]> jfr.im git - yt-dlp.git/blob - yt_dlp/YoutubeDL.py
[cleanup] Misc
[yt-dlp.git] / yt_dlp / YoutubeDL.py
1 import collections
2 import contextlib
3 import datetime
4 import errno
5 import fileinput
6 import functools
7 import io
8 import itertools
9 import json
10 import locale
11 import operator
12 import os
13 import random
14 import re
15 import shutil
16 import string
17 import subprocess
18 import sys
19 import tempfile
20 import time
21 import tokenize
22 import traceback
23 import unicodedata
24 import urllib.request
25
26 from .cache import Cache
27 from .compat import compat_os_name, compat_shlex_quote
28 from .cookies import load_cookies
29 from .downloader import FFmpegFD, get_suitable_downloader, shorten_protocol_name
30 from .downloader.rtmp import rtmpdump_version
31 from .extractor import gen_extractor_classes, get_info_extractor
32 from .extractor.common import UnsupportedURLIE
33 from .extractor.openload import PhantomJSwrapper
34 from .minicurses import format_text
35 from .plugins import directories as plugin_directories
36 from .postprocessor import _PLUGIN_CLASSES as plugin_pps
37 from .postprocessor import (
38 EmbedThumbnailPP,
39 FFmpegFixupDuplicateMoovPP,
40 FFmpegFixupDurationPP,
41 FFmpegFixupM3u8PP,
42 FFmpegFixupM4aPP,
43 FFmpegFixupStretchedPP,
44 FFmpegFixupTimestampPP,
45 FFmpegMergerPP,
46 FFmpegPostProcessor,
47 FFmpegVideoConvertorPP,
48 MoveFilesAfterDownloadPP,
49 get_postprocessor,
50 )
51 from .postprocessor.ffmpeg import resolve_mapping as resolve_recode_mapping
52 from .update import REPOSITORY, current_git_head, detect_variant
53 from .utils import (
54 DEFAULT_OUTTMPL,
55 IDENTITY,
56 LINK_TEMPLATES,
57 MEDIA_EXTENSIONS,
58 NO_DEFAULT,
59 NUMBER_RE,
60 OUTTMPL_TYPES,
61 POSTPROCESS_WHEN,
62 STR_FORMAT_RE_TMPL,
63 STR_FORMAT_TYPES,
64 ContentTooShortError,
65 DateRange,
66 DownloadCancelled,
67 DownloadError,
68 EntryNotInPlaylist,
69 ExistingVideoReached,
70 ExtractorError,
71 FormatSorter,
72 GeoRestrictedError,
73 HEADRequest,
74 ISO3166Utils,
75 LazyList,
76 MaxDownloadsReached,
77 Namespace,
78 PagedList,
79 PerRequestProxyHandler,
80 PlaylistEntries,
81 Popen,
82 PostProcessingError,
83 ReExtractInfo,
84 RejectedVideoReached,
85 SameFileError,
86 UnavailableVideoError,
87 UserNotLive,
88 YoutubeDLCookieProcessor,
89 YoutubeDLHandler,
90 YoutubeDLRedirectHandler,
91 age_restricted,
92 args_to_str,
93 bug_reports_message,
94 date_from_str,
95 deprecation_warning,
96 determine_ext,
97 determine_protocol,
98 encode_compat_str,
99 encodeFilename,
100 error_to_compat_str,
101 escapeHTML,
102 expand_path,
103 filter_dict,
104 float_or_none,
105 format_bytes,
106 format_decimal_suffix,
107 format_field,
108 formatSeconds,
109 get_compatible_ext,
110 get_domain,
111 int_or_none,
112 iri_to_uri,
113 is_path_like,
114 join_nonempty,
115 locked_file,
116 make_archive_id,
117 make_dir,
118 make_HTTPS_handler,
119 merge_headers,
120 network_exceptions,
121 number_of_digits,
122 orderedSet,
123 orderedSet_from_options,
124 parse_filesize,
125 preferredencoding,
126 prepend_extension,
127 register_socks_protocols,
128 remove_terminal_sequences,
129 render_table,
130 replace_extension,
131 sanitize_filename,
132 sanitize_path,
133 sanitize_url,
134 sanitized_Request,
135 std_headers,
136 str_or_none,
137 strftime_or_none,
138 subtitles_filename,
139 supports_terminal_sequences,
140 system_identifier,
141 timetuple_from_msec,
142 to_high_limit_path,
143 traverse_obj,
144 try_call,
145 try_get,
146 url_basename,
147 variadic,
148 version_tuple,
149 windows_enable_vt_mode,
150 write_json_file,
151 write_string,
152 )
153 from .version import CHANNEL, RELEASE_GIT_HEAD, VARIANT, __version__
154
155 if compat_os_name == 'nt':
156 import ctypes
157
158
159 class YoutubeDL:
160 """YoutubeDL class.
161
162 YoutubeDL objects are the ones responsible of downloading the
163 actual video file and writing it to disk if the user has requested
164 it, among some other tasks. In most cases there should be one per
165 program. As, given a video URL, the downloader doesn't know how to
166 extract all the needed information, task that InfoExtractors do, it
167 has to pass the URL to one of them.
168
169 For this, YoutubeDL objects have a method that allows
170 InfoExtractors to be registered in a given order. When it is passed
171 a URL, the YoutubeDL object handles it to the first InfoExtractor it
172 finds that reports being able to handle it. The InfoExtractor extracts
173 all the information about the video or videos the URL refers to, and
174 YoutubeDL process the extracted information, possibly using a File
175 Downloader to download the video.
176
177 YoutubeDL objects accept a lot of parameters. In order not to saturate
178 the object constructor with arguments, it receives a dictionary of
179 options instead. These options are available through the params
180 attribute for the InfoExtractors to use. The YoutubeDL also
181 registers itself as the downloader in charge for the InfoExtractors
182 that are added to it, so this is a "mutual registration".
183
184 Available options:
185
186 username: Username for authentication purposes.
187 password: Password for authentication purposes.
188 videopassword: Password for accessing a video.
189 ap_mso: Adobe Pass multiple-system operator identifier.
190 ap_username: Multiple-system operator account username.
191 ap_password: Multiple-system operator account password.
192 usenetrc: Use netrc for authentication instead.
193 netrc_location: Location of the netrc file. Defaults to ~/.netrc.
194 verbose: Print additional info to stdout.
195 quiet: Do not print messages to stdout.
196 no_warnings: Do not print out anything for warnings.
197 forceprint: A dict with keys WHEN mapped to a list of templates to
198 print to stdout. The allowed keys are video or any of the
199 items in utils.POSTPROCESS_WHEN.
200 For compatibility, a single list is also accepted
201 print_to_file: A dict with keys WHEN (same as forceprint) mapped to
202 a list of tuples with (template, filename)
203 forcejson: Force printing info_dict as JSON.
204 dump_single_json: Force printing the info_dict of the whole playlist
205 (or video) as a single JSON line.
206 force_write_download_archive: Force writing download archive regardless
207 of 'skip_download' or 'simulate'.
208 simulate: Do not download the video files. If unset (or None),
209 simulate only if listsubtitles, listformats or list_thumbnails is used
210 format: Video format code. see "FORMAT SELECTION" for more details.
211 You can also pass a function. The function takes 'ctx' as
212 argument and returns the formats to download.
213 See "build_format_selector" for an implementation
214 allow_unplayable_formats: Allow unplayable formats to be extracted and downloaded.
215 ignore_no_formats_error: Ignore "No video formats" error. Usefull for
216 extracting metadata even if the video is not actually
217 available for download (experimental)
218 format_sort: A list of fields by which to sort the video formats.
219 See "Sorting Formats" for more details.
220 format_sort_force: Force the given format_sort. see "Sorting Formats"
221 for more details.
222 prefer_free_formats: Whether to prefer video formats with free containers
223 over non-free ones of same quality.
224 allow_multiple_video_streams: Allow multiple video streams to be merged
225 into a single file
226 allow_multiple_audio_streams: Allow multiple audio streams to be merged
227 into a single file
228 check_formats Whether to test if the formats are downloadable.
229 Can be True (check all), False (check none),
230 'selected' (check selected formats),
231 or None (check only if requested by extractor)
232 paths: Dictionary of output paths. The allowed keys are 'home'
233 'temp' and the keys of OUTTMPL_TYPES (in utils.py)
234 outtmpl: Dictionary of templates for output names. Allowed keys
235 are 'default' and the keys of OUTTMPL_TYPES (in utils.py).
236 For compatibility with youtube-dl, a single string can also be used
237 outtmpl_na_placeholder: Placeholder for unavailable meta fields.
238 restrictfilenames: Do not allow "&" and spaces in file names
239 trim_file_name: Limit length of filename (extension excluded)
240 windowsfilenames: Force the filenames to be windows compatible
241 ignoreerrors: Do not stop on download/postprocessing errors.
242 Can be 'only_download' to ignore only download errors.
243 Default is 'only_download' for CLI, but False for API
244 skip_playlist_after_errors: Number of allowed failures until the rest of
245 the playlist is skipped
246 allowed_extractors: List of regexes to match against extractor names that are allowed
247 overwrites: Overwrite all video and metadata files if True,
248 overwrite only non-video files if None
249 and don't overwrite any file if False
250 For compatibility with youtube-dl,
251 "nooverwrites" may also be used instead
252 playlist_items: Specific indices of playlist to download.
253 playlistrandom: Download playlist items in random order.
254 lazy_playlist: Process playlist entries as they are received.
255 matchtitle: Download only matching titles.
256 rejecttitle: Reject downloads for matching titles.
257 logger: Log messages to a logging.Logger instance.
258 logtostderr: Print everything to stderr instead of stdout.
259 consoletitle: Display progress in console window's titlebar.
260 writedescription: Write the video description to a .description file
261 writeinfojson: Write the video description to a .info.json file
262 clean_infojson: Remove private fields from the infojson
263 getcomments: Extract video comments. This will not be written to disk
264 unless writeinfojson is also given
265 writeannotations: Write the video annotations to a .annotations.xml file
266 writethumbnail: Write the thumbnail image to a file
267 allow_playlist_files: Whether to write playlists' description, infojson etc
268 also to disk when using the 'write*' options
269 write_all_thumbnails: Write all thumbnail formats to files
270 writelink: Write an internet shortcut file, depending on the
271 current platform (.url/.webloc/.desktop)
272 writeurllink: Write a Windows internet shortcut file (.url)
273 writewebloclink: Write a macOS internet shortcut file (.webloc)
274 writedesktoplink: Write a Linux internet shortcut file (.desktop)
275 writesubtitles: Write the video subtitles to a file
276 writeautomaticsub: Write the automatically generated subtitles to a file
277 listsubtitles: Lists all available subtitles for the video
278 subtitlesformat: The format code for subtitles
279 subtitleslangs: List of languages of the subtitles to download (can be regex).
280 The list may contain "all" to refer to all the available
281 subtitles. The language can be prefixed with a "-" to
282 exclude it from the requested languages, e.g. ['all', '-live_chat']
283 keepvideo: Keep the video file after post-processing
284 daterange: A DateRange object, download only if the upload_date is in the range.
285 skip_download: Skip the actual download of the video file
286 cachedir: Location of the cache files in the filesystem.
287 False to disable filesystem cache.
288 noplaylist: Download single video instead of a playlist if in doubt.
289 age_limit: An integer representing the user's age in years.
290 Unsuitable videos for the given age are skipped.
291 min_views: An integer representing the minimum view count the video
292 must have in order to not be skipped.
293 Videos without view count information are always
294 downloaded. None for no limit.
295 max_views: An integer representing the maximum view count.
296 Videos that are more popular than that are not
297 downloaded.
298 Videos without view count information are always
299 downloaded. None for no limit.
300 download_archive: A set, or the name of a file where all downloads are recorded.
301 Videos already present in the file are not downloaded again.
302 break_on_existing: Stop the download process after attempting to download a
303 file that is in the archive.
304 break_per_url: Whether break_on_reject and break_on_existing
305 should act on each input URL as opposed to for the entire queue
306 cookiefile: File name or text stream from where cookies should be read and dumped to
307 cookiesfrombrowser: A tuple containing the name of the browser, the profile
308 name/path from where cookies are loaded, the name of the keyring,
309 and the container name, e.g. ('chrome', ) or
310 ('vivaldi', 'default', 'BASICTEXT') or ('firefox', 'default', None, 'Meta')
311 legacyserverconnect: Explicitly allow HTTPS connection to servers that do not
312 support RFC 5746 secure renegotiation
313 nocheckcertificate: Do not verify SSL certificates
314 client_certificate: Path to client certificate file in PEM format. May include the private key
315 client_certificate_key: Path to private key file for client certificate
316 client_certificate_password: Password for client certificate private key, if encrypted.
317 If not provided and the key is encrypted, yt-dlp will ask interactively
318 prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
319 (Only supported by some extractors)
320 enable_file_urls: Enable file:// URLs. This is disabled by default for security reasons.
321 http_headers: A dictionary of custom headers to be used for all requests
322 proxy: URL of the proxy server to use
323 geo_verification_proxy: URL of the proxy to use for IP address verification
324 on geo-restricted sites.
325 socket_timeout: Time to wait for unresponsive hosts, in seconds
326 bidi_workaround: Work around buggy terminals without bidirectional text
327 support, using fridibi
328 debug_printtraffic:Print out sent and received HTTP traffic
329 default_search: Prepend this string if an input url is not valid.
330 'auto' for elaborate guessing
331 encoding: Use this encoding instead of the system-specified.
332 extract_flat: Whether to resolve and process url_results further
333 * False: Always process (default)
334 * True: Never process
335 * 'in_playlist': Do not process inside playlist/multi_video
336 * 'discard': Always process, but don't return the result
337 from inside playlist/multi_video
338 * 'discard_in_playlist': Same as "discard", but only for
339 playlists (not multi_video)
340 wait_for_video: If given, wait for scheduled streams to become available.
341 The value should be a tuple containing the range
342 (min_secs, max_secs) to wait between retries
343 postprocessors: A list of dictionaries, each with an entry
344 * key: The name of the postprocessor. See
345 yt_dlp/postprocessor/__init__.py for a list.
346 * when: When to run the postprocessor. Allowed values are
347 the entries of utils.POSTPROCESS_WHEN
348 Assumed to be 'post_process' if not given
349 progress_hooks: A list of functions that get called on download
350 progress, with a dictionary with the entries
351 * status: One of "downloading", "error", or "finished".
352 Check this first and ignore unknown values.
353 * info_dict: The extracted info_dict
354
355 If status is one of "downloading", or "finished", the
356 following properties may also be present:
357 * filename: The final filename (always present)
358 * tmpfilename: The filename we're currently writing to
359 * downloaded_bytes: Bytes on disk
360 * total_bytes: Size of the whole file, None if unknown
361 * total_bytes_estimate: Guess of the eventual file size,
362 None if unavailable.
363 * elapsed: The number of seconds since download started.
364 * eta: The estimated time in seconds, None if unknown
365 * speed: The download speed in bytes/second, None if
366 unknown
367 * fragment_index: The counter of the currently
368 downloaded video fragment.
369 * fragment_count: The number of fragments (= individual
370 files that will be merged)
371
372 Progress hooks are guaranteed to be called at least once
373 (with status "finished") if the download is successful.
374 postprocessor_hooks: A list of functions that get called on postprocessing
375 progress, with a dictionary with the entries
376 * status: One of "started", "processing", or "finished".
377 Check this first and ignore unknown values.
378 * postprocessor: Name of the postprocessor
379 * info_dict: The extracted info_dict
380
381 Progress hooks are guaranteed to be called at least twice
382 (with status "started" and "finished") if the processing is successful.
383 merge_output_format: "/" separated list of extensions to use when merging formats.
384 final_ext: Expected final extension; used to detect when the file was
385 already downloaded and converted
386 fixup: Automatically correct known faults of the file.
387 One of:
388 - "never": do nothing
389 - "warn": only emit a warning
390 - "detect_or_warn": check whether we can do anything
391 about it, warn otherwise (default)
392 source_address: Client-side IP address to bind to.
393 sleep_interval_requests: Number of seconds to sleep between requests
394 during extraction
395 sleep_interval: Number of seconds to sleep before each download when
396 used alone or a lower bound of a range for randomized
397 sleep before each download (minimum possible number
398 of seconds to sleep) when used along with
399 max_sleep_interval.
400 max_sleep_interval:Upper bound of a range for randomized sleep before each
401 download (maximum possible number of seconds to sleep).
402 Must only be used along with sleep_interval.
403 Actual sleep time will be a random float from range
404 [sleep_interval; max_sleep_interval].
405 sleep_interval_subtitles: Number of seconds to sleep before each subtitle download
406 listformats: Print an overview of available video formats and exit.
407 list_thumbnails: Print a table of all thumbnails and exit.
408 match_filter: A function that gets called for every video with the signature
409 (info_dict, *, incomplete: bool) -> Optional[str]
410 For backward compatibility with youtube-dl, the signature
411 (info_dict) -> Optional[str] is also allowed.
412 - If it returns a message, the video is ignored.
413 - If it returns None, the video is downloaded.
414 - If it returns utils.NO_DEFAULT, the user is interactively
415 asked whether to download the video.
416 - Raise utils.DownloadCancelled(msg) to abort remaining
417 downloads when a video is rejected.
418 match_filter_func in utils.py is one example for this.
419 no_color: Do not emit color codes in output.
420 geo_bypass: Bypass geographic restriction via faking X-Forwarded-For
421 HTTP header
422 geo_bypass_country:
423 Two-letter ISO 3166-2 country code that will be used for
424 explicit geographic restriction bypassing via faking
425 X-Forwarded-For HTTP header
426 geo_bypass_ip_block:
427 IP range in CIDR notation that will be used similarly to
428 geo_bypass_country
429 external_downloader: A dictionary of protocol keys and the executable of the
430 external downloader to use for it. The allowed protocols
431 are default|http|ftp|m3u8|dash|rtsp|rtmp|mms.
432 Set the value to 'native' to use the native downloader
433 compat_opts: Compatibility options. See "Differences in default behavior".
434 The following options do not work when used through the API:
435 filename, abort-on-error, multistreams, no-live-chat, format-sort
436 no-clean-infojson, no-playlist-metafiles, no-keep-subs, no-attach-info-json.
437 Refer __init__.py for their implementation
438 progress_template: Dictionary of templates for progress outputs.
439 Allowed keys are 'download', 'postprocess',
440 'download-title' (console title) and 'postprocess-title'.
441 The template is mapped on a dictionary with keys 'progress' and 'info'
442 retry_sleep_functions: Dictionary of functions that takes the number of attempts
443 as argument and returns the time to sleep in seconds.
444 Allowed keys are 'http', 'fragment', 'file_access'
445 download_ranges: A callback function that gets called for every video with
446 the signature (info_dict, ydl) -> Iterable[Section].
447 Only the returned sections will be downloaded.
448 Each Section is a dict with the following keys:
449 * start_time: Start time of the section in seconds
450 * end_time: End time of the section in seconds
451 * title: Section title (Optional)
452 * index: Section number (Optional)
453 force_keyframes_at_cuts: Re-encode the video when downloading ranges to get precise cuts
454 noprogress: Do not print the progress bar
455 live_from_start: Whether to download livestreams videos from the start
456
457 The following parameters are not used by YoutubeDL itself, they are used by
458 the downloader (see yt_dlp/downloader/common.py):
459 nopart, updatetime, buffersize, ratelimit, throttledratelimit, min_filesize,
460 max_filesize, test, noresizebuffer, retries, file_access_retries, fragment_retries,
461 continuedl, xattr_set_filesize, hls_use_mpegts, http_chunk_size,
462 external_downloader_args, concurrent_fragment_downloads.
463
464 The following options are used by the post processors:
465 ffmpeg_location: Location of the ffmpeg/avconv binary; either the path
466 to the binary or its containing directory.
467 postprocessor_args: A dictionary of postprocessor/executable keys (in lower case)
468 and a list of additional command-line arguments for the
469 postprocessor/executable. The dict can also have "PP+EXE" keys
470 which are used when the given exe is used by the given PP.
471 Use 'default' as the name for arguments to passed to all PP
472 For compatibility with youtube-dl, a single list of args
473 can also be used
474
475 The following options are used by the extractors:
476 extractor_retries: Number of times to retry for known errors
477 dynamic_mpd: Whether to process dynamic DASH manifests (default: True)
478 hls_split_discontinuity: Split HLS playlists to different formats at
479 discontinuities such as ad breaks (default: False)
480 extractor_args: A dictionary of arguments to be passed to the extractors.
481 See "EXTRACTOR ARGUMENTS" for details.
482 E.g. {'youtube': {'skip': ['dash', 'hls']}}
483 mark_watched: Mark videos watched (even with --simulate). Only for YouTube
484
485 The following options are deprecated and may be removed in the future:
486
487 break_on_reject: Stop the download process when encountering a video that
488 has been filtered out.
489 - `raise DownloadCancelled(msg)` in match_filter instead
490 force_generic_extractor: Force downloader to use the generic extractor
491 - Use allowed_extractors = ['generic', 'default']
492 playliststart: - Use playlist_items
493 Playlist item to start at.
494 playlistend: - Use playlist_items
495 Playlist item to end at.
496 playlistreverse: - Use playlist_items
497 Download playlist items in reverse order.
498 forceurl: - Use forceprint
499 Force printing final URL.
500 forcetitle: - Use forceprint
501 Force printing title.
502 forceid: - Use forceprint
503 Force printing ID.
504 forcethumbnail: - Use forceprint
505 Force printing thumbnail URL.
506 forcedescription: - Use forceprint
507 Force printing description.
508 forcefilename: - Use forceprint
509 Force printing final filename.
510 forceduration: - Use forceprint
511 Force printing duration.
512 allsubtitles: - Use subtitleslangs = ['all']
513 Downloads all the subtitles of the video
514 (requires writesubtitles or writeautomaticsub)
515 include_ads: - Doesn't work
516 Download ads as well
517 call_home: - Not implemented
518 Boolean, true iff we are allowed to contact the
519 yt-dlp servers for debugging.
