]> jfr.im git - yt-dlp.git/blob - yt_dlp/YoutubeDL.py
[cleanup] Misc
[yt-dlp.git] / yt_dlp / YoutubeDL.py
1 import collections
2 import contextlib
3 import datetime
4 import errno
5 import fileinput
6 import functools
7 import io
8 import itertools
9 import json
10 import locale
11 import operator
12 import os
13 import random
14 import re
15 import shutil
16 import string
17 import subprocess
18 import sys
19 import tempfile
20 import time
21 import tokenize
22 import traceback
23 import unicodedata
24
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 internal metadata 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 value, replacement, default, last_field = None, None, na, ''
1275 while mobj:
1276 mobj = mobj.groupdict()
1277 default = mobj['default'] if mobj['default'] is not None else default
1278 value = get_value(mobj)
1279 last_field, replacement = mobj['fields'], mobj['replacement']
1280 if value is None and mobj['alternate']:
1281 mobj = re.match(INTERNAL_FORMAT_RE, mobj['remaining'][1:])
1282 else:
1283 break
1284
1285 fmt = outer_mobj.group('format')
1286 if fmt == 's' and value is not None and last_field in field_size_compat_map.keys():
1287 fmt = f'0{field_size_compat_map[last_field]:d}d'
1288
1289 if None not in (value, replacement):
1290 try:
1291 value = replacement_formatter.format(replacement, value)
1292 except ValueError:
1293 value, default = None, na
1294
1295 flags = outer_mobj.group('conversion') or ''
1296 str_fmt = f'{fmt[:-1]}s'
1297 if value is None:
1298 value, fmt = default, 's'
1299 elif fmt[-1] == 'l': # list
1300 delim = '\n' if '#' in flags else ', '
1301 value, fmt = delim.join(map(str, variadic(value, allowed_types=(str, bytes)))), str_fmt
1302 elif fmt[-1] == 'j': # json
1303 value, fmt = json.dumps(
1304 value, default=_dumpjson_default,
1305 indent=4 if '#' in flags else None, ensure_ascii='+' not in flags), str_fmt
1306 elif fmt[-1] == 'h': # html
1307 value, fmt = escapeHTML(str(value)), str_fmt
1308 elif fmt[-1] == 'q': # quoted
1309 value = map(str, variadic(value) if '#' in flags else [value])
1310 value, fmt = ' '.join(map(compat_shlex_quote, value)), str_fmt
1311 elif fmt[-1] == 'B': # bytes
1312 value = f'%{str_fmt}'.encode() % str(value).encode()
1313 value, fmt = value.decode('utf-8', 'ignore'), 's'
1314 elif fmt[-1] == 'U': # unicode normalized
1315 value, fmt = unicodedata.normalize(
1316 # "+" = compatibility equivalence, "#" = NFD
1317 'NF%s%s' % ('K' if '+' in flags else '', 'D' if '#' in flags else 'C'),
1318 value), str_fmt
1319 elif fmt[-1] == 'D': # decimal suffix
1320 num_fmt, fmt = fmt[:-1].replace('#', ''), 's'
1321 value = format_decimal_suffix(value, f'%{num_fmt}f%s' if num_fmt else '%d%s',
1322 factor=1024 if '#' in flags else 1000)
1323 elif fmt[-1] == 'S': # filename sanitization
1324 value, fmt = filename_sanitizer(last_field, value, restricted='#' in flags), str_fmt
1325 elif fmt[-1] == 'c':
1326 if value:
1327 value = str(value)[0]
1328 else:
1329 fmt = str_fmt
1330 elif fmt[-1] not in 'rsa': # numeric
1331 value = float_or_none(value)
1332 if value is None:
1333 value, fmt = default, 's'
1334
1335 if sanitize:
1336 # If value is an object, sanitize might convert it to a string
1337 # So we convert it to repr first
1338 if fmt[-1] == 'r':
1339 value, fmt = repr(value), str_fmt
1340 elif fmt[-1] == 'a':
1341 value, fmt = ascii(value), str_fmt
1342 if fmt[-1] in 'csra':
1343 value = sanitizer(last_field, value)
1344
1345 key = '%s\0%s' % (key.replace('%', '%\0'), outer_mobj.group('format'))
1346 TMPL_DICT[key] = value
1347 return '{prefix}%({key}){fmt}'.format(key=key, fmt=fmt, prefix=outer_mobj.group('prefix'))
1348
1349 return EXTERNAL_FORMAT_RE.sub(create_key, outtmpl), TMPL_DICT
1350
1351 def evaluate_outtmpl(self, outtmpl, info_dict, *args, **kwargs):
1352 outtmpl, info_dict = self.prepare_outtmpl(outtmpl, info_dict, *args, **kwargs)
1353 return self.escape_outtmpl(outtmpl) % info_dict
1354
1355 def _prepare_filename(self, info_dict, *, outtmpl=None, tmpl_type=None):
1356 assert None in (outtmpl, tmpl_type), 'outtmpl and tmpl_type are mutually exclusive'
1357 if outtmpl is None:
1358 outtmpl = self.params['outtmpl'].get(tmpl_type or 'default', self.params['outtmpl']['default'])
1359 try:
1360 outtmpl = self._outtmpl_expandpath(outtmpl)
1361 filename = self.evaluate_outtmpl(outtmpl, info_dict, True)
1362 if not filename:
1363 return None
1364
1365 if tmpl_type in ('', 'temp'):
1366 final_ext, ext = self.params.get('final_ext'), info_dict.get('ext')
1367 if final_ext and ext and final_ext != ext and filename.endswith(f'.{final_ext}'):
1368 filename = replace_extension(filename, ext, final_ext)
1369 elif tmpl_type:
1370 force_ext = OUTTMPL_TYPES[tmpl_type]
1371 if force_ext:
1372 filename = replace_extension(filename, force_ext, info_dict.get('ext'))
1373
1374 # https://github.com/blackjack4494/youtube-dlc/issues/85
1375 trim_file_name = self.params.get('trim_file_name', False)
1376 if trim_file_name:
1377 no_ext, *ext = filename.rsplit('.', 2)
1378 filename = join_nonempty(no_ext[:trim_file_name], *ext, delim='.')
1379
1380 return filename
1381 except ValueError as err:
1382 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
1383 return None
1384
1385 def prepare_filename(self, info_dict, dir_type='', *, outtmpl=None, warn=False):
1386 """Generate the output filename"""
1387 if outtmpl:
1388 assert not dir_type, 'outtmpl and dir_type are mutually exclusive'
1389 dir_type = None
1390 filename = self._prepare_filename(info_dict, tmpl_type=dir_type, outtmpl=outtmpl)
1391 if not filename and dir_type not in ('', 'temp'):
1392 return ''
1393
1394 if warn:
1395 if not self.params.get('paths'):
1396 pass
1397 elif filename == '-':
1398 self.report_warning('--paths is ignored when an outputting to stdout', only_once=True)
1399 elif os.path.isabs(filename):
1400 self.report_warning('--paths is ignored since an absolute path is given in output template', only_once=True)
1401 if filename == '-' or not filename:
1402 return filename
1403
1404 return self.get_output_path(dir_type, filename)
1405
1406 def _match_entry(self, info_dict, incomplete=False, silent=False):
1407 """Returns None if the file should be downloaded"""
1408 _type = 'video' if 'playlist-match-filter' in self.params['compat_opts'] else info_dict.get('_type', 'video')
1409 assert incomplete or _type == 'video', 'Only video result can be considered complete'
1410
1411 video_title = info_dict.get('title', info_dict.get('id', 'entry'))
1412
1413 def check_filter():
1414 if _type in ('playlist', 'multi_video'):
1415 return
1416 elif _type in ('url', 'url_transparent') and not try_call(
1417 lambda: self.get_info_extractor(info_dict['ie_key']).is_single_video(info_dict['url'])):
1418 return
1419
1420 if 'title' in info_dict:
1421 # This can happen when we're just evaluating the playlist
1422 title = info_dict['title']
1423 matchtitle = self.params.get('matchtitle', False)
1424 if matchtitle:
1425 if not re.search(matchtitle, title, re.IGNORECASE):
1426 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
1427 rejecttitle = self.params.get('rejecttitle', False)
1428 if rejecttitle:
1429 if re.search(rejecttitle, title, re.IGNORECASE):
1430 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
1431
1432 date = info_dict.get('upload_date')
1433 if date is not None:
1434 dateRange = self.params.get('daterange', DateRange())
1435 if date not in dateRange:
1436 return f'{date_from_str(date).isoformat()} upload date is not in range {dateRange}'
1437 view_count = info_dict.get('view_count')
1438 if view_count is not None:
1439 min_views = self.params.get('min_views')
1440 if min_views is not None and view_count < min_views:
1441 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
1442 max_views = self.params.get('max_views')
1443 if max_views is not None and view_count > max_views:
1444 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
1445 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
1446 return 'Skipping "%s" because it is age restricted' % video_title
1447
1448 match_filter = self.params.get('match_filter')
1449 if match_filter is None:
1450 return None
1451
1452 cancelled = None
1453 try:
1454 try:
1455 ret = match_filter(info_dict, incomplete=incomplete)
1456 except TypeError:
1457 # For backward compatibility
1458 ret = None if incomplete else match_filter(info_dict)
1459 except DownloadCancelled as err:
1460 if err.msg is not NO_DEFAULT:
1461 raise
1462 ret, cancelled = err.msg, err
1463
1464 if ret is NO_DEFAULT:
1465 while True:
1466 filename = self._format_screen(self.prepare_filename(info_dict), self.Styles.FILENAME)
1467 reply = input(self._format_screen(
1468 f'Download "{filename}"? (Y/n): ', self.Styles.EMPHASIS)).lower().strip()
1469 if reply in {'y', ''}:
1470 return None
1471 elif reply == 'n':
1472 if cancelled:
1473 raise type(cancelled)(f'Skipping {video_title}')
1474 return f'Skipping {video_title}'
1475 return ret
1476
1477 if self.in_download_archive(info_dict):
1478 reason = '%s has already been recorded in the archive' % video_title
1479 break_opt, break_err = 'break_on_existing', ExistingVideoReached
1480 else:
1481 try:
1482 reason = check_filter()
1483 except DownloadCancelled as e:
1484 reason, break_opt, break_err = e.msg, 'match_filter', type(e)
1485 else:
1486 break_opt, break_err = 'break_on_reject', RejectedVideoReached
1487 if reason is not None:
1488 if not silent:
1489 self.to_screen('[download] ' + reason)
1490 if self.params.get(break_opt, False):
1491 raise break_err()
1492 return reason
1493
1494 @staticmethod
1495 def add_extra_info(info_dict, extra_info):
1496 '''Set the keys from extra_info in info dict if they are missing'''
1497 for key, value in extra_info.items():
1498 info_dict.setdefault(key, value)
1499
1500 def extract_info(self, url, download=True, ie_key=None, extra_info=None,
1501 process=True, force_generic_extractor=False):
1502 """
1503 Extract and return the information dictionary of the URL
1504
1505 Arguments:
1506 @param url URL to extract
1507
1508 Keyword arguments:
1509 @param download Whether to download videos
1510 @param process Whether to resolve all unresolved references (URLs, playlist items).
1511 Must be True for download to work
1512 @param ie_key Use only the extractor with this key
1513
1514 @param extra_info Dictionary containing the extra values to add to the info (For internal use only)
1515 @force_generic_extractor Force using the generic extractor (Deprecated; use ie_key='Generic')
1516 """
1517
1518 if extra_info is None:
1519 extra_info = {}
1520
1521 if not ie_key and force_generic_extractor:
1522 ie_key = 'Generic'
1523
1524 if ie_key:
1525 ies = {ie_key: self._ies[ie_key]} if ie_key in self._ies else {}
1526 else:
1527 ies = self._ies
1528
1529 for key, ie in ies.items():
1530 if not ie.suitable(url):
1531 continue
1532
1533 if not ie.working():
1534 self.report_warning('The program functionality for this site has been marked as broken, '
1535 'and will probably not work.')
1536
1537 temp_id = ie.get_temp_id(url)
1538 if temp_id is not None and self.in_download_archive({'id': temp_id, 'ie_key': key}):
1539 self.to_screen(f'[{key}] {temp_id}: has already been recorded in the archive')
1540 if self.params.get('break_on_existing', False):
1541 raise ExistingVideoReached()
1542 break
1543 return self.__extract_info(url, self.get_info_extractor(key), download, extra_info, process)
1544 else:
1545 extractors_restricted = self.params.get('allowed_extractors') not in (None, ['default'])
1546 self.report_error(f'No suitable extractor{format_field(ie_key, None, " (%s)")} found for URL {url}',
1547 tb=False if extractors_restricted else None)
1548
1549 def _handle_extraction_exceptions(func):
1550 @functools.wraps(func)
1551 def wrapper(self, *args, **kwargs):
1552 while True:
1553 try:
1554 return func(self, *args, **kwargs)
1555 except (DownloadCancelled, LazyList.IndexError, PagedList.IndexError):
1556 raise
1557 except ReExtractInfo as e:
1558 if e.expected:
1559 self.to_screen(f'{e}; Re-extracting data')
1560 else:
1561 self.to_stderr('\r')
1562 self.report_warning(f'{e}; Re-extracting data')
1563 continue
1564 except GeoRestrictedError as e:
1565 msg = e.msg
1566 if e.countries:
1567 msg += '\nThis video is available in %s.' % ', '.join(
1568 map(ISO3166Utils.short2full, e.countries))
1569 msg += '\nYou might want to use a VPN or a proxy server (with --proxy) to workaround.'
