]> jfr.im git - yt-dlp.git/blob - yt_dlp/YoutubeDL.py
[cleanup, utils] Split into submodules (#7090)
[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 remove_terminal_sequences,
128 render_table,
129 replace_extension,
130 sanitize_filename,
131 sanitize_path,
132 sanitize_url,
133 sanitized_Request,
134 std_headers,
135 str_or_none,
136 strftime_or_none,
137 subtitles_filename,
138 supports_terminal_sequences,
139 system_identifier,
140 timetuple_from_msec,
141 to_high_limit_path,
142 traverse_obj,
143 try_call,
144 try_get,
145 url_basename,
146 variadic,
147 version_tuple,
148 windows_enable_vt_mode,
149 write_json_file,
150 write_string,
151 )
152 from .version import CHANNEL, RELEASE_GIT_HEAD, VARIANT, __version__
153
154 if compat_os_name == 'nt':
155 import ctypes
156
157
158 class YoutubeDL:
159 """YoutubeDL class.
160
161 YoutubeDL objects are the ones responsible of downloading the
162 actual video file and writing it to disk if the user has requested
163 it, among some other tasks. In most cases there should be one per
164 program. As, given a video URL, the downloader doesn't know how to
165 extract all the needed information, task that InfoExtractors do, it
166 has to pass the URL to one of them.
167
168 For this, YoutubeDL objects have a method that allows
169 InfoExtractors to be registered in a given order. When it is passed
170 a URL, the YoutubeDL object handles it to the first InfoExtractor it
171 finds that reports being able to handle it. The InfoExtractor extracts
172 all the information about the video or videos the URL refers to, and
173 YoutubeDL process the extracted information, possibly using a File
174 Downloader to download the video.
175
176 YoutubeDL objects accept a lot of parameters. In order not to saturate
177 the object constructor with arguments, it receives a dictionary of
178 options instead. These options are available through the params
179 attribute for the InfoExtractors to use. The YoutubeDL also
180 registers itself as the downloader in charge for the InfoExtractors
181 that are added to it, so this is a "mutual registration".
182
183 Available options:
184
185 username: Username for authentication purposes.
186 password: Password for authentication purposes.
187 videopassword: Password for accessing a video.
188 ap_mso: Adobe Pass multiple-system operator identifier.
189 ap_username: Multiple-system operator account username.
190 ap_password: Multiple-system operator account password.
191 usenetrc: Use netrc for authentication instead.
192 netrc_location: Location of the netrc file. Defaults to ~/.netrc.
193 verbose: Print additional info to stdout.
194 quiet: Do not print messages to stdout.
195 no_warnings: Do not print out anything for warnings.
196 forceprint: A dict with keys WHEN mapped to a list of templates to
197 print to stdout. The allowed keys are video or any of the
198 items in utils.POSTPROCESS_WHEN.
199 For compatibility, a single list is also accepted
200 print_to_file: A dict with keys WHEN (same as forceprint) mapped to
201 a list of tuples with (template, filename)
202 forcejson: Force printing info_dict as JSON.
203 dump_single_json: Force printing the info_dict of the whole playlist
204 (or video) as a single JSON line.
205 force_write_download_archive: Force writing download archive regardless
206 of 'skip_download' or 'simulate'.
207 simulate: Do not download the video files. If unset (or None),
208 simulate only if listsubtitles, listformats or list_thumbnails is used
209 format: Video format code. see "FORMAT SELECTION" for more details.
210 You can also pass a function. The function takes 'ctx' as
211 argument and returns the formats to download.
212 See "build_format_selector" for an implementation
213 allow_unplayable_formats: Allow unplayable formats to be extracted and downloaded.
214 ignore_no_formats_error: Ignore "No video formats" error. Usefull for
215 extracting metadata even if the video is not actually
216 available for download (experimental)
217 format_sort: A list of fields by which to sort the video formats.
218 See "Sorting Formats" for more details.
219 format_sort_force: Force the given format_sort. see "Sorting Formats"
220 for more details.
221 prefer_free_formats: Whether to prefer video formats with free containers
222 over non-free ones of same quality.
223 allow_multiple_video_streams: Allow multiple video streams to be merged
224 into a single file
225 allow_multiple_audio_streams: Allow multiple audio streams to be merged
226 into a single file
227 check_formats Whether to test if the formats are downloadable.
228 Can be True (check all), False (check none),
229 'selected' (check selected formats),
230 or None (check only if requested by extractor)
231 paths: Dictionary of output paths. The allowed keys are 'home'
232 'temp' and the keys of OUTTMPL_TYPES (in utils.py)
233 outtmpl: Dictionary of templates for output names. Allowed keys
234 are 'default' and the keys of OUTTMPL_TYPES (in utils.py).
235 For compatibility with youtube-dl, a single string can also be used
236 outtmpl_na_placeholder: Placeholder for unavailable meta fields.
237 restrictfilenames: Do not allow "&" and spaces in file names
238 trim_file_name: Limit length of filename (extension excluded)
239 windowsfilenames: Force the filenames to be windows compatible
240 ignoreerrors: Do not stop on download/postprocessing errors.
241 Can be 'only_download' to ignore only download errors.
242 Default is 'only_download' for CLI, but False for API
243 skip_playlist_after_errors: Number of allowed failures until the rest of
244 the playlist is skipped
245 allowed_extractors: List of regexes to match against extractor names that are allowed
246 overwrites: Overwrite all video and metadata files if True,
247 overwrite only non-video files if None
248 and don't overwrite any file if False
249 For compatibility with youtube-dl,
250 "nooverwrites" may also be used instead
251 playlist_items: Specific indices of playlist to download.
252 playlistrandom: Download playlist items in random order.
253 lazy_playlist: Process playlist entries as they are received.
254 matchtitle: Download only matching titles.
255 rejecttitle: Reject downloads for matching titles.
256 logger: Log messages to a logging.Logger instance.
257 logtostderr: Print everything to stderr instead of stdout.
258 consoletitle: Display progress in console window's titlebar.
259 writedescription: Write the video description to a .description file
260 writeinfojson: Write the video description to a .info.json file
261 clean_infojson: Remove private fields from the infojson
262 getcomments: Extract video comments. This will not be written to disk
263 unless writeinfojson is also given
264 writeannotations: Write the video annotations to a .annotations.xml file
265 writethumbnail: Write the thumbnail image to a file
266 allow_playlist_files: Whether to write playlists' description, infojson etc
267 also to disk when using the 'write*' options
268 write_all_thumbnails: Write all thumbnail formats to files
269 writelink: Write an internet shortcut file, depending on the
270 current platform (.url/.webloc/.desktop)
271 writeurllink: Write a Windows internet shortcut file (.url)
272 writewebloclink: Write a macOS internet shortcut file (.webloc)
273 writedesktoplink: Write a Linux internet shortcut file (.desktop)
274 writesubtitles: Write the video subtitles to a file
275 writeautomaticsub: Write the automatically generated subtitles to a file
276 listsubtitles: Lists all available subtitles for the video
277 subtitlesformat: The format code for subtitles
278 subtitleslangs: List of languages of the subtitles to download (can be regex).
279 The list may contain "all" to refer to all the available
280 subtitles. The language can be prefixed with a "-" to
281 exclude it from the requested languages, e.g. ['all', '-live_chat']
282 keepvideo: Keep the video file after post-processing
283 daterange: A DateRange object, download only if the upload_date is in the range.
284 skip_download: Skip the actual download of the video file
285 cachedir: Location of the cache files in the filesystem.
286 False to disable filesystem cache.
287 noplaylist: Download single video instead of a playlist if in doubt.
288 age_limit: An integer representing the user's age in years.
289 Unsuitable videos for the given age are skipped.
290 min_views: An integer representing the minimum view count the video
291 must have in order to not be skipped.
292 Videos without view count information are always
293 downloaded. None for no limit.
294 max_views: An integer representing the maximum view count.
295 Videos that are more popular than that are not
296 downloaded.
297 Videos without view count information are always
298 downloaded. None for no limit.
299 download_archive: A set, or the name of a file where all downloads are recorded.
300 Videos already present in the file are not downloaded again.
301 break_on_existing: Stop the download process after attempting to download a
302 file that is in the archive.
303 break_per_url: Whether break_on_reject and break_on_existing
304 should act on each input URL as opposed to for the entire queue
305 cookiefile: File name or text stream from where cookies should be read and dumped to
306 cookiesfrombrowser: A tuple containing the name of the browser, the profile
307 name/path from where cookies are loaded, the name of the keyring,
308 and the container name, e.g. ('chrome', ) or
309 ('vivaldi', 'default', 'BASICTEXT') or ('firefox', 'default', None, 'Meta')
310 legacyserverconnect: Explicitly allow HTTPS connection to servers that do not
311 support RFC 5746 secure renegotiation
312 nocheckcertificate: Do not verify SSL certificates
313 client_certificate: Path to client certificate file in PEM format. May include the private key
314 client_certificate_key: Path to private key file for client certificate
315 client_certificate_password: Password for client certificate private key, if encrypted.
316 If not provided and the key is encrypted, yt-dlp will ask interactively
317 prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
318 (Only supported by some extractors)
319 enable_file_urls: Enable file:// URLs. This is disabled by default for security reasons.
320 http_headers: A dictionary of custom headers to be used for all requests
321 proxy: URL of the proxy server to use
322 geo_verification_proxy: URL of the proxy to use for IP address verification
323 on geo-restricted sites.
324 socket_timeout: Time to wait for unresponsive hosts, in seconds
325 bidi_workaround: Work around buggy terminals without bidirectional text
326 support, using fridibi
327 debug_printtraffic:Print out sent and received HTTP traffic
328 default_search: Prepend this string if an input url is not valid.
329 'auto' for elaborate guessing
330 encoding: Use this encoding instead of the system-specified.
331 extract_flat: Whether to resolve and process url_results further
332 * False: Always process (default)
333 * True: Never process
334 * 'in_playlist': Do not process inside playlist/multi_video
335 * 'discard': Always process, but don't return the result
336 from inside playlist/multi_video
337 * 'discard_in_playlist': Same as "discard", but only for
338 playlists (not multi_video)
339 wait_for_video: If given, wait for scheduled streams to become available.
340 The value should be a tuple containing the range
341 (min_secs, max_secs) to wait between retries
342 postprocessors: A list of dictionaries, each with an entry
343 * key: The name of the postprocessor. See
344 yt_dlp/postprocessor/__init__.py for a list.
345 * when: When to run the postprocessor. Allowed values are
346 the entries of utils.POSTPROCESS_WHEN
347 Assumed to be 'post_process' if not given
348 progress_hooks: A list of functions that get called on download
349 progress, with a dictionary with the entries
350 * status: One of "downloading", "error", or "finished".
351 Check this first and ignore unknown values.
352 * info_dict: The extracted info_dict
353
354 If status is one of "downloading", or "finished", the
355 following properties may also be present:
356 * filename: The final filename (always present)
357 * tmpfilename: The filename we're currently writing to
358 * downloaded_bytes: Bytes on disk
359 * total_bytes: Size of the whole file, None if unknown
360 * total_bytes_estimate: Guess of the eventual file size,
361 None if unavailable.
362 * elapsed: The number of seconds since download started.
363 * eta: The estimated time in seconds, None if unknown
364 * speed: The download speed in bytes/second, None if
365 unknown
366 * fragment_index: The counter of the currently
367 downloaded video fragment.
368 * fragment_count: The number of fragments (= individual
369 files that will be merged)
370
371 Progress hooks are guaranteed to be called at least once
372 (with status "finished") if the download is successful.
373 postprocessor_hooks: A list of functions that get called on postprocessing
374 progress, with a dictionary with the entries
375 * status: One of "started", "processing", or "finished".
376 Check this first and ignore unknown values.
377 * postprocessor: Name of the postprocessor
378 * info_dict: The extracted info_dict
379
380 Progress hooks are guaranteed to be called at least twice
381 (with status "started" and "finished") if the processing is successful.
382 merge_output_format: "/" separated list of extensions to use when merging formats.
383 final_ext: Expected final extension; used to detect when the file was
384 already downloaded and converted
385 fixup: Automatically correct known faults of the file.
386 One of:
387 - "never": do nothing
388 - "warn": only emit a warning
389 - "detect_or_warn": check whether we can do anything
390 about it, warn otherwise (default)
391 source_address: Client-side IP address to bind to.
392 sleep_interval_requests: Number of seconds to sleep between requests
393 during extraction
394 sleep_interval: Number of seconds to sleep before each download when
395 used alone or a lower bound of a range for randomized
396 sleep before each download (minimum possible number
397 of seconds to sleep) when used along with
398 max_sleep_interval.
399 max_sleep_interval:Upper bound of a range for randomized sleep before each
400 download (maximum possible number of seconds to sleep).
401 Must only be used along with sleep_interval.
402 Actual sleep time will be a random float from range
403 [sleep_interval; max_sleep_interval].
404 sleep_interval_subtitles: Number of seconds to sleep before each subtitle download
405 listformats: Print an overview of available video formats and exit.
406 list_thumbnails: Print a table of all thumbnails and exit.
407 match_filter: A function that gets called for every video with the signature
408 (info_dict, *, incomplete: bool) -> Optional[str]
409 For backward compatibility with youtube-dl, the signature
410 (info_dict) -> Optional[str] is also allowed.
411 - If it returns a message, the video is ignored.
412 - If it returns None, the video is downloaded.
413 - If it returns utils.NO_DEFAULT, the user is interactively
414 asked whether to download the video.
415 - Raise utils.DownloadCancelled(msg) to abort remaining
416 downloads when a video is rejected.
417 match_filter_func in utils.py is one example for this.
418 no_color: Do not emit color codes in output.
419 geo_bypass: Bypass geographic restriction via faking X-Forwarded-For
420 HTTP header
421 geo_bypass_country:
422 Two-letter ISO 3166-2 country code that will be used for
423 explicit geographic restriction bypassing via faking
424 X-Forwarded-For HTTP header
425 geo_bypass_ip_block:
426 IP range in CIDR notation that will be used similarly to
427 geo_bypass_country
428 external_downloader: A dictionary of protocol keys and the executable of the
429 external downloader to use for it. The allowed protocols
430 are default|http|ftp|m3u8|dash|rtsp|rtmp|mms.
431 Set the value to 'native' to use the native downloader
432 compat_opts: Compatibility options. See "Differences in default behavior".
433 The following options do not work when used through the API:
434 filename, abort-on-error, multistreams, no-live-chat, format-sort
435 no-clean-infojson, no-playlist-metafiles, no-keep-subs, no-attach-info-json.
436 Refer __init__.py for their implementation
437 progress_template: Dictionary of templates for progress outputs.
438 Allowed keys are 'download', 'postprocess',
439 'download-title' (console title) and 'postprocess-title'.
440 The template is mapped on a dictionary with keys 'progress' and 'info'
441 retry_sleep_functions: Dictionary of functions that takes the number of attempts
442 as argument and returns the time to sleep in seconds.
443 Allowed keys are 'http', 'fragment', 'file_access'
444 download_ranges: A callback function that gets called for every video with
445 the signature (info_dict, ydl) -> Iterable[Section].
446 Only the returned sections will be downloaded.
447 Each Section is a dict with the following keys:
448 * start_time: Start time of the section in seconds
449 * end_time: End time of the section in seconds
450 * title: Section title (Optional)
451 * index: Section number (Optional)
452 force_keyframes_at_cuts: Re-encode the video when downloading ranges to get precise cuts
453 noprogress: Do not print the progress bar
454 live_from_start: Whether to download livestreams videos from the start
455
456 The following parameters are not used by YoutubeDL itself, they are used by
457 the downloader (see yt_dlp/downloader/common.py):
458 nopart, updatetime, buffersize, ratelimit, throttledratelimit, min_filesize,
459 max_filesize, test, noresizebuffer, retries, file_access_retries, fragment_retries,
460 continuedl, xattr_set_filesize, hls_use_mpegts, http_chunk_size,
461 external_downloader_args, concurrent_fragment_downloads.
462
463 The following options are used by the post processors:
464 ffmpeg_location: Location of the ffmpeg/avconv binary; either the path
465 to the binary or its containing directory.
466 postprocessor_args: A dictionary of postprocessor/executable keys (in lower case)
467 and a list of additional command-line arguments for the
468 postprocessor/executable. The dict can also have "PP+EXE" keys
469 which are used when the given exe is used by the given PP.
470 Use 'default' as the name for arguments to passed to all PP
471 For compatibility with youtube-dl, a single list of args
472 can also be used
473
474 The following options are used by the extractors:
475 extractor_retries: Number of times to retry for known errors
476 dynamic_mpd: Whether to process dynamic DASH manifests (default: True)
477 hls_split_discontinuity: Split HLS playlists to different formats at
478 discontinuities such as ad breaks (default: False)
479 extractor_args: A dictionary of arguments to be passed to the extractors.
480 See "EXTRACTOR ARGUMENTS" for details.
481 E.g. {'youtube': {'skip': ['dash', 'hls']}}
482 mark_watched: Mark videos watched (even with --simulate). Only for YouTube
483
484 The following options are deprecated and may be removed in the future:
485
486 break_on_reject: Stop the download process when encountering a video that
487 has been filtered out.
488 - `raise DownloadCancelled(msg)` in match_filter instead
489 force_generic_extractor: Force downloader to use the generic extractor
490 - Use allowed_extractors = ['generic', 'default']
491 playliststart: - Use playlist_items
492 Playlist item to start at.
493 playlistend: - Use playlist_items
494 Playlist item to end at.
495 playlistreverse: - Use playlist_items
496 Download playlist items in reverse order.
497 forceurl: - Use forceprint
498 Force printing final URL.
499 forcetitle: - Use forceprint
500 Force printing title.
501 forceid: - Use forceprint
502 Force printing ID.
503 forcethumbnail: - Use forceprint
504 Force printing thumbnail URL.
505 forcedescription: - Use forceprint
506 Force printing description.
507 forcefilename: - Use forceprint
508 Force printing final filename.
509 forceduration: - Use forceprint
510 Force printing duration.
511 allsubtitles: - Use subtitleslangs = ['all']
512 Downloads all the subtitles of the video
513 (requires writesubtitles or writeautomaticsub)
514 include_ads: - Doesn't work
515 Download ads as well
516 call_home: - Not implemented
517 Boolean, true iff we are allowed to contact the
518 yt-dlp servers for debugging.
519 post_hooks: - Register a custom postprocessor
520 A list of functions that get called as the final step
521 for each video file, after all postprocessors have been
522 called. The filename will be passed as the only argument.
523 hls_prefer_native: - Use external_downloader = {'m3u8': 'native'} or {'m3u8': 'ffmpeg'}.
524 Use the native HLS downloader instead of ffmpeg/avconv
525 if True, otherwise use ffmpeg/avconv if False, otherwise
526 use downloader suggested by extractor if None.
527 prefer_ffmpeg: - avconv support is deprecated
528 If False, use avconv instead of ffmpeg if both are available,
529 otherwise prefer ffmpeg.
530 youtube_include_dash_manifest: - Use extractor_args
531 If True (default), DASH manifests and related
532 data will be downloaded and processed by extractor.
533 You can reduce network I/O by disabling it if you don't
534 care about DASH. (only for youtube)
535 youtube_include_hls_manifest: - Use extractor_args
536 If True (default), HLS manifests and related
537 data will be downloaded and processed by extractor.
538 You can reduce network I/O by disabling it if you don't
539 care about HLS. (only for youtube)
540 """
541
542 _NUMERIC_FIELDS = {
543 'width', 'height', 'asr', 'audio_channels', 'fps',
544 'tbr', 'abr', 'vbr', 'filesize', 'filesize_approx',
545 'timestamp', 'release_timestamp',
546 'duration', 'view_count', 'like_count', 'dislike_count', 'repost_count',
547 'average_rating', 'comment_count', 'age_limit',
548 'start_time', 'end_time',
549 'chapter_number', 'season_number', 'episode_number',
550 'track_number', 'disc_number', 'release_year',
551 }
552
553 _format_fields = {
554 # NB: Keep in sync with the docstring of extractor/common.py
555 'url', 'manifest_url', 'manifest_stream_number', 'ext', 'format', 'format_id', 'format_note',
556 'width', 'height', 'aspect_ratio', 'resolution', 'dynamic_range', 'tbr', 'abr', 'acodec', 'asr', 'audio_channels',
557 'vbr', 'fps', 'vcodec', 'container', 'filesize', 'filesize_approx', 'rows', 'columns',
558 'player_url', 'protocol', 'fragment_base_url', 'fragments', 'is_from_start',
559 'preference', 'language', 'language_preference', 'quality', 'source_preference',
560 'http_headers', 'stretched_ratio', 'no_resume', 'has_drm', 'extra_param_to_segment_url', 'hls_aes', 'downloader_options',
561 'page_url', 'app', 'play_path', 'tc_url', 'flash_version', 'rtmp_live', 'rtmp_conn', 'rtmp_protocol', 'rtmp_real_time'
562 }
563 _format_selection_exts = {
564 'audio': set(MEDIA_EXTENSIONS.common_audio),
565 'video': set(MEDIA_EXTENSIONS.common_video + ('3gp', )),
566 'storyboards': set(MEDIA_EXTENSIONS.storyboards),
567 }
568
569 def __init__(self, params=None, auto_init=True):
570 """Create a FileDownloader object with the given options.
571 @param auto_init Whether to load the default extractors and print header (if verbose).
