]> jfr.im git - yt-dlp.git/blob - yt_dlp/YoutubeDL.py
[core] Parse `release_year` from `release_date` (#8524)
[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 }
1183 # Field is of the form key1.key2...
1184 # where keys (except first) can be string, int, slice or "{field, ...}"
1185 FIELD_INNER_RE = r'(?:\w+|%(num)s|%(num)s?(?::%(num)s?){1,2})' % {'num': r'(?:-?\d+)'}
1186 FIELD_RE = r'\w*(?:\.(?:%(inner)s|{%(field)s(?:,%(field)s)*}))*' % {
1187 'inner': FIELD_INNER_RE,
1188 'field': rf'\w*(?:\.{FIELD_INNER_RE})*'
1189 }
1190 MATH_FIELD_RE = rf'(?:{FIELD_RE}|-?{NUMBER_RE})'
1191 MATH_OPERATORS_RE = r'(?:%s)' % '|'.join(map(re.escape, MATH_FUNCTIONS.keys()))
1192 INTERNAL_FORMAT_RE = re.compile(rf'''(?xs)
1193 (?P<negate>-)?
1194 (?P<fields>{FIELD_RE})
1195 (?P<maths>(?:{MATH_OPERATORS_RE}{MATH_FIELD_RE})*)
1196 (?:>(?P<strf_format>.+?))?
1197 (?P<remaining>
1198 (?P<alternate>(?<!\\),[^|&)]+)?
1199 (?:&(?P<replacement>.*?))?
1200 (?:\|(?P<default>.*?))?
1201 )$''')
1202
1203 def _traverse_infodict(fields):
1204 fields = [f for x in re.split(r'\.({.+?})\.?', fields)
1205 for f in ([x] if x.startswith('{') else x.split('.'))]
1206 for i in (0, -1):
1207 if fields and not fields[i]:
1208 fields.pop(i)
1209
1210 for i, f in enumerate(fields):
1211 if not f.startswith('{'):
1212 continue
1213 assert f.endswith('}'), f'No closing brace for {f} in {fields}'
1214 fields[i] = {k: k.split('.') for k in f[1:-1].split(',')}
1215
1216 return traverse_obj(info_dict, fields, is_user_input=True, traverse_string=True)
1217
1218 def get_value(mdict):
1219 # Object traversal
1220 value = _traverse_infodict(mdict['fields'])
1221 # Negative
1222 if mdict['negate']:
1223 value = float_or_none(value)
1224 if value is not None:
1225 value *= -1
1226 # Do maths
1227 offset_key = mdict['maths']
1228 if offset_key:
1229 value = float_or_none(value)
1230 operator = None
1231 while offset_key:
1232 item = re.match(
1233 MATH_FIELD_RE if operator else MATH_OPERATORS_RE,
1234 offset_key).group(0)
1235 offset_key = offset_key[len(item):]
1236 if operator is None:
1237 operator = MATH_FUNCTIONS[item]
1238 continue
1239 item, multiplier = (item[1:], -1) if item[0] == '-' else (item, 1)
1240 offset = float_or_none(item)
1241 if offset is None:
1242 offset = float_or_none(_traverse_infodict(item))
1243 try:
1244 value = operator(value, multiplier * offset)
1245 except (TypeError, ZeroDivisionError):
1246 return None
1247 operator = None
1248 # Datetime formatting
1249 if mdict['strf_format']:
1250 value = strftime_or_none(value, mdict['strf_format'].replace('\\,', ','))
1251
1252 # XXX: Workaround for https://github.com/yt-dlp/yt-dlp/issues/4485
1253 if sanitize and value == '':
1254 value = None
1255 return value
1256
1257 na = self.params.get('outtmpl_na_placeholder', 'NA')
1258
1259 def filename_sanitizer(key, value, restricted=self.params.get('restrictfilenames')):
1260 return sanitize_filename(str(value), restricted=restricted, is_id=(
1261 bool(re.search(r'(^|[_.])id(\.|$)', key))
1262 if 'filename-sanitization' in self.params['compat_opts']
1263 else NO_DEFAULT))
1264
1265 sanitizer = sanitize if callable(sanitize) else filename_sanitizer
1266 sanitize = bool(sanitize)
1267
1268 def _dumpjson_default(obj):
1269 if isinstance(obj, (set, LazyList)):
1270 return list(obj)
1271 return repr(obj)
1272
1273 class _ReplacementFormatter(string.Formatter):
1274 def get_field(self, field_name, args, kwargs):
1275 if field_name.isdigit():
1276 return args[0], -1
1277 raise ValueError('Unsupported field')
1278
1279 replacement_formatter = _ReplacementFormatter()
1280
1281 def create_key(outer_mobj):
1282 if not outer_mobj.group('has_key'):
1283 return outer_mobj.group(0)
1284 key = outer_mobj.group('key')
1285 mobj = re.match(INTERNAL_FORMAT_RE, key)
1286 value, replacement, default, last_field = None, None, na, ''
1287 while mobj:
1288 mobj = mobj.groupdict()
1289 default = mobj['default'] if mobj['default'] is not None else default
1290 value = get_value(mobj)
1291 last_field, replacement = mobj['fields'], mobj['replacement']
1292 if value is None and mobj['alternate']:
1293 mobj = re.match(INTERNAL_FORMAT_RE, mobj['remaining'][1:])
1294 else:
1295 break
1296
1297 if None not in (value, replacement):
1298 try:
1299 value = replacement_formatter.format(replacement, value)
1300 except ValueError:
1301 value, default = None, na
1302
1303 fmt = outer_mobj.group('format')
1304 if fmt == 's' and last_field in field_size_compat_map.keys() and isinstance(value, int):
1305 fmt = f'0{field_size_compat_map[last_field]:d}d'
1306
1307 flags = outer_mobj.group('conversion') or ''
1308 str_fmt = f'{fmt[:-1]}s'
1309 if value is None:
1310 value, fmt = default, 's'
1311 elif fmt[-1] == 'l': # list
1312 delim = '\n' if '#' in flags else ', '
1313 value, fmt = delim.join(map(str, variadic(value, allowed_types=(str, bytes)))), str_fmt
1314 elif fmt[-1] == 'j': # json
1315 value, fmt = json.dumps(
1316 value, default=_dumpjson_default,
1317 indent=4 if '#' in flags else None, ensure_ascii='+' not in flags), str_fmt
1318 elif fmt[-1] == 'h': # html
1319 value, fmt = escapeHTML(str(value)), str_fmt
1320 elif fmt[-1] == 'q': # quoted
1321 value = map(str, variadic(value) if '#' in flags else [value])
1322 value, fmt = ' '.join(map(compat_shlex_quote, value)), str_fmt
1323 elif fmt[-1] == 'B': # bytes
1324 value = f'%{str_fmt}'.encode() % str(value).encode()
1325 value, fmt = value.decode('utf-8', 'ignore'), 's'
1326 elif fmt[-1] == 'U': # unicode normalized
1327 value, fmt = unicodedata.normalize(
1328 # "+" = compatibility equivalence, "#" = NFD
1329 'NF%s%s' % ('K' if '+' in flags else '', 'D' if '#' in flags else 'C'),
1330 value), str_fmt
1331 elif fmt[-1] == 'D': # decimal suffix
1332 num_fmt, fmt = fmt[:-1].replace('#', ''), 's'
1333 value = format_decimal_suffix(value, f'%{num_fmt}f%s' if num_fmt else '%d%s',
1334 factor=1024 if '#' in flags else 1000)
1335 elif fmt[-1] == 'S': # filename sanitization
1336 value, fmt = filename_sanitizer(last_field, value, restricted='#' in flags), str_fmt
1337 elif fmt[-1] == 'c':
1338 if value:
1339 value = str(value)[0]
1340 else:
1341 fmt = str_fmt
1342 elif fmt[-1] not in 'rsa': # numeric
1343 value = float_or_none(value)
1344 if value is None:
1345 value, fmt = default, 's'
1346
1347 if sanitize:
1348 # If value is an object, sanitize might convert it to a string
1349 # So we convert it to repr first
1350 if fmt[-1] == 'r':
1351 value, fmt = repr(value), str_fmt
1352 elif fmt[-1] == 'a':
1353 value, fmt = ascii(value), str_fmt
1354 if fmt[-1] in 'csra':
1355 value = sanitizer(last_field, value)
1356
1357 key = '%s\0%s' % (key.replace('%', '%\0'), outer_mobj.group('format'))
1358 TMPL_DICT[key] = value
1359 return '{prefix}%({key}){fmt}'.format(key=key, fmt=fmt, prefix=outer_mobj.group('prefix'))
1360
1361 return EXTERNAL_FORMAT_RE.sub(create_key, outtmpl), TMPL_DICT
1362
1363 def evaluate_outtmpl(self, outtmpl, info_dict, *args, **kwargs):
1364 outtmpl, info_dict = self.prepare_outtmpl(outtmpl, info_dict, *args, **kwargs)
1365 return self.escape_outtmpl(outtmpl) % info_dict
1366
1367 def _prepare_filename(self, info_dict, *, outtmpl=None, tmpl_type=None):
1368 assert None in (outtmpl, tmpl_type), 'outtmpl and tmpl_type are mutually exclusive'
1369 if outtmpl is None:
1370 outtmpl = self.params['outtmpl'].get(tmpl_type or 'default', self.params['outtmpl']['default'])
1371 try:
1372 outtmpl = self._outtmpl_expandpath(outtmpl)
1373 filename = self.evaluate_outtmpl(outtmpl, info_dict, True)
1374 if not filename:
1375 return None
1376
1377 if tmpl_type in ('', 'temp'):
1378 final_ext, ext = self.params.get('final_ext'), info_dict.get('ext')
1379 if final_ext and ext and final_ext != ext and filename.endswith(f'.{final_ext}'):
1380 filename = replace_extension(filename, ext, final_ext)
1381 elif tmpl_type:
1382 force_ext = OUTTMPL_TYPES[tmpl_type]
1383 if force_ext:
1384 filename = replace_extension(filename, force_ext, info_dict.get('ext'))
1385
1386 # https://github.com/blackjack4494/youtube-dlc/issues/85
1387 trim_file_name = self.params.get('trim_file_name', False)
1388 if trim_file_name:
1389 no_ext, *ext = filename.rsplit('.', 2)
1390 filename = join_nonempty(no_ext[:trim_file_name], *ext, delim='.')
1391
1392 return filename
1393 except ValueError as err:
1394 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
1395 return None
1396
1397 def prepare_filename(self, info_dict, dir_type='', *, outtmpl=None, warn=False):
1398 """Generate the output filename"""
1399 if outtmpl:
1400 assert not dir_type, 'outtmpl and dir_type are mutually exclusive'
1401 dir_type = None
1402 filename = self._prepare_filename(info_dict, tmpl_type=dir_type, outtmpl=outtmpl)
1403 if not filename and dir_type not in ('', 'temp'):
1404 return ''
1405
1406 if warn:
1407 if not self.params.get('paths'):
1408 pass
1409 elif filename == '-':
1410 self.report_warning('--paths is ignored when an outputting to stdout', only_once=True)
1411 elif os.path.isabs(filename):
1412 self.report_warning('--paths is ignored since an absolute path is given in output template', only_once=True)
1413 if filename == '-' or not filename:
1414 return filename
1415
1416 return self.get_output_path(dir_type, filename)
1417
1418 def _match_entry(self, info_dict, incomplete=False, silent=False):
1419 """Returns None if the file should be downloaded"""
1420 _type = 'video' if 'playlist-match-filter' in self.params['compat_opts'] else info_dict.get('_type', 'video')
1421 assert incomplete or _type == 'video', 'Only video result can be considered complete'
1422
1423 video_title = info_dict.get('title', info_dict.get('id', 'entry'))
1424
1425 def check_filter():
1426 if _type in ('playlist', 'multi_video'):
1427 return
1428 elif _type in ('url', 'url_transparent') and not try_call(
1429 lambda: self.get_info_extractor(info_dict['ie_key']).is_single_video(info_dict['url'])):
1430 return
1431
1432 if 'title' in info_dict:
1433 # This can happen when we're just evaluating the playlist
1434 title = info_dict['title']
1435 matchtitle = self.params.get('matchtitle', False)
1436 if matchtitle:
1437 if not re.search(matchtitle, title, re.IGNORECASE):
1438 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
1439 rejecttitle = self.params.get('rejecttitle', False)
1440 if rejecttitle:
1441 if re.search(rejecttitle, title, re.IGNORECASE):
1442 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
1443
1444 date = info_dict.get('upload_date')
1445 if date is not None:
1446 dateRange = self.params.get('daterange', DateRange())
1447 if date not in dateRange:
1448 return f'{date_from_str(date).isoformat()} upload date is not in range {dateRange}'
1449 view_count = info_dict.get('view_count')
1450 if view_count is not None:
1451 min_views = self.params.get('min_views')
1452 if min_views is not None and view_count < min_views:
1453 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
1454 max_views = self.params.get('max_views')
1455 if max_views is not None and view_count > max_views:
1456 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
1457 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
1458 return 'Skipping "%s" because it is age restricted' % video_title
1459
1460 match_filter = self.params.get('match_filter')
1461 if match_filter is None:
1462 return None
1463
1464 cancelled = None
1465 try:
1466 try:
1467 ret = match_filter(info_dict, incomplete=incomplete)
1468 except TypeError:
1469 # For backward compatibility
1470 ret = None if incomplete else match_filter(info_dict)
1471 except DownloadCancelled as err:
1472 if err.msg is not NO_DEFAULT:
1473 raise
1474 ret, cancelled = err.msg, err
1475
1476 if ret is NO_DEFAULT:
1477 while True:
1478 filename = self._format_screen(self.prepare_filename(info_dict), self.Styles.FILENAME)
1479 reply = input(self._format_screen(
1480 f'Download "{filename}"? (Y/n): ', self.Styles.EMPHASIS)).lower().strip()
1481 if reply in {'y', ''}:
1482 return None
1483 elif reply == 'n':
1484 if cancelled:
1485 raise type(cancelled)(f'Skipping {video_title}')
1486 return f'Skipping {video_title}'
1487 return ret
1488
1489 if self.in_download_archive(info_dict):
1490 reason = ''.join((
1491 format_field(info_dict, 'id', f'{self._format_screen("%s", self.Styles.ID)}: '),
1492 format_field(info_dict, 'title', f'{self._format_screen("%s", self.Styles.EMPHASIS)} '),
1493 'has already been recorded in the archive'))
1494 break_opt, break_err = 'break_on_existing', ExistingVideoReached
1495 else:
1496 try:
1497 reason = check_filter()
1498 except DownloadCancelled as e:
1499 reason, break_opt, break_err = e.msg, 'match_filter', type(e)
1500 else:
1501 break_opt, break_err = 'break_on_reject', RejectedVideoReached
1502 if reason is not None:
1503 if not silent:
1504 self.to_screen('[download] ' + reason)
1505 if self.params.get(break_opt, False):
1506 raise break_err()
1507 return reason
1508
1509 @staticmethod
1510 def add_extra_info(info_dict, extra_info):
1511 '''Set the keys from extra_info in info dict if they are missing'''
1512 for key, value in extra_info.items():
1513 info_dict.setdefault(key, value)
1514
1515 def extract_info(self, url, download=True, ie_key=None, extra_info=None,
1516 process=True, force_generic_extractor=False):
1517 """
1518 Extract and return the information dictionary of the URL
1519
1520 Arguments:
1521 @param url URL to extract
1522
1523 Keyword arguments:
1524 @param download Whether to download videos
1525 @param process Whether to resolve all unresolved references (URLs, playlist items).
1526 Must be True for download to work
1527 @param ie_key Use only the extractor with this key
1528
1529 @param extra_info Dictionary containing the extra values to add to the info (For internal use only)
1530 @force_generic_extractor Force using the generic extractor (Deprecated; use ie_key='Generic')
1531 """
1532
1533 if extra_info is None:
1534 extra_info = {}
1535
1536 if not ie_key and force_generic_extractor:
1537 ie_key = 'Generic'
1538
1539 if ie_key:
1540 ies = {ie_key: self._ies[ie_key]} if ie_key in self._ies else {}
1541 else:
1542 ies = self._ies
1543
1544 for key, ie in ies.items():
1545 if not ie.suitable(url):
1546 continue
1547
1548 if not ie.working():
1549 self.report_warning('The program functionality for this site has been marked as broken, '
1550 'and will probably not work.')