1570 self.report_error(msg)
1571 except ExtractorError as e: # An error we somewhat expected
1572 self.report_error(str(e), e.format_traceback())
1573 except Exception as e:
1574 if self.params.get('ignoreerrors'):
1575 self.report_error(str(e), tb=encode_compat_str(traceback.format_exc()))
1576 else:
1577 raise
1578 break
1579 return wrapper
1580
1581 def _wait_for_video(self, ie_result={}):
1582 if (not self.params.get('wait_for_video')
1583 or ie_result.get('_type', 'video') != 'video'
1584 or ie_result.get('formats') or ie_result.get('url')):
1585 return
1586
1587 format_dur = lambda dur: '%02d:%02d:%02d' % timetuple_from_msec(dur * 1000)[:-1]
1588 last_msg = ''
1589
1590 def progress(msg):
1591 nonlocal last_msg
1592 full_msg = f'{msg}\n'
1593 if not self.params.get('noprogress'):
1594 full_msg = msg + ' ' * (len(last_msg) - len(msg)) + '\r'
1595 elif last_msg:
1596 return
1597 self.to_screen(full_msg, skip_eol=True)
1598 last_msg = msg
1599
1600 min_wait, max_wait = self.params.get('wait_for_video')
1601 diff = try_get(ie_result, lambda x: x['release_timestamp'] - time.time())
1602 if diff is None and ie_result.get('live_status') == 'is_upcoming':
1603 diff = round(random.uniform(min_wait, max_wait) if (max_wait and min_wait) else (max_wait or min_wait), 0)
1604 self.report_warning('Release time of video is not known')
1605 elif ie_result and (diff or 0) <= 0:
1606 self.report_warning('Video should already be available according to extracted info')
1607 diff = min(max(diff or 0, min_wait or 0), max_wait or float('inf'))
1608 self.to_screen(f'[wait] Waiting for {format_dur(diff)} - Press Ctrl+C to try now')
1609
1610 wait_till = time.time() + diff
1611 try:
1612 while True:
1613 diff = wait_till - time.time()
1614 if diff <= 0:
1615 progress('')
1616 raise ReExtractInfo('[wait] Wait period ended', expected=True)
1617 progress(f'[wait] Remaining time until next attempt: {self._format_screen(format_dur(diff), self.Styles.EMPHASIS)}')
1618 time.sleep(1)
1619 except KeyboardInterrupt:
1620 progress('')
1621 raise ReExtractInfo('[wait] Interrupted by user', expected=True)
1622 except BaseException as e:
1623 if not isinstance(e, ReExtractInfo):
1624 self.to_screen('')
1625 raise
1626
1627 @_handle_extraction_exceptions
1628 def __extract_info(self, url, ie, download, extra_info, process):
1629 try:
1630 ie_result = ie.extract(url)
1631 except UserNotLive as e:
1632 if process:
1633 if self.params.get('wait_for_video'):
1634 self.report_warning(e)
1635 self._wait_for_video()
1636 raise
1637 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
1638 self.report_warning(f'Extractor {ie.IE_NAME} returned nothing{bug_reports_message()}')
1639 return
1640 if isinstance(ie_result, list):
1641 # Backwards compatibility: old IE result format
1642 ie_result = {
1643 '_type': 'compat_list',
1644 'entries': ie_result,
1645 }
1646 if extra_info.get('original_url'):
1647 ie_result.setdefault('original_url', extra_info['original_url'])
1648 self.add_default_extra_info(ie_result, ie, url)
1649 if process:
1650 self._wait_for_video(ie_result)
1651 return self.process_ie_result(ie_result, download, extra_info)
1652 else:
1653 return ie_result
1654
1655 def add_default_extra_info(self, ie_result, ie, url):
1656 if url is not None:
1657 self.add_extra_info(ie_result, {
1658 'webpage_url': url,
1659 'original_url': url,
1660 })
1661 webpage_url = ie_result.get('webpage_url')
1662 if webpage_url:
1663 self.add_extra_info(ie_result, {
1664 'webpage_url_basename': url_basename(webpage_url),
1665 'webpage_url_domain': get_domain(webpage_url),
1666 })
1667 if ie is not None:
1668 self.add_extra_info(ie_result, {
1669 'extractor': ie.IE_NAME,
1670 'extractor_key': ie.ie_key(),
1671 })
1672
1673 def process_ie_result(self, ie_result, download=True, extra_info=None):
1674 """
1675 Take the result of the ie(may be modified) and resolve all unresolved
1676 references (URLs, playlist items).
1677
1678 It will also download the videos if 'download'.
1679 Returns the resolved ie_result.
1680 """
1681 if extra_info is None:
1682 extra_info = {}
1683 result_type = ie_result.get('_type', 'video')
1684
1685 if result_type in ('url', 'url_transparent'):
1686 ie_result['url'] = sanitize_url(
1687 ie_result['url'], scheme='http' if self.params.get('prefer_insecure') else 'https')
1688 if ie_result.get('original_url') and not extra_info.get('original_url'):
1689 extra_info = {'original_url': ie_result['original_url'], **extra_info}
1690
1691 extract_flat = self.params.get('extract_flat', False)
1692 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info)
1693 or extract_flat is True):
1694 info_copy = ie_result.copy()
1695 ie = try_get(ie_result.get('ie_key'), self.get_info_extractor)
1696 if ie and not ie_result.get('id'):
1697 info_copy['id'] = ie.get_temp_id(ie_result['url'])
1698 self.add_default_extra_info(info_copy, ie, ie_result['url'])
1699 self.add_extra_info(info_copy, extra_info)
1700 info_copy, _ = self.pre_process(info_copy)
1701 self._fill_common_fields(info_copy, False)
1702 self.__forced_printings(info_copy)
1703 self._raise_pending_errors(info_copy)
1704 if self.params.get('force_write_download_archive', False):
1705 self.record_download_archive(info_copy)
1706 return ie_result
1707
1708 if result_type == 'video':
1709 self.add_extra_info(ie_result, extra_info)
1710 ie_result = self.process_video_result(ie_result, download=download)
1711 self._raise_pending_errors(ie_result)
1712 additional_urls = (ie_result or {}).get('additional_urls')
1713 if additional_urls:
1714 # TODO: Improve MetadataParserPP to allow setting a list
1715 if isinstance(additional_urls, str):
1716 additional_urls = [additional_urls]
1717 self.to_screen(
1718 '[info] %s: %d additional URL(s) requested' % (ie_result['id'], len(additional_urls)))
1719 self.write_debug('Additional URLs: "%s"' % '", "'.join(additional_urls))
1720 ie_result['additional_entries'] = [
1721 self.extract_info(
1722 url, download, extra_info=extra_info,
1723 force_generic_extractor=self.params.get('force_generic_extractor'))
1724 for url in additional_urls
1725 ]
1726 return ie_result
1727 elif result_type == 'url':
1728 # We have to add extra_info to the results because it may be
1729 # contained in a playlist
1730 return self.extract_info(
1731 ie_result['url'], download,
1732 ie_key=ie_result.get('ie_key'),
1733 extra_info=extra_info)
1734 elif result_type == 'url_transparent':
1735 # Use the information from the embedding page
1736 info = self.extract_info(
1737 ie_result['url'], ie_key=ie_result.get('ie_key'),
1738 extra_info=extra_info, download=False, process=False)
1739
1740 # extract_info may return None when ignoreerrors is enabled and
1741 # extraction failed with an error, don't crash and return early
1742 # in this case
1743 if not info:
1744 return info
1745
1746 exempted_fields = {'_type', 'url', 'ie_key'}
1747 if not ie_result.get('section_end') and ie_result.get('section_start') is None:
1748 # For video clips, the id etc of the clip extractor should be used
1749 exempted_fields |= {'id', 'extractor', 'extractor_key'}
1750
1751 new_result = info.copy()
1752 new_result.update(filter_dict(ie_result, lambda k, v: v is not None and k not in exempted_fields))
1753
1754 # Extracted info may not be a video result (i.e.
1755 # info.get('_type', 'video') != video) but rather an url or
1756 # url_transparent. In such cases outer metadata (from ie_result)
1757 # should be propagated to inner one (info). For this to happen
1758 # _type of info should be overridden with url_transparent. This
1759 # fixes issue from https://github.com/ytdl-org/youtube-dl/pull/11163.
1760 if new_result.get('_type') == 'url':
1761 new_result['_type'] = 'url_transparent'
1762
1763 return self.process_ie_result(
1764 new_result, download=download, extra_info=extra_info)
1765 elif result_type in ('playlist', 'multi_video'):
1766 # Protect from infinite recursion due to recursively nested playlists
1767 # (see https://github.com/ytdl-org/youtube-dl/issues/27833)
1768 webpage_url = ie_result.get('webpage_url') # Playlists maynot have webpage_url
1769 if webpage_url and webpage_url in self._playlist_urls:
1770 self.to_screen(
1771 '[download] Skipping already downloaded playlist: %s'
1772 % ie_result.get('title') or ie_result.get('id'))
1773 return
1774
1775 self._playlist_level += 1
1776 self._playlist_urls.add(webpage_url)
1777 self._fill_common_fields(ie_result, False)
1778 self._sanitize_thumbnails(ie_result)
1779 try:
1780 return self.__process_playlist(ie_result, download)
1781 finally:
1782 self._playlist_level -= 1
1783 if not self._playlist_level:
1784 self._playlist_urls.clear()
1785 elif result_type == 'compat_list':
1786 self.report_warning(
1787 'Extractor %s returned a compat_list result. '
1788 'It needs to be updated.' % ie_result.get('extractor'))
1789
1790 def _fixup(r):
1791 self.add_extra_info(r, {
1792 'extractor': ie_result['extractor'],
1793 'webpage_url': ie_result['webpage_url'],
1794 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1795 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1796 'extractor_key': ie_result['extractor_key'],
1797 })
1798 return r
1799 ie_result['entries'] = [
1800 self.process_ie_result(_fixup(r), download, extra_info)
1801 for r in ie_result['entries']
1802 ]
1803 return ie_result
1804 else:
1805 raise Exception('Invalid result type: %s' % result_type)
1806
1807 def _ensure_dir_exists(self, path):
1808 return make_dir(path, self.report_error)
1809
1810 @staticmethod
1811 def _playlist_infodict(ie_result, strict=False, **kwargs):
1812 info = {
1813 'playlist_count': ie_result.get('playlist_count'),
1814 'playlist': ie_result.get('title') or ie_result.get('id'),
1815 'playlist_id': ie_result.get('id'),
1816 'playlist_title': ie_result.get('title'),
1817 'playlist_uploader': ie_result.get('uploader'),
1818 'playlist_uploader_id': ie_result.get('uploader_id'),
1819 **kwargs,
1820 }
1821 if strict:
1822 return info
1823 if ie_result.get('webpage_url'):
1824 info.update({
1825 'webpage_url': ie_result['webpage_url'],
1826 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1827 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1828 })
1829 return {
1830 **info,
1831 'playlist_index': 0,
1832 '__last_playlist_index': max(ie_result.get('requested_entries') or (0, 0)),
1833 'extractor': ie_result['extractor'],
1834 'extractor_key': ie_result['extractor_key'],
1835 }
1836
1837 def __process_playlist(self, ie_result, download):
1838 """Process each entry in the playlist"""
1839 assert ie_result['_type'] in ('playlist', 'multi_video')
1840
1841 common_info = self._playlist_infodict(ie_result, strict=True)
1842 title = common_info.get('playlist') or '<Untitled>'
1843 if self._match_entry(common_info, incomplete=True) is not None:
1844 return
1845 self.to_screen(f'[download] Downloading {ie_result["_type"]}: {title}')
1846
1847 all_entries = PlaylistEntries(self, ie_result)
1848 entries = orderedSet(all_entries.get_requested_items(), lazy=True)
1849
1850 lazy = self.params.get('lazy_playlist')
1851 if lazy:
1852 resolved_entries, n_entries = [], 'N/A'
1853 ie_result['requested_entries'], ie_result['entries'] = None, None
1854 else:
1855 entries = resolved_entries = list(entries)
1856 n_entries = len(resolved_entries)
1857 ie_result['requested_entries'], ie_result['entries'] = tuple(zip(*resolved_entries)) or ([], [])
1858 if not ie_result.get('playlist_count'):
1859 # Better to do this after potentially exhausting entries
1860 ie_result['playlist_count'] = all_entries.get_full_count()
1861
1862 extra = self._playlist_infodict(ie_result, n_entries=int_or_none(n_entries))
1863 ie_copy = collections.ChainMap(ie_result, extra)
1864
1865 _infojson_written = False
1866 write_playlist_files = self.params.get('allow_playlist_files', True)
1867 if write_playlist_files and self.params.get('list_thumbnails'):
1868 self.list_thumbnails(ie_result)
1869 if write_playlist_files and not self.params.get('simulate'):
1870 _infojson_written = self._write_info_json(
1871 'playlist', ie_result, self.prepare_filename(ie_copy, 'pl_infojson'))
1872 if _infojson_written is None:
1873 return
1874 if self._write_description('playlist', ie_result,
1875 self.prepare_filename(ie_copy, 'pl_description')) is None:
1876 return
1877 # TODO: This should be passed to ThumbnailsConvertor if necessary
1878 self._write_thumbnails('playlist', ie_result, self.prepare_filename(ie_copy, 'pl_thumbnail'))
1879
1880 if lazy:
1881 if self.params.get('playlistreverse') or self.params.get('playlistrandom'):
1882 self.report_warning('playlistreverse and playlistrandom are not supported with lazy_playlist', only_once=True)
1883 elif self.params.get('playlistreverse'):
1884 entries.reverse()
1885 elif self.params.get('playlistrandom'):
1886 random.shuffle(entries)
1887
1888 self.to_screen(f'[{ie_result["extractor"]}] Playlist {title}: Downloading {n_entries} items'
1889 f'{format_field(ie_result, "playlist_count", " of %s")}')
1890
1891 keep_resolved_entries = self.params.get('extract_flat') != 'discard'
1892 if self.params.get('extract_flat') == 'discard_in_playlist':
1893 keep_resolved_entries = ie_result['_type'] != 'playlist'
1894 if keep_resolved_entries:
1895 self.write_debug('The information of all playlist entries will be held in memory')
1896
1897 failures = 0
1898 max_failures = self.params.get('skip_playlist_after_errors') or float('inf')
1899 for i, (playlist_index, entry) in enumerate(entries):
1900 if lazy:
1901 resolved_entries.append((playlist_index, entry))
1902 if not entry:
1903 continue
1904
1905 entry['__x_forwarded_for_ip'] = ie_result.get('__x_forwarded_for_ip')
1906 if not lazy and 'playlist-index' in self.params['compat_opts']:
1907 playlist_index = ie_result['requested_entries'][i]
1908
1909 entry_copy = collections.ChainMap(entry, {
1910 **common_info,
1911 'n_entries': int_or_none(n_entries),
1912 'playlist_index': playlist_index,
1913 'playlist_autonumber': i + 1,
1914 })
1915
1916 if self._match_entry(entry_copy, incomplete=True) is not None:
1917 # For compatabilty with youtube-dl. See https://github.com/yt-dlp/yt-dlp/issues/4369
1918 resolved_entries[i] = (playlist_index, NO_DEFAULT)
1919 continue
1920
1921 self.to_screen('[download] Downloading item %s of %s' % (
1922 self._format_screen(i + 1, self.Styles.ID), self._format_screen(n_entries, self.Styles.EMPHASIS)))
1923
1924 entry_result = self.__process_iterable_entry(entry, download, collections.ChainMap({
1925 'playlist_index': playlist_index,
1926 'playlist_autonumber': i + 1,
1927 }, extra))
1928 if not entry_result:
1929 failures += 1
1930 if failures >= max_failures:
1931 self.report_error(
1932 f'Skipping the remaining entries in playlist "{title}" since {failures} items failed extraction')
1933 break
1934 if keep_resolved_entries:
1935 resolved_entries[i] = (playlist_index, entry_result)
1936
1937 # Update with processed data
1938 ie_result['entries'] = [e for _, e in resolved_entries if e is not NO_DEFAULT]
1939 ie_result['requested_entries'] = [i for i, e in resolved_entries if e is not NO_DEFAULT]
1940 if ie_result['requested_entries'] == try_call(lambda: list(range(1, ie_result['playlist_count'] + 1))):
1941 # Do not set for full playlist
1942 ie_result.pop('requested_entries')
1943
1944 # Write the updated info to json
1945 if _infojson_written is True and self._write_info_json(
1946 'updated playlist', ie_result,
1947 self.prepare_filename(ie_copy, 'pl_infojson'), overwrite=True) is None:
1948 return
1949
1950 ie_result = self.run_all_pps('playlist', ie_result)
1951 self.to_screen(f'[download] Finished downloading playlist: {title}')
1952 return ie_result
1953
1954 @_handle_extraction_exceptions
1955 def __process_iterable_entry(self, entry, download, extra_info):
1956 return self.process_ie_result(
1957 entry, download=download, extra_info=extra_info)
1958
1959 def _build_format_filter(self, filter_spec):
1960 " Returns a function to filter the formats according to the filter_spec "
1961
1962 OPERATORS = {
1963 '<': operator.lt,
1964 '<=': operator.le,
1965 '>': operator.gt,
1966 '>=': operator.ge,
1967 '=': operator.eq,
1968 '!=': operator.ne,
1969 }
1970 operator_rex = re.compile(r'''(?x)\s*
1971 (?P<key>[\w.-]+)\s*
1972 (?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
1973 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)\s*
1974 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
1975 m = operator_rex.fullmatch(filter_spec)
1976 if m:
1977 try:
1978 comparison_value = int(m.group('value'))
1979 except ValueError:
1980 comparison_value = parse_filesize(m.group('value'))
1981 if comparison_value is None:
1982 comparison_value = parse_filesize(m.group('value') + 'B')
1983 if comparison_value is None:
1984 raise ValueError(
1985 'Invalid value %r in format specification %r' % (
1986 m.group('value'), filter_spec))
1987 op = OPERATORS[m.group('op')]
1988
1989 if not m:
1990 STR_OPERATORS = {
1991 '=': operator.eq,
1992 '^=': lambda attr, value: attr.startswith(value),
1993 '$=': lambda attr, value: attr.endswith(value),
1994 '*=': lambda attr, value: value in attr,
1995 '~=': lambda attr, value: value.search(attr) is not None
1996 }
1997 str_operator_rex = re.compile(r'''(?x)\s*
1998 (?P<key>[a-zA-Z0-9._-]+)\s*
1999 (?P<negation>!\s*)?(?P<op>%s)\s*(?P<none_inclusive>\?\s*)?