572 Set to 'no_verbose_header' to not print the header
573 """
574 if params is None:
575 params = {}
576 self.params = params
577 self._ies = {}
578 self._ies_instances = {}
579 self._pps = {k: [] for k in POSTPROCESS_WHEN}
580 self._printed_messages = set()
581 self._first_webpage_request = True
582 self._post_hooks = []
583 self._progress_hooks = []
584 self._postprocessor_hooks = []
585 self._download_retcode = 0
586 self._num_downloads = 0
587 self._num_videos = 0
588 self._playlist_level = 0
589 self._playlist_urls = set()
590 self.cache = Cache(self)
591
592 stdout = sys.stderr if self.params.get('logtostderr') else sys.stdout
593 self._out_files = Namespace(
594 out=stdout,
595 error=sys.stderr,
596 screen=sys.stderr if self.params.get('quiet') else stdout,
597 console=None if compat_os_name == 'nt' else next(
598 filter(supports_terminal_sequences, (sys.stderr, sys.stdout)), None)
599 )
600
601 try:
602 windows_enable_vt_mode()
603 except Exception as e:
604 self.write_debug(f'Failed to enable VT mode: {e}')
605
606 self._allow_colors = Namespace(**{
607 type_: not self.params.get('no_color') and supports_terminal_sequences(stream)
608 for type_, stream in self._out_files.items_ if type_ != 'console'
609 })
610
611 # The code is left like this to be reused for future deprecations
612 MIN_SUPPORTED, MIN_RECOMMENDED = (3, 7), (3, 7)
613 current_version = sys.version_info[:2]
614 if current_version < MIN_RECOMMENDED:
615 msg = ('Support for Python version %d.%d has been deprecated. '
616 'See https://github.com/yt-dlp/yt-dlp/issues/3764 for more details.'
617 '\n You will no longer receive updates on this version')
618 if current_version < MIN_SUPPORTED:
619 msg = 'Python version %d.%d is no longer supported'
620 self.deprecated_feature(
621 f'{msg}! Please update to Python %d.%d or above' % (*current_version, *MIN_RECOMMENDED))
622
623 if self.params.get('allow_unplayable_formats'):
624 self.report_warning(
625 f'You have asked for {self._format_err("UNPLAYABLE", self.Styles.EMPHASIS)} formats to be listed/downloaded. '
626 'This is a developer option intended for debugging. \n'
627 ' If you experience any issues while using this option, '
628 f'{self._format_err("DO NOT", self.Styles.ERROR)} open a bug report')
629
630 if self.params.get('bidi_workaround', False):
631 try:
632 import pty
633 master, slave = pty.openpty()
634 width = shutil.get_terminal_size().columns
635 width_args = [] if width is None else ['-w', str(width)]
636 sp_kwargs = {'stdin': subprocess.PIPE, 'stdout': slave, 'stderr': self._out_files.error}
637 try:
638 self._output_process = Popen(['bidiv'] + width_args, **sp_kwargs)
639 except OSError:
640 self._output_process = Popen(['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
641 self._output_channel = os.fdopen(master, 'rb')
642 except OSError as ose:
643 if ose.errno == errno.ENOENT:
644 self.report_warning(
645 'Could not find fribidi executable, ignoring --bidi-workaround. '
646 'Make sure that fribidi is an executable file in one of the directories in your $PATH.')
647 else:
648 raise
649
650 self.params['compat_opts'] = set(self.params.get('compat_opts', ()))
651 if auto_init and auto_init != 'no_verbose_header':
652 self.print_debug_header()
653
654 def check_deprecated(param, option, suggestion):
655 if self.params.get(param) is not None:
656 self.report_warning(f'{option} is deprecated. Use {suggestion} instead')
657 return True
658 return False
659
660 if check_deprecated('cn_verification_proxy', '--cn-verification-proxy', '--geo-verification-proxy'):
661 if self.params.get('geo_verification_proxy') is None:
662 self.params['geo_verification_proxy'] = self.params['cn_verification_proxy']
663
664 check_deprecated('autonumber', '--auto-number', '-o "%(autonumber)s-%(title)s.%(ext)s"')
665 check_deprecated('usetitle', '--title', '-o "%(title)s-%(id)s.%(ext)s"')
666 check_deprecated('useid', '--id', '-o "%(id)s.%(ext)s"')
667
668 for msg in self.params.get('_warnings', []):
669 self.report_warning(msg)
670 for msg in self.params.get('_deprecation_warnings', []):
671 self.deprecated_feature(msg)
672
673 if 'list-formats' in self.params['compat_opts']:
674 self.params['listformats_table'] = False
675
676 if 'overwrites' not in self.params and self.params.get('nooverwrites') is not None:
677 # nooverwrites was unnecessarily changed to overwrites
678 # in 0c3d0f51778b153f65c21906031c2e091fcfb641
679 # This ensures compatibility with both keys
680 self.params['overwrites'] = not self.params['nooverwrites']
681 elif self.params.get('overwrites') is None:
682 self.params.pop('overwrites', None)
683 else:
684 self.params['nooverwrites'] = not self.params['overwrites']
685
686 if self.params.get('simulate') is None and any((
687 self.params.get('list_thumbnails'),
688 self.params.get('listformats'),
689 self.params.get('listsubtitles'),
690 )):
691 self.params['simulate'] = 'list_only'
692
693 self.params.setdefault('forceprint', {})
694 self.params.setdefault('print_to_file', {})
695
696 # Compatibility with older syntax
697 if not isinstance(params['forceprint'], dict):
698 self.params['forceprint'] = {'video': params['forceprint']}
699
700 if auto_init:
701 self.add_default_info_extractors()
702
703 if (sys.platform != 'win32'
704 and sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968']
705 and not self.params.get('restrictfilenames', False)):
706 # Unicode filesystem API will throw errors (#1474, #13027)
707 self.report_warning(
708 'Assuming --restrict-filenames since file system encoding '
709 'cannot encode all characters. '
710 'Set the LC_ALL environment variable to fix this.')
711 self.params['restrictfilenames'] = True
712
713 self._parse_outtmpl()
714
715 # Creating format selector here allows us to catch syntax errors before the extraction
716 self.format_selector = (
717 self.params.get('format') if self.params.get('format') in (None, '-')
718 else self.params['format'] if callable(self.params['format'])
719 else self.build_format_selector(self.params['format']))
720
721 # Set http_headers defaults according to std_headers
722 self.params['http_headers'] = merge_headers(std_headers, self.params.get('http_headers', {}))
723
724 hooks = {
725 'post_hooks': self.add_post_hook,
726 'progress_hooks': self.add_progress_hook,
727 'postprocessor_hooks': self.add_postprocessor_hook,
728 }
729 for opt, fn in hooks.items():
730 for ph in self.params.get(opt, []):
731 fn(ph)
732
733 for pp_def_raw in self.params.get('postprocessors', []):
734 pp_def = dict(pp_def_raw)
735 when = pp_def.pop('when', 'post_process')
736 self.add_post_processor(
737 get_postprocessor(pp_def.pop('key'))(self, **pp_def),
738 when=when)
739
740 self._setup_opener()
741
742 def preload_download_archive(fn):
743 """Preload the archive, if any is specified"""
744 archive = set()
745 if fn is None:
746 return archive
747 elif not is_path_like(fn):
748 return fn
749
750 self.write_debug(f'Loading archive file {fn!r}')
751 try:
752 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
753 for line in archive_file:
754 archive.add(line.strip())
755 except OSError as ioe:
756 if ioe.errno != errno.ENOENT:
757 raise
758 return archive
759
760 self.archive = preload_download_archive(self.params.get('download_archive'))
761
762 def warn_if_short_id(self, argv):
763 # short YouTube ID starting with dash?
764 idxs = [
765 i for i, a in enumerate(argv)
766 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
767 if idxs:
768 correct_argv = (
769 ['yt-dlp']
770 + [a for i, a in enumerate(argv) if i not in idxs]
771 + ['--'] + [argv[i] for i in idxs]
772 )
773 self.report_warning(
774 'Long argument string detected. '
775 'Use -- to separate parameters and URLs, like this:\n%s' %
776 args_to_str(correct_argv))
777
778 def add_info_extractor(self, ie):
779 """Add an InfoExtractor object to the end of the list."""
780 ie_key = ie.ie_key()
781 self._ies[ie_key] = ie
782 if not isinstance(ie, type):
783 self._ies_instances[ie_key] = ie
784 ie.set_downloader(self)
785
786 def get_info_extractor(self, ie_key):
787 """
788 Get an instance of an IE with name ie_key, it will try to get one from
789 the _ies list, if there's no instance it will create a new one and add
790 it to the extractor list.
791 """
792 ie = self._ies_instances.get(ie_key)
793 if ie is None:
794 ie = get_info_extractor(ie_key)()
795 self.add_info_extractor(ie)
796 return ie
797
798 def add_default_info_extractors(self):
799 """
800 Add the InfoExtractors returned by gen_extractors to the end of the list
801 """
802 all_ies = {ie.IE_NAME.lower(): ie for ie in gen_extractor_classes()}
803 all_ies['end'] = UnsupportedURLIE()
804 try:
805 ie_names = orderedSet_from_options(
806 self.params.get('allowed_extractors', ['default']), {
807 'all': list(all_ies),
808 'default': [name for name, ie in all_ies.items() if ie._ENABLED],
809 }, use_regex=True)
810 except re.error as e:
811 raise ValueError(f'Wrong regex for allowed_extractors: {e.pattern}')
812 for name in ie_names:
813 self.add_info_extractor(all_ies[name])
814 self.write_debug(f'Loaded {len(ie_names)} extractors')
815
816 def add_post_processor(self, pp, when='post_process'):
817 """Add a PostProcessor object to the end of the chain."""
818 assert when in POSTPROCESS_WHEN, f'Invalid when={when}'
819 self._pps[when].append(pp)
820 pp.set_downloader(self)
821
822 def add_post_hook(self, ph):
823 """Add the post hook"""
824 self._post_hooks.append(ph)
825
826 def add_progress_hook(self, ph):
827 """Add the download progress hook"""
828 self._progress_hooks.append(ph)
829
830 def add_postprocessor_hook(self, ph):
831 """Add the postprocessing progress hook"""
832 self._postprocessor_hooks.append(ph)
833 for pps in self._pps.values():
834 for pp in pps:
835 pp.add_progress_hook(ph)
836
837 def _bidi_workaround(self, message):
838 if not hasattr(self, '_output_channel'):
839 return message
840
841 assert hasattr(self, '_output_process')
842 assert isinstance(message, str)
843 line_count = message.count('\n') + 1
844 self._output_process.stdin.write((message + '\n').encode())
845 self._output_process.stdin.flush()
846 res = ''.join(self._output_channel.readline().decode()
847 for _ in range(line_count))
848 return res[:-len('\n')]
849
850 def _write_string(self, message, out=None, only_once=False):
851 if only_once:
852 if message in self._printed_messages:
853 return
854 self._printed_messages.add(message)
855 write_string(message, out=out, encoding=self.params.get('encoding'))
856
857 def to_stdout(self, message, skip_eol=False, quiet=None):
858 """Print message to stdout"""
859 if quiet is not None:
860 self.deprecation_warning('"YoutubeDL.to_stdout" no longer accepts the argument quiet. '
861 'Use "YoutubeDL.to_screen" instead')
862 if skip_eol is not False:
863 self.deprecation_warning('"YoutubeDL.to_stdout" no longer accepts the argument skip_eol. '
864 'Use "YoutubeDL.to_screen" instead')
865 self._write_string(f'{self._bidi_workaround(message)}\n', self._out_files.out)
866
867 def to_screen(self, message, skip_eol=False, quiet=None, only_once=False):
868 """Print message to screen if not in quiet mode"""
869 if self.params.get('logger'):
870 self.params['logger'].debug(message)
871 return
872 if (self.params.get('quiet') if quiet is None else quiet) and not self.params.get('verbose'):
873 return
874 self._write_string(
875 '%s%s' % (self._bidi_workaround(message), ('' if skip_eol else '\n')),
876 self._out_files.screen, only_once=only_once)
877
878 def to_stderr(self, message, only_once=False):
879 """Print message to stderr"""
880 assert isinstance(message, str)
881 if self.params.get('logger'):
882 self.params['logger'].error(message)
883 else:
884 self._write_string(f'{self._bidi_workaround(message)}\n', self._out_files.error, only_once=only_once)
885
886 def _send_console_code(self, code):
887 if compat_os_name == 'nt' or not self._out_files.console:
888 return
889 self._write_string(code, self._out_files.console)
890
891 def to_console_title(self, message):
892 if not self.params.get('consoletitle', False):
893 return
894 message = remove_terminal_sequences(message)
895 if compat_os_name == 'nt':
896 if ctypes.windll.kernel32.GetConsoleWindow():
897 # c_wchar_p() might not be necessary if `message` is
898 # already of type unicode()
899 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
900 else:
901 self._send_console_code(f'\033]0;{message}\007')
902
903 def save_console_title(self):
904 if not self.params.get('consoletitle') or self.params.get('simulate'):
905 return
906 self._send_console_code('\033[22;0t') # Save the title on stack
907
908 def restore_console_title(self):
909 if not self.params.get('consoletitle') or self.params.get('simulate'):
910 return
911 self._send_console_code('\033[23;0t') # Restore the title from stack
912
913 def __enter__(self):
914 self.save_console_title()
915 return self
916
917 def __exit__(self, *args):
918 self.restore_console_title()
919
920 if self.params.get('cookiefile') is not None:
921 self.cookiejar.save(ignore_discard=True, ignore_expires=True)
922
923 def trouble(self, message=None, tb=None, is_error=True):
924 """Determine action to take when a download problem appears.
925
926 Depending on if the downloader has been configured to ignore
927 download errors or not, this method may throw an exception or
928 not when errors are found, after printing the message.
929
930 @param tb If given, is additional traceback information
931 @param is_error Whether to raise error according to ignorerrors
932 """
933 if message is not None:
934 self.to_stderr(message)
935 if self.params.get('verbose'):
936 if tb is None:
937 if sys.exc_info()[0]: # if .trouble has been called from an except block
938 tb = ''
939 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
940 tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
941 tb += encode_compat_str(traceback.format_exc())
942 else:
943 tb_data = traceback.format_list(traceback.extract_stack())
944 tb = ''.join(tb_data)
945 if tb:
946 self.to_stderr(tb)
947 if not is_error:
948 return
949 if not self.params.get('ignoreerrors'):
950 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
951 exc_info = sys.exc_info()[1].exc_info
952 else:
953 exc_info = sys.exc_info()
954 raise DownloadError(message, exc_info)
955 self._download_retcode = 1
956
957 Styles = Namespace(
958 HEADERS='yellow',
959 EMPHASIS='light blue',
960 FILENAME='green',
961 ID='green',
962 DELIM='blue',
963 ERROR='red',
964 WARNING='yellow',
965 SUPPRESS='light black',
966 )
967
968 def _format_text(self, handle, allow_colors, text, f, fallback=None, *, test_encoding=False):
969 text = str(text)
970 if test_encoding:
971 original_text = text
972 # handle.encoding can be None. See https://github.com/yt-dlp/yt-dlp/issues/2711
973 encoding = self.params.get('encoding') or getattr(handle, 'encoding', None) or 'ascii'
974 text = text.encode(encoding, 'ignore').decode(encoding)
975 if fallback is not None and text != original_text:
976 text = fallback
977 return format_text(text, f) if allow_colors else text if fallback is None else fallback
978
979 def _format_out(self, *args, **kwargs):
980 return self._format_text(self._out_files.out, self._allow_colors.out, *args, **kwargs)
981
982 def _format_screen(self, *args, **kwargs):
983 return self._format_text(self._out_files.screen, self._allow_colors.screen, *args, **kwargs)
984
985 def _format_err(self, *args, **kwargs):
986 return self._format_text(self._out_files.error, self._allow_colors.error, *args, **kwargs)
987
988 def report_warning(self, message, only_once=False):
989 '''
990 Print the message to stderr, it will be prefixed with 'WARNING:'
991 If stderr is a tty file the 'WARNING:' will be colored
992 '''
993 if self.params.get('logger') is not None:
994 self.params['logger'].warning(message)
995 else:
996 if self.params.get('no_warnings'):
997 return
998 self.to_stderr(f'{self._format_err("WARNING:", self.Styles.WARNING)} {message}', only_once)
999
1000 def deprecation_warning(self, message, *, stacklevel=0):
1001 deprecation_warning(
1002 message, stacklevel=stacklevel + 1, printer=self.report_error, is_error=False)
1003
1004 def deprecated_feature(self, message):
1005 if self.params.get('logger') is not None:
1006 self.params['logger'].warning(f'Deprecated Feature: {message}')
1007 self.to_stderr(f'{self._format_err("Deprecated Feature:", self.Styles.ERROR)} {message}', True)
1008
1009 def report_error(self, message, *args, **kwargs):
1010 '''
1011 Do the same as trouble, but prefixes the message with 'ERROR:', colored
1012 in red if stderr is a tty file.
1013 '''
1014 self.trouble(f'{self._format_err("ERROR:", self.Styles.ERROR)} {message}', *args, **kwargs)
1015
1016 def write_debug(self, message, only_once=False):
1017 '''Log debug message or Print message to stderr'''
1018 if not self.params.get('verbose', False):
1019 return
1020 message = f'[debug] {message}'
1021 if self.params.get('logger'):
1022 self.params['logger'].debug(message)
1023 else:
1024 self.to_stderr(message, only_once)
1025
1026 def report_file_already_downloaded(self, file_name):
1027 """Report file has already been fully downloaded."""
1028 try:
1029 self.to_screen('[download] %s has already been downloaded' % file_name)
1030 except UnicodeEncodeError:
1031 self.to_screen('[download] The file has already been downloaded')
1032
1033 def report_file_delete(self, file_name):
1034 """Report that existing file will be deleted."""
1035 try:
1036 self.to_screen('Deleting existing file %s' % file_name)
1037 except UnicodeEncodeError:
1038 self.to_screen('Deleting existing file')
1039
1040 def raise_no_formats(self, info, forced=False, *, msg=None):
1041 has_drm = info.get('_has_drm')
1042 ignored, expected = self.params.get('ignore_no_formats_error'), bool(msg)
1043 msg = msg or has_drm and 'This video is DRM protected' or 'No video formats found!'
1044 if forced or not ignored:
1045 raise ExtractorError(msg, video_id=info['id'], ie=info['extractor'],
1046 expected=has_drm or ignored or expected)
1047 else:
1048 self.report_warning(msg)
1049
1050 def parse_outtmpl(self):
1051 self.deprecation_warning('"YoutubeDL.parse_outtmpl" is deprecated and may be removed in a future version')
1052 self._parse_outtmpl()
1053 return self.params['outtmpl']
1054
1055 def _parse_outtmpl(self):
1056 sanitize = IDENTITY
1057 if self.params.get('restrictfilenames'): # Remove spaces in the default template
1058 sanitize = lambda x: x.replace(' - ', ' ').replace(' ', '-')
1059
1060 outtmpl = self.params.setdefault('outtmpl', {})
1061 if not isinstance(outtmpl, dict):
1062 self.params['outtmpl'] = outtmpl = {'default': outtmpl}
1063 outtmpl.update({k: sanitize(v) for k, v in DEFAULT_OUTTMPL.items() if outtmpl.get(k) is None})
1064
1065 def get_output_path(self, dir_type='', filename=None):
1066 paths = self.params.get('paths', {})
1067 assert isinstance(paths, dict), '"paths" parameter must be a dictionary'
1068 path = os.path.join(
1069 expand_path(paths.get('home', '').strip()),
1070 expand_path(paths.get(dir_type, '').strip()) if dir_type else '',
1071 filename or '')
1072 return sanitize_path(path, force=self.params.get('windowsfilenames'))
1073
1074 @staticmethod
1075 def _outtmpl_expandpath(outtmpl):
1076 # expand_path translates '%%' into '%' and '$$' into '$'
1077 # correspondingly that is not what we want since we need to keep
1078 # '%%' intact for template dict substitution step. Working around
1079 # with boundary-alike separator hack.
1080 sep = ''.join(random.choices(string.ascii_letters, k=32))
1081 outtmpl = outtmpl.replace('%%', f'%{sep}%').replace('$$', f'${sep}$')
1082
1083 # outtmpl should be expand_path'ed before template dict substitution
1084 # because meta fields may contain env variables we don't want to
1085 # be expanded. E.g. for outtmpl "%(title)s.%(ext)s" and
1086 # title "Hello $PATH", we don't want `$PATH` to be expanded.
1087 return expand_path(outtmpl).replace(sep, '')
1088
1089 @staticmethod
1090 def escape_outtmpl(outtmpl):
1091 ''' Escape any remaining strings like %s, %abc% etc. '''
1092 return re.sub(
1093 STR_FORMAT_RE_TMPL.format('', '(?![%(\0])'),
1094 lambda mobj: ('' if mobj.group('has_key') else '%') + mobj.group(0),
1095 outtmpl)
1096
1097 @classmethod
1098 def validate_outtmpl(cls, outtmpl):
1099 ''' @return None or Exception object '''
1100 outtmpl = re.sub(
1101 STR_FORMAT_RE_TMPL.format('[^)]*', '[ljhqBUDS]'),
1102 lambda mobj: f'{mobj.group(0)[:-1]}s',
1103 cls._outtmpl_expandpath(outtmpl))
1104 try:
1105 cls.escape_outtmpl(outtmpl) % collections.defaultdict(int)
1106 return None
1107 except ValueError as err:
1108 return err
1109
1110 @staticmethod
1111 def _copy_infodict(info_dict):
1112 info_dict = dict(info_dict)
1113 info_dict.pop('__postprocessors', None)
1114 info_dict.pop('__pending_error', None)
1115 return info_dict
1116
1117 def prepare_outtmpl(self, outtmpl, info_dict, sanitize=False):
1118 """ Make the outtmpl and info_dict suitable for substitution: ydl.escape_outtmpl(outtmpl) % info_dict
1119 @param sanitize Whether to sanitize the output as a filename.
