]> jfr.im git - yt-dlp.git/blob - yt_dlp/YoutubeDL.py
[core] Change how `Cookie` headers are handled
[yt-dlp.git] / yt_dlp / YoutubeDL.py
1 import collections
2 import contextlib
3 import copy
4 import datetime
5 import errno
6 import fileinput
7 import functools
8 import http.cookiejar
9 import io
10 import itertools
11 import json
12 import locale
13 import operator
14 import os
15 import random
16 import re
17 import shutil
18 import string
19 import subprocess
20 import sys
21 import tempfile
22 import time
23 import tokenize
24 import traceback
25 import unicodedata
26
27 from .cache import Cache
28 from .compat import urllib # isort: split
29 from .compat import compat_os_name, compat_shlex_quote
30 from .cookies import LenientSimpleCookie, load_cookies
31 from .downloader import FFmpegFD, get_suitable_downloader, shorten_protocol_name
32 from .downloader.rtmp import rtmpdump_version
33 from .extractor import gen_extractor_classes, get_info_extractor
34 from .extractor.common import UnsupportedURLIE
35 from .extractor.openload import PhantomJSwrapper
36 from .minicurses import format_text
37 from .plugins import directories as plugin_directories
38 from .postprocessor import _PLUGIN_CLASSES as plugin_pps
39 from .postprocessor import (
40 EmbedThumbnailPP,
41 FFmpegFixupDuplicateMoovPP,
42 FFmpegFixupDurationPP,
43 FFmpegFixupM3u8PP,
44 FFmpegFixupM4aPP,
45 FFmpegFixupStretchedPP,
46 FFmpegFixupTimestampPP,
47 FFmpegMergerPP,
48 FFmpegPostProcessor,
49 FFmpegVideoConvertorPP,
50 MoveFilesAfterDownloadPP,
51 get_postprocessor,
52 )
53 from .postprocessor.ffmpeg import resolve_mapping as resolve_recode_mapping
54 from .update import REPOSITORY, current_git_head, detect_variant
55 from .utils import (
56 DEFAULT_OUTTMPL,
57 IDENTITY,
58 LINK_TEMPLATES,
59 MEDIA_EXTENSIONS,
60 NO_DEFAULT,
61 NUMBER_RE,
62 OUTTMPL_TYPES,
63 POSTPROCESS_WHEN,
64 STR_FORMAT_RE_TMPL,
65 STR_FORMAT_TYPES,
66 ContentTooShortError,
67 DateRange,
68 DownloadCancelled,
69 DownloadError,
70 EntryNotInPlaylist,
71 ExistingVideoReached,
72 ExtractorError,
73 FormatSorter,
74 GeoRestrictedError,
75 HEADRequest,
76 ISO3166Utils,
77 LazyList,
78 MaxDownloadsReached,
79 Namespace,
80 PagedList,
81 PerRequestProxyHandler,
82 PlaylistEntries,
83 Popen,
84 PostProcessingError,
85 ReExtractInfo,
86 RejectedVideoReached,
87 SameFileError,
88 UnavailableVideoError,
89 UserNotLive,
90 YoutubeDLCookieProcessor,
91 YoutubeDLHandler,
92 YoutubeDLRedirectHandler,
93 age_restricted,
94 args_to_str,
95 bug_reports_message,
96 date_from_str,
97 deprecation_warning,
98 determine_ext,
99 determine_protocol,
100 encode_compat_str,
101 encodeFilename,
102 error_to_compat_str,
103 escapeHTML,
104 expand_path,
105 filter_dict,
106 float_or_none,
107 format_bytes,
108 format_decimal_suffix,
109 format_field,
110 formatSeconds,
111 get_compatible_ext,
112 get_domain,
113 int_or_none,
114 iri_to_uri,
115 is_path_like,
116 join_nonempty,
117 locked_file,
118 make_archive_id,
119 make_dir,
120 make_HTTPS_handler,
121 merge_headers,
122 network_exceptions,
123 number_of_digits,
124 orderedSet,
125 orderedSet_from_options,
126 parse_filesize,
127 preferredencoding,
128 prepend_extension,
129 remove_terminal_sequences,
130 render_table,
131 replace_extension,
132 sanitize_filename,
133 sanitize_path,
134 sanitize_url,
135 sanitized_Request,
136 std_headers,
137 str_or_none,
138 strftime_or_none,
139 subtitles_filename,
140 supports_terminal_sequences,
141 system_identifier,
142 timetuple_from_msec,
143 to_high_limit_path,
144 traverse_obj,
145 try_call,
146 try_get,
147 url_basename,
148 variadic,
149 version_tuple,
150 windows_enable_vt_mode,
151 write_json_file,
152 write_string,
153 )
154 from .version import CHANNEL, RELEASE_GIT_HEAD, VARIANT, __version__
155
156 if compat_os_name == 'nt':
157 import ctypes
158
159
160 class YoutubeDL:
161 """YoutubeDL class.
162
163 YoutubeDL objects are the ones responsible of downloading the
164 actual video file and writing it to disk if the user has requested
165 it, among some other tasks. In most cases there should be one per
166 program. As, given a video URL, the downloader doesn't know how to
167 extract all the needed information, task that InfoExtractors do, it
168 has to pass the URL to one of them.
169
170 For this, YoutubeDL objects have a method that allows
171 InfoExtractors to be registered in a given order. When it is passed
172 a URL, the YoutubeDL object handles it to the first InfoExtractor it
173 finds that reports being able to handle it. The InfoExtractor extracts
174 all the information about the video or videos the URL refers to, and
175 YoutubeDL process the extracted information, possibly using a File
176 Downloader to download the video.
177
178 YoutubeDL objects accept a lot of parameters. In order not to saturate
179 the object constructor with arguments, it receives a dictionary of
180 options instead. These options are available through the params
181 attribute for the InfoExtractors to use. The YoutubeDL also
182 registers itself as the downloader in charge for the InfoExtractors
183 that are added to it, so this is a "mutual registration".
184
185 Available options:
186
187 username: Username for authentication purposes.
188 password: Password for authentication purposes.
189 videopassword: Password for accessing a video.
190 ap_mso: Adobe Pass multiple-system operator identifier.
191 ap_username: Multiple-system operator account username.
192 ap_password: Multiple-system operator account password.
193 usenetrc: Use netrc for authentication instead.
194 netrc_location: Location of the netrc file. Defaults to ~/.netrc.
195 netrc_cmd: Use a shell command to get credentials
196 verbose: Print additional info to stdout.
197 quiet: Do not print messages to stdout.
198 no_warnings: Do not print out anything for warnings.
199 forceprint: A dict with keys WHEN mapped to a list of templates to
200 print to stdout. The allowed keys are video or any of the
201 items in utils.POSTPROCESS_WHEN.
202 For compatibility, a single list is also accepted
203 print_to_file: A dict with keys WHEN (same as forceprint) mapped to
204 a list of tuples with (template, filename)
205 forcejson: Force printing info_dict as JSON.
206 dump_single_json: Force printing the info_dict of the whole playlist
207 (or video) as a single JSON line.
208 force_write_download_archive: Force writing download archive regardless
209 of 'skip_download' or 'simulate'.
210 simulate: Do not download the video files. If unset (or None),
211 simulate only if listsubtitles, listformats or list_thumbnails is used
212 format: Video format code. see "FORMAT SELECTION" for more details.
213 You can also pass a function. The function takes 'ctx' as
214 argument and returns the formats to download.
215 See "build_format_selector" for an implementation
216 allow_unplayable_formats: Allow unplayable formats to be extracted and downloaded.
217 ignore_no_formats_error: Ignore "No video formats" error. Usefull for
218 extracting metadata even if the video is not actually
219 available for download (experimental)
220 format_sort: A list of fields by which to sort the video formats.
221 See "Sorting Formats" for more details.
222 format_sort_force: Force the given format_sort. see "Sorting Formats"
223 for more details.
224 prefer_free_formats: Whether to prefer video formats with free containers
225 over non-free ones of same quality.
226 allow_multiple_video_streams: Allow multiple video streams to be merged
227 into a single file
228 allow_multiple_audio_streams: Allow multiple audio streams to be merged
229 into a single file
230 check_formats Whether to test if the formats are downloadable.
231 Can be True (check all), False (check none),
232 'selected' (check selected formats),
233 or None (check only if requested by extractor)
234 paths: Dictionary of output paths. The allowed keys are 'home'
235 'temp' and the keys of OUTTMPL_TYPES (in utils.py)
236 outtmpl: Dictionary of templates for output names. Allowed keys
237 are 'default' and the keys of OUTTMPL_TYPES (in utils.py).
238 For compatibility with youtube-dl, a single string can also be used
239 outtmpl_na_placeholder: Placeholder for unavailable meta fields.
240 restrictfilenames: Do not allow "&" and spaces in file names
241 trim_file_name: Limit length of filename (extension excluded)
242 windowsfilenames: Force the filenames to be windows compatible
243 ignoreerrors: Do not stop on download/postprocessing errors.
244 Can be 'only_download' to ignore only download errors.
245 Default is 'only_download' for CLI, but False for API
246 skip_playlist_after_errors: Number of allowed failures until the rest of
247 the playlist is skipped
248 allowed_extractors: List of regexes to match against extractor names that are allowed
249 overwrites: Overwrite all video and metadata files if True,
250 overwrite only non-video files if None
251 and don't overwrite any file if False
252 For compatibility with youtube-dl,
253 "nooverwrites" may also be used instead
254 playlist_items: Specific indices of playlist to download.
255 playlistrandom: Download playlist items in random order.
256 lazy_playlist: Process playlist entries as they are received.
257 matchtitle: Download only matching titles.
258 rejecttitle: Reject downloads for matching titles.
259 logger: Log messages to a logging.Logger instance.
260 logtostderr: Print everything to stderr instead of stdout.
261 consoletitle: Display progress in console window's titlebar.
262 writedescription: Write the video description to a .description file
263 writeinfojson: Write the video description to a .info.json file
264 clean_infojson: Remove internal metadata from the infojson
265 getcomments: Extract video comments. This will not be written to disk
266 unless writeinfojson is also given
267 writeannotations: Write the video annotations to a .annotations.xml file
268 writethumbnail: Write the thumbnail image to a file
269 allow_playlist_files: Whether to write playlists' description, infojson etc
270 also to disk when using the 'write*' options
271 write_all_thumbnails: Write all thumbnail formats to files
272 writelink: Write an internet shortcut file, depending on the
273 current platform (.url/.webloc/.desktop)
274 writeurllink: Write a Windows internet shortcut file (.url)
275 writewebloclink: Write a macOS internet shortcut file (.webloc)
276 writedesktoplink: Write a Linux internet shortcut file (.desktop)
277 writesubtitles: Write the video subtitles to a file
278 writeautomaticsub: Write the automatically generated subtitles to a file
279 listsubtitles: Lists all available subtitles for the video
280 subtitlesformat: The format code for subtitles
281 subtitleslangs: List of languages of the subtitles to download (can be regex).
282 The list may contain "all" to refer to all the available
283 subtitles. The language can be prefixed with a "-" to
284 exclude it from the requested languages, e.g. ['all', '-live_chat']
285 keepvideo: Keep the video file after post-processing
286 daterange: A utils.DateRange object, download only if the upload_date is in the range.
287 skip_download: Skip the actual download of the video file
288 cachedir: Location of the cache files in the filesystem.
289 False to disable filesystem cache.
290 noplaylist: Download single video instead of a playlist if in doubt.
291 age_limit: An integer representing the user's age in years.
292 Unsuitable videos for the given age are skipped.
293 min_views: An integer representing the minimum view count the video
294 must have in order to not be skipped.
295 Videos without view count information are always
296 downloaded. None for no limit.
297 max_views: An integer representing the maximum view count.
298 Videos that are more popular than that are not
299 downloaded.
300 Videos without view count information are always
301 downloaded. None for no limit.
302 download_archive: A set, or the name of a file where all downloads are recorded.
303 Videos already present in the file are not downloaded again.
304 break_on_existing: Stop the download process after attempting to download a
305 file that is in the archive.
306 break_per_url: Whether break_on_reject and break_on_existing
307 should act on each input URL as opposed to for the entire queue
308 cookiefile: File name or text stream from where cookies should be read and dumped to
309 cookiesfrombrowser: A tuple containing the name of the browser, the profile
310 name/path from where cookies are loaded, the name of the keyring,
311 and the container name, e.g. ('chrome', ) or
312 ('vivaldi', 'default', 'BASICTEXT') or ('firefox', 'default', None, 'Meta')
313 legacyserverconnect: Explicitly allow HTTPS connection to servers that do not
314 support RFC 5746 secure renegotiation
315 nocheckcertificate: Do not verify SSL certificates
316 client_certificate: Path to client certificate file in PEM format. May include the private key
317 client_certificate_key: Path to private key file for client certificate
318 client_certificate_password: Password for client certificate private key, if encrypted.
319 If not provided and the key is encrypted, yt-dlp will ask interactively
320 prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
321 (Only supported by some extractors)
322 enable_file_urls: Enable file:// URLs. This is disabled by default for security reasons.
323 http_headers: A dictionary of custom headers to be used for all requests
324 proxy: URL of the proxy server to use
325 geo_verification_proxy: URL of the proxy to use for IP address verification
326 on geo-restricted sites.
327 socket_timeout: Time to wait for unresponsive hosts, in seconds
328 bidi_workaround: Work around buggy terminals without bidirectional text
329 support, using fridibi
330 debug_printtraffic:Print out sent and received HTTP traffic
331 default_search: Prepend this string if an input url is not valid.
332 'auto' for elaborate guessing
333 encoding: Use this encoding instead of the system-specified.
334 extract_flat: Whether to resolve and process url_results further
335 * False: Always process. Default for API
336 * True: Never process
337 * 'in_playlist': Do not process inside playlist/multi_video
338 * 'discard': Always process, but don't return the result
339 from inside playlist/multi_video
340 * 'discard_in_playlist': Same as "discard", but only for
341 playlists (not multi_video). Default for CLI
342 wait_for_video: If given, wait for scheduled streams to become available.
343 The value should be a tuple containing the range
344 (min_secs, max_secs) to wait between retries
345 postprocessors: A list of dictionaries, each with an entry
346 * key: The name of the postprocessor. See
347 yt_dlp/postprocessor/__init__.py for a list.
348 * when: When to run the postprocessor. Allowed values are
349 the entries of utils.POSTPROCESS_WHEN
350 Assumed to be 'post_process' if not given
351 progress_hooks: A list of functions that get called on download
352 progress, with a dictionary with the entries
353 * status: One of "downloading", "error", or "finished".
354 Check this first and ignore unknown values.
355 * info_dict: The extracted info_dict
356
357 If status is one of "downloading", or "finished", the
358 following properties may also be present:
359 * filename: The final filename (always present)
360 * tmpfilename: The filename we're currently writing to
361 * downloaded_bytes: Bytes on disk
362 * total_bytes: Size of the whole file, None if unknown
363 * total_bytes_estimate: Guess of the eventual file size,
364 None if unavailable.
365 * elapsed: The number of seconds since download started.
366 * eta: The estimated time in seconds, None if unknown
367 * speed: The download speed in bytes/second, None if
368 unknown
369 * fragment_index: The counter of the currently
370 downloaded video fragment.
371 * fragment_count: The number of fragments (= individual
372 files that will be merged)
373
374 Progress hooks are guaranteed to be called at least once
375 (with status "finished") if the download is successful.
376 postprocessor_hooks: A list of functions that get called on postprocessing
377 progress, with a dictionary with the entries
378 * status: One of "started", "processing", or "finished".
379 Check this first and ignore unknown values.
380 * postprocessor: Name of the postprocessor
381 * info_dict: The extracted info_dict
382
383 Progress hooks are guaranteed to be called at least twice
384 (with status "started" and "finished") if the processing is successful.
385 merge_output_format: "/" separated list of extensions to use when merging formats.
386 final_ext: Expected final extension; used to detect when the file was
387 already downloaded and converted
388 fixup: Automatically correct known faults of the file.
389 One of:
390 - "never": do nothing
391 - "warn": only emit a warning
392 - "detect_or_warn": check whether we can do anything
393 about it, warn otherwise (default)
394 source_address: Client-side IP address to bind to.
395 sleep_interval_requests: Number of seconds to sleep between requests
396 during extraction
397 sleep_interval: Number of seconds to sleep before each download when
398 used alone or a lower bound of a range for randomized
399 sleep before each download (minimum possible number
400 of seconds to sleep) when used along with
401 max_sleep_interval.
402 max_sleep_interval:Upper bound of a range for randomized sleep before each
403 download (maximum possible number of seconds to sleep).
404 Must only be used along with sleep_interval.
405 Actual sleep time will be a random float from range
406 [sleep_interval; max_sleep_interval].
407 sleep_interval_subtitles: Number of seconds to sleep before each subtitle download
408 listformats: Print an overview of available video formats and exit.
409 list_thumbnails: Print a table of all thumbnails and exit.
410 match_filter: A function that gets called for every video with the signature
411 (info_dict, *, incomplete: bool) -> Optional[str]
412 For backward compatibility with youtube-dl, the signature
413 (info_dict) -> Optional[str] is also allowed.
414 - If it returns a message, the video is ignored.
415 - If it returns None, the video is downloaded.
416 - If it returns utils.NO_DEFAULT, the user is interactively
417 asked whether to download the video.
418 - Raise utils.DownloadCancelled(msg) to abort remaining
419 downloads when a video is rejected.
420 match_filter_func in utils.py is one example for this.
421 color: A Dictionary with output stream names as keys
422 and their respective color policy as values.
423 Can also just be a single color policy,
424 in which case it applies to all outputs.