2000 (?P<quote>["'])?
2001 (?P<value>(?(quote)(?:(?!(?P=quote))[^\\]|\\.)+|[\w.-]+))
2002 (?(quote)(?P=quote))\s*
2003 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
2004 m = str_operator_rex.fullmatch(filter_spec)
2005 if m:
2006 if m.group('op') == '~=':
2007 comparison_value = re.compile(m.group('value'))
2008 else:
2009 comparison_value = re.sub(r'''\\([\\"'])''', r'\1', m.group('value'))
2010 str_op = STR_OPERATORS[m.group('op')]
2011 if m.group('negation'):
2012 op = lambda attr, value: not str_op(attr, value)
2013 else:
2014 op = str_op
2015
2016 if not m:
2017 raise SyntaxError('Invalid filter specification %r' % filter_spec)
2018
2019 def _filter(f):
2020 actual_value = f.get(m.group('key'))
2021 if actual_value is None:
2022 return m.group('none_inclusive')
2023 return op(actual_value, comparison_value)
2024 return _filter
2025
2026 def _check_formats(self, formats):
2027 for f in formats:
2028 self.to_screen('[info] Testing format %s' % f['format_id'])
2029 path = self.get_output_path('temp')
2030 if not self._ensure_dir_exists(f'{path}/'):
2031 continue
2032 temp_file = tempfile.NamedTemporaryFile(suffix='.tmp', delete=False, dir=path or None)
2033 temp_file.close()
2034 try:
2035 success, _ = self.dl(temp_file.name, f, test=True)
2036 except (DownloadError, OSError, ValueError) + network_exceptions:
2037 success = False
2038 finally:
2039 if os.path.exists(temp_file.name):
2040 try:
2041 os.remove(temp_file.name)
2042 except OSError:
2043 self.report_warning('Unable to delete temporary file "%s"' % temp_file.name)
2044 if success:
2045 yield f
2046 else:
2047 self.to_screen('[info] Unable to download format %s. Skipping...' % f['format_id'])
2048
2049 def _default_format_spec(self, info_dict, download=True):
2050
2051 def can_merge():
2052 merger = FFmpegMergerPP(self)
2053 return merger.available and merger.can_merge()
2054
2055 prefer_best = (
2056 not self.params.get('simulate')
2057 and download
2058 and (
2059 not can_merge()
2060 or info_dict.get('is_live') and not self.params.get('live_from_start')
2061 or self.params['outtmpl']['default'] == '-'))
2062 compat = (
2063 prefer_best
2064 or self.params.get('allow_multiple_audio_streams', False)
2065 or 'format-spec' in self.params['compat_opts'])
2066
2067 return (
2068 'best/bestvideo+bestaudio' if prefer_best
2069 else 'bestvideo*+bestaudio/best' if not compat
2070 else 'bestvideo+bestaudio/best')
2071
2072 def build_format_selector(self, format_spec):
2073 def syntax_error(note, start):
2074 message = (
2075 'Invalid format specification: '
2076 '{}\n\t{}\n\t{}^'.format(note, format_spec, ' ' * start[1]))
2077 return SyntaxError(message)
2078
2079 PICKFIRST = 'PICKFIRST'
2080 MERGE = 'MERGE'
2081 SINGLE = 'SINGLE'
2082 GROUP = 'GROUP'
2083 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
2084
2085 allow_multiple_streams = {'audio': self.params.get('allow_multiple_audio_streams', False),
2086 'video': self.params.get('allow_multiple_video_streams', False)}
2087
2088 check_formats = self.params.get('check_formats') == 'selected'
2089
2090 def _parse_filter(tokens):
2091 filter_parts = []
2092 for type, string_, start, _, _ in tokens:
2093 if type == tokenize.OP and string_ == ']':
2094 return ''.join(filter_parts)
2095 else:
2096 filter_parts.append(string_)
2097
2098 def _remove_unused_ops(tokens):
2099 # Remove operators that we don't use and join them with the surrounding strings.
2100 # E.g. 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
2101 ALLOWED_OPS = ('/', '+', ',', '(', ')')
2102 last_string, last_start, last_end, last_line = None, None, None, None
2103 for type, string_, start, end, line in tokens:
2104 if type == tokenize.OP and string_ == '[':
2105 if last_string:
2106 yield tokenize.NAME, last_string, last_start, last_end, last_line
2107 last_string = None
2108 yield type, string_, start, end, line
2109 # everything inside brackets will be handled by _parse_filter
2110 for type, string_, start, end, line in tokens:
2111 yield type, string_, start, end, line
2112 if type == tokenize.OP and string_ == ']':
2113 break
2114 elif type == tokenize.OP and string_ in ALLOWED_OPS:
2115 if last_string:
2116 yield tokenize.NAME, last_string, last_start, last_end, last_line
2117 last_string = None
2118 yield type, string_, start, end, line
2119 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
2120 if not last_string:
2121 last_string = string_
2122 last_start = start
2123 last_end = end
2124 else:
2125 last_string += string_
2126 if last_string:
2127 yield tokenize.NAME, last_string, last_start, last_end, last_line
2128
2129 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
2130 selectors = []
2131 current_selector = None
2132 for type, string_, start, _, _ in tokens:
2133 # ENCODING is only defined in python 3.x
2134 if type == getattr(tokenize, 'ENCODING', None):
2135 continue
2136 elif type in [tokenize.NAME, tokenize.NUMBER]:
2137 current_selector = FormatSelector(SINGLE, string_, [])
2138 elif type == tokenize.OP:
2139 if string_ == ')':
2140 if not inside_group:
2141 # ')' will be handled by the parentheses group
2142 tokens.restore_last_token()
2143 break
2144 elif inside_merge and string_ in ['/', ',']:
2145 tokens.restore_last_token()
2146 break
2147 elif inside_choice and string_ == ',':
2148 tokens.restore_last_token()
2149 break
2150 elif string_ == ',':
2151 if not current_selector:
2152 raise syntax_error('"," must follow a format selector', start)
2153 selectors.append(current_selector)
2154 current_selector = None
2155 elif string_ == '/':
2156 if not current_selector:
2157 raise syntax_error('"/" must follow a format selector', start)
2158 first_choice = current_selector
2159 second_choice = _parse_format_selection(tokens, inside_choice=True)
2160 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
2161 elif string_ == '[':
2162 if not current_selector:
2163 current_selector = FormatSelector(SINGLE, 'best', [])
2164 format_filter = _parse_filter(tokens)
2165 current_selector.filters.append(format_filter)
2166 elif string_ == '(':
2167 if current_selector:
2168 raise syntax_error('Unexpected "("', start)
2169 group = _parse_format_selection(tokens, inside_group=True)
2170 current_selector = FormatSelector(GROUP, group, [])
2171 elif string_ == '+':
2172 if not current_selector:
2173 raise syntax_error('Unexpected "+"', start)
2174 selector_1 = current_selector
2175 selector_2 = _parse_format_selection(tokens, inside_merge=True)
2176 if not selector_2:
2177 raise syntax_error('Expected a selector', start)
2178 current_selector = FormatSelector(MERGE, (selector_1, selector_2), [])
2179 else:
2180 raise syntax_error(f'Operator not recognized: "{string_}"', start)
2181 elif type == tokenize.ENDMARKER:
2182 break
2183 if current_selector:
2184 selectors.append(current_selector)
2185 return selectors
2186
2187 def _merge(formats_pair):
2188 format_1, format_2 = formats_pair
2189
2190 formats_info = []
2191 formats_info.extend(format_1.get('requested_formats', (format_1,)))
2192 formats_info.extend(format_2.get('requested_formats', (format_2,)))
2193
2194 if not allow_multiple_streams['video'] or not allow_multiple_streams['audio']:
2195 get_no_more = {'video': False, 'audio': False}
2196 for (i, fmt_info) in enumerate(formats_info):
2197 if fmt_info.get('acodec') == fmt_info.get('vcodec') == 'none':
2198 formats_info.pop(i)
2199 continue
2200 for aud_vid in ['audio', 'video']:
2201 if not allow_multiple_streams[aud_vid] and fmt_info.get(aud_vid[0] + 'codec') != 'none':
2202 if get_no_more[aud_vid]:
2203 formats_info.pop(i)
2204 break
2205 get_no_more[aud_vid] = True
2206
2207 if len(formats_info) == 1:
2208 return formats_info[0]
2209
2210 video_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('vcodec') != 'none']
2211 audio_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('acodec') != 'none']
2212
2213 the_only_video = video_fmts[0] if len(video_fmts) == 1 else None
2214 the_only_audio = audio_fmts[0] if len(audio_fmts) == 1 else None
2215
2216 output_ext = get_compatible_ext(
2217 vcodecs=[f.get('vcodec') for f in video_fmts],
2218 acodecs=[f.get('acodec') for f in audio_fmts],
2219 vexts=[f['ext'] for f in video_fmts],
2220 aexts=[f['ext'] for f in audio_fmts],
2221 preferences=(try_call(lambda: self.params['merge_output_format'].split('/'))
2222 or self.params.get('prefer_free_formats') and ('webm', 'mkv')))
2223
2224 filtered = lambda *keys: filter(None, (traverse_obj(fmt, *keys) for fmt in formats_info))
2225
2226 new_dict = {
2227 'requested_formats': formats_info,
2228 'format': '+'.join(filtered('format')),
2229 'format_id': '+'.join(filtered('format_id')),
2230 'ext': output_ext,
2231 'protocol': '+'.join(map(determine_protocol, formats_info)),
2232 'language': '+'.join(orderedSet(filtered('language'))) or None,
2233 'format_note': '+'.join(orderedSet(filtered('format_note'))) or None,
2234 'filesize_approx': sum(filtered('filesize', 'filesize_approx')) or None,
2235 'tbr': sum(filtered('tbr', 'vbr', 'abr')),
2236 }
2237
2238 if the_only_video:
2239 new_dict.update({
2240 'width': the_only_video.get('width'),
2241 'height': the_only_video.get('height'),
2242 'resolution': the_only_video.get('resolution') or self.format_resolution(the_only_video),
2243 'fps': the_only_video.get('fps'),
2244 'dynamic_range': the_only_video.get('dynamic_range'),
2245 'vcodec': the_only_video.get('vcodec'),
2246 'vbr': the_only_video.get('vbr'),
2247 'stretched_ratio': the_only_video.get('stretched_ratio'),
2248 'aspect_ratio': the_only_video.get('aspect_ratio'),
2249 })
2250
2251 if the_only_audio:
2252 new_dict.update({
2253 'acodec': the_only_audio.get('acodec'),
2254 'abr': the_only_audio.get('abr'),
2255 'asr': the_only_audio.get('asr'),
2256 'audio_channels': the_only_audio.get('audio_channels')
2257 })
2258
2259 return new_dict
2260
2261 def _check_formats(formats):
2262 if not check_formats:
2263 yield from formats
2264 return
2265 yield from self._check_formats(formats)
2266
2267 def _build_selector_function(selector):
2268 if isinstance(selector, list): # ,
2269 fs = [_build_selector_function(s) for s in selector]
2270
2271 def selector_function(ctx):
2272 for f in fs:
2273 yield from f(ctx)
2274 return selector_function
2275
2276 elif selector.type == GROUP: # ()
2277 selector_function = _build_selector_function(selector.selector)
2278
2279 elif selector.type == PICKFIRST: # /
2280 fs = [_build_selector_function(s) for s in selector.selector]
2281
2282 def selector_function(ctx):
2283 for f in fs:
2284 picked_formats = list(f(ctx))
2285 if picked_formats:
2286 return picked_formats
2287 return []
2288
2289 elif selector.type == MERGE: # +
2290 selector_1, selector_2 = map(_build_selector_function, selector.selector)
2291
2292 def selector_function(ctx):
2293 for pair in itertools.product(selector_1(ctx), selector_2(ctx)):
2294 yield _merge(pair)
2295
2296 elif selector.type == SINGLE: # atom
2297 format_spec = selector.selector or 'best'
2298
2299 # TODO: Add allvideo, allaudio etc by generalizing the code with best/worst selector
2300 if format_spec == 'all':
2301 def selector_function(ctx):
2302 yield from _check_formats(ctx['formats'][::-1])
2303 elif format_spec == 'mergeall':
2304 def selector_function(ctx):
2305 formats = list(_check_formats(
2306 f for f in ctx['formats'] if f.get('vcodec') != 'none' or f.get('acodec') != 'none'))
2307 if not formats:
2308 return
2309 merged_format = formats[-1]
2310 for f in formats[-2::-1]:
2311 merged_format = _merge((merged_format, f))
2312 yield merged_format
2313
2314 else:
2315 format_fallback, seperate_fallback, format_reverse, format_idx = False, None, True, 1
2316 mobj = re.match(
2317 r'(?P<bw>best|worst|b|w)(?P<type>video|audio|v|a)?(?P<mod>\*)?(?:\.(?P<n>[1-9]\d*))?$',
2318 format_spec)
2319 if mobj is not None:
2320 format_idx = int_or_none(mobj.group('n'), default=1)
2321 format_reverse = mobj.group('bw')[0] == 'b'
2322 format_type = (mobj.group('type') or [None])[0]
2323 not_format_type = {'v': 'a', 'a': 'v'}.get(format_type)
2324 format_modified = mobj.group('mod') is not None
2325
2326 format_fallback = not format_type and not format_modified # for b, w
2327 _filter_f = (
2328 (lambda f: f.get('%scodec' % format_type) != 'none')
2329 if format_type and format_modified # bv*, ba*, wv*, wa*
2330 else (lambda f: f.get('%scodec' % not_format_type) == 'none')
2331 if format_type # bv, ba, wv, wa
2332 else (lambda f: f.get('vcodec') != 'none' and f.get('acodec') != 'none')
2333 if not format_modified # b, w
2334 else lambda f: True) # b*, w*
2335 filter_f = lambda f: _filter_f(f) and (
2336 f.get('vcodec') != 'none' or f.get('acodec') != 'none')
2337 else:
2338 if format_spec in self._format_selection_exts['audio']:
2339 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none'
2340 elif format_spec in self._format_selection_exts['video']:
2341 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none' and f.get('vcodec') != 'none'
2342 seperate_fallback = lambda f: f.get('ext') == format_spec and f.get('vcodec') != 'none'
2343 elif format_spec in self._format_selection_exts['storyboards']:
2344 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') == 'none' and f.get('vcodec') == 'none'
2345 else:
2346 filter_f = lambda f: f.get('format_id') == format_spec # id
2347
2348 def selector_function(ctx):
2349 formats = list(ctx['formats'])
2350 matches = list(filter(filter_f, formats)) if filter_f is not None else formats
2351 if not matches:
2352 if format_fallback and ctx['incomplete_formats']:
2353 # for extractors with incomplete formats (audio only (soundcloud)
2354 # or video only (imgur)) best/worst will fallback to
2355 # best/worst {video,audio}-only format
2356 matches = formats
2357 elif seperate_fallback and not ctx['has_merged_format']:
2358 # for compatibility with youtube-dl when there is no pre-merged format
2359 matches = list(filter(seperate_fallback, formats))
2360 matches = LazyList(_check_formats(matches[::-1 if format_reverse else 1]))
2361 try:
2362 yield matches[format_idx - 1]
2363 except LazyList.IndexError:
2364 return
2365
2366 filters = [self._build_format_filter(f) for f in selector.filters]
2367
2368 def final_selector(ctx):
2369 ctx_copy = dict(ctx)
2370 for _filter in filters:
2371 ctx_copy['formats'] = list(filter(_filter, ctx_copy['formats']))
2372 return selector_function(ctx_copy)
2373 return final_selector
2374
2375 stream = io.BytesIO(format_spec.encode())
2376 try:
2377 tokens = list(_remove_unused_ops(tokenize.tokenize(stream.readline)))
2378 except tokenize.TokenError:
2379 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
2380
2381 class TokenIterator:
2382 def __init__(self, tokens):
2383 self.tokens = tokens
2384 self.counter = 0
2385
2386 def __iter__(self):
2387 return self
2388
2389 def __next__(self):
2390 if self.counter >= len(self.tokens):
2391 raise StopIteration()
2392 value = self.tokens[self.counter]
2393 self.counter += 1
2394 return value
2395
2396 next = __next__
2397
2398 def restore_last_token(self):
2399 self.counter -= 1
2400
2401 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
2402 return _build_selector_function(parsed_selector)
2403
2404 def _calc_headers(self, info_dict):
2405 res = merge_headers(self.params['http_headers'], info_dict.get('http_headers') or {})
2406 if 'Youtubedl-No-Compression' in res: # deprecated
2407 res.pop('Youtubedl-No-Compression', None)
2408 res['Accept-Encoding'] = 'identity'
2409 cookies = self.cookiejar.get_cookie_header(info_dict['url'])
2410 if cookies:
2411 res['Cookie'] = cookies
2412
2413 if 'X-Forwarded-For' not in res:
2414 x_forwarded_for_ip = info_dict.get('__x_forwarded_for_ip')
2415 if x_forwarded_for_ip:
2416 res['X-Forwarded-For'] = x_forwarded_for_ip
2417
2418 return res
2419
2420 def _calc_cookies(self, url):
2421 self.deprecation_warning('"YoutubeDL._calc_cookies" is deprecated and may be removed in a future version')
2422 return self.cookiejar.get_cookie_header(url)
2423
2424 def _sort_thumbnails(self, thumbnails):
2425 thumbnails.sort(key=lambda t: (
2426 t.get('preference') if t.get('preference') is not None else -1,
2427 t.get('width') if t.get('width') is not None else -1,
2428 t.get('height') if t.get('height') is not None else -1,
2429 t.get('id') if t.get('id') is not None else '',
2430 t.get('url')))
2431
2432 def _sanitize_thumbnails(self, info_dict):
2433 thumbnails = info_dict.get('thumbnails')
2434 if thumbnails is None:
2435 thumbnail = info_dict.get('thumbnail')
2436 if thumbnail:
2437 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
2438 if not thumbnails:
2439 return
2440
2441 def check_thumbnails(thumbnails):
2442 for t in thumbnails:
2443 self.to_screen(f'[info] Testing thumbnail {t["id"]}')
2444 try:
2445 self.urlopen(HEADRequest(t['url']))
2446 except network_exceptions as err:
2447 self.to_screen(f'[info] Unable to connect to thumbnail {t["id"]} URL {t["url"]!r} - {err}. Skipping...')