1120 For backward compatibility, a function can also be passed
1121 """
1122
1123 info_dict.setdefault('epoch', int(time.time())) # keep epoch consistent once set
1124
1125 info_dict = self._copy_infodict(info_dict)
1126 info_dict['duration_string'] = ( # %(duration>%H-%M-%S)s is wrong if duration > 24hrs
1127 formatSeconds(info_dict['duration'], '-' if sanitize else ':')
1128 if info_dict.get('duration', None) is not None
1129 else None)
1130 info_dict['autonumber'] = int(self.params.get('autonumber_start', 1) - 1 + self._num_downloads)
1131 info_dict['video_autonumber'] = self._num_videos
1132 if info_dict.get('resolution') is None:
1133 info_dict['resolution'] = self.format_resolution(info_dict, default=None)
1134
1135 # For fields playlist_index, playlist_autonumber and autonumber convert all occurrences
1136 # of %(field)s to %(field)0Nd for backward compatibility
1137 field_size_compat_map = {
1138 'playlist_index': number_of_digits(info_dict.get('__last_playlist_index') or 0),
1139 'playlist_autonumber': number_of_digits(info_dict.get('n_entries') or 0),
1140 'autonumber': self.params.get('autonumber_size') or 5,
1141 }
1142
1143 TMPL_DICT = {}
1144 EXTERNAL_FORMAT_RE = re.compile(STR_FORMAT_RE_TMPL.format('[^)]*', f'[{STR_FORMAT_TYPES}ljhqBUDS]'))
1145 MATH_FUNCTIONS = {
1146 '+': float.__add__,
1147 '-': float.__sub__,
1148 }
1149 # Field is of the form key1.key2...
1150 # where keys (except first) can be string, int, slice or "{field, ...}"
1151 FIELD_INNER_RE = r'(?:\w+|%(num)s|%(num)s?(?::%(num)s?){1,2})' % {'num': r'(?:-?\d+)'}
1152 FIELD_RE = r'\w*(?:\.(?:%(inner)s|{%(field)s(?:,%(field)s)*}))*' % {
1153 'inner': FIELD_INNER_RE,
1154 'field': rf'\w*(?:\.{FIELD_INNER_RE})*'
1155 }
1156 MATH_FIELD_RE = rf'(?:{FIELD_RE}|-?{NUMBER_RE})'
1157 MATH_OPERATORS_RE = r'(?:%s)' % '|'.join(map(re.escape, MATH_FUNCTIONS.keys()))
1158 INTERNAL_FORMAT_RE = re.compile(rf'''(?xs)
1159 (?P<negate>-)?
1160 (?P<fields>{FIELD_RE})
1161 (?P<maths>(?:{MATH_OPERATORS_RE}{MATH_FIELD_RE})*)
1162 (?:>(?P<strf_format>.+?))?
1163 (?P<remaining>
1164 (?P<alternate>(?<!\\),[^|&)]+)?
1165 (?:&(?P<replacement>.*?))?
1166 (?:\|(?P<default>.*?))?
1167 )$''')
1168
1169 def _traverse_infodict(fields):
1170 fields = [f for x in re.split(r'\.({.+?})\.?', fields)
1171 for f in ([x] if x.startswith('{') else x.split('.'))]
1172 for i in (0, -1):
1173 if fields and not fields[i]:
1174 fields.pop(i)
1175
1176 for i, f in enumerate(fields):
1177 if not f.startswith('{'):
1178 continue
1179 assert f.endswith('}'), f'No closing brace for {f} in {fields}'
1180 fields[i] = {k: k.split('.') for k in f[1:-1].split(',')}
1181
1182 return traverse_obj(info_dict, fields, is_user_input=True, traverse_string=True)
1183
1184 def get_value(mdict):
1185 # Object traversal
1186 value = _traverse_infodict(mdict['fields'])
1187 # Negative
1188 if mdict['negate']:
1189 value = float_or_none(value)
1190 if value is not None:
1191 value *= -1
1192 # Do maths
1193 offset_key = mdict['maths']
1194 if offset_key:
1195 value = float_or_none(value)
1196 operator = None
1197 while offset_key:
1198 item = re.match(
1199 MATH_FIELD_RE if operator else MATH_OPERATORS_RE,
1200 offset_key).group(0)
1201 offset_key = offset_key[len(item):]
1202 if operator is None:
1203 operator = MATH_FUNCTIONS[item]
1204 continue
1205 item, multiplier = (item[1:], -1) if item[0] == '-' else (item, 1)
1206 offset = float_or_none(item)
1207 if offset is None:
1208 offset = float_or_none(_traverse_infodict(item))
1209 try:
1210 value = operator(value, multiplier * offset)
1211 except (TypeError, ZeroDivisionError):
1212 return None
1213 operator = None
1214 # Datetime formatting
1215 if mdict['strf_format']:
1216 value = strftime_or_none(value, mdict['strf_format'].replace('\\,', ','))
1217
1218 # XXX: Workaround for https://github.com/yt-dlp/yt-dlp/issues/4485
1219 if sanitize and value == '':
1220 value = None
1221 return value
1222
1223 na = self.params.get('outtmpl_na_placeholder', 'NA')
1224
1225 def filename_sanitizer(key, value, restricted=self.params.get('restrictfilenames')):
1226 return sanitize_filename(str(value), restricted=restricted, is_id=(
1227 bool(re.search(r'(^|[_.])id(\.|$)', key))
1228 if 'filename-sanitization' in self.params['compat_opts']
1229 else NO_DEFAULT))
1230
1231 sanitizer = sanitize if callable(sanitize) else filename_sanitizer
1232 sanitize = bool(sanitize)
1233
1234 def _dumpjson_default(obj):
1235 if isinstance(obj, (set, LazyList)):
1236 return list(obj)
1237 return repr(obj)
1238
1239 class _ReplacementFormatter(string.Formatter):
1240 def get_field(self, field_name, args, kwargs):
1241 if field_name.isdigit():
1242 return args[0], -1
1243 raise ValueError('Unsupported field')
1244
1245 replacement_formatter = _ReplacementFormatter()
1246
1247 def create_key(outer_mobj):
1248 if not outer_mobj.group('has_key'):
1249 return outer_mobj.group(0)
1250 key = outer_mobj.group('key')
1251 mobj = re.match(INTERNAL_FORMAT_RE, key)
1252 initial_field = mobj.group('fields') if mobj else ''
1253 value, replacement, default = None, None, na
1254 while mobj:
1255 mobj = mobj.groupdict()
1256 default = mobj['default'] if mobj['default'] is not None else default
1257 value = get_value(mobj)
1258 replacement = mobj['replacement']
1259 if value is None and mobj['alternate']:
1260 mobj = re.match(INTERNAL_FORMAT_RE, mobj['remaining'][1:])
1261 else:
1262 break
1263
1264 fmt = outer_mobj.group('format')
1265 if fmt == 's' and value is not None and key in field_size_compat_map.keys():
1266 fmt = f'0{field_size_compat_map[key]:d}d'
1267
1268 if value is None:
1269 value = default
1270 elif replacement is not None:
1271 try:
1272 value = replacement_formatter.format(replacement, value)
1273 except ValueError:
1274 value = na
1275
1276 flags = outer_mobj.group('conversion') or ''
1277 str_fmt = f'{fmt[:-1]}s'
1278 if fmt[-1] == 'l': # list
1279 delim = '\n' if '#' in flags else ', '
1280 value, fmt = delim.join(map(str, variadic(value, allowed_types=(str, bytes)))), str_fmt
1281 elif fmt[-1] == 'j': # json
1282 value, fmt = json.dumps(
1283 value, default=_dumpjson_default,
1284 indent=4 if '#' in flags else None, ensure_ascii='+' not in flags), str_fmt
1285 elif fmt[-1] == 'h': # html
1286 value, fmt = escapeHTML(str(value)), str_fmt
1287 elif fmt[-1] == 'q': # quoted
1288 value = map(str, variadic(value) if '#' in flags else [value])
1289 value, fmt = ' '.join(map(compat_shlex_quote, value)), str_fmt
1290 elif fmt[-1] == 'B': # bytes
1291 value = f'%{str_fmt}'.encode() % str(value).encode()
1292 value, fmt = value.decode('utf-8', 'ignore'), 's'
1293 elif fmt[-1] == 'U': # unicode normalized
1294 value, fmt = unicodedata.normalize(
1295 # "+" = compatibility equivalence, "#" = NFD
1296 'NF%s%s' % ('K' if '+' in flags else '', 'D' if '#' in flags else 'C'),
1297 value), str_fmt
1298 elif fmt[-1] == 'D': # decimal suffix
1299 num_fmt, fmt = fmt[:-1].replace('#', ''), 's'
1300 value = format_decimal_suffix(value, f'%{num_fmt}f%s' if num_fmt else '%d%s',
1301 factor=1024 if '#' in flags else 1000)
1302 elif fmt[-1] == 'S': # filename sanitization
1303 value, fmt = filename_sanitizer(initial_field, value, restricted='#' in flags), str_fmt
1304 elif fmt[-1] == 'c':
1305 if value:
1306 value = str(value)[0]
1307 else:
1308 fmt = str_fmt
1309 elif fmt[-1] not in 'rs': # numeric
1310 value = float_or_none(value)
1311 if value is None:
1312 value, fmt = default, 's'
1313
1314 if sanitize:
1315 if fmt[-1] == 'r':
1316 # If value is an object, sanitize might convert it to a string
1317 # So we convert it to repr first
1318 value, fmt = repr(value), str_fmt
1319 if fmt[-1] in 'csr':
1320 value = sanitizer(initial_field, value)
1321
1322 key = '%s\0%s' % (key.replace('%', '%\0'), outer_mobj.group('format'))
1323 TMPL_DICT[key] = value
1324 return '{prefix}%({key}){fmt}'.format(key=key, fmt=fmt, prefix=outer_mobj.group('prefix'))
1325
1326 return EXTERNAL_FORMAT_RE.sub(create_key, outtmpl), TMPL_DICT
1327
1328 def evaluate_outtmpl(self, outtmpl, info_dict, *args, **kwargs):
1329 outtmpl, info_dict = self.prepare_outtmpl(outtmpl, info_dict, *args, **kwargs)
1330 return self.escape_outtmpl(outtmpl) % info_dict
1331
1332 def _prepare_filename(self, info_dict, *, outtmpl=None, tmpl_type=None):
1333 assert None in (outtmpl, tmpl_type), 'outtmpl and tmpl_type are mutually exclusive'
1334 if outtmpl is None:
1335 outtmpl = self.params['outtmpl'].get(tmpl_type or 'default', self.params['outtmpl']['default'])
1336 try:
1337 outtmpl = self._outtmpl_expandpath(outtmpl)
1338 filename = self.evaluate_outtmpl(outtmpl, info_dict, True)
1339 if not filename:
1340 return None
1341
1342 if tmpl_type in ('', 'temp'):
1343 final_ext, ext = self.params.get('final_ext'), info_dict.get('ext')
1344 if final_ext and ext and final_ext != ext and filename.endswith(f'.{final_ext}'):
1345 filename = replace_extension(filename, ext, final_ext)
1346 elif tmpl_type:
1347 force_ext = OUTTMPL_TYPES[tmpl_type]
1348 if force_ext:
1349 filename = replace_extension(filename, force_ext, info_dict.get('ext'))
1350
1351 # https://github.com/blackjack4494/youtube-dlc/issues/85
1352 trim_file_name = self.params.get('trim_file_name', False)
1353 if trim_file_name:
1354 no_ext, *ext = filename.rsplit('.', 2)
1355 filename = join_nonempty(no_ext[:trim_file_name], *ext, delim='.')
1356
1357 return filename
1358 except ValueError as err:
1359 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
1360 return None
1361
1362 def prepare_filename(self, info_dict, dir_type='', *, outtmpl=None, warn=False):
1363 """Generate the output filename"""
1364 if outtmpl:
1365 assert not dir_type, 'outtmpl and dir_type are mutually exclusive'
1366 dir_type = None
1367 filename = self._prepare_filename(info_dict, tmpl_type=dir_type, outtmpl=outtmpl)
1368 if not filename and dir_type not in ('', 'temp'):
1369 return ''
1370
1371 if warn:
1372 if not self.params.get('paths'):
1373 pass
1374 elif filename == '-':
1375 self.report_warning('--paths is ignored when an outputting to stdout', only_once=True)
1376 elif os.path.isabs(filename):
1377 self.report_warning('--paths is ignored since an absolute path is given in output template', only_once=True)
1378 if filename == '-' or not filename:
1379 return filename
1380
1381 return self.get_output_path(dir_type, filename)
1382
1383 def _match_entry(self, info_dict, incomplete=False, silent=False):
1384 """Returns None if the file should be downloaded"""
1385 _type = info_dict.get('_type', 'video')
1386 assert incomplete or _type == 'video', 'Only video result can be considered complete'
1387
1388 video_title = info_dict.get('title', info_dict.get('id', 'entry'))
1389
1390 def check_filter():
1391 if _type in ('playlist', 'multi_video'):
1392 return
1393 elif _type in ('url', 'url_transparent') and not try_call(
1394 lambda: self.get_info_extractor(info_dict['ie_key']).is_single_video(info_dict['url'])):
1395 return
1396
1397 if 'title' in info_dict:
1398 # This can happen when we're just evaluating the playlist
1399 title = info_dict['title']
1400 matchtitle = self.params.get('matchtitle', False)
1401 if matchtitle:
1402 if not re.search(matchtitle, title, re.IGNORECASE):
1403 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
1404 rejecttitle = self.params.get('rejecttitle', False)
1405 if rejecttitle:
1406 if re.search(rejecttitle, title, re.IGNORECASE):
1407 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
1408
1409 date = info_dict.get('upload_date')
1410 if date is not None:
1411 dateRange = self.params.get('daterange', DateRange())
1412 if date not in dateRange:
1413 return f'{date_from_str(date).isoformat()} upload date is not in range {dateRange}'
1414 view_count = info_dict.get('view_count')
1415 if view_count is not None:
1416 min_views = self.params.get('min_views')
1417 if min_views is not None and view_count < min_views:
1418 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
1419 max_views = self.params.get('max_views')
1420 if max_views is not None and view_count > max_views:
1421 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
1422 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
1423 return 'Skipping "%s" because it is age restricted' % video_title
1424
1425 match_filter = self.params.get('match_filter')
1426 if match_filter is None:
1427 return None
1428
1429 cancelled = None
1430 try:
1431 try:
1432 ret = match_filter(info_dict, incomplete=incomplete)
1433 except TypeError:
1434 # For backward compatibility
1435 ret = None if incomplete else match_filter(info_dict)
1436 except DownloadCancelled as err:
1437 if err.msg is not NO_DEFAULT:
1438 raise
1439 ret, cancelled = err.msg, err
1440
1441 if ret is NO_DEFAULT:
1442 while True:
1443 filename = self._format_screen(self.prepare_filename(info_dict), self.Styles.FILENAME)
1444 reply = input(self._format_screen(
1445 f'Download "{filename}"? (Y/n): ', self.Styles.EMPHASIS)).lower().strip()
1446 if reply in {'y', ''}:
1447 return None
1448 elif reply == 'n':
1449 if cancelled:
1450 raise type(cancelled)(f'Skipping {video_title}')
1451 return f'Skipping {video_title}'
1452 return ret
1453
1454 if self.in_download_archive(info_dict):
1455 reason = '%s has already been recorded in the archive' % video_title
1456 break_opt, break_err = 'break_on_existing', ExistingVideoReached
1457 else:
1458 try:
1459 reason = check_filter()
1460 except DownloadCancelled as e:
1461 reason, break_opt, break_err = e.msg, 'match_filter', type(e)
1462 else:
1463 break_opt, break_err = 'break_on_reject', RejectedVideoReached
1464 if reason is not None:
1465 if not silent:
1466 self.to_screen('[download] ' + reason)
1467 if self.params.get(break_opt, False):
1468 raise break_err()
1469 return reason
1470
1471 @staticmethod
1472 def add_extra_info(info_dict, extra_info):
1473 '''Set the keys from extra_info in info dict if they are missing'''
1474 for key, value in extra_info.items():
1475 info_dict.setdefault(key, value)
1476
1477 def extract_info(self, url, download=True, ie_key=None, extra_info=None,
1478 process=True, force_generic_extractor=False):
1479 """
1480 Extract and return the information dictionary of the URL
1481
1482 Arguments:
1483 @param url URL to extract
1484
1485 Keyword arguments:
1486 @param download Whether to download videos
1487 @param process Whether to resolve all unresolved references (URLs, playlist items).
1488 Must be True for download to work
1489 @param ie_key Use only the extractor with this key
1490
1491 @param extra_info Dictionary containing the extra values to add to the info (For internal use only)
1492 @force_generic_extractor Force using the generic extractor (Deprecated; use ie_key='Generic')
1493 """
1494
1495 if extra_info is None:
1496 extra_info = {}
1497
1498 if not ie_key and force_generic_extractor:
1499 ie_key = 'Generic'
1500
1501 if ie_key:
1502 ies = {ie_key: self._ies[ie_key]} if ie_key in self._ies else {}
1503 else:
1504 ies = self._ies
1505
1506 for key, ie in ies.items():
1507 if not ie.suitable(url):
1508 continue
1509
1510 if not ie.working():
1511 self.report_warning('The program functionality for this site has been marked as broken, '
1512 'and will probably not work.')
1513
1514 temp_id = ie.get_temp_id(url)
1515 if temp_id is not None and self.in_download_archive({'id': temp_id, 'ie_key': key}):
1516 self.to_screen(f'[{key}] {temp_id}: has already been recorded in the archive')
1517 if self.params.get('break_on_existing', False):
1518 raise ExistingVideoReached()
1519 break
1520 return self.__extract_info(url, self.get_info_extractor(key), download, extra_info, process)
1521 else:
1522 extractors_restricted = self.params.get('allowed_extractors') not in (None, ['default'])
1523 self.report_error(f'No suitable extractor{format_field(ie_key, None, " (%s)")} found for URL {url}',
1524 tb=False if extractors_restricted else None)
1525
1526 def _handle_extraction_exceptions(func):
1527 @functools.wraps(func)
1528 def wrapper(self, *args, **kwargs):
1529 while True:
1530 try:
1531 return func(self, *args, **kwargs)
1532 except (DownloadCancelled, LazyList.IndexError, PagedList.IndexError):
1533 raise
1534 except ReExtractInfo as e:
1535 if e.expected:
1536 self.to_screen(f'{e}; Re-extracting data')
1537 else:
1538 self.to_stderr('\r')
1539 self.report_warning(f'{e}; Re-extracting data')
1540 continue
1541 except GeoRestrictedError as e:
1542 msg = e.msg
1543 if e.countries:
1544 msg += '\nThis video is available in %s.' % ', '.join(
1545 map(ISO3166Utils.short2full, e.countries))
1546 msg += '\nYou might want to use a VPN or a proxy server (with --proxy) to workaround.'
