]> jfr.im git - yt-dlp.git/blob - yt_dlp/YoutubeDL.py
e51bceef342908e123bbf931f0bbd3d5c27f0953
[yt-dlp.git] / yt_dlp / YoutubeDL.py
1 import collections
2 import contextlib
3 import datetime
4 import errno
5 import fileinput
6 import functools
7 import io
8 import itertools
9 import json
10 import locale
11 import operator
12 import os
13 import random
14 import re
15 import shutil
16 import string
17 import subprocess
18 import sys
19 import tempfile
20 import time
21 import tokenize
22 import traceback
23 import unicodedata
24
25 from .cache import Cache
26 from .compat import urllib # isort: split
27 from .compat import compat_os_name, compat_shlex_quote
28 from .cookies import load_cookies
29 from .downloader import FFmpegFD, get_suitable_downloader, shorten_protocol_name
30 from .downloader.rtmp import rtmpdump_version
31 from .extractor import gen_extractor_classes, get_info_extractor
32 from .extractor.common import UnsupportedURLIE
33 from .extractor.openload import PhantomJSwrapper
34 from .minicurses import format_text
35 from .plugins import directories as plugin_directories
36 from .postprocessor import _PLUGIN_CLASSES as plugin_pps
37 from .postprocessor import (
38 EmbedThumbnailPP,
39 FFmpegFixupDuplicateMoovPP,
40 FFmpegFixupDurationPP,
41 FFmpegFixupM3u8PP,
42 FFmpegFixupM4aPP,
43 FFmpegFixupStretchedPP,
44 FFmpegFixupTimestampPP,
45 FFmpegMergerPP,
46 FFmpegPostProcessor,
47 FFmpegVideoConvertorPP,
48 MoveFilesAfterDownloadPP,
49 get_postprocessor,
50 )
51 from .postprocessor.ffmpeg import resolve_mapping as resolve_recode_mapping
52 from .update import REPOSITORY, current_git_head, detect_variant
53 from .utils import (
54 DEFAULT_OUTTMPL,
55 IDENTITY,
56 LINK_TEMPLATES,
57 MEDIA_EXTENSIONS,
58 NO_DEFAULT,
59 NUMBER_RE,
60 OUTTMPL_TYPES,
61 POSTPROCESS_WHEN,
62 STR_FORMAT_RE_TMPL,
63 STR_FORMAT_TYPES,
64 ContentTooShortError,
65 DateRange,
66 DownloadCancelled,
67 DownloadError,
68 EntryNotInPlaylist,
69 ExistingVideoReached,
70 ExtractorError,
71 FormatSorter,
72 GeoRestrictedError,
73 HEADRequest,
74 ISO3166Utils,
75 LazyList,
76 MaxDownloadsReached,
77 Namespace,
78 PagedList,
79 PerRequestProxyHandler,
80 PlaylistEntries,
81 Popen,
82 PostProcessingError,
83 ReExtractInfo,
84 RejectedVideoReached,
85 SameFileError,
86 UnavailableVideoError,
87 UserNotLive,
88 YoutubeDLCookieProcessor,
89 YoutubeDLHandler,
90 YoutubeDLRedirectHandler,
91 age_restricted,
92 args_to_str,
93 bug_reports_message,
94 date_from_str,
95 deprecation_warning,
96 determine_ext,
97 determine_protocol,
98 encode_compat_str,
99 encodeFilename,
100 error_to_compat_str,
101 escapeHTML,
102 expand_path,
103 filter_dict,
104 float_or_none,
105 format_bytes,
106 format_decimal_suffix,
107 format_field,
108 formatSeconds,
109 get_compatible_ext,
110 get_domain,
111 int_or_none,
112 iri_to_uri,
113 is_path_like,
114 join_nonempty,
115 locked_file,
116 make_archive_id,
117 make_dir,
118 make_HTTPS_handler,
119 merge_headers,
120 network_exceptions,
121 number_of_digits,
122 orderedSet,
123 orderedSet_from_options,
124 parse_filesize,
125 preferredencoding,
126 prepend_extension,
127 remove_terminal_sequences,
128 render_table,
129 replace_extension,
130 sanitize_filename,
131 sanitize_path,
132 sanitize_url,
133 sanitized_Request,
134 std_headers,
135 str_or_none,
136 strftime_or_none,
137 subtitles_filename,
138 supports_terminal_sequences,
139 system_identifier,
140 timetuple_from_msec,
141 to_high_limit_path,
142 traverse_obj,
143 try_call,
144 try_get,
145 url_basename,
146 variadic,
147 version_tuple,
148 windows_enable_vt_mode,
149 write_json_file,
150 write_string,
151 )
152 from .version import CHANNEL, RELEASE_GIT_HEAD, VARIANT, __version__
153
154 if compat_os_name == 'nt':
155 import ctypes
156
157
158 class YoutubeDL:
159 """YoutubeDL class.
160
161 YoutubeDL objects are the ones responsible of downloading the
162 actual video file and writing it to disk if the user has requested
163 it, among some other tasks. In most cases there should be one per
164 program. As, given a video URL, the downloader doesn't know how to
165 extract all the needed information, task that InfoExtractors do, it
166 has to pass the URL to one of them.
167
168 For this, YoutubeDL objects have a method that allows
169 InfoExtractors to be registered in a given order. When it is passed
170 a URL, the YoutubeDL object handles it to the first InfoExtractor it
171 finds that reports being able to handle it. The InfoExtractor extracts
172 all the information about the video or videos the URL refers to, and
173 YoutubeDL process the extracted information, possibly using a File
174 Downloader to download the video.
175
176 YoutubeDL objects accept a lot of parameters. In order not to saturate
177 the object constructor with arguments, it receives a dictionary of
178 options instead. These options are available through the params
179 attribute for the InfoExtractors to use. The YoutubeDL also
180 registers itself as the downloader in charge for the InfoExtractors
181 that are added to it, so this is a "mutual registration".
182
183 Available options:
184
185 username: Username for authentication purposes.
186 password: Password for authentication purposes.
187 videopassword: Password for accessing a video.
188 ap_mso: Adobe Pass multiple-system operator identifier.
189 ap_username: Multiple-system operator account username.
190 ap_password: Multiple-system operator account password.
191 usenetrc: Use netrc for authentication instead.
192 netrc_location: Location of the netrc file. Defaults to ~/.netrc.
193 netrc_cmd: Use a shell command to get credentials
194 verbose: Print additional info to stdout.
195 quiet: Do not print messages to stdout.
196 no_warnings: Do not print out anything for warnings.
197 forceprint: A dict with keys WHEN mapped to a list of templates to
198 print to stdout. The allowed keys are video or any of the
199 items in utils.POSTPROCESS_WHEN.
200 For compatibility, a single list is also accepted
201 print_to_file: A dict with keys WHEN (same as forceprint) mapped to
202 a list of tuples with (template, filename)
203 forcejson: Force printing info_dict as JSON.
204 dump_single_json: Force printing the info_dict of the whole playlist
205 (or video) as a single JSON line.
206 force_write_download_archive: Force writing download archive regardless
207 of 'skip_download' or 'simulate'.
208 simulate: Do not download the video files. If unset (or None),
209 simulate only if listsubtitles, listformats or list_thumbnails is used
210 format: Video format code. see "FORMAT SELECTION" for more details.
211 You can also pass a function. The function takes 'ctx' as
212 argument and returns the formats to download.
213 See "build_format_selector" for an implementation
214 allow_unplayable_formats: Allow unplayable formats to be extracted and downloaded.
215 ignore_no_formats_error: Ignore "No video formats" error. Usefull for
216 extracting metadata even if the video is not actually
217 available for download (experimental)
218 format_sort: A list of fields by which to sort the video formats.
219 See "Sorting Formats" for more details.
220 format_sort_force: Force the given format_sort. see "Sorting Formats"
221 for more details.
222 prefer_free_formats: Whether to prefer video formats with free containers
223 over non-free ones of same quality.
224 allow_multiple_video_streams: Allow multiple video streams to be merged
225 into a single file
226 allow_multiple_audio_streams: Allow multiple audio streams to be merged
227 into a single file
228 check_formats Whether to test if the formats are downloadable.
229 Can be True (check all), False (check none),
230 'selected' (check selected formats),
231 or None (check only if requested by extractor)
232 paths: Dictionary of output paths. The allowed keys are 'home'
233 'temp' and the keys of OUTTMPL_TYPES (in utils.py)
234 outtmpl: Dictionary of templates for output names. Allowed keys
235 are 'default' and the keys of OUTTMPL_TYPES (in utils.py).
236 For compatibility with youtube-dl, a single string can also be used
237 outtmpl_na_placeholder: Placeholder for unavailable meta fields.
238 restrictfilenames: Do not allow "&" and spaces in file names
239 trim_file_name: Limit length of filename (extension excluded)
240 windowsfilenames: Force the filenames to be windows compatible
241 ignoreerrors: Do not stop on download/postprocessing errors.
242 Can be 'only_download' to ignore only download errors.
243 Default is 'only_download' for CLI, but False for API
244 skip_playlist_after_errors: Number of allowed failures until the rest of
245 the playlist is skipped
246 allowed_extractors: List of regexes to match against extractor names that are allowed
247 overwrites: Overwrite all video and metadata files if True,
248 overwrite only non-video files if None
249 and don't overwrite any file if False
250 For compatibility with youtube-dl,
251 "nooverwrites" may also be used instead
252 playlist_items: Specific indices of playlist to download.
253 playlistrandom: Download playlist items in random order.
254 lazy_playlist: Process playlist entries as they are received.
255 matchtitle: Download only matching titles.
256 rejecttitle: Reject downloads for matching titles.
257 logger: Log messages to a logging.Logger instance.
258 logtostderr: Print everything to stderr instead of stdout.
259 consoletitle: Display progress in console window's titlebar.
260 writedescription: Write the video description to a .description file
261 writeinfojson: Write the video description to a .info.json file
262 clean_infojson: Remove private fields from the infojson
263 getcomments: Extract video comments. This will not be written to disk
264 unless writeinfojson is also given
265 writeannotations: Write the video annotations to a .annotations.xml file
266 writethumbnail: Write the thumbnail image to a file
267 allow_playlist_files: Whether to write playlists' description, infojson etc
268 also to disk when using the 'write*' options
269 write_all_thumbnails: Write all thumbnail formats to files
270 writelink: Write an internet shortcut file, depending on the
271 current platform (.url/.webloc/.desktop)
272 writeurllink: Write a Windows internet shortcut file (.url)
273 writewebloclink: Write a macOS internet shortcut file (.webloc)
274 writedesktoplink: Write a Linux internet shortcut file (.desktop)
275 writesubtitles: Write the video subtitles to a file
276 writeautomaticsub: Write the automatically generated subtitles to a file
277 listsubtitles: Lists all available subtitles for the video
278 subtitlesformat: The format code for subtitles
279 subtitleslangs: List of languages of the subtitles to download (can be regex).
280 The list may contain "all" to refer to all the available
281 subtitles. The language can be prefixed with a "-" to
282 exclude it from the requested languages, e.g. ['all', '-live_chat']
283 keepvideo: Keep the video file after post-processing
284 daterange: A utils.DateRange object, download only if the upload_date is in the range.
285 skip_download: Skip the actual download of the video file
286 cachedir: Location of the cache files in the filesystem.
287 False to disable filesystem cache.
288 noplaylist: Download single video instead of a playlist if in doubt.
289 age_limit: An integer representing the user's age in years.
290 Unsuitable videos for the given age are skipped.
291 min_views: An integer representing the minimum view count the video
292 must have in order to not be skipped.
293 Videos without view count information are always
294 downloaded. None for no limit.
295 max_views: An integer representing the maximum view count.
296 Videos that are more popular than that are not
297 downloaded.
298 Videos without view count information are always
299 downloaded. None for no limit.
300 download_archive: A set, or the name of a file where all downloads are recorded.
301 Videos already present in the file are not downloaded again.
302 break_on_existing: Stop the download process after attempting to download a
303 file that is in the archive.
304 break_per_url: Whether break_on_reject and break_on_existing
305 should act on each input URL as opposed to for the entire queue
306 cookiefile: File name or text stream from where cookies should be read and dumped to
307 cookiesfrombrowser: A tuple containing the name of the browser, the profile
308 name/path from where cookies are loaded, the name of the keyring,
309 and the container name, e.g. ('chrome', ) or
310 ('vivaldi', 'default', 'BASICTEXT') or ('firefox', 'default', None, 'Meta')
311 legacyserverconnect: Explicitly allow HTTPS connection to servers that do not
312 support RFC 5746 secure renegotiation
313 nocheckcertificate: Do not verify SSL certificates
314 client_certificate: Path to client certificate file in PEM format. May include the private key
315 client_certificate_key: Path to private key file for client certificate
316 client_certificate_password: Password for client certificate private key, if encrypted.
317 If not provided and the key is encrypted, yt-dlp will ask interactively
318 prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
319 (Only supported by some extractors)
320 enable_file_urls: Enable file:// URLs. This is disabled by default for security reasons.
321 http_headers: A dictionary of custom headers to be used for all requests
322 proxy: URL of the proxy server to use
323 geo_verification_proxy: URL of the proxy to use for IP address verification
324 on geo-restricted sites.
325 socket_timeout: Time to wait for unresponsive hosts, in seconds
326 bidi_workaround: Work around buggy terminals without bidirectional text
327 support, using fridibi
328 debug_printtraffic:Print out sent and received HTTP traffic
329 default_search: Prepend this string if an input url is not valid.
330 'auto' for elaborate guessing
331 encoding: Use this encoding instead of the system-specified.
332 extract_flat: Whether to resolve and process url_results further
333 * False: Always process. Default for API
334 * True: Never process
335 * 'in_playlist': Do not process inside playlist/multi_video
336 * 'discard': Always process, but don't return the result
337 from inside playlist/multi_video
338 * 'discard_in_playlist': Same as "discard", but only for
339 playlists (not multi_video). Default for CLI
340 wait_for_video: If given, wait for scheduled streams to become available.
341 The value should be a tuple containing the range
342 (min_secs, max_secs) to wait between retries
343 postprocessors: A list of dictionaries, each with an entry
344 * key: The name of the postprocessor. See
345 yt_dlp/postprocessor/__init__.py for a list.
346 * when: When to run the postprocessor. Allowed values are
347 the entries of utils.POSTPROCESS_WHEN
348 Assumed to be 'post_process' if not given
349 progress_hooks: A list of functions that get called on download
350 progress, with a dictionary with the entries
351 * status: One of "downloading", "error", or "finished".
352 Check this first and ignore unknown values.
353 * info_dict: The extracted info_dict
354
355 If status is one of "downloading", or "finished", the
356 following properties may also be present:
357 * filename: The final filename (always present)
358 * tmpfilename: The filename we're currently writing to
359 * downloaded_bytes: Bytes on disk
360 * total_bytes: Size of the whole file, None if unknown
361 * total_bytes_estimate: Guess of the eventual file size,
362 None if unavailable.
363 * elapsed: The number of seconds since download started.
364 * eta: The estimated time in seconds, None if unknown
365 * speed: The download speed in bytes/second, None if
366 unknown
367 * fragment_index: The counter of the currently
368 downloaded video fragment.
369 * fragment_count: The number of fragments (= individual
370 files that will be merged)
371
372 Progress hooks are guaranteed to be called at least once
373 (with status "finished") if the download is successful.
374 postprocessor_hooks: A list of functions that get called on postprocessing
375 progress, with a dictionary with the entries
376 * status: One of "started", "processing", or "finished".
377 Check this first and ignore unknown values.
378 * postprocessor: Name of the postprocessor
379 * info_dict: The extracted info_dict
380
381 Progress hooks are guaranteed to be called at least twice
382 (with status "started" and "finished") if the processing is successful.
383 merge_output_format: "/" separated list of extensions to use when merging formats.
384 final_ext: Expected final extension; used to detect when the file was
385 already downloaded and converted
386 fixup: Automatically correct known faults of the file.
387 One of:
388 - "never": do nothing
389 - "warn": only emit a warning
390 - "detect_or_warn": check whether we can do anything
391 about it, warn otherwise (default)
392 source_address: Client-side IP address to bind to.
393 sleep_interval_requests: Number of seconds to sleep between requests
394 during extraction
395 sleep_interval: Number of seconds to sleep before each download when
396 used alone or a lower bound of a range for randomized
397 sleep before each download (minimum possible number
398 of seconds to sleep) when used along with
399 max_sleep_interval.
400 max_sleep_interval:Upper bound of a range for randomized sleep before each
401 download (maximum possible number of seconds to sleep).
402 Must only be used along with sleep_interval.
403 Actual sleep time will be a random float from range
404 [sleep_interval; max_sleep_interval].
405 sleep_interval_subtitles: Number of seconds to sleep before each subtitle download
406 listformats: Print an overview of available video formats and exit.
407 list_thumbnails: Print a table of all thumbnails and exit.
408 match_filter: A function that gets called for every video with the signature
409 (info_dict, *, incomplete: bool) -> Optional[str]
410 For backward compatibility with youtube-dl, the signature
411 (info_dict) -> Optional[str] is also allowed.
412 - If it returns a message, the video is ignored.
413 - If it returns None, the video is downloaded.
414 - If it returns utils.NO_DEFAULT, the user is interactively
415 asked whether to download the video.
416 - Raise utils.DownloadCancelled(msg) to abort remaining
417 downloads when a video is rejected.
418 match_filter_func in utils.py is one example for this.
419 color: A Dictionary with output stream names as keys
420 and their respective color policy as values.
421 Can also just be a single color policy,
422 in which case it applies to all outputs.
423 Valid stream names are 'stdout' and 'stderr'.
424 Valid color policies are one of 'always', 'auto', 'no_color' or 'never'.
425 geo_bypass: Bypass geographic restriction via faking X-Forwarded-For
426 HTTP header
427 geo_bypass_country:
428 Two-letter ISO 3166-2 country code that will be used for
429 explicit geographic restriction bypassing via faking
430 X-Forwarded-For HTTP header
431 geo_bypass_ip_block:
432 IP range in CIDR notation that will be used similarly to
433 geo_bypass_country
434 external_downloader: A dictionary of protocol keys and the executable of the
435 external downloader to use for it. The allowed protocols
436 are default|http|ftp|m3u8|dash|rtsp|rtmp|mms.
437 Set the value to 'native' to use the native downloader
438 compat_opts: Compatibility options. See "Differences in default behavior".
439 The following options do not work when used through the API:
440 filename, abort-on-error, multistreams, no-live-chat, format-sort
441 no-clean-infojson, no-playlist-metafiles, no-keep-subs, no-attach-info-json.
442 Refer __init__.py for their implementation
443 progress_template: Dictionary of templates for progress outputs.
444 Allowed keys are 'download', 'postprocess',
445 'download-title' (console title) and 'postprocess-title'.
446 The template is mapped on a dictionary with keys 'progress' and 'info'
447 retry_sleep_functions: Dictionary of functions that takes the number of attempts
448 as argument and returns the time to sleep in seconds.
449 Allowed keys are 'http', 'fragment', 'file_access'
450 download_ranges: A callback function that gets called for every video with
451 the signature (info_dict, ydl) -> Iterable[Section].
452 Only the returned sections will be downloaded.
453 Each Section is a dict with the following keys:
454 * start_time: Start time of the section in seconds
455 * end_time: End time of the section in seconds
456 * title: Section title (Optional)
457 * index: Section number (Optional)
458 force_keyframes_at_cuts: Re-encode the video when downloading ranges to get precise cuts
459 noprogress: Do not print the progress bar
460 live_from_start: Whether to download livestreams videos from the start
461
462 The following parameters are not used by YoutubeDL itself, they are used by
463 the downloader (see yt_dlp/downloader/common.py):
464 nopart, updatetime, buffersize, ratelimit, throttledratelimit, min_filesize,
465 max_filesize, test, noresizebuffer, retries, file_access_retries, fragment_retries,
466 continuedl, xattr_set_filesize, hls_use_mpegts, http_chunk_size,
467 external_downloader_args, concurrent_fragment_downloads.
468
469 The following options are used by the post processors:
470 ffmpeg_location: Location of the ffmpeg/avconv binary; either the path
471 to the binary or its containing directory.
472 postprocessor_args: A dictionary of postprocessor/executable keys (in lower case)
473 and a list of additional command-line arguments for the
474 postprocessor/executable. The dict can also have "PP+EXE" keys
475 which are used when the given exe is used by the given PP.
476 Use 'default' as the name for arguments to passed to all PP
477 For compatibility with youtube-dl, a single list of args
478 can also be used
479
480 The following options are used by the extractors:
481 extractor_retries: Number of times to retry for known errors (default: 3)
482 dynamic_mpd: Whether to process dynamic DASH manifests (default: True)
483 hls_split_discontinuity: Split HLS playlists to different formats at
484 discontinuities such as ad breaks (default: False)
485 extractor_args: A dictionary of arguments to be passed to the extractors.
486 See "EXTRACTOR ARGUMENTS" for details.
487 E.g. {'youtube': {'skip': ['dash', 'hls']}}
488 mark_watched: Mark videos watched (even with --simulate). Only for YouTube
489
490 The following options are deprecated and may be removed in the future:
491
492 break_on_reject: Stop the download process when encountering a video that
493 has been filtered out.
494 - `raise DownloadCancelled(msg)` in match_filter instead
495 force_generic_extractor: Force downloader to use the generic extractor
496 - Use allowed_extractors = ['generic', 'default']
497 playliststart: - Use playlist_items
498 Playlist item to start at.
499 playlistend: - Use playlist_items
500 Playlist item to end at.
501 playlistreverse: - Use playlist_items
502 Download playlist items in reverse order.
503 forceurl: - Use forceprint
504 Force printing final URL.
505 forcetitle: - Use forceprint
506 Force printing title.
507 forceid: - Use forceprint
508 Force printing ID.
509 forcethumbnail: - Use forceprint
510 Force printing thumbnail URL.
511 forcedescription: - Use forceprint
512 Force printing description.
513 forcefilename: - Use forceprint
514 Force printing final filename.
515 forceduration: - Use forceprint
516 Force printing duration.
517 allsubtitles: - Use subtitleslangs = ['all']
518 Downloads all the subtitles of the video
519 (requires writesubtitles or writeautomaticsub)
520 include_ads: - Doesn't work
521 Download ads as well
522 call_home: - Not implemented
523 Boolean, true iff we are allowed to contact the
524 yt-dlp servers for debugging.