2448 continue
2449 yield t
2450
2451 self._sort_thumbnails(thumbnails)
2452 for i, t in enumerate(thumbnails):
2453 if t.get('id') is None:
2454 t['id'] = '%d' % i
2455 if t.get('width') and t.get('height'):
2456 t['resolution'] = '%dx%d' % (t['width'], t['height'])
2457 t['url'] = sanitize_url(t['url'])
2458
2459 if self.params.get('check_formats') is True:
2460 info_dict['thumbnails'] = LazyList(check_thumbnails(thumbnails[::-1]), reverse=True)
2461 else:
2462 info_dict['thumbnails'] = thumbnails
2463
2464 def _fill_common_fields(self, info_dict, final=True):
2465 # TODO: move sanitization here
2466 if final:
2467 title = info_dict['fulltitle'] = info_dict.get('title')
2468 if not title:
2469 if title == '':
2470 self.write_debug('Extractor gave empty title. Creating a generic title')
2471 else:
2472 self.report_warning('Extractor failed to obtain "title". Creating a generic title instead')
2473 info_dict['title'] = f'{info_dict["extractor"].replace(":", "-")} video #{info_dict["id"]}'
2474
2475 if info_dict.get('duration') is not None:
2476 info_dict['duration_string'] = formatSeconds(info_dict['duration'])
2477
2478 for ts_key, date_key in (
2479 ('timestamp', 'upload_date'),
2480 ('release_timestamp', 'release_date'),
2481 ('modified_timestamp', 'modified_date'),
2482 ):
2483 if info_dict.get(date_key) is None and info_dict.get(ts_key) is not None:
2484 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
2485 # see http://bugs.python.org/issue1646728)
2486 with contextlib.suppress(ValueError, OverflowError, OSError):
2487 upload_date = datetime.datetime.utcfromtimestamp(info_dict[ts_key])
2488 info_dict[date_key] = upload_date.strftime('%Y%m%d')
2489
2490 live_keys = ('is_live', 'was_live')
2491 live_status = info_dict.get('live_status')
2492 if live_status is None:
2493 for key in live_keys:
2494 if info_dict.get(key) is False:
2495 continue
2496 if info_dict.get(key):
2497 live_status = key
2498 break
2499 if all(info_dict.get(key) is False for key in live_keys):
2500 live_status = 'not_live'
2501 if live_status:
2502 info_dict['live_status'] = live_status
2503 for key in live_keys:
2504 if info_dict.get(key) is None:
2505 info_dict[key] = (live_status == key)
2506 if live_status == 'post_live':
2507 info_dict['was_live'] = True
2508
2509 # Auto generate title fields corresponding to the *_number fields when missing
2510 # in order to always have clean titles. This is very common for TV series.
2511 for field in ('chapter', 'season', 'episode'):
2512 if final and info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
2513 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
2514
2515 def _raise_pending_errors(self, info):
2516 err = info.pop('__pending_error', None)
2517 if err:
2518 self.report_error(err, tb=False)
2519
2520 def sort_formats(self, info_dict):
2521 formats = self._get_formats(info_dict)
2522 formats.sort(key=FormatSorter(
2523 self, info_dict.get('_format_sort_fields') or []).calculate_preference)
2524
2525 def process_video_result(self, info_dict, download=True):
2526 assert info_dict.get('_type', 'video') == 'video'
2527 self._num_videos += 1
2528
2529 if 'id' not in info_dict:
2530 raise ExtractorError('Missing "id" field in extractor result', ie=info_dict['extractor'])
2531 elif not info_dict.get('id'):
2532 raise ExtractorError('Extractor failed to obtain "id"', ie=info_dict['extractor'])
2533
2534 def report_force_conversion(field, field_not, conversion):
2535 self.report_warning(
2536 '"%s" field is not %s - forcing %s conversion, there is an error in extractor'
2537 % (field, field_not, conversion))
2538
2539 def sanitize_string_field(info, string_field):
2540 field = info.get(string_field)
2541 if field is None or isinstance(field, str):
2542 return
2543 report_force_conversion(string_field, 'a string', 'string')
2544 info[string_field] = str(field)
2545
2546 def sanitize_numeric_fields(info):
2547 for numeric_field in self._NUMERIC_FIELDS:
2548 field = info.get(numeric_field)
2549 if field is None or isinstance(field, (int, float)):
2550 continue
2551 report_force_conversion(numeric_field, 'numeric', 'int')
2552 info[numeric_field] = int_or_none(field)
2553
2554 sanitize_string_field(info_dict, 'id')
2555 sanitize_numeric_fields(info_dict)
2556 if info_dict.get('section_end') and info_dict.get('section_start') is not None:
2557 info_dict['duration'] = round(info_dict['section_end'] - info_dict['section_start'], 3)
2558 if (info_dict.get('duration') or 0) <= 0 and info_dict.pop('duration', None):
2559 self.report_warning('"duration" field is negative, there is an error in extractor')
2560
2561 chapters = info_dict.get('chapters') or []
2562 if chapters and chapters[0].get('start_time'):
2563 chapters.insert(0, {'start_time': 0})
2564
2565 dummy_chapter = {'end_time': 0, 'start_time': info_dict.get('duration')}
2566 for idx, (prev, current, next_) in enumerate(zip(
2567 (dummy_chapter, *chapters), chapters, (*chapters[1:], dummy_chapter)), 1):
2568 if current.get('start_time') is None:
2569 current['start_time'] = prev.get('end_time')
2570 if not current.get('end_time'):
2571 current['end_time'] = next_.get('start_time')
2572 if not current.get('title'):
2573 current['title'] = f'<Untitled Chapter {idx}>'
2574
2575 if 'playlist' not in info_dict:
2576 # It isn't part of a playlist
2577 info_dict['playlist'] = None
2578 info_dict['playlist_index'] = None
2579
2580 self._sanitize_thumbnails(info_dict)
2581
2582 thumbnail = info_dict.get('thumbnail')
2583 thumbnails = info_dict.get('thumbnails')
2584 if thumbnail:
2585 info_dict['thumbnail'] = sanitize_url(thumbnail)
2586 elif thumbnails:
2587 info_dict['thumbnail'] = thumbnails[-1]['url']
2588
2589 if info_dict.get('display_id') is None and 'id' in info_dict:
2590 info_dict['display_id'] = info_dict['id']
2591
2592 self._fill_common_fields(info_dict)
2593
2594 for cc_kind in ('subtitles', 'automatic_captions'):
2595 cc = info_dict.get(cc_kind)
2596 if cc:
2597 for _, subtitle in cc.items():
2598 for subtitle_format in subtitle:
2599 if subtitle_format.get('url'):
2600 subtitle_format['url'] = sanitize_url(subtitle_format['url'])
2601 if subtitle_format.get('ext') is None:
2602 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
2603
2604 automatic_captions = info_dict.get('automatic_captions')
2605 subtitles = info_dict.get('subtitles')
2606
2607 info_dict['requested_subtitles'] = self.process_subtitles(
2608 info_dict['id'], subtitles, automatic_captions)
2609
2610 formats = self._get_formats(info_dict)
2611
2612 # Backward compatibility with InfoExtractor._sort_formats
2613 field_preference = (formats or [{}])[0].pop('__sort_fields', None)
2614 if field_preference:
2615 info_dict['_format_sort_fields'] = field_preference
2616
2617 # or None ensures --clean-infojson removes it
2618 info_dict['_has_drm'] = any(f.get('has_drm') for f in formats) or None
2619 if not self.params.get('allow_unplayable_formats'):
2620 formats = [f for f in formats if not f.get('has_drm')]
2621
2622 if formats and all(f.get('acodec') == f.get('vcodec') == 'none' for f in formats):
2623 self.report_warning(
2624 f'{"This video is DRM protected and " if info_dict["_has_drm"] else ""}'
2625 'only images are available for download. Use --list-formats to see them'.capitalize())
2626
2627 get_from_start = not info_dict.get('is_live') or bool(self.params.get('live_from_start'))
2628 if not get_from_start:
2629 info_dict['title'] += ' ' + datetime.datetime.now().strftime('%Y-%m-%d %H:%M')
2630 if info_dict.get('is_live') and formats:
2631 formats = [f for f in formats if bool(f.get('is_from_start')) == get_from_start]
2632 if get_from_start and not formats:
2633 self.raise_no_formats(info_dict, msg=(
2634 '--live-from-start is passed, but there are no formats that can be downloaded from the start. '
2635 'If you want to download from the current time, use --no-live-from-start'))
2636
2637 def is_wellformed(f):
2638 url = f.get('url')
2639 if not url:
2640 self.report_warning(
2641 '"url" field is missing or empty - skipping format, '
2642 'there is an error in extractor')
2643 return False
2644 if isinstance(url, bytes):
2645 sanitize_string_field(f, 'url')
2646 return True
2647
2648 # Filter out malformed formats for better extraction robustness
2649 formats = list(filter(is_wellformed, formats or []))
2650
2651 if not formats:
2652 self.raise_no_formats(info_dict)
2653
2654 for format in formats:
2655 sanitize_string_field(format, 'format_id')
2656 sanitize_numeric_fields(format)
2657 format['url'] = sanitize_url(format['url'])
2658 if format.get('ext') is None:
2659 format['ext'] = determine_ext(format['url']).lower()
2660 if format.get('protocol') is None:
2661 format['protocol'] = determine_protocol(format)
2662 if format.get('resolution') is None:
2663 format['resolution'] = self.format_resolution(format, default=None)
2664 if format.get('dynamic_range') is None and format.get('vcodec') != 'none':
2665 format['dynamic_range'] = 'SDR'
2666 if format.get('aspect_ratio') is None:
2667 format['aspect_ratio'] = try_call(lambda: round(format['width'] / format['height'], 2))
2668 if (not format.get('manifest_url') # For fragmented formats, "tbr" is often max bitrate and not average
2669 and info_dict.get('duration') and format.get('tbr')
2670 and not format.get('filesize') and not format.get('filesize_approx')):
2671 format['filesize_approx'] = int(info_dict['duration'] * format['tbr'] * (1024 / 8))
2672 format['http_headers'] = self._calc_headers(collections.ChainMap(format, info_dict))
2673
2674 # This is copied to http_headers by the above _calc_headers and can now be removed
2675 if '__x_forwarded_for_ip' in info_dict:
2676 del info_dict['__x_forwarded_for_ip']
2677
2678 self.sort_formats({
2679 'formats': formats,
2680 '_format_sort_fields': info_dict.get('_format_sort_fields')
2681 })
2682
2683 # Sanitize and group by format_id
2684 formats_dict = {}
2685 for i, format in enumerate(formats):
2686 if not format.get('format_id'):
2687 format['format_id'] = str(i)
2688 else:
2689 # Sanitize format_id from characters used in format selector expression
2690 format['format_id'] = re.sub(r'[\s,/+\[\]()]', '_', format['format_id'])
2691 formats_dict.setdefault(format['format_id'], []).append(format)
2692
2693 # Make sure all formats have unique format_id
2694 common_exts = set(itertools.chain(*self._format_selection_exts.values()))
2695 for format_id, ambiguous_formats in formats_dict.items():
2696 ambigious_id = len(ambiguous_formats) > 1
2697 for i, format in enumerate(ambiguous_formats):
2698 if ambigious_id:
2699 format['format_id'] = '%s-%d' % (format_id, i)
2700 # Ensure there is no conflict between id and ext in format selection
2701 # See https://github.com/yt-dlp/yt-dlp/issues/1282
2702 if format['format_id'] != format['ext'] and format['format_id'] in common_exts:
2703 format['format_id'] = 'f%s' % format['format_id']
2704
2705 if format.get('format') is None:
2706 format['format'] = '{id} - {res}{note}'.format(
2707 id=format['format_id'],
2708 res=self.format_resolution(format),
2709 note=format_field(format, 'format_note', ' (%s)'),
2710 )
2711
2712 if self.params.get('check_formats') is True:
2713 formats = LazyList(self._check_formats(formats[::-1]), reverse=True)
2714
2715 if not formats or formats[0] is not info_dict:
2716 # only set the 'formats' fields if the original info_dict list them
2717 # otherwise we end up with a circular reference, the first (and unique)
2718 # element in the 'formats' field in info_dict is info_dict itself,
2719 # which can't be exported to json
2720 info_dict['formats'] = formats
2721
2722 info_dict, _ = self.pre_process(info_dict)
2723
2724 if self._match_entry(info_dict, incomplete=self._format_fields) is not None:
2725 return info_dict
2726
2727 self.post_extract(info_dict)
2728 info_dict, _ = self.pre_process(info_dict, 'after_filter')
2729
2730 # The pre-processors may have modified the formats
2731 formats = self._get_formats(info_dict)
2732
2733 list_only = self.params.get('simulate') == 'list_only'
2734 interactive_format_selection = not list_only and self.format_selector == '-'
2735 if self.params.get('list_thumbnails'):
2736 self.list_thumbnails(info_dict)
2737 if self.params.get('listsubtitles'):
2738 if 'automatic_captions' in info_dict:
2739 self.list_subtitles(
2740 info_dict['id'], automatic_captions, 'automatic captions')
2741 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
2742 if self.params.get('listformats') or interactive_format_selection:
2743 self.list_formats(info_dict)
2744 if list_only:
2745 # Without this printing, -F --print-json will not work
2746 self.__forced_printings(info_dict)
2747 return info_dict
2748
2749 format_selector = self.format_selector
2750 while True:
2751 if interactive_format_selection:
2752 req_format = input(self._format_screen('\nEnter format selector ', self.Styles.EMPHASIS)
2753 + '(Press ENTER for default, or Ctrl+C to quit)'
2754 + self._format_screen(': ', self.Styles.EMPHASIS))
2755 try:
2756 format_selector = self.build_format_selector(req_format) if req_format else None
2757 except SyntaxError as err:
2758 self.report_error(err, tb=False, is_error=False)
2759 continue
2760
2761 if format_selector is None:
2762 req_format = self._default_format_spec(info_dict, download=download)
2763 self.write_debug(f'Default format spec: {req_format}')
2764 format_selector = self.build_format_selector(req_format)
2765
2766 formats_to_download = list(format_selector({
2767 'formats': formats,
2768 'has_merged_format': any('none' not in (f.get('acodec'), f.get('vcodec')) for f in formats),
2769 'incomplete_formats': (
2770 # All formats are video-only or
2771 all(f.get('vcodec') != 'none' and f.get('acodec') == 'none' for f in formats)
2772 # all formats are audio-only
2773 or all(f.get('vcodec') == 'none' and f.get('acodec') != 'none' for f in formats)),
2774 }))
2775 if interactive_format_selection and not formats_to_download:
2776 self.report_error('Requested format is not available', tb=False, is_error=False)
2777 continue
2778 break
2779
2780 if not formats_to_download:
2781 if not self.params.get('ignore_no_formats_error'):
2782 raise ExtractorError(
2783 'Requested format is not available. Use --list-formats for a list of available formats',
2784 expected=True, video_id=info_dict['id'], ie=info_dict['extractor'])
2785 self.report_warning('Requested format is not available')
2786 # Process what we can, even without any available formats.