1547 self.report_error(msg)
1548 except ExtractorError as e: # An error we somewhat expected
1549 self.report_error(str(e), e.format_traceback())
1550 except Exception as e:
1551 if self.params.get('ignoreerrors'):
1552 self.report_error(str(e), tb=encode_compat_str(traceback.format_exc()))
1553 else:
1554 raise
1555 break
1556 return wrapper
1557
1558 def _wait_for_video(self, ie_result={}):
1559 if (not self.params.get('wait_for_video')
1560 or ie_result.get('_type', 'video') != 'video'
1561 or ie_result.get('formats') or ie_result.get('url')):
1562 return
1563
1564 format_dur = lambda dur: '%02d:%02d:%02d' % timetuple_from_msec(dur * 1000)[:-1]
1565 last_msg = ''
1566
1567 def progress(msg):
1568 nonlocal last_msg
1569 full_msg = f'{msg}\n'
1570 if not self.params.get('noprogress'):
1571 full_msg = msg + ' ' * (len(last_msg) - len(msg)) + '\r'
1572 elif last_msg:
1573 return
1574 self.to_screen(full_msg, skip_eol=True)
1575 last_msg = msg
1576
1577 min_wait, max_wait = self.params.get('wait_for_video')
1578 diff = try_get(ie_result, lambda x: x['release_timestamp'] - time.time())
1579 if diff is None and ie_result.get('live_status') == 'is_upcoming':
1580 diff = round(random.uniform(min_wait, max_wait) if (max_wait and min_wait) else (max_wait or min_wait), 0)
1581 self.report_warning('Release time of video is not known')
1582 elif ie_result and (diff or 0) <= 0:
1583 self.report_warning('Video should already be available according to extracted info')
1584 diff = min(max(diff or 0, min_wait or 0), max_wait or float('inf'))
1585 self.to_screen(f'[wait] Waiting for {format_dur(diff)} - Press Ctrl+C to try now')
1586
1587 wait_till = time.time() + diff
1588 try:
1589 while True:
1590 diff = wait_till - time.time()
1591 if diff <= 0:
1592 progress('')
1593 raise ReExtractInfo('[wait] Wait period ended', expected=True)
1594 progress(f'[wait] Remaining time until next attempt: {self._format_screen(format_dur(diff), self.Styles.EMPHASIS)}')
1595 time.sleep(1)
1596 except KeyboardInterrupt:
1597 progress('')
1598 raise ReExtractInfo('[wait] Interrupted by user', expected=True)
1599 except BaseException as e:
1600 if not isinstance(e, ReExtractInfo):
1601 self.to_screen('')
1602 raise
1603
1604 @_handle_extraction_exceptions
1605 def __extract_info(self, url, ie, download, extra_info, process):
1606 try:
1607 ie_result = ie.extract(url)
1608 except UserNotLive as e:
1609 if process:
1610 if self.params.get('wait_for_video'):
1611 self.report_warning(e)
1612 self._wait_for_video()
1613 raise
1614 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
1615 self.report_warning(f'Extractor {ie.IE_NAME} returned nothing{bug_reports_message()}')
1616 return
1617 if isinstance(ie_result, list):
1618 # Backwards compatibility: old IE result format
1619 ie_result = {
1620 '_type': 'compat_list',
1621 'entries': ie_result,
1622 }
1623 if extra_info.get('original_url'):
1624 ie_result.setdefault('original_url', extra_info['original_url'])
1625 self.add_default_extra_info(ie_result, ie, url)
1626 if process:
1627 self._wait_for_video(ie_result)
1628 return self.process_ie_result(ie_result, download, extra_info)
1629 else:
1630 return ie_result
1631
1632 def add_default_extra_info(self, ie_result, ie, url):
1633 if url is not None:
1634 self.add_extra_info(ie_result, {
1635 'webpage_url': url,
1636 'original_url': url,
1637 })
1638 webpage_url = ie_result.get('webpage_url')
1639 if webpage_url:
1640 self.add_extra_info(ie_result, {
1641 'webpage_url_basename': url_basename(webpage_url),
1642 'webpage_url_domain': get_domain(webpage_url),
1643 })
1644 if ie is not None:
1645 self.add_extra_info(ie_result, {
1646 'extractor': ie.IE_NAME,
1647 'extractor_key': ie.ie_key(),
1648 })
1649
1650 def process_ie_result(self, ie_result, download=True, extra_info=None):
1651 """
1652 Take the result of the ie(may be modified) and resolve all unresolved
1653 references (URLs, playlist items).
1654
1655 It will also download the videos if 'download'.
1656 Returns the resolved ie_result.
1657 """
1658 if extra_info is None:
1659 extra_info = {}
1660 result_type = ie_result.get('_type', 'video')
1661
1662 if result_type in ('url', 'url_transparent'):
1663 ie_result['url'] = sanitize_url(
1664 ie_result['url'], scheme='http' if self.params.get('prefer_insecure') else 'https')
1665 if ie_result.get('original_url') and not extra_info.get('original_url'):
1666 extra_info = {'original_url': ie_result['original_url'], **extra_info}
1667
1668 extract_flat = self.params.get('extract_flat', False)
1669 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info)
1670 or extract_flat is True):
1671 info_copy = ie_result.copy()
1672 ie = try_get(ie_result.get('ie_key'), self.get_info_extractor)
1673 if ie and not ie_result.get('id'):
1674 info_copy['id'] = ie.get_temp_id(ie_result['url'])
1675 self.add_default_extra_info(info_copy, ie, ie_result['url'])
1676 self.add_extra_info(info_copy, extra_info)
1677 info_copy, _ = self.pre_process(info_copy)
1678 self._fill_common_fields(info_copy, False)
1679 self.__forced_printings(info_copy)
1680 self._raise_pending_errors(info_copy)
1681 if self.params.get('force_write_download_archive', False):
1682 self.record_download_archive(info_copy)
1683 return ie_result
1684
1685 if result_type == 'video':
1686 self.add_extra_info(ie_result, extra_info)
1687 ie_result = self.process_video_result(ie_result, download=download)
1688 self._raise_pending_errors(ie_result)
1689 additional_urls = (ie_result or {}).get('additional_urls')
1690 if additional_urls:
1691 # TODO: Improve MetadataParserPP to allow setting a list
1692 if isinstance(additional_urls, str):
1693 additional_urls = [additional_urls]
1694 self.to_screen(
1695 '[info] %s: %d additional URL(s) requested' % (ie_result['id'], len(additional_urls)))
1696 self.write_debug('Additional URLs: "%s"' % '", "'.join(additional_urls))
1697 ie_result['additional_entries'] = [
1698 self.extract_info(
1699 url, download, extra_info=extra_info,
1700 force_generic_extractor=self.params.get('force_generic_extractor'))
1701 for url in additional_urls
1702 ]
1703 return ie_result
1704 elif result_type == 'url':
1705 # We have to add extra_info to the results because it may be
1706 # contained in a playlist
1707 return self.extract_info(
1708 ie_result['url'], download,
1709 ie_key=ie_result.get('ie_key'),
1710 extra_info=extra_info)
1711 elif result_type == 'url_transparent':
1712 # Use the information from the embedding page
1713 info = self.extract_info(
1714 ie_result['url'], ie_key=ie_result.get('ie_key'),
1715 extra_info=extra_info, download=False, process=False)
1716
1717 # extract_info may return None when ignoreerrors is enabled and
1718 # extraction failed with an error, don't crash and return early
1719 # in this case
1720 if not info:
1721 return info
1722
1723 exempted_fields = {'_type', 'url', 'ie_key'}
1724 if not ie_result.get('section_end') and ie_result.get('section_start') is None:
1725 # For video clips, the id etc of the clip extractor should be used
1726 exempted_fields |= {'id', 'extractor', 'extractor_key'}
1727
1728 new_result = info.copy()
1729 new_result.update(filter_dict(ie_result, lambda k, v: v is not None and k not in exempted_fields))
1730
1731 # Extracted info may not be a video result (i.e.
1732 # info.get('_type', 'video') != video) but rather an url or
1733 # url_transparent. In such cases outer metadata (from ie_result)
1734 # should be propagated to inner one (info). For this to happen
1735 # _type of info should be overridden with url_transparent. This
1736 # fixes issue from https://github.com/ytdl-org/youtube-dl/pull/11163.
1737 if new_result.get('_type') == 'url':
1738 new_result['_type'] = 'url_transparent'
1739
1740 return self.process_ie_result(
1741 new_result, download=download, extra_info=extra_info)
1742 elif result_type in ('playlist', 'multi_video'):
1743 # Protect from infinite recursion due to recursively nested playlists
1744 # (see https://github.com/ytdl-org/youtube-dl/issues/27833)
1745 webpage_url = ie_result.get('webpage_url') # Playlists maynot have webpage_url
1746 if webpage_url and webpage_url in self._playlist_urls:
1747 self.to_screen(
1748 '[download] Skipping already downloaded playlist: %s'
1749 % ie_result.get('title') or ie_result.get('id'))
1750 return
1751
1752 self._playlist_level += 1
1753 self._playlist_urls.add(webpage_url)
1754 self._fill_common_fields(ie_result, False)
1755 self._sanitize_thumbnails(ie_result)
1756 try:
1757 return self.__process_playlist(ie_result, download)
1758 finally:
1759 self._playlist_level -= 1
1760 if not self._playlist_level:
1761 self._playlist_urls.clear()
1762 elif result_type == 'compat_list':
1763 self.report_warning(
1764 'Extractor %s returned a compat_list result. '
1765 'It needs to be updated.' % ie_result.get('extractor'))
1766
1767 def _fixup(r):
1768 self.add_extra_info(r, {
1769 'extractor': ie_result['extractor'],
1770 'webpage_url': ie_result['webpage_url'],
1771 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1772 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1773 'extractor_key': ie_result['extractor_key'],
1774 })
1775 return r
1776 ie_result['entries'] = [
1777 self.process_ie_result(_fixup(r), download, extra_info)
1778 for r in ie_result['entries']
1779 ]
1780 return ie_result
1781 else:
1782 raise Exception('Invalid result type: %s' % result_type)
1783
1784 def _ensure_dir_exists(self, path):
1785 return make_dir(path, self.report_error)
1786
1787 @staticmethod
1788 def _playlist_infodict(ie_result, strict=False, **kwargs):
1789 info = {
1790 'playlist_count': ie_result.get('playlist_count'),
1791 'playlist': ie_result.get('title') or ie_result.get('id'),
1792 'playlist_id': ie_result.get('id'),
1793 'playlist_title': ie_result.get('title'),
1794 'playlist_uploader': ie_result.get('uploader'),
1795 'playlist_uploader_id': ie_result.get('uploader_id'),
1796 **kwargs,
1797 }
1798 if strict:
1799 return info
1800 if ie_result.get('webpage_url'):
1801 info.update({
1802 'webpage_url': ie_result['webpage_url'],
1803 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1804 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1805 })
1806 return {
1807 **info,
1808 'playlist_index': 0,
1809 '__last_playlist_index': max(ie_result.get('requested_entries') or (0, 0)),
1810 'extractor': ie_result['extractor'],
1811 'extractor_key': ie_result['extractor_key'],
1812 }
1813
1814 def __process_playlist(self, ie_result, download):
1815 """Process each entry in the playlist"""
1816 assert ie_result['_type'] in ('playlist', 'multi_video')
1817
1818 common_info = self._playlist_infodict(ie_result, strict=True)
1819 title = common_info.get('playlist') or '<Untitled>'
1820 if self._match_entry(common_info, incomplete=True) is not None:
1821 return
1822 self.to_screen(f'[download] Downloading {ie_result["_type"]}: {title}')
1823
1824 all_entries = PlaylistEntries(self, ie_result)
1825 entries = orderedSet(all_entries.get_requested_items(), lazy=True)
1826
1827 lazy = self.params.get('lazy_playlist')
1828 if lazy:
1829 resolved_entries, n_entries = [], 'N/A'
1830 ie_result['requested_entries'], ie_result['entries'] = None, None
1831 else:
1832 entries = resolved_entries = list(entries)
1833 n_entries = len(resolved_entries)
1834 ie_result['requested_entries'], ie_result['entries'] = tuple(zip(*resolved_entries)) or ([], [])
1835 if not ie_result.get('playlist_count'):
1836 # Better to do this after potentially exhausting entries
1837 ie_result['playlist_count'] = all_entries.get_full_count()
1838
1839 extra = self._playlist_infodict(ie_result, n_entries=int_or_none(n_entries))
1840 ie_copy = collections.ChainMap(ie_result, extra)
1841
1842 _infojson_written = False
1843 write_playlist_files = self.params.get('allow_playlist_files', True)
1844 if write_playlist_files and self.params.get('list_thumbnails'):
1845 self.list_thumbnails(ie_result)
1846 if write_playlist_files and not self.params.get('simulate'):
1847 _infojson_written = self._write_info_json(
1848 'playlist', ie_result, self.prepare_filename(ie_copy, 'pl_infojson'))
1849 if _infojson_written is None:
1850 return
1851 if self._write_description('playlist', ie_result,
1852 self.prepare_filename(ie_copy, 'pl_description')) is None:
1853 return
1854 # TODO: This should be passed to ThumbnailsConvertor if necessary
1855 self._write_thumbnails('playlist', ie_result, self.prepare_filename(ie_copy, 'pl_thumbnail'))
1856
1857 if lazy:
1858 if self.params.get('playlistreverse') or self.params.get('playlistrandom'):
1859 self.report_warning('playlistreverse and playlistrandom are not supported with lazy_playlist', only_once=True)
1860 elif self.params.get('playlistreverse'):
1861 entries.reverse()
1862 elif self.params.get('playlistrandom'):
1863 random.shuffle(entries)
1864
1865 self.to_screen(f'[{ie_result["extractor"]}] Playlist {title}: Downloading {n_entries} items'
1866 f'{format_field(ie_result, "playlist_count", " of %s")}')
1867
1868 keep_resolved_entries = self.params.get('extract_flat') != 'discard'
1869 if self.params.get('extract_flat') == 'discard_in_playlist':
1870 keep_resolved_entries = ie_result['_type'] != 'playlist'
1871 if keep_resolved_entries:
1872 self.write_debug('The information of all playlist entries will be held in memory')
1873
1874 failures = 0
1875 max_failures = self.params.get('skip_playlist_after_errors') or float('inf')
1876 for i, (playlist_index, entry) in enumerate(entries):
1877 if lazy:
1878 resolved_entries.append((playlist_index, entry))
1879 if not entry:
1880 continue
1881
1882 entry['__x_forwarded_for_ip'] = ie_result.get('__x_forwarded_for_ip')
1883 if not lazy and 'playlist-index' in self.params.get('compat_opts', []):
1884 playlist_index = ie_result['requested_entries'][i]
1885
1886 entry_copy = collections.ChainMap(entry, {
1887 **common_info,
1888 'n_entries': int_or_none(n_entries),
1889 'playlist_index': playlist_index,
1890 'playlist_autonumber': i + 1,
1891 })
1892
1893 if self._match_entry(entry_copy, incomplete=True) is not None:
1894 # For compatabilty with youtube-dl. See https://github.com/yt-dlp/yt-dlp/issues/4369
1895 resolved_entries[i] = (playlist_index, NO_DEFAULT)
1896 continue
1897
1898 self.to_screen('[download] Downloading item %s of %s' % (
1899 self._format_screen(i + 1, self.Styles.ID), self._format_screen(n_entries, self.Styles.EMPHASIS)))
1900
1901 entry_result = self.__process_iterable_entry(entry, download, collections.ChainMap({
1902 'playlist_index': playlist_index,
1903 'playlist_autonumber': i + 1,
1904 }, extra))
1905 if not entry_result:
1906 failures += 1
1907 if failures >= max_failures:
1908 self.report_error(
1909 f'Skipping the remaining entries in playlist "{title}" since {failures} items failed extraction')
1910 break
1911 if keep_resolved_entries:
1912 resolved_entries[i] = (playlist_index, entry_result)
1913
1914 # Update with processed data
1915 ie_result['entries'] = [e for _, e in resolved_entries if e is not NO_DEFAULT]
1916 ie_result['requested_entries'] = [i for i, e in resolved_entries if e is not NO_DEFAULT]
1917 if ie_result['requested_entries'] == try_call(lambda: list(range(1, ie_result['playlist_count'] + 1))):
1918 # Do not set for full playlist
1919 ie_result.pop('requested_entries')
1920
1921 # Write the updated info to json
1922 if _infojson_written is True and self._write_info_json(
1923 'updated playlist', ie_result,
1924 self.prepare_filename(ie_copy, 'pl_infojson'), overwrite=True) is None:
1925 return
1926
1927 ie_result = self.run_all_pps('playlist', ie_result)
1928 self.to_screen(f'[download] Finished downloading playlist: {title}')
1929 return ie_result
1930
1931 @_handle_extraction_exceptions
1932 def __process_iterable_entry(self, entry, download, extra_info):
1933 return self.process_ie_result(
1934 entry, download=download, extra_info=extra_info)
1935
1936 def _build_format_filter(self, filter_spec):
1937 " Returns a function to filter the formats according to the filter_spec "
1938
1939 OPERATORS = {
1940 '<': operator.lt,
1941 '<=': operator.le,
1942 '>': operator.gt,
1943 '>=': operator.ge,
1944 '=': operator.eq,
1945 '!=': operator.ne,
1946 }
1947 operator_rex = re.compile(r'''(?x)\s*
1948 (?P<key>[\w.-]+)\s*
1949 (?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
1950 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)\s*
1951 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
1952 m = operator_rex.fullmatch(filter_spec)
1953 if m:
1954 try:
1955 comparison_value = int(m.group('value'))
1956 except ValueError:
1957 comparison_value = parse_filesize(m.group('value'))
1958 if comparison_value is None:
1959 comparison_value = parse_filesize(m.group('value') + 'B')
1960 if comparison_value is None:
1961 raise ValueError(
1962 'Invalid value %r in format specification %r' % (
1963 m.group('value'), filter_spec))
1964 op = OPERATORS[m.group('op')]
1965
1966 if not m:
1967 STR_OPERATORS = {
1968 '=': operator.eq,
1969 '^=': lambda attr, value: attr.startswith(value),
1970 '$=': lambda attr, value: attr.endswith(value),
1971 '*=': lambda attr, value: value in attr,
1972 '~=': lambda attr, value: value.search(attr) is not None
1973 }
1974 str_operator_rex = re.compile(r'''(?x)\s*
1975 (?P<key>[a-zA-Z0-9._-]+)\s*
1976 (?P<negation>!\s*)?(?P<op>%s)\s*(?P<none_inclusive>\?\s*)?
1977 (?P<quote>["'])?
1978 (?P<value>(?(quote)(?:(?!(?P=quote))[^\\]|\\.)+|[\w.-]+))
1979 (?(quote)(?P=quote))\s*
1980 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
1981 m = str_operator_rex.fullmatch(filter_spec)
1982 if m:
1983 if m.group('op') == '~=':
1984 comparison_value = re.compile(m.group('value'))
1985 else:
1986 comparison_value = re.sub(r'''\\([\\"'])''', r'\1', m.group('value'))
1987 str_op = STR_OPERATORS[m.group('op')]
1988 if m.group('negation'):
1989 op = lambda attr, value: not str_op(attr, value)
1990 else:
1991 op = str_op
1992
1993 if not m:
1994 raise SyntaxError('Invalid filter specification %r' % filter_spec)
1995
1996 def _filter(f):
1997 actual_value = f.get(m.group('key'))
1998 if actual_value is None:
1999 return m.group('none_inclusive')
2000 return op(actual_value, comparison_value)
2001 return _filter
2002
2003 def _check_formats(self, formats):
2004 for f in formats:
2005 self.to_screen('[info] Testing format %s' % f['format_id'])
2006 path = self.get_output_path('temp')
2007 if not self._ensure_dir_exists(f'{path}/'):
2008 continue
2009 temp_file = tempfile.NamedTemporaryFile(suffix='.tmp', delete=False, dir=path or None)
2010 temp_file.close()
2011 try:
2012 success, _ = self.dl(temp_file.name, f, test=True)
2013 except (DownloadError, OSError, ValueError) + network_exceptions:
2014 success = False
2015 finally:
2016 if os.path.exists(temp_file.name):
2017 try:
2018 os.remove(temp_file.name)
2019 except OSError:
2020 self.report_warning('Unable to delete temporary file "%s"' % temp_file.name)
2021 if success:
2022 yield f
2023 else:
2024 self.to_screen('[info] Unable to download format %s. Skipping...' % f['format_id'])
2025
2026 def _default_format_spec(self, info_dict, download=True):
2027
2028 def can_merge():
2029 merger = FFmpegMergerPP(self)
2030 return merger.available and merger.can_merge()
2031
2032 prefer_best = (
2033 not self.params.get('simulate')
2034 and download
2035 and (
2036 not can_merge()
2037 or info_dict.get('is_live') and not self.params.get('live_from_start')
2038 or self.params['outtmpl']['default'] == '-'))
2039 compat = (
2040 prefer_best
2041 or self.params.get('allow_multiple_audio_streams', False)
2042 or 'format-spec' in self.params['compat_opts'])
2043
2044 return (
2045 'best/bestvideo+bestaudio' if prefer_best
2046 else 'bestvideo*+bestaudio/best' if not compat
2047 else 'bestvideo+bestaudio/best')
2048
2049 def build_format_selector(self, format_spec):
2050 def syntax_error(note, start):
2051 message = (
2052 'Invalid format specification: '
2053 '{}\n\t{}\n\t{}^'.format(note, format_spec, ' ' * start[1]))
2054 return SyntaxError(message)
2055
2056 PICKFIRST = 'PICKFIRST'
2057 MERGE = 'MERGE'
2058 SINGLE = 'SINGLE'
2059 GROUP = 'GROUP'
2060 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
2061
2062 allow_multiple_streams = {'audio': self.params.get('allow_multiple_audio_streams', False),
2063 'video': self.params.get('allow_multiple_video_streams', False)}
2064
2065 check_formats = self.params.get('check_formats') == 'selected'
2066
2067 def _parse_filter(tokens):
2068 filter_parts = []
2069 for type, string_, start, _, _ in tokens:
2070 if type == tokenize.OP and string_ == ']':
2071 return ''.join(filter_parts)
2072 else:
2073 filter_parts.append(string_)
2074
2075 def _remove_unused_ops(tokens):
2076 # Remove operators that we don't use and join them with the surrounding strings.