425 Valid stream names are 'stdout' and 'stderr'.
426 Valid color policies are one of 'always', 'auto', 'no_color' or 'never'.
427 geo_bypass: Bypass geographic restriction via faking X-Forwarded-For
428 HTTP header
429 geo_bypass_country:
430 Two-letter ISO 3166-2 country code that will be used for
431 explicit geographic restriction bypassing via faking
432 X-Forwarded-For HTTP header
433 geo_bypass_ip_block:
434 IP range in CIDR notation that will be used similarly to
435 geo_bypass_country
436 external_downloader: A dictionary of protocol keys and the executable of the
437 external downloader to use for it. The allowed protocols
438 are default|http|ftp|m3u8|dash|rtsp|rtmp|mms.
439 Set the value to 'native' to use the native downloader
440 compat_opts: Compatibility options. See "Differences in default behavior".
441 The following options do not work when used through the API:
442 filename, abort-on-error, multistreams, no-live-chat, format-sort
443 no-clean-infojson, no-playlist-metafiles, no-keep-subs, no-attach-info-json.
444 Refer __init__.py for their implementation
445 progress_template: Dictionary of templates for progress outputs.
446 Allowed keys are 'download', 'postprocess',
447 'download-title' (console title) and 'postprocess-title'.
448 The template is mapped on a dictionary with keys 'progress' and 'info'
449 retry_sleep_functions: Dictionary of functions that takes the number of attempts
450 as argument and returns the time to sleep in seconds.
451 Allowed keys are 'http', 'fragment', 'file_access'
452 download_ranges: A callback function that gets called for every video with
453 the signature (info_dict, ydl) -> Iterable[Section].
454 Only the returned sections will be downloaded.
455 Each Section is a dict with the following keys:
456 * start_time: Start time of the section in seconds
457 * end_time: End time of the section in seconds
458 * title: Section title (Optional)
459 * index: Section number (Optional)
460 force_keyframes_at_cuts: Re-encode the video when downloading ranges to get precise cuts
461 noprogress: Do not print the progress bar
462 live_from_start: Whether to download livestreams videos from the start
463
464 The following parameters are not used by YoutubeDL itself, they are used by
465 the downloader (see yt_dlp/downloader/common.py):
466 nopart, updatetime, buffersize, ratelimit, throttledratelimit, min_filesize,
467 max_filesize, test, noresizebuffer, retries, file_access_retries, fragment_retries,
468 continuedl, xattr_set_filesize, hls_use_mpegts, http_chunk_size,
469 external_downloader_args, concurrent_fragment_downloads.
470
471 The following options are used by the post processors:
472 ffmpeg_location: Location of the ffmpeg/avconv binary; either the path
473 to the binary or its containing directory.
474 postprocessor_args: A dictionary of postprocessor/executable keys (in lower case)
475 and a list of additional command-line arguments for the
476 postprocessor/executable. The dict can also have "PP+EXE" keys
477 which are used when the given exe is used by the given PP.
478 Use 'default' as the name for arguments to passed to all PP
479 For compatibility with youtube-dl, a single list of args
480 can also be used
481
482 The following options are used by the extractors:
483 extractor_retries: Number of times to retry for known errors (default: 3)
484 dynamic_mpd: Whether to process dynamic DASH manifests (default: True)
485 hls_split_discontinuity: Split HLS playlists to different formats at
486 discontinuities such as ad breaks (default: False)
487 extractor_args: A dictionary of arguments to be passed to the extractors.
488 See "EXTRACTOR ARGUMENTS" for details.
489 E.g. {'youtube': {'skip': ['dash', 'hls']}}
490 mark_watched: Mark videos watched (even with --simulate). Only for YouTube
491
492 The following options are deprecated and may be removed in the future:
493
494 break_on_reject: Stop the download process when encountering a video that
495 has been filtered out.
496 - `raise DownloadCancelled(msg)` in match_filter instead
497 force_generic_extractor: Force downloader to use the generic extractor
498 - Use allowed_extractors = ['generic', 'default']
499 playliststart: - Use playlist_items
500 Playlist item to start at.
501 playlistend: - Use playlist_items
502 Playlist item to end at.
503 playlistreverse: - Use playlist_items
504 Download playlist items in reverse order.
505 forceurl: - Use forceprint
506 Force printing final URL.
507 forcetitle: - Use forceprint
508 Force printing title.
509 forceid: - Use forceprint
510 Force printing ID.
511 forcethumbnail: - Use forceprint
512 Force printing thumbnail URL.
513 forcedescription: - Use forceprint
514 Force printing description.
515 forcefilename: - Use forceprint
516 Force printing final filename.
517 forceduration: - Use forceprint
518 Force printing duration.
519 allsubtitles: - Use subtitleslangs = ['all']
520 Downloads all the subtitles of the video
521 (requires writesubtitles or writeautomaticsub)
522 include_ads: - Doesn't work
523 Download ads as well
524 call_home: - Not implemented
525 Boolean, true iff we are allowed to contact the
526 yt-dlp servers for debugging.
527 post_hooks: - Register a custom postprocessor
528 A list of functions that get called as the final step
529 for each video file, after all postprocessors have been
530 called. The filename will be passed as the only argument.
531 hls_prefer_native: - Use external_downloader = {'m3u8': 'native'} or {'m3u8': 'ffmpeg'}.
532 Use the native HLS downloader instead of ffmpeg/avconv
533 if True, otherwise use ffmpeg/avconv if False, otherwise
534 use downloader suggested by extractor if None.
535 prefer_ffmpeg: - avconv support is deprecated
536 If False, use avconv instead of ffmpeg if both are available,
537 otherwise prefer ffmpeg.
538 youtube_include_dash_manifest: - Use extractor_args
539 If True (default), DASH manifests and related
540 data will be downloaded and processed by extractor.
541 You can reduce network I/O by disabling it if you don't
542 care about DASH. (only for youtube)
543 youtube_include_hls_manifest: - Use extractor_args
544 If True (default), HLS manifests and related
545 data will be downloaded and processed by extractor.
546 You can reduce network I/O by disabling it if you don't
547 care about HLS. (only for youtube)
548 no_color: Same as `color='no_color'`
549 """
550
551 _NUMERIC_FIELDS = {
552 'width', 'height', 'asr', 'audio_channels', 'fps',
553 'tbr', 'abr', 'vbr', 'filesize', 'filesize_approx',
554 'timestamp', 'release_timestamp',
555 'duration', 'view_count', 'like_count', 'dislike_count', 'repost_count',
556 'average_rating', 'comment_count', 'age_limit',
557 'start_time', 'end_time',
558 'chapter_number', 'season_number', 'episode_number',
559 'track_number', 'disc_number', 'release_year',
560 }
561
562 _format_fields = {
563 # NB: Keep in sync with the docstring of extractor/common.py
564 'url', 'manifest_url', 'manifest_stream_number', 'ext', 'format', 'format_id', 'format_note',
565 'width', 'height', 'aspect_ratio', 'resolution', 'dynamic_range', 'tbr', 'abr', 'acodec', 'asr', 'audio_channels',
566 'vbr', 'fps', 'vcodec', 'container', 'filesize', 'filesize_approx', 'rows', 'columns',
567 'player_url', 'protocol', 'fragment_base_url', 'fragments', 'is_from_start',
568 'preference', 'language', 'language_preference', 'quality', 'source_preference',
569 'http_headers', 'stretched_ratio', 'no_resume', 'has_drm', 'extra_param_to_segment_url', 'hls_aes', 'downloader_options',
570 'page_url', 'app', 'play_path', 'tc_url', 'flash_version', 'rtmp_live', 'rtmp_conn', 'rtmp_protocol', 'rtmp_real_time'
571 }
572 _format_selection_exts = {
573 'audio': set(MEDIA_EXTENSIONS.common_audio),
574 'video': set(MEDIA_EXTENSIONS.common_video + ('3gp', )),
575 'storyboards': set(MEDIA_EXTENSIONS.storyboards),
576 }
577
578 def __init__(self, params=None, auto_init=True):
579 """Create a FileDownloader object with the given options.
580 @param auto_init Whether to load the default extractors and print header (if verbose).
581 Set to 'no_verbose_header' to not print the header
582 """
583 if params is None:
584 params = {}
585 self.params = params
586 self._ies = {}
587 self._ies_instances = {}
588 self._pps = {k: [] for k in POSTPROCESS_WHEN}
589 self._printed_messages = set()
590 self._first_webpage_request = True
591 self._post_hooks = []
592 self._progress_hooks = []
593 self._postprocessor_hooks = []
594 self._download_retcode = 0
595 self._num_downloads = 0
596 self._num_videos = 0
597 self._playlist_level = 0
598 self._playlist_urls = set()
599 self.cache = Cache(self)
600
601 stdout = sys.stderr if self.params.get('logtostderr') else sys.stdout
602 self._out_files = Namespace(
603 out=stdout,
604 error=sys.stderr,
605 screen=sys.stderr if self.params.get('quiet') else stdout,
606 console=None if compat_os_name == 'nt' else next(
607 filter(supports_terminal_sequences, (sys.stderr, sys.stdout)), None)
608 )
609
610 try:
611 windows_enable_vt_mode()
612 except Exception as e:
613 self.write_debug(f'Failed to enable VT mode: {e}')
614
615 if self.params.get('no_color'):
616 if self.params.get('color') is not None:
617 self.report_warning('Overwriting params from "color" with "no_color"')
618 self.params['color'] = 'no_color'
619
620 term_allow_color = os.environ.get('TERM', '').lower() != 'dumb'
621
622 def process_color_policy(stream):
623 stream_name = {sys.stdout: 'stdout', sys.stderr: 'stderr'}[stream]
624 policy = traverse_obj(self.params, ('color', (stream_name, None), {str}), get_all=False)
625 if policy in ('auto', None):
626 return term_allow_color and supports_terminal_sequences(stream)
627 assert policy in ('always', 'never', 'no_color')
628 return {'always': True, 'never': False}.get(policy, policy)
629
630 self._allow_colors = Namespace(**{
631 name: process_color_policy(stream)
632 for name, stream in self._out_files.items_ if name != 'console'
633 })
634
635 # The code is left like this to be reused for future deprecations
636 MIN_SUPPORTED, MIN_RECOMMENDED = (3, 7), (3, 7)
637 current_version = sys.version_info[:2]
638 if current_version < MIN_RECOMMENDED:
639 msg = ('Support for Python version %d.%d has been deprecated. '
640 'See https://github.com/yt-dlp/yt-dlp/issues/3764 for more details.'
641 '\n You will no longer receive updates on this version')
642 if current_version < MIN_SUPPORTED:
643 msg = 'Python version %d.%d is no longer supported'
644 self.deprecated_feature(
645 f'{msg}! Please update to Python %d.%d or above' % (*current_version, *MIN_RECOMMENDED))
646
647 if self.params.get('allow_unplayable_formats'):
648 self.report_warning(
649 f'You have asked for {self._format_err("UNPLAYABLE", self.Styles.EMPHASIS)} formats to be listed/downloaded. '
650 'This is a developer option intended for debugging. \n'
651 ' If you experience any issues while using this option, '
652 f'{self._format_err("DO NOT", self.Styles.ERROR)} open a bug report')
653
654 if self.params.get('bidi_workaround', False):
655 try:
656 import pty
657 master, slave = pty.openpty()
658 width = shutil.get_terminal_size().columns
659 width_args = [] if width is None else ['-w', str(width)]
660 sp_kwargs = {'stdin': subprocess.PIPE, 'stdout': slave, 'stderr': self._out_files.error}
661 try:
662 self._output_process = Popen(['bidiv'] + width_args, **sp_kwargs)
663 except OSError:
664 self._output_process = Popen(['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
665 self._output_channel = os.fdopen(master, 'rb')
666 except OSError as ose:
667 if ose.errno == errno.ENOENT:
668 self.report_warning(
669 'Could not find fribidi executable, ignoring --bidi-workaround. '
670 'Make sure that fribidi is an executable file in one of the directories in your $PATH.')
671 else:
672 raise
673
674 self.params['compat_opts'] = set(self.params.get('compat_opts', ()))
675 if auto_init and auto_init != 'no_verbose_header':
676 self.print_debug_header()
677
678 self.__header_cookies = []
679 self._load_cookies(traverse_obj(self.params.get('http_headers'), 'cookie', casesense=False)) # compat
680
681 def check_deprecated(param, option, suggestion):
682 if self.params.get(param) is not None:
683 self.report_warning(f'{option} is deprecated. Use {suggestion} instead')
684 return True
685 return False
686
687 if check_deprecated('cn_verification_proxy', '--cn-verification-proxy', '--geo-verification-proxy'):
688 if self.params.get('geo_verification_proxy') is None:
689 self.params['geo_verification_proxy'] = self.params['cn_verification_proxy']
690
691 check_deprecated('autonumber', '--auto-number', '-o "%(autonumber)s-%(title)s.%(ext)s"')
692 check_deprecated('usetitle', '--title', '-o "%(title)s-%(id)s.%(ext)s"')
693 check_deprecated('useid', '--id', '-o "%(id)s.%(ext)s"')
694
695 for msg in self.params.get('_warnings', []):
696 self.report_warning(msg)
697 for msg in self.params.get('_deprecation_warnings', []):
698 self.deprecated_feature(msg)
699
700 if 'list-formats' in self.params['compat_opts']:
701 self.params['listformats_table'] = False
702
703 if 'overwrites' not in self.params and self.params.get('nooverwrites') is not None:
704 # nooverwrites was unnecessarily changed to overwrites
705 # in 0c3d0f51778b153f65c21906031c2e091fcfb641
706 # This ensures compatibility with both keys
707 self.params['overwrites'] = not self.params['nooverwrites']
708 elif self.params.get('overwrites') is None:
709 self.params.pop('overwrites', None)
710 else:
711 self.params['nooverwrites'] = not self.params['overwrites']
712
713 if self.params.get('simulate') is None and any((
714 self.params.get('list_thumbnails'),
715 self.params.get('listformats'),
716 self.params.get('listsubtitles'),
717 )):
718 self.params['simulate'] = 'list_only'
719
720 self.params.setdefault('forceprint', {})
721 self.params.setdefault('print_to_file', {})
722
723 # Compatibility with older syntax
724 if not isinstance(params['forceprint'], dict):
725 self.params['forceprint'] = {'video': params['forceprint']}
726
727 if auto_init:
728 self.add_default_info_extractors()
729
730 if (sys.platform != 'win32'
731 and sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968']
732 and not self.params.get('restrictfilenames', False)):
733 # Unicode filesystem API will throw errors (#1474, #13027)
734 self.report_warning(
735 'Assuming --restrict-filenames since file system encoding '
736 'cannot encode all characters. '
737 'Set the LC_ALL environment variable to fix this.')
738 self.params['restrictfilenames'] = True
739
740 self._parse_outtmpl()
741
742 # Creating format selector here allows us to catch syntax errors before the extraction
743 self.format_selector = (
744 self.params.get('format') if self.params.get('format') in (None, '-')
745 else self.params['format'] if callable(self.params['format'])
746 else self.build_format_selector(self.params['format']))
747
748 # Set http_headers defaults according to std_headers
749 self.params['http_headers'] = merge_headers(std_headers, self.params.get('http_headers', {}))
750
751 hooks = {
752 'post_hooks': self.add_post_hook,
753 'progress_hooks': self.add_progress_hook,
754 'postprocessor_hooks': self.add_postprocessor_hook,
755 }
756 for opt, fn in hooks.items():
757 for ph in self.params.get(opt, []):
758 fn(ph)
759
760 for pp_def_raw in self.params.get('postprocessors', []):
761 pp_def = dict(pp_def_raw)
762 when = pp_def.pop('when', 'post_process')
763 self.add_post_processor(
764 get_postprocessor(pp_def.pop('key'))(self, **pp_def),
765 when=when)
766
767 self._setup_opener()
768
769 def preload_download_archive(fn):
770 """Preload the archive, if any is specified"""
771 archive = set()
772 if fn is None:
773 return archive
774 elif not is_path_like(fn):
775 return fn
776
777 self.write_debug(f'Loading archive file {fn!r}')
778 try:
779 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
780 for line in archive_file:
781 archive.add(line.strip())
782 except OSError as ioe:
783 if ioe.errno != errno.ENOENT:
784 raise
785 return archive
786
787 self.archive = preload_download_archive(self.params.get('download_archive'))
788
789 def warn_if_short_id(self, argv):
790 # short YouTube ID starting with dash?
791 idxs = [
792 i for i, a in enumerate(argv)
793 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
794 if idxs:
795 correct_argv = (
796 ['yt-dlp']
797 + [a for i, a in enumerate(argv) if i not in idxs]
798 + ['--'] + [argv[i] for i in idxs]
799 )
800 self.report_warning(
801 'Long argument string detected. '
802 'Use -- to separate parameters and URLs, like this:\n%s' %
803 args_to_str(correct_argv))
804
805 def add_info_extractor(self, ie):
806 """Add an InfoExtractor object to the end of the list."""
807 ie_key = ie.ie_key()
808 self._ies[ie_key] = ie
809 if not isinstance(ie, type):
810 self._ies_instances[ie_key] = ie
811 ie.set_downloader(self)
812
813 def get_info_extractor(self, ie_key):
814 """
815 Get an instance of an IE with name ie_key, it will try to get one from
816 the _ies list, if there's no instance it will create a new one and add
817 it to the extractor list.
818 """
819 ie = self._ies_instances.get(ie_key)
820 if ie is None:
821 ie = get_info_extractor(ie_key)()
822 self.add_info_extractor(ie)
823 return ie
824
825 def add_default_info_extractors(self):
826 """
827 Add the InfoExtractors returned by gen_extractors to the end of the list
828 """
829 all_ies = {ie.IE_NAME.lower(): ie for ie in gen_extractor_classes()}
830 all_ies['end'] = UnsupportedURLIE()
831 try:
832 ie_names = orderedSet_from_options(
833 self.params.get('allowed_extractors', ['default']), {
834 'all': list(all_ies),
835 'default': [name for name, ie in all_ies.items() if ie._ENABLED],
836 }, use_regex=True)
837 except re.error as e:
838 raise ValueError(f'Wrong regex for allowed_extractors: {e.pattern}')
839 for name in ie_names:
840 self.add_info_extractor(all_ies[name])
841 self.write_debug(f'Loaded {len(ie_names)} extractors')
842
843 def add_post_processor(self, pp, when='post_process'):
844 """Add a PostProcessor object to the end of the chain."""
