re.compile(STR_FORMAT_RE_TMPL.format('[^)]*', f'[{STR_FORMAT_TYPES}ljhqBUDS]'))
1185 MATH_FUNCTIONS = {
1186 '+': float.__add__,
1187 '-': float.__sub__,
1188 '*': float.__mul__,
1189 }
1190 # Field is of the form key1.key2...
1191 # where keys (except first) can be string, int, slice or "{field, ...}"
1192 FIELD_INNER_RE = r'(?:\w+|%(num)s|%(num)s?(?::%(num)s?){1,2})' % {'num': r'(?:-?\d+)'}
1193 FIELD_RE = r'\w*(?:\.(?:%(inner)s|{%(field)s(?:,%(field)s)*}))*' % {
1194 'inner': FIELD_INNER_RE,
1195 'field': rf'\w*(?:\.{FIELD_INNER_RE})*'
1196 }
1197 MATH_FIELD_RE = rf'(?:{FIELD_RE}|-?{NUMBER_RE})'
1198 MATH_OPERATORS_RE = r'(?:%s)' % '|'.join(map(re.escape, MATH_FUNCTIONS.keys()))
1199 INTERNAL_FORMAT_RE = re.compile(rf'''(?xs)
1200 (?P<negate>-)?
1201 (?P<fields>{FIELD_RE})
1202 (?P<maths>(?:{MATH_OPERATORS_RE}{MATH_FIELD_RE})*)
1203 (?:>(?P<strf_format>.+?))?
1204 (?P<remaining>
1205 (?P<alternate>(?<!\\),[^|&)]+)?
1206 (?:&(?P<replacement>.*?))?
1207 (?:\|(?P<default>.*?))?
1208 )$''')
1209
1210 def _from_user_input(field):
1211 if field == ':':
1212 return ...
1213 elif ':' in field:
1214 return slice(*map(int_or_none, field.split(':')))
1215 elif int_or_none(field) is not None:
1216 return int(field)
1217 return field
1218
1219 def _traverse_infodict(fields):
1220 fields = [f for x in re.split(r'\.({.+?})\.?', fields)
1221 for f in ([x] if x.startswith('{') else x.split('.'))]
1222 for i in (0, -1):
1223 if fields and not fields[i]:
1224 fields.pop(i)
1225
1226 for i, f in enumerate(fields):
1227 if not f.startswith('{'):
1228 fields[i] = _from_user_input(f)
1229 continue
1230 assert f.endswith('}'), f'No closing brace for {f} in {fields}'
1231 fields[i] = {k: list(map(_from_user_input, k.split('.'))) for k in f[1:-1].split(',')}
1232
1233 return traverse_obj(info_dict, fields, traverse_string=True)
1234
1235 def get_value(mdict):
1236 # Object traversal
1237 value = _traverse_infodict(mdict['fields'])
1238 # Negative
1239 if mdict['negate']:
1240 value = float_or_none(value)
1241 if value is not None:
1242 value *= -1
1243 # Do maths
1244 offset_key = mdict['maths']
1245 if offset_key:
1246 value = float_or_none(value)
1247 operator = None
1248 while offset_key:
1249 item = re.match(
1250 MATH_FIELD_RE if operator else MATH_OPERATORS_RE,
1251 offset_key).group(0)
1252 offset_key = offset_key[len(item):]
1253 if operator is None:
1254 operator = MATH_FUNCTIONS[item]
1255 continue
1256 item, multiplier = (item[1:], -1) if item[0] == '-' else (item, 1)
1257 offset = float_or_none(item)
1258 if offset is None:
1259 offset = float_or_none(_traverse_infodict(item))
1260 try:
1261 value = operator(value, multiplier * offset)
1262 except (TypeError, ZeroDivisionError):
1263 return None
1264 operator = None
1265 # Datetime formatting
1266 if mdict['strf_format']:
1267 value = strftime_or_none(value, mdict['strf_format'].replace('\\,', ','))
1268
1269 # XXX: Workaround for https://github.com/yt-dlp/yt-dlp/issues/4485
1270 if sanitize and value == '':
1271 value = None
1272 return value
1273
1274 na = self.params.get('outtmpl_na_placeholder', 'NA')
1275
1276 def filename_sanitizer(key, value, restricted=self.params.get('restrictfilenames')):
1277 return sanitize_filename(str(value), restricted=restricted, is_id=(
1278 bool(re.search(r'(^|[_.])id(\.|$)', key))
1279 if 'filename-sanitization' in self.params['compat_opts']
1280 else NO_DEFAULT))
1281
1282 sanitizer = sanitize if callable(sanitize) else filename_sanitizer
1283 sanitize = bool(sanitize)
1284
1285 def _dumpjson_default(obj):
1286 if isinstance(obj, (set, LazyList)):
1287 return list(obj)
1288 return repr(obj)
1289
1290 class _ReplacementFormatter(string.Formatter):
1291 def get_field(self, field_name, args, kwargs):
1292 if field_name.isdigit():
1293 return args[0], -1
1294 raise ValueError('Unsupported field')
1295
1296 replacement_formatter = _ReplacementFormatter()
1297
1298 def create_key(outer_mobj):
1299 if not outer_mobj.group('has_key'):
1300 return outer_mobj.group(0)
1301 key = outer_mobj.group('key')
1302 mobj = re.match(INTERNAL_FORMAT_RE, key)
1303 value, replacement, default, last_field = None, None, na, ''
1304 while mobj:
1305 mobj = mobj.groupdict()
1306 default = mobj['default'] if mobj['default'] is not None else default
1307 value = get_value(mobj)
1308 last_field, replacement = mobj['fields'], mobj['replacement']
1309 if value is None and mobj['alternate']:
1310 mobj = re.match(INTERNAL_FORMAT_RE, mobj['remaining'][1:])
1311 else:
1312 break
1313
1314 if None not in (value, replacement):
1315 try:
1316 value = replacement_formatter.format(replacement, value)
1317 except ValueError:
1318 value, default = None, na
1319
1320 fmt = outer_mobj.group('format')
1321 if fmt == 's' and last_field in field_size_compat_map.keys() and isinstance(value, int):
1322 fmt = f'0{field_size_compat_map[last_field]:d}d'
1323
1324 flags = outer_mobj.group('conversion') or ''
1325 str_fmt = f'{fmt[:-1]}s'
1326 if value is None:
1327 value, fmt = default, 's'
1328 elif fmt[-1] == 'l': # list
1329 delim = '\n' if '#' in flags else ', '
1330 value, fmt = delim.join(map(str, variadic(value, allowed_types=(str, bytes)))), str_fmt
1331 elif fmt[-1] == 'j': # json
1332 value, fmt = json.dumps(
1333 value, default=_dumpjson_default,
1334 indent=4 if '#' in flags else None, ensure_ascii='+' not in flags), str_fmt
1335 elif fmt[-1] == 'h': # html
1336 value, fmt = escapeHTML(str(value)), str_fmt
1337 elif fmt[-1] == 'q': # quoted
1338 value = map(str, variadic(value) if '#' in flags else [value])
1339 value, fmt = ' '.join(map(compat_shlex_quote, value)), str_fmt
1340 elif fmt[-1] == 'B': # bytes
1341 value = f'%{str_fmt}'.encode() % str(value).encode()
1342 value, fmt = value.decode('utf-8', 'ignore'), 's'
1343 elif fmt[-1] == 'U': # unicode normalized
1344 value, fmt = unicodedata.normalize(
1345 # "+" = compatibility equivalence, "#" = NFD
1346 'NF%s%s' % ('K' if '+' in flags else '', 'D' if '#' in flags else 'C'),
1347 value), str_fmt
1348 elif fmt[-1] == 'D': # decimal suffix
1349 num_fmt, fmt = fmt[:-1].replace('#', ''), 's'
1350 value = format_decimal_suffix(value, f'%{num_fmt}f%s' if num_fmt else '%d%s',
1351 factor=1024 if '#' in flags else 1000)
1352 elif fmt[-1] == 'S': # filename sanitization
1353 value, fmt = filename_sanitizer(last_field, value, restricted='#' in flags), str_fmt
1354 elif fmt[-1] == 'c':
1355 if value:
1356 value = str(value)[0]
1357 else:
1358 fmt = str_fmt
1359 elif fmt[-1] not in 'rsa': # numeric
1360 value = float_or_none(value)
1361 if value is None:
1362 value, fmt = default, 's'
1363
1364 if sanitize:
1365 # If value is an object, sanitize might convert it to a string
1366 # So we convert it to repr first
1367 if fmt[-1] == 'r':
1368 value, fmt = repr(value), str_fmt
1369 elif fmt[-1] == 'a':
1370 value, fmt = ascii(value), str_fmt
1371 if fmt[-1] in 'csra':
1372 value = sanitizer(last_field, value)
1373
1374 key = '%s\0%s' % (key.replace('%', '%\0'), outer_mobj.group('format'))
1375 TMPL_DICT[key] = value
1376 return '{prefix}%({key}){fmt}'.format(key=key, fmt=fmt, prefix=outer_mobj.group('prefix'))
1377
1378 return EXTERNAL_FORMAT_RE.sub(create_key, outtmpl), TMPL_DICT
1379
1380 def evaluate_outtmpl(self, outtmpl, info_dict, *args, **kwargs):
1381 outtmpl, info_dict = self.prepare_outtmpl(outtmpl, info_dict, *args, **kwargs)
1382 return self.escape_outtmpl(outtmpl) % info_dict
1383
1384 def _prepare_filename(self, info_dict, *, outtmpl=None, tmpl_type=None):
1385 assert None in (outtmpl, tmpl_type), 'outtmpl and tmpl_type are mutually exclusive'
1386 if outtmpl is None:
1387 outtmpl = self.params['outtmpl'].get(tmpl_type or 'default', self.params['outtmpl']['default'])
1388 try:
1389 outtmpl = self._outtmpl_expandpath(outtmpl)
1390 filename = self.evaluate_outtmpl(outtmpl, info_dict, True)
1391 if not filename:
1392 return None
1393
1394 if tmpl_type in ('', 'temp'):
1395 final_ext, ext = self.params.get('final_ext'), info_dict.get('ext')
1396 if final_ext and ext and final_ext != ext and filename.endswith(f'.{final_ext}'):
1397 filename = replace_extension(filename, ext, final_ext)
1398 elif tmpl_type:
1399 force_ext = OUTTMPL_TYPES[tmpl_type]
1400 if force_ext:
1401 filename = replace_extension(filename, force_ext, info_dict.get('ext'))
1402
1403 # https://github.com/blackjack4494/youtube-dlc/issues/85
1404 trim_file_name = self.params.get('trim_file_name', False)
1405 if trim_file_name:
1406 no_ext, *ext = filename.rsplit('.', 2)
1407 filename = join_nonempty(no_ext[:trim_file_name], *ext, delim='.')
1408
1409 return filename
1410 except ValueError as err:
1411 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
1412 return None
1413
1414 def prepare_filename(self, info_dict, dir_type='', *, outtmpl=None, warn=False):
1415 """Generate the output filename"""
1416 if outtmpl:
1417 assert not dir_type, 'outtmpl and dir_type are mutually exclusive'
1418 dir_type = None
1419 filename = self._prepare_filename(info_dict, tmpl_type=dir_type, outtmpl=outtmpl)
1420 if not filename and dir_type not in ('', 'temp'):
1421 return ''
1422
1423 if warn:
1424 if not self.params.get('paths'):
1425 pass
1426 elif filename == '-':
1427 self.report_warning('--paths is ignored when an outputting to stdout', only_once=True)
1428 elif os.path.isabs(filename):
1429 self.report_warning('--paths is ignored since an absolute path is given in output template', only_once=True)
1430 if filename == '-' or not filename:
1431 return filename
1432
1433 return self.get_output_path(dir_type, filename)
1434
1435 def _match_entry(self, info_dict, incomplete=False, silent=False):
1436 """Returns None if the file should be downloaded"""
1437 _type = 'video' if 'playlist-match-filter' in self.params['compat_opts'] else info_dict.get('_type', 'video')
1438 assert incomplete or _type == 'video', 'Only video result can be considered complete'
1439
1440 video_title = info_dict.get('title', info_dict.get('id', 'entry'))
1441
1442 def check_filter():
1443 if _type in ('playlist', 'multi_video'):
1444 return
1445 elif _type in ('url', 'url_transparent') and not try_call(
1446 lambda: self.get_info_extractor(info_dict['ie_key']).is_single_video(info_dict['url'])):
1447 return
1448
1449 if 'title' in info_dict:
1450 # This can happen when we're just evaluating the playlist
1451 title = info_dict['title']
1452 matchtitle = self.params.get('matchtitle', False)
1453 if matchtitle:
1454 if not re.search(matchtitle, title, re.IGNORECASE):
1455 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
1456 rejecttitle = self.params.get('rejecttitle', False)
1457 if rejecttitle:
1458 if re.search(rejecttitle, title, re.IGNORECASE):
1459 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
1460
1461 date = info_dict.get('upload_date')
1462 if date is not None:
1463 dateRange = self.params.get('daterange', DateRange())
1464 if date not in dateRange:
1465 return f'{date_from_str(date).isoformat()} upload date is not in range {dateRange}'
1466 view_count = info_dict.get('view_count')
1467 if view_count is not None:
1468 min_views = self.params.get('min_views')
1469 if min_views is not None and view_count < min_views:
1470 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
1471 max_views = self.params.get('max_views')
1472 if max_views is not None and view_count > max_views:
1473 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
1474 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
1475 return 'Skipping "%s" because it is age restricted' % video_title
1476
1477 match_filter = self.params.get('match_filter')
1478 if match_filter is None:
1479 return None
1480
1481 cancelled = None
1482 try:
1483 try:
1484 ret = match_filter(info_dict, incomplete=incomplete)
1485 except TypeError:
1486 # For backward compatibility
1487 ret = None if incomplete else match_filter(info_dict)
1488 except DownloadCancelled as err:
1489 if err.msg is not NO_DEFAULT:
1490 raise
1491 ret, cancelled = err.msg, err
1492
1493 if ret is NO_DEFAULT:
1494 while True:
1495 filename = self._format_screen(self.prepare_filename(info_dict), self.Styles.FILENAME)
1496 reply = input(self._format_screen(
1497 f'Download "{filename}"? (Y/n): ', self.Styles.EMPHASIS)).lower().strip()
1498 if reply in {'y', ''}:
1499 return None
1500 elif reply == 'n':
1501 if cancelled:
1502 raise type(cancelled)(f'Skipping {video_title}')
1503 return f'Skipping {video_title}'
1504 return ret
1505
1506 if self.in_download_archive(info_dict):
1507 reason = ''.join((
1508 format_field(info_dict, 'id', f'{self._format_screen("%s", self.Styles.ID)}: '),
1509 format_field(info_dict, 'title', f'{self._format_screen("%s", self.Styles.EMPHASIS)} '),
1510 'has already been recorded in the archive'))
1511 break_opt, break_err = 'break_on_existing', ExistingVideoReached
1512 else:
1513 try:
1514 reason = check_filter()
1515 except DownloadCancelled as e:
1516 reason, break_opt, break_err = e.msg, 'match_filter', type(e)
1517 else:
1518 break_opt, break_err = 'break_on_reject', RejectedVideoReached
1519 if reason is not None:
1520 if not silent:
1521 self.to_screen('[download] ' + reason)
1522 if self.params.get(break_opt, False):
1523 raise break_err()
1524 return reason
1525
1526 @staticmethod
1527 def add_extra_info(info_dict, extra_info):
1528 '''Set the keys from extra_info in info dict if they are missing'''
1529 for key, value in extra_info.items():
1530 info_dict.setdefault(key, value)
1531
1532 def extract_info(self, url, download=True, ie_key=None, extra_info=None,
1533 process=True, force_generic_extractor=False):
1534 """
1535 Extract and return the information dictionary of the URL
1536
1537 Arguments:
1538 @param url URL to extract
1539
1540 Keyword arguments:
1541 @param download Whether to download videos
1542 @param process Whether to resolve all unresolved references (URLs, playlist items).
1543 Must be True for download to work
1544 @param ie_key Use only the extractor with this key
1545
1546 @param extra_info Dictionary containing the extra values to add to the info (For internal use only)
1547 @force_generic_extractor Force using the generic extractor (Deprecated; use ie_key='Generic')
1548 """
1549
1550 if extra_info is None:
1551 extra_info = {}
1552
1553 if not ie_key and force_generic_extractor:
1554 ie_key = 'Generic'
1555
1556 if ie_key:
1557 ies = {ie_key: self._ies[ie_key]} if ie_key in self._ies else {}
1558 else:
1559 ies = self._ies
1560
1561 for key, ie in ies.items():
1562 if not ie.suitable(url):
1563 continue
1564
1565 if not ie.working():
1566 self.report_warning('The program functionality for this site has been marked as broken, '
1567 'and will probably not work.')