2077 # E.g. 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
2078 ALLOWED_OPS = ('/', '+', ',', '(', ')')
2079 last_string, last_start, last_end, last_line = None, None, None, None
2080 for type, string_, start, end, line in tokens:
2081 if type == tokenize.OP and string_ == '[':
2082 if last_string:
2083 yield tokenize.NAME, last_string, last_start, last_end, last_line
2084 last_string = None
2085 yield type, string_, start, end, line
2086 # everything inside brackets will be handled by _parse_filter
2087 for type, string_, start, end, line in tokens:
2088 yield type, string_, start, end, line
2089 if type == tokenize.OP and string_ == ']':
2090 break
2091 elif type == tokenize.OP and string_ in ALLOWED_OPS:
2092 if last_string:
2093 yield tokenize.NAME, last_string, last_start, last_end, last_line
2094 last_string = None
2095 yield type, string_, start, end, line
2096 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
2097 if not last_string:
2098 last_string = string_
2099 last_start = start
2100 last_end = end
2101 else:
2102 last_string += string_
2103 if last_string:
2104 yield tokenize.NAME, last_string, last_start, last_end, last_line
2105
2106 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
2107 selectors = []
2108 current_selector = None
2109 for type, string_, start, _, _ in tokens:
2110 # ENCODING is only defined in python 3.x
2111 if type == getattr(tokenize, 'ENCODING', None):
2112 continue
2113 elif type in [tokenize.NAME, tokenize.NUMBER]:
2114 current_selector = FormatSelector(SINGLE, string_, [])
2115 elif type == tokenize.OP:
2116 if string_ == ')':
2117 if not inside_group:
2118 # ')' will be handled by the parentheses group
2119 tokens.restore_last_token()
2120 break
2121 elif inside_merge and string_ in ['/', ',']:
2122 tokens.restore_last_token()
2123 break
2124 elif inside_choice and string_ == ',':
2125 tokens.restore_last_token()
2126 break
2127 elif string_ == ',':
2128 if not current_selector:
2129 raise syntax_error('"," must follow a format selector', start)
2130 selectors.append(current_selector)
2131 current_selector = None
2132 elif string_ == '/':
2133 if not current_selector:
2134 raise syntax_error('"/" must follow a format selector', start)
2135 first_choice = current_selector
2136 second_choice = _parse_format_selection(tokens, inside_choice=True)
2137 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
2138 elif string_ == '[':
2139 if not current_selector:
2140 current_selector = FormatSelector(SINGLE, 'best', [])
2141 format_filter = _parse_filter(tokens)
2142 current_selector.filters.append(format_filter)
2143 elif string_ == '(':
2144 if current_selector:
2145 raise syntax_error('Unexpected "("', start)
2146 group = _parse_format_selection(tokens, inside_group=True)
2147 current_selector = FormatSelector(GROUP, group, [])
2148 elif string_ == '+':
2149 if not current_selector:
2150 raise syntax_error('Unexpected "+"', start)
2151 selector_1 = current_selector
2152 selector_2 = _parse_format_selection(tokens, inside_merge=True)
2153 if not selector_2:
2154 raise syntax_error('Expected a selector', start)
2155 current_selector = FormatSelector(MERGE, (selector_1, selector_2), [])
2156 else:
2157 raise syntax_error(f'Operator not recognized: "{string_}"', start)
2158 elif type == tokenize.ENDMARKER:
2159 break
2160 if current_selector:
2161 selectors.append(current_selector)
2162 return selectors
2163
2164 def _merge(formats_pair):
2165 format_1, format_2 = formats_pair
2166
2167 formats_info = []
2168 formats_info.extend(format_1.get('requested_formats', (format_1,)))
2169 formats_info.extend(format_2.get('requested_formats', (format_2,)))
2170
2171 if not allow_multiple_streams['video'] or not allow_multiple_streams['audio']:
2172 get_no_more = {'video': False, 'audio': False}
2173 for (i, fmt_info) in enumerate(formats_info):
2174 if fmt_info.get('acodec') == fmt_info.get('vcodec') == 'none':
2175 formats_info.pop(i)
2176 continue
2177 for aud_vid in ['audio', 'video']:
2178 if not allow_multiple_streams[aud_vid] and fmt_info.get(aud_vid[0] + 'codec') != 'none':
2179 if get_no_more[aud_vid]:
2180 formats_info.pop(i)
2181 break
2182 get_no_more[aud_vid] = True
2183
2184 if len(formats_info) == 1:
2185 return formats_info[0]
2186
2187 video_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('vcodec') != 'none']
2188 audio_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('acodec') != 'none']
2189
2190 the_only_video = video_fmts[0] if len(video_fmts) == 1 else None
2191 the_only_audio = audio_fmts[0] if len(audio_fmts) == 1 else None
2192
2193 output_ext = get_compatible_ext(
2194 vcodecs=[f.get('vcodec') for f in video_fmts],
2195 acodecs=[f.get('acodec') for f in audio_fmts],
2196 vexts=[f['ext'] for f in video_fmts],
2197 aexts=[f['ext'] for f in audio_fmts],
2198 preferences=(try_call(lambda: self.params['merge_output_format'].split('/'))
2199 or self.params.get('prefer_free_formats') and ('webm', 'mkv')))
2200
2201 filtered = lambda *keys: filter(None, (traverse_obj(fmt, *keys) for fmt in formats_info))
2202
2203 new_dict = {
2204 'requested_formats': formats_info,
2205 'format': '+'.join(filtered('format')),
2206 'format_id': '+'.join(filtered('format_id')),
2207 'ext': output_ext,
2208 'protocol': '+'.join(map(determine_protocol, formats_info)),
2209 'language': '+'.join(orderedSet(filtered('language'))) or None,
2210 'format_note': '+'.join(orderedSet(filtered('format_note'))) or None,
2211 'filesize_approx': sum(filtered('filesize', 'filesize_approx')) or None,
2212 'tbr': sum(filtered('tbr', 'vbr', 'abr')),
2213 }
2214
2215 if the_only_video:
2216 new_dict.update({
2217 'width': the_only_video.get('width'),
2218 'height': the_only_video.get('height'),
2219 'resolution': the_only_video.get('resolution') or self.format_resolution(the_only_video),
2220 'fps': the_only_video.get('fps'),
2221 'dynamic_range': the_only_video.get('dynamic_range'),
2222 'vcodec': the_only_video.get('vcodec'),
2223 'vbr': the_only_video.get('vbr'),
2224 'stretched_ratio': the_only_video.get('stretched_ratio'),
2225 'aspect_ratio': the_only_video.get('aspect_ratio'),
2226 })
2227
2228 if the_only_audio:
2229 new_dict.update({
2230 'acodec': the_only_audio.get('acodec'),
2231 'abr': the_only_audio.get('abr'),
2232 'asr': the_only_audio.get('asr'),
2233 'audio_channels': the_only_audio.get('audio_channels')
2234 })
2235
2236 return new_dict
2237
2238 def _check_formats(formats):
2239 if not check_formats:
2240 yield from formats
2241 return
2242 yield from self._check_formats(formats)
2243
2244 def _build_selector_function(selector):
2245 if isinstance(selector, list): # ,
2246 fs = [_build_selector_function(s) for s in selector]
2247
2248 def selector_function(ctx):
2249 for f in fs:
2250 yield from f(ctx)
2251 return selector_function
2252
2253 elif selector.type == GROUP: # ()
2254 selector_function = _build_selector_function(selector.selector)
2255
2256 elif selector.type == PICKFIRST: # /
2257 fs = [_build_selector_function(s) for s in selector.selector]
2258
2259 def selector_function(ctx):
2260 for f in fs:
2261 picked_formats = list(f(ctx))
2262 if picked_formats:
2263 return picked_formats
2264 return []
2265
2266 elif selector.type == MERGE: # +
2267 selector_1, selector_2 = map(_build_selector_function, selector.selector)
2268
2269 def selector_function(ctx):
2270 for pair in itertools.product(selector_1(ctx), selector_2(ctx)):
2271 yield _merge(pair)
2272
2273 elif selector.type == SINGLE: # atom
2274 format_spec = selector.selector or 'best'
2275
2276 # TODO: Add allvideo, allaudio etc by generalizing the code with best/worst selector
2277 if format_spec == 'all':
2278 def selector_function(ctx):
2279 yield from _check_formats(ctx['formats'][::-1])
2280 elif format_spec == 'mergeall':
2281 def selector_function(ctx):
2282 formats = list(_check_formats(
2283 f for f in ctx['formats'] if f.get('vcodec') != 'none' or f.get('acodec') != 'none'))
2284 if not formats:
2285 return
2286 merged_format = formats[-1]
2287 for f in formats[-2::-1]:
2288 merged_format = _merge((merged_format, f))
2289 yield merged_format
2290
2291 else:
2292 format_fallback, seperate_fallback, format_reverse, format_idx = False, None, True, 1
2293 mobj = re.match(
2294 r'(?P<bw>best|worst|b|w)(?P<type>video|audio|v|a)?(?P<mod>\*)?(?:\.(?P<n>[1-9]\d*))?$',
2295 format_spec)
2296 if mobj is not None:
2297 format_idx = int_or_none(mobj.group('n'), default=1)
2298 format_reverse = mobj.group('bw')[0] == 'b'
2299 format_type = (mobj.group('type') or [None])[0]
2300 not_format_type = {'v': 'a', 'a': 'v'}.get(format_type)
2301 format_modified = mobj.group('mod') is not None
2302
2303 format_fallback = not format_type and not format_modified # for b, w
2304 _filter_f = (
2305 (lambda f: f.get('%scodec' % format_type) != 'none')
2306 if format_type and format_modified # bv*, ba*, wv*, wa*
2307 else (lambda f: f.get('%scodec' % not_format_type) == 'none')
2308 if format_type # bv, ba, wv, wa
2309 else (lambda f: f.get('vcodec') != 'none' and f.get('acodec') != 'none')
2310 if not format_modified # b, w
2311 else lambda f: True) # b*, w*
2312 filter_f = lambda f: _filter_f(f) and (
2313 f.get('vcodec') != 'none' or f.get('acodec') != 'none')
2314 else:
2315 if format_spec in self._format_selection_exts['audio']:
2316 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none'
2317 elif format_spec in self._format_selection_exts['video']:
2318 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none' and f.get('vcodec') != 'none'
2319 seperate_fallback = lambda f: f.get('ext') == format_spec and f.get('vcodec') != 'none'
2320 elif format_spec in self._format_selection_exts['storyboards']:
2321 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') == 'none' and f.get('vcodec') == 'none'
2322 else:
2323 filter_f = lambda f: f.get('format_id') == format_spec # id
2324
2325 def selector_function(ctx):
2326 formats = list(ctx['formats'])
2327 matches = list(filter(filter_f, formats)) if filter_f is not None else formats
2328 if not matches:
2329 if format_fallback and ctx['incomplete_formats']:
2330 # for extractors with incomplete formats (audio only (soundcloud)
2331 # or video only (imgur)) best/worst will fallback to
2332 # best/worst {video,audio}-only format
2333 matches = formats
2334 elif seperate_fallback and not ctx['has_merged_format']:
2335 # for compatibility with youtube-dl when there is no pre-merged format
2336 matches = list(filter(seperate_fallback, formats))
2337 matches = LazyList(_check_formats(matches[::-1 if format_reverse else 1]))
2338 try:
2339 yield matches[format_idx - 1]
2340 except LazyList.IndexError:
2341 return
2342
2343 filters = [self._build_format_filter(f) for f in selector.filters]
2344
2345 def final_selector(ctx):
2346 ctx_copy = dict(ctx)
2347 for _filter in filters:
2348 ctx_copy['formats'] = list(filter(_filter, ctx_copy['formats']))
2349 return selector_function(ctx_copy)
2350 return final_selector
2351
2352 stream = io.BytesIO(format_spec.encode())
2353 try:
2354 tokens = list(_remove_unused_ops(tokenize.tokenize(stream.readline)))
2355 except tokenize.TokenError:
2356 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
2357
2358 class TokenIterator:
2359 def __init__(self, tokens):
2360 self.tokens = tokens
2361 self.counter = 0
2362
2363 def __iter__(self):
2364 return self
2365
2366 def __next__(self):
2367 if self.counter >= len(self.tokens):
2368 raise StopIteration()
2369 value = self.tokens[self.counter]
2370 self.counter += 1
2371 return value
2372
2373 next = __next__
2374
2375 def restore_last_token(self):
2376 self.counter -= 1
2377
2378 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
2379 return _build_selector_function(parsed_selector)
2380
2381 def _calc_headers(self, info_dict):
2382 res = merge_headers(self.params['http_headers'], info_dict.get('http_headers') or {})
2383
2384 cookies = self._calc_cookies(info_dict['url'])
2385 if cookies:
2386 res['Cookie'] = cookies
2387
2388 if 'X-Forwarded-For' not in res:
2389 x_forwarded_for_ip = info_dict.get('__x_forwarded_for_ip')
2390 if x_forwarded_for_ip:
2391 res['X-Forwarded-For'] = x_forwarded_for_ip
2392
2393 return res
2394
2395 def _calc_cookies(self, url):
2396 pr = sanitized_Request(url)
2397 self.cookiejar.add_cookie_header(pr)
2398 return pr.get_header('Cookie')
2399
2400 def _sort_thumbnails(self, thumbnails):
2401 thumbnails.sort(key=lambda t: (
2402 t.get('preference') if t.get('preference') is not None else -1,
2403 t.get('width') if t.get('width') is not None else -1,
2404 t.get('height') if t.get('height') is not None else -1,
2405 t.get('id') if t.get('id') is not None else '',
2406 t.get('url')))
2407
2408 def _sanitize_thumbnails(self, info_dict):
2409 thumbnails = info_dict.get('thumbnails')
2410 if thumbnails is None:
2411 thumbnail = info_dict.get('thumbnail')
2412 if thumbnail:
2413 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
2414 if not thumbnails:
2415 return
2416
2417 def check_thumbnails(thumbnails):
2418 for t in thumbnails:
2419 self.to_screen(f'[info] Testing thumbnail {t["id"]}')
2420 try:
2421 self.urlopen(HEADRequest(t['url']))
2422 except network_exceptions as err:
2423 self.to_screen(f'[info] Unable to connect to thumbnail {t["id"]} URL {t["url"]!r} - {err}. Skipping...')
2424 continue
2425 yield t
2426
2427 self._sort_thumbnails(thumbnails)
2428 for i, t in enumerate(thumbnails):
2429 if t.get('id') is None:
2430 t['id'] = '%d' % i
2431 if t.get('width') and t.get('height'):
2432 t['resolution'] = '%dx%d' % (t['width'], t['height'])
2433 t['url'] = sanitize_url(t['url'])
2434
2435 if self.params.get('check_formats') is True:
2436 info_dict['thumbnails'] = LazyList(check_thumbnails(thumbnails[::-1]), reverse=True)
2437 else:
2438 info_dict['thumbnails'] = thumbnails
2439
2440 def _fill_common_fields(self, info_dict, final=True):
2441 # TODO: move sanitization here
2442 if final:
2443 title = info_dict['fulltitle'] = info_dict.get('title')
2444 if not title:
2445 if title == '':
2446 self.write_debug('Extractor gave empty title. Creating a generic title')
2447 else:
2448 self.report_warning('Extractor failed to obtain "title". Creating a generic title instead')
2449 info_dict['title'] = f'{info_dict["extractor"].replace(":", "-")} video #{info_dict["id"]}'
2450
2451 if info_dict.get('duration') is not None:
2452 info_dict['duration_string'] = formatSeconds(info_dict['duration'])
2453
2454 for ts_key, date_key in (
2455 ('timestamp', 'upload_date'),
2456 ('release_timestamp', 'release_date'),
2457 ('modified_timestamp', 'modified_date'),
2458 ):
2459 if info_dict.get(date_key) is None and info_dict.get(ts_key) is not None:
2460 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
2461 # see http://bugs.python.org/issue1646728)
2462 with contextlib.suppress(ValueError, OverflowError, OSError):
2463 upload_date = datetime.datetime.utcfromtimestamp(info_dict[ts_key])
2464 info_dict[date_key] = upload_date.strftime('%Y%m%d')
2465
2466 live_keys = ('is_live', 'was_live')
2467 live_status = info_dict.get('live_status')
2468 if live_status is None:
2469 for key in live_keys:
2470 if info_dict.get(key) is False:
2471 continue
2472 if info_dict.get(key):
2473 live_status = key
2474 break
2475 if all(info_dict.get(key) is False for key in live_keys):
2476 live_status = 'not_live'
2477 if live_status:
2478 info_dict['live_status'] = live_status
2479 for key in live_keys:
2480 if info_dict.get(key) is None:
2481 info_dict[key] = (live_status == key)
2482 if live_status == 'post_live':
2483 info_dict['was_live'] = True
2484
2485 # Auto generate title fields corresponding to the *_number fields when missing
2486 # in order to always have clean titles. This is very common for TV series.