1551
1552 temp_id = ie.get_temp_id(url)
1553 if temp_id is not None and self.in_download_archive({'id': temp_id, 'ie_key': key}):
1554 self.to_screen(f'[download] {self._format_screen(temp_id, self.Styles.ID)}: '
1555 'has already been recorded in the archive')
1556 if self.params.get('break_on_existing', False):
1557 raise ExistingVideoReached()
1558 break
1559 return self.__extract_info(url, self.get_info_extractor(key), download, extra_info, process)
1560 else:
1561 extractors_restricted = self.params.get('allowed_extractors') not in (None, ['default'])
1562 self.report_error(f'No suitable extractor{format_field(ie_key, None, " (%s)")} found for URL {url}',
1563 tb=False if extractors_restricted else None)
1564
1565 def _handle_extraction_exceptions(func):
1566 @functools.wraps(func)
1567 def wrapper(self, *args, **kwargs):
1568 while True:
1569 try:
1570 return func(self, *args, **kwargs)
1571 except (DownloadCancelled, LazyList.IndexError, PagedList.IndexError):
1572 raise
1573 except ReExtractInfo as e:
1574 if e.expected:
1575 self.to_screen(f'{e}; Re-extracting data')
1576 else:
1577 self.to_stderr('\r')
1578 self.report_warning(f'{e}; Re-extracting data')
1579 continue
1580 except GeoRestrictedError as e:
1581 msg = e.msg
1582 if e.countries:
1583 msg += '\nThis video is available in %s.' % ', '.join(
1584 map(ISO3166Utils.short2full, e.countries))
1585 msg += '\nYou might want to use a VPN or a proxy server (with --proxy) to workaround.'
1586 self.report_error(msg)
1587 except ExtractorError as e: # An error we somewhat expected
1588 self.report_error(str(e), e.format_traceback())
1589 except Exception as e:
1590 if self.params.get('ignoreerrors'):
1591 self.report_error(str(e), tb=encode_compat_str(traceback.format_exc()))
1592 else:
1593 raise
1594 break
1595 return wrapper
1596
1597 def _wait_for_video(self, ie_result={}):
1598 if (not self.params.get('wait_for_video')
1599 or ie_result.get('_type', 'video') != 'video'
1600 or ie_result.get('formats') or ie_result.get('url')):
1601 return
1602
1603 format_dur = lambda dur: '%02d:%02d:%02d' % timetuple_from_msec(dur * 1000)[:-1]
1604 last_msg = ''
1605
1606 def progress(msg):
1607 nonlocal last_msg
1608 full_msg = f'{msg}\n'
1609 if not self.params.get('noprogress'):
1610 full_msg = msg + ' ' * (len(last_msg) - len(msg)) + '\r'
1611 elif last_msg:
1612 return
1613 self.to_screen(full_msg, skip_eol=True)
1614 last_msg = msg
1615
1616 min_wait, max_wait = self.params.get('wait_for_video')
1617 diff = try_get(ie_result, lambda x: x['release_timestamp'] - time.time())
1618 if diff is None and ie_result.get('live_status') == 'is_upcoming':
1619 diff = round(random.uniform(min_wait, max_wait) if (max_wait and min_wait) else (max_wait or min_wait), 0)
1620 self.report_warning('Release time of video is not known')
1621 elif ie_result and (diff or 0) <= 0:
1622 self.report_warning('Video should already be available according to extracted info')
1623 diff = min(max(diff or 0, min_wait or 0), max_wait or float('inf'))
1624 self.to_screen(f'[wait] Waiting for {format_dur(diff)} - Press Ctrl+C to try now')
1625
1626 wait_till = time.time() + diff
1627 try:
1628 while True:
1629 diff = wait_till - time.time()
1630 if diff <= 0:
1631 progress('')
1632 raise ReExtractInfo('[wait] Wait period ended', expected=True)
1633 progress(f'[wait] Remaining time until next attempt: {self._format_screen(format_dur(diff), self.Styles.EMPHASIS)}')
1634 time.sleep(1)
1635 except KeyboardInterrupt:
1636 progress('')
1637 raise ReExtractInfo('[wait] Interrupted by user', expected=True)
1638 except BaseException as e:
1639 if not isinstance(e, ReExtractInfo):
1640 self.to_screen('')
1641 raise
1642
1643 def _load_cookies(self, data, *, autoscope=True):
1644 """Loads cookies from a `Cookie` header
1645
1646 This tries to work around the security vulnerability of passing cookies to every domain.
1647 See: https://github.com/yt-dlp/yt-dlp/security/advisories/GHSA-v8mc-9377-rwjj
1648
1649 @param data The Cookie header as string to load the cookies from
1650 @param autoscope If `False`, scope cookies using Set-Cookie syntax and error for cookie without domains
1651 If `True`, save cookies for later to be stored in the jar with a limited scope
1652 If a URL, save cookies in the jar with the domain of the URL
1653 """
1654 for cookie in LenientSimpleCookie(data).values():
1655 if autoscope and any(cookie.values()):
1656 raise ValueError('Invalid syntax in Cookie Header')
1657
1658 domain = cookie.get('domain') or ''
1659 expiry = cookie.get('expires')
1660 if expiry == '': # 0 is valid
1661 expiry = None
1662 prepared_cookie = http.cookiejar.Cookie(
1663 cookie.get('version') or 0, cookie.key, cookie.value, None, False,
1664 domain, True, True, cookie.get('path') or '', bool(cookie.get('path')),
1665 cookie.get('secure') or False, expiry, False, None, None, {})
1666
1667 if domain:
1668 self.cookiejar.set_cookie(prepared_cookie)
1669 elif autoscope is True:
1670 self.deprecated_feature(
1671 'Passing cookies as a header is a potential security risk; '
1672 'they will be scoped to the domain of the downloaded urls. '
1673 'Please consider loading cookies from a file or browser instead.')
1674 self.__header_cookies.append(prepared_cookie)
1675 elif autoscope:
1676 self.report_warning(
1677 'The extractor result contains an unscoped cookie as an HTTP header. '
1678 f'If you are using yt-dlp with an input URL{bug_reports_message(before=",")}',
1679 only_once=True)
1680 self._apply_header_cookies(autoscope, [prepared_cookie])
1681 else:
1682 self.report_error('Unscoped cookies are not allowed; please specify some sort of scoping',
1683 tb=False, is_error=False)
1684
1685 def _apply_header_cookies(self, url, cookies=None):
1686 """Applies stray header cookies to the provided url
1687
1688 This loads header cookies and scopes them to the domain provided in `url`.
1689 While this is not ideal, it helps reduce the risk of them being sent
1690 to an unintended destination while mostly maintaining compatibility.
1691 """
1692 parsed = urllib.parse.urlparse(url)
1693 if not parsed.hostname:
1694 return
1695
1696 for cookie in map(copy.copy, cookies or self.__header_cookies):
1697 cookie.domain = f'.{parsed.hostname}'
1698 self.cookiejar.set_cookie(cookie)
1699
1700 @_handle_extraction_exceptions
1701 def __extract_info(self, url, ie, download, extra_info, process):
1702 self._apply_header_cookies(url)
1703
1704 try:
1705 ie_result = ie.extract(url)
1706 except UserNotLive as e:
1707 if process:
1708 if self.params.get('wait_for_video'):
1709 self.report_warning(e)
1710 self._wait_for_video()
1711 raise
1712 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
1713 self.report_warning(f'Extractor {ie.IE_NAME} returned nothing{bug_reports_message()}')
1714 return
1715 if isinstance(ie_result, list):
1716 # Backwards compatibility: old IE result format
1717 ie_result = {
1718 '_type': 'compat_list',
1719 'entries': ie_result,
1720 }
1721 if extra_info.get('original_url'):
1722 ie_result.setdefault('original_url', extra_info['original_url'])
1723 self.add_default_extra_info(ie_result, ie, url)
1724 if process:
1725 self._wait_for_video(ie_result)
1726 return self.process_ie_result(ie_result, download, extra_info)
1727 else:
1728 return ie_result
1729
1730 def add_default_extra_info(self, ie_result, ie, url):
1731 if url is not None:
1732 self.add_extra_info(ie_result, {
1733 'webpage_url': url,
1734 'original_url': url,
1735 })
1736 webpage_url = ie_result.get('webpage_url')
1737 if webpage_url:
1738 self.add_extra_info(ie_result, {
1739 'webpage_url_basename': url_basename(webpage_url),
1740 'webpage_url_domain': get_domain(webpage_url),
1741 })
1742 if ie is not None:
1743 self.add_extra_info(ie_result, {
1744 'extractor': ie.IE_NAME,
1745 'extractor_key': ie.ie_key(),
1746 })
1747
1748 def process_ie_result(self, ie_result, download=True, extra_info=None):
1749 """
1750 Take the result of the ie(may be modified) and resolve all unresolved
1751 references (URLs, playlist items).
1752
1753 It will also download the videos if 'download'.
1754 Returns the resolved ie_result.
1755 """
1756 if extra_info is None:
1757 extra_info = {}
1758 result_type = ie_result.get('_type', 'video')
1759
1760 if result_type in ('url', 'url_transparent'):
1761 ie_result['url'] = sanitize_url(
1762 ie_result['url'], scheme='http' if self.params.get('prefer_insecure') else 'https')
1763 if ie_result.get('original_url') and not extra_info.get('original_url'):
1764 extra_info = {'original_url': ie_result['original_url'], **extra_info}
1765
1766 extract_flat = self.params.get('extract_flat', False)
1767 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info)
1768 or extract_flat is True):
1769 info_copy = ie_result.copy()
1770 ie = try_get(ie_result.get('ie_key'), self.get_info_extractor)
1771 if ie and not ie_result.get('id'):
1772 info_copy['id'] = ie.get_temp_id(ie_result['url'])
1773 self.add_default_extra_info(info_copy, ie, ie_result['url'])
1774 self.add_extra_info(info_copy, extra_info)
1775 info_copy, _ = self.pre_process(info_copy)
1776 self._fill_common_fields(info_copy, False)
1777 self.__forced_printings(info_copy)
1778 self._raise_pending_errors(info_copy)
1779 if self.params.get('force_write_download_archive', False):
1780 self.record_download_archive(info_copy)
1781 return ie_result
1782
1783 if result_type == 'video':
1784 self.add_extra_info(ie_result, extra_info)
1785 ie_result = self.process_video_result(ie_result, download=download)
1786 self._raise_pending_errors(ie_result)
1787 additional_urls = (ie_result or {}).get('additional_urls')
1788 if additional_urls:
1789 # TODO: Improve MetadataParserPP to allow setting a list
1790 if isinstance(additional_urls, str):
1791 additional_urls = [additional_urls]
1792 self.to_screen(
1793 '[info] %s: %d additional URL(s) requested' % (ie_result['id'], len(additional_urls)))
1794 self.write_debug('Additional URLs: "%s"' % '", "'.join(additional_urls))
1795 ie_result['additional_entries'] = [
1796 self.extract_info(
1797 url, download, extra_info=extra_info,
1798 force_generic_extractor=self.params.get('force_generic_extractor'))
1799 for url in additional_urls
1800 ]
1801 return ie_result
1802 elif result_type == 'url':
1803 # We have to add extra_info to the results because it may be
1804 # contained in a playlist
1805 return self.extract_info(
1806 ie_result['url'], download,
1807 ie_key=ie_result.get('ie_key'),
1808 extra_info=extra_info)
1809 elif result_type == 'url_transparent':
1810 # Use the information from the embedding page
1811 info = self.extract_info(
1812 ie_result['url'], ie_key=ie_result.get('ie_key'),
1813 extra_info=extra_info, download=False, process=False)
1814
1815 # extract_info may return None when ignoreerrors is enabled and
1816 # extraction failed with an error, don't crash and return early
1817 # in this case
1818 if not info:
1819 return info
1820
1821 exempted_fields = {'_type', 'url', 'ie_key'}
1822 if not ie_result.get('section_end') and ie_result.get('section_start') is None:
1823 # For video clips, the id etc of the clip extractor should be used
1824 exempted_fields |= {'id', 'extractor', 'extractor_key'}
1825
1826 new_result = info.copy()
1827 new_result.update(filter_dict(ie_result, lambda k, v: v is not None and k not in exempted_fields))
1828
1829 # Extracted info may not be a video result (i.e.
1830 # info.get('_type', 'video') != video) but rather an url or
1831 # url_transparent. In such cases outer metadata (from ie_result)
1832 # should be propagated to inner one (info). For this to happen
1833 # _type of info should be overridden with url_transparent. This
1834 # fixes issue from https://github.com/ytdl-org/youtube-dl/pull/11163.