525 post_hooks: - Register a custom postprocessor
526 A list of functions that get called as the final step
527 for each video file, after all postprocessors have been
528 called. The filename will be passed as the only argument.
529 hls_prefer_native: - Use external_downloader = {'m3u8': 'native'} or {'m3u8': 'ffmpeg'}.
530 Use the native HLS downloader instead of ffmpeg/avconv
531 if True, otherwise use ffmpeg/avconv if False, otherwise
532 use downloader suggested by extractor if None.
533 prefer_ffmpeg: - avconv support is deprecated
534 If False, use avconv instead of ffmpeg if both are available,
535 otherwise prefer ffmpeg.
536 youtube_include_dash_manifest: - Use extractor_args
537 If True (default), DASH manifests and related
538 data will be downloaded and processed by extractor.
539 You can reduce network I/O by disabling it if you don't
540 care about DASH. (only for youtube)
541 youtube_include_hls_manifest: - Use extractor_args
542 If True (default), HLS manifests and related
543 data will be downloaded and processed by extractor.
544 You can reduce network I/O by disabling it if you don't
545 care about HLS. (only for youtube)
546 no_color: Same as `color='no_color'`
547 """
548
549 _NUMERIC_FIELDS = {
550 'width', 'height', 'asr', 'audio_channels', 'fps',
551 'tbr', 'abr', 'vbr', 'filesize', 'filesize_approx',
552 'timestamp', 'release_timestamp',
553 'duration', 'view_count', 'like_count', 'dislike_count', 'repost_count',
554 'average_rating', 'comment_count', 'age_limit',
555 'start_time', 'end_time',
556 'chapter_number', 'season_number', 'episode_number',
557 'track_number', 'disc_number', 'release_year',
558 }
559
560 _format_fields = {
561 # NB: Keep in sync with the docstring of extractor/common.py
562 'url', 'manifest_url', 'manifest_stream_number', 'ext', 'format', 'format_id', 'format_note',
563 'width', 'height', 'aspect_ratio', 'resolution', 'dynamic_range', 'tbr', 'abr', 'acodec', 'asr', 'audio_channels',
564 'vbr', 'fps', 'vcodec', 'container', 'filesize', 'filesize_approx', 'rows', 'columns',
565 'player_url', 'protocol', 'fragment_base_url', 'fragments', 'is_from_start',
566 'preference', 'language', 'language_preference', 'quality', 'source_preference',
567 'http_headers', 'stretched_ratio', 'no_resume', 'has_drm', 'extra_param_to_segment_url', 'hls_aes', 'downloader_options',
568 'page_url', 'app', 'play_path', 'tc_url', 'flash_version', 'rtmp_live', 'rtmp_conn', 'rtmp_protocol', 'rtmp_real_time'
569 }
570 _format_selection_exts = {
571 'audio': set(MEDIA_EXTENSIONS.common_audio),
572 'video': set(MEDIA_EXTENSIONS.common_video + ('3gp', )),
573 'storyboards': set(MEDIA_EXTENSIONS.storyboards),
574 }
575
576 def __init__(self, params=None, auto_init=True):
577 """Create a FileDownloader object with the given options.
578 @param auto_init Whether to load the default extractors and print header (if verbose).
579 Set to 'no_verbose_header' to not print the header
580 """
581 if params is None:
582 params = {}
583 self.params = params
584 self._ies = {}
585 self._ies_instances = {}
586 self._pps = {k: [] for k in POSTPROCESS_WHEN}
587 self._printed_messages = set()
588 self._first_webpage_request = True
589 self._post_hooks = []
590 self._progress_hooks = []
591 self._postprocessor_hooks = []
592 self._download_retcode = 0
593 self._num_downloads = 0
594 self._num_videos = 0
595 self._playlist_level = 0
596 self._playlist_urls = set()
597 self.cache = Cache(self)
598
599 stdout = sys.stderr if self.params.get('logtostderr') else sys.stdout
600 self._out_files = Namespace(
601 out=stdout,
602 error=sys.stderr,
603 screen=sys.stderr if self.params.get('quiet') else stdout,
604 console=None if compat_os_name == 'nt' else next(
605 filter(supports_terminal_sequences, (sys.stderr, sys.stdout)), None)
606 )
607
608 try:
609 windows_enable_vt_mode()
610 except Exception as e:
611 self.write_debug(f'Failed to enable VT mode: {e}')
612
613 if self.params.get('no_color'):
614 if self.params.get('color') is not None:
615 self.report_warning('Overwriting params from "color" with "no_color"')
616 self.params['color'] = 'no_color'
617
618 term_allow_color = os.environ.get('TERM', '').lower() != 'dumb'
619
620 def process_color_policy(stream):
621 stream_name = {sys.stdout: 'stdout', sys.stderr: 'stderr'}[stream]
622 policy = traverse_obj(self.params, ('color', (stream_name, None), {str}), get_all=False)
623 if policy in ('auto', None):
624 return term_allow_color and supports_terminal_sequences(stream)
625 assert policy in ('always', 'never', 'no_color')
626 return {'always': True, 'never': False}.get(policy, policy)
627
628 self._allow_colors = Namespace(**{
629 name: process_color_policy(stream)
630 for name, stream in self._out_files.items_ if name != 'console'
631 })
632
633 # The code is left like this to be reused for future deprecations
634 MIN_SUPPORTED, MIN_RECOMMENDED = (3, 7), (3, 7)
635 current_version = sys.version_info[:2]
636 if current_version < MIN_RECOMMENDED:
637 msg = ('Support for Python version %d.%d has been deprecated. '
638 'See https://github.com/yt-dlp/yt-dlp/issues/3764 for more details.'
639 '\n You will no longer receive updates on this version')
640 if current_version < MIN_SUPPORTED:
641 msg = 'Python version %d.%d is no longer supported'
642 self.deprecated_feature(
643 f'{msg}! Please update to Python %d.%d or above' % (*current_version, *MIN_RECOMMENDED))
644
645 if self.params.get('allow_unplayable_formats'):
646 self.report_warning(
647 f'You have asked for {self._format_err("UNPLAYABLE", self.Styles.EMPHASIS)} formats to be listed/downloaded. '
648 'This is a developer option intended for debugging. \n'
649 ' If you experience any issues while using this option, '
650 f'{self._format_err("DO NOT", self.Styles.ERROR)} open a bug report')
651
652 if self.params.get('bidi_workaround', False):
653 try:
654 import pty
655 master, slave = pty.openpty()
656 width = shutil.get_terminal_size().columns
657 width_args = [] if width is None else ['-w', str(width)]
658 sp_kwargs = {'stdin': subprocess.PIPE, 'stdout': slave, 'stderr': self._out_files.error}
659 try:
660 self._output_process = Popen(['bidiv'] + width_args, **sp_kwargs)
661 except OSError:
662 self._output_process = Popen(['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
663 self._output_channel = os.fdopen(master, 'rb')
664 except OSError as ose:
665 if ose.errno == errno.ENOENT:
666 self.report_warning(
667 'Could not find fribidi executable, ignoring --bidi-workaround. '
668 'Make sure that fribidi is an executable file in one of the directories in your $PATH.')
669 else:
670 raise
671
672 self.params['compat_opts'] = set(self.params.get('compat_opts', ()))
673 if auto_init and auto_init != 'no_verbose_header':
674 self.print_debug_header()
675
676 def check_deprecated(param, option, suggestion):
677 if self.params.get(param) is not None:
678 self.report_warning(f'{option} is deprecated. Use {suggestion} instead')
679 return True
680 return False
681
682 if check_deprecated('cn_verification_proxy', '--cn-verification-proxy', '--geo-verification-proxy'):
683 if self.params.get('geo_verification_proxy') is None:
684 self.params['geo_verification_proxy'] = self.params['cn_verification_proxy']
685
686 check_deprecated('autonumber', '--auto-number', '-o "%(autonumber)s-%(title)s.%(ext)s"')
687 check_deprecated('usetitle', '--title', '-o "%(title)s-%(id)s.%(ext)s"')
688 check_deprecated('useid', '--id', '-o "%(id)s.%(ext)s"')
689
690 for msg in self.params.get('_warnings', []):
691 self.report_warning(msg)
692 for msg in self.params.get('_deprecation_warnings', []):
693 self.deprecated_feature(msg)
694
695 if 'list-formats' in self.params['compat_opts']:
696 self.params['listformats_table'] = False
697
698 if 'overwrites' not in self.params and self.params.get('nooverwrites') is not None:
699 # nooverwrites was unnecessarily changed to overwrites
700 # in 0c3d0f51778b153f65c21906031c2e091fcfb641
701 # This ensures compatibility with both keys
702 self.params['overwrites'] = not self.params['nooverwrites']
703 elif self.params.get('overwrites') is None:
704 self.params.pop('overwrites', None)
705 else:
706 self.params['nooverwrites'] = not self.params['overwrites']
707
708 if self.params.get('simulate') is None and any((
709 self.params.get('list_thumbnails'),
710 self.params.get('listformats'),
711 self.params.get('listsubtitles'),
712 )):
713 self.params['simulate'] = 'list_only'
714
715 self.params.setdefault('forceprint', {})
716 self.params.setdefault('print_to_file', {})
717
718 # Compatibility with older syntax
719 if not isinstance(params['forceprint'], dict):
720 self.params['forceprint'] = {'video': params['forceprint']}
721
722 if auto_init:
723 self.add_default_info_extractors()
724
725 if (sys.platform != 'win32'
726 and sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968']
727 and not self.params.get('restrictfilenames', False)):
728 # Unicode filesystem API will throw errors (#1474, #13027)
729 self.report_warning(
730 'Assuming --restrict-filenames since file system encoding '
731 'cannot encode all characters. '
732 'Set the LC_ALL environment variable to fix this.')
733 self.params['restrictfilenames'] = True
734
735 self._parse_outtmpl()
736
737 # Creating format selector here allows us to catch syntax errors before the extraction
738 self.format_selector = (
739 self.params.get('format') if self.params.get('format') in (None, '-')
740 else self.params['format'] if callable(self.params['format'])
741 else self.build_format_selector(self.params['format']))
742
743 # Set http_headers defaults according to std_headers
744 self.params['http_headers'] = merge_headers(std_headers, self.params.get('http_headers', {}))
745
746 hooks = {
747 'post_hooks': self.add_post_hook,
748 'progress_hooks': self.add_progress_hook,
749 'postprocessor_hooks': self.add_postprocessor_hook,
750 }
751 for opt, fn in hooks.items():
752 for ph in self.params.get(opt, []):
753 fn(ph)
754
755 for pp_def_raw in self.params.get('postprocessors', []):
756 pp_def = dict(pp_def_raw)
757 when = pp_def.pop('when', 'post_process')
758 self.add_post_processor(
759 get_postprocessor(pp_def.pop('key'))(self, **pp_def),
760 when=when)
761
762 self._setup_opener()
763
764 def preload_download_archive(fn):
765 """Preload the archive, if any is specified"""
766 archive = set()
767 if fn is None:
768 return archive
769 elif not is_path_like(fn):
770 return fn
771
772 self.write_debug(f'Loading archive file {fn!r}')
773 try:
774 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
775 for line in archive_file:
776 archive.add(line.strip())
777 except OSError as ioe:
778 if ioe.errno != errno.ENOENT:
779 raise
780 return archive
781
782 self.archive = preload_download_archive(self.params.get('download_archive'))
783
784 def warn_if_short_id(self, argv):
785 # short YouTube ID starting with dash?
786 idxs = [
787 i for i, a in enumerate(argv)
788 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
789 if idxs:
790 correct_argv = (
791 ['yt-dlp']
792 + [a for i, a in enumerate(argv) if i not in idxs]
793 + ['--'] + [argv[i] for i in idxs]
794 )
795 self.report_warning(
796 'Long argument string detected. '
797 'Use -- to separate parameters and URLs, like this:\n%s' %
798 args_to_str(correct_argv))
799
800 def add_info_extractor(self, ie):
801 """Add an InfoExtractor object to the end of the list."""
802 ie_key = ie.ie_key()
803 self._ies[ie_key] = ie
804 if not isinstance(ie, type):
805 self._ies_instances[ie_key] = ie
806 ie.set_downloader(self)
807
808 def get_info_extractor(self, ie_key):
809 """
810 Get an instance of an IE with name ie_key, it will try to get one from
811 the _ies list, if there's no instance it will create a new one and add
812 it to the extractor list.
813 """
814 ie = self._ies_instances.get(ie_key)
815 if ie is None:
816 ie = get_info_extractor(ie_key)()
817 self.add_info_extractor(ie)
818 return ie
819
820 def add_default_info_extractors(self):
821 """
822 Add the InfoExtractors returned by gen_extractors to the end of the list
823 """
824 all_ies = {ie.IE_NAME.lower(): ie for ie in gen_extractor_classes()}
825 all_ies['end'] = UnsupportedURLIE()
826 try:
827 ie_names = orderedSet_from_options(
828 self.params.get('allowed_extractors', ['default']), {
829 'all': list(all_ies),
830 'default': [name for name, ie in all_ies.items() if ie._ENABLED],
831 }, use_regex=True)
832 except re.error as e:
833 raise ValueError(f'Wrong regex for allowed_extractors: {e.pattern}')
834 for name in ie_names:
835 self.add_info_extractor(all_ies[name])
836 self.write_debug(f'Loaded {len(ie_names)} extractors')
837
838 def add_post_processor(self, pp, when='post_process'):
839 """Add a PostProcessor object to the end of the chain."""
840 assert when in POSTPROCESS_WHEN, f'Invalid when={when}'
841 self._pps[when].append(pp)
842 pp.set_downloader(self)
843
844 def add_post_hook(self, ph):
845 """Add the post hook"""
846 self._post_hooks.append(ph)
847
848 def add_progress_hook(self, ph):
849 """Add the download progress hook"""
850 self._progress_hooks.append(ph)
851
852 def add_postprocessor_hook(self, ph):
853 """Add the postprocessing progress hook"""
854 self._postprocessor_hooks.append(ph)
855 for pps in self._pps.values():
856 for pp in pps:
857 pp.add_progress_hook(ph)
858
859 def _bidi_workaround(self, message):
860 if not hasattr(self, '_output_channel'):
861 return message
862
863 assert hasattr(self, '_output_process')
864 assert isinstance(message, str)
865 line_count = message.count('\n') + 1
866 self._output_process.stdin.write((message + '\n').encode())
867 self._output_process.stdin.flush()
868 res = ''.join(self._output_channel.readline().decode()
869 for _ in range(line_count))
870 return res[:-len('\n')]
871
872 def _write_string(self, message, out=None, only_once=False):
873 if only_once:
874 if message in self._printed_messages:
875 return
876 self._printed_messages.add(message)
877 write_string(message, out=out, encoding=self.params.get('encoding'))
878
879 def to_stdout(self, message, skip_eol=False, quiet=None):
880 """Print message to stdout"""
881 if quiet is not None:
882 self.deprecation_warning('"YoutubeDL.to_stdout" no longer accepts the argument quiet. '
883 'Use "YoutubeDL.to_screen" instead')
884 if skip_eol is not False:
885 self.deprecation_warning('"YoutubeDL.to_stdout" no longer accepts the argument skip_eol. '
886 'Use "YoutubeDL.to_screen" instead')
887 self._write_string(f'{self._bidi_workaround(message)}\n', self._out_files.out)
888
889 def to_screen(self, message, skip_eol=False, quiet=None, only_once=False):
890 """Print message to screen if not in quiet mode"""
891 if self.params.get('logger'):
892 self.params['logger'].debug(message)
893 return
894 if (self.params.get('quiet') if quiet is None else quiet) and not self.params.get('verbose'):
895 return
896 self._write_string(
897 '%s%s' % (self._bidi_workaround(message), ('' if skip_eol else '\n')),
898 self._out_files.screen, only_once=only_once)
899
900 def to_stderr(self, message, only_once=False):
901 """Print message to stderr"""
902 assert isinstance(message, str)
903 if self.params.get('logger'):
904 self.params['logger'].error(message)
905 else:
906 self._write_string(f'{self._bidi_workaround(message)}\n', self._out_files.error, only_once=only_once)
907
908 def _send_console_code(self, code):
909 if compat_os_name == 'nt' or not self._out_files.console:
910 return
911 self._write_string(code, self._out_files.console)
912
913 def to_console_title(self, message):
914 if not self.params.get('consoletitle', False):
915 return
916 message = remove_terminal_sequences(message)
917 if compat_os_name == 'nt':
918 if ctypes.windll.kernel32.GetConsoleWindow():
919 # c_wchar_p() might not be necessary if `message` is
920 # already of type unicode()
921 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
922 else:
923 self._send_console_code(f'\033]0;{message}\007')
924
925 def save_console_title(self):
926 if not self.params.get('consoletitle') or self.params.get('simulate'):
927 return
928 self._send_console_code('\033[22;0t') # Save the title on stack
929
930 def restore_console_title(self):
931 if not self.params.get('consoletitle') or self.params.get('simulate'):
932 return
933 self._send_console_code('\033[23;0t') # Restore the title from stack
934
935 def __enter__(self):
936 self.save_console_title()
937 return self
938
939 def __exit__(self, *args):
940 self.restore_console_title()
941
942 if self.params.get('cookiefile') is not None:
943 self.cookiejar.save(ignore_discard=True, ignore_expires=True)
944
945 def trouble(self, message=None, tb=None, is_error=True):
946 """Determine action to take when a download problem appears.
947
948 Depending on if the downloader has been configured to ignore
949 download errors or not, this method may throw an exception or
950 not when errors are found, after printing the message.
951
952 @param tb If given, is additional traceback information
953 @param is_error Whether to raise error according to ignorerrors
954 """
955 if message is not None:
956 self.to_stderr(message)
957 if self.params.get('verbose'):
958 if tb is None:
959 if sys.exc_info()[0]: # if .trouble has been called from an except block
960 tb = ''
961 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
962 tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
963 tb += encode_compat_str(traceback.format_exc())
964 else:
965 tb_data = traceback.format_list(traceback.extract_stack())
966 tb = ''.join(tb_data)
967 if tb:
968 self.to_stderr(tb)
969 if not is_error:
970 return
971 if not self.params.get('ignoreerrors'):
972 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
973 exc_info = sys.exc_info()[1].exc_info
974 else:
975 exc_info = sys.exc_info()
976 raise DownloadError(message, exc_info)
977 self._download_retcode = 1
978
979 Styles = Namespace(
980 HEADERS='yellow',
981 EMPHASIS='light blue',
982 FILENAME='green',
983 ID='green',
984 DELIM='blue',
985 ERROR='red',
986 WARNING='yellow',
987 SUPPRESS='light black',
988 )
989
990 def _format_text(self, handle, allow_colors, text, f, fallback=None, *, test_encoding=False):
991 text = str(text)
992 if test_encoding:
993 original_text = text
994 # handle.encoding can be None. See https://github.com/yt-dlp/yt-dlp/issues/2711
995 encoding = self.params.get('encoding') or getattr(handle, 'encoding', None) or 'ascii'
996 text = text.encode(encoding, 'ignore').decode(encoding)
997 if fallback is not None and text != original_text:
998 text = fallback
999 return format_text(text, f) if allow_colors is True else text if fallback is None else fallback
1000
1001 def _format_out(self, *args, **kwargs):
1002 return self._format_text(self._out_files.out, self._allow_colors.out, *args, **kwargs)
1003
1004 def _format_screen(self, *args, **kwargs):
1005 return self._format_text(self._out_files.screen, self._allow_colors.screen, *args, **kwargs)
1006
1007 def _format_err(self, *args, **kwargs):
1008 return self._format_text(self._out_files.error, self._allow_colors.error, *args, **kwargs)
1009
1010 def report_warning(self, message, only_once=False):
1011 '''
1012 Print the message to stderr, it will be prefixed with 'WARNING:'
1013 If stderr is a tty file the 'WARNING:' will be colored
1014 '''
1015 if self.params.get('logger') is not None:
1016 self.params['logger'].warning(message)
1017 else:
1018 if self.params.get('no_warnings'):
1019 return
1020 self.to_stderr(f'{self._format_err("WARNING:", self.Styles.WARNING)} {message}', only_once)
1021
1022 def deprecation_warning(self, message, *, stacklevel=0):
1023 deprecation_warning(
1024 message, stacklevel=stacklevel + 1, printer=self.report_error, is_error=False)
1025
1026 def deprecated_feature(self, message):
1027 if self.params.get('logger') is not None:
1028 self.params['logger'].warning(f'Deprecated Feature: {message}')
1029 self.to_stderr(f'{self._format_err("Deprecated Feature:", self.Styles.ERROR)} {message}', True)
1030
1031 def report_error(self, message, *args, **kwargs):
1032 '''
1033 Do the same as trouble, but prefixes the message with 'ERROR:', colored
1034 in red if stderr is a tty file.
1035 '''
1036 self.trouble(f'{self._format_err("ERROR:", self.Styles.ERROR)} {message}', *args, **kwargs)
1037
1038 def write_debug(self, message, only_once=False):
1039 '''Log debug message or Print message to stderr'''
1040 if not self.params.get('verbose', False):
1041 return
1042 message = f'[debug] {message}'
1043 if self.params.get('logger'):
1044 self.params['logger'].debug(message)
1045 else:
1046 self.to_stderr(message, only_once)
1047
1048 def report_file_already_downloaded(self, file_name):
1049 """Report file has already been fully downloaded."""
1050 try:
1051 self.to_screen('[download] %s has already been downloaded' % file_name)
1052 except UnicodeEncodeError:
1053 self.to_screen('[download] The file has already been downloaded')
1054
1055 def report_file_delete(self, file_name):
1056 """Report that existing file will be deleted."""
