]> jfr.im git - yt-dlp.git/blob - yt_dlp/YoutubeDL.py
[utils] `traverse_obj`: Move `is_user_input` into output template (#8673)
[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 REPOSITORY, _get_system_deprecation, _make_label, current_git_head, detect_variant
64 from .utils import (
65 DEFAULT_OUTTMPL,
66 IDENTITY,
67 LINK_TEMPLATES,
68 MEDIA_EXTENSIONS,
69 NO_DEFAULT,
70 NUMBER_RE,
71 OUTTMPL_TYPES,
72 POSTPROCESS_WHEN,
73 STR_FORMAT_RE_TMPL,
74 STR_FORMAT_TYPES,
75 ContentTooShortError,
76 DateRange,
77 DownloadCancelled,
78 DownloadError,
79 EntryNotInPlaylist,
80 ExistingVideoReached,
81 ExtractorError,
82 FormatSorter,
83 GeoRestrictedError,
84 ISO3166Utils,
85 LazyList,
86 MaxDownloadsReached,
87 Namespace,
88 PagedList,
89 PlaylistEntries,
90 Popen,
91 PostProcessingError,
92 ReExtractInfo,
93 RejectedVideoReached,
94 SameFileError,
95 UnavailableVideoError,
96 UserNotLive,
97 age_restricted,
98 args_to_str,
99 bug_reports_message,
100 date_from_str,
101 deprecation_warning,
102 determine_ext,
103 determine_protocol,
104 encode_compat_str,
105 encodeFilename,
106 error_to_compat_str,
107 escapeHTML,
108 expand_path,
109 extract_basic_auth,
110 filter_dict,
111 float_or_none,
112 format_bytes,
113 format_decimal_suffix,
114 format_field,
115 formatSeconds,
116 get_compatible_ext,
117 get_domain,
118 int_or_none,
119 iri_to_uri,
120 is_path_like,
121 join_nonempty,
122 locked_file,
123 make_archive_id,
124 make_dir,
125 number_of_digits,
126 orderedSet,
127 orderedSet_from_options,
128 parse_filesize,
129 preferredencoding,
130 prepend_extension,
131 remove_terminal_sequences,
132 render_table,
133 replace_extension,
134 sanitize_filename,
135 sanitize_path,
136 sanitize_url,
137 str_or_none,
138 strftime_or_none,
139 subtitles_filename,
140 supports_terminal_sequences,
141 system_identifier,
142 timetuple_from_msec,
143 to_high_limit_path,
144 traverse_obj,
145 try_call,
146 try_get,
147 url_basename,
148 variadic,
149 version_tuple,
150 windows_enable_vt_mode,
151 write_json_file,
152 write_string,
153 )
154 from .utils._utils import _YDLLogger
155 from .utils.networking import (
156 HTTPHeaderDict,
157 clean_headers,
158 clean_proxies,
159 std_headers,
160 )
161 from .version import CHANNEL, ORIGIN, RELEASE_GIT_HEAD, VARIANT, __version__
162
163 if compat_os_name == 'nt':
164 import ctypes
165
166
167 class YoutubeDL:
168 """YoutubeDL class.
169
170 YoutubeDL objects are the ones responsible of downloading the
171 actual video file and writing it to disk if the user has requested
172 it, among some other tasks. In most cases there should be one per
173 program. As, given a video URL, the downloader doesn't know how to
174 extract all the needed information, task that InfoExtractors do, it
175 has to pass the URL to one of them.
176
177 For this, YoutubeDL objects have a method that allows
178 InfoExtractors to be registered in a given order. When it is passed
179 a URL, the YoutubeDL object handles it to the first InfoExtractor it
180 finds that reports being able to handle it. The InfoExtractor extracts
181 all the information about the video or videos the URL refers to, and
182 YoutubeDL process the extracted information, possibly using a File
183 Downloader to download the video.
184
185 YoutubeDL objects accept a lot of parameters. In order not to saturate
186 the object constructor with arguments, it receives a dictionary of
187 options instead. These options are available through the params
188 attribute for the InfoExtractors to use. The YoutubeDL also
189 registers itself as the downloader in charge for the InfoExtractors
190 that are added to it, so this is a "mutual registration".
191
192 Available options:
193
194 username: Username for authentication purposes.
195 password: Password for authentication purposes.
196 videopassword: Password for accessing a video.
197 ap_mso: Adobe Pass multiple-system operator identifier.
198 ap_username: Multiple-system operator account username.
199 ap_password: Multiple-system operator account password.
200 usenetrc: Use netrc for authentication instead.
201 netrc_location: Location of the netrc file. Defaults to ~/.netrc.
202 netrc_cmd: Use a shell command to get credentials
203 verbose: Print additional info to stdout.
204 quiet: Do not print messages to stdout.
205 no_warnings: Do not print out anything for warnings.
206 forceprint: A dict with keys WHEN mapped to a list of templates to
207 print to stdout. The allowed keys are video or any of the
208 items in utils.POSTPROCESS_WHEN.
209 For compatibility, a single list is also accepted
210 print_to_file: A dict with keys WHEN (same as forceprint) mapped to
211 a list of tuples with (template, filename)
212 forcejson: Force printing info_dict as JSON.
213 dump_single_json: Force printing the info_dict of the whole playlist
214 (or video) as a single JSON line.
215 force_write_download_archive: Force writing download archive regardless
216 of 'skip_download' or 'simulate'.
217 simulate: Do not download the video files. If unset (or None),
218 simulate only if listsubtitles, listformats or list_thumbnails is used
219 format: Video format code. see "FORMAT SELECTION" for more details.
220 You can also pass a function. The function takes 'ctx' as
221 argument and returns the formats to download.
222 See "build_format_selector" for an implementation
223 allow_unplayable_formats: Allow unplayable formats to be extracted and downloaded.
224 ignore_no_formats_error: Ignore "No video formats" error. Usefull for
225 extracting metadata even if the video is not actually
226 available for download (experimental)
227 format_sort: A list of fields by which to sort the video formats.
228 See "Sorting Formats" for more details.
229 format_sort_force: Force the given format_sort. see "Sorting Formats"
230 for more details.
231 prefer_free_formats: Whether to prefer video formats with free containers
232 over non-free ones of same quality.
233 allow_multiple_video_streams: Allow multiple video streams to be merged
234 into a single file
235 allow_multiple_audio_streams: Allow multiple audio streams to be merged
236 into a single file
237 check_formats Whether to test if the formats are downloadable.
238 Can be True (check all), False (check none),
239 'selected' (check selected formats),
240 or None (check only if requested by extractor)
241 paths: Dictionary of output paths. The allowed keys are 'home'
242 'temp' and the keys of OUTTMPL_TYPES (in utils/_utils.py)
243 outtmpl: Dictionary of templates for output names. Allowed keys
244 are 'default' and the keys of OUTTMPL_TYPES (in utils/_utils.py).
245 For compatibility with youtube-dl, a single string can also be used
246 outtmpl_na_placeholder: Placeholder for unavailable meta fields.
247 restrictfilenames: Do not allow "&" and spaces in file names
248 trim_file_name: Limit length of filename (extension excluded)
249 windowsfilenames: Force the filenames to be windows compatible
250 ignoreerrors: Do not stop on download/postprocessing errors.
251 Can be 'only_download' to ignore only download errors.
252 Default is 'only_download' for CLI, but False for API
253 skip_playlist_after_errors: Number of allowed failures until the rest of
254 the playlist is skipped
255 allowed_extractors: List of regexes to match against extractor names that are allowed
256 overwrites: Overwrite all video and metadata files if True,
257 overwrite only non-video files if None
258 and don't overwrite any file if False
259 playlist_items: Specific indices of playlist to download.
260 playlistrandom: Download playlist items in random order.
261 lazy_playlist: Process playlist entries as they are received.
262 matchtitle: Download only matching titles.
263 rejecttitle: Reject downloads for matching titles.
264 logger: Log messages to a logging.Logger instance.
265 logtostderr: Print everything to stderr instead of stdout.
266 consoletitle: Display progress in console window's titlebar.
267 writedescription: Write the video description to a .description file
268 writeinfojson: Write the video description to a .info.json file
269 clean_infojson: Remove internal metadata from the infojson
270 getcomments: Extract video comments. This will not be written to disk
271 unless writeinfojson is also given
272 writeannotations: Write the video annotations to a .annotations.xml file
273 writethumbnail: Write the thumbnail image to a file
274 allow_playlist_files: Whether to write playlists' description, infojson etc
275 also to disk when using the 'write*' options
276 write_all_thumbnails: Write all thumbnail formats to files
277 writelink: Write an internet shortcut file, depending on the
278 current platform (.url/.webloc/.desktop)
279 writeurllink: Write a Windows internet shortcut file (.url)
280 writewebloclink: Write a macOS internet shortcut file (.webloc)
281 writedesktoplink: Write a Linux internet shortcut file (.desktop)
282 writesubtitles: Write the video subtitles to a file
283 writeautomaticsub: Write the automatically generated subtitles to a file
284 listsubtitles: Lists all available subtitles for the video
285 subtitlesformat: The format code for subtitles
286 subtitleslangs: List of languages of the subtitles to download (can be regex).
287 The list may contain "all" to refer to all the available
288 subtitles. The language can be prefixed with a "-" to
289 exclude it from the requested languages, e.g. ['all', '-live_chat']
290 keepvideo: Keep the video file after post-processing
291 daterange: A utils.DateRange object, download only if the upload_date is in the range.
292 skip_download: Skip the actual download of the video file
293 cachedir: Location of the cache files in the filesystem.
294 False to disable filesystem cache.
295 noplaylist: Download single video instead of a playlist if in doubt.
296 age_limit: An integer representing the user's age in years.
297 Unsuitable videos for the given age are skipped.
298 min_views: An integer representing the minimum view count the video
299 must have in order to not be skipped.
300 Videos without view count information are always
301 downloaded. None for no limit.
302 max_views: An integer representing the maximum view count.
303 Videos that are more popular than that are not
304 downloaded.
305 Videos without view count information are always
306 downloaded. None for no limit.
307 download_archive: A set, or the name of a file where all downloads are recorded.
308 Videos already present in the file are not downloaded again.
309 break_on_existing: Stop the download process after attempting to download a
310 file that is in the archive.
311 break_per_url: Whether break_on_reject and break_on_existing
312 should act on each input URL as opposed to for the entire queue
313 cookiefile: File name or text stream from where cookies should be read and dumped to
314 cookiesfrombrowser: A tuple containing the name of the browser, the profile
315 name/path from where cookies are loaded, the name of the keyring,
316 and the container name, e.g. ('chrome', ) or
317 ('vivaldi', 'default', 'BASICTEXT') or ('firefox', 'default', None, 'Meta')
318 legacyserverconnect: Explicitly allow HTTPS connection to servers that do not
319 support RFC 5746 secure renegotiation
320 nocheckcertificate: Do not verify SSL certificates
321 client_certificate: Path to client certificate file in PEM format. May include the private key
322 client_certificate_key: Path to private key file for client certificate
323 client_certificate_password: Password for client certificate private key, if encrypted.
324 If not provided and the key is encrypted, yt-dlp will ask interactively
325 prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
326 (Only supported by some extractors)
327 enable_file_urls: Enable file:// URLs. This is disabled by default for security reasons.
328 http_headers: A dictionary of custom headers to be used for all requests
329 proxy: URL of the proxy server to use
330 geo_verification_proxy: URL of the proxy to use for IP address verification
331 on geo-restricted sites.
332 socket_timeout: Time to wait for unresponsive hosts, in seconds
333 bidi_workaround: Work around buggy terminals without bidirectional text
334 support, using fridibi
335 debug_printtraffic:Print out sent and received HTTP traffic
336 default_search: Prepend this string if an input url is not valid.
337 'auto' for elaborate guessing
338 encoding: Use this encoding instead of the system-specified.
339 extract_flat: Whether to resolve and process url_results further
340 * False: Always process. Default for API
341 * True: Never process
342 * 'in_playlist': Do not process inside playlist/multi_video
343 * 'discard': Always process, but don't return the result
344 from inside playlist/multi_video
345 * 'discard_in_playlist': Same as "discard", but only for
346 playlists (not multi_video). Default for CLI
347 wait_for_video: If given, wait for scheduled streams to become available.
348 The value should be a tuple containing the range
349 (min_secs, max_secs) to wait between retries
350 postprocessors: A list of dictionaries, each with an entry
351 * key: The name of the postprocessor. See
352 yt_dlp/postprocessor/__init__.py for a list.
353 * when: When to run the postprocessor. Allowed values are
354 the entries of utils.POSTPROCESS_WHEN
355 Assumed to be 'post_process' if not given
356 progress_hooks: A list of functions that get called on download
357 progress, with a dictionary with the entries
358 * status: One of "downloading", "error", or "finished".
359 Check this first and ignore unknown values.
360 * info_dict: The extracted info_dict
361
362 If status is one of "downloading", or "finished", the
363 following properties may also be present:
364 * filename: The final filename (always present)
365 * tmpfilename: The filename we're currently writing to
366 * downloaded_bytes: Bytes on disk
367 * total_bytes: Size of the whole file, None if unknown
368 * total_bytes_estimate: Guess of the eventual file size,
369 None if unavailable.
370 * elapsed: The number of seconds since download started.
371 * eta: The estimated time in seconds, None if unknown
372 * speed: The download speed in bytes/second, None if
373 unknown
374 * fragment_index: The counter of the currently
375 downloaded video fragment.
376 * fragment_count: The number of fragments (= individual
377 files that will be merged)
378
379 Progress hooks are guaranteed to be called at least once
380 (with status "finished") if the download is successful.
381 postprocessor_hooks: A list of functions that get called on postprocessing
382 progress, with a dictionary with the entries
383 * status: One of "started", "processing", or "finished".
384 Check this first and ignore unknown values.
385 * postprocessor: Name of the postprocessor
386 * info_dict: The extracted info_dict
387
388 Progress hooks are guaranteed to be called at least twice
389 (with status "started" and "finished") if the processing is successful.
390 merge_output_format: "/" separated list of extensions to use when merging formats.
391 final_ext: Expected final extension; used to detect when the file was
392 already downloaded and converted
393 fixup: Automatically correct known faults of the file.
394 One of:
395 - "never": do nothing
396 - "warn": only emit a warning
397 - "detect_or_warn": check whether we can do anything
398 about it, warn otherwise (default)
399 source_address: Client-side IP address to bind to.
400 sleep_interval_requests: Number of seconds to sleep between requests
401 during extraction
402 sleep_interval: Number of seconds to sleep before each download when
403 used alone or a lower bound of a range for randomized
404 sleep before each download (minimum possible number
405 of seconds to sleep) when used along with
406 max_sleep_interval.
407 max_sleep_interval:Upper bound of a range for randomized sleep before each
408 download (maximum possible number of seconds to sleep).
409 Must only be used along with sleep_interval.
410 Actual sleep time will be a random float from range
411 [sleep_interval; max_sleep_interval].
412 sleep_interval_subtitles: Number of seconds to sleep before each subtitle download
413 listformats: Print an overview of available video formats and exit.
414 list_thumbnails: Print a table of all thumbnails and exit.
415 match_filter: A function that gets called for every video with the signature
416 (info_dict, *, incomplete: bool) -> Optional[str]
417 For backward compatibility with youtube-dl, the signature
418 (info_dict) -> Optional[str] is also allowed.
419 - If it returns a message, the video is ignored.
420 - If it returns None, the video is downloaded.
421 - If it returns utils.NO_DEFAULT, the user is interactively
422 asked whether to download the video.
423 - Raise utils.DownloadCancelled(msg) to abort remaining
424 downloads when a video is rejected.
425 match_filter_func in utils/_utils.py is one example for this.
426 color: A Dictionary with output stream names as keys
427 and their respective color policy as values.
428 Can also just be a single color policy,
429 in which case it applies to all outputs.
430 Valid stream names are 'stdout' and 'stderr'.
431 Valid color policies are one of 'always', 'auto', 'no_color' or 'never'.
432 geo_bypass: Bypass geographic restriction via faking X-Forwarded-For
433 HTTP header
434 geo_bypass_country:
435 Two-letter ISO 3166-2 country code that will be used for
436 explicit geographic restriction bypassing via faking
437 X-Forwarded-For HTTP header
438 geo_bypass_ip_block:
439 IP range in CIDR notation that will be used similarly to
440 geo_bypass_country
441 external_downloader: A dictionary of protocol keys and the executable of the
442 external downloader to use for it. The allowed protocols
443 are default|http|ftp|m3u8|dash|rtsp|rtmp|mms.
444 Set the value to 'native' to use the native downloader
445 compat_opts: Compatibility options. See "Differences in default behavior".
446 The following options do not work when used through the API:
447 filename, abort-on-error, multistreams, no-live-chat, format-sort
448 no-clean-infojson, no-playlist-metafiles, no-keep-subs, no-attach-info-json.
449 Refer __init__.py for their implementation
450 progress_template: Dictionary of templates for progress outputs.
451 Allowed keys are 'download', 'postprocess',
452 'download-title' (console title) and 'postprocess-title'.
453 The template is mapped on a dictionary with keys 'progress' and 'info'
454 retry_sleep_functions: Dictionary of functions that takes the number of attempts
455 as argument and returns the time to sleep in seconds.
456 Allowed keys are 'http', 'fragment', 'file_access'
457 download_ranges: A callback function that gets called for every video with
458 the signature (info_dict, ydl) -> Iterable[Section].
459 Only the returned sections will be downloaded.
460 Each Section is a dict with the following keys:
461 * start_time: Start time of the section in seconds
462 * end_time: End time of the section in seconds
463 * title: Section title (Optional)
464 * index: Section number (Optional)
465 force_keyframes_at_cuts: Re-encode the video when downloading ranges to get precise cuts
466 noprogress: Do not print the progress bar
467 live_from_start: Whether to download livestreams videos from the start
468
469 The following parameters are not used by YoutubeDL itself, they are used by
470 the downloader (see yt_dlp/downloader/common.py):
471 nopart, updatetime, buffersize, ratelimit, throttledratelimit, min_filesize,
472 max_filesize, test, noresizebuffer, retries, file_access_retries, fragment_retries,
473 continuedl, xattr_set_filesize, hls_use_mpegts, http_chunk_size,
474 external_downloader_args, concurrent_fragment_downloads.
475
476 The following options are used by the post processors:
477 ffmpeg_location: Location of the ffmpeg/avconv binary; either the path
478 to the binary or its containing directory.
479 postprocessor_args: A dictionary of postprocessor/executable keys (in lower case)
480 and a list of additional command-line arguments for the
481 postprocessor/executable. The dict can also have "PP+EXE" keys
482 which are used when the given exe is used by the given PP.
483 Use 'default' as the name for arguments to passed to all PP
484 For compatibility with youtube-dl, a single list of args
485 can also be used
486
487 The following options are used by the extractors:
488 extractor_retries: Number of times to retry for known errors (default: 3)
489 dynamic_mpd: Whether to process dynamic DASH manifests (default: True)
490 hls_split_discontinuity: Split HLS playlists to different formats at
491 discontinuities such as ad breaks (default: False)
492 extractor_args: A dictionary of arguments to be passed to the extractors.
493 See "EXTRACTOR ARGUMENTS" for details.
494 E.g. {'youtube': {'skip': ['dash', 'hls']}}
495 mark_watched: Mark videos watched (even with --simulate). Only for YouTube
496
497 The following options are deprecated and may be removed in the future:
498
499 break_on_reject: Stop the download process when encountering a video that
500 has been filtered out.
501 - `raise DownloadCancelled(msg)` in match_filter instead
502 force_generic_extractor: Force downloader to use the generic extractor
503 - Use allowed_extractors = ['generic', 'default']
504 playliststart: - Use playlist_items
505 Playlist item to start at.
506 playlistend: - Use playlist_items
507 Playlist item to end at.
508 playlistreverse: - Use playlist_items
509 Download playlist items in reverse order.
510 forceurl: - Use forceprint
511 Force printing final URL.
512 forcetitle: - Use forceprint
513 Force printing title.
514 forceid: - Use forceprint
515 Force printing ID.
516 forcethumbnail: - Use forceprint
517 Force printing thumbnail URL.
518 forcedescription: - Use forceprint
519 Force printing description.
520 forcefilename: - Use forceprint
521 Force printing final filename.
522 forceduration: - Use forceprint
523 Force printing duration.
524 allsubtitles: - Use subtitleslangs = ['all']
525 Downloads all the subtitles of the video
526 (requires writesubtitles or writeautomaticsub)
527 include_ads: - Doesn't work
528 Download ads as well
529 call_home: - Not implemented
530 Boolean, true iff we are allowed to contact the
531 yt-dlp servers for debugging.
532 post_hooks: - Register a custom postprocessor
533 A list of functions that get called as the final step
534 for each video file, after all postprocessors have been
535 called. The filename will be passed as the only argument.
536 hls_prefer_native: - Use external_downloader = {'m3u8': 'native'} or {'m3u8': 'ffmpeg'}.
537 Use the native HLS downloader instead of ffmpeg/avconv
538 if True, otherwise use ffmpeg/avconv if False, otherwise
539 use downloader suggested by extractor if None.
540 prefer_ffmpeg: - avconv support is deprecated
541 If False, use avconv instead of ffmpeg if both are available,
542 otherwise prefer ffmpeg.
543 youtube_include_dash_manifest: - Use extractor_args
544 If True (default), DASH manifests and related
545 data will be downloaded and processed by extractor.
546 You can reduce network I/O by disabling it if you don't
547 care about DASH. (only for youtube)
548 youtube_include_hls_manifest: - Use extractor_args
549 If True (default), HLS manifests and related
550 data will be downloaded and processed by extractor.
551 You can reduce network I/O by disabling it if you don't
552 care about HLS. (only for youtube)
553 no_color: Same as `color='no_color'`
554 no_overwrites: Same as `overwrites=False`
555 """
556
557 _NUMERIC_FIELDS = {
558 'width', 'height', 'asr', 'audio_channels', 'fps',
559 'tbr', 'abr', 'vbr', 'filesize', 'filesize_approx',
560 'timestamp', 'release_timestamp',
561 'duration', 'view_count', 'like_count', 'dislike_count', 'repost_count',
562 'average_rating', 'comment_count', 'age_limit',
563 'start_time', 'end_time',
564 'chapter_number', 'season_number', 'episode_number',
565 'track_number', 'disc_number', 'release_year',
566 }
567
568 _format_fields = {
569 # NB: Keep in sync with the docstring of extractor/common.py
570 'url', 'manifest_url', 'manifest_stream_number', 'ext', 'format', 'format_id', 'format_note',
571 'width', 'height', 'aspect_ratio', 'resolution', 'dynamic_range', 'tbr', 'abr', 'acodec', 'asr', 'audio_channels',
572 'vbr', 'fps', 'vcodec', 'container', 'filesize', 'filesize_approx', 'rows', 'columns',
573 'player_url', 'protocol', 'fragment_base_url', 'fragments', 'is_from_start',
574 'preference', 'language', 'language_preference', 'quality', 'source_preference', 'cookies',
575 'http_headers', 'stretched_ratio', 'no_resume', 'has_drm', 'extra_param_to_segment_url', 'hls_aes', 'downloader_options',
576 'page_url', 'app', 'play_path', 'tc_url', 'flash_version', 'rtmp_live', 'rtmp_conn', 'rtmp_protocol', 'rtmp_real_time'
577 }
578 _format_selection_exts = {
579 'audio': set(MEDIA_EXTENSIONS.common_audio),
580 'video': set(MEDIA_EXTENSIONS.common_video + ('3gp', )),
581 'storyboards': set(MEDIA_EXTENSIONS.storyboards),
582 }
583
584 def __init__(self, params=None, auto_init=True):
585 """Create a FileDownloader object with the given options.