845 assert when in POSTPROCESS_WHEN, f'Invalid when={when}'
846 self._pps[when].append(pp)
847 pp.set_downloader(self)
848
849 def add_post_hook(self, ph):
850 """Add the post hook"""
851 self._post_hooks.append(ph)
852
853 def add_progress_hook(self, ph):
854 """Add the download progress hook"""
855 self._progress_hooks.append(ph)
856
857 def add_postprocessor_hook(self, ph):
858 """Add the postprocessing progress hook"""
859 self._postprocessor_hooks.append(ph)
860 for pps in self._pps.values():
861 for pp in pps:
862 pp.add_progress_hook(ph)
863
864 def _bidi_workaround(self, message):
865 if not hasattr(self, '_output_channel'):
866 return message
867
868 assert hasattr(self, '_output_process')
869 assert isinstance(message, str)
870 line_count = message.count('\n') + 1
871 self._output_process.stdin.write((message + '\n').encode())
872 self._output_process.stdin.flush()
873 res = ''.join(self._output_channel.readline().decode()
874 for _ in range(line_count))
875 return res[:-len('\n')]
876
877 def _write_string(self, message, out=None, only_once=False):
878 if only_once:
879 if message in self._printed_messages:
880 return
881 self._printed_messages.add(message)
882 write_string(message, out=out, encoding=self.params.get('encoding'))
883
884 def to_stdout(self, message, skip_eol=False, quiet=None):
885 """Print message to stdout"""
886 if quiet is not None:
887 self.deprecation_warning('"YoutubeDL.to_stdout" no longer accepts the argument quiet. '
888 'Use "YoutubeDL.to_screen" instead')
889 if skip_eol is not False:
890 self.deprecation_warning('"YoutubeDL.to_stdout" no longer accepts the argument skip_eol. '
891 'Use "YoutubeDL.to_screen" instead')
892 self._write_string(f'{self._bidi_workaround(message)}\n', self._out_files.out)
893
894 def to_screen(self, message, skip_eol=False, quiet=None, only_once=False):
895 """Print message to screen if not in quiet mode"""
896 if self.params.get('logger'):
897 self.params['logger'].debug(message)
898 return
899 if (self.params.get('quiet') if quiet is None else quiet) and not self.params.get('verbose'):
900 return
901 self._write_string(
902 '%s%s' % (self._bidi_workaround(message), ('' if skip_eol else '\n')),
903 self._out_files.screen, only_once=only_once)
904
905 def to_stderr(self, message, only_once=False):
906 """Print message to stderr"""
907 assert isinstance(message, str)
908 if self.params.get('logger'):
909 self.params['logger'].error(message)
910 else:
911 self._write_string(f'{self._bidi_workaround(message)}\n', self._out_files.error, only_once=only_once)
912
913 def _send_console_code(self, code):
914 if compat_os_name == 'nt' or not self._out_files.console:
915 return
916 self._write_string(code, self._out_files.console)
917
918 def to_console_title(self, message):
919 if not self.params.get('consoletitle', False):
920 return
921 message = remove_terminal_sequences(message)
922 if compat_os_name == 'nt':
923 if ctypes.windll.kernel32.GetConsoleWindow():
924 # c_wchar_p() might not be necessary if `message` is
925 # already of type unicode()
926 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
927 else:
928 self._send_console_code(f'\033]0;{message}\007')
929
930 def save_console_title(self):
931 if not self.params.get('consoletitle') or self.params.get('simulate'):
932 return
933 self._send_console_code('\033[22;0t') # Save the title on stack
934
935 def restore_console_title(self):
936 if not self.params.get('consoletitle') or self.params.get('simulate'):
937 return
938 self._send_console_code('\033[23;0t') # Restore the title from stack
939
940 def __enter__(self):
941 self.save_console_title()
942 return self
943
944 def __exit__(self, *args):
945 self.restore_console_title()
946
947 if self.params.get('cookiefile') is not None:
948 self.cookiejar.save(ignore_discard=True, ignore_expires=True)
949
950 def trouble(self, message=None, tb=None, is_error=True):
951 """Determine action to take when a download problem appears.
952
953 Depending on if the downloader has been configured to ignore
954 download errors or not, this method may throw an exception or
955 not when errors are found, after printing the message.
956
957 @param tb If given, is additional traceback information
958 @param is_error Whether to raise error according to ignorerrors
959 """
960 if message is not None:
961 self.to_stderr(message)
962 if self.params.get('verbose'):
963 if tb is None:
964 if sys.exc_info()[0]: # if .trouble has been called from an except block
965 tb = ''
966 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
967 tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
968 tb += encode_compat_str(traceback.format_exc())
969 else:
970 tb_data = traceback.format_list(traceback.extract_stack())
971 tb = ''.join(tb_data)
972 if tb:
973 self.to_stderr(tb)
974 if not is_error:
975 return
976 if not self.params.get('ignoreerrors'):
977 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
978 exc_info = sys.exc_info()[1].exc_info
979 else:
980 exc_info = sys.exc_info()
981 raise DownloadError(message, exc_info)
982 self._download_retcode = 1
983
984 Styles = Namespace(
985 HEADERS='yellow',
986 EMPHASIS='light blue',
987 FILENAME='green',
988 ID='green',
989 DELIM='blue',
990 ERROR='red',
991 BAD_FORMAT='light red',
992 WARNING='yellow',
993 SUPPRESS='light black',
994 )
995
996 def _format_text(self, handle, allow_colors, text, f, fallback=None, *, test_encoding=False):
997 text = str(text)
998 if test_encoding:
999 original_text = text
1000 # handle.encoding can be None. See https://github.com/yt-dlp/yt-dlp/issues/2711
1001 encoding = self.params.get('encoding') or getattr(handle, 'encoding', None) or 'ascii'
1002 text = text.encode(encoding, 'ignore').decode(encoding)
1003 if fallback is not None and text != original_text:
1004 text = fallback
1005 return format_text(text, f) if allow_colors is True else text if fallback is None else fallback
1006
1007 def _format_out(self, *args, **kwargs):
1008 return self._format_text(self._out_files.out, self._allow_colors.out, *args, **kwargs)
1009
1010 def _format_screen(self, *args, **kwargs):
1011 return self._format_text(self._out_files.screen, self._allow_colors.screen, *args, **kwargs)
1012
1013 def _format_err(self, *args, **kwargs):
1014 return self._format_text(self._out_files.error, self._allow_colors.error, *args, **kwargs)
1015
1016 def report_warning(self, message, only_once=False):
1017 '''
1018 Print the message to stderr, it will be prefixed with 'WARNING:'
1019 If stderr is a tty file the 'WARNING:' will be colored
1020 '''
1021 if self.params.get('logger') is not None:
1022 self.params['logger'].warning(message)
1023 else:
1024 if self.params.get('no_warnings'):
1025 return
1026 self.to_stderr(f'{self._format_err("WARNING:", self.Styles.WARNING)} {message}', only_once)
1027
1028 def deprecation_warning(self, message, *, stacklevel=0):
1029 deprecation_warning(
1030 message, stacklevel=stacklevel + 1, printer=self.report_error, is_error=False)
1031
1032 def deprecated_feature(self, message):
1033 if self.params.get('logger') is not None:
1034 self.params['logger'].warning(f'Deprecated Feature: {message}')
1035 self.to_stderr(f'{self._format_err("Deprecated Feature:", self.Styles.ERROR)} {message}', True)
1036
1037 def report_error(self, message, *args, **kwargs):
1038 '''
1039 Do the same as trouble, but prefixes the message with 'ERROR:', colored
1040 in red if stderr is a tty file.
1041 '''
1042 self.trouble(f'{self._format_err("ERROR:", self.Styles.ERROR)} {message}', *args, **kwargs)
1043
1044 def write_debug(self, message, only_once=False):
1045 '''Log debug message or Print message to stderr'''
1046 if not self.params.get('verbose', False):
1047 return
1048 message = f'[debug] {message}'
1049 if self.params.get('logger'):
1050 self.params['logger'].debug(message)
1051 else:
1052 self.to_stderr(message, only_once)
1053
1054 def report_file_already_downloaded(self, file_name):
1055 """Report file has already been fully downloaded."""
1056 try:
1057 self.to_screen('[download] %s has already been downloaded' % file_name)
1058 except UnicodeEncodeError:
1059 self.to_screen('[download] The file has already been downloaded')
1060
1061 def report_file_delete(self, file_name):
1062 """Report that existing file will be deleted."""
1063 try:
1064 self.to_screen('Deleting existing file %s' % file_name)
1065 except UnicodeEncodeError:
1066 self.to_screen('Deleting existing file')
1067
1068 def raise_no_formats(self, info, forced=False, *, msg=None):
1069 has_drm = info.get('_has_drm')
1070 ignored, expected = self.params.get('ignore_no_formats_error'), bool(msg)
1071 msg = msg or has_drm and 'This video is DRM protected' or 'No video formats found!'
1072 if forced or not ignored:
1073 raise ExtractorError(msg, video_id=info['id'], ie=info['extractor'],
1074 expected=has_drm or ignored or expected)
1075 else:
1076 self.report_warning(msg)
1077
1078 def parse_outtmpl(self):
1079 self.deprecation_warning('"YoutubeDL.parse_outtmpl" is deprecated and may be removed in a future version')
1080 self._parse_outtmpl()
1081 return self.params['outtmpl']
1082
1083 def _parse_outtmpl(self):
1084 sanitize = IDENTITY
1085 if self.params.get('restrictfilenames'): # Remove spaces in the default template
1086 sanitize = lambda x: x.replace(' - ', ' ').replace(' ', '-')
1087
1088 outtmpl = self.params.setdefault('outtmpl', {})
1089 if not isinstance(outtmpl, dict):
1090 self.params['outtmpl'] = outtmpl = {'default': outtmpl}
1091 outtmpl.update({k: sanitize(v) for k, v in DEFAULT_OUTTMPL.items() if outtmpl.get(k) is None})
1092
1093 def get_output_path(self, dir_type='', filename=None):
1094 paths = self.params.get('paths', {})
1095 assert isinstance(paths, dict), '"paths" parameter must be a dictionary'
1096 path = os.path.join(
1097 expand_path(paths.get('home', '').strip()),
1098 expand_path(paths.get(dir_type, '').strip()) if dir_type else '',
1099 filename or '')
1100 return sanitize_path(path, force=self.params.get('windowsfilenames'))
1101
1102 @staticmethod
1103 def _outtmpl_expandpath(outtmpl):
1104 # expand_path translates '%%' into '%' and '$$' into '$'
1105 # correspondingly that is not what we want since we need to keep
1106 # '%%' intact for template dict substitution step. Working around
1107 # with boundary-alike separator hack.
1108 sep = ''.join(random.choices(string.ascii_letters, k=32))
1109 outtmpl = outtmpl.replace('%%', f'%{sep}%').replace('$$', f'${sep}$')
1110
1111 # outtmpl should be expand_path'ed before template dict substitution
1112 # because meta fields may contain env variables we don't want to
1113 # be expanded. E.g. for outtmpl "%(title)s.%(ext)s" and
1114 # title "Hello $PATH", we don't want `$PATH` to be expanded.
1115 return expand_path(outtmpl).replace(sep, '')
1116
1117 @staticmethod
1118 def escape_outtmpl(outtmpl):
1119 ''' Escape any remaining strings like %s, %abc% etc. '''
1120 return re.sub(
1121 STR_FORMAT_RE_TMPL.format('', '(?![%(\0])'),
1122 lambda mobj: ('' if mobj.group('has_key') else '%') + mobj.group(0),
1123 outtmpl)
1124
1125 @classmethod
1126 def validate_outtmpl(cls, outtmpl):
1127 ''' @return None or Exception object '''
1128 outtmpl = re.sub(
1129 STR_FORMAT_RE_TMPL.format('[^)]*', '[ljhqBUDS]'),
1130 lambda mobj: f'{mobj.group(0)[:-1]}s',
1131 cls._outtmpl_expandpath(outtmpl))
1132 try:
1133 cls.escape_outtmpl(outtmpl) % collections.defaultdict(int)
1134 return None
1135 except ValueError as err:
1136 return err
1137
1138 @staticmethod
1139 def _copy_infodict(info_dict):
1140 info_dict = dict(info_dict)
1141 info_dict.pop('__postprocessors', None)
1142 info_dict.pop('__pending_error', None)
1143 return info_dict
1144
1145 def prepare_outtmpl(self, outtmpl, info_dict, sanitize=False):
1146 """ Make the outtmpl and info_dict suitable for substitution: ydl.escape_outtmpl(outtmpl) % info_dict
1147 @param sanitize Whether to sanitize the output as a filename.
1148 For backward compatibility, a function can also be passed
1149 """
1150
1151 info_dict.setdefault('epoch', int(time.time())) # keep epoch consistent once set
1152
1153 info_dict = self._copy_infodict(info_dict)
1154 info_dict['duration_string'] = ( # %(duration>%H-%M-%S)s is wrong if duration > 24hrs
1155 formatSeconds(info_dict['duration'], '-' if sanitize else ':')
1156 if info_dict.get('duration', None) is not None
1157 else None)
1158 info_dict['autonumber'] = int(self.params.get('autonumber_start', 1) - 1 + self._num_downloads)
1159 info_dict['video_autonumber'] = self._num_videos
1160 if info_dict.get('resolution') is None:
1161 info_dict['resolution'] = self.format_resolution(info_dict, default=None)
1162
1163 # For fields playlist_index, playlist_autonumber and autonumber convert all occurrences
1164 # of %(field)s to %(field)0Nd for backward compatibility
1165 field_size_compat_map = {
1166 'playlist_index': number_of_digits(info_dict.get('__last_playlist_index') or 0),
1167 'playlist_autonumber': number_of_digits(info_dict.get('n_entries') or 0),
1168 'autonumber': self.params.get('autonumber_size') or 5,
1169 }
1170
1171 TMPL_DICT = {}
1172 EXTERNAL_FORMAT_RE = re.compile(STR_FORMAT_RE_TMPL.format('[^)]*', f'[{STR_FORMAT_TYPES}ljhqBUDS]'))
1173 MATH_FUNCTIONS = {
1174 '+': float.__add__,
1175 '-': float.__sub__,
1176 }
1177 # Field is of the form key1.key2...
1178 # where keys (except first) can be string, int, slice or "{field, ...}"
1179 FIELD_INNER_RE = r'(?:\w+|%(num)s|%(num)s?(?::%(num)s?){1,2})' % {'num': r'(?:-?\d+)'}
1180 FIELD_RE = r'\w*(?:\.(?:%(inner)s|{%(field)s(?:,%(field)s)*}))*' % {
1181 'inner': FIELD_INNER_RE,
1182 'field': rf'\w*(?:\.{FIELD_INNER_RE})*'
1183 }
1184 MATH_FIELD_RE = rf'(?:{FIELD_RE}|-?{NUMBER_RE})'
1185 MATH_OPERATORS_RE = r'(?:%s)' % '|'.join(map(re.escape, MATH_FUNCTIONS.keys()))
1186 INTERNAL_FORMAT_RE = re.compile(rf'''(?xs)
1187 (?P<negate>-)?
1188 (?P<fields>{FIELD_RE})
1189 (?P<maths>(?:{MATH_OPERATORS_RE}{MATH_FIELD_RE})*)
1190 (?:>(?P<strf_format>.+?))?
1191 (?P<remaining>
1192 (?P<alternate>(?<!\\),[^|&)]+)?
1193 (?:&(?P<replacement>.*?))?
1194 (?:\|(?P<default>.*?))?
1195 )$''')
1196
1197 def _traverse_infodict(fields):
1198 fields = [f for x in re.split(r'\.({.+?})\.?', fields)
1199 for f in ([x] if x.startswith('{') else x.split('.'))]