1835 if new_result.get('_type') == 'url':
1836 new_result['_type'] = 'url_transparent'
1837
1838 return self.process_ie_result(
1839 new_result, download=download, extra_info=extra_info)
1840 elif result_type in ('playlist', 'multi_video'):
1841 # Protect from infinite recursion due to recursively nested playlists
1842 # (see https://github.com/ytdl-org/youtube-dl/issues/27833)
1843 webpage_url = ie_result.get('webpage_url') # Playlists maynot have webpage_url
1844 if webpage_url and webpage_url in self._playlist_urls:
1845 self.to_screen(
1846 '[download] Skipping already downloaded playlist: %s'
1847 % ie_result.get('title') or ie_result.get('id'))
1848 return
1849
1850 self._playlist_level += 1
1851 self._playlist_urls.add(webpage_url)
1852 self._fill_common_fields(ie_result, False)
1853 self._sanitize_thumbnails(ie_result)
1854 try:
1855 return self.__process_playlist(ie_result, download)
1856 finally:
1857 self._playlist_level -= 1
1858 if not self._playlist_level:
1859 self._playlist_urls.clear()
1860 elif result_type == 'compat_list':
1861 self.report_warning(
1862 'Extractor %s returned a compat_list result. '
1863 'It needs to be updated.' % ie_result.get('extractor'))
1864
1865 def _fixup(r):
1866 self.add_extra_info(r, {
1867 'extractor': ie_result['extractor'],
1868 'webpage_url': ie_result['webpage_url'],
1869 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1870 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1871 'extractor_key': ie_result['extractor_key'],
1872 })
1873 return r
1874 ie_result['entries'] = [
1875 self.process_ie_result(_fixup(r), download, extra_info)
1876 for r in ie_result['entries']
1877 ]
1878 return ie_result
1879 else:
1880 raise Exception('Invalid result type: %s' % result_type)
1881
1882 def _ensure_dir_exists(self, path):
1883 return make_dir(path, self.report_error)
1884
1885 @staticmethod
1886 def _playlist_infodict(ie_result, strict=False, **kwargs):
1887 info = {
1888 'playlist_count': ie_result.get('playlist_count'),
1889 'playlist': ie_result.get('title') or ie_result.get('id'),
1890 'playlist_id': ie_result.get('id'),
1891 'playlist_title': ie_result.get('title'),
1892 'playlist_uploader': ie_result.get('uploader'),
1893 'playlist_uploader_id': ie_result.get('uploader_id'),
1894 **kwargs,
1895 }
1896 if strict:
1897 return info
1898 if ie_result.get('webpage_url'):
1899 info.update({
1900 'webpage_url': ie_result['webpage_url'],
1901 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1902 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1903 })
1904 return {
1905 **info,
1906 'playlist_index': 0,
1907 '__last_playlist_index': max(ie_result.get('requested_entries') or (0, 0)),
1908 'extractor': ie_result['extractor'],
1909 'extractor_key': ie_result['extractor_key'],
1910 }
1911
1912 def __process_playlist(self, ie_result, download):
1913 """Process each entry in the playlist"""
1914 assert ie_result['_type'] in ('playlist', 'multi_video')
1915
1916 common_info = self._playlist_infodict(ie_result, strict=True)
1917 title = common_info.get('playlist') or '<Untitled>'
1918 if self._match_entry(common_info, incomplete=True) is not None:
1919 return
1920 self.to_screen(f'[download] Downloading {ie_result["_type"]}: {title}')
1921
1922 all_entries = PlaylistEntries(self, ie_result)
1923 entries = orderedSet(all_entries.get_requested_items(), lazy=True)
1924
1925 lazy = self.params.get('lazy_playlist')
1926 if lazy:
1927 resolved_entries, n_entries = [], 'N/A'
1928 ie_result['requested_entries'], ie_result['entries'] = None, None
1929 else:
1930 entries = resolved_entries = list(entries)
1931 n_entries = len(resolved_entries)
1932 ie_result['requested_entries'], ie_result['entries'] = tuple(zip(*resolved_entries)) or ([], [])
1933 if not ie_result.get('playlist_count'):
1934 # Better to do this after potentially exhausting entries
1935 ie_result['playlist_count'] = all_entries.get_full_count()
1936
1937 extra = self._playlist_infodict(ie_result, n_entries=int_or_none(n_entries))
1938 ie_copy = collections.ChainMap(ie_result, extra)
1939
1940 _infojson_written = False
1941 write_playlist_files = self.params.get('allow_playlist_files', True)
1942 if write_playlist_files and self.params.get('list_thumbnails'):
1943 self.list_thumbnails(ie_result)
1944 if write_playlist_files and not self.params.get('simulate'):
1945 _infojson_written = self._write_info_json(
1946 'playlist', ie_result, self.prepare_filename(ie_copy, 'pl_infojson'))
1947 if _infojson_written is None:
1948 return
1949 if self._write_description('playlist', ie_result,
1950 self.prepare_filename(ie_copy, 'pl_description')) is None:
1951 return
1952 # TODO: This should be passed to ThumbnailsConvertor if necessary
1953 self._write_thumbnails('playlist', ie_result, self.prepare_filename(ie_copy, 'pl_thumbnail'))
1954
1955 if lazy:
1956 if self.params.get('playlistreverse') or self.params.get('playlistrandom'):
1957 self.report_warning('playlistreverse and playlistrandom are not supported with lazy_playlist', only_once=True)
1958 elif self.params.get('playlistreverse'):
1959 entries.reverse()
1960 elif self.params.get('playlistrandom'):
1961 random.shuffle(entries)
1962
1963 self.to_screen(f'[{ie_result["extractor"]}] Playlist {title}: Downloading {n_entries} items'
1964 f'{format_field(ie_result, "playlist_count", " of %s")}')
1965
1966 keep_resolved_entries = self.params.get('extract_flat') != 'discard'
1967 if self.params.get('extract_flat') == 'discard_in_playlist':
1968 keep_resolved_entries = ie_result['_type'] != 'playlist'
1969 if keep_resolved_entries:
1970 self.write_debug('The information of all playlist entries will be held in memory')
1971
1972 failures = 0
1973 max_failures = self.params.get('skip_playlist_after_errors') or float('inf')
1974 for i, (playlist_index, entry) in enumerate(entries):
1975 if lazy:
1976 resolved_entries.append((playlist_index, entry))
1977 if not entry:
1978 continue
1979
1980 entry['__x_forwarded_for_ip'] = ie_result.get('__x_forwarded_for_ip')
1981 if not lazy and 'playlist-index' in self.params['compat_opts']:
1982 playlist_index = ie_result['requested_entries'][i]
1983
1984 entry_copy = collections.ChainMap(entry, {
1985 **common_info,
1986 'n_entries': int_or_none(n_entries),
1987 'playlist_index': playlist_index,
1988 'playlist_autonumber': i + 1,
1989 })
1990
1991 if self._match_entry(entry_copy, incomplete=True) is not None:
1992 # For compatabilty with youtube-dl. See https://github.com/yt-dlp/yt-dlp/issues/4369
1993 resolved_entries[i] = (playlist_index, NO_DEFAULT)
1994 continue
1995
1996 self.to_screen('[download] Downloading item %s of %s' % (
1997 self._format_screen(i + 1, self.Styles.ID), self._format_screen(n_entries, self.Styles.EMPHASIS)))
1998
1999 entry_result = self.__process_iterable_entry(entry, download, collections.ChainMap({
2000 'playlist_index': playlist_index,
2001 'playlist_autonumber': i + 1,
2002 }, extra))
2003 if not entry_result:
2004 failures += 1
2005 if failures >= max_failures:
2006 self.report_error(
2007 f'Skipping the remaining entries in playlist "{title}" since {failures} items failed extraction')
2008 break
2009 if keep_resolved_entries:
2010 resolved_entries[i] = (playlist_index, entry_result)
2011
2012 # Update with processed data
2013 ie_result['entries'] = [e for _, e in resolved_entries if e is not NO_DEFAULT]
2014 ie_result['requested_entries'] = [i for i, e in resolved_entries if e is not NO_DEFAULT]
2015 if ie_result['requested_entries'] == try_call(lambda: list(range(1, ie_result['playlist_count'] + 1))):
2016 # Do not set for full playlist
2017 ie_result.pop('requested_entries')
2018
2019 # Write the updated info to json
2020 if _infojson_written is True and self._write_info_json(
2021 'updated playlist', ie_result,
2022 self.prepare_filename(ie_copy, 'pl_infojson'), overwrite=True) is None:
2023 return
2024
2025 ie_result = self.run_all_pps('playlist', ie_result)
2026 self.to_screen(f'[download] Finished downloading playlist: {title}')
2027 return ie_result
2028
2029 @_handle_extraction_exceptions
2030 def __process_iterable_entry(self, entry, download, extra_info):
2031 return self.process_ie_result(
2032 entry, download=download, extra_info=extra_info)
2033
2034 def _build_format_filter(self, filter_spec):
2035 " Returns a function to filter the formats according to the filter_spec "
2036
2037 OPERATORS = {
2038 '<': operator.lt,
2039 '<=': operator.le,
2040 '>': operator.gt,
2041 '>=': operator.ge,
2042 '=': operator.eq,
2043 '!=': operator.ne,
2044 }
2045 operator_rex = re.compile(r'''(?x)\s*
2046 (?P<key>[\w.-]+)\s*
2047 (?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
2048 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)\s*
2049 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
2050 m = operator_rex.fullmatch(filter_spec)
2051 if m:
2052 try:
2053 comparison_value = int(m.group('value'))
2054 except ValueError:
2055 comparison_value = parse_filesize(m.group('value'))
2056 if comparison_value is None:
2057 comparison_value = parse_filesize(m.group('value') + 'B')
2058 if comparison_value is None:
2059 raise ValueError(
2060 'Invalid value %r in format specification %r' % (
2061 m.group('value'), filter_spec))
2062 op = OPERATORS[m.group('op')]
2063
2064 if not m:
2065 STR_OPERATORS = {
2066 '=': operator.eq,
2067 '^=': lambda attr, value: attr.startswith(value),
2068 '$=': lambda attr, value: attr.endswith(value),
2069 '*=': lambda attr, value: value in attr,
2070 '~=': lambda attr, value: value.search(attr) is not None
2071 }
2072 str_operator_rex = re.compile(r'''(?x)\s*
2073 (?P<key>[a-zA-Z0-9._-]+)\s*
2074 (?P<negation>!\s*)?(?P<op>%s)\s*(?P<none_inclusive>\?\s*)?
2075 (?P<quote>["'])?
2076 (?P<value>(?(quote)(?:(?!(?P=quote))[^\\]|\\.)+|[\w.-]+))
2077 (?(quote)(?P=quote))\s*
2078 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
2079 m = str_operator_rex.fullmatch(filter_spec)
2080 if m:
2081 if m.group('op') == '~=':
2082 comparison_value = re.compile(m.group('value'))
2083 else:
2084 comparison_value = re.sub(r'''\\([\\"'])''', r'\1', m.group('value'))
2085 str_op = STR_OPERATORS[m.group('op')]
2086 if m.group('negation'):
2087 op = lambda attr, value: not str_op(attr, value)
2088 else:
2089 op = str_op
2090
2091 if not m:
2092 raise SyntaxError('Invalid filter specification %r' % filter_spec)
2093
2094 def _filter(f):
2095 actual_value = f.get(m.group('key'))
2096 if actual_value is None:
2097 return m.group('none_inclusive')
2098 return op(actual_value, comparison_value)
2099 return _filter
2100
2101 def _check_formats(self, formats):
2102 for f in formats:
2103 self.to_screen('[info] Testing format %s' % f['format_id'])
2104 path = self.get_output_path('temp')
2105 if not self._ensure_dir_exists(f'{path}/'):
2106 continue
2107 temp_file = tempfile.NamedTemporaryFile(suffix='.tmp', delete=False, dir=path or None)
2108 temp_file.close()
2109 try:
2110 success, _ = self.dl(temp_file.name, f, test=True)
2111 except (DownloadError, OSError, ValueError) + network_exceptions:
2112 success = False
2113 finally:
2114 if os.path.exists(temp_file.name):
2115 try:
2116 os.remove(temp_file.name)
2117 except OSError:
2118 self.report_warning('Unable to delete temporary file "%s"' % temp_file.name)
2119 if success:
2120 yield f
2121 else:
2122 self.to_screen('[info] Unable to download format %s. Skipping...' % f['format_id'])
2123
2124 def _default_format_spec(self, info_dict, download=True):
2125
2126 def can_merge():
2127 merger = FFmpegMergerPP(self)
2128 return merger.available and merger.can_merge()
2129
2130 prefer_best = (
2131 not self.params.get('simulate')
2132 and download
2133 and (
2134 not can_merge()
2135 or info_dict.get('is_live') and not self.params.get('live_from_start')
2136 or self.params['outtmpl']['default'] == '-'))
2137 compat = (
2138 prefer_best
2139 or self.params.get('allow_multiple_audio_streams', False)
2140 or 'format-spec' in self.params['compat_opts'])
2141
2142 return (
2143 'best/bestvideo+bestaudio' if prefer_best
2144 else 'bestvideo*+bestaudio/best' if not compat
2145 else 'bestvideo+bestaudio/best')
2146
2147 def build_format_selector(self, format_spec):
2148 def syntax_error(note, start):
2149 message = (
2150 'Invalid format specification: '
2151 '{}\n\t{}\n\t{}^'.format(note, format_spec, ' ' * start[1]))
2152 return SyntaxError(message)
2153
2154 PICKFIRST = 'PICKFIRST'
2155 MERGE = 'MERGE'
2156 SINGLE = 'SINGLE'
2157 GROUP = 'GROUP'
2158 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
2159
2160 allow_multiple_streams = {'audio': self.params.get('allow_multiple_audio_streams', False),
2161 'video': self.params.get('allow_multiple_video_streams', False)}
2162
2163 def _parse_filter(tokens):
2164 filter_parts = []
2165 for type, string_, start, _, _ in tokens:
2166 if type == tokenize.OP and string_ == ']':
2167 return ''.join(filter_parts)
2168 else:
2169 filter_parts.append(string_)
2170
2171 def _remove_unused_ops(tokens):
2172 # Remove operators that we don't use and join them with the surrounding strings.
2173 # E.g. 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
2174 ALLOWED_OPS = ('/', '+', ',', '(', ')')
2175 last_string, last_start, last_end, last_line = None, None, None, None
2176 for type, string_, start, end, line in tokens:
2177 if type == tokenize.OP and string_ == '[':
2178 if last_string:
2179 yield tokenize.NAME, last_string, last_start, last_end, last_line
2180 last_string = None
2181 yield type, string_, start, end, line
2182 # everything inside brackets will be handled by _parse_filter
2183 for type, string_, start, end, line in tokens:
2184 yield type, string_, start, end, line
2185 if type == tokenize.OP and string_ == ']':
2186 break
2187 elif type == tokenize.OP and string_ in ALLOWED_OPS:
2188 if last_string:
2189 yield tokenize.NAME, last_string, last_start, last_end, last_line
2190 last_string = None
2191 yield type, string_, start, end, line
2192 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
2193 if not last_string:
2194 last_string = string_
2195 last_start = start
2196 last_end = end
2197 else:
2198 last_string += string_
2199 if last_string:
2200 yield tokenize.NAME, last_string, last_start, last_end, last_line
2201
2202 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
2203 selectors = []
2204 current_selector = None
2205 for type, string_, start, _, _ in tokens:
2206 # ENCODING is only defined in python 3.x
2207 if type == getattr(tokenize, 'ENCODING', None):
2208 continue
2209 elif type in [tokenize.NAME, tokenize.NUMBER]:
2210 current_selector = FormatSelector(SINGLE, string_, [])
2211 elif type == tokenize.OP:
2212 if string_ == ')':
2213 if not inside_group:
2214 # ')' will be handled by the parentheses group
2215 tokens.restore_last_token()
2216 break
2217 elif inside_merge and string_ in ['/', ',']:
2218 tokens.restore_last_token()
2219 break
2220 elif inside_choice and string_ == ',':
2221 tokens.restore_last_token()
2222 break
2223 elif string_ == ',':
2224 if not current_selector:
2225 raise syntax_error('"," must follow a format selector', start)
2226 selectors.append(current_selector)
2227 current_selector = None
2228 elif string_ == '/':
2229 if not current_selector:
2230 raise syntax_error('"/" must follow a format selector', start)
2231 first_choice = current_selector
2232 second_choice = _parse_format_selection(tokens, inside_choice=True)
2233 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
2234 elif string_ == '[':
2235 if not current_selector:
2236 current_selector = FormatSelector(SINGLE, 'best', [])
2237 format_filter = _parse_filter(tokens)
2238 current_selector.filters.append(format_filter)
2239 elif string_ == '(':
2240 if current_selector:
2241 raise syntax_error('Unexpected "("', start)
2242 group = _parse_format_selection(tokens, inside_group=True)
2243 current_selector = FormatSelector(GROUP, group, [])
2244 elif string_ == '+':
2245 if not current_selector:
2246 raise syntax_error('Unexpected "+"', start)
2247 selector_1 = current_selector
2248 selector_2 = _parse_format_selection(tokens, inside_merge=True)
2249 if not selector_2:
2250 raise syntax_error('Expected a selector', start)
2251 current_selector = FormatSelector(MERGE, (selector_1, selector_2), [])
2252 else:
2253 raise syntax_error(f'Operator not recognized: "{string_}"', start)
2254 elif type == tokenize.ENDMARKER:
2255 break
2256 if current_selector:
2257 selectors.append(current_selector)
2258 return selectors
2259
2260 def _merge(formats_pair):
2261 format_1, format_2 = formats_pair
2262
2263 formats_info = []
2264 formats_info.extend(format_1.get('requested_formats', (format_1,)))
2265 formats_info.extend(format_2.get('requested_formats', (format_2,)))
2266
2267 if not allow_multiple_streams['video'] or not allow_multiple_streams['audio']:
2268 get_no_more = {'video': False, 'audio': False}
2269 for (i, fmt_info) in enumerate(formats_info):
2270 if fmt_info.get('acodec') == fmt_info.get('vcodec') == 'none':
2271 formats_info.pop(i)
2272 continue
2273 for aud_vid in ['audio', 'video']:
2274 if not allow_multiple_streams[aud_vid] and fmt_info.get(aud_vid[0] + 'codec') != 'none':