586 @param auto_init Whether to load the default extractors and print header (if verbose).
587 Set to 'no_verbose_header' to not print the header
588 """
589 if params is None:
590 params = {}
591 self.params = params
592 self._ies = {}
593 self._ies_instances = {}
594 self._pps = {k: [] for k in POSTPROCESS_WHEN}
595 self._printed_messages = set()
596 self._first_webpage_request = True
597 self._post_hooks = []
598 self._progress_hooks = []
599 self._postprocessor_hooks = []
600 self._download_retcode = 0
601 self._num_downloads = 0
602 self._num_videos = 0
603 self._playlist_level = 0
604 self._playlist_urls = set()
605 self.cache = Cache(self)
606 self.__header_cookies = []
607
608 stdout = sys.stderr if self.params.get('logtostderr') else sys.stdout
609 self._out_files = Namespace(
610 out=stdout,
611 error=sys.stderr,
612 screen=sys.stderr if self.params.get('quiet') else stdout,
613 console=None if compat_os_name == 'nt' else next(
614 filter(supports_terminal_sequences, (sys.stderr, sys.stdout)), None)
615 )
616
617 try:
618 windows_enable_vt_mode()
619 except Exception as e:
620 self.write_debug(f'Failed to enable VT mode: {e}')
621
622 if self.params.get('no_color'):
623 if self.params.get('color') is not None:
624 self.params.setdefault('_warnings', []).append(
625 'Overwriting params from "color" with "no_color"')
626 self.params['color'] = 'no_color'
627
628 term_allow_color = os.getenv('TERM', '').lower() != 'dumb'
629 no_color = bool(os.getenv('NO_COLOR'))
630
631 def process_color_policy(stream):
632 stream_name = {sys.stdout: 'stdout', sys.stderr: 'stderr'}[stream]
633 policy = traverse_obj(self.params, ('color', (stream_name, None), {str}), get_all=False)
634 if policy in ('auto', None):
635 if term_allow_color and supports_terminal_sequences(stream):
636 return 'no_color' if no_color else True
637 return False
638 assert policy in ('always', 'never', 'no_color'), policy
639 return {'always': True, 'never': False}.get(policy, policy)
640
641 self._allow_colors = Namespace(**{
642 name: process_color_policy(stream)
643 for name, stream in self._out_files.items_ if name != 'console'
644 })
645
646 system_deprecation = _get_system_deprecation()
647 if system_deprecation:
648 self.deprecated_feature(system_deprecation.replace('\n', '\n '))
649
650 if self.params.get('allow_unplayable_formats'):
651 self.report_warning(
652 f'You have asked for {self._format_err("UNPLAYABLE", self.Styles.EMPHASIS)} formats to be listed/downloaded. '
653 'This is a developer option intended for debugging. \n'
654 ' If you experience any issues while using this option, '
655 f'{self._format_err("DO NOT", self.Styles.ERROR)} open a bug report')
656
657 if self.params.get('bidi_workaround', False):
658 try:
659 import pty
660 master, slave = pty.openpty()
661 width = shutil.get_terminal_size().columns
662 width_args = [] if width is None else ['-w', str(width)]
663 sp_kwargs = {'stdin': subprocess.PIPE, 'stdout': slave, 'stderr': self._out_files.error}
664 try:
665 self._output_process = Popen(['bidiv'] + width_args, **sp_kwargs)
666 except OSError:
667 self._output_process = Popen(['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
668 self._output_channel = os.fdopen(master, 'rb')
669 except OSError as ose:
670 if ose.errno == errno.ENOENT:
671 self.report_warning(
672 'Could not find fribidi executable, ignoring --bidi-workaround. '
673 'Make sure that fribidi is an executable file in one of the directories in your $PATH.')
674 else:
675 raise
676
677 self.params['compat_opts'] = set(self.params.get('compat_opts', ()))
678 self.params['http_headers'] = HTTPHeaderDict(std_headers, self.params.get('http_headers'))
679 self._load_cookies(self.params['http_headers'].get('Cookie')) # compat
680 self.params['http_headers'].pop('Cookie', None)
681 self._request_director = self.build_request_director(_REQUEST_HANDLERS.values(), _RH_PREFERENCES)
682
683 if auto_init and auto_init != 'no_verbose_header':
684 self.print_debug_header()
685
686 def check_deprecated(param, option, suggestion):
687 if self.params.get(param) is not None:
688 self.report_warning(f'{option} is deprecated. Use {suggestion} instead')
689 return True
690 return False
691
692 if check_deprecated('cn_verification_proxy', '--cn-verification-proxy', '--geo-verification-proxy'):
693 if self.params.get('geo_verification_proxy') is None:
694 self.params['geo_verification_proxy'] = self.params['cn_verification_proxy']
695
696 check_deprecated('autonumber', '--auto-number', '-o "%(autonumber)s-%(title)s.%(ext)s"')
697 check_deprecated('usetitle', '--title', '-o "%(title)s-%(id)s.%(ext)s"')
698 check_deprecated('useid', '--id', '-o "%(id)s.%(ext)s"')
699
700 for msg in self.params.get('_warnings', []):
701 self.report_warning(msg)
702 for msg in self.params.get('_deprecation_warnings', []):
703 self.deprecated_feature(msg)
704
705 if 'list-formats' in self.params['compat_opts']:
706 self.params['listformats_table'] = False
707
708 if 'overwrites' not in self.params and self.params.get('nooverwrites') is not None:
709 # nooverwrites was unnecessarily changed to overwrites
710 # in 0c3d0f51778b153f65c21906031c2e091fcfb641
711 # This ensures compatibility with both keys
712 self.params['overwrites'] = not self.params['nooverwrites']
713 elif self.params.get('overwrites') is None:
714 self.params.pop('overwrites', None)
715 else:
716 self.params['nooverwrites'] = not self.params['overwrites']
717
718 if self.params.get('simulate') is None and any((
719 self.params.get('list_thumbnails'),
720 self.params.get('listformats'),
721 self.params.get('listsubtitles'),
722 )):
723 self.params['simulate'] = 'list_only'
724
725 self.params.setdefault('forceprint', {})
726 self.params.setdefault('print_to_file', {})
727
728 # Compatibility with older syntax
729 if not isinstance(params['forceprint'], dict):
730 self.params['forceprint'] = {'video': params['forceprint']}
731
732 if auto_init:
733 self.add_default_info_extractors()
734
735 if (sys.platform != 'win32'
736 and sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968']
737 and not self.params.get('restrictfilenames', False)):
738 # Unicode filesystem API will throw errors (#1474, #13027)
739 self.report_warning(
740 'Assuming --restrict-filenames since file system encoding '
741 'cannot encode all characters. '
742 'Set the LC_ALL environment variable to fix this.')
743 self.params['restrictfilenames'] = True
744
745 self._parse_outtmpl()
746
747 # Creating format selector here allows us to catch syntax errors before the extraction
748 self.format_selector = (
749 self.params.get('format') if self.params.get('format') in (None, '-')
750 else self.params['format'] if callable(self.params['format'])
751 else self.build_format_selector(self.params['format']))
752
753 hooks = {
754 'post_hooks': self.add_post_hook,
755 'progress_hooks': self.add_progress_hook,
756 'postprocessor_hooks': self.add_postprocessor_hook,
757 }
758 for opt, fn in hooks.items():
759 for ph in self.params.get(opt, []):
760 fn(ph)
761
762 for pp_def_raw in self.params.get('postprocessors', []):
763 pp_def = dict(pp_def_raw)
764 when = pp_def.pop('when', 'post_process')
765 self.add_post_processor(
766 get_postprocessor(pp_def.pop('key'))(self, **pp_def),
767 when=when)
768
769 def preload_download_archive(fn):
770 """Preload the archive, if any is specified"""
771 archive = set()
772 if fn is None:
773 return archive
774 elif not is_path_like(fn):
775 return fn
776
777 self.write_debug(f'Loading archive file {fn!r}')
778 try:
779 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
780 for line in archive_file:
781 archive.add(line.strip())
782 except OSError as ioe:
783 if ioe.errno != errno.ENOENT:
784 raise
785 return archive
786
787 self.archive = preload_download_archive(self.params.get('download_archive'))
788
789 def warn_if_short_id(self, argv):
790 # short YouTube ID starting with dash?
791 idxs = [
792 i for i, a in enumerate(argv)
793 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
794 if idxs:
795 correct_argv = (
796 ['yt-dlp']
797 + [a for i, a in enumerate(argv) if i not in idxs]
798 + ['--'] + [argv[i] for i in idxs]
799 )
800 self.report_warning(
801 'Long argument string detected. '
802 'Use -- to separate parameters and URLs, like this:\n%s' %
803 args_to_str(correct_argv))
804
805 def add_info_extractor(self, ie):
806 """Add an InfoExtractor object to the end of the list."""
807 ie_key = ie.ie_key()
808 self._ies[ie_key] = ie
809 if not isinstance(ie, type):
810 self._ies_instances[ie_key] = ie
811 ie.set_downloader(self)
812
813 def get_info_extractor(self, ie_key):
814 """
815 Get an instance of an IE with name ie_key, it will try to get one from
816 the _ies list, if there's no instance it will create a new one and add
817 it to the extractor list.
818 """
819 ie = self._ies_instances.get(ie_key)
820 if ie is None:
821 ie = get_info_extractor(ie_key)()
822 self.add_info_extractor(ie)
823 return ie
824
825 def add_default_info_extractors(self):
826 """
827 Add the InfoExtractors returned by gen_extractors to the end of the list
828 """
829 all_ies = {ie.IE_NAME.lower(): ie for ie in gen_extractor_classes()}
830 all_ies['end'] = UnsupportedURLIE()
831 try:
832 ie_names = orderedSet_from_options(
833 self.params.get('allowed_extractors', ['default']), {
834 'all': list(all_ies),
835 'default': [name for name, ie in all_ies.items() if ie._ENABLED],
836 }, use_regex=True)
837 except re.error as e:
838 raise ValueError(f'Wrong regex for allowed_extractors: {e.pattern}')
839 for name in ie_names:
840 self.add_info_extractor(all_ies[name])
841 self.write_debug(f'Loaded {len(ie_names)} extractors')
842
843 def add_post_processor(self, pp, when='post_process'):
844 """Add a PostProcessor object to the end of the chain."""
845 assert when in POSTPROCESS_WHEN, f'Invalid when={when}'
846 self._pps[when].append(pp)
847 pp.set_downloader(self)
848
849 def add_post_hook(self, ph):
850 """Add the post hook"""
851 self._post_hooks.append(ph)
852
853 def add_progress_hook(self, ph):
854 """Add the download progress hook"""
855 self._progress_hooks.append(ph)
856
857 def add_postprocessor_hook(self, ph):
858 """Add the postprocessing progress hook"""
859 self._postprocessor_hooks.append(ph)
860 for pps in self._pps.values():
861 for pp in pps:
862 pp.add_progress_hook(ph)
863
864 def _bidi_workaround(self, message):
865 if not hasattr(self, '_output_channel'):
866 return message
867
868 assert hasattr(self, '_output_process')
869 assert isinstance(message, str)
870 line_count = message.count('\n') + 1
871 self._output_process.stdin.write((message + '\n').encode())
872 self._output_process.stdin.flush()
873 res = ''.join(self._output_channel.readline().decode()
874 for _ in range(line_count))
875 return res[:-len('\n')]
876
877 def _write_string(self, message, out=None, only_once=False):
878 if only_once:
879 if message in self._printed_messages:
880 return
881 self._printed_messages.add(message)
882 write_string(message, out=out, encoding=self.params.get('encoding'))
883
884 def to_stdout(self, message, skip_eol=False, quiet=None):
885 """Print message to stdout"""
886 if quiet is not None:
887 self.deprecation_warning('"YoutubeDL.to_stdout" no longer accepts the argument quiet. '
888 'Use "YoutubeDL.to_screen" instead')
889 if skip_eol is not False:
890 self.deprecation_warning('"YoutubeDL.to_stdout" no longer accepts the argument skip_eol. '
891 'Use "YoutubeDL.to_screen" instead')
892 self._write_string(f'{self._bidi_workaround(message)}\n', self._out_files.out)
893
894 def to_screen(self, message, skip_eol=False, quiet=None, only_once=False):
895 """Print message to screen if not in quiet mode"""
896 if self.params.get('logger'):
897 self.params['logger'].debug(message)
898 return
899 if (self.params.get('quiet') if quiet is None else quiet) and not self.params.get('verbose'):
900 return
901 self._write_string(
902 '%s%s' % (self._bidi_workaround(message), ('' if skip_eol else '\n')),
903 self._out_files.screen, only_once=only_once)
904
905 def to_stderr(self, message, only_once=False):
906 """Print message to stderr"""
907 assert isinstance(message, str)
908 if self.params.get('logger'):
909 self.params['logger'].error(message)
910 else:
911 self._write_string(f'{self._bidi_workaround(message)}\n', self._out_files.error, only_once=only_once)
912
913 def _send_console_code(self, code):
914 if compat_os_name == 'nt' or not self._out_files.console:
915 return
916 self._write_string(code, self._out_files.console)
917
918 def to_console_title(self, message):
919 if not self.params.get('consoletitle', False):
920 return
921 message = remove_terminal_sequences(message)
922 if compat_os_name == 'nt':
923 if ctypes.windll.kernel32.GetConsoleWindow():
924 # c_wchar_p() might not be necessary if `message` is
925 # already of type unicode()
926 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
927 else:
928 self._send_console_code(f'\033]0;{message}\007')
929
930 def save_console_title(self):
931 if not self.params.get('consoletitle') or self.params.get('simulate'):
932 return
933 self._send_console_code('\033[22;0t') # Save the title on stack
934
935 def restore_console_title(self):
936 if not self.params.get('consoletitle') or self.params.get('simulate'):
937 return
938 self._send_console_code('\033[23;0t') # Restore the title from stack
939
940 def __enter__(self):
941 self.save_console_title()
942 return self
943
944 def save_cookies(self):
945 if self.params.get('cookiefile') is not None:
946 self.cookiejar.save()
947
948 def __exit__(self, *args):
949 self.restore_console_title()
950 self.close()
951
952 def close(self):
953 self.save_cookies()
954 self._request_director.close()
955
956 def trouble(self, message=None, tb=None, is_error=True):
957 """Determine action to take when a download problem appears.
958
959 Depending on if the downloader has been configured to ignore
960 download errors or not, this method may throw an exception or
961 not when errors are found, after printing the message.
962
963 @param tb If given, is additional traceback information
964 @param is_error Whether to raise error according to ignorerrors
965 """
966 if message is not None:
967 self.to_stderr(message)
968 if self.params.get('verbose'):
969 if tb is None:
970 if sys.exc_info()[0]: # if .trouble has been called from an except block
971 tb = ''
972 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
973 tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
974 tb += encode_compat_str(traceback.format_exc())
975 else:
976 tb_data = traceback.format_list(traceback.extract_stack())
977 tb = ''.join(tb_data)
978 if tb:
979 self.to_stderr(tb)
980 if not is_error:
981 return
982 if not self.params.get('ignoreerrors'):
983 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
984 exc_info = sys.exc_info()[1].exc_info
985 else:
986 exc_info = sys.exc_info()
987 raise DownloadError(message, exc_info)
988 self._download_retcode = 1
989
990 Styles = Namespace(
991 HEADERS='yellow',
992 EMPHASIS='light blue',
993 FILENAME='green',
994 ID='green',
995 DELIM='blue',
996 ERROR='red',
997 BAD_FORMAT='light red',
998 WARNING='yellow',
999 SUPPRESS='light black',
1000 )
1001
1002 def _format_text(self, handle, allow_colors, text, f, fallback=None, *, test_encoding=False):
1003 text = str(text)
1004 if test_encoding:
1005 original_text = text
1006 # handle.encoding can be None. See https://github.com/yt-dlp/yt-dlp/issues/2711
1007 encoding = self.params.get('encoding') or getattr(handle, 'encoding', None) or 'ascii'
1008 text = text.encode(encoding, 'ignore').decode(encoding)
1009 if fallback is not None and text != original_text:
1010 text = fallback
1011 return format_text(text, f) if allow_colors is True else text if fallback is None else fallback
1012
1013 def _format_out(self, *args, **kwargs):
1014 return self._format_text(self._out_files.out, self._allow_colors.out, *args, **kwargs)
1015
1016 def _format_screen(self, *args, **kwargs):
1017 return self._format_text(self._out_files.screen, self._allow_colors.screen, *args, **kwargs)
1018
1019 def _format_err(self, *args, **kwargs):
1020 return self._format_text(self._out_files.error, self._allow_colors.error, *args, **kwargs)
1021
1022 def report_warning(self, message, only_once=False):
1023 '''
1024 Print the message to stderr, it will be prefixed with 'WARNING:'
1025 If stderr is a tty file the 'WARNING:' will be colored
1026 '''
1027 if self.params.get('logger') is not None:
1028 self.params['logger'].warning(message)
1029 else:
1030 if self.params.get('no_warnings'):
1031 return
1032 self.to_stderr(f'{self._format_err("WARNING:", self.Styles.WARNING)} {message}', only_once)
1033
1034 def deprecation_warning(self, message, *, stacklevel=0):
1035 deprecation_warning(
1036 message, stacklevel=stacklevel + 1, printer=self.report_error, is_error=False)
1037
1038 def deprecated_feature(self, message):
1039 if self.params.get('logger') is not None:
1040 self.params['logger'].warning(f'Deprecated Feature: {message}')
1041 self.to_stderr(f'{self._format_err("Deprecated Feature:", self.Styles.ERROR)} {message}', True)
1042
1043 def report_error(self, message, *args, **kwargs):
1044 '''
1045 Do the same as trouble, but prefixes the message with 'ERROR:', colored
1046 in red if stderr is a tty file.
1047 '''
1048 self.trouble(f'{self._format_err("ERROR:", self.Styles.ERROR)} {message}', *args, **kwargs)
1049
1050 def write_debug(self, message, only_once=False):
1051 '''Log debug message or Print message to stderr'''
1052 if not self.params.get('verbose', False):
1053 return
1054 message = f'[debug] {message}'
1055 if self.params.get('logger'):
1056 self.params['logger'].debug(message)
1057 else:
1058 self.to_stderr(message, only_once)
1059
1060 def report_file_already_downloaded(self, file_name):
1061 """Report file has already been fully downloaded."""
1062 try:
1063 self.to_screen('[download] %s has already been downloaded' % file_name)
1064 except UnicodeEncodeError:
1065 self.to_screen('[download] The file has already been downloaded')
1066
1067 def report_file_delete(self, file_name):
1068 """Report that existing file will be deleted."""
1069 try:
1070 self.to_screen('Deleting existing file %s' % file_name)
1071 except UnicodeEncodeError:
1072 self.to_screen('Deleting existing file')
1073
1074 def raise_no_formats(self, info, forced=False, *, msg=None):
1075 has_drm = info.get('_has_drm')
1076 ignored, expected = self.params.get('ignore_no_formats_error'), bool(msg)
1077 msg = msg or has_drm and 'This video is DRM protected' or 'No video formats found!'
1078 if forced or not ignored:
1079 raise ExtractorError(msg, video_id=info['id'], ie=info['extractor'],
1080 expected=has_drm or ignored or expected)
1081 else:
1082 self.report_warning(msg)
1083
1084 def parse_outtmpl(self):
1085 self.deprecation_warning('"YoutubeDL.parse_outtmpl" is deprecated and may be removed in a future version')
1086 self._parse_outtmpl()
1087 return self.params['outtmpl']
1088
1089 def _parse_outtmpl(self):
1090 sanitize = IDENTITY
1091 if self.params.get('restrictfilenames'): # Remove spaces in the default template
1092 sanitize = lambda x: x.replace(' - ', ' ').replace(' ', '-')
1093
1094 outtmpl = self.params.setdefault('outtmpl', {})
1095 if not isinstance(outtmpl, dict):
1096 self.params['outtmpl'] = outtmpl = {'default': outtmpl}
1097 outtmpl.update({k: sanitize(v) for k, v in DEFAULT_OUTTMPL.items() if outtmpl.get(k) is None})
1098
1099 def get_output_path(self, dir_type='', filename=None):
1100 paths = self.params.get('paths', {})
1101 assert isinstance(paths, dict), '"paths" parameter must be a dictionary'
1102 path = os.path.join(
1103 expand_path(paths.get('home', '').strip()),
1104 expand_path(paths.get(dir_type, '').strip()) if dir_type else '',
1105 filename or '')
1106 return sanitize_path(path, force=self.params.get('windowsfilenames'))
1107
1108 @staticmethod
1109 def _outtmpl_expandpath(outtmpl):
1110 # expand_path translates '%%' into '%' and '$$' into '$'
1111 # correspondingly that is not what we want since we need to keep
1112 # '%%' intact for template dict substitution step. Working around
1113 # with boundary-alike separator hack.
1114 sep = ''.join(random.choices(string.ascii_letters, k=32))
1115 outtmpl = outtmpl.replace('%%', f'%{sep}%').replace('$$', f'${sep}$')
1116
1117 # outtmpl should be expand_path'ed before template dict substitution
1118 # because meta fields may contain env variables we don't want to
1119 # be expanded. E.g. for outtmpl "%(title)s.%(ext)s" and
1120 # title "Hello $PATH", we don't want `$PATH` to be expanded.
1121 return expand_path(outtmpl).replace(sep, '')
1122
1123 @staticmethod
1124 def escape_outtmpl(outtmpl):
1125 ''' Escape any remaining strings like %s, %abc% etc. '''
1126 return re.sub(
1127 STR_FORMAT_RE_TMPL.format('', '(?![%(\0])'),
1128 lambda mobj: ('' if mobj.group('has_key') else '%') + mobj.group(0),
1129 outtmpl)
1130
1131 @classmethod
1132 def validate_outtmpl(cls, outtmpl):
1133 ''' @return None or Exception object '''
1134 outtmpl = re.sub(
1135 STR_FORMAT_RE_TMPL.format('[^)]*', '[ljhqBUDS]'),
1136 lambda mobj: f'{mobj.group(0)[:-1]}s',
1137 cls._outtmpl_expandpath(outtmpl))
1138 try:
1139 cls.escape_outtmpl(outtmpl) % collections.defaultdict(int)
1140 return None
1141 except ValueError as err:
1142 return err
1143
1144 @staticmethod
1145 def _copy_infodict(info_dict):
1146 info_dict = dict(info_dict)
1147 info_dict.pop('__postprocessors', None)
1148 info_dict.pop('__pending_error', None)
1149 return info_dict
1150
1151 def prepare_outtmpl(self, outtmpl, info_dict, sanitize=False):
1152 """ Make the outtmpl and info_dict suitable for substitution: ydl.escape_outtmpl(outtmpl) % info_dict
1153 @param sanitize Whether to sanitize the output as a filename.
