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