520 post_hooks: - Register a custom postprocessor
521 A list of functions that get called as the final step
522 for each video file, after all postprocessors have been
523 called. The filename will be passed as the only argument.
524 hls_prefer_native: - Use external_downloader = {'m3u8': 'native'} or {'m3u8': 'ffmpeg'}.
525 Use the native HLS downloader instead of ffmpeg/avconv
526 if True, otherwise use ffmpeg/avconv if False, otherwise
527 use downloader suggested by extractor if None.
528 prefer_ffmpeg: - avconv support is deprecated
529 If False, use avconv instead of ffmpeg if both are available,
530 otherwise prefer ffmpeg.
531 youtube_include_dash_manifest: - Use extractor_args
532 If True (default), DASH manifests and related
533 data will be downloaded and processed by extractor.
534 You can reduce network I/O by disabling it if you don't
535 care about DASH. (only for youtube)
536 youtube_include_hls_manifest: - Use extractor_args
537 If True (default), HLS manifests and related
538 data will be downloaded and processed by extractor.
539 You can reduce network I/O by disabling it if you don't
540 care about HLS. (only for youtube)
541 """
542
543 _NUMERIC_FIELDS = {
544 'width', 'height', 'asr', 'audio_channels', 'fps',
545 'tbr', 'abr', 'vbr', 'filesize', 'filesize_approx',
546 'timestamp', 'release_timestamp',
547 'duration', 'view_count', 'like_count', 'dislike_count', 'repost_count',
548 'average_rating', 'comment_count', 'age_limit',
549 'start_time', 'end_time',
550 'chapter_number', 'season_number', 'episode_number',
551 'track_number', 'disc_number', 'release_year',
552 }
553
554 _format_fields = {
555 # NB: Keep in sync with the docstring of extractor/common.py
556 'url', 'manifest_url', 'manifest_stream_number', 'ext', 'format', 'format_id', 'format_note',
557 'width', 'height', 'aspect_ratio', 'resolution', 'dynamic_range', 'tbr', 'abr', 'acodec', 'asr', 'audio_channels',
558 'vbr', 'fps', 'vcodec', 'container', 'filesize', 'filesize_approx', 'rows', 'columns',
559 'player_url', 'protocol', 'fragment_base_url', 'fragments', 'is_from_start',
560 'preference', 'language', 'language_preference', 'quality', 'source_preference',
561 'http_headers', 'stretched_ratio', 'no_resume', 'has_drm', 'extra_param_to_segment_url', 'hls_aes', 'downloader_options',
562 'page_url', 'app', 'play_path', 'tc_url', 'flash_version', 'rtmp_live', 'rtmp_conn', 'rtmp_protocol', 'rtmp_real_time'
563 }
564 _format_selection_exts = {
565 'audio': set(MEDIA_EXTENSIONS.common_audio),
566 'video': set(MEDIA_EXTENSIONS.common_video + ('3gp', )),
567 'storyboards': set(MEDIA_EXTENSIONS.storyboards),
568 }
569
570 def __init__(self, params=None, auto_init=True):
571 """Create a FileDownloader object with the given options.
572 @param auto_init Whether to load the default extractors and print header (if verbose).
573 Set to 'no_verbose_header' to not print the header
574 """
575 if params is None:
576 params = {}
577 self.params = params
578 self._ies = {}
579 self._ies_instances = {}
580 self._pps = {k: [] for k in POSTPROCESS_WHEN}
581 self._printed_messages = set()
582 self._first_webpage_request = True
583 self._post_hooks = []
584 self._progress_hooks = []
585 self._postprocessor_hooks = []
586 self._download_retcode = 0
587 self._num_downloads = 0
588 self._num_videos = 0
589 self._playlist_level = 0
590 self._playlist_urls = set()
591 self.cache = Cache(self)
592
593 stdout = sys.stderr if self.params.get('logtostderr') else sys.stdout
594 self._out_files = Namespace(
595 out=stdout,
596 error=sys.stderr,
597 screen=sys.stderr if self.params.get('quiet') else stdout,
598 console=None if compat_os_name == 'nt' else next(
599 filter(supports_terminal_sequences, (sys.stderr, sys.stdout)), None)
600 )
601
602 try:
603 windows_enable_vt_mode()
604 except Exception as e:
605 self.write_debug(f'Failed to enable VT mode: {e}')
606
607 self._allow_colors = Namespace(**{
608 type_: not self.params.get('no_color') and supports_terminal_sequences(stream)
609 for type_, stream in self._out_files.items_ if type_ != 'console'
610 })
611
612 # The code is left like this to be reused for future deprecations
613 MIN_SUPPORTED, MIN_RECOMMENDED = (3, 7), (3, 7)
614 current_version = sys.version_info[:2]
615 if current_version < MIN_RECOMMENDED:
616 msg = ('Support for Python version %d.%d has been deprecated. '
617 'See https://github.com/yt-dlp/yt-dlp/issues/3764 for more details.'
618 '\n You will no longer receive updates on this version')
619 if current_version < MIN_SUPPORTED:
620 msg = 'Python version %d.%d is no longer supported'
621 self.deprecated_feature(
622 f'{msg}! Please update to Python %d.%d or above' % (*current_version, *MIN_RECOMMENDED))
623
624 if self.params.get('allow_unplayable_formats'):
625 self.report_warning(
626 f'You have asked for {self._format_err("UNPLAYABLE", self.Styles.EMPHASIS)} formats to be listed/downloaded. '
627 'This is a developer option intended for debugging. \n'
628 ' If you experience any issues while using this option, '
629 f'{self._format_err("DO NOT", self.Styles.ERROR)} open a bug report')
630
631 if self.params.get('bidi_workaround', False):
632 try:
633 import pty
634 master, slave = pty.openpty()
635 width = shutil.get_terminal_size().columns
636 width_args = [] if width is None else ['-w', str(width)]
637 sp_kwargs = {'stdin': subprocess.PIPE, 'stdout': slave, 'stderr': self._out_files.error}
638 try:
639 self._output_process = Popen(['bidiv'] + width_args, **sp_kwargs)
640 except OSError:
641 self._output_process = Popen(['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
642 self._output_channel = os.fdopen(master, 'rb')
643 except OSError as ose:
644 if ose.errno == errno.ENOENT:
645 self.report_warning(
646 'Could not find fribidi executable, ignoring --bidi-workaround. '
647 'Make sure that fribidi is an executable file in one of the directories in your $PATH.')
648 else:
649 raise
650
651 self.params['compat_opts'] = set(self.params.get('compat_opts', ()))
652 if auto_init and auto_init != 'no_verbose_header':
653 self.print_debug_header()
654
655 def check_deprecated(param, option, suggestion):
656 if self.params.get(param) is not None:
657 self.report_warning(f'{option} is deprecated. Use {suggestion} instead')
658 return True
659 return False
660
661 if check_deprecated('cn_verification_proxy', '--cn-verification-proxy', '--geo-verification-proxy'):
662 if self.params.get('geo_verification_proxy') is None:
663 self.params['geo_verification_proxy'] = self.params['cn_verification_proxy']
664
665 check_deprecated('autonumber', '--auto-number', '-o "%(autonumber)s-%(title)s.%(ext)s"')
666 check_deprecated('usetitle', '--title', '-o "%(title)s-%(id)s.%(ext)s"')
667 check_deprecated('useid', '--id', '-o "%(id)s.%(ext)s"')
668
669 for msg in self.params.get('_warnings', []):
670 self.report_warning(msg)
671 for msg in self.params.get('_deprecation_warnings', []):
672 self.deprecated_feature(msg)
673
674 if 'list-formats' in self.params['compat_opts']:
675 self.params['listformats_table'] = False
676
677 if 'overwrites' not in self.params and self.params.get('nooverwrites') is not None:
678 # nooverwrites was unnecessarily changed to overwrites
679 # in 0c3d0f51778b153f65c21906031c2e091fcfb641
680 # This ensures compatibility with both keys
681 self.params['overwrites'] = not self.params['nooverwrites']
682 elif self.params.get('overwrites') is None:
683 self.params.pop('overwrites', None)
684 else:
685 self.params['nooverwrites'] = not self.params['overwrites']
686
687 if self.params.get('simulate') is None and any((
688 self.params.get('list_thumbnails'),
689 self.params.get('listformats'),
690 self.params.get('listsubtitles'),
691 )):
692 self.params['simulate'] = 'list_only'
693
694 self.params.setdefault('forceprint', {})
695 self.params.setdefault('print_to_file', {})
696
697 # Compatibility with older syntax
698 if not isinstance(params['forceprint'], dict):
699 self.params['forceprint'] = {'video': params['forceprint']}
700
701 if auto_init:
702 self.add_default_info_extractors()
703
704 if (sys.platform != 'win32'
705 and sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968']
706 and not self.params.get('restrictfilenames', False)):
707 # Unicode filesystem API will throw errors (#1474, #13027)
708 self.report_warning(
709 'Assuming --restrict-filenames since file system encoding '
710 'cannot encode all characters. '
711 'Set the LC_ALL environment variable to fix this.')
712 self.params['restrictfilenames'] = True
713
714 self._parse_outtmpl()
715
716 # Creating format selector here allows us to catch syntax errors before the extraction
717 self.format_selector = (
718 self.params.get('format') if self.params.get('format') in (None, '-')
719 else self.params['format'] if callable(self.params['format'])
720 else self.build_format_selector(self.params['format']))
721
722 # Set http_headers defaults according to std_headers
723 self.params['http_headers'] = merge_headers(std_headers, self.params.get('http_headers', {}))
724
725 hooks = {
726 'post_hooks': self.add_post_hook,
727 'progress_hooks': self.add_progress_hook,
728 'postprocessor_hooks': self.add_postprocessor_hook,
729 }
730 for opt, fn in hooks.items():
731 for ph in self.params.get(opt, []):
732 fn(ph)
733
734 for pp_def_raw in self.params.get('postprocessors', []):
735 pp_def = dict(pp_def_raw)
736 when = pp_def.pop('when', 'post_process')
737 self.add_post_processor(
738 get_postprocessor(pp_def.pop('key'))(self, **pp_def),
739 when=when)
740
741 self._setup_opener()
742 register_socks_protocols()
743
744 def preload_download_archive(fn):
745 """Preload the archive, if any is specified"""
746 archive = set()
747 if fn is None:
748 return archive
749 elif not is_path_like(fn):
750 return fn
751
752 self.write_debug(f'Loading archive file {fn!r}')
753 try:
754 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
755 for line in archive_file:
756 archive.add(line.strip())
757 except OSError as ioe:
758 if ioe.errno != errno.ENOENT:
759 raise
760 return archive
761
762 self.archive = preload_download_archive(self.params.get('download_archive'))
763
764 def warn_if_short_id(self, argv):
765 # short YouTube ID starting with dash?
766 idxs = [
767 i for i, a in enumerate(argv)
768 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
769 if idxs:
770 correct_argv = (
771 ['yt-dlp']
772 + [a for i, a in enumerate(argv) if i not in idxs]
773 + ['--'] + [argv[i] for i in idxs]
774 )
775 self.report_warning(
776 'Long argument string detected. '
777 'Use -- to separate parameters and URLs, like this:\n%s' %
778 args_to_str(correct_argv))
779
780 def add_info_extractor(self, ie):
781 """Add an InfoExtractor object to the end of the list."""
782 ie_key = ie.ie_key()
783 self._ies[ie_key] = ie
784 if not isinstance(ie, type):
785 self._ies_instances[ie_key] = ie
786 ie.set_downloader(self)
787
788 def get_info_extractor(self, ie_key):
789 """
790 Get an instance of an IE with name ie_key, it will try to get one from
791 the _ies list, if there's no instance it will create a new one and add
792 it to the extractor list.
793 """
794 ie = self._ies_instances.get(ie_key)
795 if ie is None:
796 ie = get_info_extractor(ie_key)()
797 self.add_info_extractor(ie)
798 return ie
799
800 def add_default_info_extractors(self):
801 """
802 Add the InfoExtractors returned by gen_extractors to the end of the list
803 """
804 all_ies = {ie.IE_NAME.lower(): ie for ie in gen_extractor_classes()}
805 all_ies['end'] = UnsupportedURLIE()
806 try:
807 ie_names = orderedSet_from_options(
808 self.params.get('allowed_extractors', ['default']), {
809 'all': list(all_ies),
810 'default': [name for name, ie in all_ies.items() if ie._ENABLED],
811 }, use_regex=True)
812 except re.error as e:
813 raise ValueError(f'Wrong regex for allowed_extractors: {e.pattern}')
814 for name in ie_names:
815 self.add_info_extractor(all_ies[name])
816 self.write_debug(f'Loaded {len(ie_names)} extractors')
817
818 def add_post_processor(self, pp, when='post_process'):
819 """Add a PostProcessor object to the end of the chain."""
820 assert when in POSTPROCESS_WHEN, f'Invalid when={when}'
821 self._pps[when].append(pp)
822 pp.set_downloader(self)
823
824 def add_post_hook(self, ph):
825 """Add the post hook"""
826 self._post_hooks.append(ph)
827
828 def add_progress_hook(self, ph):
829 """Add the download progress hook"""
830 self._progress_hooks.append(ph)
831
832 def add_postprocessor_hook(self, ph):
833 """Add the postprocessing progress hook"""
834 self._postprocessor_hooks.append(ph)
835 for pps in self._pps.values():
836 for pp in pps:
837 pp.add_progress_hook(ph)
838
839 def _bidi_workaround(self, message):
840 if not hasattr(self, '_output_channel'):
841 return message
842
843 assert hasattr(self, '_output_process')
844 assert isinstance(message, str)
845 line_count = message.count('\n') + 1
846 self._output_process.stdin.write((message + '\n').encode())
847 self._output_process.stdin.flush()
848 res = ''.join(self._output_channel.readline().decode()
849 for _ in range(line_count))
850 return res[:-len('\n')]
851
852 def _write_string(self, message, out=None, only_once=False):
853 if only_once:
854 if message in self._printed_messages:
855 return
856 self._printed_messages.add(message)
857 write_string(message, out=out, encoding=self.params.get('encoding'))
858
859 def to_stdout(self, message, skip_eol=False, quiet=None):
860 """Print message to stdout"""
861 if quiet is not None:
862 self.deprecation_warning('"YoutubeDL.to_stdout" no longer accepts the argument quiet. '
863 'Use "YoutubeDL.to_screen" instead')
864 if skip_eol is not False:
865 self.deprecation_warning('"YoutubeDL.to_stdout" no longer accepts the argument skip_eol. '
866 'Use "YoutubeDL.to_screen" instead')
867 self._write_string(f'{self._bidi_workaround(message)}\n', self._out_files.out)
868
869 def to_screen(self, message, skip_eol=False, quiet=None, only_once=False):
870 """Print message to screen if not in quiet mode"""
871 if self.params.get('logger'):
872 self.params['logger'].debug(message)
873 return
874 if (self.params.get('quiet') if quiet is None else quiet) and not self.params.get('verbose'):
875 return
876 self._write_string(
877 '%s%s' % (self._bidi_workaround(message), ('' if skip_eol else '\n')),
878 self._out_files.screen, only_once=only_once)
879
880 def to_stderr(self, message, only_once=False):
881 """Print message to stderr"""
882 assert isinstance(message, str)
883 if self.params.get('logger'):
884 self.params['logger'].error(message)
885 else:
886 self._write_string(f'{self._bidi_workaround(message)}\n', self._out_files.error, only_once=only_once)
887
888 def _send_console_code(self, code):
889 if compat_os_name == 'nt' or not self._out_files.console:
890 return
891 self._write_string(code, self._out_files.console)
892
893 def to_console_title(self, message):
894 if not self.params.get('consoletitle', False):
895 return
896 message = remove_terminal_sequences(message)
897 if compat_os_name == 'nt':
898 if ctypes.windll.kernel32.GetConsoleWindow():
899 # c_wchar_p() might not be necessary if `message` is
900 # already of type unicode()
901 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
902 else:
903 self._send_console_code(f'\033]0;{message}\007')
904
905 def save_console_title(self):
906 if not self.params.get('consoletitle') or self.params.get('simulate'):
907 return
908 self._send_console_code('\033[22;0t') # Save the title on stack
909
910 def restore_console_title(self):
911 if not self.params.get('consoletitle') or self.params.get('simulate'):
912 return
913 self._send_console_code('\033[23;0t') # Restore the title from stack
914
915 def __enter__(self):
916 self.save_console_title()
917 return self
918
919 def __exit__(self, *args):
920 self.restore_console_title()
921
922 if self.params.get('cookiefile') is not None:
923 self.cookiejar.save(ignore_discard=True, ignore_expires=True)
924
925 def trouble(self, message=None, tb=None, is_error=True):
926 """Determine action to take when a download problem appears.
927
928 Depending on if the downloader has been configured to ignore
929 download errors or not, this method may throw an exception or
930 not when errors are found, after printing the message.
931
932 @param tb If given, is additional traceback information
933 @param is_error Whether to raise error according to ignorerrors
934 """
935 if message is not None:
936 self.to_stderr(message)
937 if self.params.get('verbose'):
938 if tb is None:
939 if sys.exc_info()[0]: # if .trouble has been called from an except block
940 tb = ''
941 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
942 tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
943 tb += encode_compat_str(traceback.format_exc())
944 else:
945 tb_data = traceback.format_list(traceback.extract_stack())
946 tb = ''.join(tb_data)
947 if tb:
948 self.to_stderr(tb)
949 if not is_error:
950 return
951 if not self.params.get('ignoreerrors'):
952 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
953 exc_info = sys.exc_info()[1].exc_info
954 else:
955 exc_info = sys.exc_info()
956 raise DownloadError(message, exc_info)
957 self._download_retcode = 1
958
959 Styles = Namespace(
960 HEADERS='yellow',
961 EMPHASIS='light blue',
962 FILENAME='green',
963 ID='green',
964 DELIM='blue',
965 ERROR='red',
966 WARNING='yellow',
967 SUPPRESS='light black',
968 )
969
970 def _format_text(self, handle, allow_colors, text, f, fallback=None, *, test_encoding=False):
971 text = str(text)
972 if test_encoding:
973 original_text = text
974 # handle.encoding can be None. See https://github.com/yt-dlp/yt-dlp/issues/2711
975 encoding = self.params.get('encoding') or getattr(handle, 'encoding', None) or 'ascii'
976 text = text.encode(encoding, 'ignore').decode(encoding)
977 if fallback is not None and text != original_text:
978 text = fallback
979 return format_text(text, f) if allow_colors else text if fallback is None else fallback
980
981 def _format_out(self, *args, **kwargs):
982 return self._format_text(self._out_files.out, self._allow_colors.out, *args, **kwargs)
983
984 def _format_screen(self, *args, **kwargs):
985 return self._format_text(self._out_files.screen, self._allow_colors.screen, *args, **kwargs)
986
987 def _format_err(self, *args, **kwargs):
988 return self._format_text(self._out_files.error, self._allow_colors.error, *args, **kwargs)
989
990 def report_warning(self, message, only_once=False):
991 '''
992 Print the message to stderr, it will be prefixed with 'WARNING:'
993 If stderr is a tty file the 'WARNING:' will be colored
994 '''
995 if self.params.get('logger') is not None:
996 self.params['logger'].warning(message)
997 else:
998 if self.params.get('no_warnings'):
999 return
1000 self.to_stderr(f'{self._format_err("WARNING:", self.Styles.WARNING)} {message}', only_once)
1001
1002 def deprecation_warning(self, message, *, stacklevel=0):
1003 deprecation_warning(
1004 message, stacklevel=stacklevel + 1, printer=self.report_error, is_error=False)
1005
1006 def deprecated_feature(self, message):
1007 if self.params.get('logger') is not None:
1008 self.params['logger'].warning(f'Deprecated Feature: {message}')
1009 self.to_stderr(f'{self._format_err("Deprecated Feature:", self.Styles.ERROR)} {message}', True)
1010
1011 def report_error(self, message, *args, **kwargs):
1012 '''
1013 Do the same as trouble, but prefixes the message with 'ERROR:', colored
1014 in red if stderr is a tty file.
1015 '''
1016 self.trouble(f'{self._format_err("ERROR:", self.Styles.ERROR)} {message}', *args, **kwargs)
1017
1018 def write_debug(self, message, only_once=False):
1019 '''Log debug message or Print message to stderr'''
1020 if not self.params.get('verbose', False):
1021 return
1022 message = f'[debug] {message}'
1023 if self.params.get('logger'):
1024 self.params['logger'].debug(message)
1025 else:
1026 self.to_stderr(message, only_once)
1027
1028 def report_file_already_downloaded(self, file_name):
1029 """Report file has already been fully downloaded."""
1030 try:
1031 self.to_screen('[download] %s has already been downloaded' % file_name)
1032 except UnicodeEncodeError:
1033 self.to_screen('[download] The file has already been downloaded')
1034
1035 def report_file_delete(self, file_name):
1036 """Report that existing file will be deleted."""
1037 try:
1038 self.to_screen('Deleting existing file %s' % file_name)
1039 except UnicodeEncodeError:
1040 self.to_screen('Deleting existing file')
1041
1042 def raise_no_formats(self, info, forced=False, *, msg=None):
1043 has_drm = info.get('_has_drm')
1044 ignored, expected = self.params.get('ignore_no_formats_error'), bool(msg)
1045 msg = msg or has_drm and 'This video is DRM protected' or 'No video formats found!'