1154 For backward compatibility, a function can also be passed
1155 """
1156
1157 info_dict.setdefault('epoch', int(time.time())) # keep epoch consistent once set
1158
1159 info_dict = self._copy_infodict(info_dict)
1160 info_dict['duration_string'] = ( # %(duration>%H-%M-%S)s is wrong if duration > 24hrs
1161 formatSeconds(info_dict['duration'], '-' if sanitize else ':')
1162 if info_dict.get('duration', None) is not None
1163 else None)
1164 info_dict['autonumber'] = int(self.params.get('autonumber_start', 1) - 1 + self._num_downloads)
1165 info_dict['video_autonumber'] = self._num_videos
1166 if info_dict.get('resolution') is None:
1167 info_dict['resolution'] = self.format_resolution(info_dict, default=None)
1168
1169 # For fields playlist_index, playlist_autonumber and autonumber convert all occurrences
1170 # of %(field)s to %(field)0Nd for backward compatibility
1171 field_size_compat_map = {
1172 'playlist_index': number_of_digits(info_dict.get('__last_playlist_index') or 0),
1173 'playlist_autonumber': number_of_digits(info_dict.get('n_entries') or 0),
1174 'autonumber': self.params.get('autonumber_size') or 5,
1175 }
1176
1177 TMPL_DICT = {}
1178 EXTERNAL_FORMAT_RE = re.compile(STR_FORMAT_RE_TMPL.format('[^)]*', f'[{STR_FORMAT_TYPES}ljhqBUDS]'))
1179 MATH_FUNCTIONS = {
1180 '+': float.__add__,
1181 '-': float.__sub__,
1182 '*': float.__mul__,
1183 }
1184 # Field is of the form key1.key2...
1185 # where keys (except first) can be string, int, slice or "{field, ...}"
1186 FIELD_INNER_RE = r'(?:\w+|%(num)s|%(num)s?(?::%(num)s?){1,2})' % {'num': r'(?:-?\d+)'}
1187 FIELD_RE = r'\w*(?:\.(?:%(inner)s|{%(field)s(?:,%(field)s)*}))*' % {
1188 'inner': FIELD_INNER_RE,
1189 'field': rf'\w*(?:\.{FIELD_INNER_RE})*'
1190 }
1191 MATH_FIELD_RE = rf'(?:{FIELD_RE}|-?{NUMBER_RE})'
1192 MATH_OPERATORS_RE = r'(?:%s)' % '|'.join(map(re.escape, MATH_FUNCTIONS.keys()))
1193 INTERNAL_FORMAT_RE = re.compile(rf'''(?xs)
1194 (?P<negate>-)?
1195 (?P<fields>{FIELD_RE})
1196 (?P<maths>(?:{MATH_OPERATORS_RE}{MATH_FIELD_RE})*)
1197 (?:>(?P<strf_format>.+?))?
1198 (?P<remaining>
1199 (?P<alternate>(?<!\\),[^|&)]+)?
1200 (?:&(?P<replacement>.*?))?
1201 (?:\|(?P<default>.*?))?
1202 )$''')
1203
1204 def _from_user_input(field):
1205 if field == ':':
1206 return ...
1207 elif ':' in field:
1208 return slice(*map(int_or_none, field.split(':')))
1209 elif int_or_none(field) is not None:
1210 return int(field)
1211 return field
1212
1213 def _traverse_infodict(fields):
1214 fields = [f for x in re.split(r'\.({.+?})\.?', fields)
1215 for f in ([x] if x.startswith('{') else x.split('.'))]
1216 for i in (0, -1):
1217 if fields and not fields[i]:
1218 fields.pop(i)
1219
1220 for i, f in enumerate(fields):
1221 if not f.startswith('{'):
1222 fields[i] = _from_user_input(f)
1223 continue
1224 assert f.endswith('}'), f'No closing brace for {f} in {fields}'
1225 fields[i] = {k: list(map(_from_user_input, k.split('.'))) for k in f[1:-1].split(',')}
1226
1227 return traverse_obj(info_dict, fields, traverse_string=True)
1228
1229 def get_value(mdict):
1230 # Object traversal
1231 value = _traverse_infodict(mdict['fields'])
1232 # Negative
1233 if mdict['negate']:
1234 value = float_or_none(value)
1235 if value is not None:
1236 value *= -1
1237 # Do maths
1238 offset_key = mdict['maths']
1239 if offset_key:
1240 value = float_or_none(value)
1241 operator = None
1242 while offset_key:
1243 item = re.match(
1244 MATH_FIELD_RE if operator else MATH_OPERATORS_RE,
1245 offset_key).group(0)
1246 offset_key = offset_key[len(item):]
1247 if operator is None:
1248 operator = MATH_FUNCTIONS[item]
1249 continue
1250 item, multiplier = (item[1:], -1) if item[0] == '-' else (item, 1)
1251 offset = float_or_none(item)
1252 if offset is None:
1253 offset = float_or_none(_traverse_infodict(item))
1254 try:
1255 value = operator(value, multiplier * offset)
1256 except (TypeError, ZeroDivisionError):
1257 return None
1258 operator = None
1259 # Datetime formatting
1260 if mdict['strf_format']:
1261 value = strftime_or_none(value, mdict['strf_format'].replace('\\,', ','))
1262
1263 # XXX: Workaround for https://github.com/yt-dlp/yt-dlp/issues/4485
1264 if sanitize and value == '':
1265 value = None
1266 return value
1267
1268 na = self.params.get('outtmpl_na_placeholder', 'NA')
1269
1270 def filename_sanitizer(key, value, restricted=self.params.get('restrictfilenames')):
1271 return sanitize_filename(str(value), restricted=restricted, is_id=(
1272 bool(re.search(r'(^|[_.])id(\.|$)', key))
1273 if 'filename-sanitization' in self.params['compat_opts']
1274 else NO_DEFAULT))
1275
1276 sanitizer = sanitize if callable(sanitize) else filename_sanitizer
1277 sanitize = bool(sanitize)
1278
1279 def _dumpjson_default(obj):
1280 if isinstance(obj, (set, LazyList)):
1281 return list(obj)
1282 return repr(obj)
1283
1284 class _ReplacementFormatter(string.Formatter):
1285 def get_field(self, field_name, args, kwargs):
1286 if field_name.isdigit():
1287 return args[0], -1
1288 raise ValueError('Unsupported field')
1289
1290 replacement_formatter = _ReplacementFormatter()
1291
1292 def create_key(outer_mobj):
1293 if not outer_mobj.group('has_key'):
1294 return outer_mobj.group(0)
1295 key = outer_mobj.group('key')
1296 mobj = re.match(INTERNAL_FORMAT_RE, key)
1297 value, replacement, default, last_field = None, None, na, ''
1298 while mobj:
1299 mobj = mobj.groupdict()
1300 default = mobj['default'] if mobj['default'] is not None else default
1301 value = get_value(mobj)
1302 last_field, replacement = mobj['fields'], mobj['replacement']
1303 if value is None and mobj['alternate']:
1304 mobj = re.match(INTERNAL_FORMAT_RE, mobj['remaining'][1:])
1305 else:
1306 break
1307
1308 if None not in (value, replacement):
1309 try:
1310 value = replacement_formatter.format(replacement, value)
1311 except ValueError:
1312 value, default = None, na
1313
1314 fmt = outer_mobj.group('format')
1315 if fmt == 's' and last_field in field_size_compat_map.keys() and isinstance(value, int):
1316 fmt = f'0{field_size_compat_map[last_field]:d}d'
1317
1318 flags = outer_mobj.group('conversion') or ''
1319 str_fmt = f'{fmt[:-1]}s'
1320 if value is None:
1321 value, fmt = default, 's'
1322 elif fmt[-1] == 'l': # list
1323 delim = '\n' if '#' in flags else ', '
1324 value, fmt = delim.join(map(str, variadic(value, allowed_types=(str, bytes)))), str_fmt
1325 elif fmt[-1] == 'j': # json
1326 value, fmt = json.dumps(
1327 value, default=_dumpjson_default,
1328 indent=4 if '#' in flags else None, ensure_ascii='+' not in flags), str_fmt
1329 elif fmt[-1] == 'h': # html
1330 value, fmt = escapeHTML(str(value)), str_fmt
1331 elif fmt[-1] == 'q': # quoted
1332 value = map(str, variadic(value) if '#' in flags else [value])
1333 value, fmt = ' '.join(map(compat_shlex_quote, value)), str_fmt
1334 elif fmt[-1] == 'B': # bytes
1335 value = f'%{str_fmt}'.encode() % str(value).encode()
1336 value, fmt = value.decode('utf-8', 'ignore'), 's'
1337 elif fmt[-1] == 'U': # unicode normalized
1338 value, fmt = unicodedata.normalize(
1339 # "+" = compatibility equivalence, "#" = NFD
1340 'NF%s%s' % ('K' if '+' in flags else '', 'D' if '#' in flags else 'C'),
1341 value), str_fmt
1342 elif fmt[-1] == 'D': # decimal suffix
1343 num_fmt, fmt = fmt[:-1].replace('#', ''), 's'
1344 value = format_decimal_suffix(value, f'%{num_fmt}f%s' if num_fmt else '%d%s',
1345 factor=1024 if '#' in flags else 1000)
1346 elif fmt[-1] == 'S': # filename sanitization
1347 value, fmt = filename_sanitizer(last_field, value, restricted='#' in flags), str_fmt
1348 elif fmt[-1] == 'c':
1349 if value:
1350 value = str(value)[0]
1351 else:
1352 fmt = str_fmt
1353 elif fmt[-1] not in 'rsa': # numeric
1354 value = float_or_none(value)
1355 if value is None:
1356 value, fmt = default, 's'
1357
1358 if sanitize:
1359 # If value is an object, sanitize might convert it to a string
1360 # So we convert it to repr first
1361 if fmt[-1] == 'r':
1362 value, fmt = repr(value), str_fmt
1363 elif fmt[-1] == 'a':
1364 value, fmt = ascii(value), str_fmt
1365 if fmt[-1] in 'csra':
1366 value = sanitizer(last_field, value)
1367
1368 key = '%s\0%s' % (key.replace('%', '%\0'), outer_mobj.group('format'))
1369 TMPL_DICT[key] = value
1370 return '{prefix}%({key}){fmt}'.format(key=key, fmt=fmt, prefix=outer_mobj.group('prefix'))
1371
1372 return EXTERNAL_FORMAT_RE.sub(create_key, outtmpl), TMPL_DICT
1373
1374 def evaluate_outtmpl(self, outtmpl, info_dict, *args, **kwargs):
1375 outtmpl, info_dict = self.prepare_outtmpl(outtmpl, info_dict, *args, **kwargs)
1376 return self.escape_outtmpl(outtmpl) % info_dict
1377
1378 def _prepare_filename(self, info_dict, *, outtmpl=None, tmpl_type=None):
1379 assert None in (outtmpl, tmpl_type), 'outtmpl and tmpl_type are mutually exclusive'
1380 if outtmpl is None:
1381 outtmpl = self.params['outtmpl'].get(tmpl_type or 'default', self.params['outtmpl']['default'])
1382 try:
1383 outtmpl = self._outtmpl_expandpath(outtmpl)
1384 filename = self.evaluate_outtmpl(outtmpl, info_dict, True)
1385 if not filename:
1386 return None
1387
1388 if tmpl_type in ('', 'temp'):
1389 final_ext, ext = self.params.get('final_ext'), info_dict.get('ext')
1390 if final_ext and ext and final_ext != ext and filename.endswith(f'.{final_ext}'):
1391 filename = replace_extension(filename, ext, final_ext)
1392 elif tmpl_type:
1393 force_ext = OUTTMPL_TYPES[tmpl_type]
1394 if force_ext:
1395 filename = replace_extension(filename, force_ext, info_dict.get('ext'))
1396
1397 # https://github.com/blackjack4494/youtube-dlc/issues/85
1398 trim_file_name = self.params.get('trim_file_name', False)
1399 if trim_file_name:
1400 no_ext, *ext = filename.rsplit('.', 2)
1401 filename = join_nonempty(no_ext[:trim_file_name], *ext, delim='.')
1402
1403 return filename
1404 except ValueError as err:
1405 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
1406 return None
1407
1408 def prepare_filename(self, info_dict, dir_type='', *, outtmpl=None, warn=False):
1409 """Generate the output filename"""
1410 if outtmpl:
1411 assert not dir_type, 'outtmpl and dir_type are mutually exclusive'
1412 dir_type = None
1413 filename = self._prepare_filename(info_dict, tmpl_type=dir_type, outtmpl=outtmpl)
1414 if not filename and dir_type not in ('', 'temp'):
1415 return ''
1416
1417 if warn:
1418 if not self.params.get('paths'):
1419 pass
1420 elif filename == '-':
1421 self.report_warning('--paths is ignored when an outputting to stdout', only_once=True)
1422 elif os.path.isabs(filename):
1423 self.report_warning('--paths is ignored since an absolute path is given in output template', only_once=True)
1424 if filename == '-' or not filename:
1425 return filename
1426
1427 return self.get_output_path(dir_type, filename)
1428
1429 def _match_entry(self, info_dict, incomplete=False, silent=False):
1430 """Returns None if the file should be downloaded"""
1431 _type = 'video' if 'playlist-match-filter' in self.params['compat_opts'] else info_dict.get('_type', 'video')
1432 assert incomplete or _type == 'video', 'Only video result can be considered complete'
1433
1434 video_title = info_dict.get('title', info_dict.get('id', 'entry'))
1435
1436 def check_filter():
1437 if _type in ('playlist', 'multi_video'):
1438 return
1439 elif _type in ('url', 'url_transparent') and not try_call(
1440 lambda: self.get_info_extractor(info_dict['ie_key']).is_single_video(info_dict['url'])):
1441 return
1442
1443 if 'title' in info_dict:
1444 # This can happen when we're just evaluating the playlist
1445 title = info_dict['title']
1446 matchtitle = self.params.get('matchtitle', False)
1447 if matchtitle:
1448 if not re.search(matchtitle, title, re.IGNORECASE):
1449 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
1450 rejecttitle = self.params.get('rejecttitle', False)
1451 if rejecttitle:
1452 if re.search(rejecttitle, title, re.IGNORECASE):
1453 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
1454
1455 date = info_dict.get('upload_date')
1456 if date is not None:
1457 dateRange = self.params.get('daterange', DateRange())
1458 if date not in dateRange:
1459 return f'{date_from_str(date).isoformat()} upload date is not in range {dateRange}'
1460 view_count = info_dict.get('view_count')
1461 if view_count is not None:
1462 min_views = self.params.get('min_views')
1463 if min_views is not None and view_count < min_views:
1464 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
1465 max_views = self.params.get('max_views')
1466 if max_views is not None and view_count > max_views:
1467 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
1468 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
1469 return 'Skipping "%s" because it is age restricted' % video_title
1470
1471 match_filter = self.params.get('match_filter')
1472 if match_filter is None:
1473 return None
1474
1475 cancelled = None
1476 try:
1477 try:
1478 ret = match_filter(info_dict, incomplete=incomplete)
1479 except TypeError:
1480 # For backward compatibility
1481 ret = None if incomplete else match_filter(info_dict)
1482 except DownloadCancelled as err:
1483 if err.msg is not NO_DEFAULT:
1484 raise
1485 ret, cancelled = err.msg, err
1486
1487 if ret is NO_DEFAULT:
1488 while True:
1489 filename = self._format_screen(self.prepare_filename(info_dict), self.Styles.FILENAME)
1490 reply = input(self._format_screen(
1491 f'Download "{filename}"? (Y/n): ', self.Styles.EMPHASIS)).lower().strip()
1492 if reply in {'y', ''}:
1493 return None
1494 elif reply == 'n':
1495 if cancelled:
1496 raise type(cancelled)(f'Skipping {video_title}')
1497 return f'Skipping {video_title}'
1498 return ret
1499
1500 if self.in_download_archive(info_dict):
1501 reason = ''.join((
1502 format_field(info_dict, 'id', f'{self._format_screen("%s", self.Styles.ID)}: '),
1503 format_field(info_dict, 'title', f'{self._format_screen("%s", self.Styles.EMPHASIS)} '),
1504 'has already been recorded in the archive'))
1505 break_opt, break_err = 'break_on_existing', ExistingVideoReached
1506 else:
1507 try:
1508 reason = check_filter()
1509 except DownloadCancelled as e:
1510 reason, break_opt, break_err = e.msg, 'match_filter', type(e)
1511 else:
1512 break_opt, break_err = 'break_on_reject', RejectedVideoReached
1513 if reason is not None:
1514 if not silent:
1515 self.to_screen('[download] ' + reason)
1516 if self.params.get(break_opt, False):
1517 raise break_err()
1518 return reason
1519
1520 @staticmethod
1521 def add_extra_info(info_dict, extra_info):
1522 '''Set the keys from extra_info in info dict if they are missing'''
1523 for key, value in extra_info.items():
1524 info_dict.setdefault(key, value)
1525
1526 def extract_info(self, url, download=True, ie_key=None, extra_info=None,
1527 process=True, force_generic_extractor=False):
1528 """
1529 Extract and return the information dictionary of the URL
1530
1531 Arguments:
1532 @param url URL to extract
1533
1534 Keyword arguments:
1535 @param download Whether to download videos
1536 @param process Whether to resolve all unresolved references (URLs, playlist items).
1537 Must be True for download to work
1538 @param ie_key Use only the extractor with this key
1539
1540 @param extra_info Dictionary containing the extra values to add to the info (For internal use only)
1541 @force_generic_extractor Force using the generic extractor (Deprecated; use ie_key='Generic')
1542 """
1543
1544 if extra_info is None:
1545 extra_info = {}
1546
1547 if not ie_key and force_generic_extractor:
1548 ie_key = 'Generic'
1549
1550 if ie_key:
1551 ies = {ie_key: self._ies[ie_key]} if ie_key in self._ies else {}
1552 else:
1553 ies = self._ies
1554
1555 for key, ie in ies.items():
1556 if not ie.suitable(url):
1557 continue
1558
1559 if not ie.working():
1560 self.report_warning('The program functionality for this site has been marked as broken, '
1561 'and will probably not work.')
1562
1563 temp_id = ie.get_temp_id(url)
1564 if temp_id is not None and self.in_download_archive({'id': temp_id, 'ie_key': key}):
1565 self.to_screen(f'[download] {self._format_screen(temp_id, self.Styles.ID)}: '
1566 'has already been recorded in the archive')
1567 if self.params.get('break_on_existing', False):
1568 raise ExistingVideoReached()
1569 break
1570 return self.__extract_info(url, self.get_info_extractor(key), download, extra_info, process)
1571 else:
1572 extractors_restricted = self.params.get('allowed_extractors') not in (None, ['default'])
1573 self.report_error(f'No suitable extractor{format_field(ie_key, None, " (%s)")} found for URL {url}',
1574 tb=False if extractors_restricted else None)
1575
1576 def _handle_extraction_exceptions(func):
1577 @functools.wraps(func)
1578 def wrapper(self, *args, **kwargs):
1579 while True:
1580 try:
1581 return func(self, *args, **kwargs)
1582 except (DownloadCancelled, LazyList.IndexError, PagedList.IndexError):
1583 raise
1584 except ReExtractInfo as e:
1585 if e.expected:
1586 self.to_screen(f'{e}; Re-extracting data')
1587 else:
1588 self.to_stderr('\r')
1589 self.report_warning(f'{e}; Re-extracting data')
1590 continue
1591 except GeoRestrictedError as e:
1592 msg = e.msg
1593 if e.countries:
1594 msg += '\nThis video is available in %s.' % ', '.join(
1595 map(ISO3166Utils.short2full, e.countries))
1596 msg += '\nYou might want to use a VPN or a proxy server (with --proxy) to workaround.'
1597 self.report_error(msg)
1598 except ExtractorError as e: # An error we somewhat expected
1599 self.report_error(str(e), e.format_traceback())
1600 except Exception as e:
1601 if self.params.get('ignoreerrors'):
1602 self.report_error(str(e), tb=encode_compat_str(traceback.format_exc()))
1603 else:
1604 raise
1605 break
1606 return wrapper
1607
1608 def _wait_for_video(self, ie_result={}):
1609 if (not self.params.get('wait_for_video')
1610 or ie_result.get('_type', 'video') != 'video'
1611 or ie_result.get('formats') or ie_result.get('url')):
1612 return
1613
1614 format_dur = lambda dur: '%02d:%02d:%02d' % timetuple_from_msec(dur * 1000)[:-1]
1615 last_msg = ''
1616
1617 def progress(msg):
1618 nonlocal last_msg
1619 full_msg = f'{msg}\n'
1620 if not self.params.get('noprogress'):
1621 full_msg = msg + ' ' * (len(last_msg) - len(msg)) + '\r'
1622 elif last_msg:
1623 return
1624 self.to_screen(full_msg, skip_eol=True)
1625 last_msg = msg
1626
1627 min_wait, max_wait = self.params.get('wait_for_video')
1628 diff = try_get(ie_result, lambda x: x['release_timestamp'] - time.time())
1629 if diff is None and ie_result.get('live_status') == 'is_upcoming':
1630 diff = round(random.uniform(min_wait, max_wait) if (max_wait and min_wait) else (max_wait or min_wait), 0)
1631 self.report_warning('Release time of video is not known')
1632 elif ie_result and (diff or 0) <= 0:
1633 self.report_warning('Video should already be available according to extracted info')
1634 diff = min(max(diff or 0, min_wait or 0), max_wait or float('inf'))
1635 self.to_screen(f'[wait] Waiting for {format_dur(diff)} - Press Ctrl+C to try now')
1636
1637 wait_till = time.time() + diff
1638 try:
1639 while True:
1640 diff = wait_till - time.time()
1641 if diff <= 0:
1642 progress('')
1643 raise ReExtractInfo('[wait] Wait period ended', expected=True)
1644 progress(f'[wait] Remaining time until next attempt: {self._format_screen(format_dur(diff), self.Styles.EMPHASIS)}')
1645 time.sleep(1)
1646 except KeyboardInterrupt:
1647 progress('')
1648 raise ReExtractInfo('[wait] Interrupted by user', expected=True)
1649 except BaseException as e:
1650 if not isinstance(e, ReExtractInfo):
1651 self.to_screen('')
1652 raise
1653
1654 def _load_cookies(self, data, *, autoscope=True):
1655 """Loads cookies from a `Cookie` header
1656
1657 This tries to work around the security vulnerability of passing cookies to every domain.
1658 See: https://github.com/yt-dlp/yt-dlp/security/advisories/GHSA-v8mc-9377-rwjj
1659
1660 @param data The Cookie header as string to load the cookies from
1661 @param autoscope If `False`, scope cookies using Set-Cookie syntax and error for cookie without domains
1662 If `True`, save cookies for later to be stored in the jar with a limited scope
1663 If a URL, save cookies in the jar with the domain of the URL
1664 """
1665 for cookie in LenientSimpleCookie(data).values():
1666 if autoscope and any(cookie.values()):
1667 raise ValueError('Invalid syntax in Cookie Header')
1668
1669 domain = cookie.get('domain') or ''
1670 expiry = cookie.get('expires')
1671 if expiry == '': # 0 is valid
1672 expiry = None
1673 prepared_cookie = http.cookiejar.Cookie(
1674 cookie.get('version') or 0, cookie.key, cookie.value, None, False,
1675 domain, True, True, cookie.get('path') or '', bool(cookie.get('path')),
1676 cookie.get('secure') or False, expiry, False, None, None, {})
1677
1678 if domain:
1679 self.cookiejar.set_cookie(prepared_cookie)
1680 elif autoscope is True:
1681 self.deprecated_feature(
1682 'Passing cookies as a header is a potential security risk; '
1683 'they will be scoped to the domain of the downloaded urls. '
1684 'Please consider loading cookies from a file or browser instead.')