2487 for field in ('chapter', 'season', 'episode'):
2488 if final and info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
2489 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
2490
2491 def _raise_pending_errors(self, info):
2492 err = info.pop('__pending_error', None)
2493 if err:
2494 self.report_error(err, tb=False)
2495
2496 def sort_formats(self, info_dict):
2497 formats = self._get_formats(info_dict)
2498 formats.sort(key=FormatSorter(
2499 self, info_dict.get('_format_sort_fields') or []).calculate_preference)
2500
2501 def process_video_result(self, info_dict, download=True):
2502 assert info_dict.get('_type', 'video') == 'video'
2503 self._num_videos += 1
2504
2505 if 'id' not in info_dict:
2506 raise ExtractorError('Missing "id" field in extractor result', ie=info_dict['extractor'])
2507 elif not info_dict.get('id'):
2508 raise ExtractorError('Extractor failed to obtain "id"', ie=info_dict['extractor'])
2509
2510 def report_force_conversion(field, field_not, conversion):
2511 self.report_warning(
2512 '"%s" field is not %s - forcing %s conversion, there is an error in extractor'
2513 % (field, field_not, conversion))
2514
2515 def sanitize_string_field(info, string_field):
2516 field = info.get(string_field)
2517 if field is None or isinstance(field, str):
2518 return
2519 report_force_conversion(string_field, 'a string', 'string')
2520 info[string_field] = str(field)
2521
2522 def sanitize_numeric_fields(info):
2523 for numeric_field in self._NUMERIC_FIELDS:
2524 field = info.get(numeric_field)
2525 if field is None or isinstance(field, (int, float)):
2526 continue
2527 report_force_conversion(numeric_field, 'numeric', 'int')
2528 info[numeric_field] = int_or_none(field)
2529
2530 sanitize_string_field(info_dict, 'id')
2531 sanitize_numeric_fields(info_dict)
2532 if info_dict.get('section_end') and info_dict.get('section_start') is not None:
2533 info_dict['duration'] = round(info_dict['section_end'] - info_dict['section_start'], 3)
2534 if (info_dict.get('duration') or 0) <= 0 and info_dict.pop('duration', None):
2535 self.report_warning('"duration" field is negative, there is an error in extractor')
2536
2537 chapters = info_dict.get('chapters') or []
2538 if chapters and chapters[0].get('start_time'):
2539 chapters.insert(0, {'start_time': 0})
2540
2541 dummy_chapter = {'end_time': 0, 'start_time': info_dict.get('duration')}
2542 for idx, (prev, current, next_) in enumerate(zip(
2543 (dummy_chapter, *chapters), chapters, (*chapters[1:], dummy_chapter)), 1):
2544 if current.get('start_time') is None:
2545 current['start_time'] = prev.get('end_time')
2546 if not current.get('end_time'):
2547 current['end_time'] = next_.get('start_time')
2548 if not current.get('title'):
2549 current['title'] = f'<Untitled Chapter {idx}>'
2550
2551 if 'playlist' not in info_dict:
2552 # It isn't part of a playlist
2553 info_dict['playlist'] = None
2554 info_dict['playlist_index'] = None
2555
2556 self._sanitize_thumbnails(info_dict)
2557
2558 thumbnail = info_dict.get('thumbnail')
2559 thumbnails = info_dict.get('thumbnails')
2560 if thumbnail:
2561 info_dict['thumbnail'] = sanitize_url(thumbnail)
2562 elif thumbnails:
2563 info_dict['thumbnail'] = thumbnails[-1]['url']
2564
2565 if info_dict.get('display_id') is None and 'id' in info_dict:
2566 info_dict['display_id'] = info_dict['id']
2567
2568 self._fill_common_fields(info_dict)
2569
2570 for cc_kind in ('subtitles', 'automatic_captions'):
2571 cc = info_dict.get(cc_kind)
2572 if cc:
2573 for _, subtitle in cc.items():
2574 for subtitle_format in subtitle:
2575 if subtitle_format.get('url'):
2576 subtitle_format['url'] = sanitize_url(subtitle_format['url'])
2577 if subtitle_format.get('ext') is None:
2578 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
2579
2580 automatic_captions = info_dict.get('automatic_captions')
2581 subtitles = info_dict.get('subtitles')
2582
2583 info_dict['requested_subtitles'] = self.process_subtitles(
2584 info_dict['id'], subtitles, automatic_captions)
2585
2586 formats = self._get_formats(info_dict)
2587
2588 # Backward compatibility with InfoExtractor._sort_formats
2589 field_preference = (formats or [{}])[0].pop('__sort_fields', None)
2590 if field_preference:
2591 info_dict['_format_sort_fields'] = field_preference
2592
2593 # or None ensures --clean-infojson removes it
2594 info_dict['_has_drm'] = any(f.get('has_drm') for f in formats) or None
2595 if not self.params.get('allow_unplayable_formats'):
2596 formats = [f for f in formats if not f.get('has_drm')]
2597
2598 if formats and all(f.get('acodec') == f.get('vcodec') == 'none' for f in formats):
2599 self.report_warning(
2600 f'{"This video is DRM protected and " if info_dict["_has_drm"] else ""}'
2601 'only images are available for download. Use --list-formats to see them'.capitalize())
2602
2603 get_from_start = not info_dict.get('is_live') or bool(self.params.get('live_from_start'))
2604 if not get_from_start:
2605 info_dict['title'] += ' ' + datetime.datetime.now().strftime('%Y-%m-%d %H:%M')
2606 if info_dict.get('is_live') and formats:
2607 formats = [f for f in formats if bool(f.get('is_from_start')) == get_from_start]
2608 if get_from_start and not formats:
2609 self.raise_no_formats(info_dict, msg=(
2610 '--live-from-start is passed, but there are no formats that can be downloaded from the start. '
2611 'If you want to download from the current time, use --no-live-from-start'))
2612
2613 def is_wellformed(f):
2614 url = f.get('url')
2615 if not url:
2616 self.report_warning(
2617 '"url" field is missing or empty - skipping format, '
2618 'there is an error in extractor')
2619 return False
2620 if isinstance(url, bytes):
2621 sanitize_string_field(f, 'url')
2622 return True
2623
2624 # Filter out malformed formats for better extraction robustness
2625 formats = list(filter(is_wellformed, formats or []))
2626
2627 if not formats:
2628 self.raise_no_formats(info_dict)
2629
2630 for format in formats:
2631 sanitize_string_field(format, 'format_id')
2632 sanitize_numeric_fields(format)
2633 format['url'] = sanitize_url(format['url'])
2634 if format.get('ext') is None:
2635 format['ext'] = determine_ext(format['url']).lower()
2636 if format.get('protocol') is None:
2637 format['protocol'] = determine_protocol(format)
2638 if format.get('resolution') is None:
2639 format['resolution'] = self.format_resolution(format, default=None)
2640 if format.get('dynamic_range') is None and format.get('vcodec') != 'none':
2641 format['dynamic_range'] = 'SDR'
2642 if format.get('aspect_ratio') is None:
2643 format['aspect_ratio'] = try_call(lambda: round(format['width'] / format['height'], 2))
2644 if (info_dict.get('duration') and format.get('tbr')
2645 and not format.get('filesize') and not format.get('filesize_approx')):
2646 format['filesize_approx'] = int(info_dict['duration'] * format['tbr'] * (1024 / 8))
2647 format['http_headers'] = self._calc_headers(collections.ChainMap(format, info_dict))
2648
2649 # This is copied to http_headers by the above _calc_headers and can now be removed
2650 if '__x_forwarded_for_ip' in info_dict:
2651 del info_dict['__x_forwarded_for_ip']
2652
2653 self.sort_formats({
2654 'formats': formats,
2655 '_format_sort_fields': info_dict.get('_format_sort_fields')
2656 })
2657
2658 # Sanitize and group by format_id
2659 formats_dict = {}
2660 for i, format in enumerate(formats):
2661 if not format.get('format_id'):
2662 format['format_id'] = str(i)
2663 else:
2664 # Sanitize format_id from characters used in format selector expression
2665 format['format_id'] = re.sub(r'[\s,/+\[\]()]', '_', format['format_id'])
2666 formats_dict.setdefault(format['format_id'], []).append(format)
2667
2668 # Make sure all formats have unique format_id
2669 common_exts = set(itertools.chain(*self._format_selection_exts.values()))
2670 for format_id, ambiguous_formats in formats_dict.items():
2671 ambigious_id = len(ambiguous_formats) > 1
2672 for i, format in enumerate(ambiguous_formats):
2673 if ambigious_id:
2674 format['format_id'] = '%s-%d' % (format_id, i)
2675 # Ensure there is no conflict between id and ext in format selection
2676 # See https://github.com/yt-dlp/yt-dlp/issues/1282
2677 if format['format_id'] != format['ext'] and format['format_id'] in common_exts:
2678 format['format_id'] = 'f%s' % format['format_id']
2679
2680 if format.get('format') is None:
2681 format['format'] = '{id} - {res}{note}'.format(
2682 id=format['format_id'],
2683 res=self.format_resolution(format),
2684 note=format_field(format, 'format_note', ' (%s)'),
2685 )
2686
2687 if self.params.get('check_formats') is True:
2688 formats = LazyList(self._check_formats(formats[::-1]), reverse=True)
2689
2690 if not formats or formats[0] is not info_dict:
2691 # only set the 'formats' fields if the original info_dict list them
2692 # otherwise we end up with a circular reference, the first (and unique)
2693 # element in the 'formats' field in info_dict is info_dict itself,
2694 # which can't be exported to json
2695 info_dict['formats'] = formats
2696
2697 info_dict, _ = self.pre_process(info_dict)
2698
2699 if self._match_entry(info_dict, incomplete=self._format_fields) is not None:
2700 return info_dict
2701
2702 self.post_extract(info_dict)
2703 info_dict, _ = self.pre_process(info_dict, 'after_filter')
2704
2705 # The pre-processors may have modified the formats
2706 formats = self._get_formats(info_dict)
2707
2708 list_only = self.params.get('simulate') == 'list_only'
2709 interactive_format_selection = not list_only and self.format_selector == '-'
2710 if self.params.get('list_thumbnails'):
2711 self.list_thumbnails(info_dict)
2712 if self.params.get('listsubtitles'):
2713 if 'automatic_captions' in info_dict:
2714 self.list_subtitles(
2715 info_dict['id'], automatic_captions, 'automatic captions')
2716 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
2717 if self.params.get('listformats') or interactive_format_selection:
2718 self.list_formats(info_dict)
2719 if list_only:
2720 # Without this printing, -F --print-json will not work
2721 self.__forced_printings(info_dict)
2722 return info_dict
2723
2724 format_selector = self.format_selector
2725 if format_selector is None:
2726 req_format = self._default_format_spec(info_dict, download=download)
2727 self.write_debug('Default format spec: %s' % req_format)
2728 format_selector = self.build_format_selector(req_format)
2729
2730 while True:
2731 if interactive_format_selection:
2732 req_format = input(
2733 self._format_screen('\nEnter format selector: ', self.Styles.EMPHASIS))
2734 try:
2735 format_selector = self.build_format_selector(req_format)
2736 except SyntaxError as err:
2737 self.report_error(err, tb=False, is_error=False)
2738 continue
2739
2740 formats_to_download = list(format_selector({
2741 'formats': formats,
2742 'has_merged_format': any('none' not in (f.get('acodec'), f.get('vcodec')) for f in formats),
2743 'incomplete_formats': (
2744 # All formats are video-only or
2745 all(f.get('vcodec') != 'none' and f.get('acodec') == 'none' for f in formats)
2746 # all formats are audio-only
2747 or all(f.get('vcodec') == 'none' and f.get('acodec') != 'none' for f in formats)),
2748 }))
2749 if interactive_format_selection and not formats_to_download:
2750 self.report_error('Requested format is not available', tb=False, is_error=False)
2751 continue
2752 break
2753
2754 if not formats_to_download:
2755 if not self.params.get('ignore_no_formats_error'):
2756 raise ExtractorError(
2757 'Requested format is not available. Use --list-formats for a list of available formats',
2758 expected=True, video_id=info_dict['id'], ie=info_dict['extractor'])
2759 self.report_warning('Requested format is not available')
2760 # Process what we can, even without any available formats.
2761 formats_to_download = [{}]
2762
2763 requested_ranges = tuple(self.params.get('download_ranges', lambda *_: [{}])(info_dict, self))
2764 best_format, downloaded_formats = formats_to_download[-1], []
2765 if download:
2766 if best_format and requested_ranges:
2767 def to_screen(*msg):
2768 self.to_screen(f'[info] {info_dict["id"]}: {" ".join(", ".join(variadic(m)) for m in msg)}')
2769
2770 to_screen(f'Downloading {len(formats_to_download)} format(s):',
2771 (f['format_id'] for f in formats_to_download))
2772 if requested_ranges != ({}, ):
2773 to_screen(f'Downloading {len(requested_ranges)} time ranges:',
2774 (f'{c["start_time"]:.1f}-{c["end_time"]:.1f}' for c in requested_ranges))
2775 max_downloads_reached = False
2776
2777 for fmt, chapter in itertools.product(formats_to_download, requested_ranges):
2778 new_info = self._copy_infodict(info_dict)
2779 new_info.update(fmt)
2780 offset, duration = info_dict.get('section_start') or 0, info_dict.get('duration') or float('inf')
2781 end_time = offset + min(chapter.get('end_time', duration), duration)
2782 if chapter or offset:
2783 new_info.update({
2784 'section_start': offset + chapter.get('start_time', 0),
2785 # duration may not be accurate. So allow deviations <1sec
2786 'section_end': end_time if end_time <= offset + duration + 1 else None,
2787 'section_title': chapter.get('title'),
2788 'section_number': chapter.get('index'),
2789 })
2790 downloaded_formats.append(new_info)
2791 try:
2792 self.process_info(new_info)
2793 except MaxDownloadsReached:
2794 max_downloads_reached = True
2795 self._raise_pending_errors(new_info)
2796 # Remove copied info
2797 for key, val in tuple(new_info.items()):
2798 if info_dict.get(key) == val:
2799 new_info.pop(key)
2800 if max_downloads_reached:
2801 break
2802
2803 write_archive = {f.get('__write_download_archive', False) for f in downloaded_formats}
2804 assert write_archive.issubset({True, False, 'ignore'})
2805 if True in write_archive and False not in write_archive:
2806 self.record_download_archive(info_dict)
2807
2808 info_dict['requested_downloads'] = downloaded_formats
2809 info_dict = self.run_all_pps('after_video', info_dict)
2810 if max_downloads_reached:
2811 raise MaxDownloadsReached()
2812
2813 # We update the info dict with the selected best quality format (backwards compatibility)
2814 info_dict.update(best_format)
2815 return info_dict
2816
2817 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
2818 """Select the requested subtitles and their format"""
2819 available_subs, normal_sub_langs = {}, []
2820 if normal_subtitles and self.params.get('writesubtitles'):
2821 available_subs.update(normal_subtitles)
2822 normal_sub_langs = tuple(normal_subtitles.keys())
2823 if automatic_captions and self.params.get('writeautomaticsub'):
2824 for lang, cap_info in automatic_captions.items():
2825 if lang not in available_subs:
2826 available_subs[lang] = cap_info
2827
2828 if not available_subs or (
2829 not self.params.get('writesubtitles')
2830 and not self.params.get('writeautomaticsub')):
2831 return None
2832
2833 all_sub_langs = tuple(available_subs.keys())
2834 if self.params.get('allsubtitles', False):
2835 requested_langs = all_sub_langs
2836 elif self.params.get('subtitleslangs', False):
2837 try:
2838 requested_langs = orderedSet_from_options(
2839 self.params.get('subtitleslangs'), {'all': all_sub_langs}, use_regex=True)
2840 except re.error as e:
2841 raise ValueError(f'Wrong regex for subtitlelangs: {e.pattern}')
2842 else:
2843 requested_langs = LazyList(itertools.chain(
2844 ['en'] if 'en' in normal_sub_langs else [],
2845 filter(lambda f: f.startswith('en'), normal_sub_langs),
2846 ['en'] if 'en' in all_sub_langs else [],
2847 filter(lambda f: f.startswith('en'), all_sub_langs),
2848 normal_sub_langs, all_sub_langs,
2849 ))[:1]
2850 if requested_langs:
2851 self.to_screen(f'[info] {video_id}: Downloading subtitles: {", ".join(requested_langs)}')
2852
2853 formats_query = self.params.get('subtitlesformat', 'best')
2854 formats_preference = formats_query.split('/') if formats_query else []
2855 subs = {}
2856 for lang in requested_langs:
2857 formats = available_subs.get(lang)
2858 if formats is None:
2859 self.report_warning(f'{lang} subtitles not available for {video_id}')
2860 continue
2861 for ext in formats_preference:
2862 if ext == 'best':
2863 f = formats[-1]
2864 break
2865 matches = list(filter(lambda f: f['ext'] == ext, formats))
2866 if matches:
2867 f = matches[-1]
2868 break
2869 else:
2870 f = formats[-1]
2871 self.report_warning(
2872 'No subtitle format found matching "%s" for language %s, '
2873 'using %s' % (formats_query, lang, f['ext']))
2874 subs[lang] = f
2875 return subs
2876
2877 def _forceprint(self, key, info_dict):
2878 if info_dict is None:
2879 return
2880 info_copy = info_dict.copy()
2881 info_copy.setdefault('filename', self.prepare_filename(info_dict))
2882 if info_dict.get('requested_formats') is not None:
2883 # For RTMP URLs, also include the playpath
2884 info_copy['urls'] = '\n'.join(f['url'] + f.get('play_path', '') for f in info_dict['requested_formats'])
2885 elif info_dict.get('url'):
2886 info_copy['urls'] = info_dict['url'] + info_dict.get('play_path', '')
2887 info_copy['formats_table'] = self.render_formats_table(info_dict)
2888 info_copy['thumbnails_table'] = self.render_thumbnails_table(info_dict)
2889 info_copy['subtitles_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('subtitles'))
2890 info_copy['automatic_captions_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('automatic_captions'))
2891
2892 def format_tmpl(tmpl):
2893 mobj = re.fullmatch(r'([\w.:,]|-\d|(?P<dict>{([\w.:,]|-\d)+}))+=?', tmpl)
2894 if not mobj:
2895 return tmpl
2896
2897 fmt = '%({})s'
2898 if tmpl.startswith('{'):
2899 tmpl, fmt = f'.{tmpl}', '%({})j'
2900 if tmpl.endswith('='):
2901 tmpl, fmt = tmpl[:-1], '{0} = %({0})#j'
2902 return '\n'.join(map(fmt.format, [tmpl] if mobj.group('dict') else tmpl.split(',')))
2903
2904 for tmpl in self.params['forceprint'].get(key, []):
2905 self.to_stdout(self.evaluate_outtmpl(format_tmpl(tmpl), info_copy))
2906
2907 for tmpl, file_tmpl in self.params['print_to_file'].get(key, []):
2908 filename = self.prepare_filename(info_dict, outtmpl=file_tmpl)
2909 tmpl = format_tmpl(tmpl)
2910 self.to_screen(f'[info] Writing {tmpl!r} to: {filename}')
2911 if self._ensure_dir_exists(filename):
2912 with open(filename, 'a', encoding='utf-8', newline='') as f:
2913 f.write(self.evaluate_outtmpl(tmpl, info_copy) + os.linesep)
2914
2915 return info_copy
2916
2917 def __forced_printings(self, info_dict, filename=None, incomplete=True):
2918 if (self.params.get('forcejson')
2919 or self.params['forceprint'].get('video')
2920 or self.params['print_to_file'].get('video')):
2921 self.post_extract(info_dict)
2922 if filename:
2923 info_dict['filename'] = filename
2924 info_copy = self._forceprint('video', info_dict)
2925
2926 def print_field(field, actual_field=None, optional=False):
2927 if actual_field is None:
2928 actual_field = field
2929 if self.params.get(f'force{field}') and (
2930 info_copy.get(field) is not None or (not optional and not incomplete)):
2931 self.to_stdout(info_copy[actual_field])
2932
2933 print_field('title')
2934 print_field('id')
2935 print_field('url', 'urls')
2936 print_field('thumbnail', optional=True)
2937 print_field('description', optional=True)
2938 if filename:
2939 print_field('filename')
2940 if self.params.get('forceduration') and info_copy.get('duration') is not None:
2941 self.to_stdout(formatSeconds(info_copy['duration']))
2942 print_field('format')
2943
2944 if self.params.get('forcejson'):
2945 self.to_stdout(json.dumps(self.sanitize_info(info_dict)))
2946
2947 def dl(self, name, info, subtitle=False, test=False):
2948 if not info.get('url'):
2949 self.raise_no_formats(info, True)
2950
2951 if test:
2952 verbose = self.params.get('verbose')
2953 params = {
2954 'test': True,
2955 'quiet': self.params.get('quiet') or not verbose,
2956 'verbose': verbose,
2957 'noprogress': not verbose,
2958 'nopart': True,
2959 'skip_unavailable_fragments': False,
2960 'keep_fragments': False,
2961 'overwrites': True,
2962 '_no_ytdl_file': True,
2963 }
2964 else:
2965 params = self.params
2966 fd = get_suitable_downloader(info, params, to_stdout=(name == '-'))(self, params)
2967 if not test:
2968 for ph in self._progress_hooks:
2969 fd.add_progress_hook(ph)
2970 urls = '", "'.join(
2971 (f['url'].split(',')[0] + ',<data>' if f['url'].startswith('data:') else f['url'])
2972 for f in info.get('requested_formats', []) or [info])
2973 self.write_debug(f'Invoking {fd.FD_NAME} downloader on "{urls}"')
2974
2975 # Note: Ideally info should be a deep-copied so that hooks cannot modify it.
2976 # But it may contain objects that are not deep-copyable
2977 new_info = self._copy_infodict(info)
2978 if new_info.get('http_headers') is None:
2979 new_info['http_headers'] = self._calc_headers(new_info)
2980 return fd.download(name, new_info, subtitle)
2981
2982 def existing_file(self, filepaths, *, default_overwrite=True):
2983 existing_files = list(filter(os.path.exists, orderedSet(filepaths)))
2984 if existing_files and not self.params.get('overwrites', default_overwrite):
2985 return existing_files[0]
2986
2987 for file in existing_files:
2988 self.report_file_delete(file)
2989 os.remove(file)
2990 return None
2991
2992 def process_info(self, info_dict):
2993 """Process a single resolved IE result. (Modifies it in-place)"""
2994
2995 assert info_dict.get('_type', 'video') == 'video'
2996 original_infodict = info_dict
2997
2998 if 'format' not in info_dict and 'ext' in info_dict:
2999 info_dict['format'] = info_dict['ext']
3000
3001 if self._match_entry(info_dict) is not None:
3002 info_dict['__write_download_archive'] = 'ignore'
3003 return
3004
3005 # Does nothing under normal operation - for backward compatibility of process_info
3006 self.post_extract(info_dict)
3007
3008 def replace_info_dict(new_info):
3009 nonlocal info_dict
3010 if new_info == info_dict:
3011 return
3012 info_dict.clear()
3013 info_dict.update(new_info)
3014
3015 new_info, _ = self.pre_process(info_dict, 'video')
3016 replace_info_dict(new_info)
3017 self._num_downloads += 1
3018
3019 # info_dict['_filename'] needs to be set for backward compatibility
3020 info_dict['_filename'] = full_filename = self.prepare_filename(info_dict, warn=True)
3021 temp_filename = self.prepare_filename(info_dict, 'temp')
3022 files_to_move = {}
3023
3024 # Forced printings
3025 self.__forced_printings(info_dict, full_filename, incomplete=('format' not in info_dict))
3026
3027 def check_max_downloads():
3028 if self._num_downloads >= float(self.params.get('max_downloads') or 'inf'):
3029 raise MaxDownloadsReached()
3030
3031 if self.params.get('simulate'):
3032 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
3033 check_max_downloads()
3034 return
3035
3036 if full_filename is None:
3037 return
3038 if not self._ensure_dir_exists(encodeFilename(full_filename)):
3039 return
3040 if not self._ensure_dir_exists(encodeFilename(temp_filename)):
3041 return
3042
3043 if self._write_description('video', info_dict,
3044 self.prepare_filename(info_dict, 'description')) is None:
3045 return
3046
3047 sub_files = self._write_subtitles(info_dict, temp_filename)
3048 if sub_files is None:
3049 return
3050 files_to_move.update(dict(sub_files))
3051
3052 thumb_files = self._write_thumbnails(
3053 'video', info_dict, temp_filename, self.prepare_filename(info_dict, 'thumbnail'))
3054 if thumb_files is None:
3055 return
3056 files_to_move.update(dict(thumb_files))
3057
3058 infofn = self.prepare_filename(info_dict, 'infojson')
3059 _infojson_written = self._write_info_json('video', info_dict, infofn)
3060 if _infojson_written:
3061 info_dict['infojson_filename'] = infofn
3062 # For backward compatibility, even though it was a private field
3063 info_dict['__infojson_filename'] = infofn
3064 elif _infojson_written is None:
3065 return
3066
3067 # Note: Annotations are deprecated
3068 annofn = None
3069 if self.params.get('writeannotations', False):
3070 annofn = self.prepare_filename(info_dict, 'annotation')
3071 if annofn:
3072 if not self._ensure_dir_exists(encodeFilename(annofn)):
3073 return
3074 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(annofn)):
3075 self.to_screen('[info] Video annotations are already present')
3076 elif not info_dict.get('annotations'):
3077 self.report_warning('There are no annotations to write.')
3078 else:
3079 try:
3080 self.to_screen('[info] Writing video annotations to: ' + annofn)
3081 with open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
3082 annofile.write(info_dict['annotations'])
3083 except (KeyError, TypeError):
3084 self.report_warning('There are no annotations to write.')