1046 if forced or not ignored:
1047 raise ExtractorError(msg, video_id=info['id'], ie=info['extractor'],
1048 expected=has_drm or ignored or expected)
1049 else:
1050 self.report_warning(msg)
1051
1052 def parse_outtmpl(self):
1053 self.deprecation_warning('"YoutubeDL.parse_outtmpl" is deprecated and may be removed in a future version')
1054 self._parse_outtmpl()
1055 return self.params['outtmpl']
1056
1057 def _parse_outtmpl(self):
1058 sanitize = IDENTITY
1059 if self.params.get('restrictfilenames'): # Remove spaces in the default template
1060 sanitize = lambda x: x.replace(' - ', ' ').replace(' ', '-')
1061
1062 outtmpl = self.params.setdefault('outtmpl', {})
1063 if not isinstance(outtmpl, dict):
1064 self.params['outtmpl'] = outtmpl = {'default': outtmpl}
1065 outtmpl.update({k: sanitize(v) for k, v in DEFAULT_OUTTMPL.items() if outtmpl.get(k) is None})
1066
1067 def get_output_path(self, dir_type='', filename=None):
1068 paths = self.params.get('paths', {})
1069 assert isinstance(paths, dict), '"paths" parameter must be a dictionary'
1070 path = os.path.join(
1071 expand_path(paths.get('home', '').strip()),
1072 expand_path(paths.get(dir_type, '').strip()) if dir_type else '',
1073 filename or '')
1074 return sanitize_path(path, force=self.params.get('windowsfilenames'))
1075
1076 @staticmethod
1077 def _outtmpl_expandpath(outtmpl):
1078 # expand_path translates '%%' into '%' and '$$' into '$'
1079 # correspondingly that is not what we want since we need to keep
1080 # '%%' intact for template dict substitution step. Working around
1081 # with boundary-alike separator hack.
1082 sep = ''.join(random.choices(string.ascii_letters, k=32))
1083 outtmpl = outtmpl.replace('%%', f'%{sep}%').replace('$$', f'${sep}$')
1084
1085 # outtmpl should be expand_path'ed before template dict substitution
1086 # because meta fields may contain env variables we don't want to
1087 # be expanded. E.g. for outtmpl "%(title)s.%(ext)s" and
1088 # title "Hello $PATH", we don't want `$PATH` to be expanded.
1089 return expand_path(outtmpl).replace(sep, '')
1090
1091 @staticmethod
1092 def escape_outtmpl(outtmpl):
1093 ''' Escape any remaining strings like %s, %abc% etc. '''
1094 return re.sub(
1095 STR_FORMAT_RE_TMPL.format('', '(?![%(\0])'),
1096 lambda mobj: ('' if mobj.group('has_key') else '%') + mobj.group(0),
1097 outtmpl)
1098
1099 @classmethod
1100 def validate_outtmpl(cls, outtmpl):
1101 ''' @return None or Exception object '''
1102 outtmpl = re.sub(
1103 STR_FORMAT_RE_TMPL.format('[^)]*', '[ljhqBUDS]'),
1104 lambda mobj: f'{mobj.group(0)[:-1]}s',
1105 cls._outtmpl_expandpath(outtmpl))
1106 try:
1107 cls.escape_outtmpl(outtmpl) % collections.defaultdict(int)
1108 return None
1109 except ValueError as err:
1110 return err
1111
1112 @staticmethod
1113 def _copy_infodict(info_dict):
1114 info_dict = dict(info_dict)
1115 info_dict.pop('__postprocessors', None)
1116 info_dict.pop('__pending_error', None)
1117 return info_dict
1118
1119 def prepare_outtmpl(self, outtmpl, info_dict, sanitize=False):
1120 """ Make the outtmpl and info_dict suitable for substitution: ydl.escape_outtmpl(outtmpl) % info_dict
1121 @param sanitize Whether to sanitize the output as a filename.
1122 For backward compatibility, a function can also be passed
1123 """
1124
1125 info_dict.setdefault('epoch', int(time.time())) # keep epoch consistent once set
1126
1127 info_dict = self._copy_infodict(info_dict)
1128 info_dict['duration_string'] = ( # %(duration>%H-%M-%S)s is wrong if duration > 24hrs
1129 formatSeconds(info_dict['duration'], '-' if sanitize else ':')
1130 if info_dict.get('duration', None) is not None
1131 else None)
1132 info_dict['autonumber'] = int(self.params.get('autonumber_start', 1) - 1 + self._num_downloads)
1133 info_dict['video_autonumber'] = self._num_videos
1134 if info_dict.get('resolution') is None:
1135 info_dict['resolution'] = self.format_resolution(info_dict, default=None)
1136
1137 # For fields playlist_index, playlist_autonumber and autonumber convert all occurrences
1138 # of %(field)s to %(field)0Nd for backward compatibility
1139 field_size_compat_map = {
1140 'playlist_index': number_of_digits(info_dict.get('__last_playlist_index') or 0),
1141 'playlist_autonumber': number_of_digits(info_dict.get('n_entries') or 0),
1142 'autonumber': self.params.get('autonumber_size') or 5,
1143 }
1144
1145 TMPL_DICT = {}
1146 EXTERNAL_FORMAT_RE = re.compile(STR_FORMAT_RE_TMPL.format('[^)]*', f'[{STR_FORMAT_TYPES}ljhqBUDS]'))
1147 MATH_FUNCTIONS = {
1148 '+': float.__add__,
1149 '-': float.__sub__,
1150 }
1151 # Field is of the form key1.key2...
1152 # where keys (except first) can be string, int, slice or "{field, ...}"
1153 FIELD_INNER_RE = r'(?:\w+|%(num)s|%(num)s?(?::%(num)s?){1,2})' % {'num': r'(?:-?\d+)'}
1154 FIELD_RE = r'\w*(?:\.(?:%(inner)s|{%(field)s(?:,%(field)s)*}))*' % {
1155 'inner': FIELD_INNER_RE,
1156 'field': rf'\w*(?:\.{FIELD_INNER_RE})*'
1157 }
1158 MATH_FIELD_RE = rf'(?:{FIELD_RE}|-?{NUMBER_RE})'
1159 MATH_OPERATORS_RE = r'(?:%s)' % '|'.join(map(re.escape, MATH_FUNCTIONS.keys()))
1160 INTERNAL_FORMAT_RE = re.compile(rf'''(?xs)
1161 (?P<negate>-)?
1162 (?P<fields>{FIELD_RE})
1163 (?P<maths>(?:{MATH_OPERATORS_RE}{MATH_FIELD_RE})*)
1164 (?:>(?P<strf_format>.+?))?
1165 (?P<remaining>
1166 (?P<alternate>(?<!\\),[^|&)]+)?
1167 (?:&(?P<replacement>.*?))?
1168 (?:\|(?P<default>.*?))?
1169 )$''')
1170
1171 def _traverse_infodict(fields):
1172 fields = [f for x in re.split(r'\.({.+?})\.?', fields)
1173 for f in ([x] if x.startswith('{') else x.split('.'))]
1174 for i in (0, -1):
1175 if fields and not fields[i]:
1176 fields.pop(i)
1177
1178 for i, f in enumerate(fields):
1179 if not f.startswith('{'):
1180 continue
1181 assert f.endswith('}'), f'No closing brace for {f} in {fields}'
1182 fields[i] = {k: k.split('.') for k in f[1:-1].split(',')}
1183
1184 return traverse_obj(info_dict, fields, is_user_input=True, traverse_string=True)
1185
1186 def get_value(mdict):
1187 # Object traversal
1188 value = _traverse_infodict(mdict['fields'])
1189 # Negative
1190 if mdict['negate']:
1191 value = float_or_none(value)
1192 if value is not None:
1193 value *= -1
1194 # Do maths
1195 offset_key = mdict['maths']
1196 if offset_key:
1197 value = float_or_none(value)
1198 operator = None
1199 while offset_key:
1200 item = re.match(
1201 MATH_FIELD_RE if operator else MATH_OPERATORS_RE,
1202 offset_key).group(0)
1203 offset_key = offset_key[len(item):]
1204 if operator is None:
1205 operator = MATH_FUNCTIONS[item]
1206 continue
1207 item, multiplier = (item[1:], -1) if item[0] == '-' else (item, 1)
1208 offset = float_or_none(item)
1209 if offset is None:
1210 offset = float_or_none(_traverse_infodict(item))
1211 try:
1212 value = operator(value, multiplier * offset)
1213 except (TypeError, ZeroDivisionError):
1214 return None
1215 operator = None
1216 # Datetime formatting
1217 if mdict['strf_format']:
1218 value = strftime_or_none(value, mdict['strf_format'].replace('\\,', ','))
1219
1220 # XXX: Workaround for https://github.com/yt-dlp/yt-dlp/issues/4485
1221 if sanitize and value == '':
1222 value = None
1223 return value
1224
1225 na = self.params.get('outtmpl_na_placeholder', 'NA')
1226
1227 def filename_sanitizer(key, value, restricted=self.params.get('restrictfilenames')):
1228 return sanitize_filename(str(value), restricted=restricted, is_id=(
1229 bool(re.search(r'(^|[_.])id(\.|$)', key))
1230 if 'filename-sanitization' in self.params['compat_opts']
1231 else NO_DEFAULT))
1232
1233 sanitizer = sanitize if callable(sanitize) else filename_sanitizer
1234 sanitize = bool(sanitize)
1235
1236 def _dumpjson_default(obj):
1237 if isinstance(obj, (set, LazyList)):
1238 return list(obj)
1239 return repr(obj)
1240
1241 class _ReplacementFormatter(string.Formatter):
1242 def get_field(self, field_name, args, kwargs):
1243 if field_name.isdigit():
1244 return args[0], -1
1245 raise ValueError('Unsupported field')
1246
1247 replacement_formatter = _ReplacementFormatter()
1248
1249 def create_key(outer_mobj):
1250 if not outer_mobj.group('has_key'):
1251 return outer_mobj.group(0)
1252 key = outer_mobj.group('key')
1253 mobj = re.match(INTERNAL_FORMAT_RE, key)
1254 initial_field = mobj.group('fields') if mobj else ''
1255 value, replacement, default = None, None, na
1256 while mobj:
1257 mobj = mobj.groupdict()
1258 default = mobj['default'] if mobj['default'] is not None else default
1259 value = get_value(mobj)
1260 replacement = mobj['replacement']
1261 if value is None and mobj['alternate']:
1262 mobj = re.match(INTERNAL_FORMAT_RE, mobj['remaining'][1:])
1263 else:
1264 break
1265
1266 fmt = outer_mobj.group('format')
1267 if fmt == 's' and value is not None and key in field_size_compat_map.keys():
1268 fmt = f'0{field_size_compat_map[key]:d}d'
1269
1270 if value is None:
1271 value = default
1272 elif replacement is not None:
1273 try:
1274 value = replacement_formatter.format(replacement, value)
1275 except ValueError:
1276 value = na
1277
1278 flags = outer_mobj.group('conversion') or ''
1279 str_fmt = f'{fmt[:-1]}s'
1280 if fmt[-1] == 'l': # list
1281 delim = '\n' if '#' in flags else ', '
1282 value, fmt = delim.join(map(str, variadic(value, allowed_types=(str, bytes)))), str_fmt
1283 elif fmt[-1] == 'j': # json
1284 value, fmt = json.dumps(
1285 value, default=_dumpjson_default,
1286 indent=4 if '#' in flags else None, ensure_ascii='+' not in flags), str_fmt
1287 elif fmt[-1] == 'h': # html
1288 value, fmt = escapeHTML(str(value)), str_fmt
1289 elif fmt[-1] == 'q': # quoted
1290 value = map(str, variadic(value) if '#' in flags else [value])
1291 value, fmt = ' '.join(map(compat_shlex_quote, value)), str_fmt
1292 elif fmt[-1] == 'B': # bytes
1293 value = f'%{str_fmt}'.encode() % str(value).encode()
1294 value, fmt = value.decode('utf-8', 'ignore'), 's'
1295 elif fmt[-1] == 'U': # unicode normalized
1296 value, fmt = unicodedata.normalize(
1297 # "+" = compatibility equivalence, "#" = NFD
1298 'NF%s%s' % ('K' if '+' in flags else '', 'D' if '#' in flags else 'C'),
1299 value), str_fmt
1300 elif fmt[-1] == 'D': # decimal suffix
1301 num_fmt, fmt = fmt[:-1].replace('#', ''), 's'
1302 value = format_decimal_suffix(value, f'%{num_fmt}f%s' if num_fmt else '%d%s',
1303 factor=1024 if '#' in flags else 1000)
1304 elif fmt[-1] == 'S': # filename sanitization
1305 value, fmt = filename_sanitizer(initial_field, value, restricted='#' in flags), str_fmt
1306 elif fmt[-1] == 'c':
1307 if value:
1308 value = str(value)[0]
1309 else:
1310 fmt = str_fmt
1311 elif fmt[-1] not in 'rs': # numeric
1312 value = float_or_none(value)
1313 if value is None:
1314 value, fmt = default, 's'
1315
1316 if sanitize:
1317 if fmt[-1] == 'r':
1318 # If value is an object, sanitize might convert it to a string
1319 # So we convert it to repr first
1320 value, fmt = repr(value), str_fmt
1321 if fmt[-1] in 'csr':
1322 value = sanitizer(initial_field, value)
1323
1324 key = '%s\0%s' % (key.replace('%', '%\0'), outer_mobj.group('format'))
1325 TMPL_DICT[key] = value
1326 return '{prefix}%({key}){fmt}'.format(key=key, fmt=fmt, prefix=outer_mobj.group('prefix'))
1327
1328 return EXTERNAL_FORMAT_RE.sub(create_key, outtmpl), TMPL_DICT
1329
1330 def evaluate_outtmpl(self, outtmpl, info_dict, *args, **kwargs):
1331 outtmpl, info_dict = self.prepare_outtmpl(outtmpl, info_dict, *args, **kwargs)
1332 return self.escape_outtmpl(outtmpl) % info_dict
1333
1334 def _prepare_filename(self, info_dict, *, outtmpl=None, tmpl_type=None):
1335 assert None in (outtmpl, tmpl_type), 'outtmpl and tmpl_type are mutually exclusive'
1336 if outtmpl is None:
1337 outtmpl = self.params['outtmpl'].get(tmpl_type or 'default', self.params['outtmpl']['default'])
1338 try:
1339 outtmpl = self._outtmpl_expandpath(outtmpl)
1340 filename = self.evaluate_outtmpl(outtmpl, info_dict, True)
1341 if not filename:
1342 return None
1343
1344 if tmpl_type in ('', 'temp'):
1345 final_ext, ext = self.params.get('final_ext'), info_dict.get('ext')
1346 if final_ext and ext and final_ext != ext and filename.endswith(f'.{final_ext}'):
1347 filename = replace_extension(filename, ext, final_ext)
1348 elif tmpl_type:
1349 force_ext = OUTTMPL_TYPES[tmpl_type]
1350 if force_ext:
1351 filename = replace_extension(filename, force_ext, info_dict.get('ext'))
1352
1353 # https://github.com/blackjack4494/youtube-dlc/issues/85
1354 trim_file_name = self.params.get('trim_file_name', False)
1355 if trim_file_name:
1356 no_ext, *ext = filename.rsplit('.', 2)
1357 filename = join_nonempty(no_ext[:trim_file_name], *ext, delim='.')
1358
1359 return filename
1360 except ValueError as err:
1361 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
1362 return None
1363
1364 def prepare_filename(self, info_dict, dir_type='', *, outtmpl=None, warn=False):
1365 """Generate the output filename"""
1366 if outtmpl:
1367 assert not dir_type, 'outtmpl and dir_type are mutually exclusive'
1368 dir_type = None
1369 filename = self._prepare_filename(info_dict, tmpl_type=dir_type, outtmpl=outtmpl)
1370 if not filename and dir_type not in ('', 'temp'):
1371 return ''
1372
1373 if warn:
1374 if not self.params.get('paths'):
1375 pass
1376 elif filename == '-':
1377 self.report_warning('--paths is ignored when an outputting to stdout', only_once=True)
1378 elif os.path.isabs(filename):
1379 self.report_warning('--paths is ignored since an absolute path is given in output template', only_once=True)
1380 if filename == '-' or not filename:
1381 return filename
1382
1383 return self.get_output_path(dir_type, filename)
1384
1385 def _match_entry(self, info_dict, incomplete=False, silent=False):
1386 """Returns None if the file should be downloaded"""
1387 _type = info_dict.get('_type', 'video')
1388 assert incomplete or _type == 'video', 'Only video result can be considered complete'
1389
1390 video_title = info_dict.get('title', info_dict.get('id', 'entry'))
1391
1392 def check_filter():
1393 if _type in ('playlist', 'multi_video'):
1394 return
1395 elif _type in ('url', 'url_transparent') and not try_call(
1396 lambda: self.get_info_extractor(info_dict['ie_key']).is_single_video(info_dict['url'])):
1397 return
1398
1399 if 'title' in info_dict:
1400 # This can happen when we're just evaluating the playlist
1401 title = info_dict['title']
1402 matchtitle = self.params.get('matchtitle', False)
1403 if matchtitle:
1404 if not re.search(matchtitle, title, re.IGNORECASE):
1405 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
1406 rejecttitle = self.params.get('rejecttitle', False)
1407 if rejecttitle:
1408 if re.search(rejecttitle, title, re.IGNORECASE):
1409 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
1410
1411 date = info_dict.get('upload_date')
1412 if date is not None:
1413 dateRange = self.params.get('daterange', DateRange())
1414 if date not in dateRange:
1415 return f'{date_from_str(date).isoformat()} upload date is not in range {dateRange}'
1416 view_count = info_dict.get('view_count')
1417 if view_count is not None:
1418 min_views = self.params.get('min_views')
1419 if min_views is not None and view_count < min_views:
1420 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
1421 max_views = self.params.get('max_views')
1422 if max_views is not None and view_count > max_views:
1423 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
1424 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
1425 return 'Skipping "%s" because it is age restricted' % video_title
1426
1427 match_filter = self.params.get('match_filter')
1428 if match_filter is None:
1429 return None
1430
1431 cancelled = None
1432 try:
1433 try:
1434 ret = match_filter(info_dict, incomplete=incomplete)
1435 except TypeError:
1436 # For backward compatibility
1437 ret = None if incomplete else match_filter(info_dict)
1438 except DownloadCancelled as err:
1439 if err.msg is not NO_DEFAULT:
1440 raise
1441 ret, cancelled = err.msg, err
1442
1443 if ret is NO_DEFAULT:
1444 while True:
1445 filename = self._format_screen(self.prepare_filename(info_dict), self.Styles.FILENAME)
1446 reply = input(self._format_screen(
1447 f'Download "{filename}"? (Y/n): ', self.Styles.EMPHASIS)).lower().strip()
1448 if reply in {'y', ''}:
1449 return None
1450 elif reply == 'n':
1451 if cancelled:
1452 raise type(cancelled)(f'Skipping {video_title}')
1453 return f'Skipping {video_title}'
1454 return ret
1455
1456 if self.in_download_archive(info_dict):
1457 reason = '%s has already been recorded in the archive' % video_title
1458 break_opt, break_err = 'break_on_existing', ExistingVideoReached
1459 else:
1460 try:
1461 reason = check_filter()
1462 except DownloadCancelled as e:
1463 reason, break_opt, break_err = e.msg, 'match_filter', type(e)
1464 else:
1465 break_opt, break_err = 'break_on_reject', RejectedVideoReached
1466 if reason is not None:
1467 if not silent:
1468 self.to_screen('[download] ' + reason)
1469 if self.params.get(break_opt, False):
1470 raise break_err()
1471 return reason
1472
1473 @staticmethod
1474 def add_extra_info(info_dict, extra_info):
1475 '''Set the keys from extra_info in info dict if they are missing'''
1476 for key, value in extra_info.items():
1477 info_dict.setdefault(key, value)
1478
1479 def extract_info(self, url, download=True, ie_key=None, extra_info=None,
1480 process=True, force_generic_extractor=False):
1481 """
1482 Extract and return the information dictionary of the URL
1483
1484 Arguments:
1485 @param url URL to extract
1486
1487 Keyword arguments:
1488 @param download Whether to download videos
1489 @param process Whether to resolve all unresolved references (URLs, playlist items).
1490 Must be True for download to work
1491 @param ie_key Use only the extractor with this key
1492
1493 @param extra_info Dictionary containing the extra values to add to the info (For internal use only)
1494 @force_generic_extractor Force using the generic extractor (Deprecated; use ie_key='Generic')
1495 """
1496
1497 if extra_info is None:
1498 extra_info = {}
1499
1500 if not ie_key and force_generic_extractor:
1501 ie_key = 'Generic'
1502
1503 if ie_key:
1504 ies = {ie_key: self._ies[ie_key]} if ie_key in self._ies else {}
1505 else:
1506 ies = self._ies
1507
1508 for key, ie in ies.items():
1509 if not ie.suitable(url):
1510 continue
1511
1512 if not ie.working():
1513 self.report_warning('The program functionality for this site has been marked as broken, '
1514 'and will probably not work.')
1515
1516 temp_id = ie.get_temp_id(url)
1517 if temp_id is not None and self.in_download_archive({'id': temp_id, 'ie_key': key}):
1518 self.to_screen(f'[{key}] {temp_id}: has already been recorded in the archive')
1519 if self.params.get('break_on_existing', False):
1520 raise ExistingVideoReached()
1521 break
1522 return self.__extract_info(url, self.get_info_extractor(key), download, extra_info, process)
1523 else:
1524 extractors_restricted = self.params.get('allowed_extractors') not in (None, ['default'])
1525 self.report_error(f'No suitable extractor{format_field(ie_key, None, " (%s)")} found for URL {url}',
1526 tb=False if extractors_restricted else None)
1527
1528 def _handle_extraction_exceptions(func):
1529 @functools.wraps(func)
1530 def wrapper(self, *args, **kwargs):
1531 while True:
1532 try:
1533 return func(self, *args, **kwargs)
1534 except (DownloadCancelled, LazyList.IndexError, PagedList.IndexError):
1535 raise
1536 except ReExtractInfo as e:
1537 if e.expected:
1538 self.to_screen(f'{e}; Re-extracting data')
1539 else:
1540 self.to_stderr('\r')
1541 self.report_warning(f'{e}; Re-extracting data')
1542 continue
1543 except GeoRestrictedError as e:
1544 msg = e.msg
1545 if e.countries:
1546 msg += '\nThis video is available in %s.' % ', '.join(
1547 map(ISO3166Utils.short2full, e.countries))
1548 msg += '\nYou might want to use a VPN or a proxy server (with --proxy) to workaround.'