2275 if get_no_more[aud_vid]:
2276 formats_info.pop(i)
2277 break
2278 get_no_more[aud_vid] = True
2279
2280 if len(formats_info) == 1:
2281 return formats_info[0]
2282
2283 video_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('vcodec') != 'none']
2284 audio_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('acodec') != 'none']
2285
2286 the_only_video = video_fmts[0] if len(video_fmts) == 1 else None
2287 the_only_audio = audio_fmts[0] if len(audio_fmts) == 1 else None
2288
2289 output_ext = get_compatible_ext(
2290 vcodecs=[f.get('vcodec') for f in video_fmts],
2291 acodecs=[f.get('acodec') for f in audio_fmts],
2292 vexts=[f['ext'] for f in video_fmts],
2293 aexts=[f['ext'] for f in audio_fmts],
2294 preferences=(try_call(lambda: self.params['merge_output_format'].split('/'))
2295 or self.params.get('prefer_free_formats') and ('webm', 'mkv')))
2296
2297 filtered = lambda *keys: filter(None, (traverse_obj(fmt, *keys) for fmt in formats_info))
2298
2299 new_dict = {
2300 'requested_formats': formats_info,
2301 'format': '+'.join(filtered('format')),
2302 'format_id': '+'.join(filtered('format_id')),
2303 'ext': output_ext,
2304 'protocol': '+'.join(map(determine_protocol, formats_info)),
2305 'language': '+'.join(orderedSet(filtered('language'))) or None,
2306 'format_note': '+'.join(orderedSet(filtered('format_note'))) or None,
2307 'filesize_approx': sum(filtered('filesize', 'filesize_approx')) or None,
2308 'tbr': sum(filtered('tbr', 'vbr', 'abr')),
2309 }
2310
2311 if the_only_video:
2312 new_dict.update({
2313 'width': the_only_video.get('width'),
2314 'height': the_only_video.get('height'),
2315 'resolution': the_only_video.get('resolution') or self.format_resolution(the_only_video),
2316 'fps': the_only_video.get('fps'),
2317 'dynamic_range': the_only_video.get('dynamic_range'),
2318 'vcodec': the_only_video.get('vcodec'),
2319 'vbr': the_only_video.get('vbr'),
2320 'stretched_ratio': the_only_video.get('stretched_ratio'),
2321 'aspect_ratio': the_only_video.get('aspect_ratio'),
2322 })
2323
2324 if the_only_audio:
2325 new_dict.update({
2326 'acodec': the_only_audio.get('acodec'),
2327 'abr': the_only_audio.get('abr'),
2328 'asr': the_only_audio.get('asr'),
2329 'audio_channels': the_only_audio.get('audio_channels')
2330 })
2331
2332 return new_dict
2333
2334 def _check_formats(formats):
2335 if self.params.get('check_formats') == 'selected':
2336 yield from self._check_formats(formats)
2337 return
2338 elif (self.params.get('check_formats') is not None
2339 or self.params.get('allow_unplayable_formats')):
2340 yield from formats
2341 return
2342
2343 for f in formats:
2344 if f.get('has_drm') or f.get('__needs_testing'):
2345 yield from self._check_formats([f])
2346 else:
2347 yield f
2348
2349 def _build_selector_function(selector):
2350 if isinstance(selector, list): # ,
2351 fs = [_build_selector_function(s) for s in selector]
2352
2353 def selector_function(ctx):
2354 for f in fs:
2355 yield from f(ctx)
2356 return selector_function
2357
2358 elif selector.type == GROUP: # ()
2359 selector_function = _build_selector_function(selector.selector)
2360
2361 elif selector.type == PICKFIRST: # /
2362 fs = [_build_selector_function(s) for s in selector.selector]
2363
2364 def selector_function(ctx):
2365 for f in fs:
2366 picked_formats = list(f(ctx))
2367 if picked_formats:
2368 return picked_formats
2369 return []
2370
2371 elif selector.type == MERGE: # +
2372 selector_1, selector_2 = map(_build_selector_function, selector.selector)
2373
2374 def selector_function(ctx):
2375 for pair in itertools.product(selector_1(ctx), selector_2(ctx)):
2376 yield _merge(pair)
2377
2378 elif selector.type == SINGLE: # atom
2379 format_spec = selector.selector or 'best'
2380
2381 # TODO: Add allvideo, allaudio etc by generalizing the code with best/worst selector
2382 if format_spec == 'all':
2383 def selector_function(ctx):
2384 yield from _check_formats(ctx['formats'][::-1])
2385 elif format_spec == 'mergeall':
2386 def selector_function(ctx):
2387 formats = list(_check_formats(
2388 f for f in ctx['formats'] if f.get('vcodec') != 'none' or f.get('acodec') != 'none'))
2389 if not formats:
2390 return
2391 merged_format = formats[-1]
2392 for f in formats[-2::-1]:
2393 merged_format = _merge((merged_format, f))
2394 yield merged_format
2395
2396 else:
2397 format_fallback, seperate_fallback, format_reverse, format_idx = False, None, True, 1
2398 mobj = re.match(
2399 r'(?P<bw>best|worst|b|w)(?P<type>video|audio|v|a)?(?P<mod>\*)?(?:\.(?P<n>[1-9]\d*))?$',
2400 format_spec)
2401 if mobj is not None:
2402 format_idx = int_or_none(mobj.group('n'), default=1)
2403 format_reverse = mobj.group('bw')[0] == 'b'
2404 format_type = (mobj.group('type') or [None])[0]
2405 not_format_type = {'v': 'a', 'a': 'v'}.get(format_type)
2406 format_modified = mobj.group('mod') is not None
2407
2408 format_fallback = not format_type and not format_modified # for b, w
2409 _filter_f = (
2410 (lambda f: f.get('%scodec' % format_type) != 'none')
2411 if format_type and format_modified # bv*, ba*, wv*, wa*
2412 else (lambda f: f.get('%scodec' % not_format_type) == 'none')
2413 if format_type # bv, ba, wv, wa
2414 else (lambda f: f.get('vcodec') != 'none' and f.get('acodec') != 'none')
2415 if not format_modified # b, w
2416 else lambda f: True) # b*, w*
2417 filter_f = lambda f: _filter_f(f) and (
2418 f.get('vcodec') != 'none' or f.get('acodec') != 'none')
2419 else:
2420 if format_spec in self._format_selection_exts['audio']:
2421 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none'
2422 elif format_spec in self._format_selection_exts['video']:
2423 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none' and f.get('vcodec') != 'none'
2424 seperate_fallback = lambda f: f.get('ext') == format_spec and f.get('vcodec') != 'none'
2425 elif format_spec in self._format_selection_exts['storyboards']:
2426 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') == 'none' and f.get('vcodec') == 'none'
2427 else:
2428 filter_f = lambda f: f.get('format_id') == format_spec # id
2429
2430 def selector_function(ctx):
2431 formats = list(ctx['formats'])
2432 matches = list(filter(filter_f, formats)) if filter_f is not None else formats
2433 if not matches:
2434 if format_fallback and ctx['incomplete_formats']:
2435 # for extractors with incomplete formats (audio only (soundcloud)
2436 # or video only (imgur)) best/worst will fallback to
2437 # best/worst {video,audio}-only format
2438 matches = formats
2439 elif seperate_fallback and not ctx['has_merged_format']:
2440 # for compatibility with youtube-dl when there is no pre-merged format
2441 matches = list(filter(seperate_fallback, formats))
2442 matches = LazyList(_check_formats(matches[::-1 if format_reverse else 1]))
2443 try:
2444 yield matches[format_idx - 1]
2445 except LazyList.IndexError:
2446 return
2447
2448 filters = [self._build_format_filter(f) for f in selector.filters]
2449
2450 def final_selector(ctx):
2451 ctx_copy = dict(ctx)
2452 for _filter in filters:
2453 ctx_copy['formats'] = list(filter(_filter, ctx_copy['formats']))
2454 return selector_function(ctx_copy)
2455 return final_selector
2456
2457 stream = io.BytesIO(format_spec.encode())
2458 try:
2459 tokens = list(_remove_unused_ops(tokenize.tokenize(stream.readline)))
2460 except tokenize.TokenError:
2461 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
2462
2463 class TokenIterator:
2464 def __init__(self, tokens):
2465 self.tokens = tokens
2466 self.counter = 0
2467
2468 def __iter__(self):
2469 return self
2470
2471 def __next__(self):
2472 if self.counter >= len(self.tokens):
2473 raise StopIteration()
2474 value = self.tokens[self.counter]
2475 self.counter += 1
2476 return value
2477
2478 next = __next__
2479
2480 def restore_last_token(self):
2481 self.counter -= 1
2482
2483 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
2484 return _build_selector_function(parsed_selector)
2485
2486 def _calc_headers(self, info_dict, load_cookies=False):
2487 res = HTTPHeaderDict(self.params['http_headers'], info_dict.get('http_headers'))
2488 clean_headers(res)
2489
2490 if load_cookies: # For --load-info-json
2491 self._load_cookies(res.get('Cookie'), autoscope=info_dict['url']) # compat
2492 self._load_cookies(info_dict.get('cookies'), autoscope=False)
2493 # The `Cookie` header is removed to prevent leaks and unscoped cookies.
2494 # See: https://github.com/yt-dlp/yt-dlp/security/advisories/GHSA-v8mc-9377-rwjj
2495 res.pop('Cookie', None)
2496 cookies = self.cookiejar.get_cookies_for_url(info_dict['url'])
2497 if cookies:
2498 encoder = LenientSimpleCookie()
2499 values = []
2500 for cookie in cookies:
2501 _, value = encoder.value_encode(cookie.value)
2502 values.append(f'{cookie.name}={value}')
2503 if cookie.domain:
2504 values.append(f'Domain={cookie.domain}')
2505 if cookie.path:
2506 values.append(f'Path={cookie.path}')
2507 if cookie.secure:
2508 values.append('Secure')
2509 if cookie.expires:
2510 values.append(f'Expires={cookie.expires}')
2511 if cookie.version:
2512 values.append(f'Version={cookie.version}')
2513 info_dict['cookies'] = '; '.join(values)
2514
2515 if 'X-Forwarded-For' not in res:
2516 x_forwarded_for_ip = info_dict.get('__x_forwarded_for_ip')
2517 if x_forwarded_for_ip:
2518 res['X-Forwarded-For'] = x_forwarded_for_ip
2519
2520 return res
2521
2522 def _calc_cookies(self, url):
2523 self.deprecation_warning('"YoutubeDL._calc_cookies" is deprecated and may be removed in a future version')
2524 return self.cookiejar.get_cookie_header(url)
2525
2526 def _sort_thumbnails(self, thumbnails):
2527 thumbnails.sort(key=lambda t: (
2528 t.get('preference') if t.get('preference') is not None else -1,
2529 t.get('width') if t.get('width') is not None else -1,
2530 t.get('height') if t.get('height') is not None else -1,
2531 t.get('id') if t.get('id') is not None else '',
2532 t.get('url')))
2533
2534 def _sanitize_thumbnails(self, info_dict):
2535 thumbnails = info_dict.get('thumbnails')
2536 if thumbnails is None:
2537 thumbnail = info_dict.get('thumbnail')
2538 if thumbnail:
2539 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
2540 if not thumbnails:
2541 return
2542
2543 def check_thumbnails(thumbnails):
2544 for t in thumbnails:
2545 self.to_screen(f'[info] Testing thumbnail {t["id"]}')
2546 try:
2547 self.urlopen(HEADRequest(t['url']))
2548 except network_exceptions as err:
2549 self.to_screen(f'[info] Unable to connect to thumbnail {t["id"]} URL {t["url"]!r} - {err}. Skipping...')
2550 continue
2551 yield t
2552
2553 self._sort_thumbnails(thumbnails)
2554 for i, t in enumerate(thumbnails):
2555 if t.get('id') is None:
2556 t['id'] = '%d' % i
2557 if t.get('width') and t.get('height'):
2558 t['resolution'] = '%dx%d' % (t['width'], t['height'])
2559 t['url'] = sanitize_url(t['url'])
2560
2561 if self.params.get('check_formats') is True:
2562 info_dict['thumbnails'] = LazyList(check_thumbnails(thumbnails[::-1]), reverse=True)
2563 else:
2564 info_dict['thumbnails'] = thumbnails
2565
2566 def _fill_common_fields(self, info_dict, final=True):
2567 # TODO: move sanitization here
2568 if final:
2569 title = info_dict['fulltitle'] = info_dict.get('title')
2570 if not title:
2571 if title == '':
2572 self.write_debug('Extractor gave empty title. Creating a generic title')
2573 else:
2574 self.report_warning('Extractor failed to obtain "title". Creating a generic title instead')
2575 info_dict['title'] = f'{info_dict["extractor"].replace(":", "-")} video #{info_dict["id"]}'
2576
2577 if info_dict.get('duration') is not None:
2578 info_dict['duration_string'] = formatSeconds(info_dict['duration'])
2579
2580 for ts_key, date_key in (
2581 ('timestamp', 'upload_date'),
2582 ('release_timestamp', 'release_date'),
2583 ('modified_timestamp', 'modified_date'),
2584 ):
2585 if info_dict.get(date_key) is None and info_dict.get(ts_key) is not None:
2586 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
2587 # see http://bugs.python.org/issue1646728)
2588 with contextlib.suppress(ValueError, OverflowError, OSError):
2589 upload_date = datetime.datetime.fromtimestamp(info_dict[ts_key], datetime.timezone.utc)
2590 info_dict[date_key] = upload_date.strftime('%Y%m%d')
2591
2592 if not info_dict.get('release_year'):
2593 info_dict['release_year'] = traverse_obj(info_dict, ('release_date', {lambda x: int(x[:4])}))
2594
2595 live_keys = ('is_live', 'was_live')
2596 live_status = info_dict.get('live_status')
2597 if live_status is None:
2598 for key in live_keys:
2599 if info_dict.get(key) is False:
2600 continue
2601 if info_dict.get(key):
2602 live_status = key
2603 break
2604 if all(info_dict.get(key) is False for key in live_keys):
2605 live_status = 'not_live'
2606 if live_status:
2607 info_dict['live_status'] = live_status
2608 for key in live_keys:
2609 if info_dict.get(key) is None:
2610 info_dict[key] = (live_status == key)
2611 if live_status == 'post_live':
2612 info_dict['was_live'] = True
2613
2614 # Auto generate title fields corresponding to the *_number fields when missing
2615 # in order to always have clean titles. This is very common for TV series.
2616 for field in ('chapter', 'season', 'episode'):
2617 if final and info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
2618 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
2619
2620 def _raise_pending_errors(self, info):
2621 err = info.pop('__pending_error', None)
2622 if err:
2623 self.report_error(err, tb=False)
2624
2625 def sort_formats(self, info_dict):
2626 formats = self._get_formats(info_dict)
2627 formats.sort(key=FormatSorter(
2628 self, info_dict.get('_format_sort_fields') or []).calculate_preference)
2629
2630 def process_video_result(self, info_dict, download=True):
2631 assert info_dict.get('_type', 'video') == 'video'
2632 self._num_videos += 1
2633
2634 if 'id' not in info_dict:
2635 raise ExtractorError('Missing "id" field in extractor result', ie=info_dict['extractor'])
2636 elif not info_dict.get('id'):
2637 raise ExtractorError('Extractor failed to obtain "id"', ie=info_dict['extractor'])
2638
2639 def report_force_conversion(field, field_not, conversion):
2640 self.report_warning(
2641 '"%s" field is not %s - forcing %s conversion, there is an error in extractor'
2642 % (field, field_not, conversion))
2643
2644 def sanitize_string_field(info, string_field):
2645 field = info.get(string_field)
2646 if field is None or isinstance(field, str):
2647 return
2648 report_force_conversion(string_field, 'a string', 'string')
2649 info[string_field] = str(field)
2650
2651 def sanitize_numeric_fields(info):
2652 for numeric_field in self._NUMERIC_FIELDS:
2653 field = info.get(numeric_field)
2654 if field is None or isinstance(field, (int, float)):
2655 continue
2656 report_force_conversion(numeric_field, 'numeric', 'int')
2657 info[numeric_field] = int_or_none(field)
2658
2659 sanitize_string_field(info_dict, 'id')
2660 sanitize_numeric_fields(info_dict)
2661 if info_dict.get('section_end') and info_dict.get('section_start') is not None:
2662 info_dict['duration'] = round(info_dict['section_end'] - info_dict['section_start'], 3)
2663 if (info_dict.get('duration') or 0) <= 0 and info_dict.pop('duration', None):
2664 self.report_warning('"duration" field is negative, there is an error in extractor')
2665
2666 chapters = info_dict.get('chapters') or []
2667 if chapters and chapters[0].get('start_time'):
2668 chapters.insert(0, {'start_time': 0})
2669
2670 dummy_chapter = {'end_time': 0, 'start_time': info_dict.get('duration')}
2671 for idx, (prev, current, next_) in enumerate(zip(
2672 (dummy_chapter, *chapters), chapters, (*chapters[1:], dummy_chapter)), 1):
2673 if current.get('start_time') is None:
2674 current['start_time'] = prev.get('end_time')
2675 if not current.get('end_time'):
2676 current['end_time'] = next_.get('start_time')
2677 if not current.get('title'):
2678 current['title'] = f'<Untitled Chapter {idx}>'
2679
2680 if 'playlist' not in info_dict:
2681 # It isn't part of a playlist
2682 info_dict['playlist'] = None
2683 info_dict['playlist_index'] = None
2684
2685 self._sanitize_thumbnails(info_dict)
2686
2687 thumbnail = info_dict.get('thumbnail')
2688 thumbnails = info_dict.get('thumbnails')
2689 if thumbnail:
2690 info_dict['thumbnail'] = sanitize_url(thumbnail)
2691 elif thumbnails:
2692 info_dict['thumbnail'] = thumbnails[-1]['url']
2693
2694 if info_dict.get('display_id') is None and 'id' in info_dict:
2695 info_dict['display_id'] = info_dict['id']
2696
2697 self._fill_common_fields(info_dict)
2698
2699 for cc_kind in ('subtitles', 'automatic_captions'):
2700 cc = info_dict.get(cc_kind)
2701 if cc:
2702 for _, subtitle in cc.items():
2703 for subtitle_format in subtitle:
2704 if subtitle_format.get('url'):
2705 subtitle_format['url'] = sanitize_url(subtitle_format['url'])
2706 if subtitle_format.get('ext') is None:
2707 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
2708
2709 automatic_captions = info_dict.get('automatic_captions')
2710 subtitles = info_dict.get('subtitles')
2711
2712 info_dict['requested_subtitles'] = self.process_subtitles(