1200 for i in (0, -1):
1201 if fields and not fields[i]:
1202 fields.pop(i)
1203
1204 for i, f in enumerate(fields):
1205 if not f.startswith('{'):
1206 continue
1207 assert f.endswith('}'), f'No closing brace for {f} in {fields}'
1208 fields[i] = {k: k.split('.') for k in f[1:-1].split(',')}
1209
1210 return traverse_obj(info_dict, fields, is_user_input=True, traverse_string=True)
1211
1212 def get_value(mdict):
1213 # Object traversal
1214 value = _traverse_infodict(mdict['fields'])
1215 # Negative
1216 if mdict['negate']:
1217 value = float_or_none(value)
1218 if value is not None:
1219 value *= -1
1220 # Do maths
1221 offset_key = mdict['maths']
1222 if offset_key:
1223 value = float_or_none(value)
1224 operator = None
1225 while offset_key:
1226 item = re.match(
1227 MATH_FIELD_RE if operator else MATH_OPERATORS_RE,
1228 offset_key).group(0)
1229 offset_key = offset_key[len(item):]
1230 if operator is None:
1231 operator = MATH_FUNCTIONS[item]
1232 continue
1233 item, multiplier = (item[1:], -1) if item[0] == '-' else (item, 1)
1234 offset = float_or_none(item)
1235 if offset is None:
1236 offset = float_or_none(_traverse_infodict(item))
1237 try:
1238 value = operator(value, multiplier * offset)
1239 except (TypeError, ZeroDivisionError):
1240 return None
1241 operator = None
1242 # Datetime formatting
1243 if mdict['strf_format']:
1244 value = strftime_or_none(value, mdict['strf_format'].replace('\\,', ','))
1245
1246 # XXX: Workaround for https://github.com/yt-dlp/yt-dlp/issues/4485
1247 if sanitize and value == '':
1248 value = None
1249 return value
1250
1251 na = self.params.get('outtmpl_na_placeholder', 'NA')
1252
1253 def filename_sanitizer(key, value, restricted=self.params.get('restrictfilenames')):
1254 return sanitize_filename(str(value), restricted=restricted, is_id=(
1255 bool(re.search(r'(^|[_.])id(\.|$)', key))
1256 if 'filename-sanitization' in self.params['compat_opts']
1257 else NO_DEFAULT))
1258
1259 sanitizer = sanitize if callable(sanitize) else filename_sanitizer
1260 sanitize = bool(sanitize)
1261
1262 def _dumpjson_default(obj):
1263 if isinstance(obj, (set, LazyList)):
1264 return list(obj)
1265 return repr(obj)
1266
1267 class _ReplacementFormatter(string.Formatter):
1268 def get_field(self, field_name, args, kwargs):
1269 if field_name.isdigit():
1270 return args[0], -1
1271 raise ValueError('Unsupported field')
1272
1273 replacement_formatter = _ReplacementFormatter()
1274
1275 def create_key(outer_mobj):
1276 if not outer_mobj.group('has_key'):
1277 return outer_mobj.group(0)
1278 key = outer_mobj.group('key')
1279 mobj = re.match(INTERNAL_FORMAT_RE, key)
1280 value, replacement, default, last_field = None, None, na, ''
1281 while mobj:
1282 mobj = mobj.groupdict()
1283 default = mobj['default'] if mobj['default'] is not None else default
1284 value = get_value(mobj)
1285 last_field, replacement = mobj['fields'], mobj['replacement']
1286 if value is None and mobj['alternate']:
1287 mobj = re.match(INTERNAL_FORMAT_RE, mobj['remaining'][1:])
1288 else:
1289 break
1290
1291 fmt = outer_mobj.group('format')
1292 if fmt == 's' and value is not None and last_field in field_size_compat_map.keys():
1293 fmt = f'0{field_size_compat_map[last_field]:d}d'
1294
1295 if None not in (value, replacement):
1296 try:
1297 value = replacement_formatter.format(replacement, value)
1298 except ValueError:
1299 value, default = None, na
1300
1301 flags = outer_mobj.group('conversion') or ''
1302 str_fmt = f'{fmt[:-1]}s'
1303 if value is None:
1304 value, fmt = default, 's'
1305 elif fmt[-1] == 'l': # list
1306 delim = '\n' if '#' in flags else ', '
1307 value, fmt = delim.join(map(str, variadic(value, allowed_types=(str, bytes)))), str_fmt
1308 elif fmt[-1] == 'j': # json
1309 value, fmt = json.dumps(
1310 value, default=_dumpjson_default,
1311 indent=4 if '#' in flags else None, ensure_ascii='+' not in flags), str_fmt
1312 elif fmt[-1] == 'h': # html
1313 value, fmt = escapeHTML(str(value)), str_fmt
1314 elif fmt[-1] == 'q': # quoted
1315 value = map(str, variadic(value) if '#' in flags else [value])
1316 value, fmt = ' '.join(map(compat_shlex_quote, value)), str_fmt
1317 elif fmt[-1] == 'B': # bytes
1318 value = f'%{str_fmt}'.encode() % str(value).encode()
1319 value, fmt = value.decode('utf-8', 'ignore'), 's'
1320 elif fmt[-1] == 'U': # unicode normalized
1321 value, fmt = unicodedata.normalize(
1322 # "+" = compatibility equivalence, "#" = NFD
1323 'NF%s%s' % ('K' if '+' in flags else '', 'D' if '#' in flags else 'C'),
1324 value), str_fmt
1325 elif fmt[-1] == 'D': # decimal suffix
1326 num_fmt, fmt = fmt[:-1].replace('#', ''), 's'
1327 value = format_decimal_suffix(value, f'%{num_fmt}f%s' if num_fmt else '%d%s',
1328 factor=1024 if '#' in flags else 1000)
1329 elif fmt[-1] == 'S': # filename sanitization
1330 value, fmt = filename_sanitizer(last_field, value, restricted='#' in flags), str_fmt
1331 elif fmt[-1] == 'c':
1332 if value:
1333 value = str(value)[0]
1334 else:
1335 fmt = str_fmt
1336 elif fmt[-1] not in 'rsa': # numeric
1337 value = float_or_none(value)
1338 if value is None:
1339 value, fmt = default, 's'
1340
1341 if sanitize:
1342 # If value is an object, sanitize might convert it to a string
1343 # So we convert it to repr first
1344 if fmt[-1] == 'r':
1345 value, fmt = repr(value), str_fmt
1346 elif fmt[-1] == 'a':
1347 value, fmt = ascii(value), str_fmt
1348 if fmt[-1] in 'csra':
1349 value = sanitizer(last_field, value)
1350
1351 key = '%s\0%s' % (key.replace('%', '%\0'), outer_mobj.group('format'))
1352 TMPL_DICT[key] = value
1353 return '{prefix}%({key}){fmt}'.format(key=key, fmt=fmt, prefix=outer_mobj.group('prefix'))
1354
1355 return EXTERNAL_FORMAT_RE.sub(create_key, outtmpl), TMPL_DICT
1356
1357 def evaluate_outtmpl(self, outtmpl, info_dict, *args, **kwargs):
1358 outtmpl, info_dict = self.prepare_outtmpl(outtmpl, info_dict, *args, **kwargs)
1359 return self.escape_outtmpl(outtmpl) % info_dict
1360
1361 def _prepare_filename(self, info_dict, *, outtmpl=None, tmpl_type=None):
1362 assert None in (outtmpl, tmpl_type), 'outtmpl and tmpl_type are mutually exclusive'
1363 if outtmpl is None:
1364 outtmpl = self.params['outtmpl'].get(tmpl_type or 'default', self.params['outtmpl']['default'])
1365 try:
1366 outtmpl = self._outtmpl_expandpath(outtmpl)
1367 filename = self.evaluate_outtmpl(outtmpl, info_dict, True)
1368 if not filename:
1369 return None
1370
1371 if tmpl_type in ('', 'temp'):
1372 final_ext, ext = self.params.get('final_ext'), info_dict.get('ext')
1373 if final_ext and ext and final_ext != ext and filename.endswith(f'.{final_ext}'):
1374 filename = replace_extension(filename, ext, final_ext)
1375 elif tmpl_type:
1376 force_ext = OUTTMPL_TYPES[tmpl_type]
1377 if force_ext:
1378 filename = replace_extension(filename, force_ext, info_dict.get('ext'))
1379
1380 # https://github.com/blackjack4494/youtube-dlc/issues/85
1381 trim_file_name = self.params.get('trim_file_name', False)
1382 if trim_file_name:
1383 no_ext, *ext = filename.rsplit('.', 2)
1384 filename = join_nonempty(no_ext[:trim_file_name], *ext, delim='.')
1385
1386 return filename
1387 except ValueError as err:
1388 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
1389 return None
1390
1391 def prepare_filename(self, info_dict, dir_type='', *, outtmpl=None, warn=False):
1392 """Generate the output filename"""
1393 if outtmpl:
1394 assert not dir_type, 'outtmpl and dir_type are mutually exclusive'
1395 dir_type = None
1396 filename = self._prepare_filename(info_dict, tmpl_type=dir_type, outtmpl=outtmpl)
1397 if not filename and dir_type not in ('', 'temp'):
1398 return ''
1399
1400 if warn:
1401 if not self.params.get('paths'):
1402 pass
1403 elif filename == '-':
1404 self.report_warning('--paths is ignored when an outputting to stdout', only_once=True)
1405 elif os.path.isabs(filename):
1406 self.report_warning('--paths is ignored since an absolute path is given in output template', only_once=True)
1407 if filename == '-' or not filename:
1408 return filename
1409
1410 return self.get_output_path(dir_type, filename)
1411
1412 def _match_entry(self, info_dict, incomplete=False, silent=False):
1413 """Returns None if the file should be downloaded"""
1414 _type = 'video' if 'playlist-match-filter' in self.params['compat_opts'] else info_dict.get('_type', 'video')
1415 assert incomplete or _type == 'video', 'Only video result can be considered complete'
1416
1417 video_title = info_dict.get('title', info_dict.get('id', 'entry'))
1418
1419 def check_filter():
1420 if _type in ('playlist', 'multi_video'):
1421 return
1422 elif _type in ('url', 'url_transparent') and not try_call(
1423 lambda: self.get_info_extractor(info_dict['ie_key']).is_single_video(info_dict['url'])):
1424 return
1425
1426 if 'title' in info_dict:
1427 # This can happen when we're just evaluating the playlist
1428 title = info_dict['title']
1429 matchtitle = self.params.get('matchtitle', False)
1430 if matchtitle:
1431 if not re.search(matchtitle, title, re.IGNORECASE):
1432 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
1433 rejecttitle = self.params.get('rejecttitle', False)
1434 if rejecttitle:
1435 if re.search(rejecttitle, title, re.IGNORECASE):
1436 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
1437
1438 date = info_dict.get('upload_date')
1439 if date is not None:
1440 dateRange = self.params.get('daterange', DateRange())
1441 if date not in dateRange:
1442 return f'{date_from_str(date).isoformat()} upload date is not in range {dateRange}'
1443 view_count = info_dict.get('view_count')
1444 if view_count is not None:
1445 min_views = self.params.get('min_views')
1446 if min_views is not None and view_count < min_views:
1447 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
1448 max_views = self.params.get('max_views')
1449 if max_views is not None and view_count > max_views:
1450 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
1451 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
1452 return 'Skipping "%s" because it is age restricted' % video_title
1453
1454 match_filter = self.params.get('match_filter')
1455 if match_filter is None:
1456 return None
1457
1458 cancelled = None
1459 try:
1460 try:
1461 ret = match_filter(info_dict, incomplete=incomplete)
1462 except TypeError:
1463 # For backward compatibility
1464 ret = None if incomplete else match_filter(info_dict)
1465 except DownloadCancelled as err:
1466 if err.msg is not NO_DEFAULT:
1467 raise
1468 ret, cancelled = err.msg, err
1469
1470 if ret is NO_DEFAULT:
1471 while True:
1472 filename = self._format_screen(self.prepare_filename(info_dict), self.Styles.FILENAME)
1473 reply = input(self._format_screen(
1474 f'Download "{filename}"? (Y/n): ', self.Styles.EMPHASIS)).lower().strip()
1475 if reply in {'y', ''}:
1476 return None
1477 elif reply == 'n':
1478 if cancelled:
1479 raise type(cancelled)(f'Skipping {video_title}')
1480 return f'Skipping {video_title}'
1481 return ret
1482
1483 if self.in_download_archive(info_dict):
1484 reason = '%s has already been recorded in the archive' % video_title
1485 break_opt, break_err = 'break_on_existing', ExistingVideoReached
1486 else:
1487 try:
1488 reason = check_filter()
1489 except DownloadCancelled as e:
1490 reason, break_opt, break_err = e.msg, 'match_filter', type(e)
1491 else:
1492 break_opt, break_err = 'break_on_reject', RejectedVideoReached
1493 if reason is not None:
1494 if not silent:
1495 self.to_screen('[download] ' + reason)
1496 if self.params.get(break_opt, False):
1497 raise break_err()
1498 return reason
1499
1500 @staticmethod
1501 def add_extra_info(info_dict, extra_info):
1502 '''Set the keys from extra_info in info dict if they are missing'''
1503 for key, value in extra_info.items():
1504 info_dict.setdefault(key, value)
1505
1506 def extract_info(self, url, download=True, ie_key=None, extra_info=None,
1507 process=True, force_generic_extractor=False):
1508 """
1509 Extract and return the information dictionary of the URL
1510
1511 Arguments:
1512 @param url URL to extract
1513
1514 Keyword arguments:
1515 @param download Whether to download videos
1516 @param process Whether to resolve all unresolved references (URLs, playlist items).
1517 Must be True for download to work
1518 @param ie_key Use only the extractor with this key
1519
1520 @param extra_info Dictionary containing the extra values to add to the info (For internal use only)
1521 @force_generic_extractor Force using the generic extractor (Deprecated; use ie_key='Generic')
1522 """
1523
1524 if extra_info is None:
1525 extra_info = {}
1526
1527 if not ie_key and force_generic_extractor:
1528 ie_key = 'Generic'
1529
1530 if ie_key:
1531 ies = {ie_key: self._ies[ie_key]} if ie_key in self._ies else {}
1532 else:
1533 ies = self._ies
1534
1535 for key, ie in ies.items():
1536 if not ie.suitable(url):
1537 continue
1538
1539 if not ie.working():
1540 self.report_warning('The program functionality for this site has been marked as broken, '
1541 'and will probably not work.')
1542
1543 temp_id = ie.get_temp_id(url)
1544 if temp_id is not None and self.in_download_archive({'id': temp_id, 'ie_key': key}):
1545 self.to_screen(f'[{key}] {temp_id}: has already been recorded in the archive')
1546 if self.params.get('break_on_existing', False):
1547 raise ExistingVideoReached()
1548 break
1549 return self.__extract_info(url, self.get_info_extractor(key), download, extra_info, process)
1550 else:
1551 extractors_restricted = self.params.get('allowed_extractors') not in (None, ['default'])
1552 self.report_error(f'No suitable extractor{format_field(ie_key, None, " (%s)")} found for URL {url}',
1553 tb=False if extractors_restricted else None)
1554
1555 def _handle_extraction_exceptions(func):
1556 @functools.wraps(func)
1557 def wrapper(self, *args, **kwargs):
1558 while True:
1559 try:
1560 return func(self, *args, **kwargs)
1561 except (DownloadCancelled, LazyList.IndexError, PagedList.IndexError):
1562 raise
1563 except ReExtractInfo as e:
1564 if e.expected:
1565 self.to_screen(f'{e}; Re-extracting data')
1566 else:
1567 self.to_stderr('\r')
1568 self.report_warning(f'{e}; Re-extracting data')
1569 continue
1570 except GeoRestrictedError as e:
1571 msg = e.msg
1572 if e.countries:
1573 msg += '\nThis video is available in %s.' % ', '.join(
1574 map(ISO3166Utils.short2full, e.countries))
1575 msg += '\nYou might want to use a VPN or a proxy server (with --proxy) to workaround.'
1576 self.report_error(msg)
1577 except ExtractorError as e: # An error we somewhat expected
1578 self.report_error(str(e), e.format_traceback())
1579 except Exception as e:
1580 if self.params.get('ignoreerrors'):
1581 self.report_error(str(e), tb=encode_compat_str(traceback.format_exc()))
1582 else:
1583 raise
1584 break
1585 return wrapper
1586
1587 def _wait_for_video(self, ie_result={}):
1588 if (not self.params.get('wait_for_video')
1589 or ie_result.get('_type', 'video') != 'video'
1590 or ie_result.get('formats') or ie_result.get('url')):
1591 return
1592
1593 format_dur = lambda dur: '%02d:%02d:%02d' % timetuple_from_msec(dur * 1000)[:-1]
1594 last_msg = ''
1595
1596 def progress(msg):
1597 nonlocal last_msg
1598 full_msg = f'{msg}\n'
1599 if not self.params.get('noprogress'):
1600 full_msg = msg + ' ' * (len(last_msg) - len(msg)) + '\r'
1601 elif last_msg:
1602 return
1603 self.to_screen(full_msg, skip_eol=True)
1604 last_msg = msg
1605
1606 min_wait, max_wait = self.params.get('wait_for_video')
1607 diff = try_get(ie_result, lambda x: x['release_timestamp'] - time.time())
1608 if diff is None and ie_result.get('live_status') == 'is_upcoming':
1609 diff = round(random.uniform(min_wait, max_wait) if (max_wait and min_wait) else (max_wait or min_wait), 0)
1610 self.report_warning('Release time of video is not known')
1611 elif ie_result and (diff or 0) <= 0:
1612 self.report_warning('Video should already be available according to extracted info')
1613 diff = min(max(diff or 0, min_wait or 0), max_wait or float('inf'))
1614 self.to_screen(f'[wait] Waiting for {format_dur(diff)} - Press Ctrl+C to try now')
1615
1616 wait_till = time.time() + diff
1617 try:
1618 while True:
1619 diff = wait_till - time.time()
1620 if diff <= 0:
1621 progress('')
1622 raise ReExtractInfo('[wait] Wait period ended', expected=True)
1623 progress(f'[wait] Remaining time until next attempt: {self._format_screen(format_dur(diff), self.Styles.EMPHASIS)}')
1624 time.sleep(1)
1625 except KeyboardInterrupt:
1626 progress('')
1627 raise ReExtractInfo('[wait] Interrupted by user', expected=True)
1628 except BaseException as e:
1629 if not isinstance(e, ReExtractInfo):
1630 self.to_screen('')
1631 raise
1632
1633 def _load_cookies(self, data, *, from_headers=True):
1634 """Loads cookies from a `Cookie` header
1635
1636 This tries to work around the security vulnerability of passing cookies to every domain.
1637 See: https://github.com/yt-dlp/yt-dlp/security/advisories/GHSA-v8mc-9377-rwjj
1638 The unscoped cookies are saved for later to be stored in the jar with a limited scope.
1639
1640 @param data The Cookie header as string to load the cookies from
1641 @param from_headers If `False`, allows Set-Cookie syntax in the cookie string (at least a domain will be required)
1642 """
1643 for cookie in LenientSimpleCookie(data).values():
1644 if from_headers and any(cookie.values()):
1645 raise ValueError('Invalid syntax in Cookie Header')
1646
1647 domain = cookie.get('domain') or ''
1648 expiry = cookie.get('expires')
1649 if expiry == '': # 0 is valid
1650 expiry = None
1651 prepared_cookie = http.cookiejar.Cookie(
1652 cookie.get('version') or 0, cookie.key, cookie.value, None, False,
1653 domain, True, True, cookie.get('path') or '', bool(cookie.get('path')),
1654 cookie.get('secure') or False, expiry, False, None, None, {})
1655
1656 if domain:
1657 self.cookiejar.set_cookie(prepared_cookie)
1658 elif from_headers:
1659 self.deprecated_feature(
1660 'Passing cookies as a header is a potential security risk; '
1661 'they will be scoped to the domain of the downloaded urls. '
1662 'Please consider loading cookies from a file or browser instead.')