1568
1569 temp_id = ie.get_temp_id(url)
1570 if temp_id is not None and self.in_download_archive({'id': temp_id, 'ie_key': key}):
1571 self.to_screen(f'[download] {self._format_screen(temp_id, self.Styles.ID)}: '
1572 'has already been recorded in the archive')
1573 if self.params.get('break_on_existing', False):
1574 raise ExistingVideoReached()
1575 break
1576 return self.__extract_info(url, self.get_info_extractor(key), download, extra_info, process)
1577 else:
1578 extractors_restricted = self.params.get('allowed_extractors') not in (None, ['default'])
1579 self.report_error(f'No suitable extractor{format_field(ie_key, None, " (%s)")} found for URL {url}',
1580 tb=False if extractors_restricted else None)
1581
1582 def _handle_extraction_exceptions(func):
1583 @functools.wraps(func)
1584 def wrapper(self, *args, **kwargs):
1585 while True:
1586 try:
1587 return func(self, *args, **kwargs)
1588 except (DownloadCancelled, LazyList.IndexError, PagedList.IndexError):
1589 raise
1590 except ReExtractInfo as e:
1591 if e.expected:
1592 self.to_screen(f'{e}; Re-extracting data')
1593 else:
1594 self.to_stderr('\r')
1595 self.report_warning(f'{e}; Re-extracting data')
1596 continue
1597 except GeoRestrictedError as e:
1598 msg = e.msg
1599 if e.countries:
1600 msg += '\nThis video is available in %s.' % ', '.join(
1601 map(ISO3166Utils.short2full, e.countries))
1602 msg += '\nYou might want to use a VPN or a proxy server (with --proxy) to workaround.'
1603 self.report_error(msg)
1604 except ExtractorError as e: # An error we somewhat expected
1605 self.report_error(str(e), e.format_traceback())
1606 except Exception as e:
1607 if self.params.get('ignoreerrors'):
1608 self.report_error(str(e), tb=encode_compat_str(traceback.format_exc()))
1609 else:
1610 raise
1611 break
1612 return wrapper
1613
1614 def _wait_for_video(self, ie_result={}):
1615 if (not self.params.get('wait_for_video')
1616 or ie_result.get('_type', 'video') != 'video'
1617 or ie_result.get('formats') or ie_result.get('url')):
1618 return
1619
1620 format_dur = lambda dur: '%02d:%02d:%02d' % timetuple_from_msec(dur * 1000)[:-1]
1621 last_msg = ''
1622
1623 def progress(msg):
1624 nonlocal last_msg
1625 full_msg = f'{msg}\n'
1626 if not self.params.get('noprogress'):
1627 full_msg = msg + ' ' * (len(last_msg) - len(msg)) + '\r'
1628 elif last_msg:
1629 return
1630 self.to_screen(full_msg, skip_eol=True)
1631 last_msg = msg
1632
1633 min_wait, max_wait = self.params.get('wait_for_video')
1634 diff = try_get(ie_result, lambda x: x['release_timestamp'] - time.time())
1635 if diff is None and ie_result.get('live_status') == 'is_upcoming':
1636 diff = round(random.uniform(min_wait, max_wait) if (max_wait and min_wait) else (max_wait or min_wait), 0)
1637 self.report_warning('Release time of video is not known')
1638 elif ie_result and (diff or 0) <= 0:
1639 self.report_warning('Video should already be available according to extracted info')
1640 diff = min(max(diff or 0, min_wait or 0), max_wait or float('inf'))
1641 self.to_screen(f'[wait] Waiting for {format_dur(diff)} - Press Ctrl+C to try now')
1642
1643 wait_till = time.time() + diff
1644 try:
1645 while True:
1646 diff = wait_till - time.time()
1647 if diff <= 0:
1648 progress('')
1649 raise ReExtractInfo('[wait] Wait period ended', expected=True)
1650 progress(f'[wait] Remaining time until next attempt: {self._format_screen(format_dur(diff), self.Styles.EMPHASIS)}')
1651 time.sleep(1)
1652 except KeyboardInterrupt:
1653 progress('')
1654 raise ReExtractInfo('[wait] Interrupted by user', expected=True)
1655 except BaseException as e:
1656 if not isinstance(e, ReExtractInfo):
1657 self.to_screen('')
1658 raise
1659
1660 def _load_cookies(self, data, *, autoscope=True):
1661 """Loads cookies from a `Cookie` header
1662
1663 This tries to work around the security vulnerability of passing cookies to every domain.
1664 See: https://github.com/yt-dlp/yt-dlp/security/advisories/GHSA-v8mc-9377-rwjj
1665
1666 @param data The Cookie header as string to load the cookies from
1667 @param autoscope If `False`, scope cookies using Set-Cookie syntax and error for cookie without domains
1668 If `True`, save cookies for later to be stored in the jar with a limited scope
1669 If a URL, save cookies in the jar with the domain of the URL
1670 """
1671 for cookie in LenientSimpleCookie(data).values():
1672 if autoscope and any(cookie.values()):
1673 raise ValueError('Invalid syntax in Cookie Header')
1674
1675 domain = cookie.get('domain') or ''
1676 expiry = cookie.get('expires')
1677 if expiry == '': # 0 is valid
1678 expiry = None
1679 prepared_cookie = http.cookiejar.Cookie(
1680 cookie.get('version') or 0, cookie.key, cookie.value, None, False,
1681 domain, True, True, cookie.get('path') or '', bool(cookie.get('path')),
1682 cookie.get('secure') or False, expiry, False, None, None, {})
1683
1684 if domain:
1685 self.cookiejar.set_cookie(prepared_cookie)
1686 elif autoscope is True:
1687 self.deprecated_feature(
1688 'Passing cookies as a header is a potential security risk; '
1689 'they will be scoped to the domain of the downloaded urls. '
1690 'Please consider loading cookies from a file or browser instead.')
1691 self.__header_cookies.append(prepared_cookie)
1692 elif autoscope:
1693 self.report_warning(
1694 'The extractor result contains an unscoped cookie as an HTTP header. '
1695 f'If you are using yt-dlp with an input URL{bug_reports_message(before=",")}',
1696 only_once=True)
1697 self._apply_header_cookies(autoscope, [prepared_cookie])
1698 else:
1699 self.report_error('Unscoped cookies are not allowed; please specify some sort of scoping',
1700 tb=False, is_error=False)
1701
1702 def _apply_header_cookies(self, url, cookies=None):
1703 """Applies stray header cookies to the provided url
1704
1705 This loads header cookies and scopes them to the domain provided in `url`.
1706 While this is not ideal, it helps reduce the risk of them being sent
1707 to an unintended destination while mostly maintaining compatibility.
1708 """
1709 parsed = urllib.parse.urlparse(url)
1710 if not parsed.hostname:
1711 return
1712
1713 for cookie in map(copy.copy, cookies or self.__header_cookies):
1714 cookie.domain = f'.{parsed.hostname}'
1715 self.cookiejar.set_cookie(cookie)
1716
1717 @_handle_extraction_exceptions
1718 def __extract_info(self, url, ie, download, extra_info, process):
1719 self._apply_header_cookies(url)
1720
1721 try:
1722 ie_result = ie.extract(url)
1723 except UserNotLive as e:
1724 if process:
1725 if self.params.get('wait_for_video'):
1726 self.report_warning(e)
1727 self._wait_for_video()
1728 raise
1729 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
1730 self.report_warning(f'Extractor {ie.IE_NAME} returned nothing{bug_reports_message()}')
1731 return
1732 if isinstance(ie_result, list):
1733 # Backwards compatibility: old IE result format
1734 ie_result = {
1735 '_type': 'compat_list',
1736 'entries': ie_result,
1737 }
1738 if extra_info.get('original_url'):
1739 ie_result.setdefault('original_url', extra_info['original_url'])
1740 self.add_default_extra_info(ie_result, ie, url)
1741 if process:
1742 self._wait_for_video(ie_result)
1743 return self.process_ie_result(ie_result, download, extra_info)
1744 else:
1745 return ie_result
1746
1747 def add_default_extra_info(self, ie_result, ie, url):
1748 if url is not None:
1749 self.add_extra_info(ie_result, {
1750 'webpage_url': url,
1751 'original_url': url,
1752 })
1753 webpage_url = ie_result.get('webpage_url')
1754 if webpage_url:
1755 self.add_extra_info(ie_result, {
1756 'webpage_url_basename': url_basename(webpage_url),
1757 'webpage_url_domain': get_domain(webpage_url),
1758 })
1759 if ie is not None:
1760 self.add_extra_info(ie_result, {
1761 'extractor': ie.IE_NAME,
1762 'extractor_key': ie.ie_key(),
1763 })
1764
1765 def process_ie_result(self, ie_result, download=True, extra_info=None):
1766 """
1767 Take the result of the ie(may be modified) and resolve all unresolved
1768 references (URLs, playlist items).
1769
1770 It will also download the videos if 'download'.
1771 Returns the resolved ie_result.
1772 """
1773 if extra_info is None:
1774 extra_info = {}
1775 result_type = ie_result.get('_type', 'video')
1776
1777 if result_type in ('url', 'url_transparent'):
1778 ie_result['url'] = sanitize_url(
1779 ie_result['url'], scheme='http' if self.params.get('prefer_insecure') else 'https')
1780 if ie_result.get('original_url') and not extra_info.get('original_url'):
1781 extra_info = {'original_url': ie_result['original_url'], **extra_info}
1782
1783 extract_flat = self.params.get('extract_flat', False)
1784 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info)
1785 or extract_flat is True):
1786 info_copy = ie_result.copy()
1787 ie = try_get(ie_result.get('ie_key'), self.get_info_extractor)
1788 if ie and not ie_result.get('id'):
1789 info_copy['id'] = ie.get_temp_id(ie_result['url'])
1790 self.add_default_extra_info(info_copy, ie, ie_result['url'])
1791 self.add_extra_info(info_copy, extra_info)
1792 info_copy, _ = self.pre_process(info_copy)
1793 self._fill_common_fields(info_copy, False)
1794 self.__forced_printings(info_copy)
1795 self._raise_pending_errors(info_copy)
1796 if self.params.get('force_write_download_archive', False):
1797 self.record_download_archive(info_copy)
1798 return ie_result
1799
1800 if result_type == 'video':
1801 self.add_extra_info(ie_result, extra_info)
1802 ie_result = self.process_video_result(ie_result, download=download)
1803 self._raise_pending_errors(ie_result)
1804 additional_urls = (ie_result or {}).get('additional_urls')
1805 if additional_urls:
1806 # TODO: Improve MetadataParserPP to allow setting a list
1807 if isinstance(additional_urls, str):
1808 additional_urls = [additional_urls]
1809 self.to_screen(
1810 '[info] %s: %d additional URL(s) requested' % (ie_result['id'], len(additional_urls)))
1811 self.write_debug('Additional URLs: "%s"' % '", "'.join(additional_urls))
1812 ie_result['additional_entries'] = [
1813 self.extract_info(
1814 url, download, extra_info=extra_info,
1815 force_generic_extractor=self.params.get('force_generic_extractor'))
1816 for url in additional_urls
1817 ]
1818 return ie_result
1819 elif result_type == 'url':
1820 # We have to add extra_info to the results because it may be
1821 # contained in a playlist
1822 return self.extract_info(
1823 ie_result['url'], download,
1824 ie_key=ie_result.get('ie_key'),
1825 extra_info=extra_info)
1826 elif result_type == 'url_transparent':
1827 # Use the information from the embedding page
1828 info = self.extract_info(
1829 ie_result['url'], ie_key=ie_result.get('ie_key'),
1830 extra_info=extra_info, download=False, process=False)
1831
1832 # extract_info may return None when ignoreerrors is enabled and
1833 # extraction failed with an error, don't crash and return early
1834 # in this case
1835 if not info:
1836 return info
1837
1838 exempted_fields = {'_type', 'url', 'ie_key'}
1839 if not ie_result.get('section_end') and ie_result.get('section_start') is None:
1840 # For video clips, the id etc of the clip extractor should be used
1841 exempted_fields |= {'id', 'extractor', 'extractor_key'}
1842
1843 new_result = info.copy()
1844 new_result.update(filter_dict(ie_result, lambda k, v: v is not None and k not in exempted_fields))
1845
1846 # Extracted info may not be a video result (i.e.
1847 # info.get('_type', 'video') != video) but rather an url or
1848 # url_transparent. In such cases outer metadata (from ie_result)
1849 # should be propagated to inner one (info). For this to happen
1850 # _type of info should be overridden with url_transparent. This
1851 # fixes issue from https://github.com/ytdl-org/youtube-dl/pull/11163.