2713 info_dict['id'], subtitles, automatic_captions)
2714
2715 formats = self._get_formats(info_dict)
2716
2717 # Backward compatibility with InfoExtractor._sort_formats
2718 field_preference = (formats or [{}])[0].pop('__sort_fields', None)
2719 if field_preference:
2720 info_dict['_format_sort_fields'] = field_preference
2721
2722 info_dict['_has_drm'] = any( # or None ensures --clean-infojson removes it
2723 f.get('has_drm') and f['has_drm'] != 'maybe' for f in formats) or None
2724 if not self.params.get('allow_unplayable_formats'):
2725 formats = [f for f in formats if not f.get('has_drm') or f['has_drm'] == 'maybe']
2726
2727 if formats and all(f.get('acodec') == f.get('vcodec') == 'none' for f in formats):
2728 self.report_warning(
2729 f'{"This video is DRM protected and " if info_dict["_has_drm"] else ""}'
2730 'only images are available for download. Use --list-formats to see them'.capitalize())
2731
2732 get_from_start = not info_dict.get('is_live') or bool(self.params.get('live_from_start'))
2733 if not get_from_start:
2734 info_dict['title'] += ' ' + datetime.datetime.now().strftime('%Y-%m-%d %H:%M')
2735 if info_dict.get('is_live') and formats:
2736 formats = [f for f in formats if bool(f.get('is_from_start')) == get_from_start]
2737 if get_from_start and not formats:
2738 self.raise_no_formats(info_dict, msg=(
2739 '--live-from-start is passed, but there are no formats that can be downloaded from the start. '
2740 'If you want to download from the current time, use --no-live-from-start'))
2741
2742 def is_wellformed(f):
2743 url = f.get('url')
2744 if not url:
2745 self.report_warning(
2746 '"url" field is missing or empty - skipping format, '
2747 'there is an error in extractor')
2748 return False
2749 if isinstance(url, bytes):
2750 sanitize_string_field(f, 'url')
2751 return True
2752
2753 # Filter out malformed formats for better extraction robustness
2754 formats = list(filter(is_wellformed, formats or []))
2755
2756 if not formats:
2757 self.raise_no_formats(info_dict)
2758
2759 for format in formats:
2760 sanitize_string_field(format, 'format_id')
2761 sanitize_numeric_fields(format)
2762 format['url'] = sanitize_url(format['url'])
2763 if format.get('ext') is None:
2764 format['ext'] = determine_ext(format['url']).lower()
2765 if format.get('protocol') is None:
2766 format['protocol'] = determine_protocol(format)
2767 if format.get('resolution') is None:
2768 format['resolution'] = self.format_resolution(format, default=None)
2769 if format.get('dynamic_range') is None and format.get('vcodec') != 'none':
2770 format['dynamic_range'] = 'SDR'
2771 if format.get('aspect_ratio') is None:
2772 format['aspect_ratio'] = try_call(lambda: round(format['width'] / format['height'], 2))
2773 # For fragmented formats, "tbr" is often max bitrate and not average
2774 if (('manifest-filesize-approx' in self.params['compat_opts'] or not format.get('manifest_url'))
2775 and info_dict.get('duration') and format.get('tbr')
2776 and not format.get('filesize') and not format.get('filesize_approx')):
2777 format['filesize_approx'] = int(info_dict['duration'] * format['tbr'] * (1024 / 8))
2778 format['http_headers'] = self._calc_headers(collections.ChainMap(format, info_dict), load_cookies=True)
2779
2780 # Safeguard against old/insecure infojson when using --load-info-json
2781 if info_dict.get('http_headers'):
2782 info_dict['http_headers'] = HTTPHeaderDict(info_dict['http_headers'])
2783 info_dict['http_headers'].pop('Cookie', None)
2784
2785 # This is copied to http_headers by the above _calc_headers and can now be removed
2786 if '__x_forwarded_for_ip' in info_dict:
2787 del info_dict['__x_forwarded_for_ip']
2788
2789 self.sort_formats({
2790 'formats': formats,
2791 '_format_sort_fields': info_dict.get('_format_sort_fields')
2792 })
2793
2794 # Sanitize and group by format_id
2795 formats_dict = {}
2796 for i, format in enumerate(formats):
2797 if not format.get('format_id'):
2798 format['format_id'] = str(i)
2799 else:
2800 # Sanitize format_id from characters used in format selector expression
2801 format['format_id'] = re.sub(r'[\s,/+\[\]()]', '_', format['format_id'])
2802 formats_dict.setdefault(format['format_id'], []).append(format)
2803
2804 # Make sure all formats have unique format_id
2805 common_exts = set(itertools.chain(*self._format_selection_exts.values()))
2806 for format_id, ambiguous_formats in formats_dict.items():
2807 ambigious_id = len(ambiguous_formats) > 1
2808 for i, format in enumerate(ambiguous_formats):
2809 if ambigious_id:
2810 format['format_id'] = '%s-%d' % (format_id, i)
2811 # Ensure there is no conflict between id and ext in format selection
2812 # See https://github.com/yt-dlp/yt-dlp/issues/1282
2813 if format['format_id'] != format['ext'] and format['format_id'] in common_exts:
2814 format['format_id'] = 'f%s' % format['format_id']
2815
2816 if format.get('format') is None:
2817 format['format'] = '{id} - {res}{note}'.format(
2818 id=format['format_id'],
2819 res=self.format_resolution(format),
2820 note=format_field(format, 'format_note', ' (%s)'),
2821 )
2822
2823 if self.params.get('check_formats') is True:
2824 formats = LazyList(self._check_formats(formats[::-1]), reverse=True)
2825
2826 if not formats or formats[0] is not info_dict:
2827 # only set the 'formats' fields if the original info_dict list them
2828 # otherwise we end up with a circular reference, the first (and unique)
2829 # element in the 'formats' field in info_dict is info_dict itself,
2830 # which can't be exported to json
2831 info_dict['formats'] = formats
2832
2833 info_dict, _ = self.pre_process(info_dict)
2834
2835 if self._match_entry(info_dict, incomplete=self._format_fields) is not None:
2836 return info_dict
2837
2838 self.post_extract(info_dict)
2839 info_dict, _ = self.pre_process(info_dict, 'after_filter')
2840
2841 # The pre-processors may have modified the formats
2842 formats = self._get_formats(info_dict)
2843
2844 list_only = self.params.get('simulate') == 'list_only'
2845 interactive_format_selection = not list_only and self.format_selector == '-'
2846 if self.params.get('list_thumbnails'):
2847 self.list_thumbnails(info_dict)
2848 if self.params.get('listsubtitles'):
2849 if 'automatic_captions' in info_dict:
2850 self.list_subtitles(
2851 info_dict['id'], automatic_captions, 'automatic captions')
2852 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
2853 if self.params.get('listformats') or interactive_format_selection:
2854 self.list_formats(info_dict)
2855 if list_only:
2856 # Without this printing, -F --print-json will not work
2857 self.__forced_printings(info_dict)
2858 return info_dict
2859
2860 format_selector = self.format_selector
2861 while True:
2862 if interactive_format_selection:
2863 req_format = input(self._format_screen('\nEnter format selector ', self.Styles.EMPHASIS)
2864 + '(Press ENTER for default, or Ctrl+C to quit)'
2865 + self._format_screen(': ', self.Styles.EMPHASIS))
2866 try:
2867 format_selector = self.build_format_selector(req_format) if req_format else None
2868 except SyntaxError as err:
2869 self.report_error(err, tb=False, is_error=False)
2870 continue
2871
2872 if format_selector is None:
2873 req_format = self._default_format_spec(info_dict, download=download)
2874 self.write_debug(f'Default format spec: {req_format}')
2875 format_selector = self.build_format_selector(req_format)
2876
2877 formats_to_download = list(format_selector({
2878 'formats': formats,
2879 'has_merged_format': any('none' not in (f.get('acodec'), f.get('vcodec')) for f in formats),
2880 'incomplete_formats': (all(f.get('vcodec') == 'none' for f in formats) # No formats with video
2881 or all(f.get('acodec') == 'none' for f in formats)), # OR, No formats with audio
2882 }))
2883 if interactive_format_selection and not formats_to_download:
2884 self.report_error('Requested format is not available', tb=False, is_error=False)
2885 continue
2886 break
2887
2888 if not formats_to_download:
2889 if not self.params.get('ignore_no_formats_error'):
2890 raise ExtractorError(
2891 'Requested format is not available. Use --list-formats for a list of available formats',
2892 expected=True, video_id=info_dict['id'], ie=info_dict['extractor'])
2893 self.report_warning('Requested format is not available')
2894 # Process what we can, even without any available formats.
2895 formats_to_download = [{}]
2896
2897 requested_ranges = tuple(self.params.get('download_ranges', lambda *_: [{}])(info_dict, self))
2898 best_format, downloaded_formats = formats_to_download[-1], []
2899 if download:
2900 if best_format and requested_ranges:
2901 def to_screen(*msg):
2902 self.to_screen(f'[info] {info_dict["id"]}: {" ".join(", ".join(variadic(m)) for m in msg)}')
2903
2904 to_screen(f'Downloading {len(formats_to_download)} format(s):',
2905 (f['format_id'] for f in formats_to_download))
2906 if requested_ranges != ({}, ):
2907 to_screen(f'Downloading {len(requested_ranges)} time ranges:',
2908 (f'{c["start_time"]:.1f}-{c["end_time"]:.1f}' for c in requested_ranges))
2909 max_downloads_reached = False
2910
2911 for fmt, chapter in itertools.product(formats_to_download, requested_ranges):
2912 new_info = self._copy_infodict(info_dict)
2913 new_info.update(fmt)
2914 offset, duration = info_dict.get('section_start') or 0, info_dict.get('duration') or float('inf')
2915 end_time = offset + min(chapter.get('end_time', duration), duration)
2916 # duration may not be accurate. So allow deviations <1sec
2917 if end_time == float('inf') or end_time > offset + duration + 1:
2918 end_time = None
2919 if chapter or offset:
2920 new_info.update({
2921 'section_start': offset + chapter.get('start_time', 0),
2922 'section_end': end_time,
2923 'section_title': chapter.get('title'),
2924 'section_number': chapter.get('index'),
2925 })
2926 downloaded_formats.append(new_info)
2927 try:
2928 self.process_info(new_info)
2929 except MaxDownloadsReached:
2930 max_downloads_reached = True
2931 self._raise_pending_errors(new_info)
2932 # Remove copied info
2933 for key, val in tuple(new_info.items()):
2934 if info_dict.get(key) == val:
2935 new_info.pop(key)
2936 if max_downloads_reached:
2937 break
2938
2939 write_archive = {f.get('__write_download_archive', False) for f in downloaded_formats}
2940 assert write_archive.issubset({True, False, 'ignore'})
2941 if True in write_archive and False not in write_archive:
2942 self.record_download_archive(info_dict)
2943
2944 info_dict['requested_downloads'] = downloaded_formats
2945 info_dict = self.run_all_pps('after_video', info_dict)
2946 if max_downloads_reached:
2947 raise MaxDownloadsReached()
2948
2949 # We update the info dict with the selected best quality format (backwards compatibility)
2950 info_dict.update(best_format)
2951 return info_dict
2952
2953 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
2954 """Select the requested subtitles and their format"""
2955 available_subs, normal_sub_langs = {}, []
2956 if normal_subtitles and self.params.get('writesubtitles'):
2957 available_subs.update(normal_subtitles)
2958 normal_sub_langs = tuple(normal_subtitles.keys())
2959 if automatic_captions and self.params.get('writeautomaticsub'):
2960 for lang, cap_info in automatic_captions.items():
2961 if lang not in available_subs:
2962 available_subs[lang] = cap_info
2963
2964 if not available_subs or (
2965 not self.params.get('writesubtitles')
2966 and not self.params.get('writeautomaticsub')):
2967 return None
2968
2969 all_sub_langs = tuple(available_subs.keys())
2970 if self.params.get('allsubtitles', False):
2971 requested_langs = all_sub_langs
2972 elif self.params.get('subtitleslangs', False):
2973 try:
2974 requested_langs = orderedSet_from_options(
2975 self.params.get('subtitleslangs'), {'all': all_sub_langs}, use_regex=True)
2976 except re.error as e:
2977 raise ValueError(f'Wrong regex for subtitlelangs: {e.pattern}')
2978 else:
2979 requested_langs = LazyList(itertools.chain(
2980 ['en'] if 'en' in normal_sub_langs else [],
2981 filter(lambda f: f.startswith('en'), normal_sub_langs),
2982 ['en'] if 'en' in all_sub_langs else [],
2983 filter(lambda f: f.startswith('en'), all_sub_langs),
2984 normal_sub_langs, all_sub_langs,
2985 ))[:1]
2986 if requested_langs:
2987 self.to_screen(f'[info] {video_id}: Downloading subtitles: {", ".join(requested_langs)}')
2988
2989 formats_query = self.params.get('subtitlesformat', 'best')
2990 formats_preference = formats_query.split('/') if formats_query else []
2991 subs = {}
2992 for lang in requested_langs:
2993 formats = available_subs.get(lang)
2994 if formats is None:
2995 self.report_warning(f'{lang} subtitles not available for {video_id}')
2996 continue
2997 for ext in formats_preference:
2998 if ext == 'best':
2999 f = formats[-1]
3000 break
3001 matches = list(filter(lambda f: f['ext'] == ext, formats))
3002 if matches:
3003 f = matches[-1]
3004 break
3005 else:
3006 f = formats[-1]
3007 self.report_warning(
3008 'No subtitle format found matching "%s" for language %s, '
3009 'using %s' % (formats_query, lang, f['ext']))
3010 subs[lang] = f
3011 return subs
3012
3013 def _forceprint(self, key, info_dict):
3014 if info_dict is None:
3015 return
3016 info_copy = info_dict.copy()
3017 info_copy.setdefault('filename', self.prepare_filename(info_dict))
3018 if info_dict.get('requested_formats') is not None:
3019 # For RTMP URLs, also include the playpath
3020 info_copy['urls'] = '\n'.join(f['url'] + f.get('play_path', '') for f in info_dict['requested_formats'])
3021 elif info_dict.get('url'):
3022 info_copy['urls'] = info_dict['url'] + info_dict.get('play_path', '')
3023 info_copy['formats_table'] = self.render_formats_table(info_dict)
3024 info_copy['thumbnails_table'] = self.render_thumbnails_table(info_dict)
3025 info_copy['subtitles_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('subtitles'))
3026 info_copy['automatic_captions_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('automatic_captions'))
3027
3028 def format_tmpl(tmpl):
3029 mobj = re.fullmatch(r'([\w.:,]|-\d|(?P<dict>{([\w.:,]|-\d)+}))+=?', tmpl)
3030 if not mobj:
3031 return tmpl
3032
3033 fmt = '%({})s'
3034 if tmpl.startswith('{'):
3035 tmpl, fmt = f'.{tmpl}', '%({})j'
3036 if tmpl.endswith('='):
3037 tmpl, fmt = tmpl[:-1], '{0} = %({0})#j'
3038 return '\n'.join(map(fmt.format, [tmpl] if mobj.group('dict') else tmpl.split(',')))
3039
3040 for tmpl in self.params['forceprint'].get(key, []):
3041 self.to_stdout(self.evaluate_outtmpl(format_tmpl(tmpl), info_copy))
3042
3043 for tmpl, file_tmpl in self.params['print_to_file'].get(key, []):
3044 filename = self.prepare_filename(info_dict, outtmpl=file_tmpl)
3045 tmpl = format_tmpl(tmpl)
3046 self.to_screen(f'[info] Writing {tmpl!r} to: {filename}')
3047 if self._ensure_dir_exists(filename):
3048 with open(filename, 'a', encoding='utf-8', newline='') as f:
3049 f.write(self.evaluate_outtmpl(tmpl, info_copy) + os.linesep)
3050
3051 return info_copy
3052
3053 def __forced_printings(self, info_dict, filename=None, incomplete=True):
3054 if (self.params.get('forcejson')
3055 or self.params['forceprint'].get('video')
3056 or self.params['print_to_file'].get('video')):
3057 self.post_extract(info_dict)
3058 if filename:
3059 info_dict['filename'] = filename
3060 info_copy = self._forceprint('video', info_dict)
3061
3062 def print_field(field, actual_field=None, optional=False):
3063 if actual_field is None:
3064 actual_field = field
3065 if self.params.get(f'force{field}') and (
3066 info_copy.get(field) is not None or (not optional and not incomplete)):
3067 self.to_stdout(info_copy[actual_field])
3068
3069 print_field('title')
3070 print_field('id')
3071 print_field('url', 'urls')
3072 print_field('thumbnail', optional=True)
3073 print_field('description', optional=True)
3074 print_field('filename')
3075 if self.params.get('forceduration') and info_copy.get('duration') is not None:
3076 self.to_stdout(formatSeconds(info_copy['duration']))
3077 print_field('format')
3078
3079 if self.params.get('forcejson'):
3080 self.to_stdout(json.dumps(self.sanitize_info(info_dict)))
3081
3082 def dl(self, name, info, subtitle=False, test=False):
3083 if not info.get('url'):
3084 self.raise_no_formats(info, True)
3085
3086 if test:
3087 verbose = self.params.get('verbose')
3088 params = {
3089 'test': True,
3090 'quiet': self.params.get('quiet') or not verbose,
3091 'verbose': verbose,
3092 'noprogress': not verbose,
3093 'nopart': True,
3094 'skip_unavailable_fragments': False,
3095 'keep_fragments': False,
3096 'overwrites': True,
3097 '_no_ytdl_file': True,
3098 }
3099 else:
3100 params = self.params
3101 fd = get_suitable_downloader(info, params, to_stdout=(name == '-'))(self, params)
3102 if not test:
3103 for ph in self._progress_hooks:
3104 fd.add_progress_hook(ph)
3105 urls = '", "'.join(
3106 (f['url'].split(',')[0] + ',<data>' if f['url'].startswith('data:') else f['url'])
3107 for f in info.get('requested_formats', []) or [info])
3108 self.write_debug(f'Invoking {fd.FD_NAME} downloader on "{urls}"')
3109
3110 # Note: Ideally info should be a deep-copied so that hooks cannot modify it.