1685 self.__header_cookies.append(prepared_cookie)
1686 elif autoscope:
1687 self.report_warning(
1688 'The extractor result contains an unscoped cookie as an HTTP header. '
1689 f'If you are using yt-dlp with an input URL{bug_reports_message(before=",")}',
1690 only_once=True)
1691 self._apply_header_cookies(autoscope, [prepared_cookie])
1692 else:
1693 self.report_error('Unscoped cookies are not allowed; please specify some sort of scoping',
1694 tb=False, is_error=False)
1695
1696 def _apply_header_cookies(self, url, cookies=None):
1697 """Applies stray header cookies to the provided url
1698
1699 This loads header cookies and scopes them to the domain provided in `url`.
1700 While this is not ideal, it helps reduce the risk of them being sent
1701 to an unintended destination while mostly maintaining compatibility.
1702 """
1703 parsed = urllib.parse.urlparse(url)
1704 if not parsed.hostname:
1705 return
1706
1707 for cookie in map(copy.copy, cookies or self.__header_cookies):
1708 cookie.domain = f'.{parsed.hostname}'
1709 self.cookiejar.set_cookie(cookie)
1710
1711 @_handle_extraction_exceptions
1712 def __extract_info(self, url, ie, download, extra_info, process):
1713 self._apply_header_cookies(url)
1714
1715 try:
1716 ie_result = ie.extract(url)
1717 except UserNotLive as e:
1718 if process:
1719 if self.params.get('wait_for_video'):
1720 self.report_warning(e)
1721 self._wait_for_video()
1722 raise
1723 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
1724 self.report_warning(f'Extractor {ie.IE_NAME} returned nothing{bug_reports_message()}')
1725 return
1726 if isinstance(ie_result, list):
1727 # Backwards compatibility: old IE result format
1728 ie_result = {
1729 '_type': 'compat_list',
1730 'entries': ie_result,
1731 }
1732 if extra_info.get('original_url'):
1733 ie_result.setdefault('original_url', extra_info['original_url'])
1734 self.add_default_extra_info(ie_result, ie, url)
1735 if process:
1736 self._wait_for_video(ie_result)
1737 return self.process_ie_result(ie_result, download, extra_info)
1738 else:
1739 return ie_result
1740
1741 def add_default_extra_info(self, ie_result, ie, url):
1742 if url is not None:
1743 self.add_extra_info(ie_result, {
1744 'webpage_url': url,
1745 'original_url': url,
1746 })
1747 webpage_url = ie_result.get('webpage_url')
1748 if webpage_url:
1749 self.add_extra_info(ie_result, {
1750 'webpage_url_basename': url_basename(webpage_url),
1751 'webpage_url_domain': get_domain(webpage_url),
1752 })
1753 if ie is not None:
1754 self.add_extra_info(ie_result, {
1755 'extractor': ie.IE_NAME,
1756 'extractor_key': ie.ie_key(),
1757 })
1758
1759 def process_ie_result(self, ie_result, download=True, extra_info=None):
1760 """
1761 Take the result of the ie(may be modified) and resolve all unresolved
1762 references (URLs, playlist items).
1763
1764 It will also download the videos if 'download'.
1765 Returns the resolved ie_result.
1766 """
1767 if extra_info is None:
1768 extra_info = {}
1769 result_type = ie_result.get('_type', 'video')
1770
1771 if result_type in ('url', 'url_transparent'):
1772 ie_result['url'] = sanitize_url(
1773 ie_result['url'], scheme='http' if self.params.get('prefer_insecure') else 'https')
1774 if ie_result.get('original_url') and not extra_info.get('original_url'):
1775 extra_info = {'original_url': ie_result['original_url'], **extra_info}
1776
1777 extract_flat = self.params.get('extract_flat', False)
1778 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info)
1779 or extract_flat is True):
1780 info_copy = ie_result.copy()
1781 ie = try_get(ie_result.get('ie_key'), self.get_info_extractor)
1782 if ie and not ie_result.get('id'):
1783 info_copy['id'] = ie.get_temp_id(ie_result['url'])
1784 self.add_default_extra_info(info_copy, ie, ie_result['url'])
1785 self.add_extra_info(info_copy, extra_info)
1786 info_copy, _ = self.pre_process(info_copy)
1787 self._fill_common_fields(info_copy, False)
1788 self.__forced_printings(info_copy)
1789 self._raise_pending_errors(info_copy)
1790 if self.params.get('force_write_download_archive', False):
1791 self.record_download_archive(info_copy)
1792 return ie_result
1793
1794 if result_type == 'video':
1795 self.add_extra_info(ie_result, extra_info)
1796 ie_result = self.process_video_result(ie_result, download=download)
1797 self._raise_pending_errors(ie_result)
1798 additional_urls = (ie_result or {}).get('additional_urls')
1799 if additional_urls:
1800 # TODO: Improve MetadataParserPP to allow setting a list
1801 if isinstance(additional_urls, str):
1802 additional_urls = [additional_urls]
1803 self.to_screen(
1804 '[info] %s: %d additional URL(s) requested' % (ie_result['id'], len(additional_urls)))
1805 self.write_debug('Additional URLs: "%s"' % '", "'.join(additional_urls))
1806 ie_result['additional_entries'] = [
1807 self.extract_info(
1808 url, download, extra_info=extra_info,
1809 force_generic_extractor=self.params.get('force_generic_extractor'))
1810 for url in additional_urls
1811 ]
1812 return ie_result
1813 elif result_type == 'url':
1814 # We have to add extra_info to the results because it may be
1815 # contained in a playlist
1816 return self.extract_info(
1817 ie_result['url'], download,
1818 ie_key=ie_result.get('ie_key'),
1819 extra_info=extra_info)
1820 elif result_type == 'url_transparent':
1821 # Use the information from the embedding page
1822 info = self.extract_info(
1823 ie_result['url'], ie_key=ie_result.get('ie_key'),
1824 extra_info=extra_info, download=False, process=False)
1825
1826 # extract_info may return None when ignoreerrors is enabled and
1827 # extraction failed with an error, don't crash and return early
1828 # in this case
1829 if not info:
1830 return info
1831
1832 exempted_fields = {'_type', 'url', 'ie_key'}
1833 if not ie_result.get('section_end') and ie_result.get('section_start') is None:
1834 # For video clips, the id etc of the clip extractor should be used
1835 exempted_fields |= {'id', 'extractor', 'extractor_key'}
1836
1837 new_result = info.copy()
1838 new_result.update(filter_dict(ie_result, lambda k, v: v is not None and k not in exempted_fields))
1839
1840 # Extracted info may not be a video result (i.e.
1841 # info.get('_type', 'video') != video) but rather an url or
1842 # url_transparent. In such cases outer metadata (from ie_result)
1843 # should be propagated to inner one (info). For this to happen
1844 # _type of info should be overridden with url_transparent. This
1845 # fixes issue from https://github.com/ytdl-org/youtube-dl/pull/11163.
1846 if new_result.get('_type') == 'url':
1847 new_result['_type'] = 'url_transparent'
1848
1849 return self.process_ie_result(
1850 new_result, download=download, extra_info=extra_info)
1851 elif result_type in ('playlist', 'multi_video'):
1852 # Protect from infinite recursion due to recursively nested playlists
1853 # (see https://github.com/ytdl-org/youtube-dl/issues/27833)
1854 webpage_url = ie_result.get('webpage_url') # Playlists maynot have webpage_url
1855 if webpage_url and webpage_url in self._playlist_urls:
1856 self.to_screen(
1857 '[download] Skipping already downloaded playlist: %s'
1858 % ie_result.get('title') or ie_result.get('id'))
1859 return
1860
1861 self._playlist_level += 1
1862 self._playlist_urls.add(webpage_url)
1863 self._fill_common_fields(ie_result, False)
1864 self._sanitize_thumbnails(ie_result)
1865 try:
1866 return self.__process_playlist(ie_result, download)
1867 finally:
1868 self._playlist_level -= 1
1869 if not self._playlist_level:
1870 self._playlist_urls.clear()
1871 elif result_type == 'compat_list':
1872 self.report_warning(
1873 'Extractor %s returned a compat_list result. '
1874 'It needs to be updated.' % ie_result.get('extractor'))
1875
1876 def _fixup(r):
1877 self.add_extra_info(r, {
1878 'extractor': ie_result['extractor'],
1879 'webpage_url': ie_result['webpage_url'],
1880 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1881 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1882 'extractor_key': ie_result['extractor_key'],
1883 })
1884 return r
1885 ie_result['entries'] = [
1886 self.process_ie_result(_fixup(r), download, extra_info)
1887 for r in ie_result['entries']
1888 ]
1889 return ie_result
1890 else:
1891 raise Exception('Invalid result type: %s' % result_type)
1892
1893 def _ensure_dir_exists(self, path):
1894 return make_dir(path, self.report_error)
1895
1896 @staticmethod
1897 def _playlist_infodict(ie_result, strict=False, **kwargs):
1898 info = {
1899 'playlist_count': ie_result.get('playlist_count'),
1900 'playlist': ie_result.get('title') or ie_result.get('id'),
1901 'playlist_id': ie_result.get('id'),
1902 'playlist_title': ie_result.get('title'),
1903 'playlist_uploader': ie_result.get('uploader'),
1904 'playlist_uploader_id': ie_result.get('uploader_id'),
1905 **kwargs,
1906 }
1907 if strict:
1908 return info
1909 if ie_result.get('webpage_url'):
1910 info.update({
1911 'webpage_url': ie_result['webpage_url'],
1912 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1913 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1914 })
1915 return {
1916 **info,
1917 'playlist_index': 0,
1918 '__last_playlist_index': max(ie_result.get('requested_entries') or (0, 0)),
1919 'extractor': ie_result['extractor'],
1920 'extractor_key': ie_result['extractor_key'],
1921 }
1922
1923 def __process_playlist(self, ie_result, download):
1924 """Process each entry in the playlist"""
1925 assert ie_result['_type'] in ('playlist', 'multi_video')
1926
1927 common_info = self._playlist_infodict(ie_result, strict=True)
1928 title = common_info.get('playlist') or '<Untitled>'
1929 if self._match_entry(common_info, incomplete=True) is not None:
1930 return
1931 self.to_screen(f'[download] Downloading {ie_result["_type"]}: {title}')
1932
1933 all_entries = PlaylistEntries(self, ie_result)
1934 entries = orderedSet(all_entries.get_requested_items(), lazy=True)
1935
1936 lazy = self.params.get('lazy_playlist')
1937 if lazy:
1938 resolved_entries, n_entries = [], 'N/A'
1939 ie_result['requested_entries'], ie_result['entries'] = None, None
1940 else:
1941 entries = resolved_entries = list(entries)
1942 n_entries = len(resolved_entries)
1943 ie_result['requested_entries'], ie_result['entries'] = tuple(zip(*resolved_entries)) or ([], [])
1944 if not ie_result.get('playlist_count'):
1945 # Better to do this after potentially exhausting entries
1946 ie_result['playlist_count'] = all_entries.get_full_count()
1947
1948 extra = self._playlist_infodict(ie_result, n_entries=int_or_none(n_entries))
1949 ie_copy = collections.ChainMap(ie_result, extra)
1950
1951 _infojson_written = False
1952 write_playlist_files = self.params.get('allow_playlist_files', True)
1953 if write_playlist_files and self.params.get('list_thumbnails'):
1954 self.list_thumbnails(ie_result)
1955 if write_playlist_files and not self.params.get('simulate'):
1956 _infojson_written = self._write_info_json(
1957 'playlist', ie_result, self.prepare_filename(ie_copy, 'pl_infojson'))
1958 if _infojson_written is None:
1959 return
1960 if self._write_description('playlist', ie_result,
1961 self.prepare_filename(ie_copy, 'pl_description')) is None:
1962 return
1963 # TODO: This should be passed to ThumbnailsConvertor if necessary
1964 self._write_thumbnails('playlist', ie_result, self.prepare_filename(ie_copy, 'pl_thumbnail'))
1965
1966 if lazy:
1967 if self.params.get('playlistreverse') or self.params.get('playlistrandom'):
1968 self.report_warning('playlistreverse and playlistrandom are not supported with lazy_playlist', only_once=True)
1969 elif self.params.get('playlistreverse'):
1970 entries.reverse()
1971 elif self.params.get('playlistrandom'):
1972 random.shuffle(entries)
1973
1974 self.to_screen(f'[{ie_result["extractor"]}] Playlist {title}: Downloading {n_entries} items'
1975 f'{format_field(ie_result, "playlist_count", " of %s")}')
1976
1977 keep_resolved_entries = self.params.get('extract_flat') != 'discard'
1978 if self.params.get('extract_flat') == 'discard_in_playlist':
1979 keep_resolved_entries = ie_result['_type'] != 'playlist'
1980 if keep_resolved_entries:
1981 self.write_debug('The information of all playlist entries will be held in memory')
1982
1983 failures = 0
1984 max_failures = self.params.get('skip_playlist_after_errors') or float('inf')
1985 for i, (playlist_index, entry) in enumerate(entries):
1986 if lazy:
1987 resolved_entries.append((playlist_index, entry))
1988 if not entry:
1989 continue
1990
1991 entry['__x_forwarded_for_ip'] = ie_result.get('__x_forwarded_for_ip')
1992 if not lazy and 'playlist-index' in self.params['compat_opts']:
1993 playlist_index = ie_result['requested_entries'][i]
1994
1995 entry_copy = collections.ChainMap(entry, {
1996 **common_info,
1997 'n_entries': int_or_none(n_entries),
1998 'playlist_index': playlist_index,
1999 'playlist_autonumber': i + 1,
2000 })
2001
2002 if self._match_entry(entry_copy, incomplete=True) is not None:
2003 # For compatabilty with youtube-dl. See https://github.com/yt-dlp/yt-dlp/issues/4369
2004 resolved_entries[i] = (playlist_index, NO_DEFAULT)
2005 continue
2006
2007 self.to_screen('[download] Downloading item %s of %s' % (
2008 self._format_screen(i + 1, self.Styles.ID), self._format_screen(n_entries, self.Styles.EMPHASIS)))
2009
2010 entry_result = self.__process_iterable_entry(entry, download, collections.ChainMap({
2011 'playlist_index': playlist_index,
2012 'playlist_autonumber': i + 1,
2013 }, extra))
2014 if not entry_result:
2015 failures += 1
2016 if failures >= max_failures:
2017 self.report_error(
2018 f'Skipping the remaining entries in playlist "{title}" since {failures} items failed extraction')
2019 break
2020 if keep_resolved_entries:
2021 resolved_entries[i] = (playlist_index, entry_result)
2022
2023 # Update with processed data
2024 ie_result['entries'] = [e for _, e in resolved_entries if e is not NO_DEFAULT]
2025 ie_result['requested_entries'] = [i for i, e in resolved_entries if e is not NO_DEFAULT]
2026 if ie_result['requested_entries'] == try_call(lambda: list(range(1, ie_result['playlist_count'] + 1))):
2027 # Do not set for full playlist
2028 ie_result.pop('requested_entries')
2029
2030 # Write the updated info to json
2031 if _infojson_written is True and self._write_info_json(
2032 'updated playlist', ie_result,
2033 self.prepare_filename(ie_copy, 'pl_infojson'), overwrite=True) is None:
2034 return
2035
2036 ie_result = self.run_all_pps('playlist', ie_result)
2037 self.to_screen(f'[download] Finished downloading playlist: {title}')
2038 return ie_result
2039
2040 @_handle_extraction_exceptions
2041 def __process_iterable_entry(self, entry, download, extra_info):
2042 return self.process_ie_result(
2043 entry, download=download, extra_info=extra_info)
2044
2045 def _build_format_filter(self, filter_spec):
2046 " Returns a function to filter the formats according to the filter_spec "
2047
2048 OPERATORS = {
2049 '<': operator.lt,
2050 '<=': operator.le,
2051 '>': operator.gt,
2052 '>=': operator.ge,
2053 '=': operator.eq,
2054 '!=': operator.ne,
2055 }
2056 operator_rex = re.compile(r'''(?x)\s*
2057 (?P<key>[\w.-]+)\s*
2058 (?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
2059 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)\s*
2060 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
2061 m = operator_rex.fullmatch(filter_spec)
2062 if m:
2063 try:
2064 comparison_value = int(m.group('value'))
2065 except ValueError:
2066 comparison_value = parse_filesize(m.group('value'))
2067 if comparison_value is None:
2068 comparison_value = parse_filesize(m.group('value') + 'B')
2069 if comparison_value is None:
2070 raise ValueError(
2071 'Invalid value %r in format specification %r' % (
2072 m.group('value'), filter_spec))
2073 op = OPERATORS[m.group('op')]
2074
2075 if not m:
2076 STR_OPERATORS = {
2077 '=': operator.eq,
2078 '^=': lambda attr, value: attr.startswith(value),
2079 '$=': lambda attr, value: attr.endswith(value),
2080 '*=': lambda attr, value: value in attr,
2081 '~=': lambda attr, value: value.search(attr) is not None
2082 }
2083 str_operator_rex = re.compile(r'''(?x)\s*
2084 (?P<key>[a-zA-Z0-9._-]+)\s*
2085 (?P<negation>!\s*)?(?P<op>%s)\s*(?P<none_inclusive>\?\s*)?
2086 (?P<quote>["'])?
2087 (?P<value>(?(quote)(?:(?!(?P=quote))[^\\]|\\.)+|[\w.-]+))
2088 (?(quote)(?P=quote))\s*
2089 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
2090 m = str_operator_rex.fullmatch(filter_spec)
2091 if m:
2092 if m.group('op') == '~=':
2093 comparison_value = re.compile(m.group('value'))
2094 else:
2095 comparison_value = re.sub(r'''\\([\\"'])''', r'\1', m.group('value'))
2096 str_op = STR_OPERATORS[m.group('op')]
2097 if m.group('negation'):
2098 op = lambda attr, value: not str_op(attr, value)
2099 else:
2100 op = str_op
2101
2102 if not m:
2103 raise SyntaxError('Invalid filter specification %r' % filter_spec)
2104
2105 def _filter(f):
2106 actual_value = f.get(m.group('key'))
2107 if actual_value is None:
2108 return m.group('none_inclusive')
2109 return op(actual_value, comparison_value)
2110 return _filter
2111
2112 def _check_formats(self, formats):
2113 for f in formats:
2114 self.to_screen('[info] Testing format %s' % f['format_id'])
2115 path = self.get_output_path('temp')
2116 if not self._ensure_dir_exists(f'{path}/'):
2117 continue
2118 temp_file = tempfile.NamedTemporaryFile(suffix='.tmp', delete=False, dir=path or None)
2119 temp_file.close()
2120 try:
2121 success, _ = self.dl(temp_file.name, f, test=True)
2122 except (DownloadError, OSError, ValueError) + network_exceptions:
2123 success = False
2124 finally:
2125 if os.path.exists(temp_file.name):
2126 try:
2127 os.remove(temp_file.name)
2128 except OSError:
2129 self.report_warning('Unable to delete temporary file "%s"' % temp_file.name)
2130 if success:
2131 yield f
2132 else:
2133 self.to_screen('[info] Unable to download format %s. Skipping...' % f['format_id'])
2134
2135 def _default_format_spec(self, info_dict, download=True):
2136
2137 def can_merge():
2138 merger = FFmpegMergerPP(self)
2139 return merger.available and merger.can_merge()
2140
2141 prefer_best = (
2142 not self.params.get('simulate')
2143 and download
2144 and (
2145 not can_merge()
2146 or info_dict.get('is_live') and not self.params.get('live_from_start')
2147 or self.params['outtmpl']['default'] == '-'))
2148 compat = (
2149 prefer_best
2150 or self.params.get('allow_multiple_audio_streams', False)
2151 or 'format-spec' in self.params['compat_opts'])
2152
2153 return (
2154 'best/bestvideo+bestaudio' if prefer_best
2155 else 'bestvideo*+bestaudio/best' if not compat
2156 else 'bestvideo+bestaudio/best')
2157
2158 def build_format_selector(self, format_spec):
2159 def syntax_error(note, start):
2160 message = (
2161 'Invalid format specification: '
2162 '{}\n\t{}\n\t{}^'.format(note, format_spec, ' ' * start[1]))
2163 return SyntaxError(message)
2164
2165 PICKFIRST = 'PICKFIRST'
2166 MERGE = 'MERGE'
2167 SINGLE = 'SINGLE'
2168 GROUP = 'GROUP'
2169 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
2170
2171 allow_multiple_streams = {'audio': self.params.get('allow_multiple_audio_streams', False),
2172 'video': self.params.get('allow_multiple_video_streams', False)}
2173
2174 def _parse_filter(tokens):
2175 filter_parts = []
2176 for type, string_, start, _, _ in tokens:
2177 if type == tokenize.OP and string_ == ']':
2178 return ''.join(filter_parts)
2179 else:
2180 filter_parts.append(string_)
2181
2182 def _remove_unused_ops(tokens):
2183 # Remove operators that we don't use and join them with the surrounding strings.