1549 self.report_error(msg)
1550 except ExtractorError as e: # An error we somewhat expected
1551 self.report_error(str(e), e.format_traceback())
1552 except Exception as e:
1553 if self.params.get('ignoreerrors'):
1554 self.report_error(str(e), tb=encode_compat_str(traceback.format_exc()))
1555 else:
1556 raise
1557 break
1558 return wrapper
1559
1560 def _wait_for_video(self, ie_result={}):
1561 if (not self.params.get('wait_for_video')
1562 or ie_result.get('_type', 'video') != 'video'
1563 or ie_result.get('formats') or ie_result.get('url')):
1564 return
1565
1566 format_dur = lambda dur: '%02d:%02d:%02d' % timetuple_from_msec(dur * 1000)[:-1]
1567 last_msg = ''
1568
1569 def progress(msg):
1570 nonlocal last_msg
1571 full_msg = f'{msg}\n'
1572 if not self.params.get('noprogress'):
1573 full_msg = msg + ' ' * (len(last_msg) - len(msg)) + '\r'
1574 elif last_msg:
1575 return
1576 self.to_screen(full_msg, skip_eol=True)
1577 last_msg = msg
1578
1579 min_wait, max_wait = self.params.get('wait_for_video')
1580 diff = try_get(ie_result, lambda x: x['release_timestamp'] - time.time())
1581 if diff is None and ie_result.get('live_status') == 'is_upcoming':
1582 diff = round(random.uniform(min_wait, max_wait) if (max_wait and min_wait) else (max_wait or min_wait), 0)
1583 self.report_warning('Release time of video is not known')
1584 elif ie_result and (diff or 0) <= 0:
1585 self.report_warning('Video should already be available according to extracted info')
1586 diff = min(max(diff or 0, min_wait or 0), max_wait or float('inf'))
1587 self.to_screen(f'[wait] Waiting for {format_dur(diff)} - Press Ctrl+C to try now')
1588
1589 wait_till = time.time() + diff
1590 try:
1591 while True:
1592 diff = wait_till - time.time()
1593 if diff <= 0:
1594 progress('')
1595 raise ReExtractInfo('[wait] Wait period ended', expected=True)
1596 progress(f'[wait] Remaining time until next attempt: {self._format_screen(format_dur(diff), self.Styles.EMPHASIS)}')
1597 time.sleep(1)
1598 except KeyboardInterrupt:
1599 progress('')
1600 raise ReExtractInfo('[wait] Interrupted by user', expected=True)
1601 except BaseException as e:
1602 if not isinstance(e, ReExtractInfo):
1603 self.to_screen('')
1604 raise
1605
1606 @_handle_extraction_exceptions
1607 def __extract_info(self, url, ie, download, extra_info, process):
1608 try:
1609 ie_result = ie.extract(url)
1610 except UserNotLive as e:
1611 if process:
1612 if self.params.get('wait_for_video'):
1613 self.report_warning(e)
1614 self._wait_for_video()
1615 raise
1616 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
1617 self.report_warning(f'Extractor {ie.IE_NAME} returned nothing{bug_reports_message()}')
1618 return
1619 if isinstance(ie_result, list):
1620 # Backwards compatibility: old IE result format
1621 ie_result = {
1622 '_type': 'compat_list',
1623 'entries': ie_result,
1624 }
1625 if extra_info.get('original_url'):
1626 ie_result.setdefault('original_url', extra_info['original_url'])
1627 self.add_default_extra_info(ie_result, ie, url)
1628 if process:
1629 self._wait_for_video(ie_result)
1630 return self.process_ie_result(ie_result, download, extra_info)
1631 else:
1632 return ie_result
1633
1634 def add_default_extra_info(self, ie_result, ie, url):
1635 if url is not None:
1636 self.add_extra_info(ie_result, {
1637 'webpage_url': url,
1638 'original_url': url,
1639 })
1640 webpage_url = ie_result.get('webpage_url')
1641 if webpage_url:
1642 self.add_extra_info(ie_result, {
1643 'webpage_url_basename': url_basename(webpage_url),
1644 'webpage_url_domain': get_domain(webpage_url),
1645 })
1646 if ie is not None:
1647 self.add_extra_info(ie_result, {
1648 'extractor': ie.IE_NAME,
1649 'extractor_key': ie.ie_key(),
1650 })
1651
1652 def process_ie_result(self, ie_result, download=True, extra_info=None):
1653 """
1654 Take the result of the ie(may be modified) and resolve all unresolved
1655 references (URLs, playlist items).
1656
1657 It will also download the videos if 'download'.
1658 Returns the resolved ie_result.
1659 """
1660 if extra_info is None:
1661 extra_info = {}
1662 result_type = ie_result.get('_type', 'video')
1663
1664 if result_type in ('url', 'url_transparent'):
1665 ie_result['url'] = sanitize_url(
1666 ie_result['url'], scheme='http' if self.params.get('prefer_insecure') else 'https')
1667 if ie_result.get('original_url') and not extra_info.get('original_url'):
1668 extra_info = {'original_url': ie_result['original_url'], **extra_info}
1669
1670 extract_flat = self.params.get('extract_flat', False)
1671 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info)
1672 or extract_flat is True):
1673 info_copy = ie_result.copy()
1674 ie = try_get(ie_result.get('ie_key'), self.get_info_extractor)
1675 if ie and not ie_result.get('id'):
1676 info_copy['id'] = ie.get_temp_id(ie_result['url'])
1677 self.add_default_extra_info(info_copy, ie, ie_result['url'])
1678 self.add_extra_info(info_copy, extra_info)
1679 info_copy, _ = self.pre_process(info_copy)
1680 self._fill_common_fields(info_copy, False)
1681 self.__forced_printings(info_copy)
1682 self._raise_pending_errors(info_copy)
1683 if self.params.get('force_write_download_archive', False):
1684 self.record_download_archive(info_copy)
1685 return ie_result
1686
1687 if result_type == 'video':
1688 self.add_extra_info(ie_result, extra_info)
1689 ie_result = self.process_video_result(ie_result, download=download)
1690 self._raise_pending_errors(ie_result)
1691 additional_urls = (ie_result or {}).get('additional_urls')
1692 if additional_urls:
1693 # TODO: Improve MetadataParserPP to allow setting a list
1694 if isinstance(additional_urls, str):
1695 additional_urls = [additional_urls]
1696 self.to_screen(
1697 '[info] %s: %d additional URL(s) requested' % (ie_result['id'], len(additional_urls)))
1698 self.write_debug('Additional URLs: "%s"' % '", "'.join(additional_urls))
1699 ie_result['additional_entries'] = [
1700 self.extract_info(
1701 url, download, extra_info=extra_info,
1702 force_generic_extractor=self.params.get('force_generic_extractor'))
1703 for url in additional_urls
1704 ]
1705 return ie_result
1706 elif result_type == 'url':
1707 # We have to add extra_info to the results because it may be
1708 # contained in a playlist
1709 return self.extract_info(
1710 ie_result['url'], download,
1711 ie_key=ie_result.get('ie_key'),
1712 extra_info=extra_info)
1713 elif result_type == 'url_transparent':
1714 # Use the information from the embedding page
1715 info = self.extract_info(
1716 ie_result['url'], ie_key=ie_result.get('ie_key'),
1717 extra_info=extra_info, download=False, process=False)
1718
1719 # extract_info may return None when ignoreerrors is enabled and
1720 # extraction failed with an error, don't crash and return early
1721 # in this case
1722 if not info:
1723 return info
1724
1725 exempted_fields = {'_type', 'url', 'ie_key'}
1726 if not ie_result.get('section_end') and ie_result.get('section_start') is None:
1727 # For video clips, the id etc of the clip extractor should be used
1728 exempted_fields |= {'id', 'extractor', 'extractor_key'}
1729
1730 new_result = info.copy()
1731 new_result.update(filter_dict(ie_result, lambda k, v: v is not None and k not in exempted_fields))
1732
1733 # Extracted info may not be a video result (i.e.
1734 # info.get('_type', 'video') != video) but rather an url or
1735 # url_transparent. In such cases outer metadata (from ie_result)
1736 # should be propagated to inner one (info). For this to happen
1737 # _type of info should be overridden with url_transparent. This
1738 # fixes issue from https://github.com/ytdl-org/youtube-dl/pull/11163.
1739 if new_result.get('_type') == 'url':
1740 new_result['_type'] = 'url_transparent'
1741
1742 return self.process_ie_result(
1743 new_result, download=download, extra_info=extra_info)
1744 elif result_type in ('playlist', 'multi_video'):
1745 # Protect from infinite recursion due to recursively nested playlists
1746 # (see https://github.com/ytdl-org/youtube-dl/issues/27833)
1747 webpage_url = ie_result.get('webpage_url') # Playlists maynot have webpage_url
1748 if webpage_url and webpage_url in self._playlist_urls:
1749 self.to_screen(
1750 '[download] Skipping already downloaded playlist: %s'
1751 % ie_result.get('title') or ie_result.get('id'))
1752 return
1753
1754 self._playlist_level += 1
1755 self._playlist_urls.add(webpage_url)
1756 self._fill_common_fields(ie_result, False)
1757 self._sanitize_thumbnails(ie_result)
1758 try:
1759 return self.__process_playlist(ie_result, download)
1760 finally:
1761 self._playlist_level -= 1
1762 if not self._playlist_level:
1763 self._playlist_urls.clear()
1764 elif result_type == 'compat_list':
1765 self.report_warning(
1766 'Extractor %s returned a compat_list result. '
1767 'It needs to be updated.' % ie_result.get('extractor'))
1768
1769 def _fixup(r):
1770 self.add_extra_info(r, {
1771 'extractor': ie_result['extractor'],
1772 'webpage_url': ie_result['webpage_url'],
1773 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1774 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1775 'extractor_key': ie_result['extractor_key'],
1776 })
1777 return r
1778 ie_result['entries'] = [
1779 self.process_ie_result(_fixup(r), download, extra_info)
1780 for r in ie_result['entries']
1781 ]
1782 return ie_result
1783 else:
1784 raise Exception('Invalid result type: %s' % result_type)
1785
1786 def _ensure_dir_exists(self, path):
1787 return make_dir(path, self.report_error)
1788
1789 @staticmethod
1790 def _playlist_infodict(ie_result, strict=False, **kwargs):
1791 info = {
1792 'playlist_count': ie_result.get('playlist_count'),
1793 'playlist': ie_result.get('title') or ie_result.get('id'),
1794 'playlist_id': ie_result.get('id'),
1795 'playlist_title': ie_result.get('title'),
1796 'playlist_uploader': ie_result.get('uploader'),
1797 'playlist_uploader_id': ie_result.get('uploader_id'),
1798 **kwargs,
1799 }
1800 if strict:
1801 return info
1802 if ie_result.get('webpage_url'):
1803 info.update({
1804 'webpage_url': ie_result['webpage_url'],
1805 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1806 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1807 })
1808 return {
1809 **info,
1810 'playlist_index': 0,
1811 '__last_playlist_index': max(ie_result.get('requested_entries') or (0, 0)),
1812 'extractor': ie_result['extractor'],
1813 'extractor_key': ie_result['extractor_key'],
1814 }
1815
1816 def __process_playlist(self, ie_result, download):
1817 """Process each entry in the playlist"""
1818 assert ie_result['_type'] in ('playlist', 'multi_video')
1819
1820 common_info = self._playlist_infodict(ie_result, strict=True)
1821 title = common_info.get('playlist') or '<Untitled>'
1822 if self._match_entry(common_info, incomplete=True) is not None:
1823 return
1824 self.to_screen(f'[download] Downloading {ie_result["_type"]}: {title}')
1825
1826 all_entries = PlaylistEntries(self, ie_result)
1827 entries = orderedSet(all_entries.get_requested_items(), lazy=True)
1828
1829 lazy = self.params.get('lazy_playlist')
1830 if lazy:
1831 resolved_entries, n_entries = [], 'N/A'
1832 ie_result['requested_entries'], ie_result['entries'] = None, None
1833 else:
1834 entries = resolved_entries = list(entries)
1835 n_entries = len(resolved_entries)
1836 ie_result['requested_entries'], ie_result['entries'] = tuple(zip(*resolved_entries)) or ([], [])
1837 if not ie_result.get('playlist_count'):
1838 # Better to do this after potentially exhausting entries
1839 ie_result['playlist_count'] = all_entries.get_full_count()
1840
1841 extra = self._playlist_infodict(ie_result, n_entries=int_or_none(n_entries))
1842 ie_copy = collections.ChainMap(ie_result, extra)
1843
1844 _infojson_written = False
1845 write_playlist_files = self.params.get('allow_playlist_files', True)
1846 if write_playlist_files and self.params.get('list_thumbnails'):
1847 self.list_thumbnails(ie_result)
1848 if write_playlist_files and not self.params.get('simulate'):
1849 _infojson_written = self._write_info_json(
1850 'playlist', ie_result, self.prepare_filename(ie_copy, 'pl_infojson'))
1851 if _infojson_written is None:
1852 return
1853 if self._write_description('playlist', ie_result,
1854 self.prepare_filename(ie_copy, 'pl_description')) is None:
1855 return
1856 # TODO: This should be passed to ThumbnailsConvertor if necessary
1857 self._write_thumbnails('playlist', ie_result, self.prepare_filename(ie_copy, 'pl_thumbnail'))
1858
1859 if lazy:
1860 if self.params.get('playlistreverse') or self.params.get('playlistrandom'):
1861 self.report_warning('playlistreverse and playlistrandom are not supported with lazy_playlist', only_once=True)
1862 elif self.params.get('playlistreverse'):
1863 entries.reverse()
1864 elif self.params.get('playlistrandom'):
1865 random.shuffle(entries)
1866
1867 self.to_screen(f'[{ie_result["extractor"]}] Playlist {title}: Downloading {n_entries} items'
1868 f'{format_field(ie_result, "playlist_count", " of %s")}')
1869
1870 keep_resolved_entries = self.params.get('extract_flat') != 'discard'
1871 if self.params.get('extract_flat') == 'discard_in_playlist':
1872 keep_resolved_entries = ie_result['_type'] != 'playlist'
1873 if keep_resolved_entries:
1874 self.write_debug('The information of all playlist entries will be held in memory')
1875
1876 failures = 0
1877 max_failures = self.params.get('skip_playlist_after_errors') or float('inf')
1878 for i, (playlist_index, entry) in enumerate(entries):
1879 if lazy:
1880 resolved_entries.append((playlist_index, entry))
1881 if not entry:
1882 continue
1883
1884 entry['__x_forwarded_for_ip'] = ie_result.get('__x_forwarded_for_ip')
1885 if not lazy and 'playlist-index' in self.params.get('compat_opts', []):
1886 playlist_index = ie_result['requested_entries'][i]
1887
1888 entry_copy = collections.ChainMap(entry, {
1889 **common_info,
1890 'n_entries': int_or_none(n_entries),
1891 'playlist_index': playlist_index,
1892 'playlist_autonumber': i + 1,
1893 })
1894
1895 if self._match_entry(entry_copy, incomplete=True) is not None:
1896 # For compatabilty with youtube-dl. See https://github.com/yt-dlp/yt-dlp/issues/4369
1897 resolved_entries[i] = (playlist_index, NO_DEFAULT)
1898 continue
1899
1900 self.to_screen('[download] Downloading item %s of %s' % (
1901 self._format_screen(i + 1, self.Styles.ID), self._format_screen(n_entries, self.Styles.EMPHASIS)))
1902
1903 entry_result = self.__process_iterable_entry(entry, download, collections.ChainMap({
1904 'playlist_index': playlist_index,
1905 'playlist_autonumber': i + 1,
1906 }, extra))
1907 if not entry_result:
1908 failures += 1
1909 if failures >= max_failures:
1910 self.report_error(
1911 f'Skipping the remaining entries in playlist "{title}" since {failures} items failed extraction')
1912 break
1913 if keep_resolved_entries:
1914 resolved_entries[i] = (playlist_index, entry_result)
1915
1916 # Update with processed data
1917 ie_result['entries'] = [e for _, e in resolved_entries if e is not NO_DEFAULT]
1918 ie_result['requested_entries'] = [i for i, e in resolved_entries if e is not NO_DEFAULT]
1919 if ie_result['requested_entries'] == try_call(lambda: list(range(1, ie_result['playlist_count'] + 1))):
1920 # Do not set for full playlist
1921 ie_result.pop('requested_entries')
1922
1923 # Write the updated info to json
1924 if _infojson_written is True and self._write_info_json(
1925 'updated playlist', ie_result,
1926 self.prepare_filename(ie_copy, 'pl_infojson'), overwrite=True) is None:
1927 return
1928
1929 ie_result = self.run_all_pps('playlist', ie_result)
1930 self.to_screen(f'[download] Finished downloading playlist: {title}')
1931 return ie_result
1932
1933 @_handle_extraction_exceptions
1934 def __process_iterable_entry(self, entry, download, extra_info):
1935 return self.process_ie_result(
1936 entry, download=download, extra_info=extra_info)
1937
1938 def _build_format_filter(self, filter_spec):
1939 " Returns a function to filter the formats according to the filter_spec "
1940
1941 OPERATORS = {
1942 '<': operator.lt,
1943 '<=': operator.le,
1944 '>': operator.gt,
1945 '>=': operator.ge,
1946 '=': operator.eq,
1947 '!=': operator.ne,
1948 }
1949 operator_rex = re.compile(r'''(?x)\s*
1950 (?P<key>[\w.-]+)\s*
1951 (?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
1952 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)\s*
1953 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
1954 m = operator_rex.fullmatch(filter_spec)
1955 if m:
1956 try:
1957 comparison_value = int(m.group('value'))
1958 except ValueError:
1959 comparison_value = parse_filesize(m.group('value'))
1960 if comparison_value is None:
1961 comparison_value = parse_filesize(m.group('value') + 'B')
1962 if comparison_value is None:
1963 raise ValueError(
1964 'Invalid value %r in format specification %r' % (
1965 m.group('value'), filter_spec))
1966 op = OPERATORS[m.group('op')]
1967
1968 if not m:
1969 STR_OPERATORS = {
1970 '=': operator.eq,
1971 '^=': lambda attr, value: attr.startswith(value),
1972 '$=': lambda attr, value: attr.endswith(value),
1973 '*=': lambda attr, value: value in attr,
1974 '~=': lambda attr, value: value.search(attr) is not None
1975 }
1976 str_operator_rex = re.compile(r'''(?x)\s*
1977 (?P<key>[a-zA-Z0-9._-]+)\s*
1978 (?P<negation>!\s*)?(?P<op>%s)\s*(?P<none_inclusive>\?\s*)?
1979 (?P<quote>["'])?
1980 (?P<value>(?(quote)(?:(?!(?P=quote))[^\\]|\\.)+|[\w.-]+))
1981 (?(quote)(?P=quote))\s*
1982 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
1983 m = str_operator_rex.fullmatch(filter_spec)
1984 if m:
1985 if m.group('op') == '~=':
1986 comparison_value = re.compile(m.group('value'))
1987 else:
1988 comparison_value = re.sub(r'''\\([\\"'])''', r'\1', m.group('value'))
1989 str_op = STR_OPERATORS[m.group('op')]
1990 if m.group('negation'):
1991 op = lambda attr, value: not str_op(attr, value)
1992 else:
1993 op = str_op
1994
1995 if not m:
1996 raise SyntaxError('Invalid filter specification %r' % filter_spec)
1997
1998 def _filter(f):
1999 actual_value = f.get(m.group('key'))
2000 if actual_value is None:
2001 return m.group('none_inclusive')
2002 return op(actual_value, comparison_value)
2003 return _filter
2004
2005 def _check_formats(self, formats):
2006 for f in formats:
2007 self.to_screen('[info] Testing format %s' % f['format_id'])
2008 path = self.get_output_path('temp')
2009 if not self._ensure_dir_exists(f'{path}/'):
2010 continue
2011 temp_file = tempfile.NamedTemporaryFile(suffix='.tmp', delete=False, dir=path or None)
2012 temp_file.close()
2013 try:
2014 success, _ = self.dl(temp_file.name, f, test=True)
2015 except (DownloadError, OSError, ValueError) + network_exceptions:
2016 success = False
2017 finally:
2018 if os.path.exists(temp_file.name):
2019 try:
2020 os.remove(temp_file.name)
2021 except OSError:
2022 self.report_warning('Unable to delete temporary file "%s"' % temp_file.name)
2023 if success:
2024 yield f
2025 else:
2026 self.to_screen('[info] Unable to download format %s. Skipping...' % f['format_id'])
2027
2028 def _default_format_spec(self, info_dict, download=True):
2029
2030 def can_merge():
2031 merger = FFmpegMergerPP(self)
2032 return merger.available and merger.can_merge()
2033
2034 prefer_best = (
2035 not self.params.get('simulate')
2036 and download
2037 and (
2038 not can_merge()
2039 or info_dict.get('is_live') and not self.params.get('live_from_start')
2040 or self.params['outtmpl']['default'] == '-'))
2041 compat = (
2042 prefer_best
2043 or self.params.get('allow_multiple_audio_streams', False)
2044 or 'format-spec' in self.params['compat_opts'])
2045
2046 return (
2047 'best/bestvideo+bestaudio' if prefer_best
2048 else 'bestvideo*+bestaudio/best' if not compat
2049 else 'bestvideo+bestaudio/best')
2050
2051 def build_format_selector(self, format_spec):
2052 def syntax_error(note, start):
2053 message = (
2054 'Invalid format specification: '
2055 '{}\n\t{}\n\t{}^'.format(note, format_spec, ' ' * start[1]))
2056 return SyntaxError(message)
2057
2058 PICKFIRST = 'PICKFIRST'
2059 MERGE = 'MERGE'
2060 SINGLE = 'SINGLE'
2061 GROUP = 'GROUP'
2062 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
2063
2064 allow_multiple_streams = {'audio': self.params.get('allow_multiple_audio_streams', False),
2065 'video': self.params.get('allow_multiple_video_streams', False)}
2066
2067 check_formats = self.params.get('check_formats') == 'selected'
2068
2069 def _parse_filter(tokens):
2070 filter_parts = []
2071 for type, string_, start, _, _ in tokens:
2072 if type == tokenize.OP and string_ == ']':
2073 return ''.join(filter_parts)
2074 else:
2075 filter_parts.append(string_)
2076
2077 def _remove_unused_ops(tokens):
2078 # Remove operators that we don't use and join them with the surrounding strings.