1852 if new_result.get('_type') == 'url':
1853 new_result['_type'] = 'url_transparent'
1854
1855 return self.process_ie_result(
1856 new_result, download=download, extra_info=extra_info)
1857 elif result_type in ('playlist', 'multi_video'):
1858 # Protect from infinite recursion due to recursively nested playlists
1859 # (see https://github.com/ytdl-org/youtube-dl/issues/27833)
1860 webpage_url = ie_result.get('webpage_url') # Playlists maynot have webpage_url
1861 if webpage_url and webpage_url in self._playlist_urls:
1862 self.to_screen(
1863 '[download] Skipping already downloaded playlist: %s'
1864 % ie_result.get('title') or ie_result.get('id'))
1865 return
1866
1867 self._playlist_level += 1
1868 self._playlist_urls.add(webpage_url)
1869 self._fill_common_fields(ie_result, False)
1870 self._sanitize_thumbnails(ie_result)
1871 try:
1872 return self.__process_playlist(ie_result, download)
1873 finally:
1874 self._playlist_level -= 1
1875 if not self._playlist_level:
1876 self._playlist_urls.clear()
1877 elif result_type == 'compat_list':
1878 self.report_warning(
1879 'Extractor %s returned a compat_list result. '
1880 'It needs to be updated.' % ie_result.get('extractor'))
1881
1882 def _fixup(r):
1883 self.add_extra_info(r, {
1884 'extractor': ie_result['extractor'],
1885 'webpage_url': ie_result['webpage_url'],
1886 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1887 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1888 'extractor_key': ie_result['extractor_key'],
1889 })
1890 return r
1891 ie_result['entries'] = [
1892 self.process_ie_result(_fixup(r), download, extra_info)
1893 for r in ie_result['entries']
1894 ]
1895 return ie_result
1896 else:
1897 raise Exception('Invalid result type: %s' % result_type)
1898
1899 def _ensure_dir_exists(self, path):
1900 return make_dir(path, self.report_error)
1901
1902 @staticmethod
1903 def _playlist_infodict(ie_result, strict=False, **kwargs):
1904 info = {
1905 'playlist_count': ie_result.get('playlist_count'),
1906 'playlist': ie_result.get('title') or ie_result.get('id'),
1907 'playlist_id': ie_result.get('id'),
1908 'playlist_title': ie_result.get('title'),
1909 'playlist_uploader': ie_result.get('uploader'),
1910 'playlist_uploader_id': ie_result.get('uploader_id'),
1911 **kwargs,
1912 }
1913 if strict:
1914 return info
1915 if ie_result.get('webpage_url'):
1916 info.update({
1917 'webpage_url': ie_result['webpage_url'],
1918 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1919 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1920 })
1921 return {
1922 **info,
1923 'playlist_index': 0,
1924 '__last_playlist_index': max(ie_result.get('requested_entries') or (0, 0)),
1925 'extractor': ie_result['extractor'],
1926 'extractor_key': ie_result['extractor_key'],
1927 }
1928
1929 def __process_playlist(self, ie_result, download):
1930 """Process each entry in the playlist"""
1931 assert ie_result['_type'] in ('playlist', 'multi_video')
1932
1933 common_info = self._playlist_infodict(ie_result, strict=True)
1934 title = common_info.get('playlist') or '<Untitled>'
1935 if self._match_entry(common_info, incomplete=True) is not None:
1936 return
1937 self.to_screen(f'[download] Downloading {ie_result["_type"]}: {title}')
1938
1939 all_entries = PlaylistEntries(self, ie_result)
1940 entries = orderedSet(all_entries.get_requested_items(), lazy=True)
1941
1942 lazy = self.params.get('lazy_playlist')
1943 if lazy:
1944 resolved_entries, n_entries = [], 'N/A'
1945 ie_result['requested_entries'], ie_result['entries'] = None, None
1946 else:
1947 entries = resolved_entries = list(entries)
1948 n_entries = len(resolved_entries)
1949 ie_result['requested_entries'], ie_result['entries'] = tuple(zip(*resolved_entries)) or ([], [])
1950 if not ie_result.get('playlist_count'):
1951 # Better to do this after potentially exhausting entries
1952 ie_result['playlist_count'] = all_entries.get_full_count()
1953
1954 extra = self._playlist_infodict(ie_result, n_entries=int_or_none(n_entries))
1955 ie_copy = collections.ChainMap(ie_result, extra)
1956
1957 _infojson_written = False
1958 write_playlist_files = self.params.get('allow_playlist_files', True)
1959 if write_playlist_files and self.params.get('list_thumbnails'):
1960 self.list_thumbnails(ie_result)
1961 if write_playlist_files and not self.params.get('simulate'):
1962 _infojson_written = self._write_info_json(
1963 'playlist', ie_result, self.prepare_filename(ie_copy, 'pl_infojson'))
1964 if _infojson_written is None:
1965 return
1966 if self._write_description('playlist', ie_result,
1967 self.prepare_filename(ie_copy, 'pl_description')) is None:
1968 return
1969 # TODO: This should be passed to ThumbnailsConvertor if necessary
1970 self._write_thumbnails('playlist', ie_result, self.prepare_filename(ie_copy, 'pl_thumbnail'))
1971
1972 if lazy:
1973 if self.params.get('playlistreverse') or self.params.get('playlistrandom'):
1974 self.report_warning('playlistreverse and playlistrandom are not supported with lazy_playlist', only_once=True)
1975 elif self.params.get('playlistreverse'):
1976 entries.reverse()
1977 elif self.params.get('playlistrandom'):
1978 random.shuffle(entries)
1979
1980 self.to_screen(f'[{ie_result["extractor"]}] Playlist {title}: Downloading {n_entries} items'
1981 f'{format_field(ie_result, "playlist_count", " of %s")}')
1982
1983 keep_resolved_entries = self.params.get('extract_flat') != 'discard'
1984 if self.params.get('extract_flat') == 'discard_in_playlist':
1985 keep_resolved_entries = ie_result['_type'] != 'playlist'
1986 if keep_resolved_entries:
1987 self.write_debug('The information of all playlist entries will be held in memory')
1988
1989 failures = 0
1990 max_failures = self.params.get('skip_playlist_after_errors') or float('inf')
1991 for i, (playlist_index, entry) in enumerate(entries):
1992 if lazy:
1993 resolved_entries.append((playlist_index, entry))
1994 if not entry:
1995 continue
1996
1997 entry['__x_forwarded_for_ip'] = ie_result.get('__x_forwarded_for_ip')
1998 if not lazy and 'playlist-index' in self.params['compat_opts']:
1999 playlist_index = ie_result['requested_entries'][i]
2000
2001 entry_copy = collections.ChainMap(entry, {
2002 **common_info,
2003 'n_entries': int_or_none(n_entries),
2004 'playlist_index': playlist_index,
2005 'playlist_autonumber': i + 1,
2006 })
2007
2008 if self._match_entry(entry_copy, incomplete=True) is not None:
2009 # For compatabilty with youtube-dl. See https://github.com/yt-dlp/yt-dlp/issues/4369
2010 resolved_entries[i] = (playlist_index, NO_DEFAULT)
2011 continue
2012
2013 self.to_screen('[download] Downloading item %s of %s' % (
2014 self._format_screen(i + 1, self.Styles.ID), self._format_screen(n_entries, self.Styles.EMPHASIS)))
2015
2016 entry_result = self.__process_iterable_entry(entry, download, collections.ChainMap({
2017 'playlist_index': playlist_index,
2018 'playlist_autonumber': i + 1,
2019 }, extra))
2020 if not entry_result:
2021 failures += 1
2022 if failures >= max_failures:
2023 self.report_error(
2024 f'Skipping the remaining entries in playlist "{title}" since {failures} items failed extraction')
2025 break
2026 if keep_resolved_entries:
2027 resolved_entries[i] = (playlist_index, entry_result)
2028
2029 # Update with processed data
2030 ie_result['entries'] = [e for _, e in resolved_entries if e is not NO_DEFAULT]
2031 ie_result['requested_entries'] = [i for i, e in resolved_entries if e is not NO_DEFAULT]
2032 if ie_result['requested_entries'] == try_call(lambda: list(range(1, ie_result['playlist_count'] + 1))):
2033 # Do not set for full playlist
2034 ie_result.pop('requested_entries')
2035
2036 # Write the updated info to json
2037 if _infojson_written is True and self._write_info_json(
2038 'updated playlist', ie_result,
2039 self.prepare_filename(ie_copy, 'pl_infojson'), overwrite=True) is None:
2040 return
2041
2042 ie_result = self.run_all_pps('playlist', ie_result)
2043 self.to_screen(f'[download] Finished downloading playlist: {title}')
2044 return ie_result
2045
2046 @_handle_extraction_exceptions
2047 def __process_iterable_entry(self, entry, download, extra_info):
2048 return self.process_ie_result(
2049 entry, download=download, extra_info=extra_info)
2050
2051 def _build_format_filter(self, filter_spec):
2052 " Returns a function to filter the formats according to the filter_spec "
2053
2054 OPERATORS = {
2055 '<': operator.lt,
2056 '<=': operator.le,
2057 '>': operator.gt,
2058 '>=': operator.ge,
2059 '=': operator.eq,
2060 '!=': operator.ne,
2061 }
2062 operator_rex = re.compile(r'''(?x)\s*
2063 (?P<key>[\w.-]+)\s*
2064 (?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
2065 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)\s*
2066 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
2067 m = operator_rex.fullmatch(filter_spec)
2068 if m:
2069 try:
2070 comparison_value = int(m.group('value'))
2071 except ValueError:
2072 comparison_value = parse_filesize(m.group('value'))
2073 if comparison_value is None:
2074 comparison_value = parse_filesize(m.group('value') + 'B')
2075 if comparison_value is None:
2076 raise ValueError(
2077 'Invalid value %r in format specification %r' % (
2078 m.group('value'), filter_spec))
2079 op = OPERATORS[m.group('op')]
2080
2081 if not m:
2082 STR_OPERATORS = {
2083 '=': operator.eq,
2084 '^=': lambda attr, value: attr.startswith(value),
2085 '$=': lambda attr, value: attr.endswith(value),
2086 '*=': lambda attr, value: value in attr,
2087 '~=': lambda attr, value: value.search(attr) is not None
2088 }
2089 str_operator_rex = re.compile(r'''(?x)\s*
2090 (?P<key>[a-zA-Z0-9._-]+)\s*
2091 (?P<negation>!\s*)?(?P<op>%s)\s*(?P<none_inclusive>\?\s*)?
2092 (?P<quote>["'])?
2093 (?P<value>(?(quote)(?:(?!(?P=quote))[^\\]|\\.)+|[\w.-]+))
2094 (?(quote)(?P=quote))\s*
2095 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
2096 m = str_operator_rex.fullmatch(filter_spec)
2097 if m:
2098 if m.group('op') == '~=':
2099 comparison_value = re.compile(m.group('value'))
2100 else:
2101 comparison_value = re.sub(r'''\\([\\"'])''', r'\1', m.group('value'))
2102 str_op = STR_OPERATORS[m.group('op')]
2103 if m.group('negation'):
2104 op = lambda attr, value: not str_op(attr, value)
2105 else:
2106 op = str_op
2107
2108 if not m:
2109 raise SyntaxError('Invalid filter specification %r' % filter_spec)
2110
2111 def _filter(f):
2112 actual_value = f.get(m.group('key'))
2113 if actual_value is None:
2114 return m.group('none_inclusive')
2115 return op(actual_value, comparison_value)
2116 return _filter
2117
2118 def _check_formats(self, formats):
2119 for f in formats:
2120 self.to_screen('[info] Testing format %s' % f['format_id'])
2121 path = self.get_output_path('temp')
2122 if not self._ensure_dir_exists(f'{path}/'):
2123 continue
2124 temp_file = tempfile.NamedTemporaryFile(suffix='.tmp', delete=False, dir=path or None)
2125 temp_file.close()
2126 try:
2127 success, _ = self.dl(temp_file.name, f, test=True)
2128 except (DownloadError, OSError, ValueError) + network_exceptions:
2129 success = False
2130 finally:
2131 if os.path.exists(temp_file.name):
2132 try:
2133 os.remove(temp_file.name)
2134 except OSError:
2135 self.report_warning('Unable to delete temporary file "%s"' % temp_file.name)
2136 if success:
2137 yield f
2138 else:
2139 self.to_screen('[info] Unable to download format %s. Skipping...' % f['format_id'])
2140
2141 def _default_format_spec(self, info_dict, download=True):
2142
2143 def can_merge():
2144 merger = FFmpegMergerPP(self)
2145 return merger.available and merger.can_merge()
2146
2147 prefer_best = (
2148 not self.params.get('simulate')
2149 and download
2150 and (
2151 not can_merge()
2152 or info_dict.get('is_live') and not self.params.get('live_from_start')
2153 or self.params['outtmpl']['default'] == '-'))
2154 compat = (
2155 prefer_best
2156 or self.params.get('allow_multiple_audio_streams', False)
2157 or 'format-spec' in self.params['compat_opts'])
2158
2159 return (
2160 'best/bestvideo+bestaudio' if prefer_best
2161 else 'bestvideo*+bestaudio/best' if not compat
2162 else 'bestvideo+bestaudio/best')
2163
2164 def build_format_selector(self, format_spec):
2165 def syntax_error(note, start):
2166 message = (
2167 'Invalid format specification: '
2168 '{}\n\t{}\n\t{}^'.format(note, format_spec, ' ' * start[1]))
2169 return SyntaxError(message)
2170
2171 PICKFIRST = 'PICKFIRST'
2172 MERGE = 'MERGE'
2173 SINGLE = 'SINGLE'
2174 GROUP = 'GROUP'
2175 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
2176
2177 allow_multiple_streams = {'audio': self.params.get('allow_multiple_audio_streams', False),
2178 'video': self.params.get('allow_multiple_video_streams', False)}
2179
2180 def _parse_filter(tokens):
2181 filter_parts = []
2182 for type, string_, start, _, _ in tokens:
2183 if type == tokenize.OP and string_ == ']':
2184 return ''.join(filter_parts)
2185 else:
2186 filter_parts.append(string_)
2187
2188 def _remove_unused_ops(tokens):
2189 # Remove operators that we don't use and join them with the surrounding strings.
2190 # E.g. 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
2191 ALLOWED_OPS = ('/', '+', ',', '(', ')')
2192 last_string, last_start, last_end, last_line = None, None, None, None
2193 for type, string_, start, end, line in tokens:
2194 if type == tokenize.OP and string_ == '[':
2195 if last_string:
2196 yield tokenize.NAME, last_string, last_start, last_end, last_line
2197 last_string = None
2198 yield type, string_, start, end, line
2199 # everything inside brackets will be handled by _parse_filter
2200 for type, string_, start, end, line in tokens:
2201 yield type, string_, start, end, line
2202 if type == tokenize.OP and string_ == ']':
2203 break
2204 elif type == tokenize.OP and string_ in ALLOWED_OPS:
2205 if last_string:
2206 yield tokenize.NAME, last_string, last_start, last_end, last_line
2207 last_string = None
2208 yield type, string_, start, end, line
2209 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
2210 if not last_string:
2211 last_string = string_
2212 last_start = start
2213 last_end = end
2214 else:
2215 last_string += string_
2216 if last_string:
2217 yield tokenize.NAME, last_string, last_start, last_end, last_line
2218
2219 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
2220 selectors = []
2221 current_selector = None
2222 for type, string_, start, _, _ in tokens:
2223 # ENCODING is only defined in python 3.x
2224 if type == getattr(tokenize, 'ENCODING', None):
2225 continue
2226 elif type in [tokenize.NAME, tokenize.NUMBER]:
2227 current_selector = FormatSelector(SINGLE, string_, [])
2228 elif type == tokenize.OP:
2229 if string_ == ')':
2230 if not inside_group:
2231 # ')' will be handled by the parentheses group
2232 tokens.restore_last_token()
2233 break
2234 elif inside_merge and string_ in ['/', ',']:
2235 tokens.restore_last_token()
2236 break
2237 elif inside_choice and string_ == ',':
2238 tokens.restore_last_token()
2239 break
2240 elif string_ == ',':
2241 if not current_selector:
2242 raise syntax_error('"," must follow a format selector', start)
2243 selectors.append(current_selector)
2244 current_selector = None
2245 elif string_ == '/':
2246 if not current_selector:
2247 raise syntax_error('"/" must follow a format selector', start)
2248 first_choice = current_selector
2249 second_choice = _parse_format_selection(tokens, inside_choice=True)
2250 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
2251 elif string_ == '[':
2252 if not current_selector:
2253 current_selector = FormatSelector(SINGLE, 'best', [])
2254 format_filter = _parse_filter(tokens)
2255 current_selector.filters.append(format_filter)
2256 elif string_ == '(':
2257 if current_selector:
2258 raise syntax_error('Unexpected "("', start)
2259 group = _parse_format_selection(tokens, inside_group=True)
2260 current_selector = FormatSelector(GROUP, group, [])
2261 elif string_ == '+':
2262 if not current_selector:
2263 raise syntax_error('Unexpected "+"', start)
2264 selector_1 = current_selector
2265 selector_2 = _parse_format_selection(tokens, inside_merge=True)
2266 if not selector_2:
2267 raise syntax_error('Expected a selector', start)
2268 current_selector = FormatSelector(MERGE, (selector_1, selector_2), [])
2269 else:
2270 raise syntax_error(f'Operator not recognized: "{string_}"', start)
2271 elif type == tokenize.ENDMARKER:
2272 break
2273 if current_selector:
2274 selectors.append(current_selector)
2275 return selectors
2276
2277 def _merge(formats_pair):
2278 format_1, format_2 = formats_pair
2279
2280 formats_info = []
2281 formats_info.extend(format_1.get('requested_formats', (format_1,)))
2282 formats_info.extend(format_2.get('requested_formats', (format_2,)))
2283
2284 if not allow_multiple_streams['video'] or not allow_multiple_streams['audio']:
2285 get_no_more = {'video': False, 'audio': False}
2286 for (i, fmt_info) in enumerate(formats_info):
2287 if fmt_info.get('acodec') == fmt_info.get('vcodec') == 'none':
2288 formats_info.pop(i)
2289 continue
2290 for aud_vid in ['audio', 'video']:
2291 if not allow_multiple_streams[aud_vid] and fmt_info.get(aud_vid[0] + 'codec') != 'none':
2292 if get_no_more[aud_vid]:
2293 formats_info.pop(i)
2294 break
2295 get_no_more[aud_vid] = True
2296
2297 if len(formats_info) == 1:
2298 return formats_info[0]
2299
2300 video_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('vcodec') != 'none']
2301 audio_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('acodec') != 'none']
2302
2303 the_only_video = video_fmts[0] if len(video_fmts) == 1 else None
2304 the_only_audio = audio_fmts[0] if len(audio_fmts) == 1 else None
2305
2306 output_ext = get_compatible_ext(
2307 vcodecs=[f.get('vcodec') for f in video_fmts],
2308 acodecs=[f.get('acodec') for f in audio_fmts],
2309 vexts=[f['ext'] for f in video_fmts],
2310 aexts=[f['ext'] for f in audio_fmts],
2311 preferences=(try_call(lambda: self.params['merge_output_format'].split('/'))
2312 or self.params.get('prefer_free_formats') and ('webm', 'mkv')))
2313
2314 filtered = lambda *keys: filter(None, (traverse_obj(fmt, *keys) for fmt in formats_info))
2315
2316 new_dict = {
2317 'requested_formats': formats_info,
2318 'format': '+'.join(filtered('format')),
2319 'format_id': '+'.join(filtered('format_id')),
2320 'ext': output_ext,
2321 'protocol': '+'.join(map(determine_protocol, formats_info)),
2322 'language': '+'.join(orderedSet(filtered('language'))) or None,
2323 'format_note': '+'.join(orderedSet(filtered('format_note'))) or None,
2324 'filesize_approx': sum(filtered('filesize', 'filesize_approx')) or None,
2325 'tbr': sum(filtered('tbr', 'vbr', 'abr')),
2326 }
2327
2328 if the_only_video:
2329 new_dict.update({
2330 'width': the_only_video.get('width'),
2331 'height': the_only_video.get('height'),
2332 'resolution': the_only_video.get('resolution') or self.format_resolution(the_only_video),
2333 'fps': the_only_video.get('fps'),
2334 'dynamic_range': the_only_video.get('dynamic_range'),
2335 'vcodec': the_only_video.get('vcodec'),
2336 'vbr': the_only_video.get('vbr'),
2337 'stretched_ratio': the_only_video.get('stretched_ratio'),
2338 'aspect_ratio': the_only_video.get('aspect_ratio'),
2339 })
2340
2341 if the_only_audio:
2342 new_dict.update({
2343 'acodec': the_only_audio.get('acodec'),
2344 'abr': the_only_audio.get('abr'),
2345 'asr': the_only_audio.get('asr'),
2346 'audio_channels': the_only_audio.get('audio_channels')
2347 })
2348
2349 return new_dict
2350
2351 def _check_formats(formats):
2352 if self.params.get('check_formats') == 'selected':
2353 yield from self._check_formats(formats)
2354 return
2355 elif (self.params.get('check_formats') is not None
2356 or self.params.get('allow_unplayable_formats')):
2357 yield from formats
2358 return
2359
2360 for f in formats:
2361 if f.get('has_drm') or f.get('__needs_testing'):
2362 yield from self._check_formats([f])
2363 else:
2364 yield f
2365
2366 def _build_selector_function(selector):
2367 if isinstance(selector, list): # ,
2368 fs = [_build_selector_function(s) for s in selector]
2369
2370 def selector_function(ctx):
2371 for f in fs:
2372 yield from f(ctx)
2373 return selector_function
2374
2375 elif selector.type == GROUP: # ()
2376 selector_function = _build_selector_function(selector.selector)
2377
2378 elif selector.type == PICKFIRST: # /
2379 fs = [_build_selector_function(s) for s in selector.selector]
2380
2381 def selector_function(ctx):
2382 for f in fs:
2383 picked_formats = list(f(ctx))
2384 if picked_formats:
2385 return picked_formats
2386 return []
2387
2388 elif selector.type == MERGE: # +
2389 selector_1, selector_2 = map(_build_selector_function, selector.selector)
2390
2391 def selector_function(ctx):
2392 for pair in itertools.product(selector_1(ctx), selector_2(ctx)):
2393 yield _merge(pair)
2394
2395 elif selector.type == SINGLE: # atom
2396 format_spec = selector.selector or 'best'
2397
2398 # TODO: Add allvideo, allaudio etc by generalizing the code with best/worst selector
2399 if format_spec == 'all':
2400 def selector_function(ctx):
2401 yield from _check_formats(ctx['formats'][::-1])
2402 elif format_spec == 'mergeall':
2403 def selector_function(ctx):
2404 formats = list(_check_formats(
2405 f for f in ctx['formats'] if f.get('vcodec') != 'none' or f.get('acodec') != 'none'))
2406 if not formats:
2407 return
2408 merged_format = formats[-1]
2409 for f in formats[-2::-1]:
2410 merged_format = _merge((merged_format, f))
2411 yield merged_format
2412
2413 else:
2414 format_fallback, seperate_fallback, format_reverse, format_idx = False, None, True, 1
2415 mobj = re.match(
2416 r'(?P<bw>best|worst|b|w)(?P<type>video|audio|v|a)?(?P<mod>\*)?(?:\.(?P<n>[1-9]\d*))?$',
2417 format_spec)
2418 if mobj is not None:
2419 format_idx = int_or_none(mobj.group('n'), default=1)
2420 format_reverse = mobj.group('bw')[0] == 'b'
2421 format_type = (mobj.group('type') or [None])[0]
2422 not_format_type = {'v': 'a', 'a': 'v'}.get(format_type)
2423 format_modified = mobj.group('mod') is not None
2424
2425 format_fallback = not format_type and not format_modified # for b, w
2426 _filter_f = (
2427 (lambda f: f.get('%scodec' % format_type) != 'none')
2428 if format_type and format_modified # bv*, ba*, wv*, wa*
2429 else (lambda f: f.get('%scodec' % not_format_type) == 'none')
2430 if format_type # bv, ba, wv, wa
2431 else (lambda f: f.get('vcodec') != 'none' and f.get('acodec') != 'none')
2432 if not format_modified # b, w
2433 else lambda f: True) # b*, w*
2434 filter_f = lambda f: _filter_f(f) and (
2435 f.get('vcodec') != 'none' or f.get('acodec') != 'none')
2436 else:
2437 if format_spec in self._format_selection_exts['audio']:
2438 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none'
2439 elif format_spec in self._format_selection_exts['video']:
2440 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none' and f.get('vcodec') != 'none'
2441 seperate_fallback = lambda f: f.get('ext') == format_spec and f.get('vcodec') != 'none'
2442 elif format_spec in self._format_selection_exts['storyboards']:
2443 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') == 'none' and f.get('vcodec') == 'none'
2444 else:
2445 filter_f = lambda f: f.get('format_id') == format_spec # id
2446
2447 def selector_function(ctx):
2448 formats = list(ctx['formats'])
2449 matches = list(filter(filter_f, formats)) if filter_f is not None else formats
2450 if not matches:
2451 if format_fallback and ctx['incomplete_formats']:
2452 # for extractors with incomplete formats (audio only (soundcloud)
2453 # or video only (imgur)) best/worst will fallback to
2454 # best/worst {video,audio}-only format
2455 matches = formats
2456 elif seperate_fallback and not ctx['has_merged_format']:
2457 # for compatibility with youtube-dl when there is no pre-merged format
2458 matches = list(filter(seperate_fallback, formats))
2459 matches = LazyList(_check_formats(matches[::-1 if format_reverse else 1]))
2460 try:
2461 yield matches[format_idx - 1]
2462 except LazyList.IndexError:
2463 return
2464
2465 filters = [self._build_format_filter(f) for f in selector.filters]
2466
2467 def final_selector(ctx):
2468 ctx_copy = dict(ctx)
2469 for _filter in filters:
2470 ctx_copy['formats'] = list(filter(_filter, ctx_copy['formats']))
2471 return selector_function(ctx_copy)
2472 return final_selector
2473
2474 # HACK: Python 3.12 changed the underlying parser, rendering '7_a' invalid
2475 # Prefix numbers with random letters to avoid it being classified as a number
2476 # See: https://github.com/yt-dlp/yt-dlp/pulls/8797
2477 # TODO: Implement parser not reliant on tokenize.tokenize
2478 prefix = ''.join(random.choices(string.ascii_letters, k=32))
2479 stream = io.BytesIO(re.sub(r'\d[_\d]*', rf'{prefix}\g<0>', format_spec).encode())
2480 try:
2481 tokens = list(_remove_unused_ops(
2482 token._replace(string=token.string.replace(prefix, ''))
2483 for token in tokenize.tokenize(stream.readline)))
2484 except tokenize.TokenError:
2485 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
2486
2487 class TokenIterator:
2488 def __init__(self, tokens):
2489 self.tokens = tokens
2490 self.counter = 0
2491
2492 def __iter__(self):
2493 return self
2494
2495 def __next__(self):
2496 if self.counter >= len(self.tokens):
2497 raise StopIteration()
2498 value = self.tokens[self.counter]
2499 self.counter += 1
2500 return value
2501
2502 next = __next__
2503
2504 def restore_last_token(self):
2505 self.counter -= 1
2506
2507 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
2508 return _build_selector_function(parsed_selector)
2509
2510 def _calc_headers(self, info_dict, load_cookies=False):
2511 res = HTTPHeaderDict(self.params['http_headers'], info_dict.get('http_headers'))
2512 clean_headers(res)
2513
2514 if load_cookies: # For --load-info-json
2515 self._load_cookies(res.get('Cookie'), autoscope=info_dict['url']) # compat
2516 self._load_cookies(info_dict.get('cookies'), autoscope=False)
2517 # The `Cookie` header is removed to prevent leaks and unscoped cookies.
2518 # See: https://github.com/yt-dlp/yt-dlp/security/advisories/GHSA-v8mc-9377-rwjj
2519 res.pop('Cookie', None)
2520 cookies = self.cookiejar.get_cookies_for_url(info_dict['url'])
2521 if cookies:
2522 encoder = LenientSimpleCookie()
2523 values = []
2524 for cookie in cookies:
2525 _, value = encoder.value_encode(cookie.value)
2526 values.append(f'{cookie.name}={value}')
2527 if cookie.domain:
2528 values.append(f'Domain={cookie.domain}')
2529 if cookie.path:
2530 values.append(f'Path={cookie.path}')
2531 if cookie.secure:
2532 values.append('Secure')
2533 if cookie.expires:
2534 values.append(f'Expires={cookie.expires}')
2535 if cookie.version:
2536 values.append(f'Version={cookie.version}')
2537 info_dict['cookies'] = '; '.join(values)
2538
2539 if 'X-Forwarded-For' not in res:
2540 x_forwarded_for_ip = info_dict.get('__x_forwarded_for_ip')
2541 if x_forwarded_for_ip:
2542 res['X-Forwarded-For'] = x_forwarded_for_ip
2543
2544 return res
2545
2546 def _calc_cookies(self, url):
2547 self.deprecation_warning('"YoutubeDL._calc_cookies" is deprecated and may be removed in a future version')
2548 return self.cookiejar.get_cookie_header(url)
2549
2550 def _sort_thumbnails(self, thumbnails):
2551 thumbnails.sort(key=lambda t: (
2552 t.get('preference') if t.get('preference') is not None else -1,
2553 t.get('width') if t.get('width') is not None else -1,
2554 t.get('height') if t.get('height') is not None else -1,
2555 t.get('id') if t.get('id') is not None else '',
2556 t.get('url')))
2557
2558 def _sanitize_thumbnails(self, info_dict):
2559 thumbnails = info_dict.get('thumbnails')
2560 if thumbnails is None:
2561 thumbnail = info_dict.get('thumbnail')
2562 if thumbnail:
2563 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
2564 if not thumbnails:
2565 return
2566
2567 def check_thumbnails(thumbnails):
2568 for t in thumbnails:
2569 self.to_screen(f'[info] Testing thumbnail {t["id"]}')
2570 try:
2571 self.urlopen(HEADRequest(t['url']))
2572 except network_exceptions as err:
2573 self.to_screen(f'[info] Unable to connect to thumbnail {t["id"]} URL {t["url"]!r} - {err}. Skipping...')
2574 continue
2575 yield t
2576
2577 self._sort_thumbnails(thumbnails)
2578 for i, t in enumerate(thumbnails):
2579 if t.get('id') is None:
2580 t['id'] = '%d' % i
2581 if t.get('width') and t.get('height'):
2582 t['resolution'] = '%dx%d' % (t['width'], t['height'])
2583 t['url'] = sanitize_url(t['url'])
2584
2585 if self.params.get('check_formats') is True:
2586 info_dict['thumbnails'] = LazyList(check_thumbnails(thumbnails[::-1]), reverse=True)
2587 else:
2588 info_dict['thumbnails'] = thumbnails
2589
2590 def _fill_common_fields(self, info_dict, final=True):
2591 # TODO: move sanitization here
2592 if final:
2593 title = info_dict['fulltitle'] = info_dict.get('title')
2594 if not title:
2595 if title == '':
2596 self.write_debug('Extractor gave empty title. Creating a generic title')
2597 else:
2598 self.report_warning('Extractor failed to obtain "title". Creating a generic title instead')
2599 info_dict['title'] = f'{info_dict["extractor"].replace(":", "-")} video #{info_dict["id"]}'
2600
2601 if info_dict.get('duration') is not None:
2602 info_dict['duration_string'] = formatSeconds(info_dict['duration'])
2603
2604 for ts_key, date_key in (
2605 ('timestamp', 'upload_date'),
2606 ('release_timestamp', 'release_date'),
2607 ('modified_timestamp', 'modified_date'),
2608 ):
2609 if info_dict.get(date_key) is None and info_dict.get(ts_key) is not None:
2610 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
2611 # see http://bugs.python.org/issue1646728)
2612 with contextlib.suppress(ValueError, OverflowError, OSError):
2613 upload_date = datetime.datetime.fromtimestamp(info_dict[ts_key], datetime.timezone.utc)
2614 info_dict[date_key] = upload_date.strftime('%Y%m%d')
2615
2616 if not info_dict.get('release_year'):
2617 info_dict['release_year'] = traverse_obj(info_dict, ('release_date', {lambda x: int(x[:4])}))
2618
2619 live_keys = ('is_live', 'was_live')
2620 live_status = info_dict.get('live_status')
2621 if live_status is None:
2622 for key in live_keys:
2623 if info_dict.get(key) is False:
2624 continue
2625 if info_dict.get(key):
2626 live_status = key
2627 break
2628 if all(info_dict.get(key) is False for key in live_keys):
2629 live_status = 'not_live'
2630 if live_status:
2631 info_dict['live_status'] = live_status
2632 for key in live_keys:
2633 if info_dict.get(key) is None:
2634 info_dict[key] = (live_status == key)
2635 if live_status == 'post_live':
2636 info_dict['was_live'] = True
2637
2638 # Auto generate title fields corresponding to the *_number fields when missing
2639 # in order to always have clean titles. This is very common for TV series.
2640 for field in ('chapter', 'season', 'episode'):
2641 if final and info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
2642 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
2643
2644 def _raise_pending_errors(self, info):
2645 err = info.pop('__pending_error', None)
2646 if err:
2647 self.report_error(err, tb=False)
2648
2649 def sort_formats(self, info_dict):
2650 formats = self._get_formats(info_dict)
2651 formats.sort(key=FormatSorter(
2652 self, info_dict.get('_format_sort_fields') or []).calculate_preference)
2653
2654 def process_video_result(self, info_dict, download=True):
2655 assert info_dict.get('_type', 'video') == 'video'
2656 self._num_videos += 1
2657
2658 if 'id' not in info_dict:
2659 raise ExtractorError('Missing "id" field in extractor result', ie=info_dict['extractor'])
2660 elif not info_dict.get('id'):
2661 raise ExtractorError('Extractor failed to obtain "id"', ie=info_dict['extractor'])
2662
2663 def report_force_conversion(field, field_not, conversion):
2664 self.report_warning(
2665 '"%s" field is not %s - forcing %s conversion, there is an error in extractor'
2666 % (field, field_not, conversion))
2667
2668 def sanitize_string_field(info, string_field):
2669 field = info.get(string_field)
2670 if field is None or isinstance(field, str):
2671 return
2672 report_force_conversion(string_field, 'a string', 'string')
2673 info[string_field] = str(field)
2674
2675 def sanitize_numeric_fields(info):
2676 for numeric_field in self._NUMERIC_FIELDS:
2677 field = info.get(numeric_field)
2678 if field is None or isinstance(field, (int, float)):
2679 continue
2680 report_force_conversion(numeric_field, 'numeric', 'int')
2681 info[numeric_field] = int_or_none(field)
2682
2683 sanitize_string_field(info_dict, 'id')
2684 sanitize_numeric_fields(info_dict)
2685 if info_dict.get('section_end') and info_dict.get('section_start') is not None:
2686 info_dict['duration'] = round(info_dict['section_end'] - info_dict['section_start'], 3)
2687 if (info_dict.get('duration') or 0) <= 0 and info_dict.pop('duration', None):
2688 self.report_warning('"duration" field is negative, there is an error in extractor')
2689
2690 chapters = info_dict.get('chapters') or []
2691 if chapters and chapters[0].get('start_time'):
2692 chapters.insert(0, {'start_time': 0})
2693
2694 dummy_chapter = {'end_time': 0, 'start_time': info_dict.get('duration')}
2695 for idx, (prev, current, next_) in enumerate(zip(
2696 (dummy_chapter, *chapters), chapters, (*chapters[1:], dummy_chapter)), 1):
2697 if current.get('start_time') is None:
2698 current['start_time'] = prev.get('end_time')
2699 if not current.get('end_time'):
2700 current['end_time'] = next_.get('start_time')
2701 if not current.get('title'):
2702 current['title'] = f'<Untitled Chapter {idx}>'
2703
2704 if 'playlist' not in info_dict:
2705 # It isn't part of a playlist
2706 info_dict['playlist'] = None
2707 info_dict['playlist_index'] = None
2708
2709 self._sanitize_thumbnails(info_dict)
2710
2711 thumbnail = info_dict.get('thumbnail')
2712 thumbnails = info_dict.get('thumbnails')
2713 if thumbnail:
2714 info_dict['thumbnail'] = sanitize_url(thumbnail)
2715 elif thumbnails:
2716 info_dict['thumbnail'] = thumbnails[-1]['url']
2717
2718 if info_dict.get('display_id') is None and 'id' in info_dict:
2719 info_dict['display_id'] = info_dict['id']
2720
2721 self._fill_common_fields(info_dict)
2722
2723 for cc_kind in ('subtitles', 'automatic_captions'):
2724 cc = info_dict.get(cc_kind)
2725 if cc:
2726 for _, subtitle in cc.items():
2727 for subtitle_format in subtitle:
2728 if subtitle_format.get('url'):
2729 subtitle_format['url'] = sanitize_url(subtitle_format['url'])
2730 if subtitle_format.get('ext') is None:
2731 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
2732