3085 except OSError:
3086 self.report_error('Cannot write annotations file: ' + annofn)
3087 return
3088
3089 # Write internet shortcut files
3090 def _write_link_file(link_type):
3091 url = try_get(info_dict['webpage_url'], iri_to_uri)
3092 if not url:
3093 self.report_warning(
3094 f'Cannot write internet shortcut file because the actual URL of "{info_dict["webpage_url"]}" is unknown')
3095 return True
3096 linkfn = replace_extension(self.prepare_filename(info_dict, 'link'), link_type, info_dict.get('ext'))
3097 if not self._ensure_dir_exists(encodeFilename(linkfn)):
3098 return False
3099 if self.params.get('overwrites', True) and os.path.exists(encodeFilename(linkfn)):
3100 self.to_screen(f'[info] Internet shortcut (.{link_type}) is already present')
3101 return True
3102 try:
3103 self.to_screen(f'[info] Writing internet shortcut (.{link_type}) to: {linkfn}')
3104 with open(encodeFilename(to_high_limit_path(linkfn)), 'w', encoding='utf-8',
3105 newline='\r\n' if link_type == 'url' else '\n') as linkfile:
3106 template_vars = {'url': url}
3107 if link_type == 'desktop':
3108 template_vars['filename'] = linkfn[:-(len(link_type) + 1)]
3109 linkfile.write(LINK_TEMPLATES[link_type] % template_vars)
3110 except OSError:
3111 self.report_error(f'Cannot write internet shortcut {linkfn}')
3112 return False
3113 return True
3114
3115 write_links = {
3116 'url': self.params.get('writeurllink'),
3117 'webloc': self.params.get('writewebloclink'),
3118 'desktop': self.params.get('writedesktoplink'),
3119 }
3120 if self.params.get('writelink'):
3121 link_type = ('webloc' if sys.platform == 'darwin'
3122 else 'desktop' if sys.platform.startswith('linux')
3123 else 'url')
3124 write_links[link_type] = True
3125
3126 if any(should_write and not _write_link_file(link_type)
3127 for link_type, should_write in write_links.items()):
3128 return
3129
3130 new_info, files_to_move = self.pre_process(info_dict, 'before_dl', files_to_move)
3131 replace_info_dict(new_info)
3132
3133 if self.params.get('skip_download'):
3134 info_dict['filepath'] = temp_filename
3135 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
3136 info_dict['__files_to_move'] = files_to_move
3137 replace_info_dict(self.run_pp(MoveFilesAfterDownloadPP(self, False), info_dict))
3138 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
3139 else:
3140 # Download
3141 info_dict.setdefault('__postprocessors', [])
3142 try:
3143
3144 def existing_video_file(*filepaths):
3145 ext = info_dict.get('ext')
3146 converted = lambda file: replace_extension(file, self.params.get('final_ext') or ext, ext)
3147 file = self.existing_file(itertools.chain(*zip(map(converted, filepaths), filepaths)),
3148 default_overwrite=False)
3149 if file:
3150 info_dict['ext'] = os.path.splitext(file)[1][1:]
3151 return file
3152
3153 fd, success = None, True
3154 if info_dict.get('protocol') or info_dict.get('url'):
3155 fd = get_suitable_downloader(info_dict, self.params, to_stdout=temp_filename == '-')
3156 if fd is not FFmpegFD and 'no-direct-merge' not in self.params['compat_opts'] and (
3157 info_dict.get('section_start') or info_dict.get('section_end')):
3158 msg = ('This format cannot be partially downloaded' if FFmpegFD.available()
3159 else 'You have requested downloading the video partially, but ffmpeg is not installed')
3160 self.report_error(f'{msg}. Aborting')
3161 return
3162
3163 if info_dict.get('requested_formats') is not None:
3164 requested_formats = info_dict['requested_formats']
3165 old_ext = info_dict['ext']
3166 if self.params.get('merge_output_format') is None:
3167 if (info_dict['ext'] == 'webm'
3168 and info_dict.get('thumbnails')
3169 # check with type instead of pp_key, __name__, or isinstance
3170 # since we dont want any custom PPs to trigger this
3171 and any(type(pp) == EmbedThumbnailPP for pp in self._pps['post_process'])): # noqa: E721
3172 info_dict['ext'] = 'mkv'
3173 self.report_warning(
3174 'webm doesn\'t support embedding a thumbnail, mkv will be used')
3175 new_ext = info_dict['ext']
3176
3177 def correct_ext(filename, ext=new_ext):
3178 if filename == '-':
3179 return filename
3180 filename_real_ext = os.path.splitext(filename)[1][1:]
3181 filename_wo_ext = (
3182 os.path.splitext(filename)[0]
3183 if filename_real_ext in (old_ext, new_ext)
3184 else filename)
3185 return f'{filename_wo_ext}.{ext}'
3186
3187 # Ensure filename always has a correct extension for successful merge
3188 full_filename = correct_ext(full_filename)
3189 temp_filename = correct_ext(temp_filename)
3190 dl_filename = existing_video_file(full_filename, temp_filename)
3191 info_dict['__real_download'] = False
3192
3193 merger = FFmpegMergerPP(self)
3194 downloaded = []
3195 if dl_filename is not None:
3196 self.report_file_already_downloaded(dl_filename)
3197 elif fd:
3198 for f in requested_formats if fd != FFmpegFD else []:
3199 f['filepath'] = fname = prepend_extension(
3200 correct_ext(temp_filename, info_dict['ext']),
3201 'f%s' % f['format_id'], info_dict['ext'])
3202 downloaded.append(fname)
3203 info_dict['url'] = '\n'.join(f['url'] for f in requested_formats)
3204 success, real_download = self.dl(temp_filename, info_dict)
3205 info_dict['__real_download'] = real_download
3206 else:
3207 if self.params.get('allow_unplayable_formats'):
3208 self.report_warning(
3209 'You have requested merging of multiple formats '
3210 'while also allowing unplayable formats to be downloaded. '
3211 'The formats won\'t be merged to prevent data corruption.')
3212 elif not merger.available:
3213 msg = 'You have requested merging of multiple formats but ffmpeg is not installed'
3214 if not self.params.get('ignoreerrors'):
3215 self.report_error(f'{msg}. Aborting due to --abort-on-error')
3216 return
3217 self.report_warning(f'{msg}. The formats won\'t be merged')
3218
3219 if temp_filename == '-':
3220 reason = ('using a downloader other than ffmpeg' if FFmpegFD.can_merge_formats(info_dict, self.params)
3221 else 'but the formats are incompatible for simultaneous download' if merger.available
3222 else 'but ffmpeg is not installed')
3223 self.report_warning(
3224 f'You have requested downloading multiple formats to stdout {reason}. '
3225 'The formats will be streamed one after the other')
3226 fname = temp_filename
3227 for f in requested_formats:
3228 new_info = dict(info_dict)
3229 del new_info['requested_formats']
3230 new_info.update(f)
3231 if temp_filename != '-':
3232 fname = prepend_extension(
3233 correct_ext(temp_filename, new_info['ext']),
3234 'f%s' % f['format_id'], new_info['ext'])
3235 if not self._ensure_dir_exists(fname):
3236 return
3237 f['filepath'] = fname
3238 downloaded.append(fname)
3239 partial_success, real_download = self.dl(fname, new_info)
3240 info_dict['__real_download'] = info_dict['__real_download'] or real_download
3241 success = success and partial_success
3242
3243 if downloaded and merger.available and not self.params.get('allow_unplayable_formats'):
3244 info_dict['__postprocessors'].append(merger)
3245 info_dict['__files_to_merge'] = downloaded
3246 # Even if there were no downloads, it is being merged only now
3247 info_dict['__real_download'] = True
3248 else:
3249 for file in downloaded:
3250 files_to_move[file] = None
3251 else:
3252 # Just a single file
3253 dl_filename = existing_video_file(full_filename, temp_filename)
3254 if dl_filename is None or dl_filename == temp_filename:
3255 # dl_filename == temp_filename could mean that the file was partially downloaded with --no-part.
3256 # So we should try to resume the download
3257 success, real_download = self.dl(temp_filename, info_dict)
3258 info_dict['__real_download'] = real_download
3259 else:
3260 self.report_file_already_downloaded(dl_filename)
3261
3262 dl_filename = dl_filename or temp_filename
3263 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
3264
3265 except network_exceptions as err:
3266 self.report_error('unable to download video data: %s' % error_to_compat_str(err))
3267 return
3268 except OSError as err:
3269 raise UnavailableVideoError(err)
3270 except (ContentTooShortError, ) as err:
3271 self.report_error(f'content too short (expected {err.expected} bytes and served {err.downloaded})')
3272 return
3273
3274 self._raise_pending_errors(info_dict)
3275 if success and full_filename != '-':
3276
3277 def fixup():
3278 do_fixup = True
3279 fixup_policy = self.params.get('fixup')
3280 vid = info_dict['id']
3281
3282 if fixup_policy in ('ignore', 'never'):
3283 return
3284 elif fixup_policy == 'warn':
3285 do_fixup = 'warn'
3286 elif fixup_policy != 'force':
3287 assert fixup_policy in ('detect_or_warn', None)
3288 if not info_dict.get('__real_download'):
3289 do_fixup = False
3290
3291 def ffmpeg_fixup(cndn, msg, cls):
3292 if not (do_fixup and cndn):
3293 return
3294 elif do_fixup == 'warn':
3295 self.report_warning(f'{vid}: {msg}')
3296 return
3297 pp = cls(self)
3298 if pp.available:
3299 info_dict['__postprocessors'].append(pp)
3300 else:
3301 self.report_warning(f'{vid}: {msg}. Install ffmpeg to fix this automatically')
3302
3303 stretched_ratio = info_dict.get('stretched_ratio')
3304 ffmpeg_fixup(stretched_ratio not in (1, None),
3305 f'Non-uniform pixel ratio {stretched_ratio}',
3306 FFmpegFixupStretchedPP)
3307
3308 downloader = get_suitable_downloader(info_dict, self.params) if 'protocol' in info_dict else None
3309 downloader = downloader.FD_NAME if downloader else None
3310
3311 ext = info_dict.get('ext')
3312 postprocessed_by_ffmpeg = info_dict.get('requested_formats') or any((
3313 isinstance(pp, FFmpegVideoConvertorPP)
3314 and resolve_recode_mapping(ext, pp.mapping)[0] not in (ext, None)
3315 ) for pp in self._pps['post_process'])
3316
3317 if not postprocessed_by_ffmpeg:
3318 ffmpeg_fixup(ext == 'm4a' and info_dict.get('container') == 'm4a_dash',
3319 'writing DASH m4a. Only some players support this container',
3320 FFmpegFixupM4aPP)
3321 ffmpeg_fixup(downloader == 'hlsnative' and not self.params.get('hls_use_mpegts')
3322 or info_dict.get('is_live') and self.params.get('hls_use_mpegts') is None,
3323 'Possible MPEG-TS in MP4 container or malformed AAC timestamps',
3324 FFmpegFixupM3u8PP)
3325 ffmpeg_fixup(info_dict.get('is_live') and downloader == 'dashsegments',
3326 'Possible duplicate MOOV atoms', FFmpegFixupDuplicateMoovPP)
3327
3328 ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed timestamps detected', FFmpegFixupTimestampPP)
3329 ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed duration detected', FFmpegFixupDurationPP)
3330
3331 fixup()
3332 try:
3333 replace_info_dict(self.post_process(dl_filename, info_dict, files_to_move))
3334 except PostProcessingError as err:
3335 self.report_error('Postprocessing: %s' % str(err))
3336 return
3337 try:
3338 for ph in self._post_hooks:
3339 ph(info_dict['filepath'])
3340 except Exception as err:
3341 self.report_error('post hooks: %s' % str(err))
3342 return
3343 info_dict['__write_download_archive'] = True
3344
3345 assert info_dict is original_infodict # Make sure the info_dict was modified in-place
3346 if self.params.get('force_write_download_archive'):
3347 info_dict['__write_download_archive'] = True
3348 check_max_downloads()
3349
3350 def __download_wrapper(self, func):
3351 @functools.wraps(func)
3352 def wrapper(*args, **kwargs):
3353 try:
3354 res = func(*args, **kwargs)
3355 except UnavailableVideoError as e:
3356 self.report_error(e)
3357 except DownloadCancelled as e:
3358 self.to_screen(f'[info] {e}')
3359 if not self.params.get('break_per_url'):
3360 raise
3361 self._num_downloads = 0
3362 else:
3363 if self.params.get('dump_single_json', False):
3364 self.post_extract(res)
3365 self.to_stdout(json.dumps(self.sanitize_info(res)))
3366 return wrapper
3367
3368 def download(self, url_list):
3369 """Download a given list of URLs."""
3370 url_list = variadic(url_list) # Passing a single URL is a common mistake
3371 outtmpl = self.params['outtmpl']['default']
3372 if (len(url_list) > 1
3373 and outtmpl != '-'
3374 and '%' not in outtmpl
3375 and self.params.get('max_downloads') != 1):
3376 raise SameFileError(outtmpl)
3377
3378 for url in url_list:
3379 self.__download_wrapper(self.extract_info)(
3380 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
3381
3382 return self._download_retcode
3383
3384 def download_with_info_file(self, info_filename):
3385 with contextlib.closing(fileinput.FileInput(
3386 [info_filename], mode='r',
3387 openhook=fileinput.hook_encoded('utf-8'))) as f:
3388 # FileInput doesn't have a read method, we can't call json.load
3389 infos = [self.sanitize_info(info, self.params.get('clean_infojson', True))
3390 for info in variadic(json.loads('\n'.join(f)))]
3391 for info in infos:
3392 try:
3393 self.__download_wrapper(self.process_ie_result)(info, download=True)
3394 except (DownloadError, EntryNotInPlaylist, ReExtractInfo) as e:
3395 if not isinstance(e, EntryNotInPlaylist):
3396 self.to_stderr('\r')
3397 webpage_url = info.get('webpage_url')
3398 if webpage_url is None:
3399 raise
3400 self.report_warning(f'The info failed to download: {e}; trying with URL {webpage_url}')
3401 self.download([webpage_url])
3402 return self._download_retcode
3403
3404 @staticmethod
3405 def sanitize_info(info_dict, remove_private_keys=False):
3406 ''' Sanitize the infodict for converting to json '''
3407 if info_dict is None:
3408 return info_dict
3409 info_dict.setdefault('epoch', int(time.time()))
3410 info_dict.setdefault('_type', 'video')
3411 info_dict.setdefault('_version', {
3412 'version': __version__,
3413 'current_git_head': current_git_head(),
3414 'release_git_head': RELEASE_GIT_HEAD,
3415 'repository': REPOSITORY,
3416 })
3417
3418 if remove_private_keys:
3419 reject = lambda k, v: v is None or k.startswith('__') or k in {
3420 'requested_downloads', 'requested_formats', 'requested_subtitles', 'requested_entries',
3421 'entries', 'filepath', '_filename', 'filename', 'infojson_filename', 'original_url',
3422 'playlist_autonumber', '_format_sort_fields',
3423 }
3424 else:
3425 reject = lambda k, v: False
3426
3427 def filter_fn(obj):
3428 if isinstance(obj, dict):
3429 return {k: filter_fn(v) for k, v in obj.items() if not reject(k, v)}
3430 elif isinstance(obj, (list, tuple, set, LazyList)):
3431 return list(map(filter_fn, obj))
3432 elif obj is None or isinstance(obj, (str, int, float, bool)):
3433 return obj
3434 else:
3435 return repr(obj)
3436
3437 return filter_fn(info_dict)
3438
3439 @staticmethod
3440 def filter_requested_info(info_dict, actually_filter=True):
3441 ''' Alias of sanitize_info for backward compatibility '''
3442 return YoutubeDL.sanitize_info(info_dict, actually_filter)
3443
3444 def _delete_downloaded_files(self, *files_to_delete, info={}, msg=None):
3445 for filename in set(filter(None, files_to_delete)):
3446 if msg:
3447 self.to_screen(msg % filename)
3448 try:
3449 os.remove(filename)
3450 except OSError:
3451 self.report_warning(f'Unable to delete file {filename}')
3452 if filename in info.get('__files_to_move', []): # NB: Delete even if None
3453 del info['__files_to_move'][filename]
3454
3455 @staticmethod
3456 def post_extract(info_dict):
3457 def actual_post_extract(info_dict):
3458 if info_dict.get('_type') in ('playlist', 'multi_video'):
3459 for video_dict in info_dict.get('entries', {}):
3460 actual_post_extract(video_dict or {})
3461 return
3462
3463 post_extractor = info_dict.pop('__post_extractor', None) or (lambda: {})
3464 info_dict.update(post_extractor())
3465
3466 actual_post_extract(info_dict or {})
3467
3468 def run_pp(self, pp, infodict):
3469 files_to_delete = []
3470 if '__files_to_move' not in infodict:
3471 infodict['__files_to_move'] = {}
3472 try:
3473 files_to_delete, infodict = pp.run(infodict)
3474 except PostProcessingError as e:
3475 # Must be True and not 'only_download'
3476 if self.params.get('ignoreerrors') is True:
3477 self.report_error(e)
3478 return infodict
3479 raise
3480
3481 if not files_to_delete:
3482 return infodict
3483 if self.params.get('keepvideo', False):
3484 for f in files_to_delete:
3485 infodict['__files_to_move'].setdefault(f, '')
3486 else:
3487 self._delete_downloaded_files(
3488 *files_to_delete, info=infodict, msg='Deleting original file %s (pass -k to keep)')
3489 return infodict
3490
3491 def run_all_pps(self, key, info, *, additional_pps=None):
3492 if key != 'video':
3493 self._forceprint(key, info)
3494 for pp in (additional_pps or []) + self._pps[key]:
3495 info = self.run_pp(pp, info)
3496 return info
3497
3498 def pre_process(self, ie_info, key='pre_process', files_to_move=None):
3499 info = dict(ie_info)
3500 info['__files_to_move'] = files_to_move or {}
3501 try:
3502 info = self.run_all_pps(key, info)
3503 except PostProcessingError as err:
3504 msg = f'Preprocessing: {err}'
3505 info.setdefault('__pending_error', msg)
3506 self.report_error(msg, is_error=False)
3507 return info, info.pop('__files_to_move', None)
3508
3509 def post_process(self, filename, info, files_to_move=None):
3510 """Run all the postprocessors on the given file."""