1663 self.__header_cookies.append(prepared_cookie)
1664 else:
1665 self.report_error('Unscoped cookies are not allowed; please specify some sort of scoping',
1666 tb=False, is_error=False)
1667
1668 def _apply_header_cookies(self, url):
1669 """Applies stray header cookies to the provided url
1670
1671 This loads header cookies and scopes them to the domain provided in `url`.
1672 While this is not ideal, it helps reduce the risk of them being sent
1673 to an unintended destination while mostly maintaining compatibility.
1674 """
1675 parsed = urllib.parse.urlparse(url)
1676 if not parsed.hostname:
1677 return
1678
1679 for cookie in map(copy.copy, self.__header_cookies):
1680 cookie.domain = f'.{parsed.hostname}'
1681 self.cookiejar.set_cookie(cookie)
1682
1683 @_handle_extraction_exceptions
1684 def __extract_info(self, url, ie, download, extra_info, process):
1685 self._apply_header_cookies(url)
1686
1687 try:
1688 ie_result = ie.extract(url)
1689 except UserNotLive as e:
1690 if process:
1691 if self.params.get('wait_for_video'):
1692 self.report_warning(e)
1693 self._wait_for_video()
1694 raise
1695 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
1696 self.report_warning(f'Extractor {ie.IE_NAME} returned nothing{bug_reports_message()}')
1697 return
1698 if isinstance(ie_result, list):
1699 # Backwards compatibility: old IE result format
1700 ie_result = {
1701 '_type': 'compat_list',
1702 'entries': ie_result,
1703 }
1704 if extra_info.get('original_url'):
1705 ie_result.setdefault('original_url', extra_info['original_url'])
1706 self.add_default_extra_info(ie_result, ie, url)
1707 if process:
1708 self._wait_for_video(ie_result)
1709 return self.process_ie_result(ie_result, download, extra_info)
1710 else:
1711 return ie_result
1712
1713 def add_default_extra_info(self, ie_result, ie, url):
1714 if url is not None:
1715 self.add_extra_info(ie_result, {
1716 'webpage_url': url,
1717 'original_url': url,
1718 })
1719 webpage_url = ie_result.get('webpage_url')
1720 if webpage_url:
1721 self.add_extra_info(ie_result, {
1722 'webpage_url_basename': url_basename(webpage_url),
1723 'webpage_url_domain': get_domain(webpage_url),
1724 })
1725 if ie is not None:
1726 self.add_extra_info(ie_result, {
1727 'extractor': ie.IE_NAME,
1728 'extractor_key': ie.ie_key(),
1729 })
1730
1731 def process_ie_result(self, ie_result, download=True, extra_info=None):
1732 """
1733 Take the result of the ie(may be modified) and resolve all unresolved
1734 references (URLs, playlist items).
1735
1736 It will also download the videos if 'download'.
1737 Returns the resolved ie_result.
1738 """
1739 if extra_info is None:
1740 extra_info = {}
1741 result_type = ie_result.get('_type', 'video')
1742
1743 if result_type in ('url', 'url_transparent'):
1744 ie_result['url'] = sanitize_url(
1745 ie_result['url'], scheme='http' if self.params.get('prefer_insecure') else 'https')
1746 if ie_result.get('original_url') and not extra_info.get('original_url'):
1747 extra_info = {'original_url': ie_result['original_url'], **extra_info}
1748
1749 extract_flat = self.params.get('extract_flat', False)
1750 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info)
1751 or extract_flat is True):
1752 info_copy = ie_result.copy()
1753 ie = try_get(ie_result.get('ie_key'), self.get_info_extractor)
1754 if ie and not ie_result.get('id'):
1755 info_copy['id'] = ie.get_temp_id(ie_result['url'])
1756 self.add_default_extra_info(info_copy, ie, ie_result['url'])
1757 self.add_extra_info(info_copy, extra_info)
1758 info_copy, _ = self.pre_process(info_copy)
1759 self._fill_common_fields(info_copy, False)
1760 self.__forced_printings(info_copy)
1761 self._raise_pending_errors(info_copy)
1762 if self.params.get('force_write_download_archive', False):
1763 self.record_download_archive(info_copy)
1764 return ie_result
1765
1766 if result_type == 'video':
1767 self.add_extra_info(ie_result, extra_info)
1768 ie_result = self.process_video_result(ie_result, download=download)
1769 self._raise_pending_errors(ie_result)
1770 additional_urls = (ie_result or {}).get('additional_urls')
1771 if additional_urls:
1772 # TODO: Improve MetadataParserPP to allow setting a list
1773 if isinstance(additional_urls, str):
1774 additional_urls = [additional_urls]
1775 self.to_screen(
1776 '[info] %s: %d additional URL(s) requested' % (ie_result['id'], len(additional_urls)))
1777 self.write_debug('Additional URLs: "%s"' % '", "'.join(additional_urls))
1778 ie_result['additional_entries'] = [
1779 self.extract_info(
1780 url, download, extra_info=extra_info,
1781 force_generic_extractor=self.params.get('force_generic_extractor'))
1782 for url in additional_urls
1783 ]
1784 return ie_result
1785 elif result_type == 'url':
1786 # We have to add extra_info to the results because it may be
1787 # contained in a playlist
1788 return self.extract_info(
1789 ie_result['url'], download,
1790 ie_key=ie_result.get('ie_key'),
1791 extra_info=extra_info)
1792 elif result_type == 'url_transparent':
1793 # Use the information from the embedding page
1794 info = self.extract_info(
1795 ie_result['url'], ie_key=ie_result.get('ie_key'),
1796 extra_info=extra_info, download=False, process=False)
1797
1798 # extract_info may return None when ignoreerrors is enabled and
1799 # extraction failed with an error, don't crash and return early
1800 # in this case
1801 if not info:
1802 return info
1803
1804 exempted_fields = {'_type', 'url', 'ie_key'}
1805 if not ie_result.get('section_end') and ie_result.get('section_start') is None:
1806 # For video clips, the id etc of the clip extractor should be used
1807 exempted_fields |= {'id', 'extractor', 'extractor_key'}
1808
1809 new_result = info.copy()
1810 new_result.update(filter_dict(ie_result, lambda k, v: v is not None and k not in exempted_fields))
1811
1812 # Extracted info may not be a video result (i.e.
1813 # info.get('_type', 'video') != video) but rather an url or
1814 # url_transparent. In such cases outer metadata (from ie_result)
1815 # should be propagated to inner one (info). For this to happen
1816 # _type of info should be overridden with url_transparent. This
1817 # fixes issue from https://github.com/ytdl-org/youtube-dl/pull/11163.
1818 if new_result.get('_type') == 'url':
1819 new_result['_type'] = 'url_transparent'
1820
1821 return self.process_ie_result(
1822 new_result, download=download, extra_info=extra_info)
1823 elif result_type in ('playlist', 'multi_video'):
1824 # Protect from infinite recursion due to recursively nested playlists
1825 # (see https://github.com/ytdl-org/youtube-dl/issues/27833)
1826 webpage_url = ie_result.get('webpage_url') # Playlists maynot have webpage_url
1827 if webpage_url and webpage_url in self._playlist_urls:
1828 self.to_screen(
1829 '[download] Skipping already downloaded playlist: %s'
1830 % ie_result.get('title') or ie_result.get('id'))
1831 return
1832
1833 self._playlist_level += 1
1834 self._playlist_urls.add(webpage_url)
1835 self._fill_common_fields(ie_result, False)
1836 self._sanitize_thumbnails(ie_result)
1837 try:
1838 return self.__process_playlist(ie_result, download)
1839 finally:
1840 self._playlist_level -= 1
1841 if not self._playlist_level:
1842 self._playlist_urls.clear()
1843 elif result_type == 'compat_list':
1844 self.report_warning(
1845 'Extractor %s returned a compat_list result. '
1846 'It needs to be updated.' % ie_result.get('extractor'))
1847
1848 def _fixup(r):
1849 self.add_extra_info(r, {
1850 'extractor': ie_result['extractor'],
1851 'webpage_url': ie_result['webpage_url'],
1852 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1853 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1854 'extractor_key': ie_result['extractor_key'],
1855 })
1856 return r
1857 ie_result['entries'] = [
1858 self.process_ie_result(_fixup(r), download, extra_info)
1859 for r in ie_result['entries']
1860 ]
1861 return ie_result
1862 else:
1863 raise Exception('Invalid result type: %s' % result_type)
1864
1865 def _ensure_dir_exists(self, path):
1866 return make_dir(path, self.report_error)
1867
1868 @staticmethod
1869 def _playlist_infodict(ie_result, strict=False, **kwargs):
1870 info = {
1871 'playlist_count': ie_result.get('playlist_count'),
1872 'playlist': ie_result.get('title') or ie_result.get('id'),
1873 'playlist_id': ie_result.get('id'),
1874 'playlist_title': ie_result.get('title'),
1875 'playlist_uploader': ie_result.get('uploader'),
1876 'playlist_uploader_id': ie_result.get('uploader_id'),
1877 **kwargs,
1878 }
1879 if strict:
1880 return info
1881 if ie_result.get('webpage_url'):
1882 info.update({
1883 'webpage_url': ie_result['webpage_url'],
1884 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1885 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1886 })
1887 return {
1888 **info,
1889 'playlist_index': 0,
1890 '__last_playlist_index': max(ie_result.get('requested_entries') or (0, 0)),
1891 'extractor': ie_result['extractor'],
1892 'extractor_key': ie_result['extractor_key'],
1893 }
1894
1895 def __process_playlist(self, ie_result, download):
1896 """Process each entry in the playlist"""
1897 assert ie_result['_type'] in ('playlist', 'multi_video')
1898
1899 common_info = self._playlist_infodict(ie_result, strict=True)
1900 title = common_info.get('playlist') or '<Untitled>'
1901 if self._match_entry(common_info, incomplete=True) is not None:
1902 return
1903 self.to_screen(f'[download] Downloading {ie_result["_type"]}: {title}')
1904
1905 all_entries = PlaylistEntries(self, ie_result)
1906 entries = orderedSet(all_entries.get_requested_items(), lazy=True)
1907
1908 lazy = self.params.get('lazy_playlist')
1909 if lazy:
1910 resolved_entries, n_entries = [], 'N/A'
1911 ie_result['requested_entries'], ie_result['entries'] = None, None
1912 else:
1913 entries = resolved_entries = list(entries)
1914 n_entries = len(resolved_entries)
1915 ie_result['requested_entries'], ie_result['entries'] = tuple(zip(*resolved_entries)) or ([], [])
1916 if not ie_result.get('playlist_count'):
1917 # Better to do this after potentially exhausting entries
1918 ie_result['playlist_count'] = all_entries.get_full_count()
1919
1920 extra = self._playlist_infodict(ie_result, n_entries=int_or_none(n_entries))
1921 ie_copy = collections.ChainMap(ie_result, extra)
1922
1923 _infojson_written = False
1924 write_playlist_files = self.params.get('allow_playlist_files', True)
1925 if write_playlist_files and self.params.get('list_thumbnails'):
1926 self.list_thumbnails(ie_result)
1927 if write_playlist_files and not self.params.get('simulate'):
1928 _infojson_written = self._write_info_json(
1929 'playlist', ie_result, self.prepare_filename(ie_copy, 'pl_infojson'))
1930 if _infojson_written is None:
1931 return
1932 if self._write_description('playlist', ie_result,
1933 self.prepare_filename(ie_copy, 'pl_description')) is None:
1934 return
1935 # TODO: This should be passed to ThumbnailsConvertor if necessary
1936 self._write_thumbnails('playlist', ie_result, self.prepare_filename(ie_copy, 'pl_thumbnail'))
1937
1938 if lazy:
1939 if self.params.get('playlistreverse') or self.params.get('playlistrandom'):
1940 self.report_warning('playlistreverse and playlistrandom are not supported with lazy_playlist', only_once=True)
1941 elif self.params.get('playlistreverse'):
1942 entries.reverse()
1943 elif self.params.get('playlistrandom'):
1944 random.shuffle(entries)
1945
1946 self.to_screen(f'[{ie_result["extractor"]}] Playlist {title}: Downloading {n_entries} items'
1947 f'{format_field(ie_result, "playlist_count", " of %s")}')
1948
1949 keep_resolved_entries = self.params.get('extract_flat') != 'discard'
1950 if self.params.get('extract_flat') == 'discard_in_playlist':
1951 keep_resolved_entries = ie_result['_type'] != 'playlist'
1952 if keep_resolved_entries:
1953 self.write_debug('The information of all playlist entries will be held in memory')
1954
1955 failures = 0
1956 max_failures = self.params.get('skip_playlist_after_errors') or float('inf')
1957 for i, (playlist_index, entry) in enumerate(entries):
1958 if lazy:
1959 resolved_entries.append((playlist_index, entry))
1960 if not entry:
1961 continue
1962
1963 entry['__x_forwarded_for_ip'] = ie_result.get('__x_forwarded_for_ip')
1964 if not lazy and 'playlist-index' in self.params['compat_opts']:
1965 playlist_index = ie_result['requested_entries'][i]
1966
1967 entry_copy = collections.ChainMap(entry, {
1968 **common_info,
1969 'n_entries': int_or_none(n_entries),
1970 'playlist_index': playlist_index,
1971 'playlist_autonumber': i + 1,
1972 })
1973
1974 if self._match_entry(entry_copy, incomplete=True) is not None:
1975 # For compatabilty with youtube-dl. See https://github.com/yt-dlp/yt-dlp/issues/4369
1976 resolved_entries[i] = (playlist_index, NO_DEFAULT)
1977 continue
1978
1979 self.to_screen('[download] Downloading item %s of %s' % (
1980 self._format_screen(i + 1, self.Styles.ID), self._format_screen(n_entries, self.Styles.EMPHASIS)))
1981
1982 entry_result = self.__process_iterable_entry(entry, download, collections.ChainMap({
1983 'playlist_index': playlist_index,
1984 'playlist_autonumber': i + 1,
1985 }, extra))
1986 if not entry_result:
1987 failures += 1
1988 if failures >= max_failures:
1989 self.report_error(
1990 f'Skipping the remaining entries in playlist "{title}" since {failures} items failed extraction')
1991 break
1992 if keep_resolved_entries:
1993 resolved_entries[i] = (playlist_index, entry_result)
1994
1995 # Update with processed data
1996 ie_result['entries'] = [e for _, e in resolved_entries if e is not NO_DEFAULT]
1997 ie_result['requested_entries'] = [i for i, e in resolved_entries if e is not NO_DEFAULT]
1998 if ie_result['requested_entries'] == try_call(lambda: list(range(1, ie_result['playlist_count'] + 1))):
1999 # Do not set for full playlist
2000 ie_result.pop('requested_entries')
2001
2002 # Write the updated info to json
2003 if _infojson_written is True and self._write_info_json(
2004 'updated playlist', ie_result,
2005 self.prepare_filename(ie_copy, 'pl_infojson'), overwrite=True) is None:
2006 return
2007
2008 ie_result = self.run_all_pps('playlist', ie_result)
2009 self.to_screen(f'[download] Finished downloading playlist: {title}')
2010 return ie_result
2011
2012 @_handle_extraction_exceptions
2013 def __process_iterable_entry(self, entry, download, extra_info):
2014 return self.process_ie_result(
2015 entry, download=download, extra_info=extra_info)
2016
2017 def _build_format_filter(self, filter_spec):
2018 " Returns a function to filter the formats according to the filter_spec "
2019
2020 OPERATORS = {
2021 '<': operator.lt,
2022 '<=': operator.le,
2023 '>': operator.gt,
2024 '>=': operator.ge,
2025 '=': operator.eq,
2026 '!=': operator.ne,
2027 }
2028 operator_rex = re.compile(r'''(?x)\s*
2029 (?P<key>[\w.-]+)\s*
2030 (?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
2031 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)\s*
2032 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
2033 m = operator_rex.fullmatch(filter_spec)
2034 if m:
2035 try:
2036 comparison_value = int(m.group('value'))
2037 except ValueError:
2038 comparison_value = parse_filesize(m.group('value'))
2039 if comparison_value is None:
2040 comparison_value = parse_filesize(m.group('value') + 'B')
2041 if comparison_value is None:
2042 raise ValueError(
2043 'Invalid value %r in format specification %r' % (
2044 m.group('value'), filter_spec))
2045 op = OPERATORS[m.group('op')]
2046
2047 if not m:
2048 STR_OPERATORS = {
2049 '=': operator.eq,
2050 '^=': lambda attr, value: attr.startswith(value),
2051 '$=': lambda attr, value: attr.endswith(value),
2052 '*=': lambda attr, value: value in attr,
2053 '~=': lambda attr, value: value.search(attr) is not None
2054 }
2055 str_operator_rex = re.compile(r'''(?x)\s*
2056 (?P<key>[a-zA-Z0-9._-]+)\s*
2057 (?P<negation>!\s*)?(?P<op>%s)\s*(?P<none_inclusive>\?\s*)?
2058 (?P<quote>["'])?