2733 automatic_captions = info_dict.get('automatic_captions')
2734 subtitles = info_dict.get('subtitles')
2735
2736 info_dict['requested_subtitles'] = self.process_subtitles(
2737 info_dict['id'], subtitles, automatic_captions)
2738
2739 formats = self._get_formats(info_dict)
2740
2741 # Backward compatibility with InfoExtractor._sort_formats
2742 field_preference = (formats or [{}])[0].pop('__sort_fields', None)
2743 if field_preference:
2744 info_dict['_format_sort_fields'] = field_preference
2745
2746 info_dict['_has_drm'] = any( # or None ensures --clean-infojson removes it
2747 f.get('has_drm') and f['has_drm'] != 'maybe' for f in formats) or None
2748 if not self.params.get('allow_unplayable_formats'):
2749 formats = [f for f in formats if not f.get('has_drm') or f['has_drm'] == 'maybe']
2750
2751 if formats and all(f.get('acodec') == f.get('vcodec') == 'none' for f in formats):
2752 self.report_warning(
2753 f'{"This video is DRM protected and " if info_dict["_has_drm"] else ""}'
2754 'only images are available for download. Use --list-formats to see them'.capitalize())
2755
2756 get_from_start = not info_dict.get('is_live') or bool(self.params.get('live_from_start'))
2757 if not get_from_start:
2758 info_dict['title'] += ' ' + datetime.datetime.now().strftime('%Y-%m-%d %H:%M')
2759 if info_dict.get('is_live') and formats:
2760 formats = [f for f in formats if bool(f.get('is_from_start')) == get_from_start]
2761 if get_from_start and not formats:
2762 self.raise_no_formats(info_dict, msg=(
2763 '--live-from-start is passed, but there are no formats that can be downloaded from the start. '
2764 'If you want to download from the current time, use --no-live-from-start'))
2765
2766 def is_wellformed(f):
2767 url = f.get('url')
2768 if not url:
2769 self.report_warning(
2770 '"url" field is missing or empty - skipping format, '
2771 'there is an error in extractor')
2772 return False
2773 if isinstance(url, bytes):
2774 sanitize_string_field(f, 'url')
2775 return True
2776
2777 # Filter out malformed formats for better extraction robustness
2778 formats = list(filter(is_wellformed, formats or []))
2779
2780 if not formats:
2781 self.raise_no_formats(info_dict)
2782
2783 for format in formats:
2784 sanitize_string_field(format, 'format_id')
2785 sanitize_numeric_fields(format)
2786 format['url'] = sanitize_url(format['url'])
2787 if format.get('ext') is None:
2788 format['ext'] = determine_ext(format['url']).lower()
2789 if format.get('protocol') is None:
2790 format['protocol'] = determine_protocol(format)
2791 if format.get('resolution') is None:
2792 format['resolution'] = self.format_resolution(format, default=None)
2793 if format.get('dynamic_range') is None and format.get('vcodec') != 'none':
2794 format['dynamic_range'] = 'SDR'
2795 if format.get('aspect_ratio') is None:
2796 format['aspect_ratio'] = try_call(lambda: round(format['width'] / format['height'], 2))
2797 # For fragmented formats, "tbr" is often max bitrate and not average
2798 if (('manifest-filesize-approx' in self.params['compat_opts'] or not format.get('manifest_url'))
2799 and info_dict.get('duration') and format.get('tbr')
2800 and not format.get('filesize') and not format.get('filesize_approx')):
2801 format['filesize_approx'] = int(info_dict['duration'] * format['tbr'] * (1024 / 8))
2802 format['http_headers'] = self._calc_headers(collections.ChainMap(format, info_dict), load_cookies=True)
2803
2804 # Safeguard against old/insecure infojson when using --load-info-json
2805 if info_dict.get('http_headers'):
2806 info_dict['http_headers'] = HTTPHeaderDict(info_dict['http_headers'])
2807 info_dict['http_headers'].pop('Cookie', None)
2808
2809 # This is copied to http_headers by the above _calc_headers and can now be removed
2810 if '__x_forwarded_for_ip' in info_dict:
2811 del info_dict['__x_forwarded_for_ip']
2812
2813 self.sort_formats({
2814 'formats': formats,
2815 '_format_sort_fields': info_dict.get('_format_sort_fields')
2816 })
2817
2818 # Sanitize and group by format_id
2819 formats_dict = {}
2820 for i, format in enumerate(formats):
2821 if not format.get('format_id'):
2822 format['format_id'] = str(i)
2823 else:
2824 # Sanitize format_id from characters used in format selector expression
2825 format['format_id'] = re.sub(r'[\s,/+\[\]()]', '_', format['format_id'])
2826 formats_dict.setdefault(format['format_id'], []).append(format)
2827
2828 # Make sure all formats have unique format_id
2829 common_exts = set(itertools.chain(*self._format_selection_exts.values()))
2830 for format_id, ambiguous_formats in formats_dict.items():
2831 ambigious_id = len(ambiguous_formats) > 1
2832 for i, format in enumerate(ambiguous_formats):
2833 if ambigious_id:
2834 format['format_id'] = '%s-%d' % (format_id, i)
2835 # Ensure there is no conflict between id and ext in format selection
2836 # See https://github.com/yt-dlp/yt-dlp/issues/1282
2837 if format['format_id'] != format['ext'] and format['format_id'] in common_exts:
2838 format['format_id'] = 'f%s' % format['format_id']
2839
2840 if format.get('format') is None:
2841 format['format'] = '{id} - {res}{note}'.format(
2842 id=format['format_id'],
2843 res=self.format_resolution(format),
2844 note=format_field(format, 'format_note', ' (%s)'),
2845 )
2846
2847 if self.params.get('check_formats') is True:
2848 formats = LazyList(self._check_formats(formats[::-1]), reverse=True)
2849
2850 if not formats or formats[0] is not info_dict:
2851 # only set the 'formats' fields if the original info_dict list them
2852 # otherwise we end up with a circular reference, the first (and unique)
2853 # element in the 'formats' field in info_dict is info_dict itself,
2854 # which can't be exported to json
2855 info_dict['formats'] = formats
2856
2857 info_dict, _ = self.pre_process(info_dict)
2858
2859 if self._match_entry(info_dict, incomplete=self._format_fields) is not None:
2860 return info_dict
2861
2862 self.post_extract(info_dict)
2863 info_dict, _ = self.pre_process(info_dict, 'after_filter')
2864
2865 # The pre-processors may have modified the formats
2866 formats = self._get_formats(info_dict)
2867
2868 list_only = self.params.get('simulate') == 'list_only'
2869 interactive_format_selection = not list_only and self.format_selector == '-'
2870 if self.params.get('list_thumbnails'):
2871 self.list_thumbnails(info_dict)
2872 if self.params.get('listsubtitles'):
2873 if 'automatic_captions' in info_dict:
2874 self.list_subtitles(
2875 info_dict['id'], automatic_captions, 'automatic captions')
2876 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
2877 if self.params.get('listformats') or interactive_format_selection:
2878 self.list_formats(info_dict)
2879 if list_only:
2880 # Without this printing, -F --print-json will not work
2881 self.__forced_printings(info_dict)
2882 return info_dict
2883
2884 format_selector = self.format_selector
2885 while True:
2886 if interactive_format_selection:
2887 req_format = input(self._format_screen('\nEnter format selector ', self.Styles.EMPHASIS)
2888 + '(Press ENTER for default, or Ctrl+C to quit)'
2889 + self._format_screen(': ', self.Styles.EMPHASIS))
2890 try:
2891 format_selector = self.build_format_selector(req_format) if req_format else None
2892 except SyntaxError as err:
2893 self.report_error(err, tb=False, is_error=False)
2894 continue
2895
2896 if format_selector is None:
2897 req_format = self._default_format_spec(info_dict, download=download)
2898 self.write_debug(f'Default format spec: {req_format}')
2899 format_selector = self.build_format_selector(req_format)
2900
2901 formats_to_download = list(format_selector({
2902 'formats': formats,
2903 'has_merged_format': any('none' not in (f.get('acodec'), f.get('vcodec')) for f in formats),
2904 'incomplete_formats': (all(f.get('vcodec') == 'none' for f in formats) # No formats with video
2905 or all(f.get('acodec') == 'none' for f in formats)), # OR, No formats with audio
2906 }))
2907 if interactive_format_selection and not formats_to_download:
2908 self.report_error('Requested format is not available', tb=False, is_error=False)
2909 continue
2910 break
2911
2912 if not formats_to_download:
2913 if not self.params.get('ignore_no_formats_error'):
2914 raise ExtractorError(
2915 'Requested format is not available. Use --list-formats for a list of available formats',
2916 expected=True, video_id=info_dict['id'], ie=info_dict['extractor'])
2917 self.report_warning('Requested format is not available')
2918 # Process what we can, even without any available formats.
2919 formats_to_download = [{}]
2920
2921 requested_ranges = tuple(self.params.get('download_ranges', lambda *_: [{}])(info_dict, self))
2922 best_format, downloaded_formats = formats_to_download[-1], []
2923 if download:
2924 if best_format and requested_ranges:
2925 def to_screen(*msg):
2926 self.to_screen(f'[info] {info_dict["id"]}: {" ".join(", ".join(variadic(m)) for m in msg)}')
2927
2928 to_screen(f'Downloading {len(formats_to_download)} format(s):',
2929 (f['format_id'] for f in formats_to_download))
2930 if requested_ranges != ({}, ):
2931 to_screen(f'Downloading {len(requested_ranges)} time ranges:',
2932 (f'{c["start_time"]:.1f}-{c["end_time"]:.1f}' for c in requested_ranges))
2933 max_downloads_reached = False
2934
2935 for fmt, chapter in itertools.product(formats_to_download, requested_ranges):
2936 new_info = self._copy_infodict(info_dict)
2937 new_info.update(fmt)
2938 offset, duration = info_dict.get('section_start') or 0, info_dict.get('duration') or float('inf')
2939 end_time = offset + min(chapter.get('end_time', duration), duration)
2940 # duration may not be accurate. So allow deviations <1sec
2941 if end_time == float('inf') or end_time > offset + duration + 1:
2942 end_time = None
2943 if chapter or offset:
2944 new_info.update({
2945 'section_start': offset + chapter.get('start_time', 0),
2946 'section_end': end_time,
2947 'section_title': chapter.get('title'),
2948 'section_number': chapter.get('index'),
2949 })
2950 downloaded_formats.append(new_info)
2951 try:
2952 self.process_info(new_info)
2953 except MaxDownloadsReached:
2954 max_downloads_reached = True
2955 self._raise_pending_errors(new_info)
2956 # Remove copied info
2957 for key, val in tuple(new_info.items()):
2958 if info_dict.get(key) == val:
2959 new_info.pop(key)
2960 if max_downloads_reached:
2961 break
2962
2963 write_archive = {f.get('__write_download_archive', False) for f in downloaded_formats}
2964 assert write_archive.issubset({True, False, 'ignore'})
2965 if True in write_archive and False not in write_archive:
2966 self.record_download_archive(info_dict)
2967
2968 info_dict['requested_downloads'] = downloaded_formats
2969 info_dict = self.run_all_pps('after_video', info_dict)
2970 if max_downloads_reached:
2971 raise MaxDownloadsReached()
2972
2973 # We update the info dict with the selected best quality format (backwards compatibility)
2974 info_dict.update(best_format)
2975 return info_dict
2976
2977 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
2978 """Select the requested subtitles and their format"""
2979 available_subs, normal_sub_langs = {}, []
2980 if normal_subtitles and self.params.get('writesubtitles'):
2981 available_subs.update(normal_subtitles)
2982 normal_sub_langs = tuple(normal_subtitles.keys())
2983 if automatic_captions and self.params.get('writeautomaticsub'):
2984 for lang, cap_info in automatic_captions.items():
2985 if lang not in available_subs:
2986 available_subs[lang] = cap_info
2987
2988 if not available_subs or (
2989 not self.params.get('writesubtitles')
2990 and not self.params.get('writeautomaticsub')):
2991 return None
2992
2993 all_sub_langs = tuple(available_subs.keys())
2994 if self.params.get('allsubtitles', False):
2995 requested_langs = all_sub_langs
2996 elif self.params.get('subtitleslangs', False):
2997 try:
2998 requested_langs = orderedSet_from_options(
2999 self.params.get('subtitleslangs'), {'all': all_sub_langs}, use_regex=True)
3000 except re.error as e:
3001 raise ValueError(f'Wrong regex for subtitlelangs: {e.pattern}')
3002 else:
3003 requested_langs = LazyList(itertools.chain(
3004 ['en'] if 'en' in normal_sub_langs else [],
3005 filter(lambda f: f.startswith('en'), normal_sub_langs),
3006 ['en'] if 'en' in all_sub_langs else [],
3007 filter(lambda f: f.startswith('en'), all_sub_langs),
3008 normal_sub_langs, all_sub_langs,
3009 ))[:1]
3010 if requested_langs:
3011 self.to_screen(f'[info] {video_id}: Downloading subtitles: {", ".join(requested_langs)}')
3012
3013 formats_query = self.params.get('subtitlesformat', 'best')
3014 formats_preference = formats_query.split('/') if formats_query else []
3015 subs = {}
3016 for lang in requested_langs:
3017 formats = available_subs.get(lang)
3018 if formats is None:
3019 self.report_warning(f'{lang} subtitles not available for {video_id}')
3020 continue
3021 for ext in formats_preference:
3022 if ext == 'best':
3023 f = formats[-1]
3024 break
3025 matches = list(filter(lambda f: f['ext'] == ext, formats))
3026 if matches:
3027 f = matches[-1]
3028 break
3029 else:
3030 f = formats[-1]
3031 self.report_warning(
3032 'No subtitle format found matching "%s" for language %s, '
3033 'using %s' % (formats_query, lang, f['ext']))
3034 subs[lang] = f
3035 return subs
3036
3037 def _forceprint(self, key, info_dict):
3038 if info_dict is None:
3039 return
3040 info_copy = info_dict.copy()
3041 info_copy.setdefault('filename', self.prepare_filename(info_dict))
3042 if info_dict.get('requested_formats') is not None:
3043 # For RTMP URLs, also include the playpath
3044 info_copy['urls'] = '\n'.join(f['url'] + f.get('play_path', '') for f in info_dict['requested_formats'])
3045 elif info_dict.get('url'):
3046 info_copy['urls'] = info_dict['url'] + info_dict.get('play_path', '')
3047 info_copy['formats_table'] = self.render_formats_table(info_dict)
3048 info_copy['thumbnails_table'] = self.render_thumbnails_table(info_dict)
3049 info_copy['subtitles_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('subtitles'))
3050 info_copy['automatic_captions_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('automatic_captions'))
3051
3052 def format_tmpl(tmpl):
3053 mobj = re.fullmatch(r'([\w.:,]|-\d|(?P<dict>{([\w.:,]|-\d)+}))+=?', tmpl)
3054 if not mobj:
3055 return tmpl
3056
3057 fmt = '%({})s'
3058 if tmpl.startswith('{'):
3059 tmpl, fmt = f'.{tmpl}', '%({})j'
3060 if tmpl.endswith('='):
3061 tmpl, fmt = tmpl[:-1], '{0} = %({0})#j'
3062 return '\n'.join(map(fmt.format, [tmpl] if mobj.group('dict') else tmpl.split(',')))
3063
3064 for tmpl in self.params['forceprint'].get(key, []):
3065 self.to_stdout(self.evaluate_outtmpl(format_tmpl(tmpl), info_copy))
3066
3067 for tmpl, file_tmpl in self.params['print_to_file'].get(key, []):
3068 filename = self.prepare_filename(info_dict, outtmpl=file_tmpl)
3069 tmpl = format_tmpl(tmpl)
3070 self.to_screen(f'[info] Writing {tmpl!r} to: {filename}')
3071 if self._ensure_dir_exists(filename):
3072 with open(filename, 'a', encoding='utf-8', newline='') as f:
3073 f.write(self.evaluate_outtmpl(tmpl, info_copy) + os.linesep)
3074
3075 return info_copy
3076
3077 def __forced_printings(self, info_dict, filename=None, incomplete=True):
3078 if (self.params.get('forcejson')
3079 or self.params['forceprint'].get('video')
3080 or self.params['print_to_file'].get('video')):
3081 self.post_extract(info_dict)
3082 if filename:
3083 info_dict['filename'] = filename
3084 info_copy = self._forceprint('video', info_dict)
3085
3086 def print_field(field, actual_field=None, optional=False):
3087 if actual_field is None:
3088 actual_field = field
3089 if self.params.get(f'force{field}') and (
3090 info_copy.get(field) is not None or (not optional and not incomplete)):
3091 self.to_stdout(info_copy[actual_field])
3092
3093 print_field('title')
3094 print_field('id')
3095 print_field('url', 'urls')
3096 print_field('thumbnail', optional=True)
3097 print_field('description', optional=True)
3098 print_field('filename')
3099 if self.params.get('forceduration') and info_copy.get('duration') is not None:
3100 self.to_stdout(formatSeconds(info_copy['duration']))
3101 print_field('format')
3102
3103 if self.params.get('forcejson'):
3104 self.to_stdout(json.dumps(self.sanitize_info(info_dict)))
3105
3106 def dl(self, name, info, subtitle=False, test=False):
3107 if not info.get('url'):
3108 self.raise_no_formats(info, True)
3109
3110 if test:
3111 verbose = self.params.get('verbose')
3112 params = {
3113 'test': True,
3114 'quiet': self.params.get('quiet') or not verbose,
3115 'verbose': verbose,
3116 'noprogress': not verbose,
3117 'nopart': True,
3118 'skip_unavailable_fragments': False,
3119 'keep_fragments': False,
3120 'overwrites': True,
3121 '_no_ytdl_file': True,
3122 }
3123 else:
3124 params = self.params
3125 fd = get_suitable_downloader(info, params, to_stdout=(name == '-'))(self, params)
3126 if not test:
3127 for ph in self._progress_hooks:
3128 fd.add_progress_hook(ph)
3129 urls = '", "'.join(
3130 (f['url'].split(',')[0] + ',<data>' if f['url'].startswith('data:') else f['url'])
3131 for f in info.get('requested_formats', []) or [info])
3132 self.write_debug(f'Invoking {fd.FD_NAME} downloader on "{urls}"')
3133
3134 # Note: Ideally info should be a deep-copied so that hooks cannot modify it.