2787 formats_to_download = [{}]
2788
2789 requested_ranges = tuple(self.params.get('download_ranges', lambda *_: [{}])(info_dict, self))
2790 best_format, downloaded_formats = formats_to_download[-1], []
2791 if download:
2792 if best_format and requested_ranges:
2793 def to_screen(*msg):
2794 self.to_screen(f'[info] {info_dict["id"]}: {" ".join(", ".join(variadic(m)) for m in msg)}')
2795
2796 to_screen(f'Downloading {len(formats_to_download)} format(s):',
2797 (f['format_id'] for f in formats_to_download))
2798 if requested_ranges != ({}, ):
2799 to_screen(f'Downloading {len(requested_ranges)} time ranges:',
2800 (f'{c["start_time"]:.1f}-{c["end_time"]:.1f}' for c in requested_ranges))
2801 max_downloads_reached = False
2802
2803 for fmt, chapter in itertools.product(formats_to_download, requested_ranges):
2804 new_info = self._copy_infodict(info_dict)
2805 new_info.update(fmt)
2806 offset, duration = info_dict.get('section_start') or 0, info_dict.get('duration') or float('inf')
2807 end_time = offset + min(chapter.get('end_time', duration), duration)
2808 # duration may not be accurate. So allow deviations <1sec
2809 if end_time == float('inf') or end_time > offset + duration + 1:
2810 end_time = None
2811 if chapter or offset:
2812 new_info.update({
2813 'section_start': offset + chapter.get('start_time', 0),
2814 'section_end': end_time,
2815 'section_title': chapter.get('title'),
2816 'section_number': chapter.get('index'),
2817 })
2818 downloaded_formats.append(new_info)
2819 try:
2820 self.process_info(new_info)
2821 except MaxDownloadsReached:
2822 max_downloads_reached = True
2823 self._raise_pending_errors(new_info)
2824 # Remove copied info
2825 for key, val in tuple(new_info.items()):
2826 if info_dict.get(key) == val:
2827 new_info.pop(key)
2828 if max_downloads_reached:
2829 break
2830
2831 write_archive = {f.get('__write_download_archive', False) for f in downloaded_formats}
2832 assert write_archive.issubset({True, False, 'ignore'})
2833 if True in write_archive and False not in write_archive:
2834 self.record_download_archive(info_dict)
2835
2836 info_dict['requested_downloads'] = downloaded_formats
2837 info_dict = self.run_all_pps('after_video', info_dict)
2838 if max_downloads_reached:
2839 raise MaxDownloadsReached()
2840
2841 # We update the info dict with the selected best quality format (backwards compatibility)
2842 info_dict.update(best_format)
2843 return info_dict
2844
2845 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
2846 """Select the requested subtitles and their format"""
2847 available_subs, normal_sub_langs = {}, []
2848 if normal_subtitles and self.params.get('writesubtitles'):
2849 available_subs.update(normal_subtitles)
2850 normal_sub_langs = tuple(normal_subtitles.keys())
2851 if automatic_captions and self.params.get('writeautomaticsub'):
2852 for lang, cap_info in automatic_captions.items():
2853 if lang not in available_subs:
2854 available_subs[lang] = cap_info
2855
2856 if not available_subs or (
2857 not self.params.get('writesubtitles')
2858 and not self.params.get('writeautomaticsub')):
2859 return None
2860
2861 all_sub_langs = tuple(available_subs.keys())
2862 if self.params.get('allsubtitles', False):
2863 requested_langs = all_sub_langs
2864 elif self.params.get('subtitleslangs', False):
2865 try:
2866 requested_langs = orderedSet_from_options(
2867 self.params.get('subtitleslangs'), {'all': all_sub_langs}, use_regex=True)
2868 except re.error as e:
2869 raise ValueError(f'Wrong regex for subtitlelangs: {e.pattern}')
2870 else:
2871 requested_langs = LazyList(itertools.chain(
2872 ['en'] if 'en' in normal_sub_langs else [],
2873 filter(lambda f: f.startswith('en'), normal_sub_langs),
2874 ['en'] if 'en' in all_sub_langs else [],
2875 filter(lambda f: f.startswith('en'), all_sub_langs),
2876 normal_sub_langs, all_sub_langs,
2877 ))[:1]
2878 if requested_langs:
2879 self.to_screen(f'[info] {video_id}: Downloading subtitles: {", ".join(requested_langs)}')
2880
2881 formats_query = self.params.get('subtitlesformat', 'best')
2882 formats_preference = formats_query.split('/') if formats_query else []
2883 subs = {}
2884 for lang in requested_langs:
2885 formats = available_subs.get(lang)
2886 if formats is None:
2887 self.report_warning(f'{lang} subtitles not available for {video_id}')
2888 continue
2889 for ext in formats_preference:
2890 if ext == 'best':
2891 f = formats[-1]
2892 break
2893 matches = list(filter(lambda f: f['ext'] == ext, formats))
2894 if matches:
2895 f = matches[-1]
2896 break
2897 else:
2898 f = formats[-1]
2899 self.report_warning(
2900 'No subtitle format found matching "%s" for language %s, '
2901 'using %s' % (formats_query, lang, f['ext']))
2902 subs[lang] = f
2903 return subs
2904
2905 def _forceprint(self, key, info_dict):
2906 if info_dict is None:
2907 return
2908 info_copy = info_dict.copy()
2909 info_copy.setdefault('filename', self.prepare_filename(info_dict))
2910 if info_dict.get('requested_formats') is not None:
2911 # For RTMP URLs, also include the playpath
2912 info_copy['urls'] = '\n'.join(f['url'] + f.get('play_path', '') for f in info_dict['requested_formats'])
2913 elif info_dict.get('url'):
2914 info_copy['urls'] = info_dict['url'] + info_dict.get('play_path', '')
2915 info_copy['formats_table'] = self.render_formats_table(info_dict)
2916 info_copy['thumbnails_table'] = self.render_thumbnails_table(info_dict)
2917 info_copy['subtitles_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('subtitles'))
2918 info_copy['automatic_captions_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('automatic_captions'))
2919
2920 def format_tmpl(tmpl):
2921 mobj = re.fullmatch(r'([\w.:,]|-\d|(?P<dict>{([\w.:,]|-\d)+}))+=?', tmpl)
2922 if not mobj:
2923 return tmpl
2924
2925 fmt = '%({})s'
2926 if tmpl.startswith('{'):
2927 tmpl, fmt = f'.{tmpl}', '%({})j'
2928 if tmpl.endswith('='):
2929 tmpl, fmt = tmpl[:-1], '{0} = %({0})#j'
2930 return '\n'.join(map(fmt.format, [tmpl] if mobj.group('dict') else tmpl.split(',')))
2931
2932 for tmpl in self.params['forceprint'].get(key, []):
2933 self.to_stdout(self.evaluate_outtmpl(format_tmpl(tmpl), info_copy))
2934
2935 for tmpl, file_tmpl in self.params['print_to_file'].get(key, []):
2936 filename = self.prepare_filename(info_dict, outtmpl=file_tmpl)
2937 tmpl = format_tmpl(tmpl)
2938 self.to_screen(f'[info] Writing {tmpl!r} to: {filename}')
2939 if self._ensure_dir_exists(filename):
2940 with open(filename, 'a', encoding='utf-8', newline='') as f:
2941 f.write(self.evaluate_outtmpl(tmpl, info_copy) + os.linesep)
2942
2943 return info_copy
2944
2945 def __forced_printings(self, info_dict, filename=None, incomplete=True):
2946 if (self.params.get('forcejson')
2947 or self.params['forceprint'].get('video')
2948 or self.params['print_to_file'].get('video')):
2949 self.post_extract(info_dict)
2950 if filename:
2951 info_dict['filename'] = filename
2952 info_copy = self._forceprint('video', info_dict)
2953
2954 def print_field(field, actual_field=None, optional=False):
2955 if actual_field is None:
2956 actual_field = field
2957 if self.params.get(f'force{field}') and (
2958 info_copy.get(field) is not None or (not optional and not incomplete)):
2959 self.to_stdout(info_copy[actual_field])
2960
2961 print_field('title')
2962 print_field('id')
2963 print_field('url', 'urls')
2964 print_field('thumbnail', optional=True)
2965 print_field('description', optional=True)
2966 print_field('filename')
2967 if self.params.get('forceduration') and info_copy.get('duration') is not None:
2968 self.to_stdout(formatSeconds(info_copy['duration']))
2969 print_field('format')
2970
2971 if self.params.get('forcejson'):
2972 self.to_stdout(json.dumps(self.sanitize_info(info_dict)))
2973
2974 def dl(self, name, info, subtitle=False, test=False):
2975 if not info.get('url'):
2976 self.raise_no_formats(info, True)
2977
2978 if test:
2979 verbose = self.params.get('verbose')
2980 params = {
2981 'test': True,
2982 'quiet': self.params.get('quiet') or not verbose,
2983 'verbose': verbose,
2984 'noprogress': not verbose,
2985 'nopart': True,
2986 'skip_unavailable_fragments': False,
2987 'keep_fragments': False,
2988 'overwrites': True,
2989 '_no_ytdl_file': True,
2990 }
2991 else:
2992 params = self.params
2993 fd = get_suitable_downloader(info, params, to_stdout=(name == '-'))(self, params)
2994 if not test:
2995 for ph in self._progress_hooks:
2996 fd.add_progress_hook(ph)
2997 urls = '", "'.join(
2998 (f['url'].split(',')[0] + ',<data>' if f['url'].startswith('data:') else f['url'])
2999 for f in info.get('requested_formats', []) or [info])
3000 self.write_debug(f'Invoking {fd.FD_NAME} downloader on "{urls}"')
3001
3002 # Note: Ideally info should be a deep-copied so that hooks cannot modify it.
3003 # But it may contain objects that are not deep-copyable
3004 new_info = self._copy_infodict(info)
3005 if new_info.get('http_headers') is None:
3006 new_info['http_headers'] = self._calc_headers(new_info)
3007 return fd.download(name, new_info, subtitle)
3008
3009 def existing_file(self, filepaths, *, default_overwrite=True):
3010 existing_files = list(filter(os.path.exists, orderedSet(filepaths)))
3011 if existing_files and not self.params.get('overwrites', default_overwrite):
3012 return existing_files[0]
3013
3014 for file in existing_files:
3015 self.report_file_delete(file)
3016 os.remove(file)
3017 return None
3018
3019 def process_info(self, info_dict):
3020 """Process a single resolved IE result. (Modifies it in-place)"""
3021
3022 assert info_dict.get('_type', 'video') == 'video'
3023 original_infodict = info_dict
3024
3025 if 'format' not in info_dict and 'ext' in info_dict:
3026 info_dict['format'] = info_dict['ext']
3027
3028 if self._match_entry(info_dict) is not None:
3029 info_dict['__write_download_archive'] = 'ignore'
3030 return
3031
3032 # Does nothing under normal operation - for backward compatibility of process_info
3033 self.post_extract(info_dict)
3034
3035 def replace_info_dict(new_info):
3036 nonlocal info_dict
3037 if new_info == info_dict:
3038 return
3039 info_dict.clear()
3040 info_dict.update(new_info)
3041
3042 new_info, _ = self.pre_process(info_dict, 'video')
3043 replace_info_dict(new_info)
3044 self._num_downloads += 1
3045
3046 # info_dict['_filename'] needs to be set for backward compatibility
3047 info_dict['_filename'] = full_filename = self.prepare_filename(info_dict, warn=True)
3048 temp_filename = self.prepare_filename(info_dict, 'temp')
3049 files_to_move = {}
3050
3051 # Forced printings
3052 self.__forced_printings(info_dict, full_filename, incomplete=('format' not in info_dict))
3053
3054 def check_max_downloads():
3055 if self._num_downloads >= float(self.params.get('max_downloads') or 'inf'):
3056 raise MaxDownloadsReached()
3057
3058 if self.params.get('simulate'):
3059 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
3060 check_max_downloads()
3061 return
3062
3063 if full_filename is None:
3064 return
3065 if not self._ensure_dir_exists(encodeFilename(full_filename)):
3066 return
3067 if not self._ensure_dir_exists(encodeFilename(temp_filename)):
3068 return
3069
3070 if self._write_description('video', info_dict,
3071 self.prepare_filename(info_dict, 'description')) is None:
3072 return
3073
3074 sub_files = self._write_subtitles(info_dict, temp_filename)
3075 if sub_files is None:
3076 return
3077 files_to_move.update(dict(sub_files))
3078
3079 thumb_files = self._write_thumbnails(
3080 'video', info_dict, temp_filename, self.prepare_filename(info_dict, 'thumbnail'))
3081 if thumb_files is None:
3082 return
3083 files_to_move.update(dict(thumb_files))
3084
3085 infofn = self.prepare_filename(info_dict, 'infojson')
3086 _infojson_written = self._write_info_json('video', info_dict, infofn)
3087 if _infojson_written:
3088 info_dict['infojson_filename'] = infofn
3089 # For backward compatibility, even though it was a private field
3090 info_dict['__infojson_filename'] = infofn
3091 elif _infojson_written is None:
3092 return
3093
3094 # Note: Annotations are deprecated
3095 annofn = None
3096 if self.params.get('writeannotations', False):
3097 annofn = self.prepare_filename(info_dict, 'annotation')
3098 if annofn:
3099 if not self._ensure_dir_exists(encodeFilename(annofn)):
3100 return
3101 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(annofn)):
3102 self.to_screen('[info] Video annotations are already present')
3103 elif not info_dict.get('annotations'):
3104 self.report_warning('There are no annotations to write.')