2184 # E.g. 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
2185 ALLOWED_OPS = ('/', '+', ',', '(', ')')
2186 last_string, last_start, last_end, last_line = None, None, None, None
2187 for type, string_, start, end, line in tokens:
2188 if type == tokenize.OP and string_ == '[':
2189 if last_string:
2190 yield tokenize.NAME, last_string, last_start, last_end, last_line
2191 last_string = None
2192 yield type, string_, start, end, line
2193 # everything inside brackets will be handled by _parse_filter
2194 for type, string_, start, end, line in tokens:
2195 yield type, string_, start, end, line
2196 if type == tokenize.OP and string_ == ']':
2197 break
2198 elif type == tokenize.OP and string_ in ALLOWED_OPS:
2199 if last_string:
2200 yield tokenize.NAME, last_string, last_start, last_end, last_line
2201 last_string = None
2202 yield type, string_, start, end, line
2203 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
2204 if not last_string:
2205 last_string = string_
2206 last_start = start
2207 last_end = end
2208 else:
2209 last_string += string_
2210 if last_string:
2211 yield tokenize.NAME, last_string, last_start, last_end, last_line
2212
2213 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
2214 selectors = []
2215 current_selector = None
2216 for type, string_, start, _, _ in tokens:
2217 # ENCODING is only defined in python 3.x
2218 if type == getattr(tokenize, 'ENCODING', None):
2219 continue
2220 elif type in [tokenize.NAME, tokenize.NUMBER]:
2221 current_selector = FormatSelector(SINGLE, string_, [])
2222 elif type == tokenize.OP:
2223 if string_ == ')':
2224 if not inside_group:
2225 # ')' will be handled by the parentheses group
2226 tokens.restore_last_token()
2227 break
2228 elif inside_merge and string_ in ['/', ',']:
2229 tokens.restore_last_token()
2230 break
2231 elif inside_choice and string_ == ',':
2232 tokens.restore_last_token()
2233 break
2234 elif string_ == ',':
2235 if not current_selector:
2236 raise syntax_error('"," must follow a format selector', start)
2237 selectors.append(current_selector)
2238 current_selector = None
2239 elif string_ == '/':
2240 if not current_selector:
2241 raise syntax_error('"/" must follow a format selector', start)
2242 first_choice = current_selector
2243 second_choice = _parse_format_selection(tokens, inside_choice=True)
2244 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
2245 elif string_ == '[':
2246 if not current_selector:
2247 current_selector = FormatSelector(SINGLE, 'best', [])
2248 format_filter = _parse_filter(tokens)
2249 current_selector.filters.append(format_filter)
2250 elif string_ == '(':
2251 if current_selector:
2252 raise syntax_error('Unexpected "("', start)
2253 group = _parse_format_selection(tokens, inside_group=True)
2254 current_selector = FormatSelector(GROUP, group, [])
2255 elif string_ == '+':
2256 if not current_selector:
2257 raise syntax_error('Unexpected "+"', start)
2258 selector_1 = current_selector
2259 selector_2 = _parse_format_selection(tokens, inside_merge=True)
2260 if not selector_2:
2261 raise syntax_error('Expected a selector', start)
2262 current_selector = FormatSelector(MERGE, (selector_1, selector_2), [])
2263 else:
2264 raise syntax_error(f'Operator not recognized: "{string_}"', start)
2265 elif type == tokenize.ENDMARKER:
2266 break
2267 if current_selector:
2268 selectors.append(current_selector)
2269 return selectors
2270
2271 def _merge(formats_pair):
2272 format_1, format_2 = formats_pair
2273
2274 formats_info = []
2275 formats_info.extend(format_1.get('requested_formats', (format_1,)))
2276 formats_info.extend(format_2.get('requested_formats', (format_2,)))
2277
2278 if not allow_multiple_streams['video'] or not allow_multiple_streams['audio']:
2279 get_no_more = {'video': False, 'audio': False}
2280 for (i, fmt_info) in enumerate(formats_info):
2281 if fmt_info.get('acodec') == fmt_info.get('vcodec') == 'none':
2282 formats_info.pop(i)
2283 continue
2284 for aud_vid in ['audio', 'video']:
2285 if not allow_multiple_streams[aud_vid] and fmt_info.get(aud_vid[0] + 'codec') != 'none':
2286 if get_no_more[aud_vid]:
2287 formats_info.pop(i)
2288 break
2289 get_no_more[aud_vid] = True
2290
2291 if len(formats_info) == 1:
2292 return formats_info[0]
2293
2294 video_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('vcodec') != 'none']
2295 audio_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('acodec') != 'none']
2296
2297 the_only_video = video_fmts[0] if len(video_fmts) == 1 else None
2298 the_only_audio = audio_fmts[0] if len(audio_fmts) == 1 else None
2299
2300 output_ext = get_compatible_ext(
2301 vcodecs=[f.get('vcodec') for f in video_fmts],
2302 acodecs=[f.get('acodec') for f in audio_fmts],
2303 vexts=[f['ext'] for f in video_fmts],
2304 aexts=[f['ext'] for f in audio_fmts],
2305 preferences=(try_call(lambda: self.params['merge_output_format'].split('/'))
2306 or self.params.get('prefer_free_formats') and ('webm', 'mkv')))
2307
2308 filtered = lambda *keys: filter(None, (traverse_obj(fmt, *keys) for fmt in formats_info))
2309
2310 new_dict = {
2311 'requested_formats': formats_info,
2312 'format': '+'.join(filtered('format')),
2313 'format_id': '+'.join(filtered('format_id')),
2314 'ext': output_ext,
2315 'protocol': '+'.join(map(determine_protocol, formats_info)),
2316 'language': '+'.join(orderedSet(filtered('language'))) or None,
2317 'format_note': '+'.join(orderedSet(filtered('format_note'))) or None,
2318 'filesize_approx': sum(filtered('filesize', 'filesize_approx')) or None,
2319 'tbr': sum(filtered('tbr', 'vbr', 'abr')),
2320 }
2321
2322 if the_only_video:
2323 new_dict.update({
2324 'width': the_only_video.get('width'),
2325 'height': the_only_video.get('height'),
2326 'resolution': the_only_video.get('resolution') or self.format_resolution(the_only_video),
2327 'fps': the_only_video.get('fps'),
2328 'dynamic_range': the_only_video.get('dynamic_range'),
2329 'vcodec': the_only_video.get('vcodec'),
2330 'vbr': the_only_video.get('vbr'),
2331 'stretched_ratio': the_only_video.get('stretched_ratio'),
2332 'aspect_ratio': the_only_video.get('aspect_ratio'),
2333 })
2334
2335 if the_only_audio:
2336 new_dict.update({
2337 'acodec': the_only_audio.get('acodec'),
2338 'abr': the_only_audio.get('abr'),
2339 'asr': the_only_audio.get('asr'),
2340 'audio_channels': the_only_audio.get('audio_channels')
2341 })
2342
2343 return new_dict
2344
2345 def _check_formats(formats):
2346 if self.params.get('check_formats') == 'selected':
2347 yield from self._check_formats(formats)
2348 return
2349 elif (self.params.get('check_formats') is not None
2350 or self.params.get('allow_unplayable_formats')):
2351 yield from formats
2352 return
2353
2354 for f in formats:
2355 if f.get('has_drm') or f.get('__needs_testing'):
2356 yield from self._check_formats([f])
2357 else:
2358 yield f
2359
2360 def _build_selector_function(selector):
2361 if isinstance(selector, list): # ,
2362 fs = [_build_selector_function(s) for s in selector]
2363
2364 def selector_function(ctx):
2365 for f in fs:
2366 yield from f(ctx)
2367 return selector_function
2368
2369 elif selector.type == GROUP: # ()
2370 selector_function = _build_selector_function(selector.selector)
2371
2372 elif selector.type == PICKFIRST: # /
2373 fs = [_build_selector_function(s) for s in selector.selector]
2374
2375 def selector_function(ctx):
2376 for f in fs:
2377 picked_formats = list(f(ctx))
2378 if picked_formats:
2379 return picked_formats
2380 return []
2381
2382 elif selector.type == MERGE: # +
2383 selector_1, selector_2 = map(_build_selector_function, selector.selector)
2384
2385 def selector_function(ctx):
2386 for pair in itertools.product(selector_1(ctx), selector_2(ctx)):
2387 yield _merge(pair)
2388
2389 elif selector.type == SINGLE: # atom
2390 format_spec = selector.selector or 'best'
2391
2392 # TODO: Add allvideo, allaudio etc by generalizing the code with best/worst selector
2393 if format_spec == 'all':
2394 def selector_function(ctx):
2395 yield from _check_formats(ctx['formats'][::-1])
2396 elif format_spec == 'mergeall':
2397 def selector_function(ctx):
2398 formats = list(_check_formats(
2399 f for f in ctx['formats'] if f.get('vcodec') != 'none' or f.get('acodec') != 'none'))
2400 if not formats:
2401 return
2402 merged_format = formats[-1]
2403 for f in formats[-2::-1]:
2404 merged_format = _merge((merged_format, f))
2405 yield merged_format
2406
2407 else:
2408 format_fallback, seperate_fallback, format_reverse, format_idx = False, None, True, 1
2409 mobj = re.match(
2410 r'(?P<bw>best|worst|b|w)(?P<type>video|audio|v|a)?(?P<mod>\*)?(?:\.(?P<n>[1-9]\d*))?$',
2411 format_spec)
2412 if mobj is not None:
2413 format_idx = int_or_none(mobj.group('n'), default=1)
2414 format_reverse = mobj.group('bw')[0] == 'b'
2415 format_type = (mobj.group('type') or [None])[0]
2416 not_format_type = {'v': 'a', 'a': 'v'}.get(format_type)
2417 format_modified = mobj.group('mod') is not None
2418
2419 format_fallback = not format_type and not format_modified # for b, w
2420 _filter_f = (
2421 (lambda f: f.get('%scodec' % format_type) != 'none')
2422 if format_type and format_modified # bv*, ba*, wv*, wa*
2423 else (lambda f: f.get('%scodec' % not_format_type) == 'none')
2424 if format_type # bv, ba, wv, wa
2425 else (lambda f: f.get('vcodec') != 'none' and f.get('acodec') != 'none')
2426 if not format_modified # b, w
2427 else lambda f: True) # b*, w*
2428 filter_f = lambda f: _filter_f(f) and (
2429 f.get('vcodec') != 'none' or f.get('acodec') != 'none')
2430 else:
2431 if format_spec in self._format_selection_exts['audio']:
2432 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none'
2433 elif format_spec in self._format_selection_exts['video']:
2434 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none' and f.get('vcodec') != 'none'
2435 seperate_fallback = lambda f: f.get('ext') == format_spec and f.get('vcodec') != 'none'
2436 elif format_spec in self._format_selection_exts['storyboards']:
2437 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') == 'none' and f.get('vcodec') == 'none'
2438 else:
2439 filter_f = lambda f: f.get('format_id') == format_spec # id
2440
2441 def selector_function(ctx):
2442 formats = list(ctx['formats'])
2443 matches = list(filter(filter_f, formats)) if filter_f is not None else formats
2444 if not matches:
2445 if format_fallback and ctx['incomplete_formats']:
2446 # for extractors with incomplete formats (audio only (soundcloud)
2447 # or video only (imgur)) best/worst will fallback to
2448 # best/worst {video,audio}-only format
2449 matches = formats
2450 elif seperate_fallback and not ctx['has_merged_format']:
2451 # for compatibility with youtube-dl when there is no pre-merged format
2452 matches = list(filter(seperate_fallback, formats))
2453 matches = LazyList(_check_formats(matches[::-1 if format_reverse else 1]))
2454 try:
2455 yield matches[format_idx - 1]
2456 except LazyList.IndexError:
2457 return
2458
2459 filters = [self._build_format_filter(f) for f in selector.filters]
2460
2461 def final_selector(ctx):
2462 ctx_copy = dict(ctx)
2463 for _filter in filters:
2464 ctx_copy['formats'] = list(filter(_filter, ctx_copy['formats']))
2465 return selector_function(ctx_copy)
2466 return final_selector
2467
2468 stream = io.BytesIO(format_spec.encode())
2469 try:
2470 tokens = list(_remove_unused_ops(tokenize.tokenize(stream.readline)))
2471 except tokenize.TokenError:
2472 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
2473
2474 class TokenIterator:
2475 def __init__(self, tokens):
2476 self.tokens = tokens
2477 self.counter = 0
2478
2479 def __iter__(self):
2480 return self
2481
2482 def __next__(self):
2483 if self.counter >= len(self.tokens):
2484 raise StopIteration()
2485 value = self.tokens[self.counter]
2486 self.counter += 1
2487 return value
2488
2489 next = __next__
2490
2491 def restore_last_token(self):
2492 self.counter -= 1
2493
2494 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
2495 return _build_selector_function(parsed_selector)
2496
2497 def _calc_headers(self, info_dict, load_cookies=False):
2498 res = HTTPHeaderDict(self.params['http_headers'], info_dict.get('http_headers'))
2499 clean_headers(res)
2500
2501 if load_cookies: # For --load-info-json
2502 self._load_cookies(res.get('Cookie'), autoscope=info_dict['url']) # compat
2503 self._load_cookies(info_dict.get('cookies'), autoscope=False)
2504 # The `Cookie` header is removed to prevent leaks and unscoped cookies.
2505 # See: https://github.com/yt-dlp/yt-dlp/security/advisories/GHSA-v8mc-9377-rwjj
2506 res.pop('Cookie', None)
2507 cookies = self.cookiejar.get_cookies_for_url(info_dict['url'])
2508 if cookies:
2509 encoder = LenientSimpleCookie()
2510 values = []
2511 for cookie in cookies:
2512 _, value = encoder.value_encode(cookie.value)
2513 values.append(f'{cookie.name}={value}')
2514 if cookie.domain:
2515 values.append(f'Domain={cookie.domain}')
2516 if cookie.path:
2517 values.append(f'Path={cookie.path}')
2518 if cookie.secure:
2519 values.append('Secure')
2520 if cookie.expires:
2521 values.append(f'Expires={cookie.expires}')
2522 if cookie.version:
2523 values.append(f'Version={cookie.version}')
2524 info_dict['cookies'] = '; '.join(values)
2525
2526 if 'X-Forwarded-For' not in res:
2527 x_forwarded_for_ip = info_dict.get('__x_forwarded_for_ip')
2528 if x_forwarded_for_ip:
2529 res['X-Forwarded-For'] = x_forwarded_for_ip
2530
2531 return res
2532
2533 def _calc_cookies(self, url):
2534 self.deprecation_warning('"YoutubeDL._calc_cookies" is deprecated and may be removed in a future version')
2535 return self.cookiejar.get_cookie_header(url)
2536
2537 def _sort_thumbnails(self, thumbnails):
2538 thumbnails.sort(key=lambda t: (
2539 t.get('preference') if t.get('preference') is not None else -1,
2540 t.get('width') if t.get('width') is not None else -1,
2541 t.get('height') if t.get('height') is not None else -1,
2542 t.get('id') if t.get('id') is not None else '',
2543 t.get('url')))
2544
2545 def _sanitize_thumbnails(self, info_dict):
2546 thumbnails = info_dict.get('thumbnails')
2547 if thumbnails is None:
2548 thumbnail = info_dict.get('thumbnail')
2549 if thumbnail:
2550 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
2551 if not thumbnails:
2552 return
2553
2554 def check_thumbnails(thumbnails):
2555 for t in thumbnails:
2556 self.to_screen(f'[info] Testing thumbnail {t["id"]}')
2557 try:
2558 self.urlopen(HEADRequest(t['url']))
2559 except network_exceptions as err:
2560 self.to_screen(f'[info] Unable to connect to thumbnail {t["id"]} URL {t["url"]!r} - {err}. Skipping...')
2561 continue
2562 yield t
2563
2564 self._sort_thumbnails(thumbnails)
2565 for i, t in enumerate(thumbnails):
2566 if t.get('id') is None:
2567 t['id'] = '%d' % i
2568 if t.get('width') and t.get('height'):
2569 t['resolution'] = '%dx%d' % (t['width'], t['height'])
2570 t['url'] = sanitize_url(t['url'])
2571
2572 if self.params.get('check_formats') is True:
2573 info_dict['thumbnails'] = LazyList(check_thumbnails(thumbnails[::-1]), reverse=True)
2574 else:
2575 info_dict['thumbnails'] = thumbnails
2576
2577 def _fill_common_fields(self, info_dict, final=True):
2578 # TODO: move sanitization here
2579 if final:
2580 title = info_dict['fulltitle'] = info_dict.get('title')
2581 if not title:
2582 if title == '':
2583 self.write_debug('Extractor gave empty title. Creating a generic title')
2584 else:
2585 self.report_warning('Extractor failed to obtain "title". Creating a generic title instead')
2586 info_dict['title'] = f'{info_dict["extractor"].replace(":", "-")} video #{info_dict["id"]}'
2587
2588 if info_dict.get('duration') is not None:
2589 info_dict['duration_string'] = formatSeconds(info_dict['duration'])
2590
2591 for ts_key, date_key in (
2592 ('timestamp', 'upload_date'),
2593 ('release_timestamp', 'release_date'),
2594 ('modified_timestamp', 'modified_date'),
2595 ):
2596 if info_dict.get(date_key) is None and info_dict.get(ts_key) is not None:
2597 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
2598 # see http://bugs.python.org/issue1646728)
2599 with contextlib.suppress(ValueError, OverflowError, OSError):
2600 upload_date = datetime.datetime.fromtimestamp(info_dict[ts_key], datetime.timezone.utc)
2601 info_dict[date_key] = upload_date.strftime('%Y%m%d')
2602
2603 if not info_dict.get('release_year'):
2604 info_dict['release_year'] = traverse_obj(info_dict, ('release_date', {lambda x: int(x[:4])}))
2605
2606 live_keys = ('is_live', 'was_live')
2607 live_status = info_dict.get('live_status')
2608 if live_status is None:
2609 for key in live_keys:
2610 if info_dict.get(key) is False:
2611 continue
2612 if info_dict.get(key):
2613 live_status = key
2614 break
2615 if all(info_dict.get(key) is False for key in live_keys):
2616 live_status = 'not_live'
2617 if live_status:
2618 info_dict['live_status'] = live_status
2619 for key in live_keys:
2620 if info_dict.get(key) is None:
2621 info_dict[key] = (live_status == key)
2622 if live_status == 'post_live':
2623 info_dict['was_live'] = True
2624
2625 # Auto generate title fields corresponding to the *_number fields when missing
2626 # in order to always have clean titles. This is very common for TV series.