1057 try:
1058 self.to_screen('Deleting existing file %s' % file_name)
1059 except UnicodeEncodeError:
1060 self.to_screen('Deleting existing file')
1061
1062 def raise_no_formats(self, info, forced=False, *, msg=None):
1063 has_drm = info.get('_has_drm')
1064 ignored, expected = self.params.get('ignore_no_formats_error'), bool(msg)
1065 msg = msg or has_drm and 'This video is DRM protected' or 'No video formats found!'
1066 if forced or not ignored:
1067 raise ExtractorError(msg, video_id=info['id'], ie=info['extractor'],
1068 expected=has_drm or ignored or expected)
1069 else:
1070 self.report_warning(msg)
1071
1072 def parse_outtmpl(self):
1073 self.deprecation_warning('"YoutubeDL.parse_outtmpl" is deprecated and may be removed in a future version')
1074 self._parse_outtmpl()
1075 return self.params['outtmpl']
1076
1077 def _parse_outtmpl(self):
1078 sanitize = IDENTITY
1079 if self.params.get('restrictfilenames'): # Remove spaces in the default template
1080 sanitize = lambda x: x.replace(' - ', ' ').replace(' ', '-')
1081
1082 outtmpl = self.params.setdefault('outtmpl', {})
1083 if not isinstance(outtmpl, dict):
1084 self.params['outtmpl'] = outtmpl = {'default': outtmpl}
1085 outtmpl.update({k: sanitize(v) for k, v in DEFAULT_OUTTMPL.items() if outtmpl.get(k) is None})
1086
1087 def get_output_path(self, dir_type='', filename=None):
1088 paths = self.params.get('paths', {})
1089 assert isinstance(paths, dict), '"paths" parameter must be a dictionary'
1090 path = os.path.join(
1091 expand_path(paths.get('home', '').strip()),
1092 expand_path(paths.get(dir_type, '').strip()) if dir_type else '',
1093 filename or '')
1094 return sanitize_path(path, force=self.params.get('windowsfilenames'))
1095
1096 @staticmethod
1097 def _outtmpl_expandpath(outtmpl):
1098 # expand_path translates '%%' into '%' and '$$' into '$'
1099 # correspondingly that is not what we want since we need to keep
1100 # '%%' intact for template dict substitution step. Working around
1101 # with boundary-alike separator hack.
1102 sep = ''.join(random.choices(string.ascii_letters, k=32))
1103 outtmpl = outtmpl.replace('%%', f'%{sep}%').replace('$$', f'${sep}$')
1104
1105 # outtmpl should be expand_path'ed before template dict substitution
1106 # because meta fields may contain env variables we don't want to
1107 # be expanded. E.g. for outtmpl "%(title)s.%(ext)s" and
1108 # title "Hello $PATH", we don't want `$PATH` to be expanded.
1109 return expand_path(outtmpl).replace(sep, '')
1110
1111 @staticmethod
1112 def escape_outtmpl(outtmpl):
1113 ''' Escape any remaining strings like %s, %abc% etc. '''
1114 return re.sub(
1115 STR_FORMAT_RE_TMPL.format('', '(?![%(\0])'),
1116 lambda mobj: ('' if mobj.group('has_key') else '%') + mobj.group(0),
1117 outtmpl)
1118
1119 @classmethod
1120 def validate_outtmpl(cls, outtmpl):
1121 ''' @return None or Exception object '''
1122 outtmpl = re.sub(
1123 STR_FORMAT_RE_TMPL.format('[^)]*', '[ljhqBUDS]'),
1124 lambda mobj: f'{mobj.group(0)[:-1]}s',
1125 cls._outtmpl_expandpath(outtmpl))
1126 try:
1127 cls.escape_outtmpl(outtmpl) % collections.defaultdict(int)
1128 return None
1129 except ValueError as err:
1130 return err
1131
1132 @staticmethod
1133 def _copy_infodict(info_dict):
1134 info_dict = dict(info_dict)
1135 info_dict.pop('__postprocessors', None)
1136 info_dict.pop('__pending_error', None)
1137 return info_dict
1138
1139 def prepare_outtmpl(self, outtmpl, info_dict, sanitize=False):
1140 """ Make the outtmpl and info_dict suitable for substitution: ydl.escape_outtmpl(outtmpl) % info_dict
1141 @param sanitize Whether to sanitize the output as a filename.
1142 For backward compatibility, a function can also be passed
1143 """
1144
1145 info_dict.setdefault('epoch', int(time.time())) # keep epoch consistent once set
1146
1147 info_dict = self._copy_infodict(info_dict)
1148 info_dict['duration_string'] = ( # %(duration>%H-%M-%S)s is wrong if duration > 24hrs
1149 formatSeconds(info_dict['duration'], '-' if sanitize else ':')
1150 if info_dict.get('duration', None) is not None
1151 else None)
1152 info_dict['autonumber'] = int(self.params.get('autonumber_start', 1) - 1 + self._num_downloads)
1153 info_dict['video_autonumber'] = self._num_videos
1154 if info_dict.get('resolution') is None:
1155 info_dict['resolution'] = self.format_resolution(info_dict, default=None)
1156
1157 # For fields playlist_index, playlist_autonumber and autonumber convert all occurrences
1158 # of %(field)s to %(field)0Nd for backward compatibility
1159 field_size_compat_map = {
1160 'playlist_index': number_of_digits(info_dict.get('__last_playlist_index') or 0),
1161 'playlist_autonumber': number_of_digits(info_dict.get('n_entries') or 0),
1162 'autonumber': self.params.get('autonumber_size') or 5,
1163 }
1164
1165 TMPL_DICT = {}
1166 EXTERNAL_FORMAT_RE = re.compile(STR_FORMAT_RE_TMPL.format('[^)]*', f'[{STR_FORMAT_TYPES}ljhqBUDS]'))
1167 MATH_FUNCTIONS = {
1168 '+': float.__add__,
1169 '-': float.__sub__,
1170 }
1171 # Field is of the form key1.key2...
1172 # where keys (except first) can be string, int, slice or "{field, ...}"
1173 FIELD_INNER_RE = r'(?:\w+|%(num)s|%(num)s?(?::%(num)s?){1,2})' % {'num': r'(?:-?\d+)'}
1174 FIELD_RE = r'\w*(?:\.(?:%(inner)s|{%(field)s(?:,%(field)s)*}))*' % {
1175 'inner': FIELD_INNER_RE,
1176 'field': rf'\w*(?:\.{FIELD_INNER_RE})*'
1177 }
1178 MATH_FIELD_RE = rf'(?:{FIELD_RE}|-?{NUMBER_RE})'
1179 MATH_OPERATORS_RE = r'(?:%s)' % '|'.join(map(re.escape, MATH_FUNCTIONS.keys()))
1180 INTERNAL_FORMAT_RE = re.compile(rf'''(?xs)
1181 (?P<negate>-)?
1182 (?P<fields>{FIELD_RE})
1183 (?P<maths>(?:{MATH_OPERATORS_RE}{MATH_FIELD_RE})*)
1184 (?:>(?P<strf_format>.+?))?
1185 (?P<remaining>
1186 (?P<alternate>(?<!\\),[^|&)]+)?
1187 (?:&(?P<replacement>.*?))?
1188 (?:\|(?P<default>.*?))?
1189 )$''')
1190
1191 def _traverse_infodict(fields):
1192 fields = [f for x in re.split(r'\.({.+?})\.?', fields)
1193 for f in ([x] if x.startswith('{') else x.split('.'))]
1194 for i in (0, -1):
1195 if fields and not fields[i]:
1196 fields.pop(i)
1197
1198 for i, f in enumerate(fields):
1199 if not f.startswith('{'):
1200 continue
1201 assert f.endswith('}'), f'No closing brace for {f} in {fields}'
1202 fields[i] = {k: k.split('.') for k in f[1:-1].split(',')}
1203
1204 return traverse_obj(info_dict, fields, is_user_input=True, traverse_string=True)
1205
1206 def get_value(mdict):
1207 # Object traversal
1208 value = _traverse_infodict(mdict['fields'])
1209 # Negative
1210 if mdict['negate']:
1211 value = float_or_none(value)
1212 if value is not None:
1213 value *= -1
1214 # Do maths
1215 offset_key = mdict['maths']
1216 if offset_key:
1217 value = float_or_none(value)
1218 operator = None
1219 while offset_key:
1220 item = re.match(
1221 MATH_FIELD_RE if operator else MATH_OPERATORS_RE,
1222 offset_key).group(0)
1223 offset_key = offset_key[len(item):]
1224 if operator is None:
1225 operator = MATH_FUNCTIONS[item]
1226 continue
1227 item, multiplier = (item[1:], -1) if item[0] == '-' else (item, 1)
1228 offset = float_or_none(item)
1229 if offset is None:
1230 offset = float_or_none(_traverse_infodict(item))
1231 try:
1232 value = operator(value, multiplier * offset)
1233 except (TypeError, ZeroDivisionError):
1234 return None
1235 operator = None
1236 # Datetime formatting
1237 if mdict['strf_format']:
1238 value = strftime_or_none(value, mdict['strf_format'].replace('\\,', ','))
1239
1240 # XXX: Workaround for https://github.com/yt-dlp/yt-dlp/issues/4485
1241 if sanitize and value == '':
1242 value = None
1243 return value
1244
1245 na = self.params.get('outtmpl_na_placeholder', 'NA')
1246
1247 def filename_sanitizer(key, value, restricted=self.params.get('restrictfilenames')):
1248 return sanitize_filename(str(value), restricted=restricted, is_id=(
1249 bool(re.search(r'(^|[_.])id(\.|$)', key))
1250 if 'filename-sanitization' in self.params['compat_opts']
1251 else NO_DEFAULT))
1252
1253 sanitizer = sanitize if callable(sanitize) else filename_sanitizer
1254 sanitize = bool(sanitize)
1255
1256 def _dumpjson_default(obj):
1257 if isinstance(obj, (set, LazyList)):
1258 return list(obj)
1259 return repr(obj)
1260
1261 class _ReplacementFormatter(string.Formatter):
1262 def get_field(self, field_name, args, kwargs):
1263 if field_name.isdigit():
1264 return args[0], -1
1265 raise ValueError('Unsupported field')
1266
1267 replacement_formatter = _ReplacementFormatter()
1268
1269 def create_key(outer_mobj):
1270 if not outer_mobj.group('has_key'):
1271 return outer_mobj.group(0)
1272 key = outer_mobj.group('key')
1273 mobj = re.match(INTERNAL_FORMAT_RE, key)
1274 initial_field = mobj.group('fields') if mobj else ''
1275 value, replacement, default = None, None, na
1276 while mobj:
1277 mobj = mobj.groupdict()
1278 default = mobj['default'] if mobj['default'] is not None else default
1279 value = get_value(mobj)
1280 replacement = mobj['replacement']
1281 if value is None and mobj['alternate']:
1282 mobj = re.match(INTERNAL_FORMAT_RE, mobj['remaining'][1:])
1283 else:
1284 break
1285
1286 fmt = outer_mobj.group('format')
1287 if fmt == 's' and value is not None and key in field_size_compat_map.keys():
1288 fmt = f'0{field_size_compat_map[key]:d}d'
1289
1290 if None not in (value, replacement):
1291 try:
1292 value = replacement_formatter.format(replacement, value)
1293 except ValueError:
1294 value, default = None, na
1295
1296 flags = outer_mobj.group('conversion') or ''
1297 str_fmt = f'{fmt[:-1]}s'
1298 if value is None:
1299 value, fmt = default, 's'
1300 elif fmt[-1] == 'l': # list
1301 delim = '\n' if '#' in flags else ', '
1302 value, fmt = delim.join(map(str, variadic(value, allowed_types=(str, bytes)))), str_fmt
1303 elif fmt[-1] == 'j': # json
1304 value, fmt = json.dumps(
1305 value, default=_dumpjson_default,
1306 indent=4 if '#' in flags else None, ensure_ascii='+' not in flags), str_fmt
1307 elif fmt[-1] == 'h': # html
1308 value, fmt = escapeHTML(str(value)), str_fmt
1309 elif fmt[-1] == 'q': # quoted
1310 value = map(str, variadic(value) if '#' in flags else [value])
1311 value, fmt = ' '.join(map(compat_shlex_quote, value)), str_fmt
1312 elif fmt[-1] == 'B': # bytes
1313 value = f'%{str_fmt}'.encode() % str(value).encode()
1314 value, fmt = value.decode('utf-8', 'ignore'), 's'
1315 elif fmt[-1] == 'U': # unicode normalized
1316 value, fmt = unicodedata.normalize(
1317 # "+" = compatibility equivalence, "#" = NFD
1318 'NF%s%s' % ('K' if '+' in flags else '', 'D' if '#' in flags else 'C'),
1319 value), str_fmt
1320 elif fmt[-1] == 'D': # decimal suffix
1321 num_fmt, fmt = fmt[:-1].replace('#', ''), 's'
1322 value = format_decimal_suffix(value, f'%{num_fmt}f%s' if num_fmt else '%d%s',
1323 factor=1024 if '#' in flags else 1000)
1324 elif fmt[-1] == 'S': # filename sanitization
1325 value, fmt = filename_sanitizer(initial_field, value, restricted='#' in flags), str_fmt
1326 elif fmt[-1] == 'c':
1327 if value:
1328 value = str(value)[0]
1329 else:
1330 fmt = str_fmt
1331 elif fmt[-1] not in 'rs': # numeric
1332 value = float_or_none(value)
1333 if value is None:
1334 value, fmt = default, 's'
1335
1336 if sanitize:
1337 if fmt[-1] == 'r':
1338 # If value is an object, sanitize might convert it to a string
1339 # So we convert it to repr first
1340 value, fmt = repr(value), str_fmt
1341 if fmt[-1] in 'csr':
1342 value = sanitizer(initial_field, value)
1343
1344 key = '%s\0%s' % (key.replace('%', '%\0'), outer_mobj.group('format'))
1345 TMPL_DICT[key] = value
1346 return '{prefix}%({key}){fmt}'.format(key=key, fmt=fmt, prefix=outer_mobj.group('prefix'))
1347
1348 return EXTERNAL_FORMAT_RE.sub(create_key, outtmpl), TMPL_DICT
1349
1350 def evaluate_outtmpl(self, outtmpl, info_dict, *args, **kwargs):
1351 outtmpl, info_dict = self.prepare_outtmpl(outtmpl, info_dict, *args, **kwargs)
1352 return self.escape_outtmpl(outtmpl) % info_dict
1353
1354 def _prepare_filename(self, info_dict, *, outtmpl=None, tmpl_type=None):
1355 assert None in (outtmpl, tmpl_type), 'outtmpl and tmpl_type are mutually exclusive'
1356 if outtmpl is None:
1357 outtmpl = self.params['outtmpl'].get(tmpl_type or 'default', self.params['outtmpl']['default'])
1358 try:
1359 outtmpl = self._outtmpl_expandpath(outtmpl)
1360 filename = self.evaluate_outtmpl(outtmpl, info_dict, True)
1361 if not filename:
1362 return None
1363
1364 if tmpl_type in ('', 'temp'):
1365 final_ext, ext = self.params.get('final_ext'), info_dict.get('ext')
1366 if final_ext and ext and final_ext != ext and filename.endswith(f'.{final_ext}'):
1367 filename = replace_extension(filename, ext, final_ext)
1368 elif tmpl_type:
1369 force_ext = OUTTMPL_TYPES[tmpl_type]
1370 if force_ext:
1371 filename = replace_extension(filename, force_ext, info_dict.get('ext'))
1372
1373 # https://github.com/blackjack4494/youtube-dlc/issues/85
1374 trim_file_name = self.params.get('trim_file_name', False)
1375 if trim_file_name:
1376 no_ext, *ext = filename.rsplit('.', 2)
1377 filename = join_nonempty(no_ext[:trim_file_name], *ext, delim='.')
1378
1379 return filename
1380 except ValueError as err:
1381 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
1382 return None
1383
1384 def prepare_filename(self, info_dict, dir_type='', *, outtmpl=None, warn=False):
1385 """Generate the output filename"""
1386 if outtmpl:
1387 assert not dir_type, 'outtmpl and dir_type are mutually exclusive'
1388 dir_type = None
1389 filename = self._prepare_filename(info_dict, tmpl_type=dir_type, outtmpl=outtmpl)
1390 if not filename and dir_type not in ('', 'temp'):
1391 return ''
1392
1393 if warn:
1394 if not self.params.get('paths'):
1395 pass
1396 elif filename == '-':
1397 self.report_warning('--paths is ignored when an outputting to stdout', only_once=True)
1398 elif os.path.isabs(filename):
1399 self.report_warning('--paths is ignored since an absolute path is given in output template', only_once=True)
1400 if filename == '-' or not filename:
1401 return filename
1402
1403 return self.get_output_path(dir_type, filename)
1404
1405 def _match_entry(self, info_dict, incomplete=False, silent=False):
1406 """Returns None if the file should be downloaded"""
1407 _type = 'video' if 'playlist-match-filter' in self.params['compat_opts'] else info_dict.get('_type', 'video')
1408 assert incomplete or _type == 'video', 'Only video result can be considered complete'
1409
1410 video_title = info_dict.get('title', info_dict.get('id', 'entry'))
1411
1412 def check_filter():
1413 if _type in ('playlist', 'multi_video'):
1414 return
1415 elif _type in ('url', 'url_transparent') and not try_call(
1416 lambda: self.get_info_extractor(info_dict['ie_key']).is_single_video(info_dict['url'])):
1417 return
1418
1419 if 'title' in info_dict:
1420 # This can happen when we're just evaluating the playlist
1421 title = info_dict['title']
1422 matchtitle = self.params.get('matchtitle', False)
1423 if matchtitle:
1424 if not re.search(matchtitle, title, re.IGNORECASE):
1425 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
1426 rejecttitle = self.params.get('rejecttitle', False)
1427 if rejecttitle:
1428 if re.search(rejecttitle, title, re.IGNORECASE):
1429 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
1430
1431 date = info_dict.get('upload_date')
1432 if date is not None:
1433 dateRange = self.params.get('daterange', DateRange())
1434 if date not in dateRange:
1435 return f'{date_from_str(date).isoformat()} upload date is not in range {dateRange}'
1436 view_count = info_dict.get('view_count')
1437 if view_count is not None:
1438 min_views = self.params.get('min_views')
1439 if min_views is not None and view_count < min_views:
1440 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
1441 max_views = self.params.get('max_views')
1442 if max_views is not None and view_count > max_views:
1443 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
1444 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
1445 return 'Skipping "%s" because it is age restricted' % video_title
1446
1447 match_filter = self.params.get('match_filter')
1448 if match_filter is None:
1449 return None
1450
1451 cancelled = None
1452 try:
1453 try:
1454 ret = match_filter(info_dict, incomplete=incomplete)
1455 except TypeError:
1456 # For backward compatibility
1457 ret = None if incomplete else match_filter(info_dict)
1458 except DownloadCancelled as err:
1459 if err.msg is not NO_DEFAULT:
1460 raise
1461 ret, cancelled = err.msg, err
1462
1463 if ret is NO_DEFAULT:
1464 while True:
1465 filename = self._format_screen(self.prepare_filename(info_dict), self.Styles.FILENAME)
1466 reply = input(self._format_screen(
1467 f'Download "{filename}"? (Y/n): ', self.Styles.EMPHASIS)).lower().strip()
1468 if reply in {'y', ''}:
1469 return None
1470 elif reply == 'n':
1471 if cancelled:
1472 raise type(cancelled)(f'Skipping {video_title}')
1473 return f'Skipping {video_title}'
1474 return ret
1475
1476 if self.in_download_archive(info_dict):
1477 reason = '%s has already been recorded in the archive' % video_title
1478 break_opt, break_err = 'break_on_existing', ExistingVideoReached
1479 else:
1480 try:
1481 reason = check_filter()
1482 except DownloadCancelled as e:
1483 reason, break_opt, break_err = e.msg, 'match_filter', type(e)
1484 else:
1485 break_opt, break_err = 'break_on_reject', RejectedVideoReached
1486 if reason is not None:
1487 if not silent:
1488 self.to_screen('[download] ' + reason)
1489 if self.params.get(break_opt, False):
1490 raise break_err()
1491 return reason
1492
1493 @staticmethod
1494 def add_extra_info(info_dict, extra_info):
1495 '''Set the keys from extra_info in info dict if they are missing'''
1496 for key, value in extra_info.items():
1497 info_dict.setdefault(key, value)
1498
1499 def extract_info(self, url, download=True, ie_key=None, extra_info=None,
1500 process=True, force_generic_extractor=False):
1501 """
1502 Extract and return the information dictionary of the URL
1503
1504 Arguments:
1505 @param url URL to extract
1506
1507 Keyword arguments:
1508 @param download Whether to download videos
1509 @param process Whether to resolve all unresolved references (URLs, playlist items).
1510 Must be True for download to work
1511 @param ie_key Use only the extractor with this key
1512
1513 @param extra_info Dictionary containing the extra values to add to the info (For internal use only)
1514 @force_generic_extractor Force using the generic extractor (Deprecated; use ie_key='Generic')
1515 """
1516
1517 if extra_info is None:
1518 extra_info = {}
1519
1520 if not ie_key and force_generic_extractor:
1521 ie_key = 'Generic'
1522
1523 if ie_key:
1524 ies = {ie_key: self._ies[ie_key]} if ie_key in self._ies else {}
1525 else:
1526 ies = self._ies
1527
1528 for key, ie in ies.items():
1529 if not ie.suitable(url):
1530 continue
1531
1532 if not ie.working():
1533 self.report_warning('The program functionality for this site has been marked as broken, '
1534 'and will probably not work.')