3105 else:
3106 try:
3107 self.to_screen('[info] Writing video annotations to: ' + annofn)
3108 with open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
3109 annofile.write(info_dict['annotations'])
3110 except (KeyError, TypeError):
3111 self.report_warning('There are no annotations to write.')
3112 except OSError:
3113 self.report_error('Cannot write annotations file: ' + annofn)
3114 return
3115
3116 # Write internet shortcut files
3117 def _write_link_file(link_type):
3118 url = try_get(info_dict['webpage_url'], iri_to_uri)
3119 if not url:
3120 self.report_warning(
3121 f'Cannot write internet shortcut file because the actual URL of "{info_dict["webpage_url"]}" is unknown')
3122 return True
3123 linkfn = replace_extension(self.prepare_filename(info_dict, 'link'), link_type, info_dict.get('ext'))
3124 if not self._ensure_dir_exists(encodeFilename(linkfn)):
3125 return False
3126 if self.params.get('overwrites', True) and os.path.exists(encodeFilename(linkfn)):
3127 self.to_screen(f'[info] Internet shortcut (.{link_type}) is already present')
3128 return True
3129 try:
3130 self.to_screen(f'[info] Writing internet shortcut (.{link_type}) to: {linkfn}')
3131 with open(encodeFilename(to_high_limit_path(linkfn)), 'w', encoding='utf-8',
3132 newline='\r\n' if link_type == 'url' else '\n') as linkfile:
3133 template_vars = {'url': url}
3134 if link_type == 'desktop':
3135 template_vars['filename'] = linkfn[:-(len(link_type) + 1)]
3136 linkfile.write(LINK_TEMPLATES[link_type] % template_vars)
3137 except OSError:
3138 self.report_error(f'Cannot write internet shortcut {linkfn}')
3139 return False
3140 return True
3141
3142 write_links = {
3143 'url': self.params.get('writeurllink'),
3144 'webloc': self.params.get('writewebloclink'),
3145 'desktop': self.params.get('writedesktoplink'),
3146 }
3147 if self.params.get('writelink'):
3148 link_type = ('webloc' if sys.platform == 'darwin'
3149 else 'desktop' if sys.platform.startswith('linux')
3150 else 'url')
3151 write_links[link_type] = True
3152
3153 if any(should_write and not _write_link_file(link_type)
3154 for link_type, should_write in write_links.items()):
3155 return
3156
3157 new_info, files_to_move = self.pre_process(info_dict, 'before_dl', files_to_move)
3158 replace_info_dict(new_info)
3159
3160 if self.params.get('skip_download'):
3161 info_dict['filepath'] = temp_filename
3162 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
3163 info_dict['__files_to_move'] = files_to_move
3164 replace_info_dict(self.run_pp(MoveFilesAfterDownloadPP(self, False), info_dict))
3165 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
3166 else:
3167 # Download
3168 info_dict.setdefault('__postprocessors', [])
3169 try:
3170
3171 def existing_video_file(*filepaths):
3172 ext = info_dict.get('ext')
3173 converted = lambda file: replace_extension(file, self.params.get('final_ext') or ext, ext)
3174 file = self.existing_file(itertools.chain(*zip(map(converted, filepaths), filepaths)),
3175 default_overwrite=False)
3176 if file:
3177 info_dict['ext'] = os.path.splitext(file)[1][1:]
3178 return file
3179
3180 fd, success = None, True
3181 if info_dict.get('protocol') or info_dict.get('url'):
3182 fd = get_suitable_downloader(info_dict, self.params, to_stdout=temp_filename == '-')
3183 if fd is not FFmpegFD and 'no-direct-merge' not in self.params['compat_opts'] and (
3184 info_dict.get('section_start') or info_dict.get('section_end')):
3185 msg = ('This format cannot be partially downloaded' if FFmpegFD.available()
3186 else 'You have requested downloading the video partially, but ffmpeg is not installed')
3187 self.report_error(f'{msg}. Aborting')
3188 return
3189
3190 if info_dict.get('requested_formats') is not None:
3191 old_ext = info_dict['ext']
3192 if self.params.get('merge_output_format') is None:
3193 if (info_dict['ext'] == 'webm'
3194 and info_dict.get('thumbnails')
3195 # check with type instead of pp_key, __name__, or isinstance
3196 # since we dont want any custom PPs to trigger this
3197 and any(type(pp) == EmbedThumbnailPP for pp in self._pps['post_process'])): # noqa: E721
3198 info_dict['ext'] = 'mkv'
3199 self.report_warning(
3200 'webm doesn\'t support embedding a thumbnail, mkv will be used')
3201 new_ext = info_dict['ext']
3202
3203 def correct_ext(filename, ext=new_ext):
3204 if filename == '-':
3205 return filename
3206 filename_real_ext = os.path.splitext(filename)[1][1:]
3207 filename_wo_ext = (
3208 os.path.splitext(filename)[0]
3209 if filename_real_ext in (old_ext, new_ext)
3210 else filename)
3211 return f'{filename_wo_ext}.{ext}'
3212
3213 # Ensure filename always has a correct extension for successful merge
3214 full_filename = correct_ext(full_filename)
3215 temp_filename = correct_ext(temp_filename)
3216 dl_filename = existing_video_file(full_filename, temp_filename)
3217
3218 info_dict['__real_download'] = False
3219 # NOTE: Copy so that original format dicts are not modified
3220 info_dict['requested_formats'] = list(map(dict, info_dict['requested_formats']))
3221
3222 merger = FFmpegMergerPP(self)
3223 downloaded = []
3224 if dl_filename is not None:
3225 self.report_file_already_downloaded(dl_filename)
3226 elif fd:
3227 for f in info_dict['requested_formats'] if fd != FFmpegFD else []:
3228 f['filepath'] = fname = prepend_extension(
3229 correct_ext(temp_filename, info_dict['ext']),
3230 'f%s' % f['format_id'], info_dict['ext'])
3231 downloaded.append(fname)
3232 info_dict['url'] = '\n'.join(f['url'] for f in info_dict['requested_formats'])
3233 success, real_download = self.dl(temp_filename, info_dict)
3234 info_dict['__real_download'] = real_download
3235 else:
3236 if self.params.get('allow_unplayable_formats'):
3237 self.report_warning(
3238 'You have requested merging of multiple formats '
3239 'while also allowing unplayable formats to be downloaded. '
3240 'The formats won\'t be merged to prevent data corruption.')
3241 elif not merger.available:
3242 msg = 'You have requested merging of multiple formats but ffmpeg is not installed'
3243 if not self.params.get('ignoreerrors'):
3244 self.report_error(f'{msg}. Aborting due to --abort-on-error')
3245 return
3246 self.report_warning(f'{msg}. The formats won\'t be merged')
3247
3248 if temp_filename == '-':
3249 reason = ('using a downloader other than ffmpeg' if FFmpegFD.can_merge_formats(info_dict, self.params)
3250 else 'but the formats are incompatible for simultaneous download' if merger.available
3251 else 'but ffmpeg is not installed')
3252 self.report_warning(
3253 f'You have requested downloading multiple formats to stdout {reason}. '
3254 'The formats will be streamed one after the other')
3255 fname = temp_filename
3256 for f in info_dict['requested_formats']:
3257 new_info = dict(info_dict)
3258 del new_info['requested_formats']
3259 new_info.update(f)
3260 if temp_filename != '-':
3261 fname = prepend_extension(
3262 correct_ext(temp_filename, new_info['ext']),
3263 'f%s' % f['format_id'], new_info['ext'])
3264 if not self._ensure_dir_exists(fname):
3265 return
3266 f['filepath'] = fname
3267 downloaded.append(fname)
3268 partial_success, real_download = self.dl(fname, new_info)
3269 info_dict['__real_download'] = info_dict['__real_download'] or real_download
3270 success = success and partial_success
3271
3272 if downloaded and merger.available and not self.params.get('allow_unplayable_formats'):
3273 info_dict['__postprocessors'].append(merger)
3274 info_dict['__files_to_merge'] = downloaded
3275 # Even if there were no downloads, it is being merged only now
3276 info_dict['__real_download'] = True
3277 else:
3278 for file in downloaded:
3279 files_to_move[file] = None
3280 else:
3281 # Just a single file
3282 dl_filename = existing_video_file(full_filename, temp_filename)
3283 if dl_filename is None or dl_filename == temp_filename:
3284 # dl_filename == temp_filename could mean that the file was partially downloaded with --no-part.
3285 # So we should try to resume the download
3286 success, real_download = self.dl(temp_filename, info_dict)
3287 info_dict['__real_download'] = real_download
3288 else:
3289 self.report_file_already_downloaded(dl_filename)
3290
3291 dl_filename = dl_filename or temp_filename
3292 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
3293
3294 except network_exceptions as err:
3295 self.report_error('unable to download video data: %s' % error_to_compat_str(err))
3296 return
3297 except OSError as err:
3298 raise UnavailableVideoError(err)
3299 except (ContentTooShortError, ) as err:
3300 self.report_error(f'content too short (expected {err.expected} bytes and served {err.downloaded})')
3301 return
3302
3303 self._raise_pending_errors(info_dict)
3304 if success and full_filename != '-':
3305
3306 def fixup():
3307 do_fixup = True
3308 fixup_policy = self.params.get('fixup')
3309 vid = info_dict['id']
3310
3311 if fixup_policy in ('ignore', 'never'):
3312 return
3313 elif fixup_policy == 'warn':
3314 do_fixup = 'warn'
3315 elif fixup_policy != 'force':
3316 assert fixup_policy in ('detect_or_warn', None)
3317 if not info_dict.get('__real_download'):
3318 do_fixup = False
3319
3320 def ffmpeg_fixup(cndn, msg, cls):
3321 if not (do_fixup and cndn):
3322 return
3323 elif do_fixup == 'warn':
3324 self.report_warning(f'{vid}: {msg}')
3325 return
3326 pp = cls(self)
3327 if pp.available:
3328 info_dict['__postprocessors'].append(pp)
3329 else:
3330 self.report_warning(f'{vid}: {msg}. Install ffmpeg to fix this automatically')
3331
3332 stretched_ratio = info_dict.get('stretched_ratio')
3333 ffmpeg_fixup(stretched_ratio not in (1, None),
3334 f'Non-uniform pixel ratio {stretched_ratio}',
3335 FFmpegFixupStretchedPP)
3336
3337 downloader = get_suitable_downloader(info_dict, self.params) if 'protocol' in info_dict else None
3338 downloader = downloader.FD_NAME if downloader else None
3339
3340 ext = info_dict.get('ext')
3341 postprocessed_by_ffmpeg = info_dict.get('requested_formats') or any((
3342 isinstance(pp, FFmpegVideoConvertorPP)
3343 and resolve_recode_mapping(ext, pp.mapping)[0] not in (ext, None)
3344 ) for pp in self._pps['post_process'])
3345
3346 if not postprocessed_by_ffmpeg:
3347 ffmpeg_fixup(ext == 'm4a' and info_dict.get('container') == 'm4a_dash',
3348 'writing DASH m4a. Only some players support this container',
3349 FFmpegFixupM4aPP)
3350 ffmpeg_fixup(downloader == 'hlsnative' and not self.params.get('hls_use_mpegts')
3351 or info_dict.get('is_live') and self.params.get('hls_use_mpegts') is None,
3352 'Possible MPEG-TS in MP4 container or malformed AAC timestamps',
3353 FFmpegFixupM3u8PP)
3354 ffmpeg_fixup(info_dict.get('is_live') and downloader == 'dashsegments',
3355 'Possible duplicate MOOV atoms', FFmpegFixupDuplicateMoovPP)
3356
3357 ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed timestamps detected', FFmpegFixupTimestampPP)
3358 ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed duration detected', FFmpegFixupDurationPP)
3359
3360 fixup()
3361 try:
3362 replace_info_dict(self.post_process(dl_filename, info_dict, files_to_move))
3363 except PostProcessingError as err:
3364 self.report_error('Postprocessing: %s' % str(err))
3365 return
3366 try:
3367 for ph in self._post_hooks:
3368 ph(info_dict['filepath'])
3369 except Exception as err:
3370 self.report_error('post hooks: %s' % str(err))
3371 return
3372 info_dict['__write_download_archive'] = True
3373
3374 assert info_dict is original_infodict # Make sure the info_dict was modified in-place
3375 if self.params.get('force_write_download_archive'):
3376 info_dict['__write_download_archive'] = True
3377 check_max_downloads()
3378
3379 def __download_wrapper(self, func):
3380 @functools.wraps(func)
3381 def wrapper(*args, **kwargs):
3382 try:
3383 res = func(*args, **kwargs)
3384 except UnavailableVideoError as e:
3385 self.report_error(e)
3386 except DownloadCancelled as e:
3387 self.to_screen(f'[info] {e}')
3388 if not self.params.get('break_per_url'):
3389 raise
3390 self._num_downloads = 0
3391 else:
3392 if self.params.get('dump_single_json', False):
3393 self.post_extract(res)
3394 self.to_stdout(json.dumps(self.sanitize_info(res)))
3395 return wrapper
3396
3397 def download(self, url_list):
3398 """Download a given list of URLs."""