2059 (?P<value>(?(quote)(?:(?!(?P=quote))[^\\]|\\.)+|[\w.-]+))
2060 (?(quote)(?P=quote))\s*
2061 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
2062 m = str_operator_rex.fullmatch(filter_spec)
2063 if m:
2064 if m.group('op') == '~=':
2065 comparison_value = re.compile(m.group('value'))
2066 else:
2067 comparison_value = re.sub(r'''\\([\\"'])''', r'\1', m.group('value'))
2068 str_op = STR_OPERATORS[m.group('op')]
2069 if m.group('negation'):
2070 op = lambda attr, value: not str_op(attr, value)
2071 else:
2072 op = str_op
2073
2074 if not m:
2075 raise SyntaxError('Invalid filter specification %r' % filter_spec)
2076
2077 def _filter(f):
2078 actual_value = f.get(m.group('key'))
2079 if actual_value is None:
2080 return m.group('none_inclusive')
2081 return op(actual_value, comparison_value)
2082 return _filter
2083
2084 def _check_formats(self, formats):
2085 for f in formats:
2086 self.to_screen('[info] Testing format %s' % f['format_id'])
2087 path = self.get_output_path('temp')
2088 if not self._ensure_dir_exists(f'{path}/'):
2089 continue
2090 temp_file = tempfile.NamedTemporaryFile(suffix='.tmp', delete=False, dir=path or None)
2091 temp_file.close()
2092 try:
2093 success, _ = self.dl(temp_file.name, f, test=True)
2094 except (DownloadError, OSError, ValueError) + network_exceptions:
2095 success = False
2096 finally:
2097 if os.path.exists(temp_file.name):
2098 try:
2099 os.remove(temp_file.name)
2100 except OSError:
2101 self.report_warning('Unable to delete temporary file "%s"' % temp_file.name)
2102 if success:
2103 yield f
2104 else:
2105 self.to_screen('[info] Unable to download format %s. Skipping...' % f['format_id'])
2106
2107 def _default_format_spec(self, info_dict, download=True):
2108
2109 def can_merge():
2110 merger = FFmpegMergerPP(self)
2111 return merger.available and merger.can_merge()
2112
2113 prefer_best = (
2114 not self.params.get('simulate')
2115 and download
2116 and (
2117 not can_merge()
2118 or info_dict.get('is_live') and not self.params.get('live_from_start')
2119 or self.params['outtmpl']['default'] == '-'))
2120 compat = (
2121 prefer_best
2122 or self.params.get('allow_multiple_audio_streams', False)
2123 or 'format-spec' in self.params['compat_opts'])
2124
2125 return (
2126 'best/bestvideo+bestaudio' if prefer_best
2127 else 'bestvideo*+bestaudio/best' if not compat
2128 else 'bestvideo+bestaudio/best')
2129
2130 def build_format_selector(self, format_spec):
2131 def syntax_error(note, start):
2132 message = (
2133 'Invalid format specification: '
2134 '{}\n\t{}\n\t{}^'.format(note, format_spec, ' ' * start[1]))
2135 return SyntaxError(message)
2136
2137 PICKFIRST = 'PICKFIRST'
2138 MERGE = 'MERGE'
2139 SINGLE = 'SINGLE'
2140 GROUP = 'GROUP'
2141 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
2142
2143 allow_multiple_streams = {'audio': self.params.get('allow_multiple_audio_streams', False),
2144 'video': self.params.get('allow_multiple_video_streams', False)}
2145
2146 def _parse_filter(tokens):
2147 filter_parts = []
2148 for type, string_, start, _, _ in tokens:
2149 if type == tokenize.OP and string_ == ']':
2150 return ''.join(filter_parts)
2151 else:
2152 filter_parts.append(string_)
2153
2154 def _remove_unused_ops(tokens):
2155 # Remove operators that we don't use and join them with the surrounding strings.
2156 # E.g. 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
2157 ALLOWED_OPS = ('/', '+', ',', '(', ')')
2158 last_string, last_start, last_end, last_line = None, None, None, None
2159 for type, string_, start, end, line in tokens:
2160 if type == tokenize.OP and string_ == '[':
2161 if last_string:
2162 yield tokenize.NAME, last_string, last_start, last_end, last_line
2163 last_string = None
2164 yield type, string_, start, end, line
2165 # everything inside brackets will be handled by _parse_filter
2166 for type, string_, start, end, line in tokens:
2167 yield type, string_, start, end, line
2168 if type == tokenize.OP and string_ == ']':
2169 break
2170 elif type == tokenize.OP and string_ in ALLOWED_OPS:
2171 if last_string:
2172 yield tokenize.NAME, last_string, last_start, last_end, last_line
2173 last_string = None
2174 yield type, string_, start, end, line
2175 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
2176 if not last_string:
2177 last_string = string_
2178 last_start = start
2179 last_end = end
2180 else:
2181 last_string += string_
2182 if last_string:
2183 yield tokenize.NAME, last_string, last_start, last_end, last_line
2184
2185 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
2186 selectors = []
2187 current_selector = None
2188 for type, string_, start, _, _ in tokens:
2189 # ENCODING is only defined in python 3.x
2190 if type == getattr(tokenize, 'ENCODING', None):
2191 continue
2192 elif type in [tokenize.NAME, tokenize.NUMBER]:
2193 current_selector = FormatSelector(SINGLE, string_, [])
2194 elif type == tokenize.OP:
2195 if string_ == ')':
2196 if not inside_group:
2197 # ')' will be handled by the parentheses group
2198 tokens.restore_last_token()
2199 break
2200 elif inside_merge and string_ in ['/', ',']:
2201 tokens.restore_last_token()
2202 break
2203 elif inside_choice and string_ == ',':
2204 tokens.restore_last_token()
2205 break
2206 elif string_ == ',':
2207 if not current_selector:
2208 raise syntax_error('"," must follow a format selector', start)
2209 selectors.append(current_selector)
2210 current_selector = None
2211 elif string_ == '/':
2212 if not current_selector:
2213 raise syntax_error('"/" must follow a format selector', start)
2214 first_choice = current_selector
2215 second_choice = _parse_format_selection(tokens, inside_choice=True)
2216 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
2217 elif string_ == '[':
2218 if not current_selector:
2219 current_selector = FormatSelector(SINGLE, 'best', [])
2220 format_filter = _parse_filter(tokens)
2221 current_selector.filters.append(format_filter)
2222 elif string_ == '(':
2223 if current_selector:
2224 raise syntax_error('Unexpected "("', start)
2225 group = _parse_format_selection(tokens, inside_group=True)
2226 current_selector = FormatSelector(GROUP, group, [])
2227 elif string_ == '+':
2228 if not current_selector:
2229 raise syntax_error('Unexpected "+"', start)
2230 selector_1 = current_selector
2231 selector_2 = _parse_format_selection(tokens, inside_merge=True)
2232 if not selector_2:
2233 raise syntax_error('Expected a selector', start)
2234 current_selector = FormatSelector(MERGE, (selector_1, selector_2), [])
2235 else:
2236 raise syntax_error(f'Operator not recognized: "{string_}"', start)
2237 elif type == tokenize.ENDMARKER:
2238 break
2239 if current_selector:
2240 selectors.append(current_selector)
2241 return selectors
2242
2243 def _merge(formats_pair):
2244 format_1, format_2 = formats_pair
2245
2246 formats_info = []
2247 formats_info.extend(format_1.get('requested_formats', (format_1,)))
2248 formats_info.extend(format_2.get('requested_formats', (format_2,)))
2249
2250 if not allow_multiple_streams['video'] or not allow_multiple_streams['audio']:
2251 get_no_more = {'video': False, 'audio': False}
2252 for (i, fmt_info) in enumerate(formats_info):
2253 if fmt_info.get('acodec') == fmt_info.get('vcodec') == 'none':
2254 formats_info.pop(i)
2255 continue
2256 for aud_vid in ['audio', 'video']:
2257 if not allow_multiple_streams[aud_vid] and fmt_info.get(aud_vid[0] + 'codec') != 'none':
2258 if get_no_more[aud_vid]:
2259 formats_info.pop(i)
2260 break
2261 get_no_more[aud_vid] = True
2262
2263 if len(formats_info) == 1:
2264 return formats_info[0]
2265
2266 video_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('vcodec') != 'none']
2267 audio_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('acodec') != 'none']
2268
2269 the_only_video = video_fmts[0] if len(video_fmts) == 1 else None
2270 the_only_audio = audio_fmts[0] if len(audio_fmts) == 1 else None
2271
2272 output_ext = get_compatible_ext(
2273 vcodecs=[f.get('vcodec') for f in video_fmts],
2274 acodecs=[f.get('acodec') for f in audio_fmts],
2275 vexts=[f['ext'] for f in video_fmts],
2276 aexts=[f['ext'] for f in audio_fmts],
2277 preferences=(try_call(lambda: self.params['merge_output_format'].split('/'))
2278 or self.params.get('prefer_free_formats') and ('webm', 'mkv')))
2279
2280 filtered = lambda *keys: filter(None, (traverse_obj(fmt, *keys) for fmt in formats_info))
2281
2282 new_dict = {
2283 'requested_formats': formats_info,
2284 'format': '+'.join(filtered('format')),
2285 'format_id': '+'.join(filtered('format_id')),
2286 'ext': output_ext,
2287 'protocol': '+'.join(map(determine_protocol, formats_info)),
2288 'language': '+'.join(orderedSet(filtered('language'))) or None,
2289 'format_note': '+'.join(orderedSet(filtered('format_note'))) or None,
2290 'filesize_approx': sum(filtered('filesize', 'filesize_approx')) or None,
2291 'tbr': sum(filtered('tbr', 'vbr', 'abr')),
2292 }
2293
2294 if the_only_video:
2295 new_dict.update({
2296 'width': the_only_video.get('width'),
2297 'height': the_only_video.get('height'),
2298 'resolution': the_only_video.get('resolution') or self.format_resolution(the_only_video),
2299 'fps': the_only_video.get('fps'),
2300 'dynamic_range': the_only_video.get('dynamic_range'),
2301 'vcodec': the_only_video.get('vcodec'),
2302 'vbr': the_only_video.get('vbr'),
2303 'stretched_ratio': the_only_video.get('stretched_ratio'),
2304 'aspect_ratio': the_only_video.get('aspect_ratio'),
2305 })
2306
2307 if the_only_audio:
2308 new_dict.update({
2309 'acodec': the_only_audio.get('acodec'),
2310 'abr': the_only_audio.get('abr'),
2311 'asr': the_only_audio.get('asr'),
2312 'audio_channels': the_only_audio.get('audio_channels')
2313 })
2314
2315 return new_dict
2316
2317 def _check_formats(formats):
2318 if (self.params.get('check_formats') is not None
2319 or self.params.get('allow_unplayable_formats')):
2320 yield from formats
2321 return
2322 elif self.params.get('check_formats') == 'selected':
2323 yield from self._check_formats(formats)
2324 return
2325
2326 for f in formats:
2327 if f.get('has_drm'):
2328 yield from self._check_formats([f])
2329 else:
2330 yield f
2331
2332 def _build_selector_function(selector):
2333 if isinstance(selector, list): # ,
2334 fs = [_build_selector_function(s) for s in selector]
2335
2336 def selector_function(ctx):
2337 for f in fs:
2338 yield from f(ctx)
2339 return selector_function
2340
2341 elif selector.type == GROUP: # ()
2342 selector_function = _build_selector_function(selector.selector)
2343
2344 elif selector.type == PICKFIRST: # /
2345 fs = [_build_selector_function(s) for s in selector.selector]
2346
2347 def selector_function(ctx):
2348 for f in fs:
2349 picked_formats = list(f(ctx))
2350 if picked_formats:
2351 return picked_formats
2352 return []
2353
2354 elif selector.type == MERGE: # +
2355 selector_1, selector_2 = map(_build_selector_function, selector.selector)
2356
2357 def selector_function(ctx):
2358 for pair in itertools.product(selector_1(ctx), selector_2(ctx)):
2359 yield _merge(pair)
2360
2361 elif selector.type == SINGLE: # atom
2362 format_spec = selector.selector or 'best'
2363
2364 # TODO: Add allvideo, allaudio etc by generalizing the code with best/worst selector
2365 if format_spec == 'all':
2366 def selector_function(ctx):
2367 yield from _check_formats(ctx['formats'][::-1])
2368 elif format_spec == 'mergeall':
2369 def selector_function(ctx):
2370 formats = list(_check_formats(
2371 f for f in ctx['formats'] if f.get('vcodec') != 'none' or f.get('acodec') != 'none'))
2372 if not formats:
2373 return
2374 merged_format = formats[-1]
2375 for f in formats[-2::-1]:
2376 merged_format = _merge((merged_format, f))
2377 yield merged_format
2378
2379 else:
2380 format_fallback, seperate_fallback, format_reverse, format_idx = False, None, True, 1
2381 mobj = re.match(
2382 r'(?P<bw>best|worst|b|w)(?P<type>video|audio|v|a)?(?P<mod>\*)?(?:\.(?P<n>[1-9]\d*))?$',
2383 format_spec)
2384 if mobj is not None:
2385 format_idx = int_or_none(mobj.group('n'), default=1)
2386 format_reverse = mobj.group('bw')[0] == 'b'
2387 format_type = (mobj.group('type') or [None])[0]
2388 not_format_type = {'v': 'a', 'a': 'v'}.get(format_type)
2389 format_modified = mobj.group('mod') is not None
2390
2391 format_fallback = not format_type and not format_modified # for b, w
2392 _filter_f = (
2393 (lambda f: f.get('%scodec' % format_type) != 'none')
2394 if format_type and format_modified # bv*, ba*, wv*, wa*
2395 else (lambda f: f.get('%scodec' % not_format_type) == 'none')
2396 if format_type # bv, ba, wv, wa
2397 else (lambda f: f.get('vcodec') != 'none' and f.get('acodec') != 'none')
2398 if not format_modified # b, w
2399 else lambda f: True) # b*, w*
2400 filter_f = lambda f: _filter_f(f) and (
2401 f.get('vcodec') != 'none' or f.get('acodec') != 'none')
2402 else:
2403 if format_spec in self._format_selection_exts['audio']:
2404 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none'
2405 elif format_spec in self._format_selection_exts['video']:
2406 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none' and f.get('vcodec') != 'none'
2407 seperate_fallback = lambda f: f.get('ext') == format_spec and f.get('vcodec') != 'none'
2408 elif format_spec in self._format_selection_exts['storyboards']:
2409 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') == 'none' and f.get('vcodec') == 'none'
2410 else:
2411 filter_f = lambda f: f.get('format_id') == format_spec # id
2412
2413 def selector_function(ctx):
2414 formats = list(ctx['formats'])
2415 matches = list(filter(filter_f, formats)) if filter_f is not None else formats
2416 if not matches:
2417 if format_fallback and ctx['incomplete_formats']:
2418 # for extractors with incomplete formats (audio only (soundcloud)
2419 # or video only (imgur)) best/worst will fallback to
2420 # best/worst {video,audio}-only format
2421 matches = formats
2422 elif seperate_fallback and not ctx['has_merged_format']:
2423 # for compatibility with youtube-dl when there is no pre-merged format
2424 matches = list(filter(seperate_fallback, formats))
2425 matches = LazyList(_check_formats(matches[::-1 if format_reverse else 1]))
2426 try:
2427 yield matches[format_idx - 1]
2428 except LazyList.IndexError:
2429 return
2430
2431 filters = [self._build_format_filter(f) for f in selector.filters]
2432
2433 def final_selector(ctx):
2434 ctx_copy = dict(ctx)
2435 for _filter in filters:
2436 ctx_copy['formats'] = list(filter(_filter, ctx_copy['formats']))
2437 return selector_function(ctx_copy)
2438 return final_selector
2439
2440 stream = io.BytesIO(format_spec.encode())
2441 try:
2442 tokens = list(_remove_unused_ops(tokenize.tokenize(stream.readline)))
2443 except tokenize.TokenError:
2444 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
2445
2446 class TokenIterator:
2447 def __init__(self, tokens):
2448 self.tokens = tokens
2449 self.counter = 0
2450
2451 def __iter__(self):
2452 return self
2453
2454 def __next__(self):
2455 if self.counter >= len(self.tokens):
2456 raise StopIteration()
2457 value = self.tokens[self.counter]
2458 self.counter += 1
2459 return value
2460
2461 next = __next__
2462
2463 def restore_last_token(self):
2464 self.counter -= 1
2465
2466 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
2467 return _build_selector_function(parsed_selector)
2468
2469 def _calc_headers(self, info_dict):
2470 res = merge_headers(self.params['http_headers'], info_dict.get('http_headers') or {})
2471 if 'Youtubedl-No-Compression' in res: # deprecated
2472 res.pop('Youtubedl-No-Compression', None)
2473 res['Accept-Encoding'] = 'identity'
2474 cookies = self.cookiejar.get_cookies_for_url(info_dict['url'])
2475 if cookies:
2476 encoder = LenientSimpleCookie()
2477 values = []
2478 for cookie in cookies:
2479 _, value = encoder.value_encode(cookie.value)
2480 values.append(f'{cookie.name}={value}')
2481 if cookie.domain:
2482 values.append(f'Domain={cookie.domain}')
2483 if cookie.path:
2484 values.append(f'Path={cookie.path}')
2485 if cookie.secure:
2486 values.append('Secure')
2487 if cookie.expires:
2488 values.append(f'Expires={cookie.expires}')
2489 if cookie.version:
2490 values.append(f'Version={cookie.version}')
2491 info_dict['cookies'] = '; '.join(values)
2492
2493 if 'X-Forwarded-For' not in res:
2494 x_forwarded_for_ip = info_dict.get('__x_forwarded_for_ip')
2495 if x_forwarded_for_ip:
2496 res['X-Forwarded-For'] = x_forwarded_for_ip
2497
2498 return res
2499
2500 def _calc_cookies(self, url):
2501 self.deprecation_warning('"YoutubeDL._calc_cookies" is deprecated and may be removed in a future version')
2502 return self.cookiejar.get_cookie_header(url)
2503
2504 def _sort_thumbnails(self, thumbnails):
2505 thumbnails.sort(key=lambda t: (
2506 t.get('preference') if t.get('preference') is not None else -1,
2507 t.get('width') if t.get('width') is not None else -1,
2508 t.get('height') if t.get('height') is not None else -1,
2509 t.get('id') if t.get('id') is not None else '',
2510 t.get('url')))
2511
2512 def _sanitize_thumbnails(self, info_dict):
2513 thumbnails = info_dict.get('thumbnails')
2514 if thumbnails is None:
2515 thumbnail = info_dict.get('thumbnail')
2516 if thumbnail:
2517 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
2518 if not thumbnails:
2519 return
2520
2521 def check_thumbnails(thumbnails):
2522 for t in thumbnails:
2523 self.to_screen(f'[info] Testing thumbnail {t["id"]}')
2524 try:
2525 self.urlopen(HEADRequest(t['url']))
2526 except network_exceptions as err:
2527 self.to_screen(f'[info] Unable to connect to thumbnail {t["id"]} URL {t["url"]!r} - {err}. Skipping...')