1535
1536 temp_id = ie.get_temp_id(url)
1537 if temp_id is not None and self.in_download_archive({'id': temp_id, 'ie_key': key}):
1538 self.to_screen(f'[{key}] {temp_id}: has already been recorded in the archive')
1539 if self.params.get('break_on_existing', False):
1540 raise ExistingVideoReached()
1541 break
1542 return self.__extract_info(url, self.get_info_extractor(key), download, extra_info, process)
1543 else:
1544 extractors_restricted = self.params.get('allowed_extractors') not in (None, ['default'])
1545 self.report_error(f'No suitable extractor{format_field(ie_key, None, " (%s)")} found for URL {url}',
1546 tb=False if extractors_restricted else None)
1547
1548 def _handle_extraction_exceptions(func):
1549 @functools.wraps(func)
1550 def wrapper(self, *args, **kwargs):
1551 while True:
1552 try:
1553 return func(self, *args, **kwargs)
1554 except (DownloadCancelled, LazyList.IndexError, PagedList.IndexError):
1555 raise
1556 except ReExtractInfo as e:
1557 if e.expected:
1558 self.to_screen(f'{e}; Re-extracting data')
1559 else:
1560 self.to_stderr('\r')
1561 self.report_warning(f'{e}; Re-extracting data')
1562 continue
1563 except GeoRestrictedError as e:
1564 msg = e.msg
1565 if e.countries:
1566 msg += '\nThis video is available in %s.' % ', '.join(
1567 map(ISO3166Utils.short2full, e.countries))
1568 msg += '\nYou might want to use a VPN or a proxy server (with --proxy) to workaround.'
1569 self.report_error(msg)
1570 except ExtractorError as e: # An error we somewhat expected
1571 self.report_error(str(e), e.format_traceback())
1572 except Exception as e:
1573 if self.params.get('ignoreerrors'):
1574 self.report_error(str(e), tb=encode_compat_str(traceback.format_exc()))
1575 else:
1576 raise
1577 break
1578 return wrapper
1579
1580 def _wait_for_video(self, ie_result={}):
1581 if (not self.params.get('wait_for_video')
1582 or ie_result.get('_type', 'video') != 'video'
1583 or ie_result.get('formats') or ie_result.get('url')):
1584 return
1585
1586 format_dur = lambda dur: '%02d:%02d:%02d' % timetuple_from_msec(dur * 1000)[:-1]
1587 last_msg = ''
1588
1589 def progress(msg):
1590 nonlocal last_msg
1591 full_msg = f'{msg}\n'
1592 if not self.params.get('noprogress'):
1593 full_msg = msg + ' ' * (len(last_msg) - len(msg)) + '\r'
1594 elif last_msg:
1595 return
1596 self.to_screen(full_msg, skip_eol=True)
1597 last_msg = msg
1598
1599 min_wait, max_wait = self.params.get('wait_for_video')
1600 diff = try_get(ie_result, lambda x: x['release_timestamp'] - time.time())
1601 if diff is None and ie_result.get('live_status') == 'is_upcoming':
1602 diff = round(random.uniform(min_wait, max_wait) if (max_wait and min_wait) else (max_wait or min_wait), 0)
1603 self.report_warning('Release time of video is not known')
1604 elif ie_result and (diff or 0) <= 0:
1605 self.report_warning('Video should already be available according to extracted info')
1606 diff = min(max(diff or 0, min_wait or 0), max_wait or float('inf'))
1607 self.to_screen(f'[wait] Waiting for {format_dur(diff)} - Press Ctrl+C to try now')
1608
1609 wait_till = time.time() + diff
1610 try:
1611 while True:
1612 diff = wait_till - time.time()
1613 if diff <= 0:
1614 progress('')
1615 raise ReExtractInfo('[wait] Wait period ended', expected=True)
1616 progress(f'[wait] Remaining time until next attempt: {self._format_screen(format_dur(diff), self.Styles.EMPHASIS)}')
1617 time.sleep(1)
1618 except KeyboardInterrupt:
1619 progress('')
1620 raise ReExtractInfo('[wait] Interrupted by user', expected=True)
1621 except BaseException as e:
1622 if not isinstance(e, ReExtractInfo):
1623 self.to_screen('')
1624 raise
1625
1626 @_handle_extraction_exceptions
1627 def __extract_info(self, url, ie, download, extra_info, process):
1628 try:
1629 ie_result = ie.extract(url)
1630 except UserNotLive as e:
1631 if process:
1632 if self.params.get('wait_for_video'):
1633 self.report_warning(e)
1634 self._wait_for_video()
1635 raise
1636 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
1637 self.report_warning(f'Extractor {ie.IE_NAME} returned nothing{bug_reports_message()}')
1638 return
1639 if isinstance(ie_result, list):
1640 # Backwards compatibility: old IE result format
1641 ie_result = {
1642 '_type': 'compat_list',
1643 'entries': ie_result,
1644 }
1645 if extra_info.get('original_url'):
1646 ie_result.setdefault('original_url', extra_info['original_url'])
1647 self.add_default_extra_info(ie_result, ie, url)
1648 if process:
1649 self._wait_for_video(ie_result)
1650 return self.process_ie_result(ie_result, download, extra_info)
1651 else:
1652 return ie_result
1653
1654 def add_default_extra_info(self, ie_result, ie, url):
1655 if url is not None:
1656 self.add_extra_info(ie_result, {
1657 'webpage_url': url,
1658 'original_url': url,
1659 })
1660 webpage_url = ie_result.get('webpage_url')
1661 if webpage_url:
1662 self.add_extra_info(ie_result, {
1663 'webpage_url_basename': url_basename(webpage_url),
1664 'webpage_url_domain': get_domain(webpage_url),
1665 })
1666 if ie is not None:
1667 self.add_extra_info(ie_result, {
1668 'extractor': ie.IE_NAME,
1669 'extractor_key': ie.ie_key(),
1670 })
1671
1672 def process_ie_result(self, ie_result, download=True, extra_info=None):
1673 """
1674 Take the result of the ie(may be modified) and resolve all unresolved
1675 references (URLs, playlist items).
1676
1677 It will also download the videos if 'download'.
1678 Returns the resolved ie_result.
1679 """
1680 if extra_info is None:
1681 extra_info = {}
1682 result_type = ie_result.get('_type', 'video')
1683
1684 if result_type in ('url', 'url_transparent'):
1685 ie_result['url'] = sanitize_url(
1686 ie_result['url'], scheme='http' if self.params.get('prefer_insecure') else 'https')
1687 if ie_result.get('original_url') and not extra_info.get('original_url'):
1688 extra_info = {'original_url': ie_result['original_url'], **extra_info}
1689
1690 extract_flat = self.params.get('extract_flat', False)
1691 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info)
1692 or extract_flat is True):
1693 info_copy = ie_result.copy()
1694 ie = try_get(ie_result.get('ie_key'), self.get_info_extractor)
1695 if ie and not ie_result.get('id'):
1696 info_copy['id'] = ie.get_temp_id(ie_result['url'])
1697 self.add_default_extra_info(info_copy, ie, ie_result['url'])
1698 self.add_extra_info(info_copy, extra_info)
1699 info_copy, _ = self.pre_process(info_copy)
1700 self._fill_common_fields(info_copy, False)
1701 self.__forced_printings(info_copy)
1702 self._raise_pending_errors(info_copy)
1703 if self.params.get('force_write_download_archive', False):
1704 self.record_download_archive(info_copy)
1705 return ie_result
1706
1707 if result_type == 'video':
1708 self.add_extra_info(ie_result, extra_info)
1709 ie_result = self.process_video_result(ie_result, download=download)
1710 self._raise_pending_errors(ie_result)
1711 additional_urls = (ie_result or {}).get('additional_urls')
1712 if additional_urls:
1713 # TODO: Improve MetadataParserPP to allow setting a list
1714 if isinstance(additional_urls, str):
1715 additional_urls = [additional_urls]
1716 self.to_screen(
1717 '[info] %s: %d additional URL(s) requested' % (ie_result['id'], len(additional_urls)))
1718 self.write_debug('Additional URLs: "%s"' % '", "'.join(additional_urls))
1719 ie_result['additional_entries'] = [
1720 self.extract_info(
1721 url, download, extra_info=extra_info,
1722 force_generic_extractor=self.params.get('force_generic_extractor'))
1723 for url in additional_urls
1724 ]
1725 return ie_result
1726 elif result_type == 'url':
1727 # We have to add extra_info to the results because it may be
1728 # contained in a playlist
1729 return self.extract_info(
1730 ie_result['url'], download,
1731 ie_key=ie_result.get('ie_key'),
1732 extra_info=extra_info)
1733 elif result_type == 'url_transparent':
1734 # Use the information from the embedding page
1735 info = self.extract_info(
1736 ie_result['url'], ie_key=ie_result.get('ie_key'),
1737 extra_info=extra_info, download=False, process=False)
1738
1739 # extract_info may return None when ignoreerrors is enabled and
1740 # extraction failed with an error, don't crash and return early
1741 # in this case
1742 if not info:
1743 return info
1744
1745 exempted_fields = {'_type', 'url', 'ie_key'}
1746 if not ie_result.get('section_end') and ie_result.get('section_start') is None:
1747 # For video clips, the id etc of the clip extractor should be used
1748 exempted_fields |= {'id', 'extractor', 'extractor_key'}
1749
1750 new_result = info.copy()
1751 new_result.update(filter_dict(ie_result, lambda k, v: v is not None and k not in exempted_fields))
1752
1753 # Extracted info may not be a video result (i.e.
1754 # info.get('_type', 'video') != video) but rather an url or
1755 # url_transparent. In such cases outer metadata (from ie_result)
1756 # should be propagated to inner one (info). For this to happen
1757 # _type of info should be overridden with url_transparent. This
1758 # fixes issue from https://github.com/ytdl-org/youtube-dl/pull/11163.
1759 if new_result.get('_type') == 'url':
1760 new_result['_type'] = 'url_transparent'
1761
1762 return self.process_ie_result(
1763 new_result, download=download, extra_info=extra_info)
1764 elif result_type in ('playlist', 'multi_video'):
1765 # Protect from infinite recursion due to recursively nested playlists
1766 # (see https://github.com/ytdl-org/youtube-dl/issues/27833)
1767 webpage_url = ie_result.get('webpage_url') # Playlists maynot have webpage_url
1768 if webpage_url and webpage_url in self._playlist_urls:
1769 self.to_screen(
1770 '[download] Skipping already downloaded playlist: %s'
1771 % ie_result.get('title') or ie_result.get('id'))
1772 return
1773
1774 self._playlist_level += 1
1775 self._playlist_urls.add(webpage_url)
1776 self._fill_common_fields(ie_result, False)
1777 self._sanitize_thumbnails(ie_result)
1778 try:
1779 return self.__process_playlist(ie_result, download)
1780 finally:
1781 self._playlist_level -= 1
1782 if not self._playlist_level:
1783 self._playlist_urls.clear()
1784 elif result_type == 'compat_list':
1785 self.report_warning(
1786 'Extractor %s returned a compat_list result. '
1787 'It needs to be updated.' % ie_result.get('extractor'))
1788
1789 def _fixup(r):
1790 self.add_extra_info(r, {
1791 'extractor': ie_result['extractor'],
1792 'webpage_url': ie_result['webpage_url'],
1793 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1794 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1795 'extractor_key': ie_result['extractor_key'],
1796 })
1797 return r
1798 ie_result['entries'] = [
1799 self.process_ie_result(_fixup(r), download, extra_info)
1800 for r in ie_result['entries']
1801 ]
1802 return ie_result
1803 else:
1804 raise Exception('Invalid result type: %s' % result_type)
1805
1806 def _ensure_dir_exists(self, path):
1807 return make_dir(path, self.report_error)
1808
1809 @staticmethod
1810 def _playlist_infodict(ie_result, strict=False, **kwargs):
1811 info = {
1812 'playlist_count': ie_result.get('playlist_count'),
1813 'playlist': ie_result.get('title') or ie_result.get('id'),
1814 'playlist_id': ie_result.get('id'),
1815 'playlist_title': ie_result.get('title'),
1816 'playlist_uploader': ie_result.get('uploader'),
1817 'playlist_uploader_id': ie_result.get('uploader_id'),
1818 **kwargs,
1819 }
1820 if strict:
1821 return info
1822 if ie_result.get('webpage_url'):
1823 info.update({
1824 'webpage_url': ie_result['webpage_url'],
1825 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1826 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1827 })
1828 return {
1829 **info,
1830 'playlist_index': 0,
1831 '__last_playlist_index': max(ie_result.get('requested_entries') or (0, 0)),
1832 'extractor': ie_result['extractor'],
1833 'extractor_key': ie_result['extractor_key'],
1834 }
1835
1836 def __process_playlist(self, ie_result, download):
1837 """Process each entry in the playlist"""
1838 assert ie_result['_type'] in ('playlist', 'multi_video')
1839
1840 common_info = self._playlist_infodict(ie_result, strict=True)
1841 title = common_info.get('playlist') or '<Untitled>'
1842 if self._match_entry(common_info, incomplete=True) is not None:
1843 return
1844 self.to_screen(f'[download] Downloading {ie_result["_type"]}: {title}')
1845
1846 all_entries = PlaylistEntries(self, ie_result)
1847 entries = orderedSet(all_entries.get_requested_items(), lazy=True)
1848
1849 lazy = self.params.get('lazy_playlist')
1850 if lazy:
1851 resolved_entries, n_entries = [], 'N/A'
1852 ie_result['requested_entries'], ie_result['entries'] = None, None
1853 else:
1854 entries = resolved_entries = list(entries)
1855 n_entries = len(resolved_entries)
1856 ie_result['requested_entries'], ie_result['entries'] = tuple(zip(*resolved_entries)) or ([], [])
1857 if not ie_result.get('playlist_count'):
1858 # Better to do this after potentially exhausting entries
1859 ie_result['playlist_count'] = all_entries.get_full_count()
1860
1861 extra = self._playlist_infodict(ie_result, n_entries=int_or_none(n_entries))
1862 ie_copy = collections.ChainMap(ie_result, extra)
1863
1864 _infojson_written = False
1865 write_playlist_files = self.params.get('allow_playlist_files', True)
1866 if write_playlist_files and self.params.get('list_thumbnails'):
1867 self.list_thumbnails(ie_result)
1868 if write_playlist_files and not self.params.get('simulate'):
1869 _infojson_written = self._write_info_json(
1870 'playlist', ie_result, self.prepare_filename(ie_copy, 'pl_infojson'))
1871 if _infojson_written is None:
1872 return
1873 if self._write_description('playlist', ie_result,
1874 self.prepare_filename(ie_copy, 'pl_description')) is None:
1875 return
1876 # TODO: This should be passed to ThumbnailsConvertor if necessary
1877 self._write_thumbnails('playlist', ie_result, self.prepare_filename(ie_copy, 'pl_thumbnail'))
1878
1879 if lazy:
1880 if self.params.get('playlistreverse') or self.params.get('playlistrandom'):
1881 self.report_warning('playlistreverse and playlistrandom are not supported with lazy_playlist', only_once=True)
1882 elif self.params.get('playlistreverse'):
1883 entries.reverse()
1884 elif self.params.get('playlistrandom'):
1885 random.shuffle(entries)
1886
1887 self.to_screen(f'[{ie_result["extractor"]}] Playlist {title}: Downloading {n_entries} items'
1888 f'{format_field(ie_result, "playlist_count", " of %s")}')
1889
1890 keep_resolved_entries = self.params.get('extract_flat') != 'discard'
1891 if self.params.get('extract_flat') == 'discard_in_playlist':
1892 keep_resolved_entries = ie_result['_type'] != 'playlist'
1893 if keep_resolved_entries:
1894 self.write_debug('The information of all playlist entries will be held in memory')
1895
1896 failures = 0
1897 max_failures = self.params.get('skip_playlist_after_errors') or float('inf')
1898 for i, (playlist_index, entry) in enumerate(entries):
1899 if lazy:
1900 resolved_entries.append((playlist_index, entry))
1901 if not entry:
1902 continue
1903
1904 entry['__x_forwarded_for_ip'] = ie_result.get('__x_forwarded_for_ip')
1905 if not lazy and 'playlist-index' in self.params.get('compat_opts', []):
1906 playlist_index = ie_result['requested_entries'][i]
1907
1908 entry_copy = collections.ChainMap(entry, {
1909 **common_info,
1910 'n_entries': int_or_none(n_entries),
1911 'playlist_index': playlist_index,
1912 'playlist_autonumber': i + 1,
1913 })
1914
1915 if self._match_entry(entry_copy, incomplete=True) is not None:
1916 # For compatabilty with youtube-dl. See https://github.com/yt-dlp/yt-dlp/issues/4369
1917 resolved_entries[i] = (playlist_index, NO_DEFAULT)
1918 continue
1919
1920 self.to_screen('[download] Downloading item %s of %s' % (
1921 self._format_screen(i + 1, self.Styles.ID), self._format_screen(n_entries, self.Styles.EMPHASIS)))
1922
1923 entry_result = self.__process_iterable_entry(entry, download, collections.ChainMap({
1924 'playlist_index': playlist_index,
1925 'playlist_autonumber': i + 1,
1926 }, extra))
1927 if not entry_result:
1928 failures += 1
1929 if failures >= max_failures:
1930 self.report_error(
1931 f'Skipping the remaining entries in playlist "{title}" since {failures} items failed extraction')
1932 break
1933 if keep_resolved_entries:
1934 resolved_entries[i] = (playlist_index, entry_result)
1935
1936 # Update with processed data
1937 ie_result['entries'] = [e for _, e in resolved_entries if e is not NO_DEFAULT]
1938 ie_result['requested_entries'] = [i for i, e in resolved_entries if e is not NO_DEFAULT]
1939 if ie_result['requested_entries'] == try_call(lambda: list(range(1, ie_result['playlist_count'] + 1))):
1940 # Do not set for full playlist
1941 ie_result.pop('requested_entries')
1942
1943 # Write the updated info to json
1944 if _infojson_written is True and self._write_info_json(
1945 'updated playlist', ie_result,
1946 self.prepare_filename(ie_copy, 'pl_infojson'), overwrite=True) is None:
1947 return
1948
1949 ie_result = self.run_all_pps('playlist', ie_result)
1950 self.to_screen(f'[download] Finished downloading playlist: {title}')
1951 return ie_result
1952
1953 @_handle_extraction_exceptions
1954 def __process_iterable_entry(self, entry, download, extra_info):
1955 return self.process_ie_result(
1956 entry, download=download, extra_info=extra_info)
1957
1958 def _build_format_filter(self, filter_spec):
1959 " Returns a function to filter the formats according to the filter_spec "
1960
1961 OPERATORS = {
1962 '<': operator.lt,
1963 '<=': operator.le,
1964 '>': operator.gt,
1965 '>=': operator.ge,
1966 '=': operator.eq,
1967 '!=': operator.ne,
1968 }
1969 operator_rex = re.compile(r'''(?x)\s*
1970 (?P<key>[\w.-]+)\s*
1971 (?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
1972 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)\s*
1973 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
1974 m = operator_rex.fullmatch(filter_spec)
1975 if m:
1976 try:
1977 comparison_value = int(m.group('value'))
1978 except ValueError:
1979 comparison_value = parse_filesize(m.group('value'))
1980 if comparison_value is None:
1981 comparison_value = parse_filesize(m.group('value') + 'B')
1982 if comparison_value is None:
1983 raise ValueError(
1984 'Invalid value %r in format specification %r' % (
1985 m.group('value'), filter_spec))
1986 op = OPERATORS[m.group('op')]
1987
1988 if not m:
1989 STR_OPERATORS = {
1990 '=': operator.eq,
1991 '^=': lambda attr, value: attr.startswith(value),
1992 '$=': lambda attr, value: attr.endswith(value),
1993 '*=': lambda attr, value: value in attr,
1994 '~=': lambda attr, value: value.search(attr) is not None
1995 }
1996 str_operator_rex = re.compile(r'''(?x)\s*
1997 (?P<key>[a-zA-Z0-9._-]+)\s*
1998 (?P<negation>!\s*)?(?P<op>%s)\s*(?P<none_inclusive>\?\s*)?
1999 (?P<quote>["'])?
2000 (?P<value>(?(quote)(?:(?!(?P=quote))[^\\]|\\.)+|[\w.-]+))
2001 (?(quote)(?P=quote))\s*
2002 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
2003 m = str_operator_rex.fullmatch(filter_spec)
2004 if m:
2005 if m.group('op') == '~=':
2006 comparison_value = re.compile(m.group('value'))
2007 else:
2008 comparison_value = re.sub(r'''\\([\\"'])''', r'\1', m.group('value'))
2009 str_op = STR_OPERATORS[m.group('op')]
2010 if m.group('negation'):
2011 op = lambda attr, value: not str_op(attr, value)
2012 else:
2013 op = str_op
2014
2015 if not m:
2016 raise SyntaxError('Invalid filter specification %r' % filter_spec)
2017
2018 def _filter(f):
2019 actual_value = f.get(m.group('key'))
2020 if actual_value is None:
2021 return m.group('none_inclusive')
2022 return op(actual_value, comparison_value)
2023 return _filter
2024
2025 def _check_formats(self, formats):
2026 for f in formats:
2027 self.to_screen('[info] Testing format %s' % f['format_id'])
2028 path = self.get_output_path('temp')
2029 if not self._ensure_dir_exists(f'{path}/'):
2030 continue
2031 temp_file = tempfile.NamedTemporaryFile(suffix='.tmp', delete=False, dir=path or None)
2032 temp_file.close()
2033 try:
2034 success, _ = self.dl(temp_file.name, f, test=True)
2035 except (DownloadError, OSError, ValueError) + network_exceptions:
2036 success = False
2037 finally:
2038 if os.path.exists(temp_file.name):
2039 try:
2040 os.remove(temp_file.name)
2041 except OSError:
2042 self.report_warning('Unable to delete temporary file "%s"' % temp_file.name)
2043 if success:
2044 yield f
2045 else:
2046 self.to_screen('[info] Unable to download format %s. Skipping...' % f['format_id'])
2047
2048 def _default_format_spec(self, info_dict, download=True):
2049
2050 def can_merge():
2051 merger = FFmpegMergerPP(self)
2052 return merger.available and merger.can_merge()
2053
2054 prefer_best = (
2055 not self.params.get('simulate')
2056 and download
2057 and (
2058 not can_merge()
2059 or info_dict.get('is_live') and not self.params.get('live_from_start')
2060 or self.params['outtmpl']['default'] == '-'))
2061 compat = (
2062 prefer_best
2063 or self.params.get('allow_multiple_audio_streams', False)
2064 or 'format-spec' in self.params['compat_opts'])
2065
2066 return (
2067 'best/bestvideo+bestaudio' if prefer_best
2068 else 'bestvideo*+bestaudio/best' if not compat
2069 else 'bestvideo+bestaudio/best')
2070
2071 def build_format_selector(self, format_spec):
2072 def syntax_error(note, start):
2073 message = (
2074 'Invalid format specification: '
2075 '{}\n\t{}\n\t{}^'.format(note, format_spec, ' ' * start[1]))
2076 return SyntaxError(message)
2077
2078 PICKFIRST = 'PICKFIRST'
2079 MERGE = 'MERGE'
2080 SINGLE = 'SINGLE'
2081 GROUP = 'GROUP'
2082 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
2083
2084 allow_multiple_streams = {'audio': self.params.get('allow_multiple_audio_streams', False),
2085 'video': self.params.get('allow_multiple_video_streams', False)}
2086
2087 check_formats = self.params.get('check_formats') == 'selected'
2088
2089 def _parse_filter(tokens):
2090 filter_parts = []
2091 for type, string_, start, _, _ in tokens:
2092 if type == tokenize.OP and string_ == ']':
2093 return ''.join(filter_parts)
2094 else:
2095 filter_parts.append(string_)
2096
2097 def _remove_unused_ops(tokens):
2098 # Remove operators that we don't use and join them with the surrounding strings.