2079 # E.g. 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
2080 ALLOWED_OPS = ('/', '+', ',', '(', ')')
2081 last_string, last_start, last_end, last_line = None, None, None, None
2082 for type, string_, start, end, line in tokens:
2083 if type == tokenize.OP and string_ == '[':
2084 if last_string:
2085 yield tokenize.NAME, last_string, last_start, last_end, last_line
2086 last_string = None
2087 yield type, string_, start, end, line
2088 # everything inside brackets will be handled by _parse_filter
2089 for type, string_, start, end, line in tokens:
2090 yield type, string_, start, end, line
2091 if type == tokenize.OP and string_ == ']':
2092 break
2093 elif type == tokenize.OP and string_ in ALLOWED_OPS:
2094 if last_string:
2095 yield tokenize.NAME, last_string, last_start, last_end, last_line
2096 last_string = None
2097 yield type, string_, start, end, line
2098 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
2099 if not last_string:
2100 last_string = string_
2101 last_start = start
2102 last_end = end
2103 else:
2104 last_string += string_
2105 if last_string:
2106 yield tokenize.NAME, last_string, last_start, last_end, last_line
2107
2108 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
2109 selectors = []
2110 current_selector = None
2111 for type, string_, start, _, _ in tokens:
2112 # ENCODING is only defined in python 3.x
2113 if type == getattr(tokenize, 'ENCODING', None):
2114 continue
2115 elif type in [tokenize.NAME, tokenize.NUMBER]:
2116 current_selector = FormatSelector(SINGLE, string_, [])
2117 elif type == tokenize.OP:
2118 if string_ == ')':
2119 if not inside_group:
2120 # ')' will be handled by the parentheses group
2121 tokens.restore_last_token()
2122 break
2123 elif inside_merge and string_ in ['/', ',']:
2124 tokens.restore_last_token()
2125 break
2126 elif inside_choice and string_ == ',':
2127 tokens.restore_last_token()
2128 break
2129 elif string_ == ',':
2130 if not current_selector:
2131 raise syntax_error('"," must follow a format selector', start)
2132 selectors.append(current_selector)
2133 current_selector = None
2134 elif string_ == '/':
2135 if not current_selector:
2136 raise syntax_error('"/" must follow a format selector', start)
2137 first_choice = current_selector
2138 second_choice = _parse_format_selection(tokens, inside_choice=True)
2139 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
2140 elif string_ == '[':
2141 if not current_selector:
2142 current_selector = FormatSelector(SINGLE, 'best', [])
2143 format_filter = _parse_filter(tokens)
2144 current_selector.filters.append(format_filter)
2145 elif string_ == '(':
2146 if current_selector:
2147 raise syntax_error('Unexpected "("', start)
2148 group = _parse_format_selection(tokens, inside_group=True)
2149 current_selector = FormatSelector(GROUP, group, [])
2150 elif string_ == '+':
2151 if not current_selector:
2152 raise syntax_error('Unexpected "+"', start)
2153 selector_1 = current_selector
2154 selector_2 = _parse_format_selection(tokens, inside_merge=True)
2155 if not selector_2:
2156 raise syntax_error('Expected a selector', start)
2157 current_selector = FormatSelector(MERGE, (selector_1, selector_2), [])
2158 else:
2159 raise syntax_error(f'Operator not recognized: "{string_}"', start)
2160 elif type == tokenize.ENDMARKER:
2161 break
2162 if current_selector:
2163 selectors.append(current_selector)
2164 return selectors
2165
2166 def _merge(formats_pair):
2167 format_1, format_2 = formats_pair
2168
2169 formats_info = []
2170 formats_info.extend(format_1.get('requested_formats', (format_1,)))
2171 formats_info.extend(format_2.get('requested_formats', (format_2,)))
2172
2173 if not allow_multiple_streams['video'] or not allow_multiple_streams['audio']:
2174 get_no_more = {'video': False, 'audio': False}
2175 for (i, fmt_info) in enumerate(formats_info):
2176 if fmt_info.get('acodec') == fmt_info.get('vcodec') == 'none':
2177 formats_info.pop(i)
2178 continue
2179 for aud_vid in ['audio', 'video']:
2180 if not allow_multiple_streams[aud_vid] and fmt_info.get(aud_vid[0] + 'codec') != 'none':
2181 if get_no_more[aud_vid]:
2182 formats_info.pop(i)
2183 break
2184 get_no_more[aud_vid] = True
2185
2186 if len(formats_info) == 1:
2187 return formats_info[0]
2188
2189 video_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('vcodec') != 'none']
2190 audio_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('acodec') != 'none']
2191
2192 the_only_video = video_fmts[0] if len(video_fmts) == 1 else None
2193 the_only_audio = audio_fmts[0] if len(audio_fmts) == 1 else None
2194
2195 output_ext = get_compatible_ext(
2196 vcodecs=[f.get('vcodec') for f in video_fmts],
2197 acodecs=[f.get('acodec') for f in audio_fmts],
2198 vexts=[f['ext'] for f in video_fmts],
2199 aexts=[f['ext'] for f in audio_fmts],
2200 preferences=(try_call(lambda: self.params['merge_output_format'].split('/'))
2201 or self.params.get('prefer_free_formats') and ('webm', 'mkv')))
2202
2203 filtered = lambda *keys: filter(None, (traverse_obj(fmt, *keys) for fmt in formats_info))
2204
2205 new_dict = {
2206 'requested_formats': formats_info,
2207 'format': '+'.join(filtered('format')),
2208 'format_id': '+'.join(filtered('format_id')),
2209 'ext': output_ext,
2210 'protocol': '+'.join(map(determine_protocol, formats_info)),
2211 'language': '+'.join(orderedSet(filtered('language'))) or None,
2212 'format_note': '+'.join(orderedSet(filtered('format_note'))) or None,
2213 'filesize_approx': sum(filtered('filesize', 'filesize_approx')) or None,
2214 'tbr': sum(filtered('tbr', 'vbr', 'abr')),
2215 }
2216
2217 if the_only_video:
2218 new_dict.update({
2219 'width': the_only_video.get('width'),
2220 'height': the_only_video.get('height'),
2221 'resolution': the_only_video.get('resolution') or self.format_resolution(the_only_video),
2222 'fps': the_only_video.get('fps'),
2223 'dynamic_range': the_only_video.get('dynamic_range'),
2224 'vcodec': the_only_video.get('vcodec'),
2225 'vbr': the_only_video.get('vbr'),
2226 'stretched_ratio': the_only_video.get('stretched_ratio'),
2227 'aspect_ratio': the_only_video.get('aspect_ratio'),
2228 })
2229
2230 if the_only_audio:
2231 new_dict.update({
2232 'acodec': the_only_audio.get('acodec'),
2233 'abr': the_only_audio.get('abr'),
2234 'asr': the_only_audio.get('asr'),
2235 'audio_channels': the_only_audio.get('audio_channels')
2236 })
2237
2238 return new_dict
2239
2240 def _check_formats(formats):
2241 if not check_formats:
2242 yield from formats
2243 return
2244 yield from self._check_formats(formats)
2245
2246 def _build_selector_function(selector):
2247 if isinstance(selector, list): # ,
2248 fs = [_build_selector_function(s) for s in selector]
2249
2250 def selector_function(ctx):
2251 for f in fs:
2252 yield from f(ctx)
2253 return selector_function
2254
2255 elif selector.type == GROUP: # ()
2256 selector_function = _build_selector_function(selector.selector)
2257
2258 elif selector.type == PICKFIRST: # /
2259 fs = [_build_selector_function(s) for s in selector.selector]
2260
2261 def selector_function(ctx):
2262 for f in fs:
2263 picked_formats = list(f(ctx))
2264 if picked_formats:
2265 return picked_formats
2266 return []
2267
2268 elif selector.type == MERGE: # +
2269 selector_1, selector_2 = map(_build_selector_function, selector.selector)
2270
2271 def selector_function(ctx):
2272 for pair in itertools.product(selector_1(ctx), selector_2(ctx)):
2273 yield _merge(pair)
2274
2275 elif selector.type == SINGLE: # atom
2276 format_spec = selector.selector or 'best'
2277
2278 # TODO: Add allvideo, allaudio etc by generalizing the code with best/worst selector
2279 if format_spec == 'all':
2280 def selector_function(ctx):
2281 yield from _check_formats(ctx['formats'][::-1])
2282 elif format_spec == 'mergeall':
2283 def selector_function(ctx):
2284 formats = list(_check_formats(
2285 f for f in ctx['formats'] if f.get('vcodec') != 'none' or f.get('acodec') != 'none'))
2286 if not formats:
2287 return
2288 merged_format = formats[-1]
2289 for f in formats[-2::-1]:
2290 merged_format = _merge((merged_format, f))
2291 yield merged_format
2292
2293 else:
2294 format_fallback, seperate_fallback, format_reverse, format_idx = False, None, True, 1
2295 mobj = re.match(
2296 r'(?P<bw>best|worst|b|w)(?P<type>video|audio|v|a)?(?P<mod>\*)?(?:\.(?P<n>[1-9]\d*))?$',
2297 format_spec)
2298 if mobj is not None:
2299 format_idx = int_or_none(mobj.group('n'), default=1)
2300 format_reverse = mobj.group('bw')[0] == 'b'
2301 format_type = (mobj.group('type') or [None])[0]
2302 not_format_type = {'v': 'a', 'a': 'v'}.get(format_type)
2303 format_modified = mobj.group('mod') is not None
2304
2305 format_fallback = not format_type and not format_modified # for b, w
2306 _filter_f = (
2307 (lambda f: f.get('%scodec' % format_type) != 'none')
2308 if format_type and format_modified # bv*, ba*, wv*, wa*
2309 else (lambda f: f.get('%scodec' % not_format_type) == 'none')
2310 if format_type # bv, ba, wv, wa
2311 else (lambda f: f.get('vcodec') != 'none' and f.get('acodec') != 'none')
2312 if not format_modified # b, w
2313 else lambda f: True) # b*, w*
2314 filter_f = lambda f: _filter_f(f) and (
2315 f.get('vcodec') != 'none' or f.get('acodec') != 'none')
2316 else:
2317 if format_spec in self._format_selection_exts['audio']:
2318 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none'
2319 elif format_spec in self._format_selection_exts['video']:
2320 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none' and f.get('vcodec') != 'none'
2321 seperate_fallback = lambda f: f.get('ext') == format_spec and f.get('vcodec') != 'none'
2322 elif format_spec in self._format_selection_exts['storyboards']:
2323 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') == 'none' and f.get('vcodec') == 'none'
2324 else:
2325 filter_f = lambda f: f.get('format_id') == format_spec # id
2326
2327 def selector_function(ctx):
2328 formats = list(ctx['formats'])
2329 matches = list(filter(filter_f, formats)) if filter_f is not None else formats
2330 if not matches:
2331 if format_fallback and ctx['incomplete_formats']:
2332 # for extractors with incomplete formats (audio only (soundcloud)
2333 # or video only (imgur)) best/worst will fallback to
2334 # best/worst {video,audio}-only format
2335 matches = formats
2336 elif seperate_fallback and not ctx['has_merged_format']:
2337 # for compatibility with youtube-dl when there is no pre-merged format
2338 matches = list(filter(seperate_fallback, formats))
2339 matches = LazyList(_check_formats(matches[::-1 if format_reverse else 1]))
2340 try:
2341 yield matches[format_idx - 1]
2342 except LazyList.IndexError:
2343 return
2344
2345 filters = [self._build_format_filter(f) for f in selector.filters]
2346
2347 def final_selector(ctx):
2348 ctx_copy = dict(ctx)
2349 for _filter in filters:
2350 ctx_copy['formats'] = list(filter(_filter, ctx_copy['formats']))
2351 return selector_function(ctx_copy)
2352 return final_selector
2353
2354 stream = io.BytesIO(format_spec.encode())
2355 try:
2356 tokens = list(_remove_unused_ops(tokenize.tokenize(stream.readline)))
2357 except tokenize.TokenError:
2358 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
2359
2360 class TokenIterator:
2361 def __init__(self, tokens):
2362 self.tokens = tokens
2363 self.counter = 0
2364
2365 def __iter__(self):
2366 return self
2367
2368 def __next__(self):
2369 if self.counter >= len(self.tokens):
2370 raise StopIteration()
2371 value = self.tokens[self.counter]
2372 self.counter += 1
2373 return value
2374
2375 next = __next__
2376
2377 def restore_last_token(self):
2378 self.counter -= 1
2379
2380 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
2381 return _build_selector_function(parsed_selector)
2382
2383 def _calc_headers(self, info_dict):
2384 res = merge_headers(self.params['http_headers'], info_dict.get('http_headers') or {})
2385
2386 cookies = self._calc_cookies(info_dict['url'])
2387 if cookies:
2388 res['Cookie'] = cookies
2389
2390 if 'X-Forwarded-For' not in res:
2391 x_forwarded_for_ip = info_dict.get('__x_forwarded_for_ip')
2392 if x_forwarded_for_ip:
2393 res['X-Forwarded-For'] = x_forwarded_for_ip
2394
2395 return res
2396
2397 def _calc_cookies(self, url):
2398 pr = sanitized_Request(url)
2399 self.cookiejar.add_cookie_header(pr)
2400 return pr.get_header('Cookie')
2401
2402 def _sort_thumbnails(self, thumbnails):
2403 thumbnails.sort(key=lambda t: (
2404 t.get('preference') if t.get('preference') is not None else -1,
2405 t.get('width') if t.get('width') is not None else -1,
2406 t.get('height') if t.get('height') is not None else -1,
2407 t.get('id') if t.get('id') is not None else '',
2408 t.get('url')))
2409
2410 def _sanitize_thumbnails(self, info_dict):
2411 thumbnails = info_dict.get('thumbnails')
2412 if thumbnails is None:
2413 thumbnail = info_dict.get('thumbnail')
2414 if thumbnail:
2415 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
2416 if not thumbnails:
2417 return
2418
2419 def check_thumbnails(thumbnails):
2420 for t in thumbnails:
2421 self.to_screen(f'[info] Testing thumbnail {t["id"]}')
2422 try:
2423 self.urlopen(HEADRequest(t['url']))
2424 except network_exceptions as err:
2425 self.to_screen(f'[info] Unable to connect to thumbnail {t["id"]} URL {t["url"]!r} - {err}. Skipping...')
2426 continue
2427 yield t
2428
2429 self._sort_thumbnails(thumbnails)
2430 for i, t in enumerate(thumbnails):
2431 if t.get('id') is None:
2432 t['id'] = '%d' % i
2433 if t.get('width') and t.get('height'):
2434 t['resolution'] = '%dx%d' % (t['width'], t['height'])
2435 t['url'] = sanitize_url(t['url'])
2436
2437 if self.params.get('check_formats') is True:
2438 info_dict['thumbnails'] = LazyList(check_thumbnails(thumbnails[::-1]), reverse=True)
2439 else:
2440 info_dict['thumbnails'] = thumbnails
2441
2442 def _fill_common_fields(self, info_dict, final=True):
2443 # TODO: move sanitization here
2444 if final:
2445 title = info_dict['fulltitle'] = info_dict.get('title')
2446 if not title:
2447 if title == '':
2448 self.write_debug('Extractor gave empty title. Creating a generic title')
2449 else:
2450 self.report_warning('Extractor failed to obtain "title". Creating a generic title instead')
2451 info_dict['title'] = f'{info_dict["extractor"].replace(":", "-")} video #{info_dict["id"]}'
2452
2453 if info_dict.get('duration') is not None:
2454 info_dict['duration_string'] = formatSeconds(info_dict['duration'])
2455
2456 for ts_key, date_key in (
2457 ('timestamp', 'upload_date'),
2458 ('release_timestamp', 'release_date'),
2459 ('modified_timestamp', 'modified_date'),
2460 ):
2461 if info_dict.get(date_key) is None and info_dict.get(ts_key) is not None:
2462 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
2463 # see http://bugs.python.org/issue1646728)
2464 with contextlib.suppress(ValueError, OverflowError, OSError):
2465 upload_date = datetime.datetime.utcfromtimestamp(info_dict[ts_key])
2466 info_dict[date_key] = upload_date.strftime('%Y%m%d')
2467
2468 live_keys = ('is_live', 'was_live')
2469 live_status = info_dict.get('live_status')
2470 if live_status is None:
2471 for key in live_keys:
2472 if info_dict.get(key) is False:
2473 continue
2474 if info_dict.get(key):
2475 live_status = key
2476 break
2477 if all(info_dict.get(key) is False for key in live_keys):
2478 live_status = 'not_live'
2479 if live_status:
2480 info_dict['live_status'] = live_status
2481 for key in live_keys:
2482 if info_dict.get(key) is None:
2483 info_dict[key] = (live_status == key)
2484 if live_status == 'post_live':
2485 info_dict['was_live'] = True
2486
2487 # Auto generate title fields corresponding to the *_number fields when missing
2488 # in order to always have clean titles. This is very common for TV series.