3111 # But it may contain objects that are not deep-copyable
3112 new_info = self._copy_infodict(info)
3113 if new_info.get('http_headers') is None:
3114 new_info['http_headers'] = self._calc_headers(new_info)
3115 return fd.download(name, new_info, subtitle)
3116
3117 def existing_file(self, filepaths, *, default_overwrite=True):
3118 existing_files = list(filter(os.path.exists, orderedSet(filepaths)))
3119 if existing_files and not self.params.get('overwrites', default_overwrite):
3120 return existing_files[0]
3121
3122 for file in existing_files:
3123 self.report_file_delete(file)
3124 os.remove(file)
3125 return None
3126
3127 def process_info(self, info_dict):
3128 """Process a single resolved IE result. (Modifies it in-place)"""
3129
3130 assert info_dict.get('_type', 'video') == 'video'
3131 original_infodict = info_dict
3132
3133 if 'format' not in info_dict and 'ext' in info_dict:
3134 info_dict['format'] = info_dict['ext']
3135
3136 if self._match_entry(info_dict) is not None:
3137 info_dict['__write_download_archive'] = 'ignore'
3138 return
3139
3140 # Does nothing under normal operation - for backward compatibility of process_info
3141 self.post_extract(info_dict)
3142
3143 def replace_info_dict(new_info):
3144 nonlocal info_dict
3145 if new_info == info_dict:
3146 return
3147 info_dict.clear()
3148 info_dict.update(new_info)
3149
3150 new_info, _ = self.pre_process(info_dict, 'video')
3151 replace_info_dict(new_info)
3152 self._num_downloads += 1
3153
3154 # info_dict['_filename'] needs to be set for backward compatibility
3155 info_dict['_filename'] = full_filename = self.prepare_filename(info_dict, warn=True)
3156 temp_filename = self.prepare_filename(info_dict, 'temp')
3157 files_to_move = {}
3158
3159 # Forced printings
3160 self.__forced_printings(info_dict, full_filename, incomplete=('format' not in info_dict))
3161
3162 def check_max_downloads():
3163 if self._num_downloads >= float(self.params.get('max_downloads') or 'inf'):
3164 raise MaxDownloadsReached()
3165
3166 if self.params.get('simulate'):
3167 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
3168 check_max_downloads()
3169 return
3170
3171 if full_filename is None:
3172 return
3173 if not self._ensure_dir_exists(encodeFilename(full_filename)):
3174 return
3175 if not self._ensure_dir_exists(encodeFilename(temp_filename)):
3176 return
3177
3178 if self._write_description('video', info_dict,
3179 self.prepare_filename(info_dict, 'description')) is None:
3180 return
3181
3182 sub_files = self._write_subtitles(info_dict, temp_filename)
3183 if sub_files is None:
3184 return
3185 files_to_move.update(dict(sub_files))
3186
3187 thumb_files = self._write_thumbnails(
3188 'video', info_dict, temp_filename, self.prepare_filename(info_dict, 'thumbnail'))
3189 if thumb_files is None:
3190 return
3191 files_to_move.update(dict(thumb_files))
3192
3193 infofn = self.prepare_filename(info_dict, 'infojson')
3194 _infojson_written = self._write_info_json('video', info_dict, infofn)
3195 if _infojson_written:
3196 info_dict['infojson_filename'] = infofn
3197 # For backward compatibility, even though it was a private field
3198 info_dict['__infojson_filename'] = infofn
3199 elif _infojson_written is None:
3200 return
3201
3202 # Note: Annotations are deprecated
3203 annofn = None
3204 if self.params.get('writeannotations', False):
3205 annofn = self.prepare_filename(info_dict, 'annotation')
3206 if annofn:
3207 if not self._ensure_dir_exists(encodeFilename(annofn)):
3208 return
3209 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(annofn)):
3210 self.to_screen('[info] Video annotations are already present')
3211 elif not info_dict.get('annotations'):
3212 self.report_warning('There are no annotations to write.')
3213 else:
3214 try:
3215 self.to_screen('[info] Writing video annotations to: ' + annofn)
3216 with open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
3217 annofile.write(info_dict['annotations'])
3218 except (KeyError, TypeError):
3219 self.report_warning('There are no annotations to write.')
3220 except OSError:
3221 self.report_error('Cannot write annotations file: ' + annofn)
3222 return
3223
3224 # Write internet shortcut files
3225 def _write_link_file(link_type):
3226 url = try_get(info_dict['webpage_url'], iri_to_uri)
3227 if not url:
3228 self.report_warning(
3229 f'Cannot write internet shortcut file because the actual URL of "{info_dict["webpage_url"]}" is unknown')
3230 return True
3231 linkfn = replace_extension(self.prepare_filename(info_dict, 'link'), link_type, info_dict.get('ext'))
3232 if not self._ensure_dir_exists(encodeFilename(linkfn)):
3233 return False
3234 if self.params.get('overwrites', True) and os.path.exists(encodeFilename(linkfn)):
3235 self.to_screen(f'[info] Internet shortcut (.{link_type}) is already present')
3236 return True
3237 try:
3238 self.to_screen(f'[info] Writing internet shortcut (.{link_type}) to: {linkfn}')
3239 with open(encodeFilename(to_high_limit_path(linkfn)), 'w', encoding='utf-8',
3240 newline='\r\n' if link_type == 'url' else '\n') as linkfile:
3241 template_vars = {'url': url}
3242 if link_type == 'desktop':
3243 template_vars['filename'] = linkfn[:-(len(link_type) + 1)]
3244 linkfile.write(LINK_TEMPLATES[link_type] % template_vars)
3245 except OSError:
3246 self.report_error(f'Cannot write internet shortcut {linkfn}')
3247 return False
3248 return True
3249
3250 write_links = {
3251 'url': self.params.get('writeurllink'),
3252 'webloc': self.params.get('writewebloclink'),
3253 'desktop': self.params.get('writedesktoplink'),
3254 }
3255 if self.params.get('writelink'):
3256 link_type = ('webloc' if sys.platform == 'darwin'
3257 else 'desktop' if sys.platform.startswith('linux')
3258 else 'url')
3259 write_links[link_type] = True
3260
3261 if any(should_write and not _write_link_file(link_type)
3262 for link_type, should_write in write_links.items()):
3263 return
3264
3265 new_info, files_to_move = self.pre_process(info_dict, 'before_dl', files_to_move)
3266 replace_info_dict(new_info)
3267
3268 if self.params.get('skip_download'):
3269 info_dict['filepath'] = temp_filename
3270 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
3271 info_dict['__files_to_move'] = files_to_move
3272 replace_info_dict(self.run_pp(MoveFilesAfterDownloadPP(self, False), info_dict))
3273 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
3274 else:
3275 # Download
3276 info_dict.setdefault('__postprocessors', [])
3277 try:
3278
3279 def existing_video_file(*filepaths):
3280 ext = info_dict.get('ext')
3281 converted = lambda file: replace_extension(file, self.params.get('final_ext') or ext, ext)
3282 file = self.existing_file(itertools.chain(*zip(map(converted, filepaths), filepaths)),
3283 default_overwrite=False)
3284 if file:
3285 info_dict['ext'] = os.path.splitext(file)[1][1:]
3286 return file
3287
3288 fd, success = None, True
3289 if info_dict.get('protocol') or info_dict.get('url'):
3290 fd = get_suitable_downloader(info_dict, self.params, to_stdout=temp_filename == '-')
3291 if fd != FFmpegFD and 'no-direct-merge' not in self.params['compat_opts'] and (
3292 info_dict.get('section_start') or info_dict.get('section_end')):
3293 msg = ('This format cannot be partially downloaded' if FFmpegFD.available()
3294 else 'You have requested downloading the video partially, but ffmpeg is not installed')
3295 self.report_error(f'{msg}. Aborting')
3296 return
3297
3298 if info_dict.get('requested_formats') is not None:
3299 old_ext = info_dict['ext']
3300 if self.params.get('merge_output_format') is None:
3301 if (info_dict['ext'] == 'webm'
3302 and info_dict.get('thumbnails')
3303 # check with type instead of pp_key, __name__, or isinstance
3304 # since we dont want any custom PPs to trigger this
3305 and any(type(pp) == EmbedThumbnailPP for pp in self._pps['post_process'])): # noqa: E721
3306 info_dict['ext'] = 'mkv'
3307 self.report_warning(
3308 'webm doesn\'t support embedding a thumbnail, mkv will be used')
3309 new_ext = info_dict['ext']
3310
3311 def correct_ext(filename, ext=new_ext):
3312 if filename == '-':
3313 return filename
3314 filename_real_ext = os.path.splitext(filename)[1][1:]
3315 filename_wo_ext = (
3316 os.path.splitext(filename)[0]
3317 if filename_real_ext in (old_ext, new_ext)
3318 else filename)
3319 return f'{filename_wo_ext}.{ext}'
3320
3321 # Ensure filename always has a correct extension for successful merge
3322 full_filename = correct_ext(full_filename)
3323 temp_filename = correct_ext(temp_filename)
3324 dl_filename = existing_video_file(full_filename, temp_filename)
3325
3326 info_dict['__real_download'] = False
3327 # NOTE: Copy so that original format dicts are not modified
3328 info_dict['requested_formats'] = list(map(dict, info_dict['requested_formats']))
3329
3330 merger = FFmpegMergerPP(self)
3331 downloaded = []
3332 if dl_filename is not None:
3333 self.report_file_already_downloaded(dl_filename)
3334 elif fd:
3335 for f in info_dict['requested_formats'] if fd != FFmpegFD else []:
3336 f['filepath'] = fname = prepend_extension(
3337 correct_ext(temp_filename, info_dict['ext']),
3338 'f%s' % f['format_id'], info_dict['ext'])
3339 downloaded.append(fname)
3340 info_dict['url'] = '\n'.join(f['url'] for f in info_dict['requested_formats'])
3341 success, real_download = self.dl(temp_filename, info_dict)
3342 info_dict['__real_download'] = real_download
3343 else:
3344 if self.params.get('allow_unplayable_formats'):
3345 self.report_warning(
3346 'You have requested merging of multiple formats '
3347 'while also allowing unplayable formats to be downloaded. '
3348 'The formats won\'t be merged to prevent data corruption.')
3349 elif not merger.available:
3350 msg = 'You have requested merging of multiple formats but ffmpeg is not installed'
3351 if not self.params.get('ignoreerrors'):
3352 self.report_error(f'{msg}. Aborting due to --abort-on-error')
3353 return
3354 self.report_warning(f'{msg}. The formats won\'t be merged')
3355
3356 if temp_filename == '-':
3357 reason = ('using a downloader other than ffmpeg' if FFmpegFD.can_merge_formats(info_dict, self.params)
3358 else 'but the formats are incompatible for simultaneous download' if merger.available
3359 else 'but ffmpeg is not installed')
3360 self.report_warning(
3361 f'You have requested downloading multiple formats to stdout {reason}. '
3362 'The formats will be streamed one after the other')
3363 fname = temp_filename
3364 for f in info_dict['requested_formats']:
3365 new_info = dict(info_dict)
3366 del new_info['requested_formats']
3367 new_info.update(f)
3368 if temp_filename != '-':
3369 fname = prepend_extension(
3370 correct_ext(temp_filename, new_info['ext']),
3371 'f%s' % f['format_id'], new_info['ext'])
3372 if not self._ensure_dir_exists(fname):
3373 return
3374 f['filepath'] = fname
3375 downloaded.append(fname)
3376 partial_success, real_download = self.dl(fname, new_info)
3377 info_dict['__real_download'] = info_dict['__real_download'] or real_download
3378 success = success and partial_success
3379
3380 if downloaded and merger.available and not self.params.get('allow_unplayable_formats'):
3381 info_dict['__postprocessors'].append(merger)
3382 info_dict['__files_to_merge'] = downloaded
3383 # Even if there were no downloads, it is being merged only now
3384 info_dict['__real_download'] = True
3385 else:
3386 for file in downloaded:
3387 files_to_move[file] = None
3388 else:
3389 # Just a single file
3390 dl_filename = existing_video_file(full_filename, temp_filename)
3391 if dl_filename is None or dl_filename == temp_filename:
3392 # dl_filename == temp_filename could mean that the file was partially downloaded with --no-part.
3393 # So we should try to resume the download
3394 success, real_download = self.dl(temp_filename, info_dict)
3395 info_dict['__real_download'] = real_download
3396 else:
3397 self.report_file_already_downloaded(dl_filename)
3398
3399 dl_filename = dl_filename or temp_filename
3400 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
3401
3402 except network_exceptions as err:
3403 self.report_error('unable to download video data: %s' % error_to_compat_str(err))
3404 return
3405 except OSError as err:
3406 raise UnavailableVideoError(err)
3407 except (ContentTooShortError, ) as err:
3408 self.report_error(f'content too short (expected {err.expected} bytes and served {err.downloaded})')
3409 return
3410
3411 self._raise_pending_errors(info_dict)
3412 if success and full_filename != '-':
3413
3414 def fixup():
3415 do_fixup = True
3416 fixup_policy = self.params.get('fixup')
3417 vid = info_dict['id']
3418
3419 if fixup_policy in ('ignore', 'never'):
3420 return
3421 elif fixup_policy == 'warn':
3422 do_fixup = 'warn'
3423 elif fixup_policy != 'force':
3424 assert fixup_policy in ('detect_or_warn', None)
3425 if not info_dict.get('__real_download'):
3426 do_fixup = False
3427
3428 def ffmpeg_fixup(cndn, msg, cls):
3429 if not (do_fixup and cndn):
3430 return
3431 elif do_fixup == 'warn':
3432 self.report_warning(f'{vid}: {msg}')
3433 return
3434 pp = cls(self)
3435 if pp.available:
3436 info_dict['__postprocessors'].append(pp)
3437 else:
3438 self.report_warning(f'{vid}: {msg}. Install ffmpeg to fix this automatically')
3439
3440 stretched_ratio = info_dict.get('stretched_ratio')
3441 ffmpeg_fixup(stretched_ratio not in (1, None),
3442 f'Non-uniform pixel ratio {stretched_ratio}',
3443 FFmpegFixupStretchedPP)
3444
3445 downloader = get_suitable_downloader(info_dict, self.params) if 'protocol' in info_dict else None
3446 downloader = downloader.FD_NAME if downloader else None
3447
3448 ext = info_dict.get('ext')
3449 postprocessed_by_ffmpeg = info_dict.get('requested_formats') or any((
3450 isinstance(pp, FFmpegVideoConvertorPP)
3451 and resolve_recode_mapping(ext, pp.mapping)[0] not in (ext, None)
3452 ) for pp in self._pps['post_process'])
3453
3454 if not postprocessed_by_ffmpeg:
3455 ffmpeg_fixup(fd != FFmpegFD and ext == 'm4a'
3456 and info_dict.get('container') == 'm4a_dash',
3457 'writing DASH m4a. Only some players support this container',
3458 FFmpegFixupM4aPP)
3459 ffmpeg_fixup(downloader == 'hlsnative' and not self.params.get('hls_use_mpegts')
3460 or info_dict.get('is_live') and self.params.get('hls_use_mpegts') is None,
3461 'Possible MPEG-TS in MP4 container or malformed AAC timestamps',
3462 FFmpegFixupM3u8PP)
3463 ffmpeg_fixup(info_dict.get('is_live') and downloader == 'dashsegments',
3464 'Possible duplicate MOOV atoms', FFmpegFixupDuplicateMoovPP)
3465
3466 ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed timestamps detected', FFmpegFixupTimestampPP)
3467 ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed duration detected', FFmpegFixupDurationPP)
3468
3469 fixup()
3470 try:
3471 replace_info_dict(self.post_process(dl_filename, info_dict, files_to_move))
3472 except PostProcessingError as err:
3473 self.report_error('Postprocessing: %s' % str(err))
3474 return
3475 try:
3476 for ph in self._post_hooks:
3477 ph(info_dict['filepath'])
3478 except Exception as err:
3479 self.report_error('post hooks: %s' % str(err))
3480 return
3481 info_dict['__write_download_archive'] = True
3482
3483 assert info_dict is original_infodict # Make sure the info_dict was modified in-place
3484 if self.params.get('force_write_download_archive'):
3485 info_dict['__write_download_archive'] = True
3486 check_max_downloads()
3487
3488 def __download_wrapper(self, func):
3489 @functools.wraps(func)
3490 def wrapper(*args, **kwargs):
3491 try:
3492 res = func(*args, **kwargs)
3493 except UnavailableVideoError as e:
3494 self.report_error(e)
3495 except DownloadCancelled as e:
3496 self.to_screen(f'[info] {e}')
3497 if not self.params.get('break_per_url'):
3498 raise
3499 self._num_downloads = 0
3500 else:
3501 if self.params.get('dump_single_json', False):
3502 self.post_extract(res)
3503 self.to_stdout(json.dumps(self.sanitize_info(res)))
3504 return wrapper
3505
3506 def download(self, url_list):
3507 """Download a given list of URLs."""