2627 for field in ('chapter', 'season', 'episode'):
2628 if final and info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
2629 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
2630
2631 def _raise_pending_errors(self, info):
2632 err = info.pop('__pending_error', None)
2633 if err:
2634 self.report_error(err, tb=False)
2635
2636 def sort_formats(self, info_dict):
2637 formats = self._get_formats(info_dict)
2638 formats.sort(key=FormatSorter(
2639 self, info_dict.get('_format_sort_fields') or []).calculate_preference)
2640
2641 def process_video_result(self, info_dict, download=True):
2642 assert info_dict.get('_type', 'video') == 'video'
2643 self._num_videos += 1
2644
2645 if 'id' not in info_dict:
2646 raise ExtractorError('Missing "id" field in extractor result', ie=info_dict['extractor'])
2647 elif not info_dict.get('id'):
2648 raise ExtractorError('Extractor failed to obtain "id"', ie=info_dict['extractor'])
2649
2650 def report_force_conversion(field, field_not, conversion):
2651 self.report_warning(
2652 '"%s" field is not %s - forcing %s conversion, there is an error in extractor'
2653 % (field, field_not, conversion))
2654
2655 def sanitize_string_field(info, string_field):
2656 field = info.get(string_field)
2657 if field is None or isinstance(field, str):
2658 return
2659 report_force_conversion(string_field, 'a string', 'string')
2660 info[string_field] = str(field)
2661
2662 def sanitize_numeric_fields(info):
2663 for numeric_field in self._NUMERIC_FIELDS:
2664 field = info.get(numeric_field)
2665 if field is None or isinstance(field, (int, float)):
2666 continue
2667 report_force_conversion(numeric_field, 'numeric', 'int')
2668 info[numeric_field] = int_or_none(field)
2669
2670 sanitize_string_field(info_dict, 'id')
2671 sanitize_numeric_fields(info_dict)
2672 if info_dict.get('section_end') and info_dict.get('section_start') is not None:
2673 info_dict['duration'] = round(info_dict['section_end'] - info_dict['section_start'], 3)
2674 if (info_dict.get('duration') or 0) <= 0 and info_dict.pop('duration', None):
2675 self.report_warning('"duration" field is negative, there is an error in extractor')
2676
2677 chapters = info_dict.get('chapters') or []
2678 if chapters and chapters[0].get('start_time'):
2679 chapters.insert(0, {'start_time': 0})
2680
2681 dummy_chapter = {'end_time': 0, 'start_time': info_dict.get('duration')}
2682 for idx, (prev, current, next_) in enumerate(zip(
2683 (dummy_chapter, *chapters), chapters, (*chapters[1:], dummy_chapter)), 1):
2684 if current.get('start_time') is None:
2685 current['start_time'] = prev.get('end_time')
2686 if not current.get('end_time'):
2687 current['end_time'] = next_.get('start_time')
2688 if not current.get('title'):
2689 current['title'] = f'<Untitled Chapter {idx}>'
2690
2691 if 'playlist' not in info_dict:
2692 # It isn't part of a playlist
2693 info_dict['playlist'] = None
2694 info_dict['playlist_index'] = None
2695
2696 self._sanitize_thumbnails(info_dict)
2697
2698 thumbnail = info_dict.get('thumbnail')
2699 thumbnails = info_dict.get('thumbnails')
2700 if thumbnail:
2701 info_dict['thumbnail'] = sanitize_url(thumbnail)
2702 elif thumbnails:
2703 info_dict['thumbnail'] = thumbnails[-1]['url']
2704
2705 if info_dict.get('display_id') is None and 'id' in info_dict:
2706 info_dict['display_id'] = info_dict['id']
2707
2708 self._fill_common_fields(info_dict)
2709
2710 for cc_kind in ('subtitles', 'automatic_captions'):
2711 cc = info_dict.get(cc_kind)
2712 if cc:
2713 for _, subtitle in cc.items():
2714 for subtitle_format in subtitle:
2715 if subtitle_format.get('url'):
2716 subtitle_format['url'] = sanitize_url(subtitle_format['url'])
2717 if subtitle_format.get('ext') is None:
2718 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
2719
2720 automatic_captions = info_dict.get('automatic_captions')
2721 subtitles = info_dict.get('subtitles')
2722
2723 info_dict['requested_subtitles'] = self.process_subtitles(
2724 info_dict['id'], subtitles, automatic_captions)
2725
2726 formats = self._get_formats(info_dict)
2727
2728 # Backward compatibility with InfoExtractor._sort_formats
2729 field_preference = (formats or [{}])[0].pop('__sort_fields', None)
2730 if field_preference:
2731 info_dict['_format_sort_fields'] = field_preference
2732
2733 info_dict['_has_drm'] = any( # or None ensures --clean-infojson removes it
2734 f.get('has_drm') and f['has_drm'] != 'maybe' for f in formats) or None
2735 if not self.params.get('allow_unplayable_formats'):
2736 formats = [f for f in formats if not f.get('has_drm') or f['has_drm'] == 'maybe']
2737
2738 if formats and all(f.get('acodec') == f.get('vcodec') == 'none' for f in formats):
2739 self.report_warning(
2740 f'{"This video is DRM protected and " if info_dict["_has_drm"] else ""}'
2741 'only images are available for download. Use --list-formats to see them'.capitalize())
2742
2743 get_from_start = not info_dict.get('is_live') or bool(self.params.get('live_from_start'))
2744 if not get_from_start:
2745 info_dict['title'] += ' ' + datetime.datetime.now().strftime('%Y-%m-%d %H:%M')
2746 if info_dict.get('is_live') and formats:
2747 formats = [f for f in formats if bool(f.get('is_from_start')) == get_from_start]
2748 if get_from_start and not formats:
2749 self.raise_no_formats(info_dict, msg=(
2750 '--live-from-start is passed, but there are no formats that can be downloaded from the start. '
2751 'If you want to download from the current time, use --no-live-from-start'))
2752
2753 def is_wellformed(f):
2754 url = f.get('url')
2755 if not url:
2756 self.report_warning(
2757 '"url" field is missing or empty - skipping format, '
2758 'there is an error in extractor')
2759 return False
2760 if isinstance(url, bytes):
2761 sanitize_string_field(f, 'url')
2762 return True
2763
2764 # Filter out malformed formats for better extraction robustness
2765 formats = list(filter(is_wellformed, formats or []))
2766
2767 if not formats:
2768 self.raise_no_formats(info_dict)
2769
2770 for format in formats:
2771 sanitize_string_field(format, 'format_id')
2772 sanitize_numeric_fields(format)
2773 format['url'] = sanitize_url(format['url'])
2774 if format.get('ext') is None:
2775 format['ext'] = determine_ext(format['url']).lower()
2776 if format.get('protocol') is None:
2777 format['protocol'] = determine_protocol(format)
2778 if format.get('resolution') is None:
2779 format['resolution'] = self.format_resolution(format, default=None)
2780 if format.get('dynamic_range') is None and format.get('vcodec') != 'none':
2781 format['dynamic_range'] = 'SDR'
2782 if format.get('aspect_ratio') is None:
2783 format['aspect_ratio'] = try_call(lambda: round(format['width'] / format['height'], 2))
2784 # For fragmented formats, "tbr" is often max bitrate and not average
2785 if (('manifest-filesize-approx' in self.params['compat_opts'] or not format.get('manifest_url'))
2786 and info_dict.get('duration') and format.get('tbr')
2787 and not format.get('filesize') and not format.get('filesize_approx')):
2788 format['filesize_approx'] = int(info_dict['duration'] * format['tbr'] * (1024 / 8))
2789 format['http_headers'] = self._calc_headers(collections.ChainMap(format, info_dict), load_cookies=True)
2790
2791 # Safeguard against old/insecure infojson when using --load-info-json
2792 if info_dict.get('http_headers'):
2793 info_dict['http_headers'] = HTTPHeaderDict(info_dict['http_headers'])
2794 info_dict['http_headers'].pop('Cookie', None)
2795
2796 # This is copied to http_headers by the above _calc_headers and can now be removed
2797 if '__x_forwarded_for_ip' in info_dict:
2798 del info_dict['__x_forwarded_for_ip']
2799
2800 self.sort_formats({
2801 'formats': formats,
2802 '_format_sort_fields': info_dict.get('_format_sort_fields')
2803 })
2804
2805 # Sanitize and group by format_id
2806 formats_dict = {}
2807 for i, format in enumerate(formats):
2808 if not format.get('format_id'):
2809 format['format_id'] = str(i)
2810 else:
2811 # Sanitize format_id from characters used in format selector expression
2812 format['format_id'] = re.sub(r'[\s,/+\[\]()]', '_', format['format_id'])
2813 formats_dict.setdefault(format['format_id'], []).append(format)
2814
2815 # Make sure all formats have unique format_id
2816 common_exts = set(itertools.chain(*self._format_selection_exts.values()))
2817 for format_id, ambiguous_formats in formats_dict.items():
2818 ambigious_id = len(ambiguous_formats) > 1
2819 for i, format in enumerate(ambiguous_formats):
2820 if ambigious_id:
2821 format['format_id'] = '%s-%d' % (format_id, i)
2822 # Ensure there is no conflict between id and ext in format selection
2823 # See https://github.com/yt-dlp/yt-dlp/issues/1282
2824 if format['format_id'] != format['ext'] and format['format_id'] in common_exts:
2825 format['format_id'] = 'f%s' % format['format_id']
2826
2827 if format.get('format') is None:
2828 format['format'] = '{id} - {res}{note}'.format(
2829 id=format['format_id'],
2830 res=self.format_resolution(format),
2831 note=format_field(format, 'format_note', ' (%s)'),
2832 )
2833
2834 if self.params.get('check_formats') is True:
2835 formats = LazyList(self._check_formats(formats[::-1]), reverse=True)
2836
2837 if not formats or formats[0] is not info_dict:
2838 # only set the 'formats' fields if the original info_dict list them
2839 # otherwise we end up with a circular reference, the first (and unique)
2840 # element in the 'formats' field in info_dict is info_dict itself,
2841 # which can't be exported to json
2842 info_dict['formats'] = formats
2843
2844 info_dict, _ = self.pre_process(info_dict)
2845
2846 if self._match_entry(info_dict, incomplete=self._format_fields) is not None:
2847 return info_dict
2848
2849 self.post_extract(info_dict)
2850 info_dict, _ = self.pre_process(info_dict, 'after_filter')
2851
2852 # The pre-processors may have modified the formats
2853 formats = self._get_formats(info_dict)
2854
2855 list_only = self.params.get('simulate') == 'list_only'
2856 interactive_format_selection = not list_only and self.format_selector == '-'
2857 if self.params.get('list_thumbnails'):
2858 self.list_thumbnails(info_dict)
2859 if self.params.get('listsubtitles'):
2860 if 'automatic_captions' in info_dict:
2861 self.list_subtitles(
2862 info_dict['id'], automatic_captions, 'automatic captions')
2863 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
2864 if self.params.get('listformats') or interactive_format_selection:
2865 self.list_formats(info_dict)
2866 if list_only:
2867 # Without this printing, -F --print-json will not work
2868 self.__forced_printings(info_dict)
2869 return info_dict
2870
2871 format_selector = self.format_selector
2872 while True:
2873 if interactive_format_selection:
2874 req_format = input(self._format_screen('\nEnter format selector ', self.Styles.EMPHASIS)
2875 + '(Press ENTER for default, or Ctrl+C to quit)'
2876 + self._format_screen(': ', self.Styles.EMPHASIS))
2877 try:
2878 format_selector = self.build_format_selector(req_format) if req_format else None
2879 except SyntaxError as err:
2880 self.report_error(err, tb=False, is_error=False)
2881 continue
2882
2883 if format_selector is None:
2884 req_format = self._default_format_spec(info_dict, download=download)
2885 self.write_debug(f'Default format spec: {req_format}')
2886 format_selector = self.build_format_selector(req_format)
2887
2888 formats_to_download = list(format_selector({
2889 'formats': formats,
2890 'has_merged_format': any('none' not in (f.get('acodec'), f.get('vcodec')) for f in formats),
2891 'incomplete_formats': (all(f.get('vcodec') == 'none' for f in formats) # No formats with video
2892 or all(f.get('acodec') == 'none' for f in formats)), # OR, No formats with audio
2893 }))
2894 if interactive_format_selection and not formats_to_download:
2895 self.report_error('Requested format is not available', tb=False, is_error=False)
2896 continue
2897 break
2898
2899 if not formats_to_download:
2900 if not self.params.get('ignore_no_formats_error'):
2901 raise ExtractorError(
2902 'Requested format is not available. Use --list-formats for a list of available formats',
2903 expected=True, video_id=info_dict['id'], ie=info_dict['extractor'])
2904 self.report_warning('Requested format is not available')
2905 # Process what we can, even without any available formats.
2906 formats_to_download = [{}]
2907
2908 requested_ranges = tuple(self.params.get('download_ranges', lambda *_: [{}])(info_dict, self))
2909 best_format, downloaded_formats = formats_to_download[-1], []
2910 if download:
2911 if best_format and requested_ranges:
2912 def to_screen(*msg):
2913 self.to_screen(f'[info] {info_dict["id"]}: {" ".join(", ".join(variadic(m)) for m in msg)}')
2914
2915 to_screen(f'Downloading {len(formats_to_download)} format(s):',
2916 (f['format_id'] for f in formats_to_download))
2917 if requested_ranges != ({}, ):
2918 to_screen(f'Downloading {len(requested_ranges)} time ranges:',
2919 (f'{c["start_time"]:.1f}-{c["end_time"]:.1f}' for c in requested_ranges))
2920 max_downloads_reached = False
2921
2922 for fmt, chapter in itertools.product(formats_to_download, requested_ranges):
2923 new_info = self._copy_infodict(info_dict)
2924 new_info.update(fmt)
2925 offset, duration = info_dict.get('section_start') or 0, info_dict.get('duration') or float('inf')
2926 end_time = offset + min(chapter.get('end_time', duration), duration)
2927 # duration may not be accurate. So allow deviations <1sec
2928 if end_time == float('inf') or end_time > offset + duration + 1:
2929 end_time = None
2930 if chapter or offset:
2931 new_info.update({
2932 'section_start': offset + chapter.get('start_time', 0),
2933 'section_end': end_time,
2934 'section_title': chapter.get('title'),
2935 'section_number': chapter.get('index'),
2936 })
2937 downloaded_formats.append(new_info)
2938 try:
2939 self.process_info(new_info)
2940 except MaxDownloadsReached:
2941 max_downloads_reached = True
2942 self._raise_pending_errors(new_info)
2943 # Remove copied info
2944 for key, val in tuple(new_info.items()):
2945 if info_dict.get(key) == val:
2946 new_info.pop(key)
2947 if max_downloads_reached:
2948 break
2949
2950 write_archive = {f.get('__write_download_archive', False) for f in downloaded_formats}
2951 assert write_archive.issubset({True, False, 'ignore'})
2952 if True in write_archive and False not in write_archive:
2953 self.record_download_archive(info_dict)
2954
2955 info_dict['requested_downloads'] = downloaded_formats
2956 info_dict = self.run_all_pps('after_video', info_dict)
2957 if max_downloads_reached:
2958 raise MaxDownloadsReached()
2959
2960 # We update the info dict with the selected best quality format (backwards compatibility)
2961 info_dict.update(best_format)
2962 return info_dict
2963
2964 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
2965 """Select the requested subtitles and their format"""
2966 available_subs, normal_sub_langs = {}, []
2967 if normal_subtitles and self.params.get('writesubtitles'):
2968 available_subs.update(normal_subtitles)
2969 normal_sub_langs = tuple(normal_subtitles.keys())
2970 if automatic_captions and self.params.get('writeautomaticsub'):
2971 for lang, cap_info in automatic_captions.items():
2972 if lang not in available_subs:
2973 available_subs[lang] = cap_info
2974
2975 if not available_subs or (
2976 not self.params.get('writesubtitles')
2977 and not self.params.get('writeautomaticsub')):
2978 return None
2979
2980 all_sub_langs = tuple(available_subs.keys())
2981 if self.params.get('allsubtitles', False):
2982 requested_langs = all_sub_langs
2983 elif self.params.get('subtitleslangs', False):
2984 try:
2985 requested_langs = orderedSet_from_options(
2986 self.params.get('subtitleslangs'), {'all': all_sub_langs}, use_regex=True)
2987 except re.error as e:
2988 raise ValueError(f'Wrong regex for subtitlelangs: {e.pattern}')
2989 else:
2990 requested_langs = LazyList(itertools.chain(
2991 ['en'] if 'en' in normal_sub_langs else [],
2992 filter(lambda f: f.startswith('en'), normal_sub_langs),
2993 ['en'] if 'en' in all_sub_langs else [],
2994 filter(lambda f: f.startswith('en'), all_sub_langs),
2995 normal_sub_langs, all_sub_langs,
2996 ))[:1]
2997 if requested_langs:
2998 self.to_screen(f'[info] {video_id}: Downloading subtitles: {", ".join(requested_langs)}')
2999
3000 formats_query = self.params.get('subtitlesformat', 'best')
3001 formats_preference = formats_query.split('/') if formats_query else []
3002 subs = {}
3003 for lang in requested_langs:
3004 formats = available_subs.get(lang)
3005 if formats is None:
3006 self.report_warning(f'{lang} subtitles not available for {video_id}')
3007 continue
3008 for ext in formats_preference:
3009 if ext == 'best':
3010 f = formats[-1]
3011 break
3012 matches = list(filter(lambda f: f['ext'] == ext, formats))
3013 if matches:
3014 f = matches[-1]
3015 break
3016 else:
3017 f = formats[-1]
3018 self.report_warning(
3019 'No subtitle format found matching "%s" for language %s, '
3020 'using %s' % (formats_query, lang, f['ext']))
3021 subs[lang] = f
3022 return subs
3023
3024 def _forceprint(self, key, info_dict):
3025 if info_dict is None:
3026 return
3027 info_copy = info_dict.copy()
3028 info_copy.setdefault('filename', self.prepare_filename(info_dict))
3029 if info_dict.get('requested_formats') is not None:
3030 # For RTMP URLs, also include the playpath
3031 info_copy['urls'] = '\n'.join(f['url'] + f.get('play_path', '') for f in info_dict['requested_formats'])
3032 elif info_dict.get('url'):
3033 info_copy['urls'] = info_dict['url'] + info_dict.get('play_path', '')
3034 info_copy['formats_table'] = self.render_formats_table(info_dict)
3035 info_copy['thumbnails_table'] = self.render_thumbnails_table(info_dict)
3036 info_copy['subtitles_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('subtitles'))
3037 info_copy['automatic_captions_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('automatic_captions'))
3038
3039 def format_tmpl(tmpl):
3040 mobj = re.fullmatch(r'([\w.:,]|-\d|(?P<dict>{([\w.:,]|-\d)+}))+=?', tmpl)
3041 if not mobj:
3042 return tmpl
3043
3044 fmt = '%({})s'
3045 if tmpl.startswith('{'):
3046 tmpl, fmt = f'.{tmpl}', '%({})j'
3047 if tmpl.endswith('='):
3048 tmpl, fmt = tmpl[:-1], '{0} = %({0})#j'
3049 return '\n'.join(map(fmt.format, [tmpl] if mobj.group('dict') else tmpl.split(',')))
3050
3051 for tmpl in self.params['forceprint'].get(key, []):
3052 self.to_stdout(self.evaluate_outtmpl(format_tmpl(tmpl), info_copy))
3053
3054 for tmpl, file_tmpl in self.params['print_to_file'].get(key, []):
3055 filename = self.prepare_filename(info_dict, outtmpl=file_tmpl)
3056 tmpl = format_tmpl(tmpl)
3057 self.to_screen(f'[info] Writing {tmpl!r} to: {filename}')
3058 if self._ensure_dir_exists(filename):
3059 with open(filename, 'a', encoding='utf-8', newline='') as f:
3060 f.write(self.evaluate_outtmpl(tmpl, info_copy) + os.linesep)
3061
3062 return info_copy
3063
3064 def __forced_printings(self, info_dict, filename=None, incomplete=True):
3065 if (self.params.get('forcejson')
3066 or self.params['forceprint'].get('video')
3067 or self.params['print_to_file'].get('video')):
3068 self.post_extract(info_dict)
3069 if filename:
3070 info_dict['filename'] = filename
3071 info_copy = self._forceprint('video', info_dict)
3072
3073 def print_field(field, actual_field=None, optional=False):
3074 if actual_field is None:
3075 actual_field = field
3076 if self.params.get(f'force{field}') and (
3077 info_copy.get(field) is not None or (not optional and not incomplete)):
3078 self.to_stdout(info_copy[actual_field])
3079
3080 print_field('title')
3081 print_field('id')
3082 print_field('url', 'urls')
3083 print_field('thumbnail', optional=True)
3084 print_field('description', optional=True)
3085 print_field('filename')
3086 if self.params.get('forceduration') and info_copy.get('duration') is not None:
3087 self.to_stdout(formatSeconds(info_copy['duration']))
3088 print_field('format')
3089
3090 if self.params.get('forcejson'):
3091 self.to_stdout(json.dumps(self.sanitize_info(info_dict)))
3092
3093 def dl(self, name, info, subtitle=False, test=False):
3094 if not info.get('url'):
3095 self.raise_no_formats(info, True)
3096
3097 if test:
3098 verbose = self.params.get('verbose')
3099 params = {
3100 'test': True,
3101 'quiet': self.params.get('quiet') or not verbose,
3102 'verbose': verbose,
3103 'noprogress': not verbose,
3104 'nopart': True,
3105 'skip_unavailable_fragments': False,
3106 'keep_fragments': False,
3107 'overwrites': True,
3108 '_no_ytdl_file': True,
3109 }
3110 else:
3111 params = self.params
3112 fd = get_suitable_downloader(info, params, to_stdout=(name == '-'))(self, params)
3113 if not test:
3114 for ph in self._progress_hooks:
3115 fd.add_progress_hook(ph)
3116 urls = '", "'.join(
3117 (f['url'].split(',')[0] + ',<data>' if f['url'].startswith('data:') else f['url'])
3118 for f in info.get('requested_formats', []) or [info])
3119 self.write_debug(f'Invoking {fd.FD_NAME} downloader on "{urls}"')
3120
3121 # Note: Ideally info should be a deep-copied so that hooks cannot modify it.
3122 # But it may contain objects that are not deep-copyable
3123 new_info = self._copy_infodict(info)
3124 if new_info.get('http_headers') is None:
3125 new_info['http_headers'] = self._calc_headers(new_info)
3126 return fd.download(name, new_info, subtitle)
3127
3128 def existing_file(self, filepaths, *, default_overwrite=True):
3129 existing_files = list(filter(os.path.exists, orderedSet(filepaths)))
3130 if existing_files and not self.params.get('overwrites', default_overwrite):
3131 return existing_files[0]
3132
3133 for file in existing_files:
3134 self.report_file_delete(file)
3135 os.remove(file)
3136 return None
3137
3138 def process_info(self, info_dict):
3139 """Process a single resolved IE result. (Modifies it in-place)"""
3140
3141 assert info_dict.get('_type', 'video') == 'video'
3142 original_infodict = info_dict
3143
3144 if 'format' not in info_dict and 'ext' in info_dict:
3145 info_dict['format'] = info_dict['ext']
3146
3147 if self._match_entry(info_dict) is not None:
3148 info_dict['__write_download_archive'] = 'ignore'
3149 return
3150
3151 # Does nothing under normal operation - for backward compatibility of process_info
3152 self.post_extract(info_dict)
3153
3154 def replace_info_dict(new_info):
3155 nonlocal info_dict
3156 if new_info == info_dict:
3157 return
3158 info_dict.clear()
3159 info_dict.update(new_info)
3160
3161 new_info, _ = self.pre_process(info_dict, 'video')
3162 replace_info_dict(new_info)
3163 self._num_downloads += 1
3164
3165 # info_dict['_filename'] needs to be set for backward compatibility
3166 info_dict['_filename'] = full_filename = self.prepare_filename(info_dict, warn=True)
3167 temp_filename = self.prepare_filename(info_dict, 'temp')
3168 files_to_move = {}
3169
3170 # Forced printings
3171 self.__forced_printings(info_dict, full_filename, incomplete=('format' not in info_dict))
3172
3173 def check_max_downloads():
3174 if self._num_downloads >= float(self.params.get('max_downloads') or 'inf'):
3175 raise MaxDownloadsReached()
3176
3177 if self.params.get('simulate'):
3178 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
3179 check_max_downloads()
3180 return
3181
3182 if full_filename is None:
3183 return
3184 if not self._ensure_dir_exists(encodeFilename(full_filename)):
3185 return
3186 if not self._ensure_dir_exists(encodeFilename(temp_filename)):
3187 return
3188
3189 if self._write_description('video', info_dict,
3190 self.prepare_filename(info_dict, 'description')) is None:
3191 return
3192
3193 sub_files = self._write_subtitles(info_dict, temp_filename)
3194 if sub_files is None:
3195 return
3196 files_to_move.update(dict(sub_files))
3197
3198 thumb_files = self._write_thumbnails(
3199 'video', info_dict, temp_filename, self.prepare_filename(info_dict, 'thumbnail'))
3200 if thumb_files is None:
3201 return
3202 files_to_move.update(dict(thumb_files))
3203
3204 infofn = self.prepare_filename(info_dict, 'infojson')
3205 _infojson_written = self._write_info_json('video', info_dict, infofn)
3206 if _infojson_written:
3207 info_dict['infojson_filename'] = infofn
3208 # For backward compatibility, even though it was a private field
3209 info_dict['__infojson_filename'] = infofn
3210 elif _infojson_written is None:
3211 return
3212
3213 # Note: Annotations are deprecated
3214 annofn = None
3215 if self.params.get('writeannotations', False):
3216 annofn = self.prepare_filename(info_dict, 'annotation')
3217 if annofn:
3218 if not self._ensure_dir_exists(encodeFilename(annofn)):
3219 return
3220 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(annofn)):
3221 self.to_screen('[info] Video annotations are already present')
3222 elif not info_dict.get('annotations'):
3223 self.report_warning('There are no annotations to write.')
3224 else:
3225 try:
3226 self.to_screen('[info] Writing video annotations to: ' + annofn)
3227 with open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
3228 annofile.write(info_dict['annotations'])
3229 except (KeyError, TypeError):
3230 self.report_warning('There are no annotations to write.')