2099 # E.g. 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
2100 ALLOWED_OPS = ('/', '+', ',', '(', ')')
2101 last_string, last_start, last_end, last_line = None, None, None, None
2102 for type, string_, start, end, line in tokens:
2103 if type == tokenize.OP and string_ == '[':
2104 if last_string:
2105 yield tokenize.NAME, last_string, last_start, last_end, last_line
2106 last_string = None
2107 yield type, string_, start, end, line
2108 # everything inside brackets will be handled by _parse_filter
2109 for type, string_, start, end, line in tokens:
2110 yield type, string_, start, end, line
2111 if type == tokenize.OP and string_ == ']':
2112 break
2113 elif type == tokenize.OP and string_ in ALLOWED_OPS:
2114 if last_string:
2115 yield tokenize.NAME, last_string, last_start, last_end, last_line
2116 last_string = None
2117 yield type, string_, start, end, line
2118 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
2119 if not last_string:
2120 last_string = string_
2121 last_start = start
2122 last_end = end
2123 else:
2124 last_string += string_
2125 if last_string:
2126 yield tokenize.NAME, last_string, last_start, last_end, last_line
2127
2128 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
2129 selectors = []
2130 current_selector = None
2131 for type, string_, start, _, _ in tokens:
2132 # ENCODING is only defined in python 3.x
2133 if type == getattr(tokenize, 'ENCODING', None):
2134 continue
2135 elif type in [tokenize.NAME, tokenize.NUMBER]:
2136 current_selector = FormatSelector(SINGLE, string_, [])
2137 elif type == tokenize.OP:
2138 if string_ == ')':
2139 if not inside_group:
2140 # ')' will be handled by the parentheses group
2141 tokens.restore_last_token()
2142 break
2143 elif inside_merge and string_ in ['/', ',']:
2144 tokens.restore_last_token()
2145 break
2146 elif inside_choice and string_ == ',':
2147 tokens.restore_last_token()
2148 break
2149 elif string_ == ',':
2150 if not current_selector:
2151 raise syntax_error('"," must follow a format selector', start)
2152 selectors.append(current_selector)
2153 current_selector = None
2154 elif string_ == '/':
2155 if not current_selector:
2156 raise syntax_error('"/" must follow a format selector', start)
2157 first_choice = current_selector
2158 second_choice = _parse_format_selection(tokens, inside_choice=True)
2159 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
2160 elif string_ == '[':
2161 if not current_selector:
2162 current_selector = FormatSelector(SINGLE, 'best', [])
2163 format_filter = _parse_filter(tokens)
2164 current_selector.filters.append(format_filter)
2165 elif string_ == '(':
2166 if current_selector:
2167 raise syntax_error('Unexpected "("', start)
2168 group = _parse_format_selection(tokens, inside_group=True)
2169 current_selector = FormatSelector(GROUP, group, [])
2170 elif string_ == '+':
2171 if not current_selector:
2172 raise syntax_error('Unexpected "+"', start)
2173 selector_1 = current_selector
2174 selector_2 = _parse_format_selection(tokens, inside_merge=True)
2175 if not selector_2:
2176 raise syntax_error('Expected a selector', start)
2177 current_selector = FormatSelector(MERGE, (selector_1, selector_2), [])
2178 else:
2179 raise syntax_error(f'Operator not recognized: "{string_}"', start)
2180 elif type == tokenize.ENDMARKER:
2181 break
2182 if current_selector:
2183 selectors.append(current_selector)
2184 return selectors
2185
2186 def _merge(formats_pair):
2187 format_1, format_2 = formats_pair
2188
2189 formats_info = []
2190 formats_info.extend(format_1.get('requested_formats', (format_1,)))
2191 formats_info.extend(format_2.get('requested_formats', (format_2,)))
2192
2193 if not allow_multiple_streams['video'] or not allow_multiple_streams['audio']:
2194 get_no_more = {'video': False, 'audio': False}
2195 for (i, fmt_info) in enumerate(formats_info):
2196 if fmt_info.get('acodec') == fmt_info.get('vcodec') == 'none':
2197 formats_info.pop(i)
2198 continue
2199 for aud_vid in ['audio', 'video']:
2200 if not allow_multiple_streams[aud_vid] and fmt_info.get(aud_vid[0] + 'codec') != 'none':
2201 if get_no_more[aud_vid]:
2202 formats_info.pop(i)
2203 break
2204 get_no_more[aud_vid] = True
2205
2206 if len(formats_info) == 1:
2207 return formats_info[0]
2208
2209 video_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('vcodec') != 'none']
2210 audio_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('acodec') != 'none']
2211
2212 the_only_video = video_fmts[0] if len(video_fmts) == 1 else None
2213 the_only_audio = audio_fmts[0] if len(audio_fmts) == 1 else None
2214
2215 output_ext = get_compatible_ext(
2216 vcodecs=[f.get('vcodec') for f in video_fmts],
2217 acodecs=[f.get('acodec') for f in audio_fmts],
2218 vexts=[f['ext'] for f in video_fmts],
2219 aexts=[f['ext'] for f in audio_fmts],
2220 preferences=(try_call(lambda: self.params['merge_output_format'].split('/'))
2221 or self.params.get('prefer_free_formats') and ('webm', 'mkv')))
2222
2223 filtered = lambda *keys: filter(None, (traverse_obj(fmt, *keys) for fmt in formats_info))
2224
2225 new_dict = {
2226 'requested_formats': formats_info,
2227 'format': '+'.join(filtered('format')),
2228 'format_id': '+'.join(filtered('format_id')),
2229 'ext': output_ext,
2230 'protocol': '+'.join(map(determine_protocol, formats_info)),
2231 'language': '+'.join(orderedSet(filtered('language'))) or None,
2232 'format_note': '+'.join(orderedSet(filtered('format_note'))) or None,
2233 'filesize_approx': sum(filtered('filesize', 'filesize_approx')) or None,
2234 'tbr': sum(filtered('tbr', 'vbr', 'abr')),
2235 }
2236
2237 if the_only_video:
2238 new_dict.update({
2239 'width': the_only_video.get('width'),
2240 'height': the_only_video.get('height'),
2241 'resolution': the_only_video.get('resolution') or self.format_resolution(the_only_video),
2242 'fps': the_only_video.get('fps'),
2243 'dynamic_range': the_only_video.get('dynamic_range'),
2244 'vcodec': the_only_video.get('vcodec'),
2245 'vbr': the_only_video.get('vbr'),
2246 'stretched_ratio': the_only_video.get('stretched_ratio'),
2247 'aspect_ratio': the_only_video.get('aspect_ratio'),
2248 })
2249
2250 if the_only_audio:
2251 new_dict.update({
2252 'acodec': the_only_audio.get('acodec'),
2253 'abr': the_only_audio.get('abr'),
2254 'asr': the_only_audio.get('asr'),
2255 'audio_channels': the_only_audio.get('audio_channels')
2256 })
2257
2258 return new_dict
2259
2260 def _check_formats(formats):
2261 if not check_formats:
2262 yield from formats
2263 return
2264 yield from self._check_formats(formats)
2265
2266 def _build_selector_function(selector):
2267 if isinstance(selector, list): # ,
2268 fs = [_build_selector_function(s) for s in selector]
2269
2270 def selector_function(ctx):
2271 for f in fs:
2272 yield from f(ctx)
2273 return selector_function
2274
2275 elif selector.type == GROUP: # ()
2276 selector_function = _build_selector_function(selector.selector)
2277
2278 elif selector.type == PICKFIRST: # /
2279 fs = [_build_selector_function(s) for s in selector.selector]
2280
2281 def selector_function(ctx):
2282 for f in fs:
2283 picked_formats = list(f(ctx))
2284 if picked_formats:
2285 return picked_formats
2286 return []
2287
2288 elif selector.type == MERGE: # +
2289 selector_1, selector_2 = map(_build_selector_function, selector.selector)
2290
2291 def selector_function(ctx):
2292 for pair in itertools.product(selector_1(ctx), selector_2(ctx)):
2293 yield _merge(pair)
2294
2295 elif selector.type == SINGLE: # atom
2296 format_spec = selector.selector or 'best'
2297
2298 # TODO: Add allvideo, allaudio etc by generalizing the code with best/worst selector
2299 if format_spec == 'all':
2300 def selector_function(ctx):
2301 yield from _check_formats(ctx['formats'][::-1])
2302 elif format_spec == 'mergeall':
2303 def selector_function(ctx):
2304 formats = list(_check_formats(
2305 f for f in ctx['formats'] if f.get('vcodec') != 'none' or f.get('acodec') != 'none'))
2306 if not formats:
2307 return
2308 merged_format = formats[-1]
2309 for f in formats[-2::-1]:
2310 merged_format = _merge((merged_format, f))
2311 yield merged_format
2312
2313 else:
2314 format_fallback, seperate_fallback, format_reverse, format_idx = False, None, True, 1
2315 mobj = re.match(
2316 r'(?P<bw>best|worst|b|w)(?P<type>video|audio|v|a)?(?P<mod>\*)?(?:\.(?P<n>[1-9]\d*))?$',
2317 format_spec)
2318 if mobj is not None:
2319 format_idx = int_or_none(mobj.group('n'), default=1)
2320 format_reverse = mobj.group('bw')[0] == 'b'
2321 format_type = (mobj.group('type') or [None])[0]
2322 not_format_type = {'v': 'a', 'a': 'v'}.get(format_type)
2323 format_modified = mobj.group('mod') is not None
2324
2325 format_fallback = not format_type and not format_modified # for b, w
2326 _filter_f = (
2327 (lambda f: f.get('%scodec' % format_type) != 'none')
2328 if format_type and format_modified # bv*, ba*, wv*, wa*
2329 else (lambda f: f.get('%scodec' % not_format_type) == 'none')
2330 if format_type # bv, ba, wv, wa
2331 else (lambda f: f.get('vcodec') != 'none' and f.get('acodec') != 'none')
2332 if not format_modified # b, w
2333 else lambda f: True) # b*, w*
2334 filter_f = lambda f: _filter_f(f) and (
2335 f.get('vcodec') != 'none' or f.get('acodec') != 'none')
2336 else:
2337 if format_spec in self._format_selection_exts['audio']:
2338 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none'
2339 elif format_spec in self._format_selection_exts['video']:
2340 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none' and f.get('vcodec') != 'none'
2341 seperate_fallback = lambda f: f.get('ext') == format_spec and f.get('vcodec') != 'none'
2342 elif format_spec in self._format_selection_exts['storyboards']:
2343 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') == 'none' and f.get('vcodec') == 'none'
2344 else:
2345 filter_f = lambda f: f.get('format_id') == format_spec # id
2346
2347 def selector_function(ctx):
2348 formats = list(ctx['formats'])
2349 matches = list(filter(filter_f, formats)) if filter_f is not None else formats
2350 if not matches:
2351 if format_fallback and ctx['incomplete_formats']:
2352 # for extractors with incomplete formats (audio only (soundcloud)
2353 # or video only (imgur)) best/worst will fallback to
2354 # best/worst {video,audio}-only format
2355 matches = formats
2356 elif seperate_fallback and not ctx['has_merged_format']:
2357 # for compatibility with youtube-dl when there is no pre-merged format
2358 matches = list(filter(seperate_fallback, formats))
2359 matches = LazyList(_check_formats(matches[::-1 if format_reverse else 1]))
2360 try:
2361 yield matches[format_idx - 1]
2362 except LazyList.IndexError:
2363 return
2364
2365 filters = [self._build_format_filter(f) for f in selector.filters]
2366
2367 def final_selector(ctx):
2368 ctx_copy = dict(ctx)
2369 for _filter in filters:
2370 ctx_copy['formats'] = list(filter(_filter, ctx_copy['formats']))
2371 return selector_function(ctx_copy)
2372 return final_selector
2373
2374 stream = io.BytesIO(format_spec.encode())
2375 try:
2376 tokens = list(_remove_unused_ops(tokenize.tokenize(stream.readline)))
2377 except tokenize.TokenError:
2378 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
2379
2380 class TokenIterator:
2381 def __init__(self, tokens):
2382 self.tokens = tokens
2383 self.counter = 0
2384
2385 def __iter__(self):
2386 return self
2387
2388 def __next__(self):
2389 if self.counter >= len(self.tokens):
2390 raise StopIteration()
2391 value = self.tokens[self.counter]
2392 self.counter += 1
2393 return value
2394
2395 next = __next__
2396
2397 def restore_last_token(self):
2398 self.counter -= 1
2399
2400 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
2401 return _build_selector_function(parsed_selector)
2402
2403 def _calc_headers(self, info_dict):
2404 res = merge_headers(self.params['http_headers'], info_dict.get('http_headers') or {})
2405 if 'Youtubedl-No-Compression' in res: # deprecated
2406 res.pop('Youtubedl-No-Compression', None)
2407 res['Accept-Encoding'] = 'identity'
2408 cookies = self.cookiejar.get_cookie_header(info_dict['url'])
2409 if cookies:
2410 res['Cookie'] = cookies
2411
2412 if 'X-Forwarded-For' not in res:
2413 x_forwarded_for_ip = info_dict.get('__x_forwarded_for_ip')
2414 if x_forwarded_for_ip:
2415 res['X-Forwarded-For'] = x_forwarded_for_ip
2416
2417 return res
2418
2419 def _calc_cookies(self, url):
2420 self.deprecation_warning('"YoutubeDL._calc_cookies" is deprecated and may be removed in a future version')
2421 return self.cookiejar.get_cookie_header(url)
2422
2423 def _sort_thumbnails(self, thumbnails):
2424 thumbnails.sort(key=lambda t: (
2425 t.get('preference') if t.get('preference') is not None else -1,
2426 t.get('width') if t.get('width') is not None else -1,
2427 t.get('height') if t.get('height') is not None else -1,
2428 t.get('id') if t.get('id') is not None else '',
2429 t.get('url')))
2430
2431 def _sanitize_thumbnails(self, info_dict):
2432 thumbnails = info_dict.get('thumbnails')
2433 if thumbnails is None:
2434 thumbnail = info_dict.get('thumbnail')
2435 if thumbnail:
2436 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
2437 if not thumbnails:
2438 return
2439
2440 def check_thumbnails(thumbnails):
2441 for t in thumbnails:
2442 self.to_screen(f'[info] Testing thumbnail {t["id"]}')
2443 try:
2444 self.urlopen(HEADRequest(t['url']))
2445 except network_exceptions as err:
2446 self.to_screen(f'[info] Unable to connect to thumbnail {t["id"]} URL {t["url"]!r} - {err}. Skipping...')
2447 continue
2448 yield t
2449
2450 self._sort_thumbnails(thumbnails)
2451 for i, t in enumerate(thumbnails):
2452 if t.get('id') is None:
2453 t['id'] = '%d' % i
2454 if t.get('width') and t.get('height'):
2455 t['resolution'] = '%dx%d' % (t['width'], t['height'])
2456 t['url'] = sanitize_url(t['url'])
2457
2458 if self.params.get('check_formats') is True:
2459 info_dict['thumbnails'] = LazyList(check_thumbnails(thumbnails[::-1]), reverse=True)
2460 else:
2461 info_dict['thumbnails'] = thumbnails
2462
2463 def _fill_common_fields(self, info_dict, final=True):
2464 # TODO: move sanitization here
2465 if final:
2466 title = info_dict['fulltitle'] = info_dict.get('title')
2467 if not title:
2468 if title == '':
2469 self.write_debug('Extractor gave empty title. Creating a generic title')
2470 else:
2471 self.report_warning('Extractor failed to obtain "title". Creating a generic title instead')
2472 info_dict['title'] = f'{info_dict["extractor"].replace(":", "-")} video #{info_dict["id"]}'
2473
2474 if info_dict.get('duration') is not None:
2475 info_dict['duration_string'] = formatSeconds(info_dict['duration'])
2476
2477 for ts_key, date_key in (
2478 ('timestamp', 'upload_date'),
2479 ('release_timestamp', 'release_date'),
2480 ('modified_timestamp', 'modified_date'),
2481 ):
2482 if info_dict.get(date_key) is None and info_dict.get(ts_key) is not None:
2483 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
2484 # see http://bugs.python.org/issue1646728)
2485 with contextlib.suppress(ValueError, OverflowError, OSError):
2486 upload_date = datetime.datetime.utcfromtimestamp(info_dict[ts_key])
2487 info_dict[date_key] = upload_date.strftime('%Y%m%d')
2488
2489 live_keys = ('is_live', 'was_live')
2490 live_status = info_dict.get('live_status')
2491 if live_status is None:
2492 for key in live_keys:
2493 if info_dict.get(key) is False:
2494 continue
2495 if info_dict.get(key):
2496 live_status = key
2497 break
2498 if all(info_dict.get(key) is False for key in live_keys):
2499 live_status = 'not_live'
2500 if live_status:
2501 info_dict['live_status'] = live_status
2502 for key in live_keys:
2503 if info_dict.get(key) is None:
2504 info_dict[key] = (live_status == key)
2505 if live_status == 'post_live':
2506 info_dict['was_live'] = True
2507
2508 # Auto generate title fields corresponding to the *_number fields when missing
2509 # in order to always have clean titles. This is very common for TV series.