3135 # But it may contain objects that are not deep-copyable
3136 new_info = self._copy_infodict(info)
3137 if new_info.get('http_headers') is None:
3138 new_info['http_headers'] = self._calc_headers(new_info)
3139 return fd.download(name, new_info, subtitle)
3140
3141 def existing_file(self, filepaths, *, default_overwrite=True):
3142 existing_files = list(filter(os.path.exists, orderedSet(filepaths)))
3143 if existing_files and not self.params.get('overwrites', default_overwrite):
3144 return existing_files[0]
3145
3146 for file in existing_files:
3147 self.report_file_delete(file)
3148 os.remove(file)
3149 return None
3150
3151 def process_info(self, info_dict):
3152 """Process a single resolved IE result. (Modifies it in-place)"""
3153
3154 assert info_dict.get('_type', 'video') == 'video'
3155 original_infodict = info_dict
3156
3157 if 'format' not in info_dict and 'ext' in info_dict:
3158 info_dict['format'] = info_dict['ext']
3159
3160 if self._match_entry(info_dict) is not None:
3161 info_dict['__write_download_archive'] = 'ignore'
3162 return
3163
3164 # Does nothing under normal operation - for backward compatibility of process_info
3165 self.post_extract(info_dict)
3166
3167 def replace_info_dict(new_info):
3168 nonlocal info_dict
3169 if new_info == info_dict:
3170 return
3171 info_dict.clear()
3172 info_dict.update(new_info)
3173
3174 new_info, _ = self.pre_process(info_dict, 'video')
3175 replace_info_dict(new_info)
3176 self._num_downloads += 1
3177
3178 # info_dict['_filename'] needs to be set for backward compatibility
3179 info_dict['_filename'] = full_filename = self.prepare_filename(info_dict, warn=True)
3180 temp_filename = self.prepare_filename(info_dict, 'temp')
3181 files_to_move = {}
3182
3183 # Forced printings
3184 self.__forced_printings(info_dict, full_filename, incomplete=('format' not in info_dict))
3185
3186 def check_max_downloads():
3187 if self._num_downloads >= float(self.params.get('max_downloads') or 'inf'):
3188 raise MaxDownloadsReached()
3189
3190 if self.params.get('simulate'):
3191 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
3192 check_max_downloads()
3193 return
3194
3195 if full_filename is None:
3196 return
3197 if not self._ensure_dir_exists(encodeFilename(full_filename)):
3198 return
3199 if not self._ensure_dir_exists(encodeFilename(temp_filename)):
3200 return
3201
3202 if self._write_description('video', info_dict,
3203 self.prepare_filename(info_dict, 'description')) is None:
3204 return
3205
3206 sub_files = self._write_subtitles(info_dict, temp_filename)
3207 if sub_files is None:
3208 return
3209 files_to_move.update(dict(sub_files))
3210
3211 thumb_files = self._write_thumbnails(
3212 'video', info_dict, temp_filename, self.prepare_filename(info_dict, 'thumbnail'))
3213 if thumb_files is None:
3214 return
3215 files_to_move.update(dict(thumb_files))
3216
3217 infofn = self.prepare_filename(info_dict, 'infojson')
3218 _infojson_written = self._write_info_json('video', info_dict, infofn)
3219 if _infojson_written:
3220 info_dict['infojson_filename'] = infofn
3221 # For backward compatibility, even though it was a private field
3222 info_dict['__infojson_filename'] = infofn
3223 elif _infojson_written is None:
3224 return
3225
3226 # Note: Annotations are deprecated
3227 annofn = None
3228 if self.params.get('writeannotations', False):
3229 annofn = self.prepare_filename(info_dict, 'annotation')
3230 if annofn:
3231 if not self._ensure_dir_exists(encodeFilename(annofn)):
3232 return
3233 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(annofn)):
3234 self.to_screen('[info] Video annotations are already present')
3235 elif not info_dict.get('annotations'):
3236 self.report_warning('There are no annotations to write.')
3237 else:
3238 try:
3239 self.to_screen('[info] Writing video annotations to: ' + annofn)
3240 with open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
3241 annofile.write(info_dict['annotations'])
3242 except (KeyError, TypeError):
3243 self.report_warning('There are no annotations to write.')
3244 except OSError:
3245 self.report_error('Cannot write annotations file: ' + annofn)
3246 return
3247
3248 # Write internet shortcut files
3249 def _write_link_file(link_type):
3250 url = try_get(info_dict['webpage_url'], iri_to_uri)
3251 if not url:
3252 self.report_warning(
3253 f'Cannot write internet shortcut file because the actual URL of "{info_dict["webpage_url"]}" is unknown')
3254 return True
3255 linkfn = replace_extension(self.prepare_filename(info_dict, 'link'), link_type, info_dict.get('ext'))
3256 if not self._ensure_dir_exists(encodeFilename(linkfn)):
3257 return False
3258 if self.params.get('overwrites', True) and os.path.exists(encodeFilename(linkfn)):
3259 self.to_screen(f'[info] Internet shortcut (.{link_type}) is already present')
3260 return True
3261 try:
3262 self.to_screen(f'[info] Writing internet shortcut (.{link_type}) to: {linkfn}')
3263 with open(encodeFilename(to_high_limit_path(linkfn)), 'w', encoding='utf-8',
3264 newline='\r\n' if link_type == 'url' else '\n') as linkfile:
3265 template_vars = {'url': url}
3266 if link_type == 'desktop':
3267 template_vars['filename'] = linkfn[:-(len(link_type) + 1)]
3268 linkfile.write(LINK_TEMPLATES[link_type] % template_vars)
3269 except OSError:
3270 self.report_error(f'Cannot write internet shortcut {linkfn}')
3271 return False
3272 return True
3273
3274 write_links = {
3275 'url': self.params.get('writeurllink'),
3276 'webloc': self.params.get('writewebloclink'),
3277 'desktop': self.params.get('writedesktoplink'),
3278 }
3279 if self.params.get('writelink'):
3280 link_type = ('webloc' if sys.platform == 'darwin'
3281 else 'desktop' if sys.platform.startswith('linux')
3282 else 'url')
3283 write_links[link_type] = True
3284
3285 if any(should_write and not _write_link_file(link_type)
3286 for link_type, should_write in write_links.items()):
3287 return
3288
3289 new_info, files_to_move = self.pre_process(info_dict, 'before_dl', files_to_move)
3290 replace_info_dict(new_info)
3291
3292 if self.params.get('skip_download'):
3293 info_dict['filepath'] = temp_filename
3294 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
3295 info_dict['__files_to_move'] = files_to_move
3296 replace_info_dict(self.run_pp(MoveFilesAfterDownloadPP(self, False), info_dict))
3297 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
3298 else:
3299 # Download
3300 info_dict.setdefault('__postprocessors', [])
3301 try:
3302
3303 def existing_video_file(*filepaths):
3304 ext = info_dict.get('ext')
3305 converted = lambda file: replace_extension(file, self.params.get('final_ext') or ext, ext)
3306 file = self.existing_file(itertools.chain(*zip(map(converted, filepaths), filepaths)),
3307 default_overwrite=False)
3308 if file:
3309 info_dict['ext'] = os.path.splitext(file)[1][1:]
3310 return file
3311
3312 fd, success = None, True
3313 if info_dict.get('protocol') or info_dict.get('url'):
3314 fd = get_suitable_downloader(info_dict, self.params, to_stdout=temp_filename == '-')
3315 if fd != FFmpegFD and 'no-direct-merge' not in self.params['compat_opts'] and (
3316 info_dict.get('section_start') or info_dict.get('section_end')):
3317 msg = ('This format cannot be partially downloaded' if FFmpegFD.available()
3318 else 'You have requested downloading the video partially, but ffmpeg is not installed')
3319 self.report_error(f'{msg}. Aborting')
3320 return
3321
3322 if info_dict.get('requested_formats') is not None:
3323 old_ext = info_dict['ext']
3324 if self.params.get('merge_output_format') is None:
3325 if (info_dict['ext'] == 'webm'
3326 and info_dict.get('thumbnails')
3327 # check with type instead of pp_key, __name__, or isinstance
3328 # since we dont want any custom PPs to trigger this
3329 and any(type(pp) == EmbedThumbnailPP for pp in self._pps['post_process'])): # noqa: E721
3330 info_dict['ext'] = 'mkv'
3331 self.report_warning(
3332 'webm doesn\'t support embedding a thumbnail, mkv will be used')
3333 new_ext = info_dict['ext']
3334
3335 def correct_ext(filename, ext=new_ext):
3336 if filename == '-':
3337 return filename
3338 filename_real_ext = os.path.splitext(filename)[1][1:]
3339 filename_wo_ext = (
3340 os.path.splitext(filename)[0]
3341 if filename_real_ext in (old_ext, new_ext)
3342 else filename)
3343 return f'{filename_wo_ext}.{ext}'
3344
3345 # Ensure filename always has a correct extension for successful merge
3346 full_filename = correct_ext(full_filename)
3347 temp_filename = correct_ext(temp_filename)
3348 dl_filename = existing_video_file(full_filename, temp_filename)
3349
3350 info_dict['__real_download'] = False
3351 # NOTE: Copy so that original format dicts are not modified
3352 info_dict['requested_formats'] = list(map(dict, info_dict['requested_formats']))
3353
3354 merger = FFmpegMergerPP(self)
3355 downloaded = []
3356 if dl_filename is not None:
3357 self.report_file_already_downloaded(dl_filename)
3358 elif fd:
3359 for f in info_dict['requested_formats'] if fd != FFmpegFD else []:
3360 f['filepath'] = fname = prepend_extension(
3361 correct_ext(temp_filename, info_dict['ext']),
3362 'f%s' % f['format_id'], info_dict['ext'])
3363 downloaded.append(fname)
3364 info_dict['url'] = '\n'.join(f['url'] for f in info_dict['requested_formats'])
3365 success, real_download = self.dl(temp_filename, info_dict)
3366 info_dict['__real_download'] = real_download
3367 else:
3368 if self.params.get('allow_unplayable_formats'):
3369 self.report_warning(
3370 'You have requested merging of multiple formats '
3371 'while also allowing unplayable formats to be downloaded. '
3372 'The formats won\'t be merged to prevent data corruption.')
3373 elif not merger.available:
3374 msg = 'You have requested merging of multiple formats but ffmpeg is not installed'
3375 if not self.params.get('ignoreerrors'):
3376 self.report_error(f'{msg}. Aborting due to --abort-on-error')
3377 return
3378 self.report_warning(f'{msg}. The formats won\'t be merged')
3379
3380 if temp_filename == '-':
3381 reason = ('using a downloader other than ffmpeg' if FFmpegFD.can_merge_formats(info_dict, self.params)
3382 else 'but the formats are incompatible for simultaneous download' if merger.available
3383 else 'but ffmpeg is not installed')
3384 self.report_warning(
3385 f'You have requested downloading multiple formats to stdout {reason}. '
3386 'The formats will be streamed one after the other')
3387 fname = temp_filename
3388 for f in info_dict['requested_formats']:
3389 new_info = dict(info_dict)
3390 del new_info['requested_formats']
3391 new_info.update(f)
3392 if temp_filename != '-':
3393 fname = prepend_extension(
3394 correct_ext(temp_filename, new_info['ext']),
3395 'f%s' % f['format_id'], new_info['ext'])
3396 if not self._ensure_dir_exists(fname):
3397 return
3398 f['filepath'] = fname
3399 downloaded.append(fname)
3400 partial_success, real_download = self.dl(fname, new_info)
3401 info_dict['__real_download'] = info_dict['__real_download'] or real_download
3402 success = success and partial_success
3403
3404 if downloaded and merger.available and not self.params.get('allow_unplayable_formats'):
3405 info_dict['__postprocessors'].append(merger)
3406 info_dict['__files_to_merge'] = downloaded
3407 # Even if there were no downloads, it is being merged only now
3408 info_dict['__real_download'] = True
3409 else:
3410 for file in downloaded:
3411 files_to_move[file] = None
3412 else:
3413 # Just a single file
3414 dl_filename = existing_video_file(full_filename, temp_filename)
3415 if dl_filename is None or dl_filename == temp_filename:
3416 # dl_filename == temp_filename could mean that the file was partially downloaded with --no-part.
3417 # So we should try to resume the download
3418 success, real_download = self.dl(temp_filename, info_dict)
3419 info_dict['__real_download'] = real_download
3420 else:
3421 self.report_file_already_downloaded(dl_filename)
3422
3423 dl_filename = dl_filename or temp_filename
3424 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
3425
3426 except network_exceptions as err:
3427 self.report_error('unable to download video data: %s' % error_to_compat_str(err))
3428 return
3429 except OSError as err:
3430 raise UnavailableVideoError(err)
3431 except (ContentTooShortError, ) as err:
3432 self.report_error(f'content too short (expected {err.expected} bytes and served {err.downloaded})')
3433 return
3434
3435 self._raise_pending_errors(info_dict)
3436 if success and full_filename != '-':
3437
3438 def fixup():
3439 do_fixup = True
3440 fixup_policy = self.params.get('fixup')
3441 vid = info_dict['id']
3442
3443 if fixup_policy in ('ignore', 'never'):
3444 return
3445 elif fixup_policy == 'warn':
3446 do_fixup = 'warn'
3447 elif fixup_policy != 'force':
3448 assert fixup_policy in ('detect_or_warn', None)
3449 if not info_dict.get('__real_download'):
3450 do_fixup = False
3451
3452 def ffmpeg_fixup(cndn, msg, cls):
3453 if not (do_fixup and cndn):
3454 return
3455 elif do_fixup == 'warn':
3456 self.report_warning(f'{vid}: {msg}')
3457 return
3458 pp = cls(self)
3459 if pp.available:
3460 info_dict['__postprocessors'].append(pp)
3461 else:
3462 self.report_warning(f'{vid}: {msg}. Install ffmpeg to fix this automatically')
3463
3464 stretched_ratio = info_dict.get('stretched_ratio')
3465 ffmpeg_fixup(stretched_ratio not in (1, None),
3466 f'Non-uniform pixel ratio {stretched_ratio}',
3467 FFmpegFixupStretchedPP)
3468
3469 downloader = get_suitable_downloader(info_dict, self.params) if 'protocol' in info_dict else None
3470 downloader = downloader.FD_NAME if downloader else None
3471
3472 ext = info_dict.get('ext')
3473 postprocessed_by_ffmpeg = info_dict.get('requested_formats') or any((
3474 isinstance(pp, FFmpegVideoConvertorPP)
3475 and resolve_recode_mapping(ext, pp.mapping)[0] not in (ext, None)
3476 ) for pp in self._pps['post_process'])
3477
3478 if not postprocessed_by_ffmpeg:
3479 ffmpeg_fixup(fd != FFmpegFD and ext == 'm4a'
3480 and info_dict.get('container') == 'm4a_dash',
3481 'writing DASH m4a. Only some players support this container',
3482 FFmpegFixupM4aPP)
3483 ffmpeg_fixup(downloader == 'hlsnative' and not self.params.get('hls_use_mpegts')
3484 or info_dict.get('is_live') and self.params.get('hls_use_mpegts') is None,
3485 'Possible MPEG-TS in MP4 container or malformed AAC timestamps',
3486 FFmpegFixupM3u8PP)
3487 ffmpeg_fixup(info_dict.get('is_live') and downloader == 'dashsegments',
3488 'Possible duplicate MOOV atoms', FFmpegFixupDuplicateMoovPP)
3489
3490 ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed timestamps detected', FFmpegFixupTimestampPP)
3491 ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed duration detected', FFmpegFixupDurationPP)
3492
3493 fixup()
3494 try:
3495 replace_info_dict(self.post_process(dl_filename, info_dict, files_to_move))
3496 except PostProcessingError as err:
3497 self.report_error('Postprocessing: %s' % str(err))
3498 return
3499 try:
3500 for ph in self._post_hooks:
3501 ph(info_dict['filepath'])
3502 except Exception as err:
3503 self.report_error('post hooks: %s' % str(err))
3504 return
3505 info_dict['__write_download_archive'] = True
3506
3507 assert info_dict is original_infodict # Make sure the info_dict was modified in-place
3508 if self.params.get('force_write_download_archive'):
3509 info_dict['__write_download_archive'] = True
3510 check_max_downloads()
3511
3512 def __download_wrapper(self, func):
3513 @functools.wraps(func)
3514 def wrapper(*args, **kwargs):
3515 try:
3516 res = func(*args, **kwargs)
3517 except UnavailableVideoError as e:
3518 self.report_error(e)
3519 except DownloadCancelled as e:
3520 self.to_screen(f'[info] {e}')
3521 if not self.params.get('break_per_url'):
3522 raise
3523 self._num_downloads = 0
3524 else:
3525 if self.params.get('dump_single_json', False):
3526 self.post_extract(res)
3527 self.to_stdout(json.dumps(self.sanitize_info(res)))
3528 return wrapper
3529
3530 def download(self, url_list):
3531 """Download a given list of URLs."""