2489 for field in ('chapter', 'season', 'episode'):
2490 if final and info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
2491 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
2492
2493 def _raise_pending_errors(self, info):
2494 err = info.pop('__pending_error', None)
2495 if err:
2496 self.report_error(err, tb=False)
2497
2498 def sort_formats(self, info_dict):
2499 formats = self._get_formats(info_dict)
2500 formats.sort(key=FormatSorter(
2501 self, info_dict.get('_format_sort_fields') or []).calculate_preference)
2502
2503 def process_video_result(self, info_dict, download=True):
2504 assert info_dict.get('_type', 'video') == 'video'
2505 self._num_videos += 1
2506
2507 if 'id' not in info_dict:
2508 raise ExtractorError('Missing "id" field in extractor result', ie=info_dict['extractor'])
2509 elif not info_dict.get('id'):
2510 raise ExtractorError('Extractor failed to obtain "id"', ie=info_dict['extractor'])
2511
2512 def report_force_conversion(field, field_not, conversion):
2513 self.report_warning(
2514 '"%s" field is not %s - forcing %s conversion, there is an error in extractor'
2515 % (field, field_not, conversion))
2516
2517 def sanitize_string_field(info, string_field):
2518 field = info.get(string_field)
2519 if field is None or isinstance(field, str):
2520 return
2521 report_force_conversion(string_field, 'a string', 'string')
2522 info[string_field] = str(field)
2523
2524 def sanitize_numeric_fields(info):
2525 for numeric_field in self._NUMERIC_FIELDS:
2526 field = info.get(numeric_field)
2527 if field is None or isinstance(field, (int, float)):
2528 continue
2529 report_force_conversion(numeric_field, 'numeric', 'int')
2530 info[numeric_field] = int_or_none(field)
2531
2532 sanitize_string_field(info_dict, 'id')
2533 sanitize_numeric_fields(info_dict)
2534 if info_dict.get('section_end') and info_dict.get('section_start') is not None:
2535 info_dict['duration'] = round(info_dict['section_end'] - info_dict['section_start'], 3)
2536 if (info_dict.get('duration') or 0) <= 0 and info_dict.pop('duration', None):
2537 self.report_warning('"duration" field is negative, there is an error in extractor')
2538
2539 chapters = info_dict.get('chapters') or []
2540 if chapters and chapters[0].get('start_time'):
2541 chapters.insert(0, {'start_time': 0})
2542
2543 dummy_chapter = {'end_time': 0, 'start_time': info_dict.get('duration')}
2544 for idx, (prev, current, next_) in enumerate(zip(
2545 (dummy_chapter, *chapters), chapters, (*chapters[1:], dummy_chapter)), 1):
2546 if current.get('start_time') is None:
2547 current['start_time'] = prev.get('end_time')
2548 if not current.get('end_time'):
2549 current['end_time'] = next_.get('start_time')
2550 if not current.get('title'):
2551 current['title'] = f'<Untitled Chapter {idx}>'
2552
2553 if 'playlist' not in info_dict:
2554 # It isn't part of a playlist
2555 info_dict['playlist'] = None
2556 info_dict['playlist_index'] = None
2557
2558 self._sanitize_thumbnails(info_dict)
2559
2560 thumbnail = info_dict.get('thumbnail')
2561 thumbnails = info_dict.get('thumbnails')
2562 if thumbnail:
2563 info_dict['thumbnail'] = sanitize_url(thumbnail)
2564 elif thumbnails:
2565 info_dict['thumbnail'] = thumbnails[-1]['url']
2566
2567 if info_dict.get('display_id') is None and 'id' in info_dict:
2568 info_dict['display_id'] = info_dict['id']
2569
2570 self._fill_common_fields(info_dict)
2571
2572 for cc_kind in ('subtitles', 'automatic_captions'):
2573 cc = info_dict.get(cc_kind)
2574 if cc:
2575 for _, subtitle in cc.items():
2576 for subtitle_format in subtitle:
2577 if subtitle_format.get('url'):
2578 subtitle_format['url'] = sanitize_url(subtitle_format['url'])
2579 if subtitle_format.get('ext') is None:
2580 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
2581
2582 automatic_captions = info_dict.get('automatic_captions')
2583 subtitles = info_dict.get('subtitles')
2584
2585 info_dict['requested_subtitles'] = self.process_subtitles(
2586 info_dict['id'], subtitles, automatic_captions)
2587
2588 formats = self._get_formats(info_dict)
2589
2590 # Backward compatibility with InfoExtractor._sort_formats
2591 field_preference = (formats or [{}])[0].pop('__sort_fields', None)
2592 if field_preference:
2593 info_dict['_format_sort_fields'] = field_preference
2594
2595 # or None ensures --clean-infojson removes it
2596 info_dict['_has_drm'] = any(f.get('has_drm') for f in formats) or None
2597 if not self.params.get('allow_unplayable_formats'):
2598 formats = [f for f in formats if not f.get('has_drm')]
2599
2600 if formats and all(f.get('acodec') == f.get('vcodec') == 'none' for f in formats):
2601 self.report_warning(
2602 f'{"This video is DRM protected and " if info_dict["_has_drm"] else ""}'
2603 'only images are available for download. Use --list-formats to see them'.capitalize())
2604
2605 get_from_start = not info_dict.get('is_live') or bool(self.params.get('live_from_start'))
2606 if not get_from_start:
2607 info_dict['title'] += ' ' + datetime.datetime.now().strftime('%Y-%m-%d %H:%M')
2608 if info_dict.get('is_live') and formats:
2609 formats = [f for f in formats if bool(f.get('is_from_start')) == get_from_start]
2610 if get_from_start and not formats:
2611 self.raise_no_formats(info_dict, msg=(
2612 '--live-from-start is passed, but there are no formats that can be downloaded from the start. '
2613 'If you want to download from the current time, use --no-live-from-start'))
2614
2615 def is_wellformed(f):
2616 url = f.get('url')
2617 if not url:
2618 self.report_warning(
2619 '"url" field is missing or empty - skipping format, '
2620 'there is an error in extractor')
2621 return False
2622 if isinstance(url, bytes):
2623 sanitize_string_field(f, 'url')
2624 return True
2625
2626 # Filter out malformed formats for better extraction robustness
2627 formats = list(filter(is_wellformed, formats or []))
2628
2629 if not formats:
2630 self.raise_no_formats(info_dict)
2631
2632 for format in formats:
2633 sanitize_string_field(format, 'format_id')
2634 sanitize_numeric_fields(format)
2635 format['url'] = sanitize_url(format['url'])
2636 if format.get('ext') is None:
2637 format['ext'] = determine_ext(format['url']).lower()
2638 if format.get('protocol') is None:
2639 format['protocol'] = determine_protocol(format)
2640 if format.get('resolution') is None:
2641 format['resolution'] = self.format_resolution(format, default=None)
2642 if format.get('dynamic_range') is None and format.get('vcodec') != 'none':
2643 format['dynamic_range'] = 'SDR'
2644 if format.get('aspect_ratio') is None:
2645 format['aspect_ratio'] = try_call(lambda: round(format['width'] / format['height'], 2))
2646 if (info_dict.get('duration') and format.get('tbr')
2647 and not format.get('filesize') and not format.get('filesize_approx')):
2648 format['filesize_approx'] = int(info_dict['duration'] * format['tbr'] * (1024 / 8))
2649 format['http_headers'] = self._calc_headers(collections.ChainMap(format, info_dict))
2650
2651 # This is copied to http_headers by the above _calc_headers and can now be removed
2652 if '__x_forwarded_for_ip' in info_dict:
2653 del info_dict['__x_forwarded_for_ip']
2654
2655 self.sort_formats({
2656 'formats': formats,
2657 '_format_sort_fields': info_dict.get('_format_sort_fields')
2658 })
2659
2660 # Sanitize and group by format_id
2661 formats_dict = {}
2662 for i, format in enumerate(formats):
2663 if not format.get('format_id'):
2664 format['format_id'] = str(i)
2665 else:
2666 # Sanitize format_id from characters used in format selector expression
2667 format['format_id'] = re.sub(r'[\s,/+\[\]()]', '_', format['format_id'])
2668 formats_dict.setdefault(format['format_id'], []).append(format)
2669
2670 # Make sure all formats have unique format_id
2671 common_exts = set(itertools.chain(*self._format_selection_exts.values()))
2672 for format_id, ambiguous_formats in formats_dict.items():
2673 ambigious_id = len(ambiguous_formats) > 1
2674 for i, format in enumerate(ambiguous_formats):
2675 if ambigious_id:
2676 format['format_id'] = '%s-%d' % (format_id, i)
2677 # Ensure there is no conflict between id and ext in format selection
2678 # See https://github.com/yt-dlp/yt-dlp/issues/1282
2679 if format['format_id'] != format['ext'] and format['format_id'] in common_exts:
2680 format['format_id'] = 'f%s' % format['format_id']
2681
2682 if format.get('format') is None:
2683 format['format'] = '{id} - {res}{note}'.format(
2684 id=format['format_id'],
2685 res=self.format_resolution(format),
2686 note=format_field(format, 'format_note', ' (%s)'),
2687 )
2688
2689 if self.params.get('check_formats') is True:
2690 formats = LazyList(self._check_formats(formats[::-1]), reverse=True)
2691
2692 if not formats or formats[0] is not info_dict:
2693 # only set the 'formats' fields if the original info_dict list them
2694 # otherwise we end up with a circular reference, the first (and unique)
2695 # element in the 'formats' field in info_dict is info_dict itself,
2696 # which can't be exported to json
2697 info_dict['formats'] = formats
2698
2699 info_dict, _ = self.pre_process(info_dict)
2700
2701 if self._match_entry(info_dict, incomplete=self._format_fields) is not None:
2702 return info_dict
2703
2704 self.post_extract(info_dict)
2705 info_dict, _ = self.pre_process(info_dict, 'after_filter')
2706
2707 # The pre-processors may have modified the formats
2708 formats = self._get_formats(info_dict)
2709
2710 list_only = self.params.get('simulate') == 'list_only'
2711 interactive_format_selection = not list_only and self.format_selector == '-'
2712 if self.params.get('list_thumbnails'):
2713 self.list_thumbnails(info_dict)
2714 if self.params.get('listsubtitles'):
2715 if 'automatic_captions' in info_dict:
2716 self.list_subtitles(
2717 info_dict['id'], automatic_captions, 'automatic captions')
2718 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
2719 if self.params.get('listformats') or interactive_format_selection:
2720 self.list_formats(info_dict)
2721 if list_only:
2722 # Without this printing, -F --print-json will not work
2723 self.__forced_printings(info_dict)
2724 return info_dict
2725
2726 format_selector = self.format_selector
2727 if format_selector is None:
2728 req_format = self._default_format_spec(info_dict, download=download)
2729 self.write_debug('Default format spec: %s' % req_format)
2730 format_selector = self.build_format_selector(req_format)
2731
2732 while True:
2733 if interactive_format_selection:
2734 req_format = input(
2735 self._format_screen('\nEnter format selector: ', self.Styles.EMPHASIS))
2736 try:
2737 format_selector = self.build_format_selector(req_format)
2738 except SyntaxError as err:
2739 self.report_error(err, tb=False, is_error=False)
2740 continue
2741
2742 formats_to_download = list(format_selector({
2743 'formats': formats,
2744 'has_merged_format': any('none' not in (f.get('acodec'), f.get('vcodec')) for f in formats),
2745 'incomplete_formats': (
2746 # All formats are video-only or
2747 all(f.get('vcodec') != 'none' and f.get('acodec') == 'none' for f in formats)
2748 # all formats are audio-only
2749 or all(f.get('vcodec') == 'none' and f.get('acodec') != 'none' for f in formats)),
2750 }))
2751 if interactive_format_selection and not formats_to_download:
2752 self.report_error('Requested format is not available', tb=False, is_error=False)
2753 continue
2754 break
2755
2756 if not formats_to_download:
2757 if not self.params.get('ignore_no_formats_error'):
2758 raise ExtractorError(
2759 'Requested format is not available. Use --list-formats for a list of available formats',
2760 expected=True, video_id=info_dict['id'], ie=info_dict['extractor'])
2761 self.report_warning('Requested format is not available')
2762 # Process what we can, even without any available formats.
2763 formats_to_download = [{}]
2764
2765 requested_ranges = tuple(self.params.get('download_ranges', lambda *_: [{}])(info_dict, self))
2766 best_format, downloaded_formats = formats_to_download[-1], []
2767 if download:
2768 if best_format and requested_ranges:
2769 def to_screen(*msg):
2770 self.to_screen(f'[info] {info_dict["id"]}: {" ".join(", ".join(variadic(m)) for m in msg)}')
2771
2772 to_screen(f'Downloading {len(formats_to_download)} format(s):',
2773 (f['format_id'] for f in formats_to_download))
2774 if requested_ranges != ({}, ):
2775 to_screen(f'Downloading {len(requested_ranges)} time ranges:',
2776 (f'{c["start_time"]:.1f}-{c["end_time"]:.1f}' for c in requested_ranges))
2777 max_downloads_reached = False
2778
2779 for fmt, chapter in itertools.product(formats_to_download, requested_ranges):
2780 new_info = self._copy_infodict(info_dict)
2781 new_info.update(fmt)
2782 offset, duration = info_dict.get('section_start') or 0, info_dict.get('duration') or float('inf')
2783 end_time = offset + min(chapter.get('end_time', duration), duration)
2784 if chapter or offset:
2785 new_info.update({
2786 'section_start': offset + chapter.get('start_time', 0),
2787 # duration may not be accurate. So allow deviations <1sec
2788 'section_end': end_time if end_time <= offset + duration + 1 else None,
2789 'section_title': chapter.get('title'),
2790 'section_number': chapter.get('index'),
2791 })
2792 downloaded_formats.append(new_info)
2793 try:
2794 self.process_info(new_info)
2795 except MaxDownloadsReached:
2796 max_downloads_reached = True
2797 self._raise_pending_errors(new_info)
2798 # Remove copied info
2799 for key, val in tuple(new_info.items()):
2800 if info_dict.get(key) == val:
2801 new_info.pop(key)
2802 if max_downloads_reached:
2803 break
2804
2805 write_archive = {f.get('__write_download_archive', False) for f in downloaded_formats}
2806 assert write_archive.issubset({True, False, 'ignore'})
2807 if True in write_archive and False not in write_archive:
2808 self.record_download_archive(info_dict)
2809
2810 info_dict['requested_downloads'] = downloaded_formats
2811 info_dict = self.run_all_pps('after_video', info_dict)
2812 if max_downloads_reached:
2813 raise MaxDownloadsReached()
2814
2815 # We update the info dict with the selected best quality format (backwards compatibility)
2816 info_dict.update(best_format)
2817 return info_dict
2818
2819 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
2820 """Select the requested subtitles and their format"""
2821 available_subs, normal_sub_langs = {}, []
2822 if normal_subtitles and self.params.get('writesubtitles'):
2823 available_subs.update(normal_subtitles)
2824 normal_sub_langs = tuple(normal_subtitles.keys())
2825 if automatic_captions and self.params.get('writeautomaticsub'):
2826 for lang, cap_info in automatic_captions.items():
2827 if lang not in available_subs:
2828 available_subs[lang] = cap_info
2829
2830 if not available_subs or (
2831 not self.params.get('writesubtitles')
2832 and not self.params.get('writeautomaticsub')):
2833 return None
2834
2835 all_sub_langs = tuple(available_subs.keys())
2836 if self.params.get('allsubtitles', False):
2837 requested_langs = all_sub_langs
2838 elif self.params.get('subtitleslangs', False):
2839 try:
2840 requested_langs = orderedSet_from_options(
2841 self.params.get('subtitleslangs'), {'all': all_sub_langs}, use_regex=True)
2842 except re.error as e:
2843 raise ValueError(f'Wrong regex for subtitlelangs: {e.pattern}')
2844 else:
2845 requested_langs = LazyList(itertools.chain(
2846 ['en'] if 'en' in normal_sub_langs else [],
2847 filter(lambda f: f.startswith('en'), normal_sub_langs),
2848 ['en'] if 'en' in all_sub_langs else [],
2849 filter(lambda f: f.startswith('en'), all_sub_langs),
2850 normal_sub_langs, all_sub_langs,
2851 ))[:1]
2852 if requested_langs:
2853 self.to_screen(f'[info] {video_id}: Downloading subtitles: {", ".join(requested_langs)}')
2854
2855 formats_query = self.params.get('subtitlesformat', 'best')
2856 formats_preference = formats_query.split('/') if formats_query else []
2857 subs = {}
2858 for lang in requested_langs:
2859 formats = available_subs.get(lang)
2860 if formats is None:
2861 self.report_warning(f'{lang} subtitles not available for {video_id}')
2862 continue
2863 for ext in formats_preference:
2864 if ext == 'best':
2865 f = formats[-1]
2866 break
2867 matches = list(filter(lambda f: f['ext'] == ext, formats))
2868 if matches:
2869 f = matches[-1]
2870 break
2871 else:
2872 f = formats[-1]
2873 self.report_warning(
2874 'No subtitle format found matching "%s" for language %s, '
2875 'using %s' % (formats_query, lang, f['ext']))
2876 subs[lang] = f
2877 return subs
2878
2879 def _forceprint(self, key, info_dict):
2880 if info_dict is None:
2881 return
2882 info_copy = info_dict.copy()
2883 info_copy.setdefault('filename', self.prepare_filename(info_dict))
2884 if info_dict.get('requested_formats') is not None:
2885 # For RTMP URLs, also include the playpath
2886 info_copy['urls'] = '\n'.join(f['url'] + f.get('play_path', '') for f in info_dict['requested_formats'])
2887 elif info_dict.get('url'):
2888 info_copy['urls'] = info_dict['url'] + info_dict.get('play_path', '')
2889 info_copy['formats_table'] = self.render_formats_table(info_dict)
2890 info_copy['thumbnails_table'] = self.render_thumbnails_table(info_dict)
2891 info_copy['subtitles_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('subtitles'))
2892 info_copy['automatic_captions_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('automatic_captions'))
2893
2894 def format_tmpl(tmpl):
2895 mobj = re.fullmatch(r'([\w.:,]|-\d|(?P<dict>{([\w.:,]|-\d)+}))+=?', tmpl)
2896 if not mobj:
2897 return tmpl
2898
2899 fmt = '%({})s'
2900 if tmpl.startswith('{'):
2901 tmpl, fmt = f'.{tmpl}', '%({})j'
2902 if tmpl.endswith('='):
2903 tmpl, fmt = tmpl[:-1], '{0} = %({0})#j'
2904 return '\n'.join(map(fmt.format, [tmpl] if mobj.group('dict') else tmpl.split(',')))
2905
2906 for tmpl in self.params['forceprint'].get(key, []):
2907 self.to_stdout(self.evaluate_outtmpl(format_tmpl(tmpl), info_copy))
2908
2909 for tmpl, file_tmpl in self.params['print_to_file'].get(key, []):
2910 filename = self.prepare_filename(info_dict, outtmpl=file_tmpl)
2911 tmpl = format_tmpl(tmpl)
2912 self.to_screen(f'[info] Writing {tmpl!r} to: {filename}')
2913 if self._ensure_dir_exists(filename):
2914 with open(filename, 'a', encoding='utf-8', newline='') as f:
2915 f.write(self.evaluate_outtmpl(tmpl, info_copy) + os.linesep)
2916
2917 return info_copy
2918
2919 def __forced_printings(self, info_dict, filename=None, incomplete=True):
2920 if (self.params.get('forcejson')
2921 or self.params['forceprint'].get('video')
2922 or self.params['print_to_file'].get('video')):
2923 self.post_extract(info_dict)
2924 if filename:
2925 info_dict['filename'] = filename
2926 info_copy = self._forceprint('video', info_dict)
2927
2928 def print_field(field, actual_field=None, optional=False):
2929 if actual_field is None:
2930 actual_field = field
2931 if self.params.get(f'force{field}') and (
2932 info_copy.get(field) is not None or (not optional and not incomplete)):
2933 self.to_stdout(info_copy[actual_field])
2934
2935 print_field('title')
2936 print_field('id')
2937 print_field('url', 'urls')
2938 print_field('thumbnail', optional=True)
2939 print_field('description', optional=True)
2940 if filename:
2941 print_field('filename')
2942 if self.params.get('forceduration') and info_copy.get('duration') is not None:
2943 self.to_stdout(formatSeconds(info_copy['duration']))
2944 print_field('format')
2945
2946 if self.params.get('forcejson'):
2947 self.to_stdout(json.dumps(self.sanitize_info(info_dict)))
2948
2949 def dl(self, name, info, subtitle=False, test=False):
2950 if not info.get('url'):
2951 self.raise_no_formats(info, True)
2952
2953 if test:
2954 verbose = self.params.get('verbose')
2955 params = {
2956 'test': True,
2957 'quiet': self.params.get('quiet') or not verbose,
2958 'verbose': verbose,
2959 'noprogress': not verbose,
2960 'nopart': True,
2961 'skip_unavailable_fragments': False,
2962 'keep_fragments': False,
2963 'overwrites': True,
2964 '_no_ytdl_file': True,
2965 }
2966 else:
2967 params = self.params
2968 fd = get_suitable_downloader(info, params, to_stdout=(name == '-'))(self, params)
2969 if not test:
2970 for ph in self._progress_hooks:
2971 fd.add_progress_hook(ph)
2972 urls = '", "'.join(
2973 (f['url'].split(',')[0] + ',<data>' if f['url'].startswith('data:') else f['url'])
2974 for f in info.get('requested_formats', []) or [info])
2975 self.write_debug(f'Invoking {fd.FD_NAME} downloader on "{urls}"')
2976
2977 # Note: Ideally info should be a deep-copied so that hooks cannot modify it.
2978 # But it may contain objects that are not deep-copyable
2979 new_info = self._copy_infodict(info)
2980 if new_info.get('http_headers') is None:
2981 new_info['http_headers'] = self._calc_headers(new_info)
2982 return fd.download(name, new_info, subtitle)
2983
2984 def existing_file(self, filepaths, *, default_overwrite=True):
2985 existing_files = list(filter(os.path.exists, orderedSet(filepaths)))
2986 if existing_files and not self.params.get('overwrites', default_overwrite):
2987 return existing_files[0]
2988
2989 for file in existing_files:
2990 self.report_file_delete(file)
2991 os.remove(file)
2992 return None
2993
2994 def process_info(self, info_dict):
2995 """Process a single resolved IE result. (Modifies it in-place)"""
2996
2997 assert info_dict.get('_type', 'video') == 'video'
2998 original_infodict = info_dict
2999
3000 if 'format' not in info_dict and 'ext' in info_dict:
3001 info_dict['format'] = info_dict['ext']
3002
3003 if self._match_entry(info_dict) is not None:
3004 info_dict['__write_download_archive'] = 'ignore'
3005 return
3006
3007 # Does nothing under normal operation - for backward compatibility of process_info
3008 self.post_extract(info_dict)
3009
3010 def replace_info_dict(new_info):
3011 nonlocal info_dict
3012 if new_info == info_dict:
3013 return
3014 info_dict.clear()
3015 info_dict.update(new_info)
3016
3017 new_info, _ = self.pre_process(info_dict, 'video')
3018 replace_info_dict(new_info)
3019 self._num_downloads += 1
3020
3021 # info_dict['_filename'] needs to be set for backward compatibility
3022 info_dict['_filename'] = full_filename = self.prepare_filename(info_dict, warn=True)
3023 temp_filename = self.prepare_filename(info_dict, 'temp')
3024 files_to_move = {}
3025
3026 # Forced printings
3027 self.__forced_printings(info_dict, full_filename, incomplete=('format' not in info_dict))
3028
3029 def check_max_downloads():
3030 if self._num_downloads >= float(self.params.get('max_downloads') or 'inf'):
3031 raise MaxDownloadsReached()
3032
3033 if self.params.get('simulate'):
3034 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
3035 check_max_downloads()
3036 return
3037
3038 if full_filename is None:
3039 return
3040 if not self._ensure_dir_exists(encodeFilename(full_filename)):
3041 return
3042 if not self._ensure_dir_exists(encodeFilename(temp_filename)):
3043 return
3044
3045 if self._write_description('video', info_dict,
3046 self.prepare_filename(info_dict, 'description')) is None:
3047 return
3048
3049 sub_files = self._write_subtitles(info_dict, temp_filename)
3050 if sub_files is None:
3051 return
3052 files_to_move.update(dict(sub_files))
3053
3054 thumb_files = self._write_thumbnails(
3055 'video', info_dict, temp_filename, self.prepare_filename(info_dict, 'thumbnail'))
3056 if thumb_files is None:
3057 return
3058 files_to_move.update(dict(thumb_files))
3059
3060 infofn = self.prepare_filename(info_dict, 'infojson')
3061 _infojson_written = self._write_info_json('video', info_dict, infofn)
3062 if _infojson_written:
3063 info_dict['infojson_filename'] = infofn
3064 # For backward compatibility, even though it was a private field
3065 info_dict['__infojson_filename'] = infofn
3066 elif _infojson_written is None:
3067 return
3068
3069 # Note: Annotations are deprecated
3070 annofn = None
3071 if self.params.get('writeannotations', False):
3072 annofn = self.prepare_filename(info_dict, 'annotation')
3073 if annofn:
3074 if not self._ensure_dir_exists(encodeFilename(annofn)):
3075 return
3076 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(annofn)):
3077 self.to_screen('[info] Video annotations are already present')
3078 elif not info_dict.get('annotations'):
3079 self.report_warning('There are no annotations to write.')
3080 else:
3081 try:
3082 self.to_screen('[info] Writing video annotations to: ' + annofn)
3083 with open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
3084 annofile.write(info_dict['annotations'])
3085 except (KeyError, TypeError):
3086 self.report_warning('There are no annotations to write.')