2510 for field in ('chapter', 'season', 'episode'):
2511 if final and info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
2512 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
2513
2514 def _raise_pending_errors(self, info):
2515 err = info.pop('__pending_error', None)
2516 if err:
2517 self.report_error(err, tb=False)
2518
2519 def sort_formats(self, info_dict):
2520 formats = self._get_formats(info_dict)
2521 formats.sort(key=FormatSorter(
2522 self, info_dict.get('_format_sort_fields') or []).calculate_preference)
2523
2524 def process_video_result(self, info_dict, download=True):
2525 assert info_dict.get('_type', 'video') == 'video'
2526 self._num_videos += 1
2527
2528 if 'id' not in info_dict:
2529 raise ExtractorError('Missing "id" field in extractor result', ie=info_dict['extractor'])
2530 elif not info_dict.get('id'):
2531 raise ExtractorError('Extractor failed to obtain "id"', ie=info_dict['extractor'])
2532
2533 def report_force_conversion(field, field_not, conversion):
2534 self.report_warning(
2535 '"%s" field is not %s - forcing %s conversion, there is an error in extractor'
2536 % (field, field_not, conversion))
2537
2538 def sanitize_string_field(info, string_field):
2539 field = info.get(string_field)
2540 if field is None or isinstance(field, str):
2541 return
2542 report_force_conversion(string_field, 'a string', 'string')
2543 info[string_field] = str(field)
2544
2545 def sanitize_numeric_fields(info):
2546 for numeric_field in self._NUMERIC_FIELDS:
2547 field = info.get(numeric_field)
2548 if field is None or isinstance(field, (int, float)):
2549 continue
2550 report_force_conversion(numeric_field, 'numeric', 'int')
2551 info[numeric_field] = int_or_none(field)
2552
2553 sanitize_string_field(info_dict, 'id')
2554 sanitize_numeric_fields(info_dict)
2555 if info_dict.get('section_end') and info_dict.get('section_start') is not None:
2556 info_dict['duration'] = round(info_dict['section_end'] - info_dict['section_start'], 3)
2557 if (info_dict.get('duration') or 0) <= 0 and info_dict.pop('duration', None):
2558 self.report_warning('"duration" field is negative, there is an error in extractor')
2559
2560 chapters = info_dict.get('chapters') or []
2561 if chapters and chapters[0].get('start_time'):
2562 chapters.insert(0, {'start_time': 0})
2563
2564 dummy_chapter = {'end_time': 0, 'start_time': info_dict.get('duration')}
2565 for idx, (prev, current, next_) in enumerate(zip(
2566 (dummy_chapter, *chapters), chapters, (*chapters[1:], dummy_chapter)), 1):
2567 if current.get('start_time') is None:
2568 current['start_time'] = prev.get('end_time')
2569 if not current.get('end_time'):
2570 current['end_time'] = next_.get('start_time')
2571 if not current.get('title'):
2572 current['title'] = f'<Untitled Chapter {idx}>'
2573
2574 if 'playlist' not in info_dict:
2575 # It isn't part of a playlist
2576 info_dict['playlist'] = None
2577 info_dict['playlist_index'] = None
2578
2579 self._sanitize_thumbnails(info_dict)
2580
2581 thumbnail = info_dict.get('thumbnail')
2582 thumbnails = info_dict.get('thumbnails')
2583 if thumbnail:
2584 info_dict['thumbnail'] = sanitize_url(thumbnail)
2585 elif thumbnails:
2586 info_dict['thumbnail'] = thumbnails[-1]['url']
2587
2588 if info_dict.get('display_id') is None and 'id' in info_dict:
2589 info_dict['display_id'] = info_dict['id']
2590
2591 self._fill_common_fields(info_dict)
2592
2593 for cc_kind in ('subtitles', 'automatic_captions'):
2594 cc = info_dict.get(cc_kind)
2595 if cc:
2596 for _, subtitle in cc.items():
2597 for subtitle_format in subtitle:
2598 if subtitle_format.get('url'):
2599 subtitle_format['url'] = sanitize_url(subtitle_format['url'])
2600 if subtitle_format.get('ext') is None:
2601 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
2602
2603 automatic_captions = info_dict.get('automatic_captions')
2604 subtitles = info_dict.get('subtitles')
2605
2606 info_dict['requested_subtitles'] = self.process_subtitles(
2607 info_dict['id'], subtitles, automatic_captions)
2608
2609 formats = self._get_formats(info_dict)
2610
2611 # Backward compatibility with InfoExtractor._sort_formats
2612 field_preference = (formats or [{}])[0].pop('__sort_fields', None)
2613 if field_preference:
2614 info_dict['_format_sort_fields'] = field_preference
2615
2616 # or None ensures --clean-infojson removes it
2617 info_dict['_has_drm'] = any(f.get('has_drm') for f in formats) or None
2618 if not self.params.get('allow_unplayable_formats'):
2619 formats = [f for f in formats if not f.get('has_drm')]
2620
2621 if formats and all(f.get('acodec') == f.get('vcodec') == 'none' for f in formats):
2622 self.report_warning(
2623 f'{"This video is DRM protected and " if info_dict["_has_drm"] else ""}'
2624 'only images are available for download. Use --list-formats to see them'.capitalize())
2625
2626 get_from_start = not info_dict.get('is_live') or bool(self.params.get('live_from_start'))
2627 if not get_from_start:
2628 info_dict['title'] += ' ' + datetime.datetime.now().strftime('%Y-%m-%d %H:%M')
2629 if info_dict.get('is_live') and formats:
2630 formats = [f for f in formats if bool(f.get('is_from_start')) == get_from_start]
2631 if get_from_start and not formats:
2632 self.raise_no_formats(info_dict, msg=(
2633 '--live-from-start is passed, but there are no formats that can be downloaded from the start. '
2634 'If you want to download from the current time, use --no-live-from-start'))
2635
2636 def is_wellformed(f):
2637 url = f.get('url')
2638 if not url:
2639 self.report_warning(
2640 '"url" field is missing or empty - skipping format, '
2641 'there is an error in extractor')
2642 return False
2643 if isinstance(url, bytes):
2644 sanitize_string_field(f, 'url')
2645 return True
2646
2647 # Filter out malformed formats for better extraction robustness
2648 formats = list(filter(is_wellformed, formats or []))
2649
2650 if not formats:
2651 self.raise_no_formats(info_dict)
2652
2653 for format in formats:
2654 sanitize_string_field(format, 'format_id')
2655 sanitize_numeric_fields(format)
2656 format['url'] = sanitize_url(format['url'])
2657 if format.get('ext') is None:
2658 format['ext'] = determine_ext(format['url']).lower()
2659 if format.get('protocol') is None:
2660 format['protocol'] = determine_protocol(format)
2661 if format.get('resolution') is None:
2662 format['resolution'] = self.format_resolution(format, default=None)
2663 if format.get('dynamic_range') is None and format.get('vcodec') != 'none':
2664 format['dynamic_range'] = 'SDR'
2665 if format.get('aspect_ratio') is None:
2666 format['aspect_ratio'] = try_call(lambda: round(format['width'] / format['height'], 2))
2667 if (info_dict.get('duration') and format.get('tbr')
2668 and not format.get('filesize') and not format.get('filesize_approx')):
2669 format['filesize_approx'] = int(info_dict['duration'] * format['tbr'] * (1024 / 8))
2670 format['http_headers'] = self._calc_headers(collections.ChainMap(format, info_dict))
2671
2672 # This is copied to http_headers by the above _calc_headers and can now be removed
2673 if '__x_forwarded_for_ip' in info_dict:
2674 del info_dict['__x_forwarded_for_ip']
2675
2676 self.sort_formats({
2677 'formats': formats,
2678 '_format_sort_fields': info_dict.get('_format_sort_fields')
2679 })
2680
2681 # Sanitize and group by format_id
2682 formats_dict = {}
2683 for i, format in enumerate(formats):
2684 if not format.get('format_id'):
2685 format['format_id'] = str(i)
2686 else:
2687 # Sanitize format_id from characters used in format selector expression
2688 format['format_id'] = re.sub(r'[\s,/+\[\]()]', '_', format['format_id'])
2689 formats_dict.setdefault(format['format_id'], []).append(format)
2690
2691 # Make sure all formats have unique format_id
2692 common_exts = set(itertools.chain(*self._format_selection_exts.values()))
2693 for format_id, ambiguous_formats in formats_dict.items():
2694 ambigious_id = len(ambiguous_formats) > 1
2695 for i, format in enumerate(ambiguous_formats):
2696 if ambigious_id:
2697 format['format_id'] = '%s-%d' % (format_id, i)
2698 # Ensure there is no conflict between id and ext in format selection
2699 # See https://github.com/yt-dlp/yt-dlp/issues/1282
2700 if format['format_id'] != format['ext'] and format['format_id'] in common_exts:
2701 format['format_id'] = 'f%s' % format['format_id']
2702
2703 if format.get('format') is None:
2704 format['format'] = '{id} - {res}{note}'.format(
2705 id=format['format_id'],
2706 res=self.format_resolution(format),
2707 note=format_field(format, 'format_note', ' (%s)'),
2708 )
2709
2710 if self.params.get('check_formats') is True:
2711 formats = LazyList(self._check_formats(formats[::-1]), reverse=True)
2712
2713 if not formats or formats[0] is not info_dict:
2714 # only set the 'formats' fields if the original info_dict list them
2715 # otherwise we end up with a circular reference, the first (and unique)
2716 # element in the 'formats' field in info_dict is info_dict itself,
2717 # which can't be exported to json
2718 info_dict['formats'] = formats
2719
2720 info_dict, _ = self.pre_process(info_dict)
2721
2722 if self._match_entry(info_dict, incomplete=self._format_fields) is not None:
2723 return info_dict
2724
2725 self.post_extract(info_dict)
2726 info_dict, _ = self.pre_process(info_dict, 'after_filter')
2727
2728 # The pre-processors may have modified the formats
2729 formats = self._get_formats(info_dict)
2730
2731 list_only = self.params.get('simulate') == 'list_only'
2732 interactive_format_selection = not list_only and self.format_selector == '-'
2733 if self.params.get('list_thumbnails'):
2734 self.list_thumbnails(info_dict)
2735 if self.params.get('listsubtitles'):
2736 if 'automatic_captions' in info_dict:
2737 self.list_subtitles(
2738 info_dict['id'], automatic_captions, 'automatic captions')
2739 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
2740 if self.params.get('listformats') or interactive_format_selection:
2741 self.list_formats(info_dict)
2742 if list_only:
2743 # Without this printing, -F --print-json will not work
2744 self.__forced_printings(info_dict)
2745 return info_dict
2746
2747 format_selector = self.format_selector
2748 while True:
2749 if interactive_format_selection:
2750 req_format = input(self._format_screen('\nEnter format selector ', self.Styles.EMPHASIS)
2751 + '(Press ENTER for default, or Ctrl+C to quit)'
2752 + self._format_screen(': ', self.Styles.EMPHASIS))
2753 try:
2754 format_selector = self.build_format_selector(req_format) if req_format else None
2755 except SyntaxError as err:
2756 self.report_error(err, tb=False, is_error=False)
2757 continue
2758
2759 if format_selector is None:
2760 req_format = self._default_format_spec(info_dict, download=download)
2761 self.write_debug(f'Default format spec: {req_format}')
2762 format_selector = self.build_format_selector(req_format)
2763
2764 formats_to_download = list(format_selector({
2765 'formats': formats,
2766 'has_merged_format': any('none' not in (f.get('acodec'), f.get('vcodec')) for f in formats),
2767 'incomplete_formats': (
2768 # All formats are video-only or
2769 all(f.get('vcodec') != 'none' and f.get('acodec') == 'none' for f in formats)
2770 # all formats are audio-only
2771 or all(f.get('vcodec') == 'none' and f.get('acodec') != 'none' for f in formats)),
2772 }))
2773 if interactive_format_selection and not formats_to_download:
2774 self.report_error('Requested format is not available', tb=False, is_error=False)
2775 continue
2776 break
2777
2778 if not formats_to_download:
2779 if not self.params.get('ignore_no_formats_error'):
2780 raise ExtractorError(
2781 'Requested format is not available. Use --list-formats for a list of available formats',
2782 expected=True, video_id=info_dict['id'], ie=info_dict['extractor'])
2783 self.report_warning('Requested format is not available')
2784 # Process what we can, even without any available formats.
2785 formats_to_download = [{}]
2786
2787 requested_ranges = tuple(self.params.get('download_ranges', lambda *_: [{}])(info_dict, self))
2788 best_format, downloaded_formats = formats_to_download[-1], []
2789 if download:
2790 if best_format and requested_ranges:
2791 def to_screen(*msg):
2792 self.to_screen(f'[info] {info_dict["id"]}: {" ".join(", ".join(variadic(m)) for m in msg)}')
2793
2794 to_screen(f'Downloading {len(formats_to_download)} format(s):',
2795 (f['format_id'] for f in formats_to_download))
2796 if requested_ranges != ({}, ):
2797 to_screen(f'Downloading {len(requested_ranges)} time ranges:',
2798 (f'{c["start_time"]:.1f}-{c["end_time"]:.1f}' for c in requested_ranges))
2799 max_downloads_reached = False
2800
2801 for fmt, chapter in itertools.product(formats_to_download, requested_ranges):
2802 new_info = self._copy_infodict(info_dict)
2803 new_info.update(fmt)
2804 offset, duration = info_dict.get('section_start') or 0, info_dict.get('duration') or float('inf')
2805 end_time = offset + min(chapter.get('end_time', duration), duration)
2806 if chapter or offset:
2807 new_info.update({
2808 'section_start': offset + chapter.get('start_time', 0),
2809 # duration may not be accurate. So allow deviations <1sec
2810 'section_end': end_time if end_time <= offset + duration + 1 else None,
2811 'section_title': chapter.get('title'),
2812 'section_number': chapter.get('index'),
2813 })
2814 downloaded_formats.append(new_info)
2815 try:
2816 self.process_info(new_info)
2817 except MaxDownloadsReached:
2818 max_downloads_reached = True
2819 self._raise_pending_errors(new_info)
2820 # Remove copied info
2821 for key, val in tuple(new_info.items()):
2822 if info_dict.get(key) == val:
2823 new_info.pop(key)
2824 if max_downloads_reached:
2825 break
2826
2827 write_archive = {f.get('__write_download_archive', False) for f in downloaded_formats}
2828 assert write_archive.issubset({True, False, 'ignore'})
2829 if True in write_archive and False not in write_archive:
2830 self.record_download_archive(info_dict)
2831
2832 info_dict['requested_downloads'] = downloaded_formats
2833 info_dict = self.run_all_pps('after_video', info_dict)
2834 if max_downloads_reached:
2835 raise MaxDownloadsReached()
2836
2837 # We update the info dict with the selected best quality format (backwards compatibility)
2838 info_dict.update(best_format)
2839 return info_dict
2840
2841 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
2842 """Select the requested subtitles and their format"""
2843 available_subs, normal_sub_langs = {}, []
2844 if normal_subtitles and self.params.get('writesubtitles'):
2845 available_subs.update(normal_subtitles)
2846 normal_sub_langs = tuple(normal_subtitles.keys())
2847 if automatic_captions and self.params.get('writeautomaticsub'):
2848 for lang, cap_info in automatic_captions.items():
2849 if lang not in available_subs:
2850 available_subs[lang] = cap_info
2851
2852 if not available_subs or (
2853 not self.params.get('writesubtitles')
2854 and not self.params.get('writeautomaticsub')):
2855 return None
2856
2857 all_sub_langs = tuple(available_subs.keys())
2858 if self.params.get('allsubtitles', False):
2859 requested_langs = all_sub_langs
2860 elif self.params.get('subtitleslangs', False):
2861 try:
2862 requested_langs = orderedSet_from_options(
2863 self.params.get('subtitleslangs'), {'all': all_sub_langs}, use_regex=True)
2864 except re.error as e:
2865 raise ValueError(f'Wrong regex for subtitlelangs: {e.pattern}')
2866 else:
2867 requested_langs = LazyList(itertools.chain(
2868 ['en'] if 'en' in normal_sub_langs else [],
2869 filter(lambda f: f.startswith('en'), normal_sub_langs),
2870 ['en'] if 'en' in all_sub_langs else [],
2871 filter(lambda f: f.startswith('en'), all_sub_langs),
2872 normal_sub_langs, all_sub_langs,
2873 ))[:1]
2874 if requested_langs:
2875 self.to_screen(f'[info] {video_id}: Downloading subtitles: {", ".join(requested_langs)}')
2876
2877 formats_query = self.params.get('subtitlesformat', 'best')
2878 formats_preference = formats_query.split('/') if formats_query else []
2879 subs = {}
2880 for lang in requested_langs:
2881 formats = available_subs.get(lang)
2882 if formats is None:
2883 self.report_warning(f'{lang} subtitles not available for {video_id}')
2884 continue
2885 for ext in formats_preference:
2886 if ext == 'best':
2887 f = formats[-1]
2888 break
2889 matches = list(filter(lambda f: f['ext'] == ext, formats))
2890 if matches:
2891 f = matches[-1]
2892 break
2893 else:
2894 f = formats[-1]
2895 self.report_warning(
2896 'No subtitle format found matching "%s" for language %s, '
2897 'using %s' % (formats_query, lang, f['ext']))
2898 subs[lang] = f
2899 return subs
2900
2901 def _forceprint(self, key, info_dict):
2902 if info_dict is None:
2903 return
2904 info_copy = info_dict.copy()
2905 info_copy.setdefault('filename', self.prepare_filename(info_dict))
2906 if info_dict.get('requested_formats') is not None:
2907 # For RTMP URLs, also include the playpath
2908 info_copy['urls'] = '\n'.join(f['url'] + f.get('play_path', '') for f in info_dict['requested_formats'])
2909 elif info_dict.get('url'):
2910 info_copy['urls'] = info_dict['url'] + info_dict.get('play_path', '')
2911 info_copy['formats_table'] = self.render_formats_table(info_dict)
2912 info_copy['thumbnails_table'] = self.render_thumbnails_table(info_dict)
2913 info_copy['subtitles_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('subtitles'))
2914 info_copy['automatic_captions_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('automatic_captions'))
2915
2916 def format_tmpl(tmpl):
2917 mobj = re.fullmatch(r'([\w.:,]|-\d|(?P<dict>{([\w.:,]|-\d)+}))+=?', tmpl)
2918 if not mobj:
2919 return tmpl
2920
2921 fmt = '%({})s'
2922 if tmpl.startswith('{'):
2923 tmpl, fmt = f'.{tmpl}', '%({})j'
2924 if tmpl.endswith('='):
2925 tmpl, fmt = tmpl[:-1], '{0} = %({0})#j'
2926 return '\n'.join(map(fmt.format, [tmpl] if mobj.group('dict') else tmpl.split(',')))
2927
2928 for tmpl in self.params['forceprint'].get(key, []):
2929 self.to_stdout(self.evaluate_outtmpl(format_tmpl(tmpl), info_copy))
2930
2931 for tmpl, file_tmpl in self.params['print_to_file'].get(key, []):
2932 filename = self.prepare_filename(info_dict, outtmpl=file_tmpl)
2933 tmpl = format_tmpl(tmpl)
2934 self.to_screen(f'[info] Writing {tmpl!r} to: {filename}')
2935 if self._ensure_dir_exists(filename):
2936 with open(filename, 'a', encoding='utf-8', newline='') as f:
2937 f.write(self.evaluate_outtmpl(tmpl, info_copy) + os.linesep)
2938
2939 return info_copy
2940
2941 def __forced_printings(self, info_dict, filename=None, incomplete=True):
2942 if (self.params.get('forcejson')
2943 or self.params['forceprint'].get('video')
2944 or self.params['print_to_file'].get('video')):
2945 self.post_extract(info_dict)
2946 if filename:
2947 info_dict['filename'] = filename
2948 info_copy = self._forceprint('video', info_dict)
2949
2950 def print_field(field, actual_field=None, optional=False):
2951 if actual_field is None:
2952 actual_field = field
2953 if self.params.get(f'force{field}') and (
2954 info_copy.get(field) is not None or (not optional and not incomplete)):
2955 self.to_stdout(info_copy[actual_field])
2956
2957 print_field('title')
2958 print_field('id')
2959 print_field('url', 'urls')
2960 print_field('thumbnail', optional=True)
2961 print_field('description', optional=True)
2962 if filename:
2963 print_field('filename')
2964 if self.params.get('forceduration') and info_copy.get('duration') is not None:
2965 self.to_stdout(formatSeconds(info_copy['duration']))
2966 print_field('format')
2967
2968 if self.params.get('forcejson'):
2969 self.to_stdout(json.dumps(self.sanitize_info(info_dict)))
2970
2971 def dl(self, name, info, subtitle=False, test=False):
2972 if not info.get('url'):
2973 self.raise_no_formats(info, True)
2974
2975 if test:
2976 verbose = self.params.get('verbose')
2977 params = {
2978 'test': True,
2979 'quiet': self.params.get('quiet') or not verbose,
2980 'verbose': verbose,
2981 'noprogress': not verbose,
2982 'nopart': True,
2983 'skip_unavailable_fragments': False,
2984 'keep_fragments': False,
2985 'overwrites': True,
2986 '_no_ytdl_file': True,
2987 }
2988 else:
2989 params = self.params
2990 fd = get_suitable_downloader(info, params, to_stdout=(name == '-'))(self, params)
2991 if not test:
2992 for ph in self._progress_hooks:
2993 fd.add_progress_hook(ph)
2994 urls = '", "'.join(
2995 (f['url'].split(',')[0] + ',<data>' if f['url'].startswith('data:') else f['url'])
2996 for f in info.get('requested_formats', []) or [info])
2997 self.write_debug(f'Invoking {fd.FD_NAME} downloader on "{urls}"')
2998
2999 # Note: Ideally info should be a deep-copied so that hooks cannot modify it.
3000 # But it may contain objects that are not deep-copyable
3001 new_info = self._copy_infodict(info)
3002 if new_info.get('http_headers') is None:
3003 new_info['http_headers'] = self._calc_headers(new_info)
3004 return fd.download(name, new_info, subtitle)
3005
3006 def existing_file(self, filepaths, *, default_overwrite=True):
3007 existing_files = list(filter(os.path.exists, orderedSet(filepaths)))
3008 if existing_files and not self.params.get('overwrites', default_overwrite):
3009 return existing_files[0]
3010
3011 for file in existing_files:
3012 self.report_file_delete(file)
3013 os.remove(file)
3014 return None
3015
3016 def process_info(self, info_dict):
3017 """Process a single resolved IE result. (Modifies it in-place)"""
3018
3019 assert info_dict.get('_type', 'video') == 'video'
3020 original_infodict = info_dict
3021
3022 if 'format' not in info_dict and 'ext' in info_dict:
3023 info_dict['format'] = info_dict['ext']
3024
3025 if self._match_entry(info_dict) is not None:
3026 info_dict['__write_download_archive'] = 'ignore'
3027 return
3028
3029 # Does nothing under normal operation - for backward compatibility of process_info
3030 self.post_extract(info_dict)
3031
3032 def replace_info_dict(new_info):
3033 nonlocal info_dict
3034 if new_info == info_dict:
3035 return
3036 info_dict.clear()
3037 info_dict.update(new_info)
3038
3039 new_info, _ = self.pre_process(info_dict, 'video')
3040 replace_info_dict(new_info)
3041 self._num_downloads += 1
3042
3043 # info_dict['_filename'] needs to be set for backward compatibility
3044 info_dict['_filename'] = full_filename = self.prepare_filename(info_dict, warn=True)
3045 temp_filename = self.prepare_filename(info_dict, 'temp')
3046 files_to_move = {}
3047
3048 # Forced printings
3049 self.__forced_printings(info_dict, full_filename, incomplete=('format' not in info_dict))
3050
3051 def check_max_downloads():
3052 if self._num_downloads >= float(self.params.get('max_downloads') or 'inf'):
3053 raise MaxDownloadsReached()
3054
3055 if self.params.get('simulate'):
3056 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
3057 check_max_downloads()
3058 return
3059
3060 if full_filename is None:
3061 return
3062 if not self._ensure_dir_exists(encodeFilename(full_filename)):
3063 return
3064 if not self._ensure_dir_exists(encodeFilename(temp_filename)):
3065 return
3066
3067 if self._write_description('video', info_dict,
3068 self.prepare_filename(info_dict, 'description')) is None:
3069 return
3070
3071 sub_files = self._write_subtitles(info_dict, temp_filename)
3072 if sub_files is None:
3073 return
3074 files_to_move.update(dict(sub_files))
3075
3076 thumb_files = self._write_thumbnails(
3077 'video', info_dict, temp_filename, self.prepare_filename(info_dict, 'thumbnail'))
3078 if thumb_files is None:
3079 return
3080 files_to_move.update(dict(thumb_files))
3081
3082 infofn = self.prepare_filename(info_dict, 'infojson')
3083 _infojson_written = self._write_info_json('video', info_dict, infofn)
3084 if _infojson_written:
3085 info_dict['infojson_filename'] = infofn
3086 # For backward compatibility, even though it was a private field
3087 info_dict['__infojson_filename'] = infofn
3088 elif _infojson_written is None:
3089 return
3090
3091 # Note: Annotations are deprecated
3092 annofn = None
3093 if self.params.get('writeannotations', False):
3094 annofn = self.prepare_filename(info_dict, 'annotation')
3095 if annofn:
3096 if not self._ensure_dir_exists(encodeFilename(annofn)):
3097 return
3098 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(annofn)):
3099 self.to_screen('[info] Video annotations are already present')
3100 elif not info_dict.get('annotations'):
3101 self.report_warning('There are no annotations to write.')