3399 url_list = variadic(url_list) # Passing a single URL is a common mistake
3400 outtmpl = self.params['outtmpl']['default']
3401 if (len(url_list) > 1
3402 and outtmpl != '-'
3403 and '%' not in outtmpl
3404 and self.params.get('max_downloads') != 1):
3405 raise SameFileError(outtmpl)
3406
3407 for url in url_list:
3408 self.__download_wrapper(self.extract_info)(
3409 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
3410
3411 return self._download_retcode
3412
3413 def download_with_info_file(self, info_filename):
3414 with contextlib.closing(fileinput.FileInput(
3415 [info_filename], mode='r',
3416 openhook=fileinput.hook_encoded('utf-8'))) as f:
3417 # FileInput doesn't have a read method, we can't call json.load
3418 infos = [self.sanitize_info(info, self.params.get('clean_infojson', True))
3419 for info in variadic(json.loads('\n'.join(f)))]
3420 for info in infos:
3421 try:
3422 self.__download_wrapper(self.process_ie_result)(info, download=True)
3423 except (DownloadError, EntryNotInPlaylist, ReExtractInfo) as e:
3424 if not isinstance(e, EntryNotInPlaylist):
3425 self.to_stderr('\r')
3426 webpage_url = info.get('webpage_url')
3427 if webpage_url is None:
3428 raise
3429 self.report_warning(f'The info failed to download: {e}; trying with URL {webpage_url}')
3430 self.download([webpage_url])
3431 return self._download_retcode
3432
3433 @staticmethod
3434 def sanitize_info(info_dict, remove_private_keys=False):
3435 ''' Sanitize the infodict for converting to json '''
3436 if info_dict is None:
3437 return info_dict
3438 info_dict.setdefault('epoch', int(time.time()))
3439 info_dict.setdefault('_type', 'video')
3440 info_dict.setdefault('_version', {
3441 'version': __version__,
3442 'current_git_head': current_git_head(),
3443 'release_git_head': RELEASE_GIT_HEAD,
3444 'repository': REPOSITORY,
3445 })
3446
3447 if remove_private_keys:
3448 reject = lambda k, v: v is None or k.startswith('__') or k in {
3449 'requested_downloads', 'requested_formats', 'requested_subtitles', 'requested_entries',
3450 'entries', 'filepath', '_filename', 'filename', 'infojson_filename', 'original_url',
3451 'playlist_autonumber', '_format_sort_fields',
3452 }
3453 else:
3454 reject = lambda k, v: False
3455
3456 def filter_fn(obj):
3457 if isinstance(obj, dict):
3458 return {k: filter_fn(v) for k, v in obj.items() if not reject(k, v)}
3459 elif isinstance(obj, (list, tuple, set, LazyList)):
3460 return list(map(filter_fn, obj))
3461 elif obj is None or isinstance(obj, (str, int, float, bool)):
3462 return obj
3463 else:
3464 return repr(obj)
3465
3466 return filter_fn(info_dict)
3467
3468 @staticmethod
3469 def filter_requested_info(info_dict, actually_filter=True):
3470 ''' Alias of sanitize_info for backward compatibility '''
3471 return YoutubeDL.sanitize_info(info_dict, actually_filter)
3472
3473 def _delete_downloaded_files(self, *files_to_delete, info={}, msg=None):
3474 for filename in set(filter(None, files_to_delete)):
3475 if msg:
3476 self.to_screen(msg % filename)
3477 try:
3478 os.remove(filename)
3479 except OSError:
3480 self.report_warning(f'Unable to delete file {filename}')
3481 if filename in info.get('__files_to_move', []): # NB: Delete even if None
3482 del info['__files_to_move'][filename]
3483
3484 @staticmethod
3485 def post_extract(info_dict):
3486 def actual_post_extract(info_dict):
3487 if info_dict.get('_type') in ('playlist', 'multi_video'):
3488 for video_dict in info_dict.get('entries', {}):
3489 actual_post_extract(video_dict or {})
3490 return
3491
3492 post_extractor = info_dict.pop('__post_extractor', None) or (lambda: {})
3493 info_dict.update(post_extractor())
3494
3495 actual_post_extract(info_dict or {})
3496
3497 def run_pp(self, pp, infodict):
3498 files_to_delete = []
3499 if '__files_to_move' not in infodict:
3500 infodict['__files_to_move'] = {}
3501 try:
3502 files_to_delete, infodict = pp.run(infodict)
3503 except PostProcessingError as e:
3504 # Must be True and not 'only_download'
3505 if self.params.get('ignoreerrors') is True:
3506 self.report_error(e)
3507 return infodict
3508 raise
3509
3510 if not files_to_delete:
3511 return infodict
3512 if self.params.get('keepvideo', False):
3513 for f in files_to_delete:
3514 infodict['__files_to_move'].setdefault(f, '')
3515 else:
3516 self._delete_downloaded_files(
3517 *files_to_delete, info=infodict, msg='Deleting original file %s (pass -k to keep)')
3518 return infodict
3519
3520 def run_all_pps(self, key, info, *, additional_pps=None):
3521 if key != 'video':
3522 self._forceprint(key, info)
3523 for pp in (additional_pps or []) + self._pps[key]:
3524 info = self.run_pp(pp, info)
3525 return info
3526
3527 def pre_process(self, ie_info, key='pre_process', files_to_move=None):
3528 info = dict(ie_info)
3529 info['__files_to_move'] = files_to_move or {}
3530 try:
3531 info = self.run_all_pps(key, info)
3532 except PostProcessingError as err:
3533 msg = f'Preprocessing: {err}'
3534 info.setdefault('__pending_error', msg)
3535 self.report_error(msg, is_error=False)
3536 return info, info.pop('__files_to_move', None)
3537
3538 def post_process(self, filename, info, files_to_move=None):
3539 """Run all the postprocessors on the given file."""
3540 info['filepath'] = filename
3541 info['__files_to_move'] = files_to_move or {}
3542 info = self.run_all_pps('post_process', info, additional_pps=info.get('__postprocessors'))
3543 info = self.run_pp(MoveFilesAfterDownloadPP(self), info)
3544 del info['__files_to_move']
3545 return self.run_all_pps('after_move', info)
3546
3547 def _make_archive_id(self, info_dict):
3548 video_id = info_dict.get('id')
3549 if not video_id:
3550 return
3551 # Future-proof against any change in case
3552 # and backwards compatibility with prior versions
3553 extractor = info_dict.get('extractor_key') or info_dict.get('ie_key') # key in a playlist
3554 if extractor is None:
3555 url = str_or_none(info_dict.get('url'))
3556 if not url:
3557 return
3558 # Try to find matching extractor for the URL and take its ie_key
3559 for ie_key, ie in self._ies.items():
3560 if ie.suitable(url):
3561 extractor = ie_key
3562 break
3563 else:
3564 return
3565 return make_archive_id(extractor, video_id)
3566
3567 def in_download_archive(self, info_dict):
3568 if not self.archive:
3569 return False
3570
3571 vid_ids = [self._make_archive_id(info_dict)]
3572 vid_ids.extend(info_dict.get('_old_archive_ids') or [])
3573 return any(id_ in self.archive for id_ in vid_ids)
3574
3575 def record_download_archive(self, info_dict):
3576 fn = self.params.get('download_archive')
3577 if fn is None:
3578 return
3579 vid_id = self._make_archive_id(info_dict)
3580 assert vid_id
3581
3582 self.write_debug(f'Adding to archive: {vid_id}')
3583 if is_path_like(fn):
3584 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
3585 archive_file.write(vid_id + '\n')
3586 self.archive.add(vid_id)
3587
3588 @staticmethod
3589 def format_resolution(format, default='unknown'):
3590 if format.get('vcodec') == 'none' and format.get('acodec') != 'none':
3591 return 'audio only'
3592 if format.get('resolution') is not None:
3593 return format['resolution']
3594 if format.get('width') and format.get('height'):
3595 return '%dx%d' % (format['width'], format['height'])
3596 elif format.get('height'):
3597 return '%sp' % format['height']
3598 elif format.get('width'):
3599 return '%dx?' % format['width']
3600 return default
3601
3602 def _list_format_headers(self, *headers):
3603 if self.params.get('listformats_table', True) is not False:
3604 return [self._format_out(header, self.Styles.HEADERS) for header in headers]
3605 return headers
3606
3607 def _format_note(self, fdict):
3608 res = ''
3609 if fdict.get('ext') in ['f4f', 'f4m']:
3610 res += '(unsupported)'
3611 if fdict.get('language'):
3612 if res:
3613 res += ' '
3614 res += '[%s]' % fdict['language']
3615 if fdict.get('format_note') is not None:
3616 if res:
3617 res += ' '
3618 res += fdict['format_note']
3619 if fdict.get('tbr') is not None:
3620 if res:
3621 res += ', '
3622 res += '%4dk' % fdict['tbr']
3623 if fdict.get('container') is not None:
3624 if res:
3625 res += ', '
3626 res += '%s container' % fdict['container']
3627 if (fdict.get('vcodec') is not None
3628 and fdict.get('vcodec') != 'none'):
3629 if res:
3630 res += ', '
3631 res += fdict['vcodec']
3632 if fdict.get('vbr') is not None:
3633 res += '@'
3634 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
3635 res += 'video@'
3636 if fdict.get('vbr') is not None:
3637 res += '%4dk' % fdict['vbr']
3638 if fdict.get('fps') is not None:
3639 if res:
3640 res += ', '
3641 res += '%sfps' % fdict['fps']
3642 if fdict.get('acodec') is not None:
3643 if res:
3644 res += ', '
3645 if fdict['acodec'] == 'none':
3646 res += 'video only'
3647 else:
3648 res += '%-5s' % fdict['acodec']
3649 elif fdict.get('abr') is not None:
3650 if res:
3651 res += ', '
3652 res += 'audio'
3653 if fdict.get('abr') is not None:
3654 res += '@%3dk' % fdict['abr']
3655 if fdict.get('asr') is not None:
3656 res += ' (%5dHz)' % fdict['asr']
3657 if fdict.get('filesize') is not None:
3658 if res:
3659 res += ', '
3660 res += format_bytes(fdict['filesize'])
3661 elif fdict.get('filesize_approx') is not None:
3662 if res:
3663 res += ', '
3664 res += '~' + format_bytes(fdict['filesize_approx'])
3665 return res
3666
3667 def _get_formats(self, info_dict):
3668 if info_dict.get('formats') is None:
3669 if info_dict.get('url') and info_dict.get('_type', 'video') == 'video':
3670 return [info_dict]
3671 return []
3672 return info_dict['formats']
3673
3674 def render_formats_table(self, info_dict):
3675 formats = self._get_formats(info_dict)
3676 if not formats:
3677 return
3678 if not self.params.get('listformats_table', True) is not False:
3679 table = [
3680 [
3681 format_field(f, 'format_id'),
3682 format_field(f, 'ext'),
3683 self.format_resolution(f),
3684 self._format_note(f)
3685 ] for f in formats if (f.get('preference') or 0) >= -1000]
3686 return render_table(['format code', 'extension', 'resolution', 'note'], table, extra_gap=1)
3687
3688 def simplified_codec(f, field):
3689 assert field in ('acodec', 'vcodec')
3690 codec = f.get(field)
3691 if not codec:
3692 return 'unknown'
3693 elif codec != 'none':
3694 return '.'.join(codec.split('.')[:4])
3695
3696 if field == 'vcodec' and f.get('acodec') == 'none':
3697 return 'images'
3698 elif field == 'acodec' and f.get('vcodec') == 'none':
3699 return ''
3700 return self._format_out('audio only' if field == 'vcodec' else 'video only',
3701 self.Styles.SUPPRESS)
3702
3703 delim = self._format_out('\u2502', self.Styles.DELIM, '|', test_encoding=True)
3704 table = [
3705 [
3706 self._format_out(format_field(f, 'format_id'), self.Styles.ID),
3707 format_field(f, 'ext'),
3708 format_field(f, func=self.format_resolution, ignore=('audio only', 'images')),
3709 format_field(f, 'fps', '\t%d', func=round),
3710 format_field(f, 'dynamic_range', '%s', ignore=(None, 'SDR')).replace('HDR', ''),
3711 format_field(f, 'audio_channels', '\t%s'),
3712 delim, (
3713 format_field(f, 'filesize', ' \t%s', func=format_bytes)
3714 or format_field(f, 'filesize_approx', '≈\t%s', func=format_bytes)
3715 or format_field(try_call(lambda: format_bytes(int(info_dict['duration'] * f['tbr'] * (1024 / 8)))),
3716 None, self._format_out('~\t%s', self.Styles.SUPPRESS))),
3717 format_field(f, 'tbr', '\t%dk', func=round),
3718 shorten_protocol_name(f.get('protocol', '')),
3719 delim,
3720 simplified_codec(f, 'vcodec'),
3721 format_field(f, 'vbr', '\t%dk', func=round),
3722 simplified_codec(f, 'acodec'),
3723 format_field(f, 'abr', '\t%dk', func=round),
3724 format_field(f, 'asr', '\t%s', func=format_decimal_suffix),
3725 join_nonempty(
3726 self._format_out('UNSUPPORTED', 'light red') if f.get('ext') in ('f4f', 'f4m') else None,
3727 self._format_out('DRM', 'light red') if f.get('has_drm') else None,
3728 format_field(f, 'language', '[%s]'),
3729 join_nonempty(format_field(f, 'format_note'),
3730 format_field(f, 'container', ignore=(None, f.get('ext'))),
3731 delim=', '),
3732 delim=' '),
3733 ] for f in formats if f.get('preference') is None or f['preference'] >= -1000]
3734 header_line = self._list_format_headers(
3735 'ID', 'EXT', 'RESOLUTION', '\tFPS', 'HDR', 'CH', delim, '\tFILESIZE', '\tTBR', 'PROTO',
3736 delim, 'VCODEC', '\tVBR', 'ACODEC', '\tABR', '\tASR', 'MORE INFO')
3737
3738 return render_table(
3739 header_line, table, hide_empty=True,
3740 delim=self._format_out('\u2500', self.Styles.DELIM, '-', test_encoding=True))
3741
3742 def render_thumbnails_table(self, info_dict):
3743 thumbnails = list(info_dict.get('thumbnails') or [])
3744 if not thumbnails:
3745 return None
3746 return render_table(
3747 self._list_format_headers('ID', 'Width', 'Height', 'URL'),
3748 [[t.get('id'), t.get('width') or 'unknown', t.get('height') or 'unknown', t['url']] for t in thumbnails])
3749
3750 def render_subtitles_table(self, video_id, subtitles):
3751 def _row(lang, formats):
3752 exts, names = zip(*((f['ext'], f.get('name') or 'unknown') for f in reversed(formats)))
3753 if len(set(names)) == 1:
3754 names = [] if names[0] == 'unknown' else names[:1]
3755 return [lang, ', '.join(names), ', '.join(exts)]
3756
3757 if not subtitles:
3758 return None
3759 return render_table(
3760 self._list_format_headers('Language', 'Name', 'Formats'),
3761 [_row(lang, formats) for lang, formats in subtitles.items()],
3762 hide_empty=True)
3763
3764 def __list_table(self, video_id, name, func, *args):
3765 table = func(*args)
3766 if not table:
3767 self.to_screen(f'{video_id} has no {name}')
3768 return
3769 self.to_screen(f'[info] Available {name} for {video_id}:')
3770 self.to_stdout(table)
3771
3772 def list_formats(self, info_dict):
3773 self.__list_table(info_dict['id'], 'formats', self.render_formats_table, info_dict)
3774
3775 def list_thumbnails(self, info_dict):
3776 self.__list_table(info_dict['id'], 'thumbnails', self.render_thumbnails_table, info_dict)
3777
3778 def list_subtitles(self, video_id, subtitles, name='subtitles'):
3779 self.__list_table(video_id, name, self.render_subtitles_table, video_id, subtitles)
3780
3781 def urlopen(self, req):
3782 """ Start an HTTP download """
3783 if isinstance(req, str):
3784 req = sanitized_Request(req)
3785 return self._opener.open(req, timeout=self._socket_timeout)
3786
3787 def print_debug_header(self):