3532 url_list = variadic(url_list) # Passing a single URL is a common mistake
3533 outtmpl = self.params['outtmpl']['default']
3534 if (len(url_list) > 1
3535 and outtmpl != '-'
3536 and '%' not in outtmpl
3537 and self.params.get('max_downloads') != 1):
3538 raise SameFileError(outtmpl)
3539
3540 for url in url_list:
3541 self.__download_wrapper(self.extract_info)(
3542 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
3543
3544 return self._download_retcode
3545
3546 def download_with_info_file(self, info_filename):
3547 with contextlib.closing(fileinput.FileInput(
3548 [info_filename], mode='r',
3549 openhook=fileinput.hook_encoded('utf-8'))) as f:
3550 # FileInput doesn't have a read method, we can't call json.load
3551 infos = [self.sanitize_info(info, self.params.get('clean_infojson', True))
3552 for info in variadic(json.loads('\n'.join(f)))]
3553 for info in infos:
3554 try:
3555 self.__download_wrapper(self.process_ie_result)(info, download=True)
3556 except (DownloadError, EntryNotInPlaylist, ReExtractInfo) as e:
3557 if not isinstance(e, EntryNotInPlaylist):
3558 self.to_stderr('\r')
3559 webpage_url = info.get('webpage_url')
3560 if webpage_url is None:
3561 raise
3562 self.report_warning(f'The info failed to download: {e}; trying with URL {webpage_url}')
3563 self.download([webpage_url])
3564 return self._download_retcode
3565
3566 @staticmethod
3567 def sanitize_info(info_dict, remove_private_keys=False):
3568 ''' Sanitize the infodict for converting to json '''
3569 if info_dict is None:
3570 return info_dict
3571 info_dict.setdefault('epoch', int(time.time()))
3572 info_dict.setdefault('_type', 'video')
3573 info_dict.setdefault('_version', {
3574 'version': __version__,
3575 'current_git_head': current_git_head(),
3576 'release_git_head': RELEASE_GIT_HEAD,
3577 'repository': ORIGIN,
3578 })
3579
3580 if remove_private_keys:
3581 reject = lambda k, v: v is None or k.startswith('__') or k in {
3582 'requested_downloads', 'requested_formats', 'requested_subtitles', 'requested_entries',
3583 'entries', 'filepath', '_filename', 'filename', 'infojson_filename', 'original_url',
3584 'playlist_autonumber',
3585 }
3586 else:
3587 reject = lambda k, v: False
3588
3589 def filter_fn(obj):
3590 if isinstance(obj, dict):
3591 return {k: filter_fn(v) for k, v in obj.items() if not reject(k, v)}
3592 elif isinstance(obj, (list, tuple, set, LazyList)):
3593 return list(map(filter_fn, obj))
3594 elif obj is None or isinstance(obj, (str, int, float, bool)):
3595 return obj
3596 else:
3597 return repr(obj)
3598
3599 return filter_fn(info_dict)
3600
3601 @staticmethod
3602 def filter_requested_info(info_dict, actually_filter=True):
3603 ''' Alias of sanitize_info for backward compatibility '''
3604 return YoutubeDL.sanitize_info(info_dict, actually_filter)
3605
3606 def _delete_downloaded_files(self, *files_to_delete, info={}, msg=None):
3607 for filename in set(filter(None, files_to_delete)):
3608 if msg:
3609 self.to_screen(msg % filename)
3610 try:
3611 os.remove(filename)
3612 except OSError:
3613 self.report_warning(f'Unable to delete file {filename}')
3614 if filename in info.get('__files_to_move', []): # NB: Delete even if None
3615 del info['__files_to_move'][filename]
3616
3617 @staticmethod
3618 def post_extract(info_dict):
3619 def actual_post_extract(info_dict):
3620 if info_dict.get('_type') in ('playlist', 'multi_video'):
3621 for video_dict in info_dict.get('entries', {}):
3622 actual_post_extract(video_dict or {})
3623 return
3624
3625 post_extractor = info_dict.pop('__post_extractor', None) or (lambda: {})
3626 info_dict.update(post_extractor())
3627
3628 actual_post_extract(info_dict or {})
3629
3630 def run_pp(self, pp, infodict):
3631 files_to_delete = []
3632 if '__files_to_move' not in infodict:
3633 infodict['__files_to_move'] = {}
3634 try:
3635 files_to_delete, infodict = pp.run(infodict)
3636 except PostProcessingError as e:
3637 # Must be True and not 'only_download'
3638 if self.params.get('ignoreerrors') is True:
3639 self.report_error(e)
3640 return infodict
3641 raise
3642
3643 if not files_to_delete:
3644 return infodict
3645 if self.params.get('keepvideo', False):
3646 for f in files_to_delete:
3647 infodict['__files_to_move'].setdefault(f, '')
3648 else:
3649 self._delete_downloaded_files(
3650 *files_to_delete, info=infodict, msg='Deleting original file %s (pass -k to keep)')
3651 return infodict
3652
3653 def run_all_pps(self, key, info, *, additional_pps=None):
3654 if key != 'video':
3655 self._forceprint(key, info)
3656 for pp in (additional_pps or []) + self._pps[key]:
3657 info = self.run_pp(pp, info)
3658 return info
3659
3660 def pre_process(self, ie_info, key='pre_process', files_to_move=None):
3661 info = dict(ie_info)
3662 info['__files_to_move'] = files_to_move or {}
3663 try:
3664 info = self.run_all_pps(key, info)
3665 except PostProcessingError as err:
3666 msg = f'Preprocessing: {err}'
3667 info.setdefault('__pending_error', msg)
3668 self.report_error(msg, is_error=False)
3669 return info, info.pop('__files_to_move', None)
3670
3671 def post_process(self, filename, info, files_to_move=None):
3672 """Run all the postprocessors on the given file."""
3673 info['filepath'] = filename
3674 info['__files_to_move'] = files_to_move or {}
3675 info = self.run_all_pps('post_process', info, additional_pps=info.get('__postprocessors'))
3676 info = self.run_pp(MoveFilesAfterDownloadPP(self), info)
3677 del info['__files_to_move']
3678 return self.run_all_pps('after_move', info)
3679
3680 def _make_archive_id(self, info_dict):
3681 video_id = info_dict.get('id')
3682 if not video_id:
3683 return
3684 # Future-proof against any change in case
3685 # and backwards compatibility with prior versions
3686 extractor = info_dict.get('extractor_key') or info_dict.get('ie_key') # key in a playlist
3687 if extractor is None:
3688 url = str_or_none(info_dict.get('url'))
3689 if not url:
3690 return
3691 # Try to find matching extractor for the URL and take its ie_key
3692 for ie_key, ie in self._ies.items():
3693 if ie.suitable(url):
3694 extractor = ie_key
3695 break
3696 else:
3697 return
3698 return make_archive_id(extractor, video_id)
3699
3700 def in_download_archive(self, info_dict):
3701 if not self.archive:
3702 return False
3703
3704 vid_ids = [self._make_archive_id(info_dict)]
3705 vid_ids.extend(info_dict.get('_old_archive_ids') or [])
3706 return any(id_ in self.archive for id_ in vid_ids)
3707
3708 def record_download_archive(self, info_dict):
3709 fn = self.params.get('download_archive')
3710 if fn is None:
3711 return
3712 vid_id = self._make_archive_id(info_dict)
3713 assert vid_id
3714
3715 self.write_debug(f'Adding to archive: {vid_id}')
3716 if is_path_like(fn):
3717 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
3718 archive_file.write(vid_id + '\n')
3719 self.archive.add(vid_id)
3720
3721 @staticmethod
3722 def format_resolution(format, default='unknown'):
3723 if format.get('vcodec') == 'none' and format.get('acodec') != 'none':
3724 return 'audio only'
3725 if format.get('resolution') is not None:
3726 return format['resolution']
3727 if format.get('width') and format.get('height'):
3728 return '%dx%d' % (format['width'], format['height'])
3729 elif format.get('height'):
3730 return '%sp' % format['height']
3731 elif format.get('width'):
3732 return '%dx?' % format['width']
3733 return default
3734
3735 def _list_format_headers(self, *headers):
3736 if self.params.get('listformats_table', True) is not False:
3737 return [self._format_out(header, self.Styles.HEADERS) for header in headers]
3738 return headers
3739
3740 def _format_note(self, fdict):
3741 res = ''
3742 if fdict.get('ext') in ['f4f', 'f4m']:
3743 res += '(unsupported)'
3744 if fdict.get('language'):
3745 if res:
3746 res += ' '
3747 res += '[%s]' % fdict['language']
3748 if fdict.get('format_note') is not None:
3749 if res:
3750 res += ' '
3751 res += fdict['format_note']
3752 if fdict.get('tbr') is not None:
3753 if res:
3754 res += ', '
3755 res += '%4dk' % fdict['tbr']
3756 if fdict.get('container') is not None:
3757 if res:
3758 res += ', '
3759 res += '%s container' % fdict['container']
3760 if (fdict.get('vcodec') is not None
3761 and fdict.get('vcodec') != 'none'):
3762 if res:
3763 res += ', '
3764 res += fdict['vcodec']
3765 if fdict.get('vbr') is not None:
3766 res += '@'
3767 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
3768 res += 'video@'
3769 if fdict.get('vbr') is not None:
3770 res += '%4dk' % fdict['vbr']
3771 if fdict.get('fps') is not None:
3772 if res:
3773 res += ', '
3774 res += '%sfps' % fdict['fps']
3775 if fdict.get('acodec') is not None:
3776 if res:
3777 res += ', '
3778 if fdict['acodec'] == 'none':
3779 res += 'video only'
3780 else:
3781 res += '%-5s' % fdict['acodec']
3782 elif fdict.get('abr') is not None:
3783 if res:
3784 res += ', '
3785 res += 'audio'
3786 if fdict.get('abr') is not None:
3787 res += '@%3dk' % fdict['abr']
3788 if fdict.get('asr') is not None:
3789 res += ' (%5dHz)' % fdict['asr']
3790 if fdict.get('filesize') is not None:
3791 if res:
3792 res += ', '
3793 res += format_bytes(fdict['filesize'])
3794 elif fdict.get('filesize_approx') is not None:
3795 if res:
3796 res += ', '
3797 res += '~' + format_bytes(fdict['filesize_approx'])
3798 return res
3799
3800 def _get_formats(self, info_dict):
3801 if info_dict.get('formats') is None:
3802 if info_dict.get('url') and info_dict.get('_type', 'video') == 'video':
3803 return [info_dict]
3804 return []
3805 return info_dict['formats']
3806
3807 def render_formats_table(self, info_dict):
3808 formats = self._get_formats(info_dict)
3809 if not formats:
3810 return
3811 if not self.params.get('listformats_table', True) is not False:
3812 table = [
3813 [
3814 format_field(f, 'format_id'),
3815 format_field(f, 'ext'),
3816 self.format_resolution(f),
3817 self._format_note(f)
3818 ] for f in formats if (f.get('preference') or 0) >= -1000]
3819 return render_table(['format code', 'extension', 'resolution', 'note'], table, extra_gap=1)
3820
3821 def simplified_codec(f, field):
3822 assert field in ('acodec', 'vcodec')
3823 codec = f.get(field)
3824 if not codec:
3825 return 'unknown'
3826 elif codec != 'none':
3827 return '.'.join(codec.split('.')[:4])
3828
3829 if field == 'vcodec' and f.get('acodec') == 'none':
3830 return 'images'
3831 elif field == 'acodec' and f.get('vcodec') == 'none':
3832 return ''
3833 return self._format_out('audio only' if field == 'vcodec' else 'video only',
3834 self.Styles.SUPPRESS)
3835
3836 delim = self._format_out('\u2502', self.Styles.DELIM, '|', test_encoding=True)
3837 table = [
3838 [
3839 self._format_out(format_field(f, 'format_id'), self.Styles.ID),
3840 format_field(f, 'ext'),
3841 format_field(f, func=self.format_resolution, ignore=('audio only', 'images')),
3842 format_field(f, 'fps', '\t%d', func=round),
3843 format_field(f, 'dynamic_range', '%s', ignore=(None, 'SDR')).replace('HDR', ''),
3844 format_field(f, 'audio_channels', '\t%s'),
3845 delim, (
3846 format_field(f, 'filesize', ' \t%s', func=format_bytes)
3847 or format_field(f, 'filesize_approx', '≈\t%s', func=format_bytes)
3848 or format_field(try_call(lambda: format_bytes(int(info_dict['duration'] * f['tbr'] * (1024 / 8)))),
3849 None, self._format_out('~\t%s', self.Styles.SUPPRESS))),
3850 format_field(f, 'tbr', '\t%dk', func=round),
3851 shorten_protocol_name(f.get('protocol', '')),
3852 delim,
3853 simplified_codec(f, 'vcodec'),
3854 format_field(f, 'vbr', '\t%dk', func=round),
3855 simplified_codec(f, 'acodec'),
3856 format_field(f, 'abr', '\t%dk', func=round),
3857 format_field(f, 'asr', '\t%s', func=format_decimal_suffix),
3858 join_nonempty(format_field(f, 'language', '[%s]'), join_nonempty(
3859 self._format_out('UNSUPPORTED', self.Styles.BAD_FORMAT) if f.get('ext') in ('f4f', 'f4m') else None,
3860 (self._format_out('Maybe DRM', self.Styles.WARNING) if f.get('has_drm') == 'maybe'
3861 else self._format_out('DRM', self.Styles.BAD_FORMAT) if f.get('has_drm') else None),
3862 format_field(f, 'format_note'),
3863 format_field(f, 'container', ignore=(None, f.get('ext'))),
3864 delim=', '), delim=' '),
3865 ] for f in formats if f.get('preference') is None or f['preference'] >= -1000]
3866 header_line = self._list_format_headers(
3867 'ID', 'EXT', 'RESOLUTION', '\tFPS', 'HDR', 'CH', delim, '\tFILESIZE', '\tTBR', 'PROTO',
3868 delim, 'VCODEC', '\tVBR', 'ACODEC', '\tABR', '\tASR', 'MORE INFO')
3869
3870 return render_table(
3871 header_line, table, hide_empty=True,
3872 delim=self._format_out('\u2500', self.Styles.DELIM, '-', test_encoding=True))
3873
3874 def render_thumbnails_table(self, info_dict):
3875 thumbnails = list(info_dict.get('thumbnails') or [])
3876 if not thumbnails:
3877 return None
3878 return render_table(
3879 self._list_format_headers('ID', 'Width', 'Height', 'URL'),
3880 [[t.get('id'), t.get('width') or 'unknown', t.get('height') or 'unknown', t['url']] for t in thumbnails])
3881
3882 def render_subtitles_table(self, video_id, subtitles):
3883 def _row(lang, formats):
3884 exts, names = zip(*((f['ext'], f.get('name') or 'unknown') for f in reversed(formats)))
3885 if len(set(names)) == 1:
3886 names = [] if names[0] == 'unknown' else names[:1]
3887 return [lang, ', '.join(names), ', '.join(exts)]
3888
3889 if not subtitles:
3890 return None
3891 return render_table(
3892 self._list_format_headers('Language', 'Name', 'Formats'),
3893 [_row(lang, formats) for lang, formats in subtitles.items()],
3894 hide_empty=True)
3895
3896 def __list_table(self, video_id, name, func, *args):
3897 table = func(*args)
3898 if not table:
3899 self.to_screen(f'{video_id} has no {name}')
3900 return
3901 self.to_screen(f'[info] Available {name} for {video_id}:')
3902 self.to_stdout(table)
3903
3904 def list_formats(self, info_dict):
3905 self.__list_table(info_dict['id'], 'formats', self.render_formats_table, info_dict)
3906
3907 def list_thumbnails(self, info_dict):
3908 self.__list_table(info_dict['id'], 'thumbnails', self.render_thumbnails_table, info_dict)
3909
3910 def list_subtitles(self, video_id, subtitles, name='subtitles'):
3911 self.__list_table(video_id, name, self.render_subtitles_table, video_id, subtitles)
3912
3913 def print_debug_header(self):
3914 if not self.params.get('verbose'):
3915 return
3916
3917 from . import _IN_CLI # Must be delayed import
3918
3919 # These imports can be slow. So import them only as needed
3920 from .extractor.extractors import _LAZY_LOADER
3921 from .extractor.extractors import (
3922 _PLUGIN_CLASSES as plugin_ies,
3923 _PLUGIN_OVERRIDES as plugin_ie_overrides
3924 )
3925
3926 def get_encoding(stream):
3927 ret = str(getattr(stream, 'encoding', 'missing (%s)' % type(stream).__name__))
3928 additional_info = []
3929 if os.environ.get('TERM', '').lower() == 'dumb':
3930 additional_info.append('dumb')
3931 if not supports_terminal_sequences(stream):
3932 from .utils import WINDOWS_VT_MODE # Must be imported locally
3933 additional_info.append('No VT' if WINDOWS_VT_MODE is False else 'No ANSI')
3934 if additional_info:
3935 ret = f'{ret} ({",".join(additional_info)})'
3936 return ret
3937
3938 encoding_str = 'Encodings: locale %s, fs %s, pref %s, %s' % (
3939 locale.getpreferredencoding(),
3940 sys.getfilesystemencoding(),
3941 self.get_encoding(),
3942 ', '.join(
3943 f'{key} {get_encoding(stream)}' for key, stream in self._out_files.items_
3944 if stream is not None and key != 'console')
3945 )
3946
3947 logger = self.params.get('logger')
3948 if logger:
3949 write_debug = lambda msg: logger.debug(f'[debug] {msg}')
3950 write_debug(encoding_str)
3951 else:
3952 write_string(f'[debug] {encoding_str}\n', encoding=None)
3953 write_debug = lambda msg: self._write_string(f'[debug] {msg}\n')
3954
3955 source = detect_variant()
3956 if VARIANT not in (None, 'pip'):
3957 source += '*'
3958 klass = type(self)
3959 write_debug(join_nonempty(
3960 f'{REPOSITORY.rpartition("/")[2]} version',
3961 _make_label(ORIGIN, CHANNEL.partition('@')[2] or __version__, __version__),
3962 f'[{RELEASE_GIT_HEAD[:9]}]' if RELEASE_GIT_HEAD else '',
3963 '' if source == 'unknown' else f'({source})',
3964 '' if _IN_CLI else 'API' if klass == YoutubeDL else f'API:{self.__module__}.{klass.__qualname__}',
3965 delim=' '))
3966
3967 if not _IN_CLI:
3968 write_debug(f'params: {self.params}')
3969
3970 if not _LAZY_LOADER:
3971 if os.environ.get('YTDLP_NO_LAZY_EXTRACTORS'):
3972 write_debug('Lazy loading extractors is forcibly disabled')
3973 else:
3974 write_debug('Lazy loading extractors is disabled')
3975 if self.params['compat_opts']:
3976 write_debug('Compatibility options: %s' % ', '.join(self.params['compat_opts']))
3977
3978 if current_git_head():
3979 write_debug(f'Git HEAD: {current_git_head()}')
3980 write_debug(system_identifier())
3981
3982 exe_versions, ffmpeg_features = FFmpegPostProcessor.get_versions_and_features(self)
3983 ffmpeg_features = {key for key, val in ffmpeg_features.items() if val}
3984 if ffmpeg_features:
3985 exe_versions['ffmpeg'] += ' (%s)' % ','.join(sorted(ffmpeg_features))
3986
3987 exe_versions['rtmpdump'] = rtmpdump_version()
3988 exe_versions['phantomjs'] = PhantomJSwrapper._version()
3989 exe_str = ', '.join(
3990 f'{exe} {v}' for exe, v in sorted(exe_versions.items()) if v
3991 ) or 'none'
3992 write_debug('exe versions: %s' % exe_str)
3993
3994 from .compat.compat_utils import get_package_info
3995 from .dependencies import available_dependencies
3996
3997 write_debug('Optional libraries: %s' % (', '.join(sorted({
3998 join_nonempty(*get_package_info(m)) for m in available_dependencies.values()
3999 })) or 'none'))
4000
4001 write_debug(f'Proxy map: {self.proxies}')
4002 write_debug(f'Request Handlers: {", ".join(rh.RH_NAME for rh in self._request_director.handlers.values())}')
4003 for plugin_type, plugins in {'Extractor': plugin_ies, 'Post-Processor': plugin_pps}.items():
4004 display_list = ['%s%s' % (
4005 klass.__name__, '' if klass.__name__ == name else f' as {name}')
4006 for name, klass in plugins.items()]
4007 if plugin_type == 'Extractor':
4008 display_list.extend(f'{plugins[-1].IE_NAME.partition("+")[2]} ({parent.__name__})'
4009 for parent, plugins in plugin_ie_overrides.items())
4010 if not display_list:
4011 continue
4012 write_debug(f'{plugin_type} Plugins: {", ".join(sorted(display_list))}')
4013
4014 plugin_dirs = plugin_directories()
4015 if plugin_dirs:
4016 write_debug(f'Plugin directories: {plugin_dirs}')
4017
4018 # Not implemented
4019 if False and self.params.get('call_home'):
4020 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode()
4021 write_debug('Public IP address: %s' % ipaddr)
4022 latest_version = self.urlopen(
4023 'https://yt-dl.org/latest/version').read().decode()
4024 if version_tuple(latest_version) > version_tuple(__version__):
4025 self.report_warning(
4026 'You are using an outdated version (newest version: %s)! '
4027 'See https://yt-dl.org/update if you need help updating.' %
4028 latest_version)
4029
4030 @functools.cached_property
4031 def proxies(self):
4032 """Global proxy configuration"""
4033 opts_proxy = self.params.get('proxy')
4034 if opts_proxy is not None:
4035 if opts_proxy == '':
4036 opts_proxy = '__noproxy__'
4037 proxies = {'all': opts_proxy}
4038 else:
4039 proxies = urllib.request.getproxies()
4040 # compat. Set HTTPS_PROXY to __noproxy__ to revert
4041 if 'http' in proxies and 'https' not in proxies:
4042 proxies['https'] = proxies['http']
4043
4044 return proxies
4045
4046 @functools.cached_property
4047 def cookiejar(self):
4048 """Global cookiejar instance"""
4049 return load_cookies(
4050 self.params.get('cookiefile'), self.params.get('cookiesfrombrowser'), self)
4051
4052 @property
4053 def _opener(self):
4054 """