2528 continue
2529 yield t
2530
2531 self._sort_thumbnails(thumbnails)
2532 for i, t in enumerate(thumbnails):
2533 if t.get('id') is None:
2534 t['id'] = '%d' % i
2535 if t.get('width') and t.get('height'):
2536 t['resolution'] = '%dx%d' % (t['width'], t['height'])
2537 t['url'] = sanitize_url(t['url'])
2538
2539 if self.params.get('check_formats') is True:
2540 info_dict['thumbnails'] = LazyList(check_thumbnails(thumbnails[::-1]), reverse=True)
2541 else:
2542 info_dict['thumbnails'] = thumbnails
2543
2544 def _fill_common_fields(self, info_dict, final=True):
2545 # TODO: move sanitization here
2546 if final:
2547 title = info_dict['fulltitle'] = info_dict.get('title')
2548 if not title:
2549 if title == '':
2550 self.write_debug('Extractor gave empty title. Creating a generic title')
2551 else:
2552 self.report_warning('Extractor failed to obtain "title". Creating a generic title instead')
2553 info_dict['title'] = f'{info_dict["extractor"].replace(":", "-")} video #{info_dict["id"]}'
2554
2555 if info_dict.get('duration') is not None:
2556 info_dict['duration_string'] = formatSeconds(info_dict['duration'])
2557
2558 for ts_key, date_key in (
2559 ('timestamp', 'upload_date'),
2560 ('release_timestamp', 'release_date'),
2561 ('modified_timestamp', 'modified_date'),
2562 ):
2563 if info_dict.get(date_key) is None and info_dict.get(ts_key) is not None:
2564 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
2565 # see http://bugs.python.org/issue1646728)
2566 with contextlib.suppress(ValueError, OverflowError, OSError):
2567 upload_date = datetime.datetime.utcfromtimestamp(info_dict[ts_key])
2568 info_dict[date_key] = upload_date.strftime('%Y%m%d')
2569
2570 live_keys = ('is_live', 'was_live')
2571 live_status = info_dict.get('live_status')
2572 if live_status is None:
2573 for key in live_keys:
2574 if info_dict.get(key) is False:
2575 continue
2576 if info_dict.get(key):
2577 live_status = key
2578 break
2579 if all(info_dict.get(key) is False for key in live_keys):
2580 live_status = 'not_live'
2581 if live_status:
2582 info_dict['live_status'] = live_status
2583 for key in live_keys:
2584 if info_dict.get(key) is None:
2585 info_dict[key] = (live_status == key)
2586 if live_status == 'post_live':
2587 info_dict['was_live'] = True
2588
2589 # Auto generate title fields corresponding to the *_number fields when missing
2590 # in order to always have clean titles. This is very common for TV series.
2591 for field in ('chapter', 'season', 'episode'):
2592 if final and info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
2593 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
2594
2595 def _raise_pending_errors(self, info):
2596 err = info.pop('__pending_error', None)
2597 if err:
2598 self.report_error(err, tb=False)
2599
2600 def sort_formats(self, info_dict):
2601 formats = self._get_formats(info_dict)
2602 formats.sort(key=FormatSorter(
2603 self, info_dict.get('_format_sort_fields') or []).calculate_preference)
2604
2605 def process_video_result(self, info_dict, download=True):
2606 assert info_dict.get('_type', 'video') == 'video'
2607 self._num_videos += 1
2608
2609 if 'id' not in info_dict:
2610 raise ExtractorError('Missing "id" field in extractor result', ie=info_dict['extractor'])
2611 elif not info_dict.get('id'):
2612 raise ExtractorError('Extractor failed to obtain "id"', ie=info_dict['extractor'])
2613
2614 def report_force_conversion(field, field_not, conversion):
2615 self.report_warning(
2616 '"%s" field is not %s - forcing %s conversion, there is an error in extractor'
2617 % (field, field_not, conversion))
2618
2619 def sanitize_string_field(info, string_field):
2620 field = info.get(string_field)
2621 if field is None or isinstance(field, str):
2622 return
2623 report_force_conversion(string_field, 'a string', 'string')
2624 info[string_field] = str(field)
2625
2626 def sanitize_numeric_fields(info):
2627 for numeric_field in self._NUMERIC_FIELDS:
2628 field = info.get(numeric_field)
2629 if field is None or isinstance(field, (int, float)):
2630 continue
2631 report_force_conversion(numeric_field, 'numeric', 'int')
2632 info[numeric_field] = int_or_none(field)
2633
2634 sanitize_string_field(info_dict, 'id')
2635 sanitize_numeric_fields(info_dict)
2636 if info_dict.get('section_end') and info_dict.get('section_start') is not None:
2637 info_dict['duration'] = round(info_dict['section_end'] - info_dict['section_start'], 3)
2638 if (info_dict.get('duration') or 0) <= 0 and info_dict.pop('duration', None):
2639 self.report_warning('"duration" field is negative, there is an error in extractor')
2640
2641 chapters = info_dict.get('chapters') or []
2642 if chapters and chapters[0].get('start_time'):
2643 chapters.insert(0, {'start_time': 0})
2644
2645 dummy_chapter = {'end_time': 0, 'start_time': info_dict.get('duration')}
2646 for idx, (prev, current, next_) in enumerate(zip(
2647 (dummy_chapter, *chapters), chapters, (*chapters[1:], dummy_chapter)), 1):
2648 if current.get('start_time') is None:
2649 current['start_time'] = prev.get('end_time')
2650 if not current.get('end_time'):
2651 current['end_time'] = next_.get('start_time')
2652 if not current.get('title'):
2653 current['title'] = f'<Untitled Chapter {idx}>'
2654
2655 if 'playlist' not in info_dict:
2656 # It isn't part of a playlist
2657 info_dict['playlist'] = None
2658 info_dict['playlist_index'] = None
2659
2660 self._sanitize_thumbnails(info_dict)
2661
2662 thumbnail = info_dict.get('thumbnail')
2663 thumbnails = info_dict.get('thumbnails')
2664 if thumbnail:
2665 info_dict['thumbnail'] = sanitize_url(thumbnail)
2666 elif thumbnails:
2667 info_dict['thumbnail'] = thumbnails[-1]['url']
2668
2669 if info_dict.get('display_id') is None and 'id' in info_dict:
2670 info_dict['display_id'] = info_dict['id']
2671
2672 self._fill_common_fields(info_dict)
2673
2674 for cc_kind in ('subtitles', 'automatic_captions'):
2675 cc = info_dict.get(cc_kind)
2676 if cc:
2677 for _, subtitle in cc.items():
2678 for subtitle_format in subtitle:
2679 if subtitle_format.get('url'):
2680 subtitle_format['url'] = sanitize_url(subtitle_format['url'])
2681 if subtitle_format.get('ext') is None:
2682 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
2683
2684 automatic_captions = info_dict.get('automatic_captions')
2685 subtitles = info_dict.get('subtitles')
2686
2687 info_dict['requested_subtitles'] = self.process_subtitles(
2688 info_dict['id'], subtitles, automatic_captions)
2689
2690 formats = self._get_formats(info_dict)
2691
2692 # Backward compatibility with InfoExtractor._sort_formats
2693 field_preference = (formats or [{}])[0].pop('__sort_fields', None)
2694 if field_preference:
2695 info_dict['_format_sort_fields'] = field_preference
2696
2697 info_dict['_has_drm'] = any( # or None ensures --clean-infojson removes it
2698 f.get('has_drm') and f['has_drm'] != 'maybe' for f in formats) or None
2699 if not self.params.get('allow_unplayable_formats'):
2700 formats = [f for f in formats if not f.get('has_drm') or f['has_drm'] == 'maybe']
2701
2702 if formats and all(f.get('acodec') == f.get('vcodec') == 'none' for f in formats):
2703 self.report_warning(
2704 f'{"This video is DRM protected and " if info_dict["_has_drm"] else ""}'
2705 'only images are available for download. Use --list-formats to see them'.capitalize())
2706
2707 get_from_start = not info_dict.get('is_live') or bool(self.params.get('live_from_start'))
2708 if not get_from_start:
2709 info_dict['title'] += ' ' + datetime.datetime.now().strftime('%Y-%m-%d %H:%M')
2710 if info_dict.get('is_live') and formats:
2711 formats = [f for f in formats if bool(f.get('is_from_start')) == get_from_start]
2712 if get_from_start and not formats:
2713 self.raise_no_formats(info_dict, msg=(
2714 '--live-from-start is passed, but there are no formats that can be downloaded from the start. '
2715 'If you want to download from the current time, use --no-live-from-start'))
2716
2717 def is_wellformed(f):
2718 url = f.get('url')
2719 if not url:
2720 self.report_warning(
2721 '"url" field is missing or empty - skipping format, '
2722 'there is an error in extractor')
2723 return False
2724 if isinstance(url, bytes):
2725 sanitize_string_field(f, 'url')
2726 return True
2727
2728 # Filter out malformed formats for better extraction robustness
2729 formats = list(filter(is_wellformed, formats or []))
2730
2731 if not formats:
2732 self.raise_no_formats(info_dict)
2733
2734 for format in formats:
2735 sanitize_string_field(format, 'format_id')
2736 sanitize_numeric_fields(format)
2737 format['url'] = sanitize_url(format['url'])
2738 if format.get('ext') is None:
2739 format['ext'] = determine_ext(format['url']).lower()
2740 if format.get('protocol') is None:
2741 format['protocol'] = determine_protocol(format)
2742 if format.get('resolution') is None:
2743 format['resolution'] = self.format_resolution(format, default=None)
2744 if format.get('dynamic_range') is None and format.get('vcodec') != 'none':
2745 format['dynamic_range'] = 'SDR'
2746 if format.get('aspect_ratio') is None:
2747 format['aspect_ratio'] = try_call(lambda: round(format['width'] / format['height'], 2))
2748 if (not format.get('manifest_url') # For fragmented formats, "tbr" is often max bitrate and not average
2749 and info_dict.get('duration') and format.get('tbr')
2750 and not format.get('filesize') and not format.get('filesize_approx')):
2751 format['filesize_approx'] = int(info_dict['duration'] * format['tbr'] * (1024 / 8))
2752 format['http_headers'] = self._calc_headers(collections.ChainMap(format, info_dict))
2753
2754 # This is copied to http_headers by the above _calc_headers and can now be removed
2755 if '__x_forwarded_for_ip' in info_dict:
2756 del info_dict['__x_forwarded_for_ip']
2757
2758 self.sort_formats({
2759 'formats': formats,
2760 '_format_sort_fields': info_dict.get('_format_sort_fields')
2761 })
2762
2763 # Sanitize and group by format_id
2764 formats_dict = {}
2765 for i, format in enumerate(formats):
2766 if not format.get('format_id'):
2767 format['format_id'] = str(i)
2768 else:
2769 # Sanitize format_id from characters used in format selector expression
2770 format['format_id'] = re.sub(r'[\s,/+\[\]()]', '_', format['format_id'])
2771 formats_dict.setdefault(format['format_id'], []).append(format)
2772
2773 # Make sure all formats have unique format_id
2774 common_exts = set(itertools.chain(*self._format_selection_exts.values()))
2775 for format_id, ambiguous_formats in formats_dict.items():
2776 ambigious_id = len(ambiguous_formats) > 1
2777 for i, format in enumerate(ambiguous_formats):
2778 if ambigious_id:
2779 format['format_id'] = '%s-%d' % (format_id, i)
2780 # Ensure there is no conflict between id and ext in format selection
2781 # See https://github.com/yt-dlp/yt-dlp/issues/1282
2782 if format['format_id'] != format['ext'] and format['format_id'] in common_exts:
2783 format['format_id'] = 'f%s' % format['format_id']
2784
2785 if format.get('format') is None:
2786 format['format'] = '{id} - {res}{note}'.format(
2787 id=format['format_id'],
2788 res=self.format_resolution(format),
2789 note=format_field(format, 'format_note', ' (%s)'),
2790 )
2791
2792 if self.params.get('check_formats') is True:
2793 formats = LazyList(self._check_formats(formats[::-1]), reverse=True)
2794
2795 if not formats or formats[0] is not info_dict:
2796 # only set the 'formats' fields if the original info_dict list them
2797 # otherwise we end up with a circular reference, the first (and unique)
2798 # element in the 'formats' field in info_dict is info_dict itself,
2799 # which can't be exported to json
2800 info_dict['formats'] = formats
2801
2802 info_dict, _ = self.pre_process(info_dict)
2803
2804 if self._match_entry(info_dict, incomplete=self._format_fields) is not None:
2805 return info_dict
2806
2807 self.post_extract(info_dict)
2808 info_dict, _ = self.pre_process(info_dict, 'after_filter')
2809
2810 # The pre-processors may have modified the formats
2811 formats = self._get_formats(info_dict)
2812
2813 list_only = self.params.get('simulate') == 'list_only'
2814 interactive_format_selection = not list_only and self.format_selector == '-'
2815 if self.params.get('list_thumbnails'):
2816 self.list_thumbnails(info_dict)
2817 if self.params.get('listsubtitles'):
2818 if 'automatic_captions' in info_dict:
2819 self.list_subtitles(
2820 info_dict['id'], automatic_captions, 'automatic captions')
2821 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
2822 if self.params.get('listformats') or interactive_format_selection:
2823 self.list_formats(info_dict)
2824 if list_only:
2825 # Without this printing, -F --print-json will not work
2826 self.__forced_printings(info_dict)
2827 return info_dict
2828
2829 format_selector = self.format_selector
2830 while True:
2831 if interactive_format_selection:
2832 req_format = input(self._format_screen('\nEnter format selector ', self.Styles.EMPHASIS)
2833 + '(Press ENTER for default, or Ctrl+C to quit)'
2834 + self._format_screen(': ', self.Styles.EMPHASIS))
2835 try:
2836 format_selector = self.build_format_selector(req_format) if req_format else None
2837 except SyntaxError as err:
2838 self.report_error(err, tb=False, is_error=False)
2839 continue
2840
2841 if format_selector is None:
2842 req_format = self._default_format_spec(info_dict, download=download)
2843 self.write_debug(f'Default format spec: {req_format}')
2844 format_selector = self.build_format_selector(req_format)
2845
2846 formats_to_download = list(format_selector({
2847 'formats': formats,
2848 'has_merged_format': any('none' not in (f.get('acodec'), f.get('vcodec')) for f in formats),
2849 'incomplete_formats': (all(f.get('vcodec') == 'none' for f in formats) # No formats with video
2850 or all(f.get('acodec') == 'none' for f in formats)), # OR, No formats with audio
2851 }))
2852 if interactive_format_selection and not formats_to_download:
2853 self.report_error('Requested format is not available', tb=False, is_error=False)
2854 continue
2855 break
2856
2857 if not formats_to_download:
2858 if not self.params.get('ignore_no_formats_error'):
2859 raise ExtractorError(
2860 'Requested format is not available. Use --list-formats for a list of available formats',
2861 expected=True, video_id=info_dict['id'], ie=info_dict['extractor'])
2862 self.report_warning('Requested format is not available')
2863 # Process what we can, even without any available formats.