3508 url_list = variadic(url_list) # Passing a single URL is a common mistake
3509 outtmpl = self.params['outtmpl']['default']
3510 if (len(url_list) > 1
3511 and outtmpl != '-'
3512 and '%' not in outtmpl
3513 and self.params.get('max_downloads') != 1):
3514 raise SameFileError(outtmpl)
3515
3516 for url in url_list:
3517 self.__download_wrapper(self.extract_info)(
3518 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
3519
3520 return self._download_retcode
3521
3522 def download_with_info_file(self, info_filename):
3523 with contextlib.closing(fileinput.FileInput(
3524 [info_filename], mode='r',
3525 openhook=fileinput.hook_encoded('utf-8'))) as f:
3526 # FileInput doesn't have a read method, we can't call json.load
3527 infos = [self.sanitize_info(info, self.params.get('clean_infojson', True))
3528 for info in variadic(json.loads('\n'.join(f)))]
3529 for info in infos:
3530 try:
3531 self.__download_wrapper(self.process_ie_result)(info, download=True)
3532 except (DownloadError, EntryNotInPlaylist, ReExtractInfo) as e:
3533 if not isinstance(e, EntryNotInPlaylist):
3534 self.to_stderr('\r')
3535 webpage_url = info.get('webpage_url')
3536 if webpage_url is None:
3537 raise
3538 self.report_warning(f'The info failed to download: {e}; trying with URL {webpage_url}')
3539 self.download([webpage_url])
3540 return self._download_retcode
3541
3542 @staticmethod
3543 def sanitize_info(info_dict, remove_private_keys=False):
3544 ''' Sanitize the infodict for converting to json '''
3545 if info_dict is None:
3546 return info_dict
3547 info_dict.setdefault('epoch', int(time.time()))
3548 info_dict.setdefault('_type', 'video')
3549 info_dict.setdefault('_version', {
3550 'version': __version__,
3551 'current_git_head': current_git_head(),
3552 'release_git_head': RELEASE_GIT_HEAD,
3553 'repository': ORIGIN,
3554 })
3555
3556 if remove_private_keys:
3557 reject = lambda k, v: v is None or k.startswith('__') or k in {
3558 'requested_downloads', 'requested_formats', 'requested_subtitles', 'requested_entries',
3559 'entries', 'filepath', '_filename', 'filename', 'infojson_filename', 'original_url',
3560 'playlist_autonumber',
3561 }
3562 else:
3563 reject = lambda k, v: False
3564
3565 def filter_fn(obj):
3566 if isinstance(obj, dict):
3567 return {k: filter_fn(v) for k, v in obj.items() if not reject(k, v)}
3568 elif isinstance(obj, (list, tuple, set, LazyList)):
3569 return list(map(filter_fn, obj))
3570 elif obj is None or isinstance(obj, (str, int, float, bool)):
3571 return obj
3572 else:
3573 return repr(obj)
3574
3575 return filter_fn(info_dict)
3576
3577 @staticmethod
3578 def filter_requested_info(info_dict, actually_filter=True):
3579 ''' Alias of sanitize_info for backward compatibility '''
3580 return YoutubeDL.sanitize_info(info_dict, actually_filter)
3581
3582 def _delete_downloaded_files(self, *files_to_delete, info={}, msg=None):
3583 for filename in set(filter(None, files_to_delete)):
3584 if msg:
3585 self.to_screen(msg % filename)
3586 try:
3587 os.remove(filename)
3588 except OSError:
3589 self.report_warning(f'Unable to delete file {filename}')
3590 if filename in info.get('__files_to_move', []): # NB: Delete even if None
3591 del info['__files_to_move'][filename]
3592
3593 @staticmethod
3594 def post_extract(info_dict):
3595 def actual_post_extract(info_dict):
3596 if info_dict.get('_type') in ('playlist', 'multi_video'):
3597 for video_dict in info_dict.get('entries', {}):
3598 actual_post_extract(video_dict or {})
3599 return
3600
3601 post_extractor = info_dict.pop('__post_extractor', None) or (lambda: {})
3602 info_dict.update(post_extractor())
3603
3604 actual_post_extract(info_dict or {})
3605
3606 def run_pp(self, pp, infodict):
3607 files_to_delete = []
3608 if '__files_to_move' not in infodict:
3609 infodict['__files_to_move'] = {}
3610 try:
3611 files_to_delete, infodict = pp.run(infodict)
3612 except PostProcessingError as e:
3613 # Must be True and not 'only_download'
3614 if self.params.get('ignoreerrors') is True:
3615 self.report_error(e)
3616 return infodict
3617 raise
3618
3619 if not files_to_delete:
3620 return infodict
3621 if self.params.get('keepvideo', False):
3622 for f in files_to_delete:
3623 infodict['__files_to_move'].setdefault(f, '')
3624 else:
3625 self._delete_downloaded_files(
3626 *files_to_delete, info=infodict, msg='Deleting original file %s (pass -k to keep)')
3627 return infodict
3628
3629 def run_all_pps(self, key, info, *, additional_pps=None):
3630 if key != 'video':
3631 self._forceprint(key, info)
3632 for pp in (additional_pps or []) + self._pps[key]:
3633 info = self.run_pp(pp, info)
3634 return info
3635
3636 def pre_process(self, ie_info, key='pre_process', files_to_move=None):
3637 info = dict(ie_info)
3638 info['__files_to_move'] = files_to_move or {}
3639 try:
3640 info = self.run_all_pps(key, info)
3641 except PostProcessingError as err:
3642 msg = f'Preprocessing: {err}'
3643 info.setdefault('__pending_error', msg)
3644 self.report_error(msg, is_error=False)
3645 return info, info.pop('__files_to_move', None)
3646
3647 def post_process(self, filename, info, files_to_move=None):
3648 """Run all the postprocessors on the given file."""
3649 info['filepath'] = filename
3650 info['__files_to_move'] = files_to_move or {}
3651 info = self.run_all_pps('post_process', info, additional_pps=info.get('__postprocessors'))
3652 info = self.run_pp(MoveFilesAfterDownloadPP(self), info)
3653 del info['__files_to_move']
3654 return self.run_all_pps('after_move', info)
3655
3656 def _make_archive_id(self, info_dict):
3657 video_id = info_dict.get('id')
3658 if not video_id:
3659 return
3660 # Future-proof against any change in case
3661 # and backwards compatibility with prior versions
3662 extractor = info_dict.get('extractor_key') or info_dict.get('ie_key') # key in a playlist
3663 if extractor is None:
3664 url = str_or_none(info_dict.get('url'))
3665 if not url:
3666 return
3667 # Try to find matching extractor for the URL and take its ie_key
3668 for ie_key, ie in self._ies.items():
3669 if ie.suitable(url):
3670 extractor = ie_key
3671 break
3672 else:
3673 return
3674 return make_archive_id(extractor, video_id)
3675
3676 def in_download_archive(self, info_dict):
3677 if not self.archive:
3678 return False
3679
3680 vid_ids = [self._make_archive_id(info_dict)]
3681 vid_ids.extend(info_dict.get('_old_archive_ids') or [])
3682 return any(id_ in self.archive for id_ in vid_ids)
3683
3684 def record_download_archive(self, info_dict):
3685 fn = self.params.get('download_archive')
3686 if fn is None:
3687 return
3688 vid_id = self._make_archive_id(info_dict)
3689 assert vid_id
3690
3691 self.write_debug(f'Adding to archive: {vid_id}')
3692 if is_path_like(fn):
3693 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
3694 archive_file.write(vid_id + '\n')
3695 self.archive.add(vid_id)
3696
3697 @staticmethod
3698 def format_resolution(format, default='unknown'):
3699 if format.get('vcodec') == 'none' and format.get('acodec') != 'none':
3700 return 'audio only'
3701 if format.get('resolution') is not None:
3702 return format['resolution']
3703 if format.get('width') and format.get('height'):
3704 return '%dx%d' % (format['width'], format['height'])
3705 elif format.get('height'):
3706 return '%sp' % format['height']
3707 elif format.get('width'):
3708 return '%dx?' % format['width']
3709 return default
3710
3711 def _list_format_headers(self, *headers):
3712 if self.params.get('listformats_table', True) is not False:
3713 return [self._format_out(header, self.Styles.HEADERS) for header in headers]
3714 return headers
3715
3716 def _format_note(self, fdict):
3717 res = ''
3718 if fdict.get('ext') in ['f4f', 'f4m']:
3719 res += '(unsupported)'
3720 if fdict.get('language'):
3721 if res:
3722 res += ' '
3723 res += '[%s]' % fdict['language']
3724 if fdict.get('format_note') is not None:
3725 if res:
3726 res += ' '
3727 res += fdict['format_note']
3728 if fdict.get('tbr') is not None:
3729 if res:
3730 res += ', '
3731 res += '%4dk' % fdict['tbr']
3732 if fdict.get('container') is not None:
3733 if res:
3734 res += ', '
3735 res += '%s container' % fdict['container']
3736 if (fdict.get('vcodec') is not None
3737 and fdict.get('vcodec') != 'none'):
3738 if res:
3739 res += ', '
3740 res += fdict['vcodec']
3741 if fdict.get('vbr') is not None:
3742 res += '@'
3743 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
3744 res += 'video@'
3745 if fdict.get('vbr') is not None:
3746 res += '%4dk' % fdict['vbr']
3747 if fdict.get('fps') is not None:
3748 if res:
3749 res += ', '
3750 res += '%sfps' % fdict['fps']
3751 if fdict.get('acodec') is not None:
3752 if res:
3753 res += ', '
3754 if fdict['acodec'] == 'none':
3755 res += 'video only'
3756 else:
3757 res += '%-5s' % fdict['acodec']
3758 elif fdict.get('abr') is not None:
3759 if res:
3760 res += ', '
3761 res += 'audio'
3762 if fdict.get('abr') is not None:
3763 res += '@%3dk' % fdict['abr']
3764 if fdict.get('asr') is not None:
3765 res += ' (%5dHz)' % fdict['asr']
3766 if fdict.get('filesize') is not None:
3767 if res:
3768 res += ', '
3769 res += format_bytes(fdict['filesize'])
3770 elif fdict.get('filesize_approx') is not None:
3771 if res:
3772 res += ', '
3773 res += '~' + format_bytes(fdict['filesize_approx'])
3774 return res
3775
3776 def _get_formats(self, info_dict):
3777 if info_dict.get('formats') is None:
3778 if info_dict.get('url') and info_dict.get('_type', 'video') == 'video':
3779 return [info_dict]
3780 return []
3781 return info_dict['formats']
3782
3783 def render_formats_table(self, info_dict):
3784 formats = self._get_formats(info_dict)
3785 if not formats:
3786 return
3787 if not self.params.get('listformats_table', True) is not False:
3788 table = [
3789 [
3790 format_field(f, 'format_id'),
3791 format_field(f, 'ext'),
3792 self.format_resolution(f),
3793 self._format_note(f)
3794 ] for f in formats if (f.get('preference') or 0) >= -1000]
3795 return render_table(['format code', 'extension', 'resolution', 'note'], table, extra_gap=1)
3796
3797 def simplified_codec(f, field):
3798 assert field in ('acodec', 'vcodec')
3799 codec = f.get(field)
3800 if not codec:
3801 return 'unknown'
3802 elif codec != 'none':
3803 return '.'.join(codec.split('.')[:4])
3804
3805 if field == 'vcodec' and f.get('acodec') == 'none':
3806 return 'images'
3807 elif field == 'acodec' and f.get('vcodec') == 'none':
3808 return ''
3809 return self._format_out('audio only' if field == 'vcodec' else 'video only',
3810 self.Styles.SUPPRESS)
3811
3812 delim = self._format_out('\u2502', self.Styles.DELIM, '|', test_encoding=True)
3813 table = [
3814 [
3815 self._format_out(format_field(f, 'format_id'), self.Styles.ID),
3816 format_field(f, 'ext'),
3817 format_field(f, func=self.format_resolution, ignore=('audio only', 'images')),
3818 format_field(f, 'fps', '\t%d', func=round),
3819 format_field(f, 'dynamic_range', '%s', ignore=(None, 'SDR')).replace('HDR', ''),
3820 format_field(f, 'audio_channels', '\t%s'),
3821 delim, (
3822 format_field(f, 'filesize', ' \t%s', func=format_bytes)
3823 or format_field(f, 'filesize_approx', '≈\t%s', func=format_bytes)
3824 or format_field(try_call(lambda: format_bytes(int(info_dict['duration'] * f['tbr'] * (1024 / 8)))),
3825 None, self._format_out('~\t%s', self.Styles.SUPPRESS))),
3826 format_field(f, 'tbr', '\t%dk', func=round),
3827 shorten_protocol_name(f.get('protocol', '')),
3828 delim,
3829 simplified_codec(f, 'vcodec'),
3830 format_field(f, 'vbr', '\t%dk', func=round),
3831 simplified_codec(f, 'acodec'),
3832 format_field(f, 'abr', '\t%dk', func=round),
3833 format_field(f, 'asr', '\t%s', func=format_decimal_suffix),
3834 join_nonempty(format_field(f, 'language', '[%s]'), join_nonempty(
3835 self._format_out('UNSUPPORTED', self.Styles.BAD_FORMAT) if f.get('ext') in ('f4f', 'f4m') else None,
3836 (self._format_out('Maybe DRM', self.Styles.WARNING) if f.get('has_drm') == 'maybe'
3837 else self._format_out('DRM', self.Styles.BAD_FORMAT) if f.get('has_drm') else None),
3838 format_field(f, 'format_note'),
3839 format_field(f, 'container', ignore=(None, f.get('ext'))),
3840 delim=', '), delim=' '),
3841 ] for f in formats if f.get('preference') is None or f['preference'] >= -1000]
3842 header_line = self._list_format_headers(
3843 'ID', 'EXT', 'RESOLUTION', '\tFPS', 'HDR', 'CH', delim, '\tFILESIZE', '\tTBR', 'PROTO',
3844 delim, 'VCODEC', '\tVBR', 'ACODEC', '\tABR', '\tASR', 'MORE INFO')
3845
3846 return render_table(
3847 header_line, table, hide_empty=True,
3848 delim=self._format_out('\u2500', self.Styles.DELIM, '-', test_encoding=True))
3849
3850 def render_thumbnails_table(self, info_dict):
3851 thumbnails = list(info_dict.get('thumbnails') or [])
3852 if not thumbnails:
3853 return None
3854 return render_table(
3855 self._list_format_headers('ID', 'Width', 'Height', 'URL'),
3856 [[t.get('id'), t.get('width') or 'unknown', t.get('height') or 'unknown', t['url']] for t in thumbnails])
3857
3858 def render_subtitles_table(self, video_id, subtitles):
3859 def _row(lang, formats):
3860 exts, names = zip(*((f['ext'], f.get('name') or 'unknown') for f in reversed(formats)))
3861 if len(set(names)) == 1:
3862 names = [] if names[0] == 'unknown' else names[:1]
3863 return [lang, ', '.join(names), ', '.join(exts)]
3864
3865 if not subtitles:
3866 return None
3867 return render_table(
3868 self._list_format_headers('Language', 'Name', 'Formats'),
3869 [_row(lang, formats) for lang, formats in subtitles.items()],
3870 hide_empty=True)
3871
3872 def __list_table(self, video_id, name, func, *args):
3873 table = func(*args)
3874 if not table:
3875 self.to_screen(f'{video_id} has no {name}')
3876 return
3877 self.to_screen(f'[info] Available {name} for {video_id}:')
3878 self.to_stdout(table)
3879
3880 def list_formats(self, info_dict):
3881 self.__list_table(info_dict['id'], 'formats', self.render_formats_table, info_dict)
3882
3883 def list_thumbnails(self, info_dict):
3884 self.__list_table(info_dict['id'], 'thumbnails', self.render_thumbnails_table, info_dict)
3885
3886 def list_subtitles(self, video_id, subtitles, name='subtitles'):
3887 self.__list_table(video_id, name, self.render_subtitles_table, video_id, subtitles)
3888
3889 def print_debug_header(self):
3890 if not self.params.get('verbose'):
3891 return
3892
3893 from . import _IN_CLI # Must be delayed import
3894
3895 # These imports can be slow. So import them only as needed
3896 from .extractor.extractors import _LAZY_LOADER
3897 from .extractor.extractors import (
3898 _PLUGIN_CLASSES as plugin_ies,
3899 _PLUGIN_OVERRIDES as plugin_ie_overrides
3900 )
3901
3902 def get_encoding(stream):
3903 ret = str(getattr(stream, 'encoding', 'missing (%s)' % type(stream).__name__))
3904 additional_info = []
3905 if os.environ.get('TERM', '').lower() == 'dumb':
3906 additional_info.append('dumb')
3907 if not supports_terminal_sequences(stream):
3908 from .utils import WINDOWS_VT_MODE # Must be imported locally
3909 additional_info.append('No VT' if WINDOWS_VT_MODE is False else 'No ANSI')
3910 if additional_info:
3911 ret = f'{ret} ({",".join(additional_info)})'
3912 return ret
3913
3914 encoding_str = 'Encodings: locale %s, fs %s, pref %s, %s' % (
3915 locale.getpreferredencoding(),
3916 sys.getfilesystemencoding(),
3917 self.get_encoding(),
3918 ', '.join(
3919 f'{key} {get_encoding(stream)}' for key, stream in self._out_files.items_
3920 if stream is not None and key != 'console')
3921 )
3922
3923 logger = self.params.get('logger')
3924 if logger:
3925 write_debug = lambda msg: logger.debug(f'[debug] {msg}')
3926 write_debug(encoding_str)
3927 else:
3928 write_string(f'[debug] {encoding_str}\n', encoding=None)
3929 write_debug = lambda msg: self._write_string(f'[debug] {msg}\n')
3930
3931 source = detect_variant()
3932 if VARIANT not in (None, 'pip'):
3933 source += '*'
3934 klass = type(self)
3935 write_debug(join_nonempty(
3936 f'{REPOSITORY.rpartition("/")[2]} version',
3937 _make_label(ORIGIN, CHANNEL.partition('@')[2] or __version__, __version__),
3938 f'[{RELEASE_GIT_HEAD[:9]}]' if RELEASE_GIT_HEAD else '',
3939 '' if source == 'unknown' else f'({source})',
3940 '' if _IN_CLI else 'API' if klass == YoutubeDL else f'API:{self.__module__}.{klass.__qualname__}',
3941 delim=' '))
3942
3943 if not _IN_CLI:
3944 write_debug(f'params: {self.params}')
3945
3946 if not _LAZY_LOADER:
3947 if os.environ.get('YTDLP_NO_LAZY_EXTRACTORS'):
3948 write_debug('Lazy loading extractors is forcibly disabled')
3949 else:
3950 write_debug('Lazy loading extractors is disabled')
3951 if self.params['compat_opts']:
3952 write_debug('Compatibility options: %s' % ', '.join(self.params['compat_opts']))
3953
3954 if current_git_head():
3955 write_debug(f'Git HEAD: {current_git_head()}')
3956 write_debug(system_identifier())
3957
3958 exe_versions, ffmpeg_features = FFmpegPostProcessor.get_versions_and_features(self)
3959 ffmpeg_features = {key for key, val in ffmpeg_features.items() if val}
3960 if ffmpeg_features:
3961 exe_versions['ffmpeg'] += ' (%s)' % ','.join(sorted(ffmpeg_features))
3962
3963 exe_versions['rtmpdump'] = rtmpdump_version()
3964 exe_versions['phantomjs'] = PhantomJSwrapper._version()
3965 exe_str = ', '.join(
3966 f'{exe} {v}' for exe, v in sorted(exe_versions.items()) if v
3967 ) or 'none'
3968 write_debug('exe versions: %s' % exe_str)
3969
3970 from .compat.compat_utils import get_package_info
3971 from .dependencies import available_dependencies
3972
3973 write_debug('Optional libraries: %s' % (', '.join(sorted({
3974 join_nonempty(*get_package_info(m)) for m in available_dependencies.values()
3975 })) or 'none'))
3976
3977 write_debug(f'Proxy map: {self.proxies}')
3978 write_debug(f'Request Handlers: {", ".join(rh.RH_NAME for rh in self._request_director.handlers.values())}')
3979 for plugin_type, plugins in {'Extractor': plugin_ies, 'Post-Processor': plugin_pps}.items():
3980 display_list = ['%s%s' % (
3981 klass.__name__, '' if klass.__name__ == name else f' as {name}')
3982 for name, klass in plugins.items()]
3983 if plugin_type == 'Extractor':
3984 display_list.extend(f'{plugins[-1].IE_NAME.partition("+")[2]} ({parent.__name__})'
3985 for parent, plugins in plugin_ie_overrides.items())
3986 if not display_list:
3987 continue
3988 write_debug(f'{plugin_type} Plugins: {", ".join(sorted(display_list))}')
3989
3990 plugin_dirs = plugin_directories()
3991 if plugin_dirs:
3992 write_debug(f'Plugin directories: {plugin_dirs}')
3993
3994 # Not implemented
3995 if False and self.params.get('call_home'):
3996 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode()
3997 write_debug('Public IP address: %s' % ipaddr)
3998 latest_version = self.urlopen(
3999 'https://yt-dl.org/latest/version').read().decode()
4000 if version_tuple(latest_version) > version_tuple(__version__):
4001 self.report_warning(
4002 'You are using an outdated version (newest version: %s)! '
4003 'See https://yt-dl.org/update if you need help updating.' %
4004 latest_version)
4005
4006 @functools.cached_property
4007 def proxies(self):
4008 """Global proxy configuration"""
4009 opts_proxy = self.params.get('proxy')
4010 if opts_proxy is not None:
4011 if opts_proxy == '':
4012 opts_proxy = '__noproxy__'
4013 proxies = {'all': opts_proxy}
4014 else:
4015 proxies = urllib.request.getproxies()
4016 # compat. Set HTTPS_PROXY to __noproxy__ to revert
4017 if 'http' in proxies and 'https' not in proxies:
4018 proxies['https'] = proxies['http']
4019
4020 return proxies
4021
4022 @functools.cached_property
4023 def cookiejar(self):
4024 """Global cookiejar instance"""
4025 return load_cookies(
4026 self.params.get('cookiefile'), self.params.get('cookiesfrombrowser'), self)
4027
4028 @property
4029 def _opener(self):
4030 """