4055 Get a urllib OpenerDirector from the Urllib handler (deprecated).
4056 """
4057 self.deprecation_warning('YoutubeDL._opener is deprecated, use YoutubeDL.urlopen()')
4058 handler = self._request_director.handlers['Urllib']
4059 return handler._get_instance(cookiejar=self.cookiejar, proxies=self.proxies)
4060
4061 def urlopen(self, req):
4062 """ Start an HTTP download """
4063 if isinstance(req, str):
4064 req = Request(req)
4065 elif isinstance(req, urllib.request.Request):
4066 self.deprecation_warning(
4067 'Passing a urllib.request.Request object to YoutubeDL.urlopen() is deprecated. '
4068 'Use yt_dlp.networking.common.Request instead.')
4069 req = urllib_req_to_req(req)
4070 assert isinstance(req, Request)
4071
4072 # compat: Assume user:pass url params are basic auth
4073 url, basic_auth_header = extract_basic_auth(req.url)
4074 if basic_auth_header:
4075 req.headers['Authorization'] = basic_auth_header
4076 req.url = sanitize_url(url)
4077
4078 clean_proxies(proxies=req.proxies, headers=req.headers)
4079 clean_headers(req.headers)
4080
4081 try:
4082 return self._request_director.send(req)
4083 except NoSupportingHandlers as e:
4084 for ue in e.unsupported_errors:
4085 # FIXME: This depends on the order of errors.
4086 if not (ue.handler and ue.msg):
4087 continue
4088 if ue.handler.RH_KEY == 'Urllib' and 'unsupported url scheme: "file"' in ue.msg.lower():
4089 raise RequestError(
4090 'file:// URLs are disabled by default in yt-dlp for security reasons. '
4091 'Use --enable-file-urls to enable at your own risk.', cause=ue) from ue
4092 if 'unsupported proxy type: "https"' in ue.msg.lower():
4093 raise RequestError(
4094 'To use an HTTPS proxy for this request, one of the following dependencies needs to be installed: requests')
4095
4096 elif (
4097 re.match(r'unsupported url scheme: "wss?"', ue.msg.lower())
4098 and 'websockets' not in self._request_director.handlers
4099 ):
4100 raise RequestError(
4101 'This request requires WebSocket support. '
4102 'Ensure one of the following dependencies are installed: websockets',
4103 cause=ue) from ue
4104 raise
4105 except SSLError as e:
4106 if 'UNSAFE_LEGACY_RENEGOTIATION_DISABLED' in str(e):
4107 raise RequestError('UNSAFE_LEGACY_RENEGOTIATION_DISABLED: Try using --legacy-server-connect', cause=e) from e
4108 elif 'SSLV3_ALERT_HANDSHAKE_FAILURE' in str(e):
4109 raise RequestError(
4110 'SSLV3_ALERT_HANDSHAKE_FAILURE: The server may not support the current cipher list. '
4111 'Try using --legacy-server-connect', cause=e) from e
4112 raise
4113 except HTTPError as e: # TODO: Remove in a future release
4114 raise _CompatHTTPError(e) from e
4115
4116 def build_request_director(self, handlers, preferences=None):
4117 logger = _YDLLogger(self)
4118 headers = self.params['http_headers'].copy()
4119 proxies = self.proxies.copy()
4120 clean_headers(headers)
4121 clean_proxies(proxies, headers)
4122
4123 director = RequestDirector(logger=logger, verbose=self.params.get('debug_printtraffic'))
4124 for handler in handlers:
4125 director.add_handler(handler(
4126 logger=logger,
4127 headers=headers,
4128 cookiejar=self.cookiejar,
4129 proxies=proxies,
4130 prefer_system_certs='no-certifi' in self.params['compat_opts'],
4131 verify=not self.params.get('nocheckcertificate'),
4132 **traverse_obj(self.params, {
4133 'verbose': 'debug_printtraffic',
4134 'source_address': 'source_address',
4135 'timeout': 'socket_timeout',
4136 'legacy_ssl_support': 'legacyserverconnect',
4137 'enable_file_urls': 'enable_file_urls',
4138 'client_cert': {
4139 'client_certificate': 'client_certificate',
4140 'client_certificate_key': 'client_certificate_key',
4141 'client_certificate_password': 'client_certificate_password',
4142 },
4143 }),
4144 ))
4145 director.preferences.update(preferences or [])
4146 if 'prefer-legacy-http-handler' in self.params['compat_opts']:
4147 director.preferences.add(lambda rh, _: 500 if rh.RH_KEY == 'Urllib' else 0)
4148 return director
4149
4150 def encode(self, s):
4151 if isinstance(s, bytes):
4152 return s # Already encoded
4153
4154 try:
4155 return s.encode(self.get_encoding())
4156 except UnicodeEncodeError as err:
4157 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
4158 raise
4159
4160 def get_encoding(self):
4161 encoding = self.params.get('encoding')
4162 if encoding is None:
4163 encoding = preferredencoding()
4164 return encoding
4165
4166 def _write_info_json(self, label, ie_result, infofn, overwrite=None):
4167 ''' Write infojson and returns True = written, 'exists' = Already exists, False = skip, None = error '''
4168 if overwrite is None:
4169 overwrite = self.params.get('overwrites', True)
4170 if not self.params.get('writeinfojson'):
4171 return False
4172 elif not infofn:
4173 self.write_debug(f'Skipping writing {label} infojson')
4174 return False
4175 elif not self._ensure_dir_exists(infofn):
4176 return None
4177 elif not overwrite and os.path.exists(infofn):
4178 self.to_screen(f'[info] {label.title()} metadata is already present')
4179 return 'exists'
4180
4181 self.to_screen(f'[info] Writing {label} metadata as JSON to: {infofn}')
4182 try:
4183 write_json_file(self.sanitize_info(ie_result, self.params.get('clean_infojson', True)), infofn)
4184 return True
4185 except OSError:
4186 self.report_error(f'Cannot write {label} metadata to JSON file {infofn}')
4187 return None
4188
4189 def _write_description(self, label, ie_result, descfn):
4190 ''' Write description and returns True = written, False = skip, None = error '''
4191 if not self.params.get('writedescription'):
4192 return False
4193 elif not descfn:
4194 self.write_debug(f'Skipping writing {label} description')
4195 return False
4196 elif not self._ensure_dir_exists(descfn):
4197 return None
4198 elif not self.params.get('overwrites', True) and os.path.exists(descfn):
4199 self.to_screen(f'[info] {label.title()} description is already present')
4200 elif ie_result.get('description') is None:
4201 self.to_screen(f'[info] There\'s no {label} description to write')
4202 return False
4203 else:
4204 try:
4205 self.to_screen(f'[info] Writing {label} description to: {descfn}')
4206 with open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
4207 descfile.write(ie_result['description'])
4208 except OSError:
4209 self.report_error(f'Cannot write {label} description file {descfn}')
4210 return None
4211 return True
4212
4213 def _write_subtitles(self, info_dict, filename):
4214 ''' Write subtitles to file and return list of (sub_filename, final_sub_filename); or None if error'''
4215 ret = []
4216 subtitles = info_dict.get('requested_subtitles')
4217 if not (self.params.get('writesubtitles') or self.params.get('writeautomaticsub')):
4218 # subtitles download errors are already managed as troubles in relevant IE
4219 # that way it will silently go on when used with unsupporting IE
4220 return ret
4221 elif not subtitles:
4222 self.to_screen('[info] There are no subtitles for the requested languages')
4223 return ret
4224 sub_filename_base = self.prepare_filename(info_dict, 'subtitle')
4225 if not sub_filename_base:
4226 self.to_screen('[info] Skipping writing video subtitles')
4227 return ret
4228
4229 for sub_lang, sub_info in subtitles.items():
4230 sub_format = sub_info['ext']
4231 sub_filename = subtitles_filename(filename, sub_lang, sub_format, info_dict.get('ext'))
4232 sub_filename_final = subtitles_filename(sub_filename_base, sub_lang, sub_format, info_dict.get('ext'))
4233 existing_sub = self.existing_file((sub_filename_final, sub_filename))
4234 if existing_sub:
4235 self.to_screen(f'[info] Video subtitle {sub_lang}.{sub_format} is already present')
4236 sub_info['filepath'] = existing_sub
4237 ret.append((existing_sub, sub_filename_final))
4238 continue
4239
4240 self.to_screen(f'[info] Writing video subtitles to: {sub_filename}')
4241 if sub_info.get('data') is not None:
4242 try:
4243 # Use newline='' to prevent conversion of newline characters
4244 # See https://github.com/ytdl-org/youtube-dl/issues/10268
4245 with open(sub_filename, 'w', encoding='utf-8', newline='') as subfile:
4246 subfile.write(sub_info['data'])
4247 sub_info['filepath'] = sub_filename
4248 ret.append((sub_filename, sub_filename_final))
4249 continue
4250 except OSError:
4251 self.report_error(f'Cannot write video subtitles file {sub_filename}')
4252 return None
4253
4254 try:
4255 sub_copy = sub_info.copy()
4256 sub_copy.setdefault('http_headers', info_dict.get('http_headers'))
4257 self.dl(sub_filename, sub_copy, subtitle=True)
4258 sub_info['filepath'] = sub_filename
4259 ret.append((sub_filename, sub_filename_final))
4260 except (DownloadError, ExtractorError, IOError, OSError, ValueError) + network_exceptions as err:
4261 msg = f'Unable to download video subtitles for {sub_lang!r}: {err}'
4262 if self.params.get('ignoreerrors') is not True: # False or 'only_download'
4263 if not self.params.get('ignoreerrors'):
4264 self.report_error(msg)
4265 raise DownloadError(msg)
4266 self.report_warning(msg)
4267 return ret
4268
4269 def _write_thumbnails(self, label, info_dict, filename, thumb_filename_base=None):
4270 ''' Write thumbnails to file and return list of (thumb_filename, final_thumb_filename); or None if error '''
4271 write_all = self.params.get('write_all_thumbnails', False)
4272 thumbnails, ret = [], []
4273 if write_all or self.params.get('writethumbnail', False):
4274 thumbnails = info_dict.get('thumbnails') or []
4275 if not thumbnails:
4276 self.to_screen(f'[info] There are no {label} thumbnails to download')
4277 return ret
4278 multiple = write_all and len(thumbnails) > 1
4279
4280 if thumb_filename_base is None:
4281 thumb_filename_base = filename
4282 if thumbnails and not thumb_filename_base:
4283 self.write_debug(f'Skipping writing {label} thumbnail')
4284 return ret
4285
4286 if thumbnails and not self._ensure_dir_exists(filename):
4287 return None
4288
4289 for idx, t in list(enumerate(thumbnails))[::-1]:
4290 thumb_ext = (f'{t["id"]}.' if multiple else '') + determine_ext(t['url'], 'jpg')
4291 thumb_display_id = f'{label} thumbnail {t["id"]}'
4292 thumb_filename = replace_extension(filename, thumb_ext, info_dict.get('ext'))
4293 thumb_filename_final = replace_extension(thumb_filename_base, thumb_ext, info_dict.get('ext'))
4294
4295 existing_thumb = self.existing_file((thumb_filename_final, thumb_filename))
4296 if existing_thumb:
4297 self.to_screen('[info] %s is already present' % (
4298 thumb_display_id if multiple else f'{label} thumbnail').capitalize())
4299 t['filepath'] = existing_thumb
4300 ret.append((existing_thumb, thumb_filename_final))
4301 else:
4302 self.to_screen(f'[info] Downloading {thumb_display_id} ...')
4303 try:
4304 uf = self.urlopen(Request(t['url'], headers=t.get('http_headers', {})))
4305 self.to_screen(f'[info] Writing {thumb_display_id} to: {thumb_filename}')
4306 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
4307 shutil.copyfileobj(uf, thumbf)
4308 ret.append((thumb_filename, thumb_filename_final))
4309 t['filepath'] = thumb_filename
4310 except network_exceptions as err:
4311 if isinstance(err, HTTPError) and err.status == 404:
4312 self.to_screen(f'[info] {thumb_display_id.title()} does not exist')
4313 else:
4314 self.report_warning(f'Unable to download {thumb_display_id}: {err}')
4315 thumbnails.pop(idx)
4316 if ret and not write_all:
4317 break
4318 return ret