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