3102 else:
3103 try:
3104 self.to_screen('[info] Writing video annotations to: ' + annofn)
3105 with open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
3106 annofile.write(info_dict['annotations'])
3107 except (KeyError, TypeError):
3108 self.report_warning('There are no annotations to write.')
3109 except OSError:
3110 self.report_error('Cannot write annotations file: ' + annofn)
3111 return
3112
3113 # Write internet shortcut files
3114 def _write_link_file(link_type):
3115 url = try_get(info_dict['webpage_url'], iri_to_uri)
3116 if not url:
3117 self.report_warning(
3118 f'Cannot write internet shortcut file because the actual URL of "{info_dict["webpage_url"]}" is unknown')
3119 return True
3120 linkfn = replace_extension(self.prepare_filename(info_dict, 'link'), link_type, info_dict.get('ext'))
3121 if not self._ensure_dir_exists(encodeFilename(linkfn)):
3122 return False
3123 if self.params.get('overwrites', True) and os.path.exists(encodeFilename(linkfn)):
3124 self.to_screen(f'[info] Internet shortcut (.{link_type}) is already present')
3125 return True
3126 try:
3127 self.to_screen(f'[info] Writing internet shortcut (.{link_type}) to: {linkfn}')
3128 with open(encodeFilename(to_high_limit_path(linkfn)), 'w', encoding='utf-8',
3129 newline='\r\n' if link_type == 'url' else '\n') as linkfile:
3130 template_vars = {'url': url}
3131 if link_type == 'desktop':
3132 template_vars['filename'] = linkfn[:-(len(link_type) + 1)]
3133 linkfile.write(LINK_TEMPLATES[link_type] % template_vars)
3134 except OSError:
3135 self.report_error(f'Cannot write internet shortcut {linkfn}')
3136 return False
3137 return True
3138
3139 write_links = {
3140 'url': self.params.get('writeurllink'),
3141 'webloc': self.params.get('writewebloclink'),
3142 'desktop': self.params.get('writedesktoplink'),
3143 }
3144 if self.params.get('writelink'):
3145 link_type = ('webloc' if sys.platform == 'darwin'
3146 else 'desktop' if sys.platform.startswith('linux')
3147 else 'url')
3148 write_links[link_type] = True
3149
3150 if any(should_write and not _write_link_file(link_type)
3151 for link_type, should_write in write_links.items()):
3152 return
3153
3154 new_info, files_to_move = self.pre_process(info_dict, 'before_dl', files_to_move)
3155 replace_info_dict(new_info)
3156
3157 if self.params.get('skip_download'):
3158 info_dict['filepath'] = temp_filename
3159 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
3160 info_dict['__files_to_move'] = files_to_move
3161 replace_info_dict(self.run_pp(MoveFilesAfterDownloadPP(self, False), info_dict))
3162 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
3163 else:
3164 # Download
3165 info_dict.setdefault('__postprocessors', [])
3166 try:
3167
3168 def existing_video_file(*filepaths):
3169 ext = info_dict.get('ext')
3170 converted = lambda file: replace_extension(file, self.params.get('final_ext') or ext, ext)
3171 file = self.existing_file(itertools.chain(*zip(map(converted, filepaths), filepaths)),
3172 default_overwrite=False)
3173 if file:
3174 info_dict['ext'] = os.path.splitext(file)[1][1:]
3175 return file
3176
3177 fd, success = None, True
3178 if info_dict.get('protocol') or info_dict.get('url'):
3179 fd = get_suitable_downloader(info_dict, self.params, to_stdout=temp_filename == '-')
3180 if fd is not FFmpegFD and 'no-direct-merge' not in self.params['compat_opts'] and (
3181 info_dict.get('section_start') or info_dict.get('section_end')):
3182 msg = ('This format cannot be partially downloaded' if FFmpegFD.available()
3183 else 'You have requested downloading the video partially, but ffmpeg is not installed')
3184 self.report_error(f'{msg}. Aborting')
3185 return
3186
3187 if info_dict.get('requested_formats') is not None:
3188 requested_formats = info_dict['requested_formats']
3189 old_ext = info_dict['ext']
3190 if self.params.get('merge_output_format') is None:
3191 if (info_dict['ext'] == 'webm'
3192 and info_dict.get('thumbnails')
3193 # check with type instead of pp_key, __name__, or isinstance
3194 # since we dont want any custom PPs to trigger this
3195 and any(type(pp) == EmbedThumbnailPP for pp in self._pps['post_process'])): # noqa: E721
3196 info_dict['ext'] = 'mkv'
3197 self.report_warning(
3198 'webm doesn\'t support embedding a thumbnail, mkv will be used')
3199 new_ext = info_dict['ext']
3200
3201 def correct_ext(filename, ext=new_ext):
3202 if filename == '-':
3203 return filename
3204 filename_real_ext = os.path.splitext(filename)[1][1:]
3205 filename_wo_ext = (
3206 os.path.splitext(filename)[0]
3207 if filename_real_ext in (old_ext, new_ext)
3208 else filename)
3209 return f'{filename_wo_ext}.{ext}'
3210
3211 # Ensure filename always has a correct extension for successful merge
3212 full_filename = correct_ext(full_filename)
3213 temp_filename = correct_ext(temp_filename)
3214 dl_filename = existing_video_file(full_filename, temp_filename)
3215 info_dict['__real_download'] = False
3216
3217 merger = FFmpegMergerPP(self)
3218 downloaded = []
3219 if dl_filename is not None:
3220 self.report_file_already_downloaded(dl_filename)
3221 elif fd:
3222 for f in requested_formats if fd != FFmpegFD else []:
3223 f['filepath'] = fname = prepend_extension(
3224 correct_ext(temp_filename, info_dict['ext']),
3225 'f%s' % f['format_id'], info_dict['ext'])
3226 downloaded.append(fname)
3227 info_dict['url'] = '\n'.join(f['url'] for f in requested_formats)
3228 success, real_download = self.dl(temp_filename, info_dict)
3229 info_dict['__real_download'] = real_download
3230 else:
3231 if self.params.get('allow_unplayable_formats'):
3232 self.report_warning(
3233 'You have requested merging of multiple formats '
3234 'while also allowing unplayable formats to be downloaded. '
3235 'The formats won\'t be merged to prevent data corruption.')
3236 elif not merger.available:
3237 msg = 'You have requested merging of multiple formats but ffmpeg is not installed'
3238 if not self.params.get('ignoreerrors'):
3239 self.report_error(f'{msg}. Aborting due to --abort-on-error')
3240 return
3241 self.report_warning(f'{msg}. The formats won\'t be merged')
3242
3243 if temp_filename == '-':
3244 reason = ('using a downloader other than ffmpeg' if FFmpegFD.can_merge_formats(info_dict, self.params)
3245 else 'but the formats are incompatible for simultaneous download' if merger.available
3246 else 'but ffmpeg is not installed')
3247 self.report_warning(
3248 f'You have requested downloading multiple formats to stdout {reason}. '
3249 'The formats will be streamed one after the other')
3250 fname = temp_filename
3251 for f in requested_formats:
3252 new_info = dict(info_dict)
3253 del new_info['requested_formats']
3254 new_info.update(f)
3255 if temp_filename != '-':
3256 fname = prepend_extension(
3257 correct_ext(temp_filename, new_info['ext']),
3258 'f%s' % f['format_id'], new_info['ext'])
3259 if not self._ensure_dir_exists(fname):
3260 return
3261 f['filepath'] = fname
3262 downloaded.append(fname)
3263 partial_success, real_download = self.dl(fname, new_info)
3264 info_dict['__real_download'] = info_dict['__real_download'] or real_download
3265 success = success and partial_success
3266
3267 if downloaded and merger.available and not self.params.get('allow_unplayable_formats'):
3268 info_dict['__postprocessors'].append(merger)
3269 info_dict['__files_to_merge'] = downloaded
3270 # Even if there were no downloads, it is being merged only now
3271 info_dict['__real_download'] = True
3272 else:
3273 for file in downloaded:
3274 files_to_move[file] = None
3275 else:
3276 # Just a single file
3277 dl_filename = existing_video_file(full_filename, temp_filename)
3278 if dl_filename is None or dl_filename == temp_filename:
3279 # dl_filename == temp_filename could mean that the file was partially downloaded with --no-part.
3280 # So we should try to resume the download
3281 success, real_download = self.dl(temp_filename, info_dict)
3282 info_dict['__real_download'] = real_download
3283 else:
3284 self.report_file_already_downloaded(dl_filename)
3285
3286 dl_filename = dl_filename or temp_filename
3287 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
3288
3289 except network_exceptions as err:
3290 self.report_error('unable to download video data: %s' % error_to_compat_str(err))
3291 return
3292 except OSError as err:
3293 raise UnavailableVideoError(err)
3294 except (ContentTooShortError, ) as err:
3295 self.report_error(f'content too short (expected {err.expected} bytes and served {err.downloaded})')
3296 return
3297
3298 self._raise_pending_errors(info_dict)
3299 if success and full_filename != '-':
3300
3301 def fixup():
3302 do_fixup = True
3303 fixup_policy = self.params.get('fixup')
3304 vid = info_dict['id']
3305
3306 if fixup_policy in ('ignore', 'never'):
3307 return
3308 elif fixup_policy == 'warn':
3309 do_fixup = 'warn'
3310 elif fixup_policy != 'force':
3311 assert fixup_policy in ('detect_or_warn', None)
3312 if not info_dict.get('__real_download'):
3313 do_fixup = False
3314
3315 def ffmpeg_fixup(cndn, msg, cls):
3316 if not (do_fixup and cndn):
3317 return
3318 elif do_fixup == 'warn':
3319 self.report_warning(f'{vid}: {msg}')
3320 return
3321 pp = cls(self)
3322 if pp.available:
3323 info_dict['__postprocessors'].append(pp)
3324 else:
3325 self.report_warning(f'{vid}: {msg}. Install ffmpeg to fix this automatically')
3326
3327 stretched_ratio = info_dict.get('stretched_ratio')
3328 ffmpeg_fixup(stretched_ratio not in (1, None),
3329 f'Non-uniform pixel ratio {stretched_ratio}',
3330 FFmpegFixupStretchedPP)
3331
3332 downloader = get_suitable_downloader(info_dict, self.params) if 'protocol' in info_dict else None
3333 downloader = downloader.FD_NAME if downloader else None
3334
3335 ext = info_dict.get('ext')
3336 postprocessed_by_ffmpeg = info_dict.get('requested_formats') or any((
3337 isinstance(pp, FFmpegVideoConvertorPP)
3338 and resolve_recode_mapping(ext, pp.mapping)[0] not in (ext, None)
3339 ) for pp in self._pps['post_process'])
3340
3341 if not postprocessed_by_ffmpeg:
3342 ffmpeg_fixup(ext == 'm4a' and info_dict.get('container') == 'm4a_dash',
3343 'writing DASH m4a. Only some players support this container',
3344 FFmpegFixupM4aPP)
3345 ffmpeg_fixup(downloader == 'hlsnative' and not self.params.get('hls_use_mpegts')
3346 or info_dict.get('is_live') and self.params.get('hls_use_mpegts') is None,
3347 'Possible MPEG-TS in MP4 container or malformed AAC timestamps',
3348 FFmpegFixupM3u8PP)
3349 ffmpeg_fixup(info_dict.get('is_live') and downloader == 'dashsegments',
3350 'Possible duplicate MOOV atoms', FFmpegFixupDuplicateMoovPP)
3351
3352 ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed timestamps detected', FFmpegFixupTimestampPP)
3353 ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed duration detected', FFmpegFixupDurationPP)
3354
3355 fixup()
3356 try:
3357 replace_info_dict(self.post_process(dl_filename, info_dict, files_to_move))
3358 except PostProcessingError as err:
3359 self.report_error('Postprocessing: %s' % str(err))
3360 return
3361 try:
3362 for ph in self._post_hooks:
3363 ph(info_dict['filepath'])
3364 except Exception as err:
3365 self.report_error('post hooks: %s' % str(err))
3366 return
3367 info_dict['__write_download_archive'] = True
3368
3369 assert info_dict is original_infodict # Make sure the info_dict was modified in-place
3370 if self.params.get('force_write_download_archive'):
3371 info_dict['__write_download_archive'] = True
3372 check_max_downloads()
3373
3374 def __download_wrapper(self, func):
3375 @functools.wraps(func)
3376 def wrapper(*args, **kwargs):
3377 try:
3378 res = func(*args, **kwargs)
3379 except UnavailableVideoError as e:
3380 self.report_error(e)
3381 except DownloadCancelled as e:
3382 self.to_screen(f'[info] {e}')
3383 if not self.params.get('break_per_url'):
3384 raise
3385 self._num_downloads = 0
3386 else:
3387 if self.params.get('dump_single_json', False):
3388 self.post_extract(res)
3389 self.to_stdout(json.dumps(self.sanitize_info(res)))
3390 return wrapper
3391
3392 def download(self, url_list):
3393 """Download a given list of URLs."""
3394 url_list = variadic(url_list) # Passing a single URL is a common mistake
3395 outtmpl = self.params['outtmpl']['default']
3396 if (len(url_list) > 1
3397 and outtmpl != '-'
3398 and '%' not in outtmpl
3399 and self.params.get('max_downloads') != 1):
3400 raise SameFileError(outtmpl)
3401
3402 for url in url_list:
3403 self.__download_wrapper(self.extract_info)(
3404 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
3405
3406 return self._download_retcode
3407
3408 def download_with_info_file(self, info_filename):
3409 with contextlib.closing(fileinput.FileInput(
3410 [info_filename], mode='r',
3411 openhook=fileinput.hook_encoded('utf-8'))) as f:
3412 # FileInput doesn't have a read method, we can't call json.load
3413 infos = [self.sanitize_info(info, self.params.get('clean_infojson', True))
3414 for info in variadic(json.loads('\n'.join(f)))]
3415 for info in infos:
3416 try:
3417 self.__download_wrapper(self.process_ie_result)(info, download=True)
3418 except (DownloadError, EntryNotInPlaylist, ReExtractInfo) as e:
3419 if not isinstance(e, EntryNotInPlaylist):
3420 self.to_stderr('\r')
3421 webpage_url = info.get('webpage_url')
3422 if webpage_url is None:
3423 raise
3424 self.report_warning(f'The info failed to download: {e}; trying with URL {webpage_url}')
3425 self.download([webpage_url])
3426 return self._download_retcode
3427
3428 @staticmethod
3429 def sanitize_info(info_dict, remove_private_keys=False):
3430 ''' Sanitize the infodict for converting to json '''
3431 if info_dict is None:
3432 return info_dict
3433 info_dict.setdefault('epoch', int(time.time()))
3434 info_dict.setdefault('_type', 'video')
3435 info_dict.setdefault('_version', {
3436 'version': __version__,
3437 'current_git_head': current_git_head(),
3438 'release_git_head': RELEASE_GIT_HEAD,
3439 'repository': REPOSITORY,
3440 })
3441
3442 if remove_private_keys:
3443 reject = lambda k, v: v is None or k.startswith('__') or k in {
3444 'requested_downloads', 'requested_formats', 'requested_subtitles', 'requested_entries',
3445 'entries', 'filepath', '_filename', 'filename', 'infojson_filename', 'original_url',
3446 'playlist_autonumber', '_format_sort_fields',
3447 }
3448 else:
3449 reject = lambda k, v: False
3450
3451 def filter_fn(obj):
3452 if isinstance(obj, dict):
3453 return {k: filter_fn(v) for k, v in obj.items() if not reject(k, v)}
3454 elif isinstance(obj, (list, tuple, set, LazyList)):
3455 return list(map(filter_fn, obj))
3456 elif obj is None or isinstance(obj, (str, int, float, bool)):
3457 return obj
3458 else:
3459 return repr(obj)
3460
3461 return filter_fn(info_dict)
3462
3463 @staticmethod
3464 def filter_requested_info(info_dict, actually_filter=True):
3465 ''' Alias of sanitize_info for backward compatibility '''
3466 return YoutubeDL.sanitize_info(info_dict, actually_filter)
3467
3468 def _delete_downloaded_files(self, *files_to_delete, info={}, msg=None):
3469 for filename in set(filter(None, files_to_delete)):
3470 if msg:
3471 self.to_screen(msg % filename)
3472 try:
3473 os.remove(filename)
3474 except OSError:
3475 self.report_warning(f'Unable to delete file {filename}')
3476 if filename in info.get('__files_to_move', []): # NB: Delete even if None
3477 del info['__files_to_move'][filename]
3478
3479 @staticmethod
3480 def post_extract(info_dict):
3481 def actual_post_extract(info_dict):
3482 if info_dict.get('_type') in ('playlist', 'multi_video'):
3483 for video_dict in info_dict.get('entries', {}):
3484 actual_post_extract(video_dict or {})
3485 return
3486
3487 post_extractor = info_dict.pop('__post_extractor', None) or (lambda: {})
3488 info_dict.update(post_extractor())
3489
3490 actual_post_extract(info_dict or {})
3491
3492 def run_pp(self, pp, infodict):
3493 files_to_delete = []
3494 if '__files_to_move' not in infodict:
3495 infodict['__files_to_move'] = {}
3496 try:
3497 files_to_delete, infodict = pp.run(infodict)
3498 except PostProcessingError as e:
3499 # Must be True and not 'only_download'
3500 if self.params.get('ignoreerrors') is True:
3501 self.report_error(e)
3502 return infodict
3503 raise
3504
3505 if not files_to_delete:
3506 return infodict
3507 if self.params.get('keepvideo', False):
3508 for f in files_to_delete:
3509 infodict['__files_to_move'].setdefault(f, '')
3510 else:
3511 self._delete_downloaded_files(
3512 *files_to_delete, info=infodict, msg='Deleting original file %s (pass -k to keep)')
3513 return infodict
3514
3515 def run_all_pps(self, key, info, *, additional_pps=None):
3516 if key != 'video':
3517 self._forceprint(key, info)
3518 for pp in (additional_pps or []) + self._pps[key]:
3519 info = self.run_pp(pp, info)
3520 return info
3521
3522 def pre_process(self, ie_info, key='pre_process', files_to_move=None):
3523 info = dict(ie_info)
3524 info['__files_to_move'] = files_to_move or {}
3525 try:
3526 info = self.run_all_pps(key, info)
3527 except PostProcessingError as err:
3528 msg = f'Preprocessing: {err}'
3529 info.setdefault('__pending_error', msg)
3530 self.report_error(msg, is_error=False)
3531 return info, info.pop('__files_to_move', None)
3532
3533 def post_process(self, filename, info, files_to_move=None):
3534 """Run all the postprocessors on the given file."""