3087 except OSError:
3088 self.report_error('Cannot write annotations file: ' + annofn)
3089 return
3090
3091 # Write internet shortcut files
3092 def _write_link_file(link_type):
3093 url = try_get(info_dict['webpage_url'], iri_to_uri)
3094 if not url:
3095 self.report_warning(
3096 f'Cannot write internet shortcut file because the actual URL of "{info_dict["webpage_url"]}" is unknown')
3097 return True
3098 linkfn = replace_extension(self.prepare_filename(info_dict, 'link'), link_type, info_dict.get('ext'))
3099 if not self._ensure_dir_exists(encodeFilename(linkfn)):
3100 return False
3101 if self.params.get('overwrites', True) and os.path.exists(encodeFilename(linkfn)):
3102 self.to_screen(f'[info] Internet shortcut (.{link_type}) is already present')
3103 return True
3104 try:
3105 self.to_screen(f'[info] Writing internet shortcut (.{link_type}) to: {linkfn}')
3106 with open(encodeFilename(to_high_limit_path(linkfn)), 'w', encoding='utf-8',
3107 newline='\r\n' if link_type == 'url' else '\n') as linkfile:
3108 template_vars = {'url': url}
3109 if link_type == 'desktop':
3110 template_vars['filename'] = linkfn[:-(len(link_type) + 1)]
3111 linkfile.write(LINK_TEMPLATES[link_type] % template_vars)
3112 except OSError:
3113 self.report_error(f'Cannot write internet shortcut {linkfn}')
3114 return False
3115 return True
3116
3117 write_links = {
3118 'url': self.params.get('writeurllink'),
3119 'webloc': self.params.get('writewebloclink'),
3120 'desktop': self.params.get('writedesktoplink'),
3121 }
3122 if self.params.get('writelink'):
3123 link_type = ('webloc' if sys.platform == 'darwin'
3124 else 'desktop' if sys.platform.startswith('linux')
3125 else 'url')
3126 write_links[link_type] = True
3127
3128 if any(should_write and not _write_link_file(link_type)
3129 for link_type, should_write in write_links.items()):
3130 return
3131
3132 new_info, files_to_move = self.pre_process(info_dict, 'before_dl', files_to_move)
3133 replace_info_dict(new_info)
3134
3135 if self.params.get('skip_download'):
3136 info_dict['filepath'] = temp_filename
3137 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
3138 info_dict['__files_to_move'] = files_to_move
3139 replace_info_dict(self.run_pp(MoveFilesAfterDownloadPP(self, False), info_dict))
3140 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
3141 else:
3142 # Download
3143 info_dict.setdefault('__postprocessors', [])
3144 try:
3145
3146 def existing_video_file(*filepaths):
3147 ext = info_dict.get('ext')
3148 converted = lambda file: replace_extension(file, self.params.get('final_ext') or ext, ext)
3149 file = self.existing_file(itertools.chain(*zip(map(converted, filepaths), filepaths)),
3150 default_overwrite=False)
3151 if file:
3152 info_dict['ext'] = os.path.splitext(file)[1][1:]
3153 return file
3154
3155 fd, success = None, True
3156 if info_dict.get('protocol') or info_dict.get('url'):
3157 fd = get_suitable_downloader(info_dict, self.params, to_stdout=temp_filename == '-')
3158 if fd is not FFmpegFD and 'no-direct-merge' not in self.params['compat_opts'] and (
3159 info_dict.get('section_start') or info_dict.get('section_end')):
3160 msg = ('This format cannot be partially downloaded' if FFmpegFD.available()
3161 else 'You have requested downloading the video partially, but ffmpeg is not installed')
3162 self.report_error(f'{msg}. Aborting')
3163 return
3164
3165 if info_dict.get('requested_formats') is not None:
3166 requested_formats = info_dict['requested_formats']
3167 old_ext = info_dict['ext']
3168 if self.params.get('merge_output_format') is None:
3169 if (info_dict['ext'] == 'webm'
3170 and info_dict.get('thumbnails')
3171 # check with type instead of pp_key, __name__, or isinstance
3172 # since we dont want any custom PPs to trigger this
3173 and any(type(pp) == EmbedThumbnailPP for pp in self._pps['post_process'])): # noqa: E721
3174 info_dict['ext'] = 'mkv'
3175 self.report_warning(
3176 'webm doesn\'t support embedding a thumbnail, mkv will be used')
3177 new_ext = info_dict['ext']
3178
3179 def correct_ext(filename, ext=new_ext):
3180 if filename == '-':
3181 return filename
3182 filename_real_ext = os.path.splitext(filename)[1][1:]
3183 filename_wo_ext = (
3184 os.path.splitext(filename)[0]
3185 if filename_real_ext in (old_ext, new_ext)
3186 else filename)
3187 return f'{filename_wo_ext}.{ext}'
3188
3189 # Ensure filename always has a correct extension for successful merge
3190 full_filename = correct_ext(full_filename)
3191 temp_filename = correct_ext(temp_filename)
3192 dl_filename = existing_video_file(full_filename, temp_filename)
3193 info_dict['__real_download'] = False
3194
3195 merger = FFmpegMergerPP(self)
3196 downloaded = []
3197 if dl_filename is not None:
3198 self.report_file_already_downloaded(dl_filename)
3199 elif fd:
3200 for f in requested_formats if fd != FFmpegFD else []:
3201 f['filepath'] = fname = prepend_extension(
3202 correct_ext(temp_filename, info_dict['ext']),
3203 'f%s' % f['format_id'], info_dict['ext'])
3204 downloaded.append(fname)
3205 info_dict['url'] = '\n'.join(f['url'] for f in requested_formats)
3206 success, real_download = self.dl(temp_filename, info_dict)
3207 info_dict['__real_download'] = real_download
3208 else:
3209 if self.params.get('allow_unplayable_formats'):
3210 self.report_warning(
3211 'You have requested merging of multiple formats '
3212 'while also allowing unplayable formats to be downloaded. '
3213 'The formats won\'t be merged to prevent data corruption.')
3214 elif not merger.available:
3215 msg = 'You have requested merging of multiple formats but ffmpeg is not installed'
3216 if not self.params.get('ignoreerrors'):
3217 self.report_error(f'{msg}. Aborting due to --abort-on-error')
3218 return
3219 self.report_warning(f'{msg}. The formats won\'t be merged')
3220
3221 if temp_filename == '-':
3222 reason = ('using a downloader other than ffmpeg' if FFmpegFD.can_merge_formats(info_dict, self.params)
3223 else 'but the formats are incompatible for simultaneous download' if merger.available
3224 else 'but ffmpeg is not installed')
3225 self.report_warning(
3226 f'You have requested downloading multiple formats to stdout {reason}. '
3227 'The formats will be streamed one after the other')
3228 fname = temp_filename
3229 for f in requested_formats:
3230 new_info = dict(info_dict)
3231 del new_info['requested_formats']
3232 new_info.update(f)
3233 if temp_filename != '-':
3234 fname = prepend_extension(
3235 correct_ext(temp_filename, new_info['ext']),
3236 'f%s' % f['format_id'], new_info['ext'])
3237 if not self._ensure_dir_exists(fname):
3238 return
3239 f['filepath'] = fname
3240 downloaded.append(fname)
3241 partial_success, real_download = self.dl(fname, new_info)
3242 info_dict['__real_download'] = info_dict['__real_download'] or real_download
3243 success = success and partial_success
3244
3245 if downloaded and merger.available and not self.params.get('allow_unplayable_formats'):
3246 info_dict['__postprocessors'].append(merger)
3247 info_dict['__files_to_merge'] = downloaded
3248 # Even if there were no downloads, it is being merged only now
3249 info_dict['__real_download'] = True
3250 else:
3251 for file in downloaded:
3252 files_to_move[file] = None
3253 else:
3254 # Just a single file
3255 dl_filename = existing_video_file(full_filename, temp_filename)
3256 if dl_filename is None or dl_filename == temp_filename:
3257 # dl_filename == temp_filename could mean that the file was partially downloaded with --no-part.
3258 # So we should try to resume the download
3259 success, real_download = self.dl(temp_filename, info_dict)
3260 info_dict['__real_download'] = real_download
3261 else:
3262 self.report_file_already_downloaded(dl_filename)
3263
3264 dl_filename = dl_filename or temp_filename
3265 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
3266
3267 except network_exceptions as err:
3268 self.report_error('unable to download video data: %s' % error_to_compat_str(err))
3269 return
3270 except OSError as err:
3271 raise UnavailableVideoError(err)
3272 except (ContentTooShortError, ) as err:
3273 self.report_error(f'content too short (expected {err.expected} bytes and served {err.downloaded})')
3274 return
3275
3276 self._raise_pending_errors(info_dict)
3277 if success and full_filename != '-':
3278
3279 def fixup():
3280 do_fixup = True
3281 fixup_policy = self.params.get('fixup')
3282 vid = info_dict['id']
3283
3284 if fixup_policy in ('ignore', 'never'):
3285 return
3286 elif fixup_policy == 'warn':
3287 do_fixup = 'warn'
3288 elif fixup_policy != 'force':
3289 assert fixup_policy in ('detect_or_warn', None)
3290 if not info_dict.get('__real_download'):
3291 do_fixup = False
3292
3293 def ffmpeg_fixup(cndn, msg, cls):
3294 if not (do_fixup and cndn):
3295 return
3296 elif do_fixup == 'warn':
3297 self.report_warning(f'{vid}: {msg}')
3298 return
3299 pp = cls(self)
3300 if pp.available:
3301 info_dict['__postprocessors'].append(pp)
3302 else:
3303 self.report_warning(f'{vid}: {msg}. Install ffmpeg to fix this automatically')
3304
3305 stretched_ratio = info_dict.get('stretched_ratio')
3306 ffmpeg_fixup(stretched_ratio not in (1, None),
3307 f'Non-uniform pixel ratio {stretched_ratio}',
3308 FFmpegFixupStretchedPP)
3309
3310 downloader = get_suitable_downloader(info_dict, self.params) if 'protocol' in info_dict else None
3311 downloader = downloader.FD_NAME if downloader else None
3312
3313 ext = info_dict.get('ext')
3314 postprocessed_by_ffmpeg = info_dict.get('requested_formats') or any((
3315 isinstance(pp, FFmpegVideoConvertorPP)
3316 and resolve_recode_mapping(ext, pp.mapping)[0] not in (ext, None)
3317 ) for pp in self._pps['post_process'])
3318
3319 if not postprocessed_by_ffmpeg:
3320 ffmpeg_fixup(ext == 'm4a' and info_dict.get('container') == 'm4a_dash',
3321 'writing DASH m4a. Only some players support this container',
3322 FFmpegFixupM4aPP)
3323 ffmpeg_fixup(downloader == 'hlsnative' and not self.params.get('hls_use_mpegts')
3324 or info_dict.get('is_live') and self.params.get('hls_use_mpegts') is None,
3325 'Possible MPEG-TS in MP4 container or malformed AAC timestamps',
3326 FFmpegFixupM3u8PP)
3327 ffmpeg_fixup(info_dict.get('is_live') and downloader == 'dashsegments',
3328 'Possible duplicate MOOV atoms', FFmpegFixupDuplicateMoovPP)
3329
3330 ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed timestamps detected', FFmpegFixupTimestampPP)
3331 ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed duration detected', FFmpegFixupDurationPP)
3332
3333 fixup()
3334 try:
3335 replace_info_dict(self.post_process(dl_filename, info_dict, files_to_move))
3336 except PostProcessingError as err:
3337 self.report_error('Postprocessing: %s' % str(err))
3338 return
3339 try:
3340 for ph in self._post_hooks:
3341 ph(info_dict['filepath'])
3342 except Exception as err:
3343 self.report_error('post hooks: %s' % str(err))
3344 return
3345 info_dict['__write_download_archive'] = True
3346
3347 assert info_dict is original_infodict # Make sure the info_dict was modified in-place
3348 if self.params.get('force_write_download_archive'):
3349 info_dict['__write_download_archive'] = True
3350 check_max_downloads()
3351
3352 def __download_wrapper(self, func):
3353 @functools.wraps(func)
3354 def wrapper(*args, **kwargs):
3355 try:
3356 res = func(*args, **kwargs)
3357 except UnavailableVideoError as e:
3358 self.report_error(e)
3359 except DownloadCancelled as e:
3360 self.to_screen(f'[info] {e}')
3361 if not self.params.get('break_per_url'):
3362 raise
3363 self._num_downloads = 0
3364 else:
3365 if self.params.get('dump_single_json', False):
3366 self.post_extract(res)
3367 self.to_stdout(json.dumps(self.sanitize_info(res)))
3368 return wrapper
3369
3370 def download(self, url_list):
3371 """Download a given list of URLs."""
3372 url_list = variadic(url_list) # Passing a single URL is a common mistake
3373 outtmpl = self.params['outtmpl']['default']
3374 if (len(url_list) > 1
3375 and outtmpl != '-'
3376 and '%' not in outtmpl
3377 and self.params.get('max_downloads') != 1):
3378 raise SameFileError(outtmpl)
3379
3380 for url in url_list:
3381 self.__download_wrapper(self.extract_info)(
3382 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
3383
3384 return self._download_retcode
3385
3386 def download_with_info_file(self, info_filename):
3387 with contextlib.closing(fileinput.FileInput(
3388 [info_filename], mode='r',
3389 openhook=fileinput.hook_encoded('utf-8'))) as f:
3390 # FileInput doesn't have a read method, we can't call json.load
3391 infos = [self.sanitize_info(info, self.params.get('clean_infojson', True))
3392 for info in variadic(json.loads('\n'.join(f)))]
3393 for info in infos:
3394 try:
3395 self.__download_wrapper(self.process_ie_result)(info, download=True)
3396 except (DownloadError, EntryNotInPlaylist, ReExtractInfo) as e:
3397 if not isinstance(e, EntryNotInPlaylist):
3398 self.to_stderr('\r')
3399 webpage_url = info.get('webpage_url')
3400 if webpage_url is None:
3401 raise
3402 self.report_warning(f'The info failed to download: {e}; trying with URL {webpage_url}')
3403 self.download([webpage_url])
3404 return self._download_retcode
3405
3406 @staticmethod
3407 def sanitize_info(info_dict, remove_private_keys=False):
3408 ''' Sanitize the infodict for converting to json '''
3409 if info_dict is None:
3410 return info_dict
3411 info_dict.setdefault('epoch', int(time.time()))
3412 info_dict.setdefault('_type', 'video')
3413 info_dict.setdefault('_version', {
3414 'version': __version__,
3415 'current_git_head': current_git_head(),
3416 'release_git_head': RELEASE_GIT_HEAD,
3417 'repository': REPOSITORY,
3418 })
3419
3420 if remove_private_keys:
3421 reject = lambda k, v: v is None or k.startswith('__') or k in {
3422 'requested_downloads', 'requested_formats', 'requested_subtitles', 'requested_entries',
3423 'entries', 'filepath', '_filename', 'filename', 'infojson_filename', 'original_url',
3424 'playlist_autonumber', '_format_sort_fields',
3425 }
3426 else:
3427 reject = lambda k, v: False
3428
3429 def filter_fn(obj):
3430 if isinstance(obj, dict):
3431 return {k: filter_fn(v) for k, v in obj.items() if not reject(k, v)}
3432 elif isinstance(obj, (list, tuple, set, LazyList)):
3433 return list(map(filter_fn, obj))
3434 elif obj is None or isinstance(obj, (str, int, float, bool)):
3435 return obj
3436 else:
3437 return repr(obj)
3438
3439 return filter_fn(info_dict)
3440
3441 @staticmethod
3442 def filter_requested_info(info_dict, actually_filter=True):
3443 ''' Alias of sanitize_info for backward compatibility '''
3444 return YoutubeDL.sanitize_info(info_dict, actually_filter)
3445
3446 def _delete_downloaded_files(self, *files_to_delete, info={}, msg=None):
3447 for filename in set(filter(None, files_to_delete)):
3448 if msg:
3449 self.to_screen(msg % filename)
3450 try:
3451 os.remove(filename)
3452 except OSError:
3453 self.report_warning(f'Unable to delete file {filename}')
3454 if filename in info.get('__files_to_move', []): # NB: Delete even if None
3455 del info['__files_to_move'][filename]
3456
3457 @staticmethod
3458 def post_extract(info_dict):
3459 def actual_post_extract(info_dict):
3460 if info_dict.get('_type') in ('playlist', 'multi_video'):
3461 for video_dict in info_dict.get('entries', {}):
3462 actual_post_extract(video_dict or {})
3463 return
3464
3465 post_extractor = info_dict.pop('__post_extractor', None) or (lambda: {})
3466 info_dict.update(post_extractor())
3467
3468 actual_post_extract(info_dict or {})
3469
3470 def run_pp(self, pp, infodict):
3471 files_to_delete = []
3472 if '__files_to_move' not in infodict:
3473 infodict['__files_to_move'] = {}
3474 try:
3475 files_to_delete, infodict = pp.run(infodict)
3476 except PostProcessingError as e:
3477 # Must be True and not 'only_download'
3478 if self.params.get('ignoreerrors') is True:
3479 self.report_error(e)
3480 return infodict
3481 raise
3482
3483 if not files_to_delete:
3484 return infodict
3485 if self.params.get('keepvideo', False):
3486 for f in files_to_delete:
3487 infodict['__files_to_move'].setdefault(f, '')
3488 else:
3489 self._delete_downloaded_files(
3490 *files_to_delete, info=infodict, msg='Deleting original file %s (pass -k to keep)')
3491 return infodict
3492
3493 def run_all_pps(self, key, info, *, additional_pps=None):
3494 if key != 'video':
3495 self._forceprint(key, info)
3496 for pp in (additional_pps or []) + self._pps[key]:
3497 info = self.run_pp(pp, info)
3498 return info
3499
3500 def pre_process(self, ie_info, key='pre_process', files_to_move=None):
3501 info = dict(ie_info)
3502 info['__files_to_move'] = files_to_move or {}
3503 try:
3504 info = self.run_all_pps(key, info)
3505 except PostProcessingError as err:
3506 msg = f'Preprocessing: {err}'
3507 info.setdefault('__pending_error', msg)
3508 self.report_error(msg, is_error=False)
3509 return info, info.pop('__files_to_move', None)
3510
3511 def post_process(self, filename, info, files_to_move=None):
3512 """Run all the postprocessors on the given file."""
