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