3535 info['filepath'] = filename
3536 info['__files_to_move'] = files_to_move or {}
3537 info = self.run_all_pps('post_process', info, additional_pps=info.get('__postprocessors'))
3538 info = self.run_pp(MoveFilesAfterDownloadPP(self), info)
3539 del info['__files_to_move']
3540 return self.run_all_pps('after_move', info)
3541
3542 def _make_archive_id(self, info_dict):
3543 video_id = info_dict.get('id')
3544 if not video_id:
3545 return
3546 # Future-proof against any change in case
3547 # and backwards compatibility with prior versions
3548 extractor = info_dict.get('extractor_key') or info_dict.get('ie_key') # key in a playlist
3549 if extractor is None:
3550 url = str_or_none(info_dict.get('url'))
3551 if not url:
3552 return
3553 # Try to find matching extractor for the URL and take its ie_key
3554 for ie_key, ie in self._ies.items():
3555 if ie.suitable(url):
3556 extractor = ie_key
3557 break
3558 else:
3559 return
3560 return make_archive_id(extractor, video_id)
3561
3562 def in_download_archive(self, info_dict):
3563 if not self.archive:
3564 return False
3565
3566 vid_ids = [self._make_archive_id(info_dict)]
3567 vid_ids.extend(info_dict.get('_old_archive_ids') or [])
3568 return any(id_ in self.archive for id_ in vid_ids)
3569
3570 def record_download_archive(self, info_dict):
3571 fn = self.params.get('download_archive')
3572 if fn is None:
3573 return
3574 vid_id = self._make_archive_id(info_dict)
3575 assert vid_id
3576
3577 self.write_debug(f'Adding to archive: {vid_id}')
3578 if is_path_like(fn):
3579 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
3580 archive_file.write(vid_id + '\n')
3581 self.archive.add(vid_id)
3582
3583 @staticmethod
3584 def format_resolution(format, default='unknown'):
3585 if format.get('vcodec') == 'none' and format.get('acodec') != 'none':
3586 return 'audio only'
3587 if format.get('resolution') is not None:
3588 return format['resolution']
3589 if format.get('width') and format.get('height'):
3590 return '%dx%d' % (format['width'], format['height'])
3591 elif format.get('height'):
3592 return '%sp' % format['height']
3593 elif format.get('width'):
3594 return '%dx?' % format['width']
3595 return default
3596
3597 def _list_format_headers(self, *headers):
3598 if self.params.get('listformats_table', True) is not False:
3599 return [self._format_out(header, self.Styles.HEADERS) for header in headers]
3600 return headers
3601
3602 def _format_note(self, fdict):
3603 res = ''
3604 if fdict.get('ext') in ['f4f', 'f4m']:
3605 res += '(unsupported)'
3606 if fdict.get('language'):
3607 if res:
3608 res += ' '
3609 res += '[%s]' % fdict['language']
3610 if fdict.get('format_note') is not None:
3611 if res:
3612 res += ' '
3613 res += fdict['format_note']
3614 if fdict.get('tbr') is not None:
3615 if res:
3616 res += ', '
3617 res += '%4dk' % fdict['tbr']
3618 if fdict.get('container') is not None:
3619 if res:
3620 res += ', '
3621 res += '%s container' % fdict['container']
3622 if (fdict.get('vcodec') is not None
3623 and fdict.get('vcodec') != 'none'):
3624 if res:
3625 res += ', '
3626 res += fdict['vcodec']
3627 if fdict.get('vbr') is not None:
3628 res += '@'
3629 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
3630 res += 'video@'
3631 if fdict.get('vbr') is not None:
3632 res += '%4dk' % fdict['vbr']
3633 if fdict.get('fps') is not None:
3634 if res:
3635 res += ', '
3636 res += '%sfps' % fdict['fps']
3637 if fdict.get('acodec') is not None:
3638 if res:
3639 res += ', '
3640 if fdict['acodec'] == 'none':
3641 res += 'video only'
3642 else:
3643 res += '%-5s' % fdict['acodec']
3644 elif fdict.get('abr') is not None:
3645 if res:
3646 res += ', '
3647 res += 'audio'
3648 if fdict.get('abr') is not None:
3649 res += '@%3dk' % fdict['abr']
3650 if fdict.get('asr') is not None:
3651 res += ' (%5dHz)' % fdict['asr']
3652 if fdict.get('filesize') is not None:
3653 if res:
3654 res += ', '
3655 res += format_bytes(fdict['filesize'])
3656 elif fdict.get('filesize_approx') is not None:
3657 if res:
3658 res += ', '
3659 res += '~' + format_bytes(fdict['filesize_approx'])
3660 return res
3661
3662 def _get_formats(self, info_dict):
3663 if info_dict.get('formats') is None:
3664 if info_dict.get('url') and info_dict.get('_type', 'video') == 'video':
3665 return [info_dict]
3666 return []
3667 return info_dict['formats']
3668
3669 def render_formats_table(self, info_dict):
3670 formats = self._get_formats(info_dict)
3671 if not formats:
3672 return
3673 if not self.params.get('listformats_table', True) is not False:
3674 table = [
3675 [
3676 format_field(f, 'format_id'),
3677 format_field(f, 'ext'),
3678 self.format_resolution(f),
3679 self._format_note(f)
3680 ] for f in formats if (f.get('preference') or 0) >= -1000]
3681 return render_table(['format code', 'extension', 'resolution', 'note'], table, extra_gap=1)
3682
3683 def simplified_codec(f, field):
3684 assert field in ('acodec', 'vcodec')
3685 codec = f.get(field, 'unknown')
3686 if not codec:
3687 return 'unknown'
3688 elif codec != 'none':
3689 return '.'.join(codec.split('.')[:4])
3690
3691 if field == 'vcodec' and f.get('acodec') == 'none':
3692 return 'images'
3693 elif field == 'acodec' and f.get('vcodec') == 'none':
3694 return ''
3695 return self._format_out('audio only' if field == 'vcodec' else 'video only',
3696 self.Styles.SUPPRESS)
3697
3698 delim = self._format_out('\u2502', self.Styles.DELIM, '|', test_encoding=True)
3699 table = [
3700 [
3701 self._format_out(format_field(f, 'format_id'), self.Styles.ID),
3702 format_field(f, 'ext'),
3703 format_field(f, func=self.format_resolution, ignore=('audio only', 'images')),
3704 format_field(f, 'fps', '\t%d', func=round),
3705 format_field(f, 'dynamic_range', '%s', ignore=(None, 'SDR')).replace('HDR', ''),
3706 format_field(f, 'audio_channels', '\t%s'),
3707 delim,
3708 format_field(f, 'filesize', ' \t%s', func=format_bytes) + format_field(f, 'filesize_approx', '~\t%s', func=format_bytes),
3709 format_field(f, 'tbr', '\t%dk', func=round),
3710 shorten_protocol_name(f.get('protocol', '')),
3711 delim,
3712 simplified_codec(f, 'vcodec'),
3713 format_field(f, 'vbr', '\t%dk', func=round),
3714 simplified_codec(f, 'acodec'),
3715 format_field(f, 'abr', '\t%dk', func=round),
3716 format_field(f, 'asr', '\t%s', func=format_decimal_suffix),
3717 join_nonempty(
3718 self._format_out('UNSUPPORTED', 'light red') if f.get('ext') in ('f4f', 'f4m') else None,
3719 self._format_out('DRM', 'light red') if f.get('has_drm') else None,
3720 format_field(f, 'language', '[%s]'),
3721 join_nonempty(format_field(f, 'format_note'),
3722 format_field(f, 'container', ignore=(None, f.get('ext'))),
3723 delim=', '),
3724 delim=' '),
3725 ] for f in formats if f.get('preference') is None or f['preference'] >= -1000]
3726 header_line = self._list_format_headers(
3727 'ID', 'EXT', 'RESOLUTION', '\tFPS', 'HDR', 'CH', delim, '\tFILESIZE', '\tTBR', 'PROTO',
3728 delim, 'VCODEC', '\tVBR', 'ACODEC', '\tABR', '\tASR', 'MORE INFO')
3729
3730 return render_table(
3731 header_line, table, hide_empty=True,
3732 delim=self._format_out('\u2500', self.Styles.DELIM, '-', test_encoding=True))
3733
3734 def render_thumbnails_table(self, info_dict):
3735 thumbnails = list(info_dict.get('thumbnails') or [])
3736 if not thumbnails:
3737 return None
3738 return render_table(
3739 self._list_format_headers('ID', 'Width', 'Height', 'URL'),
3740 [[t.get('id'), t.get('width') or 'unknown', t.get('height') or 'unknown', t['url']] for t in thumbnails])
3741
3742 def render_subtitles_table(self, video_id, subtitles):
3743 def _row(lang, formats):
3744 exts, names = zip(*((f['ext'], f.get('name') or 'unknown') for f in reversed(formats)))
3745 if len(set(names)) == 1:
3746 names = [] if names[0] == 'unknown' else names[:1]
3747 return [lang, ', '.join(names), ', '.join(exts)]
3748
3749 if not subtitles:
3750 return None
3751 return render_table(
3752 self._list_format_headers('Language', 'Name', 'Formats'),
3753 [_row(lang, formats) for lang, formats in subtitles.items()],
3754 hide_empty=True)
3755
3756 def __list_table(self, video_id, name, func, *args):
3757 table = func(*args)
3758 if not table:
3759 self.to_screen(f'{video_id} has no {name}')
3760 return
3761 self.to_screen(f'[info] Available {name} for {video_id}:')
3762 self.to_stdout(table)
3763
3764 def list_formats(self, info_dict):
3765 self.__list_table(info_dict['id'], 'formats', self.render_formats_table, info_dict)
3766
3767 def list_thumbnails(self, info_dict):
3768 self.__list_table(info_dict['id'], 'thumbnails', self.render_thumbnails_table, info_dict)
3769
3770 def list_subtitles(self, video_id, subtitles, name='subtitles'):
3771 self.__list_table(video_id, name, self.render_subtitles_table, video_id, subtitles)
3772
3773 def urlopen(self, req):
3774 """ Start an HTTP download """
3775 if isinstance(req, str):
3776 req = sanitized_Request(req)
3777 return self._opener.open(req, timeout=self._socket_timeout)
3778
3779 def print_debug_header(self):
3780 if not self.params.get('verbose'):
3781 return
3782
3783 from . import _IN_CLI # Must be delayed import
3784
3785 # These imports can be slow. So import them only as needed
3786 from .extractor.extractors import _LAZY_LOADER
3787 from .extractor.extractors import (
3788 _PLUGIN_CLASSES as plugin_ies,
3789 _PLUGIN_OVERRIDES as plugin_ie_overrides
3790 )
3791
3792 def get_encoding(stream):
3793 ret = str(getattr(stream, 'encoding', 'missing (%s)' % type(stream).__name__))
3794 additional_info = []
3795 if os.environ.get('TERM', '').lower() == 'dumb':
3796 additional_info.append('dumb')
3797 if not supports_terminal_sequences(stream):
3798 from .utils import WINDOWS_VT_MODE # Must be imported locally
3799 additional_info.append('No VT' if WINDOWS_VT_MODE is False else 'No ANSI')
3800 if additional_info:
3801 ret = f'{ret} ({",".join(additional_info)})'
3802 return ret
3803
3804 encoding_str = 'Encodings: locale %s, fs %s, pref %s, %s' % (
3805 locale.getpreferredencoding(),
3806 sys.getfilesystemencoding(),
3807 self.get_encoding(),
3808 ', '.join(
3809 f'{key} {get_encoding(stream)}' for key, stream in self._out_files.items_
3810 if stream is not None and key != 'console')
3811 )
3812
3813 logger = self.params.get('logger')
3814 if logger:
3815 write_debug = lambda msg: logger.debug(f'[debug] {msg}')
3816 write_debug(encoding_str)
3817 else:
3818 write_string(f'[debug] {encoding_str}\n', encoding=None)
3819 write_debug = lambda msg: self._write_string(f'[debug] {msg}\n')
3820
3821 source = detect_variant()
3822 if VARIANT not in (None, 'pip'):
3823 source += '*'
3824 klass = type(self)
3825 write_debug(join_nonempty(
3826 f'{"yt-dlp" if REPOSITORY == "yt-dlp/yt-dlp" else REPOSITORY} version',
3827 f'{CHANNEL}@{__version__}',
3828 f'[{RELEASE_GIT_HEAD[:9]}]' if RELEASE_GIT_HEAD else '',
3829 '' if source == 'unknown' else f'({source})',
3830 '' if _IN_CLI else 'API' if klass == YoutubeDL else f'API:{self.__module__}.{klass.__qualname__}',
3831 delim=' '))
3832
3833 if not _IN_CLI:
3834 write_debug(f'params: {self.params}')
3835
3836 if not _LAZY_LOADER:
3837 if os.environ.get('YTDLP_NO_LAZY_EXTRACTORS'):
3838 write_debug('Lazy loading extractors is forcibly disabled')
3839 else:
3840 write_debug('Lazy loading extractors is disabled')
3841 if self.params['compat_opts']:
3842 write_debug('Compatibility options: %s' % ', '.join(self.params['compat_opts']))
3843
3844 if current_git_head():
3845 write_debug(f'Git HEAD: {current_git_head()}')
3846 write_debug(system_identifier())
3847
3848 exe_versions, ffmpeg_features = FFmpegPostProcessor.get_versions_and_features(self)
3849 ffmpeg_features = {key for key, val in ffmpeg_features.items() if val}
3850 if ffmpeg_features:
3851 exe_versions['ffmpeg'] += ' (%s)' % ','.join(sorted(ffmpeg_features))
3852
3853 exe_versions['rtmpdump'] = rtmpdump_version()
3854 exe_versions['phantomjs'] = PhantomJSwrapper._version()
3855 exe_str = ', '.join(
3856 f'{exe} {v}' for exe, v in sorted(exe_versions.items()) if v
3857 ) or 'none'
3858 write_debug('exe versions: %s' % exe_str)
3859
3860 from .compat.compat_utils import get_package_info
3861 from .dependencies import available_dependencies
3862
3863 write_debug('Optional libraries: %s' % (', '.join(sorted({
3864 join_nonempty(*get_package_info(m)) for m in available_dependencies.values()
3865 })) or 'none'))
3866
3867 self._setup_opener()
3868 proxy_map = {}
3869 for handler in self._opener.handlers:
3870 if hasattr(handler, 'proxies'):
3871 proxy_map.update(handler.proxies)
3872 write_debug(f'Proxy map: {proxy_map}')
3873
3874 for plugin_type, plugins in {'Extractor': plugin_ies, 'Post-Processor': plugin_pps}.items():
3875 display_list = ['%s%s' % (
3876 klass.__name__, '' if klass.__name__ == name else f' as {name}')
3877 for name, klass in plugins.items()]
3878 if plugin_type == 'Extractor':
3879 display_list.extend(f'{plugins[-1].IE_NAME.partition("+")[2]} ({parent.__name__})'
3880 for parent, plugins in plugin_ie_overrides.items())
3881 if not display_list:
3882 continue
3883 write_debug(f'{plugin_type} Plugins: {", ".join(sorted(display_list))}')
3884
3885 plugin_dirs = plugin_directories()
3886 if plugin_dirs:
3887 write_debug(f'Plugin directories: {plugin_dirs}')
3888
3889 # Not implemented
3890 if False and self.params.get('call_home'):
3891 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode()
3892 write_debug('Public IP address: %s' % ipaddr)
3893 latest_version = self.urlopen(
3894 'https://yt-dl.org/latest/version').read().decode()
3895 if version_tuple(latest_version) > version_tuple(__version__):
3896 self.report_warning(
3897 'You are using an outdated version (newest version: %s)! '
3898 'See https://yt-dl.org/update if you need help updating.' %
3899 latest_version)
3900
3901 def _setup_opener(self):
3902 if hasattr(self, '_opener'):
3903 return
3904 timeout_val = self.params.get('socket_timeout')
3905 self._socket_timeout = 20 if timeout_val is None else float(timeout_val)
3906
3907 opts_cookiesfrombrowser = self.params.get('cookiesfrombrowser')
3908 opts_cookiefile = self.params.get('cookiefile')
3909 opts_proxy = self.params.get('proxy')
3910
3911 self.cookiejar = load_cookies(opts_cookiefile, opts_cookiesfrombrowser, self)
3912
3913 cookie_processor = YoutubeDLCookieProcessor(self.cookiejar)
3914 if opts_proxy is not None:
3915 if opts_proxy == '':
3916 proxies = {}
3917 else:
3918 proxies = {'http': opts_proxy, 'https': opts_proxy}
3919 else:
3920 proxies = urllib.request.getproxies()
3921 # Set HTTPS proxy to HTTP one if given (https://github.com/ytdl-org/youtube-dl/issues/805)
3922 if 'http' in proxies and 'https' not in proxies:
3923 proxies['https'] = proxies['http']
3924 proxy_handler = PerRequestProxyHandler(proxies)
3925
3926 debuglevel = 1 if self.params.get('debug_printtraffic') else 0
3927 https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
3928 ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
3929 redirect_handler = YoutubeDLRedirectHandler()
3930 data_handler = urllib.request.DataHandler()
3931
3932 # When passing our own FileHandler instance, build_opener won't add the
3933 # default FileHandler and allows us to disable the file protocol, which
3934 # can be used for malicious purposes (see
3935 # https://github.com/ytdl-org/youtube-dl/issues/8227)
3936 file_handler = urllib.request.FileHandler()
3937
3938 if not self.params.get('enable_file_urls'):
3939 def file_open(*args, **kwargs):
3940 raise urllib.error.URLError(
3941 'file:// URLs are explicitly disabled in yt-dlp for security reasons. '
3942 'Use --enable-file-urls to enable at your own risk.')
3943 file_handler.file_open = file_open
3944
3945 opener = urllib.request.build_opener(
3946 proxy_handler, https_handler, cookie_processor, ydlh, redirect_handler, data_handler, file_handler)
3947
3948 # Delete the default user-agent header, which would otherwise apply in
3949 # cases where our custom HTTP handler doesn't come into play
3950 # (See https://github.com/ytdl-org/youtube-dl/issues/1309 for details)
3951 opener.addheaders = []
3952 self._opener = opener
3953
3954 def encode(self, s):
3955 if isinstance(s, bytes):
3956 return s # Already encoded
3957
3958 try:
3959 return s.encode(self.get_encoding())
3960 except UnicodeEncodeError as err:
3961 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
3962 raise
3963
3964 def get_encoding(self):
3965 encoding = self.params.get('encoding')
3966 if encoding is None:
3967 encoding = preferredencoding()
3968 return encoding
3969
3970 def _write_info_json(self, label, ie_result, infofn, overwrite=None):
3971 ''' Write infojson and returns True = written, 'exists' = Already exists, False = skip, None = error '''
3972 if overwrite is None:
3973 overwrite = self.params.get('overwrites', True)
3974 if not self.params.get('writeinfojson'):
3975 return False
3976 elif not infofn:
3977 self.write_debug(f'Skipping writing {label} infojson')
3978 return False
3979 elif not self._ensure_dir_exists(infofn):
3980 return None
3981 elif not overwrite and os.path.exists(infofn):
3982 self.to_screen(f'[info] {label.title()} metadata is already present')
3983 return 'exists'
3984
3985 self.to_screen(f'[info] Writing {label} metadata as JSON to: {infofn}')
3986 try:
3987 write_json_file(self.sanitize_info(ie_result, self.params.get('clean_infojson', True)), infofn)
3988 return True
3989 except OSError:
3990 self.report_error(f'Cannot write {label} metadata to JSON file {infofn}')
3991 return None
3992
3993 def _write_description(self, label, ie_result, descfn):
3994 ''' Write description and returns True = written, False = skip, None = error '''
3995 if not self.params.get('writedescription'):
3996 return False
3997 elif not descfn:
3998 self.write_debug(f'Skipping writing {label} description')
3999 return False
4000 elif not self._ensure_dir_exists(descfn):
4001 return None
4002 elif not self.params.get('overwrites', True) and os.path.exists(descfn):
4003 self.to_screen(f'[info] {label.title()} description is already present')
4004 elif ie_result.get('description') is None:
4005 self.to_screen(f'[info] There\'s no {label} description to write')
4006 return False
4007 else:
4008 try:
4009 self.to_screen(f'[info] Writing {label} description to: {descfn}')
4010 with open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
4011 descfile.write(ie_result['description'])
4012 except OSError:
4013 self.report_error(f'Cannot write {label} description file {descfn}')
4014 return None
4015 return True
4016
4017 def _write_subtitles(self, info_dict, filename):
4018 ''' Write subtitles to file and return list of (sub_filename, final_sub_filename); or None if error'''
4019 ret = []
4020 subtitles = info_dict.get('requested_subtitles')
4021 if not (self.params.get('writesubtitles') or self.params.get('writeautomaticsub')):
4022 # subtitles download errors are already managed as troubles in relevant IE
4023 # that way it will silently go on when used with unsupporting IE
4024 return ret
4025 elif not subtitles:
4026 self.to_screen('[info] There are no subtitles for the requested languages')
4027 return ret
4028 sub_filename_base = self.prepare_filename(info_dict, 'subtitle')
4029 if not sub_filename_base:
4030 self.to_screen('[info] Skipping writing video subtitles')
4031 return ret
4032
4033 for sub_lang, sub_info in subtitles.items():
4034 sub_format = sub_info['ext']
4035 sub_filename = subtitles_filename(filename, sub_lang, sub_format, info_dict.get('ext'))
4036 sub_filename_final = subtitles_filename(sub_filename_base, sub_lang, sub_format, info_dict.get('ext'))
4037 existing_sub = self.existing_file((sub_filename_final, sub_filename))
4038 if existing_sub:
4039 self.to_screen(f'[info] Video subtitle {sub_lang}.{sub_format} is already present')
4040 sub_info['filepath'] = existing_sub
4041 ret.append((existing_sub, sub_filename_final))
4042 continue
4043
4044 self.to_screen(f'[info] Writing video subtitles to: {sub_filename}')
4045 if sub_info.get('data') is not None:
4046 try:
4047 # Use newline='' to prevent conversion of newline characters
4048 # See https://github.com/ytdl-org/youtube-dl/issues/10268
4049 with open(sub_filename, 'w', encoding='utf-8', newline='') as subfile:
4050 subfile.write(sub_info['data'])
4051 sub_info['filepath'] = sub_filename
4052 ret.append((sub_filename, sub_filename_final))
4053 continue
4054 except OSError:
4055 self.report_error(f'Cannot write video subtitles file {sub_filename}')
4056 return None
4057
4058 try:
4059 sub_copy = sub_info.copy()
4060 sub_copy.setdefault('http_headers', info_dict.get('http_headers'))
4061 self.dl(sub_filename, sub_copy, subtitle=True)
4062 sub_info['filepath'] = sub_filename
4063 ret.append((sub_filename, sub_filename_final))
4064 except (DownloadError, ExtractorError, IOError, OSError, ValueError) + network_exceptions as err:
4065 msg = f'Unable to download video subtitles for {sub_lang!r}: {err}'
4066 if self.params.get('ignoreerrors') is not True: # False or 'only_download'
4067 if not self.params.get('ignoreerrors'):
4068 self.report_error(msg)
4069 raise DownloadError(msg)
4070 self.report_warning(msg)
4071 return ret
4072
4073 def _write_thumbnails(self, label, info_dict, filename, thumb_filename_base=None):
4074 ''' Write thumbnails to file and return list of (thumb_filename, final_thumb_filename) '''
4075 write_all = self.params.get('write_all_thumbnails', False)
4076 thumbnails, ret = [], []
4077 if write_all or self.params.get('writethumbnail', False):
4078 thumbnails = info_dict.get('thumbnails') or []
4079 if not thumbnails:
4080 self.to_screen(f'[info] There are no {label} thumbnails to download')
4081 return ret
4082 multiple = write_all and len(thumbnails) > 1
4083
4084 if thumb_filename_base is None:
4085 thumb_filename_base = filename
4086 if thumbnails and not thumb_filename_base:
4087 self.write_debug(f'Skipping writing {label} thumbnail')
4088 return ret
4089
4090 for idx, t in list(enumerate(thumbnails))[::-1]:
4091 thumb_ext = (f'{t["id"]}.' if multiple else '') + determine_ext(t['url'], 'jpg')
4092 thumb_display_id = f'{label} thumbnail {t["id"]}'
4093 thumb_filename = replace_extension(filename, thumb_ext, info_dict.get('ext'))
4094 thumb_filename_final = replace_extension(thumb_filename_base, thumb_ext, info_dict.get('ext'))
4095
4096 existing_thumb = self.existing_file((thumb_filename_final, thumb_filename))
4097 if existing_thumb:
4098 self.to_screen('[info] %s is already present' % (
4099 thumb_display_id if multiple else f'{label} thumbnail').capitalize())
4100 t['filepath'] = existing_thumb
4101 ret.append((existing_thumb, thumb_filename_final))
4102 else:
4103 self.to_screen(f'[info] Downloading {thumb_display_id} ...')
4104 try:
4105 uf = self.urlopen(sanitized_Request(t['url'], headers=t.get('http_headers', {})))
4106 self.to_screen(f'[info] Writing {thumb_display_id} to: {thumb_filename}')
4107 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
4108 shutil.copyfileobj(uf, thumbf)
4109 ret.append((thumb_filename, thumb_filename_final))
4110 t['filepath'] = thumb_filename
4111 except network_exceptions as err:
4112 thumbnails.pop(idx)
4113 self.report_warning(f'Unable to download {thumb_display_id}: {err}')
4114 if ret and not write_all:
4115 break
4116 return ret