3788 if not self.params.get('verbose'):
3789 return
3790
3791 from . import _IN_CLI # Must be delayed import
3792
3793 # These imports can be slow. So import them only as needed
3794 from .extractor.extractors import _LAZY_LOADER
3795 from .extractor.extractors import (
3796 _PLUGIN_CLASSES as plugin_ies,
3797 _PLUGIN_OVERRIDES as plugin_ie_overrides
3798 )
3799
3800 def get_encoding(stream):
3801 ret = str(getattr(stream, 'encoding', 'missing (%s)' % type(stream).__name__))
3802 additional_info = []
3803 if os.environ.get('TERM', '').lower() == 'dumb':
3804 additional_info.append('dumb')
3805 if not supports_terminal_sequences(stream):
3806 from .utils import WINDOWS_VT_MODE # Must be imported locally
3807 additional_info.append('No VT' if WINDOWS_VT_MODE is False else 'No ANSI')
3808 if additional_info:
3809 ret = f'{ret} ({",".join(additional_info)})'
3810 return ret
3811
3812 encoding_str = 'Encodings: locale %s, fs %s, pref %s, %s' % (
3813 locale.getpreferredencoding(),
3814 sys.getfilesystemencoding(),
3815 self.get_encoding(),
3816 ', '.join(
3817 f'{key} {get_encoding(stream)}' for key, stream in self._out_files.items_
3818 if stream is not None and key != 'console')
3819 )
3820
3821 logger = self.params.get('logger')
3822 if logger:
3823 write_debug = lambda msg: logger.debug(f'[debug] {msg}')
3824 write_debug(encoding_str)
3825 else:
3826 write_string(f'[debug] {encoding_str}\n', encoding=None)
3827 write_debug = lambda msg: self._write_string(f'[debug] {msg}\n')
3828
3829 source = detect_variant()
3830 if VARIANT not in (None, 'pip'):
3831 source += '*'
3832 klass = type(self)
3833 write_debug(join_nonempty(
3834 f'{"yt-dlp" if REPOSITORY == "yt-dlp/yt-dlp" else REPOSITORY} version',
3835 f'{CHANNEL}@{__version__}',
3836 f'[{RELEASE_GIT_HEAD[:9]}]' if RELEASE_GIT_HEAD else '',
3837 '' if source == 'unknown' else f'({source})',
3838 '' if _IN_CLI else 'API' if klass == YoutubeDL else f'API:{self.__module__}.{klass.__qualname__}',
3839 delim=' '))
3840
3841 if not _IN_CLI:
3842 write_debug(f'params: {self.params}')
3843
3844 if not _LAZY_LOADER:
3845 if os.environ.get('YTDLP_NO_LAZY_EXTRACTORS'):
3846 write_debug('Lazy loading extractors is forcibly disabled')
3847 else:
3848 write_debug('Lazy loading extractors is disabled')
3849 if self.params['compat_opts']:
3850 write_debug('Compatibility options: %s' % ', '.join(self.params['compat_opts']))
3851
3852 if current_git_head():
3853 write_debug(f'Git HEAD: {current_git_head()}')
3854 write_debug(system_identifier())
3855
3856 exe_versions, ffmpeg_features = FFmpegPostProcessor.get_versions_and_features(self)
3857 ffmpeg_features = {key for key, val in ffmpeg_features.items() if val}
3858 if ffmpeg_features:
3859 exe_versions['ffmpeg'] += ' (%s)' % ','.join(sorted(ffmpeg_features))
3860
3861 exe_versions['rtmpdump'] = rtmpdump_version()
3862 exe_versions['phantomjs'] = PhantomJSwrapper._version()
3863 exe_str = ', '.join(
3864 f'{exe} {v}' for exe, v in sorted(exe_versions.items()) if v
3865 ) or 'none'
3866 write_debug('exe versions: %s' % exe_str)
3867
3868 from .compat.compat_utils import get_package_info
3869 from .dependencies import available_dependencies
3870
3871 write_debug('Optional libraries: %s' % (', '.join(sorted({
3872 join_nonempty(*get_package_info(m)) for m in available_dependencies.values()
3873 })) or 'none'))
3874
3875 self._setup_opener()
3876 proxy_map = {}
3877 for handler in self._opener.handlers:
3878 if hasattr(handler, 'proxies'):
3879 proxy_map.update(handler.proxies)
3880 write_debug(f'Proxy map: {proxy_map}')
3881
3882 for plugin_type, plugins in {'Extractor': plugin_ies, 'Post-Processor': plugin_pps}.items():
3883 display_list = ['%s%s' % (
3884 klass.__name__, '' if klass.__name__ == name else f' as {name}')
3885 for name, klass in plugins.items()]
3886 if plugin_type == 'Extractor':
3887 display_list.extend(f'{plugins[-1].IE_NAME.partition("+")[2]} ({parent.__name__})'
3888 for parent, plugins in plugin_ie_overrides.items())
3889 if not display_list:
3890 continue
3891 write_debug(f'{plugin_type} Plugins: {", ".join(sorted(display_list))}')
3892
3893 plugin_dirs = plugin_directories()
3894 if plugin_dirs:
3895 write_debug(f'Plugin directories: {plugin_dirs}')
3896
3897 # Not implemented
3898 if False and self.params.get('call_home'):
3899 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode()
3900 write_debug('Public IP address: %s' % ipaddr)
3901 latest_version = self.urlopen(
3902 'https://yt-dl.org/latest/version').read().decode()
3903 if version_tuple(latest_version) > version_tuple(__version__):
3904 self.report_warning(
3905 'You are using an outdated version (newest version: %s)! '
3906 'See https://yt-dl.org/update if you need help updating.' %
3907 latest_version)
3908
3909 def _setup_opener(self):
3910 if hasattr(self, '_opener'):
3911 return
3912 timeout_val = self.params.get('socket_timeout')
3913 self._socket_timeout = 20 if timeout_val is None else float(timeout_val)
3914
3915 opts_cookiesfrombrowser = self.params.get('cookiesfrombrowser')
3916 opts_cookiefile = self.params.get('cookiefile')
3917 opts_proxy = self.params.get('proxy')
3918
3919 self.cookiejar = load_cookies(opts_cookiefile, opts_cookiesfrombrowser, self)
3920
3921 cookie_processor = YoutubeDLCookieProcessor(self.cookiejar)
3922 if opts_proxy is not None:
3923 if opts_proxy == '':
3924 proxies = {}
3925 else:
3926 proxies = {'http': opts_proxy, 'https': opts_proxy}
3927 else:
3928 proxies = urllib.request.getproxies()
3929 # Set HTTPS proxy to HTTP one if given (https://github.com/ytdl-org/youtube-dl/issues/805)
3930 if 'http' in proxies and 'https' not in proxies:
3931 proxies['https'] = proxies['http']
3932 proxy_handler = PerRequestProxyHandler(proxies)
3933
3934 debuglevel = 1 if self.params.get('debug_printtraffic') else 0
3935 https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
3936 ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
3937 redirect_handler = YoutubeDLRedirectHandler()
3938 data_handler = urllib.request.DataHandler()
3939
3940 # When passing our own FileHandler instance, build_opener won't add the
3941 # default FileHandler and allows us to disable the file protocol, which
3942 # can be used for malicious purposes (see
3943 # https://github.com/ytdl-org/youtube-dl/issues/8227)
3944 file_handler = urllib.request.FileHandler()
3945
3946 if not self.params.get('enable_file_urls'):
3947 def file_open(*args, **kwargs):
3948 raise urllib.error.URLError(
3949 'file:// URLs are explicitly disabled in yt-dlp for security reasons. '
3950 'Use --enable-file-urls to enable at your own risk.')
3951 file_handler.file_open = file_open
3952
3953 opener = urllib.request.build_opener(
3954 proxy_handler, https_handler, cookie_processor, ydlh, redirect_handler, data_handler, file_handler)
3955
3956 # Delete the default user-agent header, which would otherwise apply in
3957 # cases where our custom HTTP handler doesn't come into play
3958 # (See https://github.com/ytdl-org/youtube-dl/issues/1309 for details)
3959 opener.addheaders = []
3960 self._opener = opener
3961
3962 def encode(self, s):
3963 if isinstance(s, bytes):
3964 return s # Already encoded
3965
3966 try:
3967 return s.encode(self.get_encoding())
3968 except UnicodeEncodeError as err:
3969 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
3970 raise
3971
3972 def get_encoding(self):
3973 encoding = self.params.get('encoding')
3974 if encoding is None:
3975 encoding = preferredencoding()
3976 return encoding
3977
3978 def _write_info_json(self, label, ie_result, infofn, overwrite=None):
3979 ''' Write infojson and returns True = written, 'exists' = Already exists, False = skip, None = error '''
3980 if overwrite is None:
3981 overwrite = self.params.get('overwrites', True)
3982 if not self.params.get('writeinfojson'):
3983 return False
3984 elif not infofn:
3985 self.write_debug(f'Skipping writing {label} infojson')
3986 return False
3987 elif not self._ensure_dir_exists(infofn):
3988 return None
3989 elif not overwrite and os.path.exists(infofn):
3990 self.to_screen(f'[info] {label.title()} metadata is already present')
3991 return 'exists'
3992
3993 self.to_screen(f'[info] Writing {label} metadata as JSON to: {infofn}')
3994 try:
3995 write_json_file(self.sanitize_info(ie_result, self.params.get('clean_infojson', True)), infofn)
3996 return True
3997 except OSError:
3998 self.report_error(f'Cannot write {label} metadata to JSON file {infofn}')
3999 return None
4000
4001 def _write_description(self, label, ie_result, descfn):
4002 ''' Write description and returns True = written, False = skip, None = error '''
4003 if not self.params.get('writedescription'):
4004 return False
4005 elif not descfn:
4006 self.write_debug(f'Skipping writing {label} description')
4007 return False
4008 elif not self._ensure_dir_exists(descfn):
4009 return None
4010 elif not self.params.get('overwrites', True) and os.path.exists(descfn):
4011 self.to_screen(f'[info] {label.title()} description is already present')
4012 elif ie_result.get('description') is None:
4013 self.to_screen(f'[info] There\'s no {label} description to write')
4014 return False
4015 else:
4016 try:
4017 self.to_screen(f'[info] Writing {label} description to: {descfn}')
4018 with open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
4019 descfile.write(ie_result['description'])
4020 except OSError:
4021 self.report_error(f'Cannot write {label} description file {descfn}')
4022 return None
4023 return True
4024
4025 def _write_subtitles(self, info_dict, filename):
4026 ''' Write subtitles to file and return list of (sub_filename, final_sub_filename); or None if error'''
4027 ret = []
4028 subtitles = info_dict.get('requested_subtitles')
4029 if not (self.params.get('writesubtitles') or self.params.get('writeautomaticsub')):
4030 # subtitles download errors are already managed as troubles in relevant IE
4031 # that way it will silently go on when used with unsupporting IE
4032 return ret
4033 elif not subtitles:
4034 self.to_screen('[info] There are no subtitles for the requested languages')
4035 return ret
4036 sub_filename_base = self.prepare_filename(info_dict, 'subtitle')
4037 if not sub_filename_base:
4038 self.to_screen('[info] Skipping writing video subtitles')
4039 return ret
4040
4041 for sub_lang, sub_info in subtitles.items():
4042 sub_format = sub_info['ext']
4043 sub_filename = subtitles_filename(filename, sub_lang, sub_format, info_dict.get('ext'))
4044 sub_filename_final = subtitles_filename(sub_filename_base, sub_lang, sub_format, info_dict.get('ext'))
4045 existing_sub = self.existing_file((sub_filename_final, sub_filename))
4046 if existing_sub:
4047 self.to_screen(f'[info] Video subtitle {sub_lang}.{sub_format} is already present')
4048 sub_info['filepath'] = existing_sub
4049 ret.append((existing_sub, sub_filename_final))
4050 continue
4051
4052 self.to_screen(f'[info] Writing video subtitles to: {sub_filename}')
4053 if sub_info.get('data') is not None:
4054 try:
4055 # Use newline='' to prevent conversion of newline characters
4056 # See https://github.com/ytdl-org/youtube-dl/issues/10268
4057 with open(sub_filename, 'w', encoding='utf-8', newline='') as subfile:
4058 subfile.write(sub_info['data'])
4059 sub_info['filepath'] = sub_filename
4060 ret.append((sub_filename, sub_filename_final))
4061 continue
4062 except OSError:
4063 self.report_error(f'Cannot write video subtitles file {sub_filename}')
4064 return None
4065
4066 try:
4067 sub_copy = sub_info.copy()
4068 sub_copy.setdefault('http_headers', info_dict.get('http_headers'))
4069 self.dl(sub_filename, sub_copy, subtitle=True)
4070 sub_info['filepath'] = sub_filename
4071 ret.append((sub_filename, sub_filename_final))
4072 except (DownloadError, ExtractorError, IOError, OSError, ValueError) + network_exceptions as err:
4073 msg = f'Unable to download video subtitles for {sub_lang!r}: {err}'
4074 if self.params.get('ignoreerrors') is not True: # False or 'only_download'
4075 if not self.params.get('ignoreerrors'):
4076 self.report_error(msg)
4077 raise DownloadError(msg)
4078 self.report_warning(msg)
4079 return ret
4080
4081 def _write_thumbnails(self, label, info_dict, filename, thumb_filename_base=None):
4082 ''' Write thumbnails to file and return list of (thumb_filename, final_thumb_filename) '''
4083 write_all = self.params.get('write_all_thumbnails', False)
4084 thumbnails, ret = [], []
4085 if write_all or self.params.get('writethumbnail', False):
4086 thumbnails = info_dict.get('thumbnails') or []
4087 if not thumbnails:
4088 self.to_screen(f'[info] There are no {label} thumbnails to download')
4089 return ret
4090 multiple = write_all and len(thumbnails) > 1
4091
4092 if thumb_filename_base is None:
4093 thumb_filename_base = filename
4094 if thumbnails and not thumb_filename_base:
4095 self.write_debug(f'Skipping writing {label} thumbnail')
4096 return ret
4097
4098 for idx, t in list(enumerate(thumbnails))[::-1]:
4099 thumb_ext = (f'{t["id"]}.' if multiple else '') + determine_ext(t['url'], 'jpg')
4100 thumb_display_id = f'{label} thumbnail {t["id"]}'
4101 thumb_filename = replace_extension(filename, thumb_ext, info_dict.get('ext'))
4102 thumb_filename_final = replace_extension(thumb_filename_base, thumb_ext, info_dict.get('ext'))
4103
4104 existing_thumb = self.existing_file((thumb_filename_final, thumb_filename))
4105 if existing_thumb:
4106 self.to_screen('[info] %s is already present' % (
4107 thumb_display_id if multiple else f'{label} thumbnail').capitalize())
4108 t['filepath'] = existing_thumb
4109 ret.append((existing_thumb, thumb_filename_final))
4110 else:
4111 self.to_screen(f'[info] Downloading {thumb_display_id} ...')
4112 try:
4113 uf = self.urlopen(sanitized_Request(t['url'], headers=t.get('http_headers', {})))
4114 self.to_screen(f'[info] Writing {thumb_display_id} to: {thumb_filename}')
4115 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
4116 shutil.copyfileobj(uf, thumbf)
4117 ret.append((thumb_filename, thumb_filename_final))
4118 t['filepath'] = thumb_filename
4119 except network_exceptions as err:
4120 if isinstance(err, urllib.error.HTTPError) and err.code == 404:
4121 self.to_screen(f'[info] {thumb_display_id.title()} does not exist')
4122 else:
4123 self.report_warning(f'Unable to download {thumb_display_id}: {err}')
4124 thumbnails.pop(idx)
4125 if ret and not write_all:
4126 break
4127 return ret