3513 info['filepath'] = filename
3514 info['__files_to_move'] = files_to_move or {}
3515 info = self.run_all_pps('post_process', info, additional_pps=info.get('__postprocessors'))
3516 info = self.run_pp(MoveFilesAfterDownloadPP(self), info)
3517 del info['__files_to_move']
3518 return self.run_all_pps('after_move', info)
3519
3520 def _make_archive_id(self, info_dict):
3521 video_id = info_dict.get('id')
3522 if not video_id:
3523 return
3524 # Future-proof against any change in case
3525 # and backwards compatibility with prior versions
3526 extractor = info_dict.get('extractor_key') or info_dict.get('ie_key') # key in a playlist
3527 if extractor is None:
3528 url = str_or_none(info_dict.get('url'))
3529 if not url:
3530 return
3531 # Try to find matching extractor for the URL and take its ie_key
3532 for ie_key, ie in self._ies.items():
3533 if ie.suitable(url):
3534 extractor = ie_key
3535 break
3536 else:
3537 return
3538 return make_archive_id(extractor, video_id)
3539
3540 def in_download_archive(self, info_dict):
3541 if not self.archive:
3542 return False
3543
3544 vid_ids = [self._make_archive_id(info_dict)]
3545 vid_ids.extend(info_dict.get('_old_archive_ids') or [])
3546 return any(id_ in self.archive for id_ in vid_ids)
3547
3548 def record_download_archive(self, info_dict):
3549 fn = self.params.get('download_archive')
3550 if fn is None:
3551 return
3552 vid_id = self._make_archive_id(info_dict)
3553 assert vid_id
3554
3555 self.write_debug(f'Adding to archive: {vid_id}')
3556 if is_path_like(fn):
3557 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
3558 archive_file.write(vid_id + '\n')
3559 self.archive.add(vid_id)
3560
3561 @staticmethod
3562 def format_resolution(format, default='unknown'):
3563 if format.get('vcodec') == 'none' and format.get('acodec') != 'none':
3564 return 'audio only'
3565 if format.get('resolution') is not None:
3566 return format['resolution']
3567 if format.get('width') and format.get('height'):
3568 return '%dx%d' % (format['width'], format['height'])
3569 elif format.get('height'):
3570 return '%sp' % format['height']
3571 elif format.get('width'):
3572 return '%dx?' % format['width']
3573 return default
3574
3575 def _list_format_headers(self, *headers):
3576 if self.params.get('listformats_table', True) is not False:
3577 return [self._format_out(header, self.Styles.HEADERS) for header in headers]
3578 return headers
3579
3580 def _format_note(self, fdict):
3581 res = ''
3582 if fdict.get('ext') in ['f4f', 'f4m']:
3583 res += '(unsupported)'
3584 if fdict.get('language'):
3585 if res:
3586 res += ' '
3587 res += '[%s]' % fdict['language']
3588 if fdict.get('format_note') is not None:
3589 if res:
3590 res += ' '
3591 res += fdict['format_note']
3592 if fdict.get('tbr') is not None:
3593 if res:
3594 res += ', '
3595 res += '%4dk' % fdict['tbr']
3596 if fdict.get('container') is not None:
3597 if res:
3598 res += ', '
3599 res += '%s container' % fdict['container']
3600 if (fdict.get('vcodec') is not None
3601 and fdict.get('vcodec') != 'none'):
3602 if res:
3603 res += ', '
3604 res += fdict['vcodec']
3605 if fdict.get('vbr') is not None:
3606 res += '@'
3607 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
3608 res += 'video@'
3609 if fdict.get('vbr') is not None:
3610 res += '%4dk' % fdict['vbr']
3611 if fdict.get('fps') is not None:
3612 if res:
3613 res += ', '
3614 res += '%sfps' % fdict['fps']
3615 if fdict.get('acodec') is not None:
3616 if res:
3617 res += ', '
3618 if fdict['acodec'] == 'none':
3619 res += 'video only'
3620 else:
3621 res += '%-5s' % fdict['acodec']
3622 elif fdict.get('abr') is not None:
3623 if res:
3624 res += ', '
3625 res += 'audio'
3626 if fdict.get('abr') is not None:
3627 res += '@%3dk' % fdict['abr']
3628 if fdict.get('asr') is not None:
3629 res += ' (%5dHz)' % fdict['asr']
3630 if fdict.get('filesize') is not None:
3631 if res:
3632 res += ', '
3633 res += format_bytes(fdict['filesize'])
3634 elif fdict.get('filesize_approx') is not None:
3635 if res:
3636 res += ', '
3637 res += '~' + format_bytes(fdict['filesize_approx'])
3638 return res
3639
3640 def _get_formats(self, info_dict):
3641 if info_dict.get('formats') is None:
3642 if info_dict.get('url') and info_dict.get('_type', 'video') == 'video':
3643 return [info_dict]
3644 return []
3645 return info_dict['formats']
3646
3647 def render_formats_table(self, info_dict):
3648 formats = self._get_formats(info_dict)
3649 if not formats:
3650 return
3651 if not self.params.get('listformats_table', True) is not False:
3652 table = [
3653 [
3654 format_field(f, 'format_id'),
3655 format_field(f, 'ext'),
3656 self.format_resolution(f),
3657 self._format_note(f)
3658 ] for f in formats if (f.get('preference') or 0) >= -1000]
3659 return render_table(['format code', 'extension', 'resolution', 'note'], table, extra_gap=1)
3660
3661 def simplified_codec(f, field):
3662 assert field in ('acodec', 'vcodec')
3663 codec = f.get(field, 'unknown')
3664 if not codec:
3665 return 'unknown'
3666 elif codec != 'none':
3667 return '.'.join(codec.split('.')[:4])
3668
3669 if field == 'vcodec' and f.get('acodec') == 'none':
3670 return 'images'
3671 elif field == 'acodec' and f.get('vcodec') == 'none':
3672 return ''
3673 return self._format_out('audio only' if field == 'vcodec' else 'video only',
3674 self.Styles.SUPPRESS)
3675
3676 delim = self._format_out('\u2502', self.Styles.DELIM, '|', test_encoding=True)
3677 table = [
3678 [
3679 self._format_out(format_field(f, 'format_id'), self.Styles.ID),
3680 format_field(f, 'ext'),
3681 format_field(f, func=self.format_resolution, ignore=('audio only', 'images')),
3682 format_field(f, 'fps', '\t%d', func=round),
3683 format_field(f, 'dynamic_range', '%s', ignore=(None, 'SDR')).replace('HDR', ''),
3684 format_field(f, 'audio_channels', '\t%s'),
3685 delim,
3686 format_field(f, 'filesize', ' \t%s', func=format_bytes) + format_field(f, 'filesize_approx', '~\t%s', func=format_bytes),
3687 format_field(f, 'tbr', '\t%dk', func=round),
3688 shorten_protocol_name(f.get('protocol', '')),
3689 delim,
3690 simplified_codec(f, 'vcodec'),
3691 format_field(f, 'vbr', '\t%dk', func=round),
3692 simplified_codec(f, 'acodec'),
3693 format_field(f, 'abr', '\t%dk', func=round),
3694 format_field(f, 'asr', '\t%s', func=format_decimal_suffix),
3695 join_nonempty(
3696 self._format_out('UNSUPPORTED', 'light red') if f.get('ext') in ('f4f', 'f4m') else None,
3697 self._format_out('DRM', 'light red') if f.get('has_drm') else None,
3698 format_field(f, 'language', '[%s]'),
3699 join_nonempty(format_field(f, 'format_note'),
3700 format_field(f, 'container', ignore=(None, f.get('ext'))),
3701 delim=', '),
3702 delim=' '),
3703 ] for f in formats if f.get('preference') is None or f['preference'] >= -1000]
3704 header_line = self._list_format_headers(
3705 'ID', 'EXT', 'RESOLUTION', '\tFPS', 'HDR', 'CH', delim, '\tFILESIZE', '\tTBR', 'PROTO',
3706 delim, 'VCODEC', '\tVBR', 'ACODEC', '\tABR', '\tASR', 'MORE INFO')
3707
3708 return render_table(
3709 header_line, table, hide_empty=True,
3710 delim=self._format_out('\u2500', self.Styles.DELIM, '-', test_encoding=True))
3711
3712 def render_thumbnails_table(self, info_dict):
3713 thumbnails = list(info_dict.get('thumbnails') or [])
3714 if not thumbnails:
3715 return None
3716 return render_table(
3717 self._list_format_headers('ID', 'Width', 'Height', 'URL'),
3718 [[t.get('id'), t.get('width') or 'unknown', t.get('height') or 'unknown', t['url']] for t in thumbnails])
3719
3720 def render_subtitles_table(self, video_id, subtitles):
3721 def _row(lang, formats):
3722 exts, names = zip(*((f['ext'], f.get('name') or 'unknown') for f in reversed(formats)))
3723 if len(set(names)) == 1:
3724 names = [] if names[0] == 'unknown' else names[:1]
3725 return [lang, ', '.join(names), ', '.join(exts)]
3726
3727 if not subtitles:
3728 return None
3729 return render_table(
3730 self._list_format_headers('Language', 'Name', 'Formats'),
3731 [_row(lang, formats) for lang, formats in subtitles.items()],
3732 hide_empty=True)
3733
3734 def __list_table(self, video_id, name, func, *args):
3735 table = func(*args)
3736 if not table:
3737 self.to_screen(f'{video_id} has no {name}')
3738 return
3739 self.to_screen(f'[info] Available {name} for {video_id}:')
3740 self.to_stdout(table)
3741
3742 def list_formats(self, info_dict):
3743 self.__list_table(info_dict['id'], 'formats', self.render_formats_table, info_dict)
3744
3745 def list_thumbnails(self, info_dict):
3746 self.__list_table(info_dict['id'], 'thumbnails', self.render_thumbnails_table, info_dict)
3747
3748 def list_subtitles(self, video_id, subtitles, name='subtitles'):
3749 self.__list_table(video_id, name, self.render_subtitles_table, video_id, subtitles)
3750
3751 def urlopen(self, req):
3752 """ Start an HTTP download """
3753 if isinstance(req, str):
3754 req = sanitized_Request(req)
3755 return self._opener.open(req, timeout=self._socket_timeout)
3756
3757 def print_debug_header(self):
3758 if not self.params.get('verbose'):
3759 return
3760
3761 from . import _IN_CLI # Must be delayed import
3762
3763 # These imports can be slow. So import them only as needed
3764 from .extractor.extractors import _LAZY_LOADER
3765 from .extractor.extractors import (
3766 _PLUGIN_CLASSES as plugin_ies,
3767 _PLUGIN_OVERRIDES as plugin_ie_overrides
3768 )
3769
3770 def get_encoding(stream):
3771 ret = str(getattr(stream, 'encoding', 'missing (%s)' % type(stream).__name__))
3772 if not supports_terminal_sequences(stream):
3773 from .utils import WINDOWS_VT_MODE # Must be imported locally
3774 ret += ' (No VT)' if WINDOWS_VT_MODE is False else ' (No ANSI)'
3775 return ret
3776
3777 encoding_str = 'Encodings: locale %s, fs %s, pref %s, %s' % (
3778 locale.getpreferredencoding(),
3779 sys.getfilesystemencoding(),
3780 self.get_encoding(),
3781 ', '.join(
3782 f'{key} {get_encoding(stream)}' for key, stream in self._out_files.items_
3783 if stream is not None and key != 'console')
3784 )
3785
3786 logger = self.params.get('logger')
3787 if logger:
3788 write_debug = lambda msg: logger.debug(f'[debug] {msg}')
3789 write_debug(encoding_str)
3790 else:
3791 write_string(f'[debug] {encoding_str}\n', encoding=None)
3792 write_debug = lambda msg: self._write_string(f'[debug] {msg}\n')
3793
3794 source = detect_variant()
3795 if VARIANT not in (None, 'pip'):
3796 source += '*'
3797 klass = type(self)
3798 write_debug(join_nonempty(
3799 f'{"yt-dlp" if REPOSITORY == "yt-dlp/yt-dlp" else REPOSITORY} version',
3800 f'{CHANNEL}@{__version__}',
3801 f'[{RELEASE_GIT_HEAD[:9]}]' if RELEASE_GIT_HEAD else '',
3802 '' if source == 'unknown' else f'({source})',
3803 '' if _IN_CLI else 'API' if klass == YoutubeDL else f'API:{self.__module__}.{klass.__qualname__}',
3804 delim=' '))
3805
3806 if not _IN_CLI:
3807 write_debug(f'params: {self.params}')
3808
3809 if not _LAZY_LOADER:
3810 if os.environ.get('YTDLP_NO_LAZY_EXTRACTORS'):
3811 write_debug('Lazy loading extractors is forcibly disabled')
3812 else:
3813 write_debug('Lazy loading extractors is disabled')
3814 if self.params['compat_opts']:
3815 write_debug('Compatibility options: %s' % ', '.join(self.params['compat_opts']))
3816
3817 if current_git_head():
3818 write_debug(f'Git HEAD: {current_git_head()}')
3819 write_debug(system_identifier())
3820
3821 exe_versions, ffmpeg_features = FFmpegPostProcessor.get_versions_and_features(self)
3822 ffmpeg_features = {key for key, val in ffmpeg_features.items() if val}
3823 if ffmpeg_features:
3824 exe_versions['ffmpeg'] += ' (%s)' % ','.join(sorted(ffmpeg_features))
3825
3826 exe_versions['rtmpdump'] = rtmpdump_version()
3827 exe_versions['phantomjs'] = PhantomJSwrapper._version()
3828 exe_str = ', '.join(
3829 f'{exe} {v}' for exe, v in sorted(exe_versions.items()) if v
3830 ) or 'none'
3831 write_debug('exe versions: %s' % exe_str)
3832
3833 from .compat.compat_utils import get_package_info
3834 from .dependencies import available_dependencies
3835
3836 write_debug('Optional libraries: %s' % (', '.join(sorted({
3837 join_nonempty(*get_package_info(m)) for m in available_dependencies.values()
3838 })) or 'none'))
3839
3840 self._setup_opener()
3841 proxy_map = {}
3842 for handler in self._opener.handlers:
3843 if hasattr(handler, 'proxies'):
3844 proxy_map.update(handler.proxies)
3845 write_debug(f'Proxy map: {proxy_map}')
3846
3847 for plugin_type, plugins in {'Extractor': plugin_ies, 'Post-Processor': plugin_pps}.items():
3848 display_list = ['%s%s' % (
3849 klass.__name__, '' if klass.__name__ == name else f' as {name}')
3850 for name, klass in plugins.items()]
3851 if plugin_type == 'Extractor':
3852 display_list.extend(f'{plugins[-1].IE_NAME.partition("+")[2]} ({parent.__name__})'
3853 for parent, plugins in plugin_ie_overrides.items())
3854 if not display_list:
3855 continue
3856 write_debug(f'{plugin_type} Plugins: {", ".join(sorted(display_list))}')
3857
3858 plugin_dirs = plugin_directories()
3859 if plugin_dirs:
3860 write_debug(f'Plugin directories: {plugin_dirs}')
3861
3862 # Not implemented
3863 if False and self.params.get('call_home'):
3864 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode()
3865 write_debug('Public IP address: %s' % ipaddr)
3866 latest_version = self.urlopen(
3867 'https://yt-dl.org/latest/version').read().decode()
3868 if version_tuple(latest_version) > version_tuple(__version__):
3869 self.report_warning(
3870 'You are using an outdated version (newest version: %s)! '
3871 'See https://yt-dl.org/update if you need help updating.' %
3872 latest_version)
3873
3874 def _setup_opener(self):
3875 if hasattr(self, '_opener'):
3876 return
3877 timeout_val = self.params.get('socket_timeout')
3878 self._socket_timeout = 20 if timeout_val is None else float(timeout_val)
3879
3880 opts_cookiesfrombrowser = self.params.get('cookiesfrombrowser')
3881 opts_cookiefile = self.params.get('cookiefile')
3882 opts_proxy = self.params.get('proxy')
3883
3884 self.cookiejar = load_cookies(opts_cookiefile, opts_cookiesfrombrowser, self)
3885
3886 cookie_processor = YoutubeDLCookieProcessor(self.cookiejar)
3887 if opts_proxy is not None:
3888 if opts_proxy == '':
3889 proxies = {}
3890 else:
3891 proxies = {'http': opts_proxy, 'https': opts_proxy}
3892 else:
3893 proxies = urllib.request.getproxies()
3894 # Set HTTPS proxy to HTTP one if given (https://github.com/ytdl-org/youtube-dl/issues/805)
3895 if 'http' in proxies and 'https' not in proxies:
3896 proxies['https'] = proxies['http']
3897 proxy_handler = PerRequestProxyHandler(proxies)
3898
3899 debuglevel = 1 if self.params.get('debug_printtraffic') else 0
3900 https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
3901 ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
3902 redirect_handler = YoutubeDLRedirectHandler()
3903 data_handler = urllib.request.DataHandler()
3904
3905 # When passing our own FileHandler instance, build_opener won't add the
3906 # default FileHandler and allows us to disable the file protocol, which
3907 # can be used for malicious purposes (see
3908 # https://github.com/ytdl-org/youtube-dl/issues/8227)
3909 file_handler = urllib.request.FileHandler()
3910
3911 if not self.params.get('enable_file_urls'):
3912 def file_open(*args, **kwargs):
3913 raise urllib.error.URLError(
3914 'file:// URLs are explicitly disabled in yt-dlp for security reasons. '
3915 'Use --enable-file-urls to enable at your own risk.')
3916 file_handler.file_open = file_open
3917
3918 opener = urllib.request.build_opener(
3919 proxy_handler, https_handler, cookie_processor, ydlh, redirect_handler, data_handler, file_handler)
3920
3921 # Delete the default user-agent header, which would otherwise apply in
3922 # cases where our custom HTTP handler doesn't come into play
3923 # (See https://github.com/ytdl-org/youtube-dl/issues/1309 for details)
3924 opener.addheaders = []
3925 self._opener = opener
3926
3927 def encode(self, s):
3928 if isinstance(s, bytes):
3929 return s # Already encoded
3930
3931 try:
3932 return s.encode(self.get_encoding())
3933 except UnicodeEncodeError as err:
3934 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
3935 raise
3936
3937 def get_encoding(self):
3938 encoding = self.params.get('encoding')
3939 if encoding is None:
3940 encoding = preferredencoding()
3941 return encoding
3942
3943 def _write_info_json(self, label, ie_result, infofn, overwrite=None):
3944 ''' Write infojson and returns True = written, 'exists' = Already exists, False = skip, None = error '''
3945 if overwrite is None:
3946 overwrite = self.params.get('overwrites', True)
3947 if not self.params.get('writeinfojson'):
3948 return False
3949 elif not infofn:
3950 self.write_debug(f'Skipping writing {label} infojson')
3951 return False
3952 elif not self._ensure_dir_exists(infofn):
3953 return None
3954 elif not overwrite and os.path.exists(infofn):
3955 self.to_screen(f'[info] {label.title()} metadata is already present')
3956 return 'exists'
3957
3958 self.to_screen(f'[info] Writing {label} metadata as JSON to: {infofn}')
3959 try:
3960 write_json_file(self.sanitize_info(ie_result, self.params.get('clean_infojson', True)), infofn)
3961 return True
3962 except OSError:
3963 self.report_error(f'Cannot write {label} metadata to JSON file {infofn}')
3964 return None
3965
3966 def _write_description(self, label, ie_result, descfn):
3967 ''' Write description and returns True = written, False = skip, None = error '''
3968 if not self.params.get('writedescription'):
3969 return False
3970 elif not descfn:
3971 self.write_debug(f'Skipping writing {label} description')
3972 return False
3973 elif not self._ensure_dir_exists(descfn):
3974 return None
3975 elif not self.params.get('overwrites', True) and os.path.exists(descfn):
3976 self.to_screen(f'[info] {label.title()} description is already present')
3977 elif ie_result.get('description') is None:
3978 self.to_screen(f'[info] There\'s no {label} description to write')
3979 return False
3980 else:
3981 try:
3982 self.to_screen(f'[info] Writing {label} description to: {descfn}')
3983 with open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
3984 descfile.write(ie_result['description'])
3985 except OSError:
3986 self.report_error(f'Cannot write {label} description file {descfn}')
3987 return None
3988 return True
3989
3990 def _write_subtitles(self, info_dict, filename):
3991 ''' Write subtitles to file and return list of (sub_filename, final_sub_filename); or None if error'''
3992 ret = []
3993 subtitles = info_dict.get('requested_subtitles')
3994 if not (self.params.get('writesubtitles') or self.params.get('writeautomaticsub')):
3995 # subtitles download errors are already managed as troubles in relevant IE
3996 # that way it will silently go on when used with unsupporting IE
3997 return ret
3998 elif not subtitles:
3999 self.to_screen('[info] There are no subtitles for the requested languages')
4000 return ret
4001 sub_filename_base = self.prepare_filename(info_dict, 'subtitle')
4002 if not sub_filename_base:
4003 self.to_screen('[info] Skipping writing video subtitles')
4004 return ret
4005
4006 for sub_lang, sub_info in subtitles.items():
4007 sub_format = sub_info['ext']
4008 sub_filename = subtitles_filename(filename, sub_lang, sub_format, info_dict.get('ext'))
4009 sub_filename_final = subtitles_filename(sub_filename_base, sub_lang, sub_format, info_dict.get('ext'))
4010 existing_sub = self.existing_file((sub_filename_final, sub_filename))
4011 if existing_sub:
4012 self.to_screen(f'[info] Video subtitle {sub_lang}.{sub_format} is already present')
4013 sub_info['filepath'] = existing_sub
4014 ret.append((existing_sub, sub_filename_final))
4015 continue
4016
4017 self.to_screen(f'[info] Writing video subtitles to: {sub_filename}')
4018 if sub_info.get('data') is not None:
4019 try:
4020 # Use newline='' to prevent conversion of newline characters
4021 # See https://github.com/ytdl-org/youtube-dl/issues/10268
4022 with open(sub_filename, 'w', encoding='utf-8', newline='') as subfile:
4023 subfile.write(sub_info['data'])
4024 sub_info['filepath'] = sub_filename
4025 ret.append((sub_filename, sub_filename_final))
4026 continue
4027 except OSError:
4028 self.report_error(f'Cannot write video subtitles file {sub_filename}')
4029 return None
4030
4031 try:
4032 sub_copy = sub_info.copy()
4033 sub_copy.setdefault('http_headers', info_dict.get('http_headers'))
4034 self.dl(sub_filename, sub_copy, subtitle=True)
4035 sub_info['filepath'] = sub_filename
4036 ret.append((sub_filename, sub_filename_final))
4037 except (DownloadError, ExtractorError, IOError, OSError, ValueError) + network_exceptions as err:
4038 msg = f'Unable to download video subtitles for {sub_lang!r}: {err}'
4039 if self.params.get('ignoreerrors') is not True: # False or 'only_download'
4040 if not self.params.get('ignoreerrors'):
4041 self.report_error(msg)
4042 raise DownloadError(msg)
4043 self.report_warning(msg)
4044 return ret
4045
4046 def _write_thumbnails(self, label, info_dict, filename, thumb_filename_base=None):
4047 ''' Write thumbnails to file and return list of (thumb_filename, final_thumb_filename) '''
4048 write_all = self.params.get('write_all_thumbnails', False)
4049 thumbnails, ret = [], []
4050 if write_all or self.params.get('writethumbnail', False):
4051 thumbnails = info_dict.get('thumbnails') or []
4052 if not thumbnails:
4053 self.to_screen(f'[info] There are no {label} thumbnails to download')
4054 return ret
4055 multiple = write_all and len(thumbnails) > 1
4056
4057 if thumb_filename_base is None:
4058 thumb_filename_base = filename
4059 if thumbnails and not thumb_filename_base:
4060 self.write_debug(f'Skipping writing {label} thumbnail')
4061 return ret
4062
4063 for idx, t in list(enumerate(thumbnails))[::-1]:
4064 thumb_ext = (f'{t["id"]}.' if multiple else '') + determine_ext(t['url'], 'jpg')
4065 thumb_display_id = f'{label} thumbnail {t["id"]}'
4066 thumb_filename = replace_extension(filename, thumb_ext, info_dict.get('ext'))
4067 thumb_filename_final = replace_extension(thumb_filename_base, thumb_ext, info_dict.get('ext'))
4068
4069 existing_thumb = self.existing_file((thumb_filename_final, thumb_filename))
4070 if existing_thumb:
4071 self.to_screen('[info] %s is already present' % (
4072 thumb_display_id if multiple else f'{label} thumbnail').capitalize())
4073 t['filepath'] = existing_thumb
4074 ret.append((existing_thumb, thumb_filename_final))
4075 else:
4076 self.to_screen(f'[info] Downloading {thumb_display_id} ...')
4077 try:
4078 uf = self.urlopen(sanitized_Request(t['url'], headers=t.get('http_headers', {})))
4079 self.to_screen(f'[info] Writing {thumb_display_id} to: {thumb_filename}')
4080 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
4081 shutil.copyfileobj(uf, thumbf)
4082 ret.append((thumb_filename, thumb_filename_final))
4083 t['filepath'] = thumb_filename
4084 except network_exceptions as err:
4085 thumbnails.pop(idx)
4086 self.report_warning(f'Unable to download {thumb_display_id}: {err}')
4087 if ret and not write_all:
4088 break
4089 return ret