3231 except OSError:
3232 self.report_error('Cannot write annotations file: ' + annofn)
3233 return
3234
3235 # Write internet shortcut files
3236 def _write_link_file(link_type):
3237 url = try_get(info_dict['webpage_url'], iri_to_uri)
3238 if not url:
3239 self.report_warning(
3240 f'Cannot write internet shortcut file because the actual URL of "{info_dict["webpage_url"]}" is unknown')
3241 return True
3242 linkfn = replace_extension(self.prepare_filename(info_dict, 'link'), link_type, info_dict.get('ext'))
3243 if not self._ensure_dir_exists(encodeFilename(linkfn)):
3244 return False
3245 if self.params.get('overwrites', True) and os.path.exists(encodeFilename(linkfn)):
3246 self.to_screen(f'[info] Internet shortcut (.{link_type}) is already present')
3247 return True
3248 try:
3249 self.to_screen(f'[info] Writing internet shortcut (.{link_type}) to: {linkfn}')
3250 with open(encodeFilename(to_high_limit_path(linkfn)), 'w', encoding='utf-8',
3251 newline='\r\n' if link_type == 'url' else '\n') as linkfile:
3252 template_vars = {'url': url}
3253 if link_type == 'desktop':
3254 template_vars['filename'] = linkfn[:-(len(link_type) + 1)]
3255 linkfile.write(LINK_TEMPLATES[link_type] % template_vars)
3256 except OSError:
3257 self.report_error(f'Cannot write internet shortcut {linkfn}')
3258 return False
3259 return True
3260
3261 write_links = {
3262 'url': self.params.get('writeurllink'),
3263 'webloc': self.params.get('writewebloclink'),
3264 'desktop': self.params.get('writedesktoplink'),
3265 }
3266 if self.params.get('writelink'):
3267 link_type = ('webloc' if sys.platform == 'darwin'
3268 else 'desktop' if sys.platform.startswith('linux')
3269 else 'url')
3270 write_links[link_type] = True
3271
3272 if any(should_write and not _write_link_file(link_type)
3273 for link_type, should_write in write_links.items()):
3274 return
3275
3276 new_info, files_to_move = self.pre_process(info_dict, 'before_dl', files_to_move)
3277 replace_info_dict(new_info)
3278
3279 if self.params.get('skip_download'):
3280 info_dict['filepath'] = temp_filename
3281 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
3282 info_dict['__files_to_move'] = files_to_move
3283 replace_info_dict(self.run_pp(MoveFilesAfterDownloadPP(self, False), info_dict))
3284 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
3285 else:
3286 # Download
3287 info_dict.setdefault('__postprocessors', [])
3288 try:
3289
3290 def existing_video_file(*filepaths):
3291 ext = info_dict.get('ext')
3292 converted = lambda file: replace_extension(file, self.params.get('final_ext') or ext, ext)
3293 file = self.existing_file(itertools.chain(*zip(map(converted, filepaths), filepaths)),
3294 default_overwrite=False)
3295 if file:
3296 info_dict['ext'] = os.path.splitext(file)[1][1:]
3297 return file
3298
3299 fd, success = None, True
3300 if info_dict.get('protocol') or info_dict.get('url'):
3301 fd = get_suitable_downloader(info_dict, self.params, to_stdout=temp_filename == '-')
3302 if fd != FFmpegFD and 'no-direct-merge' not in self.params['compat_opts'] and (
3303 info_dict.get('section_start') or info_dict.get('section_end')):
3304 msg = ('This format cannot be partially downloaded' if FFmpegFD.available()
3305 else 'You have requested downloading the video partially, but ffmpeg is not installed')
3306 self.report_error(f'{msg}. Aborting')
3307 return
3308
3309 if info_dict.get('requested_formats') is not None:
3310 old_ext = info_dict['ext']
3311 if self.params.get('merge_output_format') is None:
3312 if (info_dict['ext'] == 'webm'
3313 and info_dict.get('thumbnails')
3314 # check with type instead of pp_key, __name__, or isinstance
3315 # since we dont want any custom PPs to trigger this
3316 and any(type(pp) == EmbedThumbnailPP for pp in self._pps['post_process'])): # noqa: E721
3317 info_dict['ext'] = 'mkv'
3318 self.report_warning(
3319 'webm doesn\'t support embedding a thumbnail, mkv will be used')
3320 new_ext = info_dict['ext']
3321
3322 def correct_ext(filename, ext=new_ext):
3323 if filename == '-':
3324 return filename
3325 filename_real_ext = os.path.splitext(filename)[1][1:]
3326 filename_wo_ext = (
3327 os.path.splitext(filename)[0]
3328 if filename_real_ext in (old_ext, new_ext)
3329 else filename)
3330 return f'{filename_wo_ext}.{ext}'
3331
3332 # Ensure filename always has a correct extension for successful merge
3333 full_filename = correct_ext(full_filename)
3334 temp_filename = correct_ext(temp_filename)
3335 dl_filename = existing_video_file(full_filename, temp_filename)
3336
3337 info_dict['__real_download'] = False
3338 # NOTE: Copy so that original format dicts are not modified
3339 info_dict['requested_formats'] = list(map(dict, info_dict['requested_formats']))
3340
3341 merger = FFmpegMergerPP(self)
3342 downloaded = []
3343 if dl_filename is not None:
3344 self.report_file_already_downloaded(dl_filename)
3345 elif fd:
3346 for f in info_dict['requested_formats'] if fd != FFmpegFD else []:
3347 f['filepath'] = fname = prepend_extension(
3348 correct_ext(temp_filename, info_dict['ext']),
3349 'f%s' % f['format_id'], info_dict['ext'])
3350 downloaded.append(fname)
3351 info_dict['url'] = '\n'.join(f['url'] for f in info_dict['requested_formats'])
3352 success, real_download = self.dl(temp_filename, info_dict)
3353 info_dict['__real_download'] = real_download
3354 else:
3355 if self.params.get('allow_unplayable_formats'):
3356 self.report_warning(
3357 'You have requested merging of multiple formats '
3358 'while also allowing unplayable formats to be downloaded. '
3359 'The formats won\'t be merged to prevent data corruption.')
3360 elif not merger.available:
3361 msg = 'You have requested merging of multiple formats but ffmpeg is not installed'
3362 if not self.params.get('ignoreerrors'):
3363 self.report_error(f'{msg}. Aborting due to --abort-on-error')
3364 return
3365 self.report_warning(f'{msg}. The formats won\'t be merged')
3366
3367 if temp_filename == '-':
3368 reason = ('using a downloader other than ffmpeg' if FFmpegFD.can_merge_formats(info_dict, self.params)
3369 else 'but the formats are incompatible for simultaneous download' if merger.available
3370 else 'but ffmpeg is not installed')
3371 self.report_warning(
3372 f'You have requested downloading multiple formats to stdout {reason}. '
3373 'The formats will be streamed one after the other')
3374 fname = temp_filename
3375 for f in info_dict['requested_formats']:
3376 new_info = dict(info_dict)
3377 del new_info['requested_formats']
3378 new_info.update(f)
3379 if temp_filename != '-':
3380 fname = prepend_extension(
3381 correct_ext(temp_filename, new_info['ext']),
3382 'f%s' % f['format_id'], new_info['ext'])
3383 if not self._ensure_dir_exists(fname):
3384 return
3385 f['filepath'] = fname
3386 downloaded.append(fname)
3387 partial_success, real_download = self.dl(fname, new_info)
3388 info_dict['__real_download'] = info_dict['__real_download'] or real_download
3389 success = success and partial_success
3390
3391 if downloaded and merger.available and not self.params.get('allow_unplayable_formats'):
3392 info_dict['__postprocessors'].append(merger)
3393 info_dict['__files_to_merge'] = downloaded
3394 # Even if there were no downloads, it is being merged only now
3395 info_dict['__real_download'] = True
3396 else:
3397 for file in downloaded:
3398 files_to_move[file] = None
3399 else:
3400 # Just a single file
3401 dl_filename = existing_video_file(full_filename, temp_filename)
3402 if dl_filename is None or dl_filename == temp_filename:
3403 # dl_filename == temp_filename could mean that the file was partially downloaded with --no-part.
3404 # So we should try to resume the download
3405 success, real_download = self.dl(temp_filename, info_dict)
3406 info_dict['__real_download'] = real_download
3407 else:
3408 self.report_file_already_downloaded(dl_filename)
3409
3410 dl_filename = dl_filename or temp_filename
3411 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
3412
3413 except network_exceptions as err:
3414 self.report_error('unable to download video data: %s' % error_to_compat_str(err))
3415 return
3416 except OSError as err:
3417 raise UnavailableVideoError(err)
3418 except (ContentTooShortError, ) as err:
3419 self.report_error(f'content too short (expected {err.expected} bytes and served {err.downloaded})')
3420 return
3421
3422 self._raise_pending_errors(info_dict)
3423 if success and full_filename != '-':
3424
3425 def fixup():
3426 do_fixup = True
3427 fixup_policy = self.params.get('fixup')
3428 vid = info_dict['id']
3429
3430 if fixup_policy in ('ignore', 'never'):
3431 return
3432 elif fixup_policy == 'warn':
3433 do_fixup = 'warn'
3434 elif fixup_policy != 'force':
3435 assert fixup_policy in ('detect_or_warn', None)
3436 if not info_dict.get('__real_download'):
3437 do_fixup = False
3438
3439 def ffmpeg_fixup(cndn, msg, cls):
3440 if not (do_fixup and cndn):
3441 return
3442 elif do_fixup == 'warn':
3443 self.report_warning(f'{vid}: {msg}')
3444 return
3445 pp = cls(self)
3446 if pp.available:
3447 info_dict['__postprocessors'].append(pp)
3448 else:
3449 self.report_warning(f'{vid}: {msg}. Install ffmpeg to fix this automatically')
3450
3451 stretched_ratio = info_dict.get('stretched_ratio')
3452 ffmpeg_fixup(stretched_ratio not in (1, None),
3453 f'Non-uniform pixel ratio {stretched_ratio}',
3454 FFmpegFixupStretchedPP)
3455
3456 downloader = get_suitable_downloader(info_dict, self.params) if 'protocol' in info_dict else None
3457 downloader = downloader.FD_NAME if downloader else None
3458
3459 ext = info_dict.get('ext')
3460 postprocessed_by_ffmpeg = info_dict.get('requested_formats') or any((
3461 isinstance(pp, FFmpegVideoConvertorPP)
3462 and resolve_recode_mapping(ext, pp.mapping)[0] not in (ext, None)
3463 ) for pp in self._pps['post_process'])
3464
3465 if not postprocessed_by_ffmpeg:
3466 ffmpeg_fixup(fd != FFmpegFD and ext == 'm4a'
3467 and info_dict.get('container') == 'm4a_dash',
3468 'writing DASH m4a. Only some players support this container',
3469 FFmpegFixupM4aPP)
3470 ffmpeg_fixup(downloader == 'hlsnative' and not self.params.get('hls_use_mpegts')
3471 or info_dict.get('is_live') and self.params.get('hls_use_mpegts') is None,
3472 'Possible MPEG-TS in MP4 container or malformed AAC timestamps',
3473 FFmpegFixupM3u8PP)
3474 ffmpeg_fixup(info_dict.get('is_live') and downloader == 'dashsegments',
3475 'Possible duplicate MOOV atoms', FFmpegFixupDuplicateMoovPP)
3476
3477 ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed timestamps detected', FFmpegFixupTimestampPP)
3478 ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed duration detected', FFmpegFixupDurationPP)
3479
3480 fixup()
3481 try:
3482 replace_info_dict(self.post_process(dl_filename, info_dict, files_to_move))
3483 except PostProcessingError as err:
3484 self.report_error('Postprocessing: %s' % str(err))
3485 return
3486 try:
3487 for ph in self._post_hooks:
3488 ph(info_dict['filepath'])
3489 except Exception as err:
3490 self.report_error('post hooks: %s' % str(err))
3491 return
3492 info_dict['__write_download_archive'] = True
3493
3494 assert info_dict is original_infodict # Make sure the info_dict was modified in-place
3495 if self.params.get('force_write_download_archive'):
3496 info_dict['__write_download_archive'] = True
3497 check_max_downloads()
3498
3499 def __download_wrapper(self, func):
3500 @functools.wraps(func)
3501 def wrapper(*args, **kwargs):
3502 try:
3503 res = func(*args, **kwargs)
3504 except UnavailableVideoError as e:
3505 self.report_error(e)
3506 except DownloadCancelled as e:
3507 self.to_screen(f'[info] {e}')
3508 if not self.params.get('break_per_url'):
3509 raise
3510 self._num_downloads = 0
3511 else:
3512 if self.params.get('dump_single_json', False):
3513 self.post_extract(res)
3514 self.to_stdout(json.dumps(self.sanitize_info(res)))
3515 return wrapper
3516
3517 def download(self, url_list):
3518 """Download a given list of URLs."""
3519 url_list = variadic(url_list) # Passing a single URL is a common mistake
3520 outtmpl = self.params['outtmpl']['default']
3521 if (len(url_list) > 1
3522 and outtmpl != '-'
3523 and '%' not in outtmpl
3524 and self.params.get('max_downloads') != 1):
3525 raise SameFileError(outtmpl)
3526
3527 for url in url_list:
3528 self.__download_wrapper(self.extract_info)(
3529 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
3530
3531 return self._download_retcode
3532
3533 def download_with_info_file(self, info_filename):
3534 with contextlib.closing(fileinput.FileInput(
3535 [info_filename], mode='r',
3536 openhook=fileinput.hook_encoded('utf-8'))) as f:
3537 # FileInput doesn't have a read method, we can't call json.load
3538 infos = [self.sanitize_info(info, self.params.get('clean_infojson', True))
3539 for info in variadic(json.loads('\n'.join(f)))]
3540 for info in infos:
3541 try:
3542 self.__download_wrapper(self.process_ie_result)(info, download=True)
3543 except (DownloadError, EntryNotInPlaylist, ReExtractInfo) as e:
3544 if not isinstance(e, EntryNotInPlaylist):
3545 self.to_stderr('\r')
3546 webpage_url = info.get('webpage_url')
3547 if webpage_url is None:
3548 raise
3549 self.report_warning(f'The info failed to download: {e}; trying with URL {webpage_url}')
3550 self.download([webpage_url])
3551 return self._download_retcode
3552
3553 @staticmethod
3554 def sanitize_info(info_dict, remove_private_keys=False):
3555 ''' Sanitize the infodict for converting to json '''
3556 if info_dict is None:
3557 return info_dict
3558 info_dict.setdefault('epoch', int(time.time()))
3559 info_dict.setdefault('_type', 'video')
3560 info_dict.setdefault('_version', {
3561 'version': __version__,
3562 'current_git_head': current_git_head(),
3563 'release_git_head': RELEASE_GIT_HEAD,
3564 'repository': ORIGIN,
3565 })
3566
3567 if remove_private_keys:
3568 reject = lambda k, v: v is None or k.startswith('__') or k in {
3569 'requested_downloads', 'requested_formats', 'requested_subtitles', 'requested_entries',
3570 'entries', 'filepath', '_filename', 'filename', 'infojson_filename', 'original_url',
3571 'playlist_autonumber',
3572 }
3573 else:
3574 reject = lambda k, v: False
3575
3576 def filter_fn(obj):
3577 if isinstance(obj, dict):
3578 return {k: filter_fn(v) for k, v in obj.items() if not reject(k, v)}
3579 elif isinstance(obj, (list, tuple, set, LazyList)):
3580 return list(map(filter_fn, obj))
3581 elif obj is None or isinstance(obj, (str, int, float, bool)):
3582 return obj
3583 else:
3584 return repr(obj)
3585
3586 return filter_fn(info_dict)
3587
3588 @staticmethod
3589 def filter_requested_info(info_dict, actually_filter=True):
3590 ''' Alias of sanitize_info for backward compatibility '''
3591 return YoutubeDL.sanitize_info(info_dict, actually_filter)
3592
3593 def _delete_downloaded_files(self, *files_to_delete, info={}, msg=None):
3594 for filename in set(filter(None, files_to_delete)):
3595 if msg:
3596 self.to_screen(msg % filename)
3597 try:
3598 os.remove(filename)
3599 except OSError:
3600 self.report_warning(f'Unable to delete file {filename}')
3601 if filename in info.get('__files_to_move', []): # NB: Delete even if None
3602 del info['__files_to_move'][filename]
3603
3604 @staticmethod
3605 def post_extract(info_dict):
3606 def actual_post_extract(info_dict):
3607 if info_dict.get('_type') in ('playlist', 'multi_video'):
3608 for video_dict in info_dict.get('entries', {}):
3609 actual_post_extract(video_dict or {})
3610 return
3611
3612 post_extractor = info_dict.pop('__post_extractor', None) or (lambda: {})
3613 info_dict.update(post_extractor())
3614
3615 actual_post_extract(info_dict or {})
3616
3617 def run_pp(self, pp, infodict):
3618 files_to_delete = []
3619 if '__files_to_move' not in infodict:
3620 infodict['__files_to_move'] = {}
3621 try:
3622 files_to_delete, infodict = pp.run(infodict)
3623 except PostProcessingError as e:
3624 # Must be True and not 'only_download'
3625 if self.params.get('ignoreerrors') is True:
3626 self.report_error(e)
3627 return infodict
3628 raise
3629
3630 if not files_to_delete:
3631 return infodict
3632 if self.params.get('keepvideo', False):
3633 for f in files_to_delete:
3634 infodict['__files_to_move'].setdefault(f, '')
3635 else:
3636 self._delete_downloaded_files(
3637 *files_to_delete, info=infodict, msg='Deleting original file %s (pass -k to keep)')
3638 return infodict
3639
3640 def run_all_pps(self, key, info, *, additional_pps=None):
3641 if key != 'video':
3642 self._forceprint(key, info)
3643 for pp in (additional_pps or []) + self._pps[key]:
3644 info = self.run_pp(pp, info)
3645 return info
3646
3647 def pre_process(self, ie_info, key='pre_process', files_to_move=None):
3648 info = dict(ie_info)
3649 info['__files_to_move'] = files_to_move or {}
3650 try:
3651 info = self.run_all_pps(key, info)
3652 except PostProcessingError as err:
3653 msg = f'Preprocessing: {err}'
3654 info.setdefault('__pending_error', msg)
3655 self.report_error(msg, is_error=False)
3656 return info, info.pop('__files_to_move', None)
3657
3658 def post_process(self, filename, info, files_to_move=None):
3659 """Run all the postprocessors on the given file."""