4031 Get a urllib OpenerDirector from the Urllib handler (deprecated).
4032 """
4033 self.deprecation_warning('YoutubeDL._opener is deprecated, use YoutubeDL.urlopen()')
4034 handler = self._request_director.handlers['Urllib']
4035 return handler._get_instance(cookiejar=self.cookiejar, proxies=self.proxies)
4036
4037 def urlopen(self, req):
4038 """ Start an HTTP download """
4039 if isinstance(req, str):
4040 req = Request(req)
4041 elif isinstance(req, urllib.request.Request):
4042 self.deprecation_warning(
4043 'Passing a urllib.request.Request object to YoutubeDL.urlopen() is deprecated. '
4044 'Use yt_dlp.networking.common.Request instead.')
4045 req = urllib_req_to_req(req)
4046 assert isinstance(req, Request)
4047
4048 # compat: Assume user:pass url params are basic auth
4049 url, basic_auth_header = extract_basic_auth(req.url)
4050 if basic_auth_header:
4051 req.headers['Authorization'] = basic_auth_header
4052 req.url = sanitize_url(url)
4053
4054 clean_proxies(proxies=req.proxies, headers=req.headers)
4055 clean_headers(req.headers)
4056
4057 try:
4058 return self._request_director.send(req)
4059 except NoSupportingHandlers as e:
4060 for ue in e.unsupported_errors:
4061 # FIXME: This depends on the order of errors.
4062 if not (ue.handler and ue.msg):
4063 continue
4064 if ue.handler.RH_KEY == 'Urllib' and 'unsupported url scheme: "file"' in ue.msg.lower():
4065 raise RequestError(
4066 'file:// URLs are disabled by default in yt-dlp for security reasons. '
4067 'Use --enable-file-urls to enable at your own risk.', cause=ue) from ue
4068 if 'unsupported proxy type: "https"' in ue.msg.lower():
4069 raise RequestError(
4070 'To use an HTTPS proxy for this request, one of the following dependencies needs to be installed: requests')
4071
4072 elif (
4073 re.match(r'unsupported url scheme: "wss?"', ue.msg.lower())
4074 and 'websockets' not in self._request_director.handlers
4075 ):
4076 raise RequestError(
4077 'This request requires WebSocket support. '
4078 'Ensure one of the following dependencies are installed: websockets',
4079 cause=ue) from ue
4080 raise
4081 except SSLError as e:
4082 if 'UNSAFE_LEGACY_RENEGOTIATION_DISABLED' in str(e):
4083 raise RequestError('UNSAFE_LEGACY_RENEGOTIATION_DISABLED: Try using --legacy-server-connect', cause=e) from e
4084 elif 'SSLV3_ALERT_HANDSHAKE_FAILURE' in str(e):
4085 raise RequestError(
4086 'SSLV3_ALERT_HANDSHAKE_FAILURE: The server may not support the current cipher list. '
4087 'Try using --legacy-server-connect', cause=e) from e
4088 raise
4089 except HTTPError as e: # TODO: Remove in a future release
4090 raise _CompatHTTPError(e) from e
4091
4092 def build_request_director(self, handlers, preferences=None):
4093 logger = _YDLLogger(self)
4094 headers = self.params['http_headers'].copy()
4095 proxies = self.proxies.copy()
4096 clean_headers(headers)
4097 clean_proxies(proxies, headers)
4098
4099 director = RequestDirector(logger=logger, verbose=self.params.get('debug_printtraffic'))
4100 for handler in handlers:
4101 director.add_handler(handler(
4102 logger=logger,
4103 headers=headers,
4104 cookiejar=self.cookiejar,
4105 proxies=proxies,
4106 prefer_system_certs='no-certifi' in self.params['compat_opts'],
4107 verify=not self.params.get('nocheckcertificate'),
4108 **traverse_obj(self.params, {
4109 'verbose': 'debug_printtraffic',
4110 'source_address': 'source_address',
4111 'timeout': 'socket_timeout',
4112 'legacy_ssl_support': 'legacyserverconnect',
4113 'enable_file_urls': 'enable_file_urls',
4114 'client_cert': {
4115 'client_certificate': 'client_certificate',
4116 'client_certificate_key': 'client_certificate_key',
4117 'client_certificate_password': 'client_certificate_password',
4118 },
4119 }),
4120 ))
4121 director.preferences.update(preferences or [])
4122 if 'prefer-legacy-http-handler' in self.params['compat_opts']:
4123 director.preferences.add(lambda rh, _: 500 if rh.RH_KEY == 'Urllib' else 0)
4124 return director
4125
4126 def encode(self, s):
4127 if isinstance(s, bytes):
4128 return s # Already encoded
4129
4130 try:
4131 return s.encode(self.get_encoding())
4132 except UnicodeEncodeError as err:
4133 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
4134 raise
4135
4136 def get_encoding(self):
4137 encoding = self.params.get('encoding')
4138 if encoding is None:
4139 encoding = preferredencoding()
4140 return encoding
4141
4142 def _write_info_json(self, label, ie_result, infofn, overwrite=None):
4143 ''' Write infojson and returns True = written, 'exists' = Already exists, False = skip, None = error '''
4144 if overwrite is None:
4145 overwrite = self.params.get('overwrites', True)
4146 if not self.params.get('writeinfojson'):
4147 return False
4148 elif not infofn:
4149 self.write_debug(f'Skipping writing {label} infojson')
4150 return False
4151 elif not self._ensure_dir_exists(infofn):
4152 return None
4153 elif not overwrite and os.path.exists(infofn):
4154 self.to_screen(f'[info] {label.title()} metadata is already present')
4155 return 'exists'
4156
4157 self.to_screen(f'[info] Writing {label} metadata as JSON to: {infofn}')
4158 try:
4159 write_json_file(self.sanitize_info(ie_result, self.params.get('clean_infojson', True)), infofn)
4160 return True
4161 except OSError:
4162 self.report_error(f'Cannot write {label} metadata to JSON file {infofn}')
4163 return None
4164
4165 def _write_description(self, label, ie_result, descfn):
4166 ''' Write description and returns True = written, False = skip, None = error '''
4167 if not self.params.get('writedescription'):
4168 return False
4169 elif not descfn:
4170 self.write_debug(f'Skipping writing {label} description')
4171 return False
4172 elif not self._ensure_dir_exists(descfn):
4173 return None
4174 elif not self.params.get('overwrites', True) and os.path.exists(descfn):
4175 self.to_screen(f'[info] {label.title()} description is already present')
4176 elif ie_result.get('description') is None:
4177 self.to_screen(f'[info] There\'s no {label} description to write')
4178 return False
4179 else:
4180 try:
4181 self.to_screen(f'[info] Writing {label} description to: {descfn}')
4182 with open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
4183 descfile.write(ie_result['description'])
4184 except OSError:
4185 self.report_error(f'Cannot write {label} description file {descfn}')
4186 return None
4187 return True
4188
4189 def _write_subtitles(self, info_dict, filename):
4190 ''' Write subtitles to file and return list of (sub_filename, final_sub_filename); or None if error'''
4191 ret = []
4192 subtitles = info_dict.get('requested_subtitles')
4193 if not (self.params.get('writesubtitles') or self.params.get('writeautomaticsub')):
4194 # subtitles download errors are already managed as troubles in relevant IE
4195 # that way it will silently go on when used with unsupporting IE
4196 return ret
4197 elif not subtitles:
4198 self.to_screen('[info] There are no subtitles for the requested languages')
4199 return ret
4200 sub_filename_base = self.prepare_filename(info_dict, 'subtitle')
4201 if not sub_filename_base:
4202 self.to_screen('[info] Skipping writing video subtitles')
4203 return ret
4204
4205 for sub_lang, sub_info in subtitles.items():
4206 sub_format = sub_info['ext']
4207 sub_filename = subtitles_filename(filename, sub_lang, sub_format, info_dict.get('ext'))
4208 sub_filename_final = subtitles_filename(sub_filename_base, sub_lang, sub_format, info_dict.get('ext'))
4209 existing_sub = self.existing_file((sub_filename_final, sub_filename))
4210 if existing_sub:
4211 self.to_screen(f'[info] Video subtitle {sub_lang}.{sub_format} is already present')
4212 sub_info['filepath'] = existing_sub
4213 ret.append((existing_sub, sub_filename_final))
4214 continue
4215
4216 self.to_screen(f'[info] Writing video subtitles to: {sub_filename}')
4217 if sub_info.get('data') is not None:
4218 try:
4219 # Use newline='' to prevent conversion of newline characters
4220 # See https://github.com/ytdl-org/youtube-dl/issues/10268
4221 with open(sub_filename, 'w', encoding='utf-8', newline='') as subfile:
4222 subfile.write(sub_info['data'])
4223 sub_info['filepath'] = sub_filename
4224 ret.append((sub_filename, sub_filename_final))
4225 continue
4226 except OSError:
4227 self.report_error(f'Cannot write video subtitles file {sub_filename}')
4228 return None
4229
4230 try:
4231 sub_copy = sub_info.copy()
4232 sub_copy.setdefault('http_headers', info_dict.get('http_headers'))
4233 self.dl(sub_filename, sub_copy, subtitle=True)
4234 sub_info['filepath'] = sub_filename
4235 ret.append((sub_filename, sub_filename_final))
4236 except (DownloadError, ExtractorError, IOError, OSError, ValueError) + network_exceptions as err:
4237 msg = f'Unable to download video subtitles for {sub_lang!r}: {err}'
4238 if self.params.get('ignoreerrors') is not True: # False or 'only_download'
4239 if not self.params.get('ignoreerrors'):
4240 self.report_error(msg)
4241 raise DownloadError(msg)
4242 self.report_warning(msg)
4243 return ret
4244
4245 def _write_thumbnails(self, label, info_dict, filename, thumb_filename_base=None):
4246 ''' Write thumbnails to file and return list of (thumb_filename, final_thumb_filename); or None if error '''
4247 write_all = self.params.get('write_all_thumbnails', False)
4248 thumbnails, ret = [], []
4249 if write_all or self.params.get('writethumbnail', False):
4250 thumbnails = info_dict.get('thumbnails') or []
4251 if not thumbnails:
4252 self.to_screen(f'[info] There are no {label} thumbnails to download')
4253 return ret
4254 multiple = write_all and len(thumbnails) > 1
4255
4256 if thumb_filename_base is None:
4257 thumb_filename_base = filename
4258 if thumbnails and not thumb_filename_base:
4259 self.write_debug(f'Skipping writing {label} thumbnail')
4260 return ret
4261
4262 if thumbnails and not self._ensure_dir_exists(filename):
4263 return None
4264
4265 for idx, t in list(enumerate(thumbnails))[::-1]:
4266 thumb_ext = (f'{t["id"]}.' if multiple else '') + determine_ext(t['url'], 'jpg')
4267 thumb_display_id = f'{label} thumbnail {t["id"]}'
4268 thumb_filename = replace_extension(filename, thumb_ext, info_dict.get('ext'))
4269 thumb_filename_final = replace_extension(thumb_filename_base, thumb_ext, info_dict.get('ext'))
4270
4271 existing_thumb = self.existing_file((thumb_filename_final, thumb_filename))
4272 if existing_thumb:
4273 self.to_screen('[info] %s is already present' % (
4274 thumb_display_id if multiple else f'{label} thumbnail').capitalize())
4275 t['filepath'] = existing_thumb
4276 ret.append((existing_thumb, thumb_filename_final))
4277 else:
4278 self.to_screen(f'[info] Downloading {thumb_display_id} ...')
4279 try:
4280 uf = self.urlopen(Request(t['url'], headers=t.get('http_headers', {})))
4281 self.to_screen(f'[info] Writing {thumb_display_id} to: {thumb_filename}')
4282 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
4283 shutil.copyfileobj(uf, thumbf)
4284 ret.append((thumb_filename, thumb_filename_final))
4285 t['filepath'] = thumb_filename
4286 except network_exceptions as err:
4287 if isinstance(err, HTTPError) and err.status == 404:
4288 self.to_screen(f'[info] {thumb_display_id.title()} does not exist')
4289 else:
4290 self.report_warning(f'Unable to download {thumb_display_id}: {err}')
4291 thumbnails.pop(idx)
4292 if ret and not write_all:
4293 break
4294 return ret