3511 info['filepath'] = filename
3512 info['__files_to_move'] = files_to_move or {}
3513 info = self.run_all_pps('post_process', info, additional_pps=info.get('__postprocessors'))
3514 info = self.run_pp(MoveFilesAfterDownloadPP(self), info)
3515 del info['__files_to_move']
3516 return self.run_all_pps('after_move', info)
3517
3518 def _make_archive_id(self, info_dict):
3519 video_id = info_dict.get('id')
3520 if not video_id:
3521 return
3522 # Future-proof against any change in case
3523 # and backwards compatibility with prior versions
3524 extractor = info_dict.get('extractor_key') or info_dict.get('ie_key') # key in a playlist
3525 if extractor is None:
3526 url = str_or_none(info_dict.get('url'))
3527 if not url:
3528 return
3529 # Try to find matching extractor for the URL and take its ie_key
3530 for ie_key, ie in self._ies.items():
3531 if ie.suitable(url):
3532 extractor = ie_key
3533 break
3534 else:
3535 return
3536 return make_archive_id(extractor, video_id)
3537
3538 def in_download_archive(self, info_dict):
3539 if not self.archive:
3540 return False
3541
3542 vid_ids = [self._make_archive_id(info_dict)]
3543 vid_ids.extend(info_dict.get('_old_archive_ids') or [])
3544 return any(id_ in self.archive for id_ in vid_ids)
3545
3546 def record_download_archive(self, info_dict):
3547 fn = self.params.get('download_archive')
3548 if fn is None:
3549 return
3550 vid_id = self._make_archive_id(info_dict)
3551 assert vid_id
3552
3553 self.write_debug(f'Adding to archive: {vid_id}')
3554 if is_path_like(fn):
3555 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
3556 archive_file.write(vid_id + '\n')
3557 self.archive.add(vid_id)
3558
3559 @staticmethod
3560 def format_resolution(format, default='unknown'):
3561 if format.get('vcodec') == 'none' and format.get('acodec') != 'none':
3562 return 'audio only'
3563 if format.get('resolution') is not None:
3564 return format['resolution']
3565 if format.get('width') and format.get('height'):
3566 return '%dx%d' % (format['width'], format['height'])
3567 elif format.get('height'):
3568 return '%sp' % format['height']
3569 elif format.get('width'):
3570 return '%dx?' % format['width']
3571 return default
3572
3573 def _list_format_headers(self, *headers):
3574 if self.params.get('listformats_table', True) is not False:
3575 return [self._format_out(header, self.Styles.HEADERS) for header in headers]
3576 return headers
3577
3578 def _format_note(self, fdict):
3579 res = ''
3580 if fdict.get('ext') in ['f4f', 'f4m']:
3581 res += '(unsupported)'
3582 if fdict.get('language'):
3583 if res:
3584 res += ' '
3585 res += '[%s]' % fdict['language']
3586 if fdict.get('format_note') is not None:
3587 if res:
3588 res += ' '
3589 res += fdict['format_note']
3590 if fdict.get('tbr') is not None:
3591 if res:
3592 res += ', '
3593 res += '%4dk' % fdict['tbr']
3594 if fdict.get('container') is not None:
3595 if res:
3596 res += ', '
3597 res += '%s container' % fdict['container']
3598 if (fdict.get('vcodec') is not None
3599 and fdict.get('vcodec') != 'none'):
3600 if res:
3601 res += ', '
3602 res += fdict['vcodec']
3603 if fdict.get('vbr') is not None:
3604 res += '@'
3605 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
3606 res += 'video@'
3607 if fdict.get('vbr') is not None:
3608 res += '%4dk' % fdict['vbr']
3609 if fdict.get('fps') is not None:
3610 if res:
3611 res += ', '
3612 res += '%sfps' % fdict['fps']
3613 if fdict.get('acodec') is not None:
3614 if res:
3615 res += ', '
3616 if fdict['acodec'] == 'none':
3617 res += 'video only'
3618 else:
3619 res += '%-5s' % fdict['acodec']
3620 elif fdict.get('abr') is not None:
3621 if res:
3622 res += ', '
3623 res += 'audio'
3624 if fdict.get('abr') is not None:
3625 res += '@%3dk' % fdict['abr']
3626 if fdict.get('asr') is not None:
3627 res += ' (%5dHz)' % fdict['asr']
3628 if fdict.get('filesize') is not None:
3629 if res:
3630 res += ', '
3631 res += format_bytes(fdict['filesize'])
3632 elif fdict.get('filesize_approx') is not None:
3633 if res:
3634 res += ', '
3635 res += '~' + format_bytes(fdict['filesize_approx'])
3636 return res
3637
3638 def _get_formats(self, info_dict):
3639 if info_dict.get('formats') is None:
3640 if info_dict.get('url') and info_dict.get('_type', 'video') == 'video':
3641 return [info_dict]
3642 return []
3643 return info_dict['formats']
3644
3645 def render_formats_table(self, info_dict):
3646 formats = self._get_formats(info_dict)
3647 if not formats:
3648 return
3649 if not self.params.get('listformats_table', True) is not False:
3650 table = [
3651 [
3652 format_field(f, 'format_id'),
3653 format_field(f, 'ext'),
3654 self.format_resolution(f),
3655 self._format_note(f)
3656 ] for f in formats if (f.get('preference') or 0) >= -1000]
3657 return render_table(['format code', 'extension', 'resolution', 'note'], table, extra_gap=1)
3658
3659 def simplified_codec(f, field):
3660 assert field in ('acodec', 'vcodec')
3661 codec = f.get(field, 'unknown')
3662 if not codec:
3663 return 'unknown'
3664 elif codec != 'none':
3665 return '.'.join(codec.split('.')[:4])
3666
3667 if field == 'vcodec' and f.get('acodec') == 'none':
3668 return 'images'
3669 elif field == 'acodec' and f.get('vcodec') == 'none':
3670 return ''
3671 return self._format_out('audio only' if field == 'vcodec' else 'video only',
3672 self.Styles.SUPPRESS)
3673
3674 delim = self._format_out('\u2502', self.Styles.DELIM, '|', test_encoding=True)
3675 table = [
3676 [
3677 self._format_out(format_field(f, 'format_id'), self.Styles.ID),
3678 format_field(f, 'ext'),
3679 format_field(f, func=self.format_resolution, ignore=('audio only', 'images')),
3680 format_field(f, 'fps', '\t%d', func=round),
3681 format_field(f, 'dynamic_range', '%s', ignore=(None, 'SDR')).replace('HDR', ''),
3682 format_field(f, 'audio_channels', '\t%s'),
3683 delim,
3684 format_field(f, 'filesize', ' \t%s', func=format_bytes) + format_field(f, 'filesize_approx', '~\t%s', func=format_bytes),
3685 format_field(f, 'tbr', '\t%dk', func=round),
3686 shorten_protocol_name(f.get('protocol', '')),
3687 delim,
3688 simplified_codec(f, 'vcodec'),
3689 format_field(f, 'vbr', '\t%dk', func=round),
3690 simplified_codec(f, 'acodec'),
3691 format_field(f, 'abr', '\t%dk', func=round),
3692 format_field(f, 'asr', '\t%s', func=format_decimal_suffix),
3693 join_nonempty(
3694 self._format_out('UNSUPPORTED', 'light red') if f.get('ext') in ('f4f', 'f4m') else None,
3695 self._format_out('DRM', 'light red') if f.get('has_drm') else None,
3696 format_field(f, 'language', '[%s]'),
3697 join_nonempty(format_field(f, 'format_note'),
3698 format_field(f, 'container', ignore=(None, f.get('ext'))),
3699 delim=', '),
3700 delim=' '),
3701 ] for f in formats if f.get('preference') is None or f['preference'] >= -1000]
3702 header_line = self._list_format_headers(
3703 'ID', 'EXT', 'RESOLUTION', '\tFPS', 'HDR', 'CH', delim, '\tFILESIZE', '\tTBR', 'PROTO',
3704 delim, 'VCODEC', '\tVBR', 'ACODEC', '\tABR', '\tASR', 'MORE INFO')
3705
3706 return render_table(
3707 header_line, table, hide_empty=True,
3708 delim=self._format_out('\u2500', self.Styles.DELIM, '-', test_encoding=True))
3709
3710 def render_thumbnails_table(self, info_dict):
3711 thumbnails = list(info_dict.get('thumbnails') or [])
3712 if not thumbnails:
3713 return None
3714 return render_table(
3715 self._list_format_headers('ID', 'Width', 'Height', 'URL'),
3716 [[t.get('id'), t.get('width') or 'unknown', t.get('height') or 'unknown', t['url']] for t in thumbnails])
3717
3718 def render_subtitles_table(self, video_id, subtitles):
3719 def _row(lang, formats):
3720 exts, names = zip(*((f['ext'], f.get('name') or 'unknown') for f in reversed(formats)))
3721 if len(set(names)) == 1:
3722 names = [] if names[0] == 'unknown' else names[:1]
3723 return [lang, ', '.join(names), ', '.join(exts)]
3724
3725 if not subtitles:
3726 return None
3727 return render_table(
3728 self._list_format_headers('Language', 'Name', 'Formats'),
3729 [_row(lang, formats) for lang, formats in subtitles.items()],
3730 hide_empty=True)
3731
3732 def __list_table(self, video_id, name, func, *args):
3733 table = func(*args)
3734 if not table:
3735 self.to_screen(f'{video_id} has no {name}')
3736 return
3737 self.to_screen(f'[info] Available {name} for {video_id}:')
3738 self.to_stdout(table)
3739
3740 def list_formats(self, info_dict):
3741 self.__list_table(info_dict['id'], 'formats', self.render_formats_table, info_dict)
3742
3743 def list_thumbnails(self, info_dict):
3744 self.__list_table(info_dict['id'], 'thumbnails', self.render_thumbnails_table, info_dict)
3745
3746 def list_subtitles(self, video_id, subtitles, name='subtitles'):
3747 self.__list_table(video_id, name, self.render_subtitles_table, video_id, subtitles)
3748
3749 def urlopen(self, req):
3750 """ Start an HTTP download """
3751 if isinstance(req, str):
3752 req = sanitized_Request(req)
3753 return self._opener.open(req, timeout=self._socket_timeout)
3754
3755 def print_debug_header(self):
3756 if not self.params.get('verbose'):
3757 return
3758
3759 from . import _IN_CLI # Must be delayed import
3760
3761 # These imports can be slow. So import them only as needed
3762 from .extractor.extractors import _LAZY_LOADER
3763 from .extractor.extractors import (
3764 _PLUGIN_CLASSES as plugin_ies,
3765 _PLUGIN_OVERRIDES as plugin_ie_overrides
3766 )
3767
3768 def get_encoding(stream):
3769 ret = str(getattr(stream, 'encoding', 'missing (%s)' % type(stream).__name__))
3770 if not supports_terminal_sequences(stream):
3771 from .utils import WINDOWS_VT_MODE # Must be imported locally
3772 ret += ' (No VT)' if WINDOWS_VT_MODE is False else ' (No ANSI)'
3773 return ret
3774
3775 encoding_str = 'Encodings: locale %s, fs %s, pref %s, %s' % (
3776 locale.getpreferredencoding(),
3777 sys.getfilesystemencoding(),
3778 self.get_encoding(),
3779 ', '.join(
3780 f'{key} {get_encoding(stream)}' for key, stream in self._out_files.items_
3781 if stream is not None and key != 'console')
3782 )
3783
3784 logger = self.params.get('logger')
3785 if logger:
3786 write_debug = lambda msg: logger.debug(f'[debug] {msg}')
3787 write_debug(encoding_str)
3788 else:
3789 write_string(f'[debug] {encoding_str}\n', encoding=None)
3790 write_debug = lambda msg: self._write_string(f'[debug] {msg}\n')
3791
3792 source = detect_variant()
3793 if VARIANT not in (None, 'pip'):
3794 source += '*'
3795 klass = type(self)
3796 write_debug(join_nonempty(
3797 f'{"yt-dlp" if REPOSITORY == "yt-dlp/yt-dlp" else REPOSITORY} version',
3798 f'{CHANNEL}@{__version__}',
3799 f'[{RELEASE_GIT_HEAD[:9]}]' if RELEASE_GIT_HEAD else '',
3800 '' if source == 'unknown' else f'({source})',
3801 '' if _IN_CLI else 'API' if klass == YoutubeDL else f'API:{self.__module__}.{klass.__qualname__}',
3802 delim=' '))
3803
3804 if not _IN_CLI:
3805 write_debug(f'params: {self.params}')
3806
3807 if not _LAZY_LOADER:
3808 if os.environ.get('YTDLP_NO_LAZY_EXTRACTORS'):
3809 write_debug('Lazy loading extractors is forcibly disabled')
3810 else:
3811 write_debug('Lazy loading extractors is disabled')
3812 if self.params['compat_opts']:
3813 write_debug('Compatibility options: %s' % ', '.join(self.params['compat_opts']))
3814
3815 if current_git_head():
3816 write_debug(f'Git HEAD: {current_git_head()}')
3817 write_debug(system_identifier())
3818
3819 exe_versions, ffmpeg_features = FFmpegPostProcessor.get_versions_and_features(self)
3820 ffmpeg_features = {key for key, val in ffmpeg_features.items() if val}
3821 if ffmpeg_features:
3822 exe_versions['ffmpeg'] += ' (%s)' % ','.join(sorted(ffmpeg_features))
3823
3824 exe_versions['rtmpdump'] = rtmpdump_version()
3825 exe_versions['phantomjs'] = PhantomJSwrapper._version()
3826 exe_str = ', '.join(
3827 f'{exe} {v}' for exe, v in sorted(exe_versions.items()) if v
3828 ) or 'none'
3829 write_debug('exe versions: %s' % exe_str)
3830
3831 from .compat.compat_utils import get_package_info
3832 from .dependencies import available_dependencies
3833
3834 write_debug('Optional libraries: %s' % (', '.join(sorted({
3835 join_nonempty(*get_package_info(m)) for m in available_dependencies.values()
3836 })) or 'none'))
3837
3838 self._setup_opener()
3839 proxy_map = {}
3840 for handler in self._opener.handlers:
3841 if hasattr(handler, 'proxies'):
3842 proxy_map.update(handler.proxies)
3843 write_debug(f'Proxy map: {proxy_map}')
3844
3845 for plugin_type, plugins in {'Extractor': plugin_ies, 'Post-Processor': plugin_pps}.items():
3846 display_list = ['%s%s' % (
3847 klass.__name__, '' if klass.__name__ == name else f' as {name}')
3848 for name, klass in plugins.items()]
3849 if plugin_type == 'Extractor':
3850 display_list.extend(f'{plugins[-1].IE_NAME.partition("+")[2]} ({parent.__name__})'
3851 for parent, plugins in plugin_ie_overrides.items())
3852 if not display_list:
3853 continue
3854 write_debug(f'{plugin_type} Plugins: {", ".join(sorted(display_list))}')
3855
3856 plugin_dirs = plugin_directories()
3857 if plugin_dirs:
3858 write_debug(f'Plugin directories: {plugin_dirs}')
3859
3860 # Not implemented
3861 if False and self.params.get('call_home'):
3862 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode()
3863 write_debug('Public IP address: %s' % ipaddr)
3864 latest_version = self.urlopen(
3865 'https://yt-dl.org/latest/version').read().decode()
3866 if version_tuple(latest_version) > version_tuple(__version__):
3867 self.report_warning(
3868 'You are using an outdated version (newest version: %s)! '
3869 'See https://yt-dl.org/update if you need help updating.' %
3870 latest_version)
3871
3872 def _setup_opener(self):
3873 if hasattr(self, '_opener'):
3874 return
3875 timeout_val = self.params.get('socket_timeout')
3876 self._socket_timeout = 20 if timeout_val is None else float(timeout_val)
3877
3878 opts_cookiesfrombrowser = self.params.get('cookiesfrombrowser')
3879 opts_cookiefile = self.params.get('cookiefile')
3880 opts_proxy = self.params.get('proxy')
3881
3882 self.cookiejar = load_cookies(opts_cookiefile, opts_cookiesfrombrowser, self)
3883
3884 cookie_processor = YoutubeDLCookieProcessor(self.cookiejar)
3885 if opts_proxy is not None:
3886 if opts_proxy == '':
3887 proxies = {}
3888 else:
3889 proxies = {'http': opts_proxy, 'https': opts_proxy}
3890 else:
3891 proxies = urllib.request.getproxies()
3892 # Set HTTPS proxy to HTTP one if given (https://github.com/ytdl-org/youtube-dl/issues/805)
3893 if 'http' in proxies and 'https' not in proxies:
3894 proxies['https'] = proxies['http']
3895 proxy_handler = PerRequestProxyHandler(proxies)
3896
3897 debuglevel = 1 if self.params.get('debug_printtraffic') else 0
3898 https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
3899 ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
3900 redirect_handler = YoutubeDLRedirectHandler()
3901 data_handler = urllib.request.DataHandler()
3902
3903 # When passing our own FileHandler instance, build_opener won't add the
3904 # default FileHandler and allows us to disable the file protocol, which
3905 # can be used for malicious purposes (see
3906 # https://github.com/ytdl-org/youtube-dl/issues/8227)
3907 file_handler = urllib.request.FileHandler()
3908
3909 if not self.params.get('enable_file_urls'):
3910 def file_open(*args, **kwargs):
3911 raise urllib.error.URLError(
3912 'file:// URLs are explicitly disabled in yt-dlp for security reasons. '
3913 'Use --enable-file-urls to enable at your own risk.')
3914 file_handler.file_open = file_open
3915
3916 opener = urllib.request.build_opener(
3917 proxy_handler, https_handler, cookie_processor, ydlh, redirect_handler, data_handler, file_handler)
3918
3919 # Delete the default user-agent header, which would otherwise apply in
3920 # cases where our custom HTTP handler doesn't come into play
3921 # (See https://github.com/ytdl-org/youtube-dl/issues/1309 for details)
3922 opener.addheaders = []
3923 self._opener = opener
3924
3925 def encode(self, s):
3926 if isinstance(s, bytes):
3927 return s # Already encoded
3928
3929 try:
3930 return s.encode(self.get_encoding())
3931 except UnicodeEncodeError as err:
3932 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
3933 raise
3934
3935 def get_encoding(self):
3936 encoding = self.params.get('encoding')
3937 if encoding is None:
3938 encoding = preferredencoding()
3939 return encoding
3940
3941 def _write_info_json(self, label, ie_result, infofn, overwrite=None):
3942 ''' Write infojson and returns True = written, 'exists' = Already exists, False = skip, None = error '''
3943 if overwrite is None:
3944 overwrite = self.params.get('overwrites', True)
3945 if not self.params.get('writeinfojson'):
3946 return False
3947 elif not infofn:
3948 self.write_debug(f'Skipping writing {label} infojson')
3949 return False
3950 elif not self._ensure_dir_exists(infofn):
3951 return None
3952 elif not overwrite and os.path.exists(infofn):
3953 self.to_screen(f'[info] {label.title()} metadata is already present')
3954 return 'exists'
3955
3956 self.to_screen(f'[info] Writing {label} metadata as JSON to: {infofn}')
3957 try:
3958 write_json_file(self.sanitize_info(ie_result, self.params.get('clean_infojson', True)), infofn)
3959 return True
3960 except OSError:
3961 self.report_error(f'Cannot write {label} metadata to JSON file {infofn}')
3962 return None
3963
3964 def _write_description(self, label, ie_result, descfn):
3965 ''' Write description and returns True = written, False = skip, None = error '''
3966 if not self.params.get('writedescription'):
3967 return False
3968 elif not descfn:
3969 self.write_debug(f'Skipping writing {label} description')
3970 return False
3971 elif not self._ensure_dir_exists(descfn):
3972 return None
3973 elif not self.params.get('overwrites', True) and os.path.exists(descfn):
3974 self.to_screen(f'[info] {label.title()} description is already present')
3975 elif ie_result.get('description') is None:
3976 self.to_screen(f'[info] There\'s no {label} description to write')
3977 return False
3978 else:
3979 try:
3980 self.to_screen(f'[info] Writing {label} description to: {descfn}')
3981 with open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
3982 descfile.write(ie_result['description'])
3983 except OSError:
3984 self.report_error(f'Cannot write {label} description file {descfn}')
3985 return None
3986 return True
3987
3988 def _write_subtitles(self, info_dict, filename):
3989 ''' Write subtitles to file and return list of (sub_filename, final_sub_filename); or None if error'''
3990 ret = []
3991 subtitles = info_dict.get('requested_subtitles')
3992 if not (self.params.get('writesubtitles') or self.params.get('writeautomaticsub')):
3993 # subtitles download errors are already managed as troubles in relevant IE
3994 # that way it will silently go on when used with unsupporting IE
3995 return ret
3996 elif not subtitles:
3997 self.to_screen('[info] There are no subtitles for the requested languages')
3998 return ret
3999 sub_filename_base = self.prepare_filename(info_dict, 'subtitle')
4000 if not sub_filename_base:
4001 self.to_screen('[info] Skipping writing video subtitles')
4002 return ret
4003
4004 for sub_lang, sub_info in subtitles.items():
4005 sub_format = sub_info['ext']
4006 sub_filename = subtitles_filename(filename, sub_lang, sub_format, info_dict.get('ext'))
4007 sub_filename_final = subtitles_filename(sub_filename_base, sub_lang, sub_format, info_dict.get('ext'))
4008 existing_sub = self.existing_file((sub_filename_final, sub_filename))
4009 if existing_sub:
4010 self.to_screen(f'[info] Video subtitle {sub_lang}.{sub_format} is already present')
4011 sub_info['filepath'] = existing_sub
4012 ret.append((existing_sub, sub_filename_final))
4013 continue
4014
4015 self.to_screen(f'[info] Writing video subtitles to: {sub_filename}')
4016 if sub_info.get('data') is not None:
4017 try:
4018 # Use newline='' to prevent conversion of newline characters
4019 # See https://github.com/ytdl-org/youtube-dl/issues/10268
4020 with open(sub_filename, 'w', encoding='utf-8', newline='') as subfile:
4021 subfile.write(sub_info['data'])
4022 sub_info['filepath'] = sub_filename
4023 ret.append((sub_filename, sub_filename_final))
4024 continue
4025 except OSError:
4026 self.report_error(f'Cannot write video subtitles file {sub_filename}')
4027 return None
4028
4029 try:
4030 sub_copy = sub_info.copy()
4031 sub_copy.setdefault('http_headers', info_dict.get('http_headers'))
4032 self.dl(sub_filename, sub_copy, subtitle=True)
4033 sub_info['filepath'] = sub_filename
4034 ret.append((sub_filename, sub_filename_final))
4035 except (DownloadError, ExtractorError, IOError, OSError, ValueError) + network_exceptions as err:
4036 msg = f'Unable to download video subtitles for {sub_lang!r}: {err}'
4037 if self.params.get('ignoreerrors') is not True: # False or 'only_download'
4038 if not self.params.get('ignoreerrors'):
4039 self.report_error(msg)
4040 raise DownloadError(msg)
4041 self.report_warning(msg)
4042 return ret
4043
4044 def _write_thumbnails(self, label, info_dict, filename, thumb_filename_base=None):
4045 ''' Write thumbnails to file and return list of (thumb_filename, final_thumb_filename) '''
4046 write_all = self.params.get('write_all_thumbnails', False)
4047 thumbnails, ret = [], []
4048 if write_all or self.params.get('writethumbnail', False):
4049 thumbnails = info_dict.get('thumbnails') or []
4050 if not thumbnails:
4051 self.to_screen(f'[info] There are no {label} thumbnails to download')
4052 return ret
4053 multiple = write_all and len(thumbnails) > 1
4054
4055 if thumb_filename_base is None:
4056 thumb_filename_base = filename
4057 if thumbnails and not thumb_filename_base:
4058 self.write_debug(f'Skipping writing {label} thumbnail')
4059 return ret
4060
4061 for idx, t in list(enumerate(thumbnails))[::-1]:
4062 thumb_ext = (f'{t["id"]}.' if multiple else '') + determine_ext(t['url'], 'jpg')
4063 thumb_display_id = f'{label} thumbnail {t["id"]}'
4064 thumb_filename = replace_extension(filename, thumb_ext, info_dict.get('ext'))
4065 thumb_filename_final = replace_extension(thumb_filename_base, thumb_ext, info_dict.get('ext'))
4066
4067 existing_thumb = self.existing_file((thumb_filename_final, thumb_filename))
4068 if existing_thumb:
4069 self.to_screen('[info] %s is already present' % (
4070 thumb_display_id if multiple else f'{label} thumbnail').capitalize())
4071 t['filepath'] = existing_thumb
4072 ret.append((existing_thumb, thumb_filename_final))
4073 else:
4074 self.to_screen(f'[info] Downloading {thumb_display_id} ...')
4075 try:
4076 uf = self.urlopen(sanitized_Request(t['url'], headers=t.get('http_headers', {})))
4077 self.to_screen(f'[info] Writing {thumb_display_id} to: {thumb_filename}')
4078 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
4079 shutil.copyfileobj(uf, thumbf)
4080 ret.append((thumb_filename, thumb_filename_final))
4081 t['filepath'] = thumb_filename
4082 except network_exceptions as err:
4083 thumbnails.pop(idx)
4084 self.report_warning(f'Unable to download {thumb_display_id}: {err}')
4085 if ret and not write_all:
4086 break
4087 return ret