3660 info['filepath'] = filename
3661 info['__files_to_move'] = files_to_move or {}
3662 info = self.run_all_pps('post_process', info, additional_pps=info.get('__postprocessors'))
3663 info = self.run_pp(MoveFilesAfterDownloadPP(self), info)
3664 del info['__files_to_move']
3665 return self.run_all_pps('after_move', info)
3666
3667 def _make_archive_id(self, info_dict):
3668 video_id = info_dict.get('id')
3669 if not video_id:
3670 return
3671 # Future-proof against any change in case
3672 # and backwards compatibility with prior versions
3673 extractor = info_dict.get('extractor_key') or info_dict.get('ie_key') # key in a playlist
3674 if extractor is None:
3675 url = str_or_none(info_dict.get('url'))
3676 if not url:
3677 return
3678 # Try to find matching extractor for the URL and take its ie_key
3679 for ie_key, ie in self._ies.items():
3680 if ie.suitable(url):
3681 extractor = ie_key
3682 break
3683 else:
3684 return
3685 return make_archive_id(extractor, video_id)
3686
3687 def in_download_archive(self, info_dict):
3688 if not self.archive:
3689 return False
3690
3691 vid_ids = [self._make_archive_id(info_dict)]
3692 vid_ids.extend(info_dict.get('_old_archive_ids') or [])
3693 return any(id_ in self.archive for id_ in vid_ids)
3694
3695 def record_download_archive(self, info_dict):
3696 fn = self.params.get('download_archive')
3697 if fn is None:
3698 return
3699 vid_id = self._make_archive_id(info_dict)
3700 assert vid_id
3701
3702 self.write_debug(f'Adding to archive: {vid_id}')
3703 if is_path_like(fn):
3704 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
3705 archive_file.write(vid_id + '\n')
3706 self.archive.add(vid_id)
3707
3708 @staticmethod
3709 def format_resolution(format, default='unknown'):
3710 if format.get('vcodec') == 'none' and format.get('acodec') != 'none':
3711 return 'audio only'
3712 if format.get('resolution') is not None:
3713 return format['resolution']
3714 if format.get('width') and format.get('height'):
3715 return '%dx%d' % (format['width'], format['height'])
3716 elif format.get('height'):
3717 return '%sp' % format['height']
3718 elif format.get('width'):
3719 return '%dx?' % format['width']
3720 return default
3721
3722 def _list_format_headers(self, *headers):
3723 if self.params.get('listformats_table', True) is not False:
3724 return [self._format_out(header, self.Styles.HEADERS) for header in headers]
3725 return headers
3726
3727 def _format_note(self, fdict):
3728 res = ''
3729 if fdict.get('ext') in ['f4f', 'f4m']:
3730 res += '(unsupported)'
3731 if fdict.get('language'):
3732 if res:
3733 res += ' '
3734 res += '[%s]' % fdict['language']
3735 if fdict.get('format_note') is not None:
3736 if res:
3737 res += ' '
3738 res += fdict['format_note']
3739 if fdict.get('tbr') is not None:
3740 if res:
3741 res += ', '
3742 res += '%4dk' % fdict['tbr']
3743 if fdict.get('container') is not None:
3744 if res:
3745 res += ', '
3746 res += '%s container' % fdict['container']
3747 if (fdict.get('vcodec') is not None
3748 and fdict.get('vcodec') != 'none'):
3749 if res:
3750 res += ', '
3751 res += fdict['vcodec']
3752 if fdict.get('vbr') is not None:
3753 res += '@'
3754 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
3755 res += 'video@'
3756 if fdict.get('vbr') is not None:
3757 res += '%4dk' % fdict['vbr']
3758 if fdict.get('fps') is not None:
3759 if res:
3760 res += ', '
3761 res += '%sfps' % fdict['fps']
3762 if fdict.get('acodec') is not None:
3763 if res:
3764 res += ', '
3765 if fdict['acodec'] == 'none':
3766 res += 'video only'
3767 else:
3768 res += '%-5s' % fdict['acodec']
3769 elif fdict.get('abr') is not None:
3770 if res:
3771 res += ', '
3772 res += 'audio'
3773 if fdict.get('abr') is not None:
3774 res += '@%3dk' % fdict['abr']
3775 if fdict.get('asr') is not None:
3776 res += ' (%5dHz)' % fdict['asr']
3777 if fdict.get('filesize') is not None:
3778 if res:
3779 res += ', '
3780 res += format_bytes(fdict['filesize'])
3781 elif fdict.get('filesize_approx') is not None:
3782 if res:
3783 res += ', '
3784 res += '~' + format_bytes(fdict['filesize_approx'])
3785 return res
3786
3787 def _get_formats(self, info_dict):
3788 if info_dict.get('formats') is None:
3789 if info_dict.get('url') and info_dict.get('_type', 'video') == 'video':
3790 return [info_dict]
3791 return []
3792 return info_dict['formats']
3793
3794 def render_formats_table(self, info_dict):
3795 formats = self._get_formats(info_dict)
3796 if not formats:
3797 return
3798 if not self.params.get('listformats_table', True) is not False:
3799 table = [
3800 [
3801 format_field(f, 'format_id'),
3802 format_field(f, 'ext'),
3803 self.format_resolution(f),
3804 self._format_note(f)
3805 ] for f in formats if (f.get('preference') or 0) >= -1000]
3806 return render_table(['format code', 'extension', 'resolution', 'note'], table, extra_gap=1)
3807
3808 def simplified_codec(f, field):
3809 assert field in ('acodec', 'vcodec')
3810 codec = f.get(field)
3811 if not codec:
3812 return 'unknown'
3813 elif codec != 'none':
3814 return '.'.join(codec.split('.')[:4])
3815
3816 if field == 'vcodec' and f.get('acodec') == 'none':
3817 return 'images'
3818 elif field == 'acodec' and f.get('vcodec') == 'none':
3819 return ''
3820 return self._format_out('audio only' if field == 'vcodec' else 'video only',
3821 self.Styles.SUPPRESS)
3822
3823 delim = self._format_out('\u2502', self.Styles.DELIM, '|', test_encoding=True)
3824 table = [
3825 [
3826 self._format_out(format_field(f, 'format_id'), self.Styles.ID),
3827 format_field(f, 'ext'),
3828 format_field(f, func=self.format_resolution, ignore=('audio only', 'images')),
3829 format_field(f, 'fps', '\t%d', func=round),
3830 format_field(f, 'dynamic_range', '%s', ignore=(None, 'SDR')).replace('HDR', ''),
3831 format_field(f, 'audio_channels', '\t%s'),
3832 delim, (
3833 format_field(f, 'filesize', ' \t%s', func=format_bytes)
3834 or format_field(f, 'filesize_approx', '≈\t%s', func=format_bytes)
3835 or format_field(try_call(lambda: format_bytes(int(info_dict['duration'] * f['tbr'] * (1024 / 8)))),
3836 None, self._format_out('~\t%s', self.Styles.SUPPRESS))),
3837 format_field(f, 'tbr', '\t%dk', func=round),
3838 shorten_protocol_name(f.get('protocol', '')),
3839 delim,
3840 simplified_codec(f, 'vcodec'),
3841 format_field(f, 'vbr', '\t%dk', func=round),
3842 simplified_codec(f, 'acodec'),
3843 format_field(f, 'abr', '\t%dk', func=round),
3844 format_field(f, 'asr', '\t%s', func=format_decimal_suffix),
3845 join_nonempty(format_field(f, 'language', '[%s]'), join_nonempty(
3846 self._format_out('UNSUPPORTED', self.Styles.BAD_FORMAT) if f.get('ext') in ('f4f', 'f4m') else None,
3847 (self._format_out('Maybe DRM', self.Styles.WARNING) if f.get('has_drm') == 'maybe'
3848 else self._format_out('DRM', self.Styles.BAD_FORMAT) if f.get('has_drm') else None),
3849 format_field(f, 'format_note'),
3850 format_field(f, 'container', ignore=(None, f.get('ext'))),
3851 delim=', '), delim=' '),
3852 ] for f in formats if f.get('preference') is None or f['preference'] >= -1000]
3853 header_line = self._list_format_headers(
3854 'ID', 'EXT', 'RESOLUTION', '\tFPS', 'HDR', 'CH', delim, '\tFILESIZE', '\tTBR', 'PROTO',
3855 delim, 'VCODEC', '\tVBR', 'ACODEC', '\tABR', '\tASR', 'MORE INFO')
3856
3857 return render_table(
3858 header_line, table, hide_empty=True,
3859 delim=self._format_out('\u2500', self.Styles.DELIM, '-', test_encoding=True))
3860
3861 def render_thumbnails_table(self, info_dict):
3862 thumbnails = list(info_dict.get('thumbnails') or [])
3863 if not thumbnails:
3864 return None
3865 return render_table(
3866 self._list_format_headers('ID', 'Width', 'Height', 'URL'),
3867 [[t.get('id'), t.get('width') or 'unknown', t.get('height') or 'unknown', t['url']] for t in thumbnails])
3868
3869 def render_subtitles_table(self, video_id, subtitles):
3870 def _row(lang, formats):
3871 exts, names = zip(*((f['ext'], f.get('name') or 'unknown') for f in reversed(formats)))
3872 if len(set(names)) == 1:
3873 names = [] if names[0] == 'unknown' else names[:1]
3874 return [lang, ', '.join(names), ', '.join(exts)]
3875
3876 if not subtitles:
3877 return None
3878 return render_table(
3879 self._list_format_headers('Language', 'Name', 'Formats'),
3880 [_row(lang, formats) for lang, formats in subtitles.items()],
3881 hide_empty=True)
3882
3883 def __list_table(self, video_id, name, func, *args):
3884 table = func(*args)
3885 if not table:
3886 self.to_screen(f'{video_id} has no {name}')
3887 return
3888 self.to_screen(f'[info] Available {name} for {video_id}:')
3889 self.to_stdout(table)
3890
3891 def list_formats(self, info_dict):
3892 self.__list_table(info_dict['id'], 'formats', self.render_formats_table, info_dict)
3893
3894 def list_thumbnails(self, info_dict):
3895 self.__list_table(info_dict['id'], 'thumbnails', self.render_thumbnails_table, info_dict)
3896
3897 def list_subtitles(self, video_id, subtitles, name='subtitles'):
3898 self.__list_table(video_id, name, self.render_subtitles_table, video_id, subtitles)
3899
3900 def print_debug_header(self):
3901 if not self.params.get('verbose'):
3902 return
3903
3904 from . import _IN_CLI # Must be delayed import
3905
3906 # These imports can be slow. So import them only as needed
3907 from .extractor.extractors import _LAZY_LOADER
3908 from .extractor.extractors import (
3909 _PLUGIN_CLASSES as plugin_ies,
3910 _PLUGIN_OVERRIDES as plugin_ie_overrides
3911 )
3912
3913 def get_encoding(stream):
3914 ret = str(getattr(stream, 'encoding', 'missing (%s)' % type(stream).__name__))
3915 additional_info = []
3916 if os.environ.get('TERM', '').lower() == 'dumb':
3917 additional_info.append('dumb')
3918 if not supports_terminal_sequences(stream):
3919 from .utils import WINDOWS_VT_MODE # Must be imported locally
3920 additional_info.append('No VT' if WINDOWS_VT_MODE is False else 'No ANSI')
3921 if additional_info:
3922 ret = f'{ret} ({",".join(additional_info)})'
3923 return ret
3924
3925 encoding_str = 'Encodings: locale %s, fs %s, pref %s, %s' % (
3926 locale.getpreferredencoding(),
3927 sys.getfilesystemencoding(),
3928 self.get_encoding(),
3929 ', '.join(
3930 f'{key} {get_encoding(stream)}' for key, stream in self._out_files.items_
3931 if stream is not None and key != 'console')
3932 )
3933
3934 logger = self.params.get('logger')
3935 if logger:
3936 write_debug = lambda msg: logger.debug(f'[debug] {msg}')
3937 write_debug(encoding_str)
3938 else:
3939 write_string(f'[debug] {encoding_str}\n', encoding=None)
3940 write_debug = lambda msg: self._write_string(f'[debug] {msg}\n')
3941
3942 source = detect_variant()
3943 if VARIANT not in (None, 'pip'):
3944 source += '*'
3945 klass = type(self)
3946 write_debug(join_nonempty(
3947 f'{REPOSITORY.rpartition("/")[2]} version',
3948 _make_label(ORIGIN, CHANNEL.partition('@')[2] or __version__, __version__),
3949 f'[{RELEASE_GIT_HEAD[:9]}]' if RELEASE_GIT_HEAD else '',
3950 '' if source == 'unknown' else f'({source})',
3951 '' if _IN_CLI else 'API' if klass == YoutubeDL else f'API:{self.__module__}.{klass.__qualname__}',
3952 delim=' '))
3953
3954 if not _IN_CLI:
3955 write_debug(f'params: {self.params}')
3956
3957 if not _LAZY_LOADER:
3958 if os.environ.get('YTDLP_NO_LAZY_EXTRACTORS'):
3959 write_debug('Lazy loading extractors is forcibly disabled')
3960 else:
3961 write_debug('Lazy loading extractors is disabled')
3962 if self.params['compat_opts']:
3963 write_debug('Compatibility options: %s' % ', '.join(self.params['compat_opts']))
3964
3965 if current_git_head():
3966 write_debug(f'Git HEAD: {current_git_head()}')
3967 write_debug(system_identifier())
3968
3969 exe_versions, ffmpeg_features = FFmpegPostProcessor.get_versions_and_features(self)
3970 ffmpeg_features = {key for key, val in ffmpeg_features.items() if val}
3971 if ffmpeg_features:
3972 exe_versions['ffmpeg'] += ' (%s)' % ','.join(sorted(ffmpeg_features))
3973
3974 exe_versions['rtmpdump'] = rtmpdump_version()
3975 exe_versions['phantomjs'] = PhantomJSwrapper._version()
3976 exe_str = ', '.join(
3977 f'{exe} {v}' for exe, v in sorted(exe_versions.items()) if v
3978 ) or 'none'
3979 write_debug('exe versions: %s' % exe_str)
3980
3981 from .compat.compat_utils import get_package_info
3982 from .dependencies import available_dependencies
3983
3984 write_debug('Optional libraries: %s' % (', '.join(sorted({
3985 join_nonempty(*get_package_info(m)) for m in available_dependencies.values()
3986 })) or 'none'))
3987
3988 write_debug(f'Proxy map: {self.proxies}')
3989 write_debug(f'Request Handlers: {", ".join(rh.RH_NAME for rh in self._request_director.handlers.values())}')
3990 for plugin_type, plugins in {'Extractor': plugin_ies, 'Post-Processor': plugin_pps}.items():
3991 display_list = ['%s%s' % (
3992 klass.__name__, '' if klass.__name__ == name else f' as {name}')
3993 for name, klass in plugins.items()]
3994 if plugin_type == 'Extractor':
3995 display_list.extend(f'{plugins[-1].IE_NAME.partition("+")[2]} ({parent.__name__})'
3996 for parent, plugins in plugin_ie_overrides.items())
3997 if not display_list:
3998 continue
3999 write_debug(f'{plugin_type} Plugins: {", ".join(sorted(display_list))}')
4000
4001 plugin_dirs = plugin_directories()
4002 if plugin_dirs:
4003 write_debug(f'Plugin directories: {plugin_dirs}')
4004
4005 # Not implemented
4006 if False and self.params.get('call_home'):
4007 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode()
4008 write_debug('Public IP address: %s' % ipaddr)
4009 latest_version = self.urlopen(
4010 'https://yt-dl.org/latest/version').read().decode()
4011 if version_tuple(latest_version) > version_tuple(__version__):
4012 self.report_warning(
4013 'You are using an outdated version (newest version: %s)! '
4014 'See https://yt-dl.org/update if you need help updating.' %
4015 latest_version)
4016
4017 @functools.cached_property
4018 def proxies(self):
4019 """Global proxy configuration"""
4020 opts_proxy = self.params.get('proxy')
4021 if opts_proxy is not None:
4022 if opts_proxy == '':
4023 opts_proxy = '__noproxy__'
4024 proxies = {'all': opts_proxy}
4025 else:
4026 proxies = urllib.request.getproxies()
4027 # compat. Set HTTPS_PROXY to __noproxy__ to revert
4028 if 'http' in proxies and 'https' not in proxies:
4029 proxies['https'] = proxies['http']
4030
4031 return proxies
4032
4033 @functools.cached_property
4034 def cookiejar(self):
4035 """Global cookiejar instance"""
4036 return load_cookies(
4037 self.params.get('cookiefile'), self.params.get('cookiesfrombrowser'), self)
4038
4039 @property
4040 def _opener(self):
4041 """
4042 Get a urllib OpenerDirector from the Urllib handler (deprecated).
4043 """
4044 self.deprecation_warning('YoutubeDL._opener is deprecated, use YoutubeDL.urlopen()')
4045 handler = self._request_director.handlers['Urllib']
4046 return handler._get_instance(cookiejar=self.cookiejar, proxies=self.proxies)
4047
4048 def urlopen(self, req):
4049 """ Start an HTTP download """
4050 if isinstance(req, str):
4051 req = Request(req)
4052 elif isinstance(req, urllib.request.Request):
4053 self.deprecation_warning(
4054 'Passing a urllib.request.Request object to YoutubeDL.urlopen() is deprecated. '
4055 'Use yt_dlp.networking.common.Request instead.')
4056 req = urllib_req_to_req(req)
4057 assert isinstance(req, Request)
4058
4059 # compat: Assume user:pass url params are basic auth
4060 url, basic_auth_header = extract_basic_auth(req.url)
4061 if basic_auth_header:
4062 req.headers['Authorization'] = basic_auth_header
4063 req.url = sanitize_url(url)
4064
4065 clean_proxies(proxies=req.proxies, headers=req.headers)
4066 clean_headers(req.headers)
4067
4068 try:
4069 return self._request_director.send(req)
4070 except NoSupportingHandlers as e:
4071 for ue in e.unsupported_errors:
4072 # FIXME: This depends on the order of errors.
4073 if not (ue.handler and ue.msg):
4074 continue
4075 if ue.handler.RH_KEY == 'Urllib' and 'unsupported url scheme: "file"' in ue.msg.lower():
4076 raise RequestError(
4077 'file:// URLs are disabled by default in yt-dlp for security reasons. '
4078 'Use --enable-file-urls to enable at your own risk.', cause=ue) from ue
4079 if 'unsupported proxy type: "https"' in ue.msg.lower():
4080 raise RequestError(
4081 'To use an HTTPS proxy for this request, one of the following dependencies needs to be installed: requests')
4082
4083 elif (
4084 re.match(r'unsupported url scheme: "wss?"', ue.msg.lower())
4085 and 'websockets' not in self._request_director.handlers
4086 ):
4087 raise RequestError(
4088 'This request requires WebSocket support. '
4089 'Ensure one of the following dependencies are installed: websockets',
4090 cause=ue) from ue
4091 raise
4092 except SSLError as e:
4093 if 'UNSAFE_LEGACY_RENEGOTIATION_DISABLED' in str(e):
4094 raise RequestError('UNSAFE_LEGACY_RENEGOTIATION_DISABLED: Try using --legacy-server-connect', cause=e) from e
4095 elif 'SSLV3_ALERT_HANDSHAKE_FAILURE' in str(e):
4096 raise RequestError(
4097 'SSLV3_ALERT_HANDSHAKE_FAILURE: The server may not support the current cipher list. '
4098 'Try using --legacy-server-connect', cause=e) from e
4099 raise
4100 except HTTPError as e: # TODO: Remove in a future release
4101 raise _CompatHTTPError(e) from e
4102
4103 def build_request_director(self, handlers, preferences=None):
4104 logger = _YDLLogger(self)
4105 headers = self.params['http_headers'].copy()
4106 proxies = self.proxies.copy()
4107 clean_headers(headers)
4108 clean_proxies(proxies, headers)
4109
4110 director = RequestDirector(logger=logger, verbose=self.params.get('debug_printtraffic'))
4111 for handler in handlers:
4112 director.add_handler(handler(
4113 logger=logger,
4114 headers=headers,
4115 cookiejar=self.cookiejar,
4116 proxies=proxies,
4117 prefer_system_certs='no-certifi' in self.params['compat_opts'],
4118 verify=not self.params.get('nocheckcertificate'),
4119 **traverse_obj(self.params, {
4120 'verbose': 'debug_printtraffic',
4121 'source_address': 'source_address',
4122 'timeout': 'socket_timeout',
4123 'legacy_ssl_support': 'legacyserverconnect',
4124 'enable_file_urls': 'enable_file_urls',
4125 'client_cert': {
4126 'client_certificate': 'client_certificate',
4127 'client_certificate_key': 'client_certificate_key',
4128 'client_certificate_password': 'client_certificate_password',
4129 },
4130 }),
4131 ))
4132 director.preferences.update(preferences or [])
4133 if 'prefer-legacy-http-handler' in self.params['compat_opts']:
4134 director.preferences.add(lambda rh, _: 500 if rh.RH_KEY == 'Urllib' else 0)
4135 return director
4136
4137 def encode(self, s):
4138 if isinstance(s, bytes):
4139 return s # Already encoded
4140
4141 try:
4142 return s.encode(self.get_encoding())
4143 except UnicodeEncodeError as err:
4144 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
4145 raise
4146
4147 def get_encoding(self):
4148 encoding = self.params.get('encoding')
4149 if encoding is None:
4150 encoding = preferredencoding()
4151 return encoding
4152
4153 def _write_info_json(self, label, ie_result, infofn, overwrite=None):
4154 ''' Write infojson and returns True = written, 'exists' = Already exists, False = skip, None = error '''
4155 if overwrite is None:
4156 overwrite = self.params.get('overwrites', True)
4157 if not self.params.get('writeinfojson'):
4158 return False
4159 elif not infofn:
4160 self.write_debug(f'Skipping writing {label} infojson')
4161 return False
4162 elif not self._ensure_dir_exists(infofn):
4163 return None
4164 elif not overwrite and os.path.exists(infofn):
4165 self.to_screen(f'[info] {label.title()} metadata is already present')
4166 return 'exists'
4167
4168 self.to_screen(f'[info] Writing {label} metadata as JSON to: {infofn}')
4169 try:
4170 write_json_file(self.sanitize_info(ie_result, self.params.get('clean_infojson', True)), infofn)
4171 return True
4172 except OSError:
4173 self.report_error(f'Cannot write {label} metadata to JSON file {infofn}')
4174 return None
4175
4176 def _write_description(self, label, ie_result, descfn):
4177 ''' Write description and returns True = written, False = skip, None = error '''
4178 if not self.params.get('writedescription'):
4179 return False
4180 elif not descfn:
4181 self.write_debug(f'Skipping writing {label} description')
4182 return False
4183 elif not self._ensure_dir_exists(descfn):
4184 return None
4185 elif not self.params.get('overwrites', True) and os.path.exists(descfn):
4186 self.to_screen(f'[info] {label.title()} description is already present')
4187 elif ie_result.get('description') is None:
4188 self.to_screen(f'[info] There\'s no {label} description to write')
4189 return False
4190 else:
4191 try:
4192 self.to_screen(f'[info] Writing {label} description to: {descfn}')
4193 with open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
4194 descfile.write(ie_result['description'])
4195 except OSError:
4196 self.report_error(f'Cannot write {label} description file {descfn}')
4197 return None
4198 return True
4199
4200 def _write_subtitles(self, info_dict, filename):
4201 ''' Write subtitles to file and return list of (sub_filename, final_sub_filename); or None if error'''
4202 ret = []
4203 subtitles = info_dict.get('requested_subtitles')
4204 if not (self.params.get('writesubtitles') or self.params.get('writeautomaticsub')):
4205 # subtitles download errors are already managed as troubles in relevant IE
4206 # that way it will silently go on when used with unsupporting IE
4207 return ret
4208 elif not subtitles:
4209 self.to_screen('[info] There are no subtitles for the requested languages')
4210 return ret
4211 sub_filename_base = self.prepare_filename(info_dict, 'subtitle')
4212 if not sub_filename_base:
4213 self.to_screen('[info] Skipping writing video subtitles')
4214 return ret
4215
4216 for sub_lang, sub_info in subtitles.items():
4217 sub_format = sub_info['ext']
4218 sub_filename = subtitles_filename(filename, sub_lang, sub_format, info_dict.get('ext'))
4219 sub_filename_final = subtitles_filename(sub_filename_base, sub_lang, sub_format, info_dict.get('ext'))
4220 existing_sub = self.existing_file((sub_filename_final, sub_filename))
4221 if existing_sub:
4222 self.to_screen(f'[info] Video subtitle {sub_lang}.{sub_format} is already present')
4223 sub_info['filepath'] = existing_sub
4224 ret.append((existing_sub, sub_filename_final))
4225 continue
4226
4227 self.to_screen(f'[info] Writing video subtitles to: {sub_filename}')
4228 if sub_info.get('data') is not None:
4229 try:
4230 # Use newline='' to prevent conversion of newline characters
4231 # See https://github.com/ytdl-org/youtube-dl/issues/10268
4232 with open(sub_filename, 'w', encoding='utf-8', newline='') as subfile:
4233 subfile.write(sub_info['data'])
4234 sub_info['filepath'] = sub_filename
4235 ret.append((sub_filename, sub_filename_final))
4236 continue
4237 except OSError:
4238 self.report_error(f'Cannot write video subtitles file {sub_filename}')
4239 return None
4240
4241 try:
4242 sub_copy = sub_info.copy()
4243 sub_copy.setdefault('http_headers', info_dict.get('http_headers'))
4244 self.dl(sub_filename, sub_copy, subtitle=True)
4245 sub_info['filepath'] = sub_filename
4246 ret.append((sub_filename, sub_filename_final))
4247 except (DownloadError, ExtractorError, IOError, OSError, ValueError) + network_exceptions as err:
4248 msg = f'Unable to download video subtitles for {sub_lang!r}: {err}'
4249 if self.params.get('ignoreerrors') is not True: # False or 'only_download'
4250 if not self.params.get('ignoreerrors'):
4251 self.report_error(msg)
4252 raise DownloadError(msg)
4253 self.report_warning(msg)
4254 return ret
4255
4256 def _write_thumbnails(self, label, info_dict, filename, thumb_filename_base=None):
4257 ''' Write thumbnails to file and return list of (thumb_filename, final_thumb_filename); or None if error '''
4258 write_all = self.params.get('write_all_thumbnails', False)
4259 thumbnails, ret = [], []
4260 if write_all or self.params.get('writethumbnail', False):
4261 thumbnails = info_dict.get('thumbnails') or []
4262 if not thumbnails:
4263 self.to_screen(f'[info] There are no {label} thumbnails to download')
4264 return ret
4265 multiple = write_all and len(thumbnails) > 1
4266
4267 if thumb_filename_base is None:
4268 thumb_filename_base = filename
4269 if thumbnails and not thumb_filename_base:
4270 self.write_debug(f'Skipping writing {label} thumbnail')
4271 return ret
4272
4273 if thumbnails and not self._ensure_dir_exists(filename):
4274 return None
4275
4276 for idx, t in list(enumerate(thumbnails))[::-1]:
4277 thumb_ext = (f'{t["id"]}.' if multiple else '') + determine_ext(t['url'], 'jpg')
4278 thumb_display_id = f'{label} thumbnail {t["id"]}'
4279 thumb_filename = replace_extension(filename, thumb_ext, info_dict.get('ext'))
4280 thumb_filename_final = replace_extension(thumb_filename_base, thumb_ext, info_dict.get('ext'))
4281
4282 existing_thumb = self.existing_file((thumb_filename_final, thumb_filename))
4283 if existing_thumb:
4284 self.to_screen('[info] %s is already present' % (
4285 thumb_display_id if multiple else f'{label} thumbnail').capitalize())
4286 t['filepath'] = existing_thumb
4287 ret.append((existing_thumb, thumb_filename_final))
4288 else:
4289 self.to_screen(f'[info] Downloading {thumb_display_id} ...')
4290 try:
4291 uf = self.urlopen(Request(t['url'], headers=t.get('http_headers', {})))
4292 self.to_screen(f'[info] Writing {thumb_display_id} to: {thumb_filename}')
4293 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
4294 shutil.copyfileobj(uf, thumbf)
4295 ret.append((thumb_filename, thumb_filename_final))
4296 t['filepath'] = thumb_filename
4297 except network_exceptions as err:
4298 if isinstance(err, HTTPError) and err.status == 404:
4299 self.to_screen(f'[info] {thumb_display_id.title()} does not exist')
4300 else:
4301 self.report_warning(f'Unable to download {thumb_display_id}: {err}')
4302 thumbnails.pop(idx)
4303 if ret and not write_all:
4304 break
4305 return ret