2864 formats_to_download = [{}]
2865
2866 requested_ranges = tuple(self.params.get('download_ranges', lambda *_: [{}])(info_dict, self))
2867 best_format, downloaded_formats = formats_to_download[-1], []
2868 if download:
2869 if best_format and requested_ranges:
2870 def to_screen(*msg):
2871 self.to_screen(f'[info] {info_dict["id"]}: {" ".join(", ".join(variadic(m)) for m in msg)}')
2872
2873 to_screen(f'Downloading {len(formats_to_download)} format(s):',
2874 (f['format_id'] for f in formats_to_download))
2875 if requested_ranges != ({}, ):
2876 to_screen(f'Downloading {len(requested_ranges)} time ranges:',
2877 (f'{c["start_time"]:.1f}-{c["end_time"]:.1f}' for c in requested_ranges))
2878 max_downloads_reached = False
2879
2880 for fmt, chapter in itertools.product(formats_to_download, requested_ranges):
2881 new_info = self._copy_infodict(info_dict)
2882 new_info.update(fmt)
2883 offset, duration = info_dict.get('section_start') or 0, info_dict.get('duration') or float('inf')
2884 end_time = offset + min(chapter.get('end_time', duration), duration)
2885 # duration may not be accurate. So allow deviations <1sec
2886 if end_time == float('inf') or end_time > offset + duration + 1:
2887 end_time = None
2888 if chapter or offset:
2889 new_info.update({
2890 'section_start': offset + chapter.get('start_time', 0),
2891 'section_end': end_time,
2892 'section_title': chapter.get('title'),
2893 'section_number': chapter.get('index'),
2894 })
2895 downloaded_formats.append(new_info)
2896 try:
2897 self.process_info(new_info)
2898 except MaxDownloadsReached:
2899 max_downloads_reached = True
2900 self._raise_pending_errors(new_info)
2901 # Remove copied info
2902 for key, val in tuple(new_info.items()):
2903 if info_dict.get(key) == val:
2904 new_info.pop(key)
2905 if max_downloads_reached:
2906 break
2907
2908 write_archive = {f.get('__write_download_archive', False) for f in downloaded_formats}
2909 assert write_archive.issubset({True, False, 'ignore'})
2910 if True in write_archive and False not in write_archive:
2911 self.record_download_archive(info_dict)
2912
2913 info_dict['requested_downloads'] = downloaded_formats
2914 info_dict = self.run_all_pps('after_video', info_dict)
2915 if max_downloads_reached:
2916 raise MaxDownloadsReached()
2917
2918 # We update the info dict with the selected best quality format (backwards compatibility)
2919 info_dict.update(best_format)
2920 return info_dict
2921
2922 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
2923 """Select the requested subtitles and their format"""
2924 available_subs, normal_sub_langs = {}, []
2925 if normal_subtitles and self.params.get('writesubtitles'):
2926 available_subs.update(normal_subtitles)
2927 normal_sub_langs = tuple(normal_subtitles.keys())
2928 if automatic_captions and self.params.get('writeautomaticsub'):
2929 for lang, cap_info in automatic_captions.items():
2930 if lang not in available_subs:
2931 available_subs[lang] = cap_info
2932
2933 if not available_subs or (
2934 not self.params.get('writesubtitles')
2935 and not self.params.get('writeautomaticsub')):
2936 return None
2937
2938 all_sub_langs = tuple(available_subs.keys())
2939 if self.params.get('allsubtitles', False):
2940 requested_langs = all_sub_langs
2941 elif self.params.get('subtitleslangs', False):
2942 try:
2943 requested_langs = orderedSet_from_options(
2944 self.params.get('subtitleslangs'), {'all': all_sub_langs}, use_regex=True)
2945 except re.error as e:
2946 raise ValueError(f'Wrong regex for subtitlelangs: {e.pattern}')
2947 else:
2948 requested_langs = LazyList(itertools.chain(
2949 ['en'] if 'en' in normal_sub_langs else [],
2950 filter(lambda f: f.startswith('en'), normal_sub_langs),
2951 ['en'] if 'en' in all_sub_langs else [],
2952 filter(lambda f: f.startswith('en'), all_sub_langs),
2953 normal_sub_langs, all_sub_langs,
2954 ))[:1]
2955 if requested_langs:
2956 self.to_screen(f'[info] {video_id}: Downloading subtitles: {", ".join(requested_langs)}')
2957
2958 formats_query = self.params.get('subtitlesformat', 'best')
2959 formats_preference = formats_query.split('/') if formats_query else []
2960 subs = {}
2961 for lang in requested_langs:
2962 formats = available_subs.get(lang)
2963 if formats is None:
2964 self.report_warning(f'{lang} subtitles not available for {video_id}')
2965 continue
2966 for ext in formats_preference:
2967 if ext == 'best':
2968 f = formats[-1]
2969 break
2970 matches = list(filter(lambda f: f['ext'] == ext, formats))
2971 if matches:
2972 f = matches[-1]
2973 break
2974 else:
2975 f = formats[-1]
2976 self.report_warning(
2977 'No subtitle format found matching "%s" for language %s, '
2978 'using %s' % (formats_query, lang, f['ext']))
2979 subs[lang] = f
2980 return subs
2981
2982 def _forceprint(self, key, info_dict):
2983 if info_dict is None:
2984 return
2985 info_copy = info_dict.copy()
2986 info_copy.setdefault('filename', self.prepare_filename(info_dict))
2987 if info_dict.get('requested_formats') is not None:
2988 # For RTMP URLs, also include the playpath
2989 info_copy['urls'] = '\n'.join(f['url'] + f.get('play_path', '') for f in info_dict['requested_formats'])
2990 elif info_dict.get('url'):
2991 info_copy['urls'] = info_dict['url'] + info_dict.get('play_path', '')
2992 info_copy['formats_table'] = self.render_formats_table(info_dict)
2993 info_copy['thumbnails_table'] = self.render_thumbnails_table(info_dict)
2994 info_copy['subtitles_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('subtitles'))
2995 info_copy['automatic_captions_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('automatic_captions'))
2996
2997 def format_tmpl(tmpl):
2998 mobj = re.fullmatch(r'([\w.:,]|-\d|(?P<dict>{([\w.:,]|-\d)+}))+=?', tmpl)
2999 if not mobj:
3000 return tmpl
3001
3002 fmt = '%({})s'
3003 if tmpl.startswith('{'):
3004 tmpl, fmt = f'.{tmpl}', '%({})j'
3005 if tmpl.endswith('='):
3006 tmpl, fmt = tmpl[:-1], '{0} = %({0})#j'
3007 return '\n'.join(map(fmt.format, [tmpl] if mobj.group('dict') else tmpl.split(',')))
3008
3009 for tmpl in self.params['forceprint'].get(key, []):
3010 self.to_stdout(self.evaluate_outtmpl(format_tmpl(tmpl), info_copy))
3011
3012 for tmpl, file_tmpl in self.params['print_to_file'].get(key, []):
3013 filename = self.prepare_filename(info_dict, outtmpl=file_tmpl)
3014 tmpl = format_tmpl(tmpl)
3015 self.to_screen(f'[info] Writing {tmpl!r} to: {filename}')
3016 if self._ensure_dir_exists(filename):
3017 with open(filename, 'a', encoding='utf-8', newline='') as f:
3018 f.write(self.evaluate_outtmpl(tmpl, info_copy) + os.linesep)
3019
3020 return info_copy
3021
3022 def __forced_printings(self, info_dict, filename=None, incomplete=True):
3023 if (self.params.get('forcejson')
3024 or self.params['forceprint'].get('video')
3025 or self.params['print_to_file'].get('video')):
3026 self.post_extract(info_dict)
3027 if filename:
3028 info_dict['filename'] = filename
3029 info_copy = self._forceprint('video', info_dict)
3030
3031 def print_field(field, actual_field=None, optional=False):
3032 if actual_field is None:
3033 actual_field = field
3034 if self.params.get(f'force{field}') and (
3035 info_copy.get(field) is not None or (not optional and not incomplete)):
3036 self.to_stdout(info_copy[actual_field])
3037
3038 print_field('title')
3039 print_field('id')
3040 print_field('url', 'urls')
3041 print_field('thumbnail', optional=True)
3042 print_field('description', optional=True)
3043 print_field('filename')
3044 if self.params.get('forceduration') and info_copy.get('duration') is not None:
3045 self.to_stdout(formatSeconds(info_copy['duration']))
3046 print_field('format')
3047
3048 if self.params.get('forcejson'):
3049 self.to_stdout(json.dumps(self.sanitize_info(info_dict)))
3050
3051 def dl(self, name, info, subtitle=False, test=False):
3052 if not info.get('url'):
3053 self.raise_no_formats(info, True)
3054
3055 if test:
3056 verbose = self.params.get('verbose')
3057 params = {
3058 'test': True,
3059 'quiet': self.params.get('quiet') or not verbose,
3060 'verbose': verbose,
3061 'noprogress': not verbose,
3062 'nopart': True,
3063 'skip_unavailable_fragments': False,
3064 'keep_fragments': False,
3065 'overwrites': True,
3066 '_no_ytdl_file': True,
3067 }
3068 else:
3069 params = self.params
3070 fd = get_suitable_downloader(info, params, to_stdout=(name == '-'))(self, params)
3071 if not test:
3072 for ph in self._progress_hooks:
3073 fd.add_progress_hook(ph)
3074 urls = '", "'.join(
3075 (f['url'].split(',')[0] + ',<data>' if f['url'].startswith('data:') else f['url'])
3076 for f in info.get('requested_formats', []) or [info])
3077 self.write_debug(f'Invoking {fd.FD_NAME} downloader on "{urls}"')
3078
3079 # Note: Ideally info should be a deep-copied so that hooks cannot modify it.
3080 # But it may contain objects that are not deep-copyable
3081 new_info = self._copy_infodict(info)
3082 if new_info.get('http_headers') is None:
3083 new_info['http_headers'] = self._calc_headers(new_info)
3084 return fd.download(name, new_info, subtitle)
3085
3086 def existing_file(self, filepaths, *, default_overwrite=True):
3087 existing_files = list(filter(os.path.exists, orderedSet(filepaths)))
3088 if existing_files and not self.params.get('overwrites', default_overwrite):
3089 return existing_files[0]
3090
3091 for file in existing_files:
3092 self.report_file_delete(file)
3093 os.remove(file)
3094 return None
3095
3096 def process_info(self, info_dict):
3097 """Process a single resolved IE result. (Modifies it in-place)"""
3098
3099 assert info_dict.get('_type', 'video') == 'video'
3100 original_infodict = info_dict
3101
3102 if 'format' not in info_dict and 'ext' in info_dict:
3103 info_dict['format'] = info_dict['ext']
3104
3105 if self._match_entry(info_dict) is not None:
3106 info_dict['__write_download_archive'] = 'ignore'
3107 return
3108
3109 # Does nothing under normal operation - for backward compatibility of process_info
3110 self.post_extract(info_dict)
3111
3112 def replace_info_dict(new_info):
3113 nonlocal info_dict
3114 if new_info == info_dict:
3115 return
3116 info_dict.clear()
3117 info_dict.update(new_info)
3118
3119 new_info, _ = self.pre_process(info_dict, 'video')
3120 replace_info_dict(new_info)
3121 self._num_downloads += 1
3122
3123 # info_dict['_filename'] needs to be set for backward compatibility
3124 info_dict['_filename'] = full_filename = self.prepare_filename(info_dict, warn=True)
3125 temp_filename = self.prepare_filename(info_dict, 'temp')
3126 files_to_move = {}
3127
3128 # Forced printings
3129 self.__forced_printings(info_dict, full_filename, incomplete=('format' not in info_dict))
3130
3131 def check_max_downloads():
3132 if self._num_downloads >= float(self.params.get('max_downloads') or 'inf'):
3133 raise MaxDownloadsReached()
3134
3135 if self.params.get('simulate'):
3136 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
3137 check_max_downloads()
3138 return
3139
3140 if full_filename is None:
3141 return
3142 if not self._ensure_dir_exists(encodeFilename(full_filename)):
3143 return
3144 if not self._ensure_dir_exists(encodeFilename(temp_filename)):
3145 return
3146
3147 if self._write_description('video', info_dict,
3148 self.prepare_filename(info_dict, 'description')) is None:
3149 return
3150
3151 sub_files = self._write_subtitles(info_dict, temp_filename)
3152 if sub_files is None:
3153 return
3154 files_to_move.update(dict(sub_files))
3155
3156 thumb_files = self._write_thumbnails(
3157 'video', info_dict, temp_filename, self.prepare_filename(info_dict, 'thumbnail'))
3158 if thumb_files is None:
3159 return
3160 files_to_move.update(dict(thumb_files))
3161
3162 infofn = self.prepare_filename(info_dict, 'infojson')
3163 _infojson_written = self._write_info_json('video', info_dict, infofn)
3164 if _infojson_written:
3165 info_dict['infojson_filename'] = infofn
3166 # For backward compatibility, even though it was a private field
3167 info_dict['__infojson_filename'] = infofn
3168 elif _infojson_written is None:
3169 return
3170
3171 # Note: Annotations are deprecated
3172 annofn = None
3173 if self.params.get('writeannotations', False):
3174 annofn = self.prepare_filename(info_dict, 'annotation')
3175 if annofn:
3176 if not self._ensure_dir_exists(encodeFilename(annofn)):
3177 return
3178 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(annofn)):
3179 self.to_screen('[info] Video annotations are already present')
3180 elif not info_dict.get('annotations'):
3181 self.report_warning('There are no annotations to write.')
3182 else:
3183 try:
3184 self.to_screen('[info] Writing video annotations to: ' + annofn)
3185 with open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
3186 annofile.write(info_dict['annotations'])
3187 except (KeyError, TypeError):
3188 self.report_warning('There are no annotations to write.')
3189 except OSError:
3190 self.report_error('Cannot write annotations file: ' + annofn)
3191 return
3192
3193 # Write internet shortcut files
3194 def _write_link_file(link_type):
3195 url = try_get(info_dict['webpage_url'], iri_to_uri)
3196 if not url:
3197 self.report_warning(
3198 f'Cannot write internet shortcut file because the actual URL of "{info_dict["webpage_url"]}" is unknown')
3199 return True
3200 linkfn = replace_extension(self.prepare_filename(info_dict, 'link'), link_type, info_dict.get('ext'))
3201 if not self._ensure_dir_exists(encodeFilename(linkfn)):
3202 return False
3203 if self.params.get('overwrites', True) and os.path.exists(encodeFilename(linkfn)):
3204 self.to_screen(f'[info] Internet shortcut (.{link_type}) is already present')
3205 return True
3206 try:
3207 self.to_screen(f'[info] Writing internet shortcut (.{link_type}) to: {linkfn}')
3208 with open(encodeFilename(to_high_limit_path(linkfn)), 'w', encoding='utf-8',
3209 newline='\r\n' if link_type == 'url' else '\n') as linkfile:
3210 template_vars = {'url': url}
3211 if link_type == 'desktop':
3212 template_vars['filename'] = linkfn[:-(len(link_type) + 1)]
3213 linkfile.write(LINK_TEMPLATES[link_type] % template_vars)
3214 except OSError:
3215 self.report_error(f'Cannot write internet shortcut {linkfn}')
3216 return False
3217 return True
3218
3219 write_links = {
3220 'url': self.params.get('writeurllink'),
3221 'webloc': self.params.get('writewebloclink'),
3222 'desktop': self.params.get('writedesktoplink'),
3223 }
3224 if self.params.get('writelink'):
3225 link_type = ('webloc' if sys.platform == 'darwin'
3226 else 'desktop' if sys.platform.startswith('linux')
3227 else 'url')
3228 write_links[link_type] = True
3229
3230 if any(should_write and not _write_link_file(link_type)
3231 for link_type, should_write in write_links.items()):
3232 return
3233
3234 new_info, files_to_move = self.pre_process(info_dict, 'before_dl', files_to_move)
3235 replace_info_dict(new_info)
3236
3237 if self.params.get('skip_download'):
3238 info_dict['filepath'] = temp_filename
3239 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
3240 info_dict['__files_to_move'] = files_to_move
3241 replace_info_dict(self.run_pp(MoveFilesAfterDownloadPP(self, False), info_dict))
3242 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
3243 else:
3244 # Download
3245 info_dict.setdefault('__postprocessors', [])
3246 try:
3247
3248 def existing_video_file(*filepaths):
3249 ext = info_dict.get('ext')
3250 converted = lambda file: replace_extension(file, self.params.get('final_ext') or ext, ext)
3251 file = self.existing_file(itertools.chain(*zip(map(converted, filepaths), filepaths)),
3252 default_overwrite=False)
3253 if file:
3254 info_dict['ext'] = os.path.splitext(file)[1][1:]
3255 return file
3256
3257 fd, success = None, True
3258 if info_dict.get('protocol') or info_dict.get('url'):
3259 fd = get_suitable_downloader(info_dict, self.params, to_stdout=temp_filename == '-')
3260 if fd is not FFmpegFD and 'no-direct-merge' not in self.params['compat_opts'] and (
3261 info_dict.get('section_start') or info_dict.get('section_end')):
3262 msg = ('This format cannot be partially downloaded' if FFmpegFD.available()
3263 else 'You have requested downloading the video partially, but ffmpeg is not installed')
3264 self.report_error(f'{msg}. Aborting')
3265 return
3266
3267 if info_dict.get('requested_formats') is not None:
3268 old_ext = info_dict['ext']
3269 if self.params.get('merge_output_format') is None:
3270 if (info_dict['ext'] == 'webm'
3271 and info_dict.get('thumbnails')
3272 # check with type instead of pp_key, __name__, or isinstance
3273 # since we dont want any custom PPs to trigger this
3274 and any(type(pp) == EmbedThumbnailPP for pp in self._pps['post_process'])): # noqa: E721
3275 info_dict['ext'] = 'mkv'
3276 self.report_warning(
3277 'webm doesn\'t support embedding a thumbnail, mkv will be used')
3278 new_ext = info_dict['ext']
3279
3280 def correct_ext(filename, ext=new_ext):
3281 if filename == '-':
3282 return filename
3283 filename_real_ext = os.path.splitext(filename)[1][1:]
3284 filename_wo_ext = (
3285 os.path.splitext(filename)[0]
3286 if filename_real_ext in (old_ext, new_ext)
3287 else filename)
3288 return f'{filename_wo_ext}.{ext}'
3289
3290 # Ensure filename always has a correct extension for successful merge
3291 full_filename = correct_ext(full_filename)
3292 temp_filename = correct_ext(temp_filename)
3293 dl_filename = existing_video_file(full_filename, temp_filename)
3294
3295 info_dict['__real_download'] = False
3296 # NOTE: Copy so that original format dicts are not modified
3297 info_dict['requested_formats'] = list(map(dict, info_dict['requested_formats']))
3298
3299 merger = FFmpegMergerPP(self)
3300 downloaded = []
3301 if dl_filename is not None:
3302 self.report_file_already_downloaded(dl_filename)
3303 elif fd:
3304 for f in info_dict['requested_formats'] if fd != FFmpegFD else []:
3305 f['filepath'] = fname = prepend_extension(
3306 correct_ext(temp_filename, info_dict['ext']),
3307 'f%s' % f['format_id'], info_dict['ext'])
3308 downloaded.append(fname)
3309 info_dict['url'] = '\n'.join(f['url'] for f in info_dict['requested_formats'])
3310 success, real_download = self.dl(temp_filename, info_dict)
3311 info_dict['__real_download'] = real_download
3312 else:
3313 if self.params.get('allow_unplayable_formats'):
3314 self.report_warning(
3315 'You have requested merging of multiple formats '
3316 'while also allowing unplayable formats to be downloaded. '
3317 'The formats won\'t be merged to prevent data corruption.')
3318 elif not merger.available:
3319 msg = 'You have requested merging of multiple formats but ffmpeg is not installed'
3320 if not self.params.get('ignoreerrors'):
3321 self.report_error(f'{msg}. Aborting due to --abort-on-error')
3322 return
3323 self.report_warning(f'{msg}. The formats won\'t be merged')
3324
3325 if temp_filename == '-':
3326 reason = ('using a downloader other than ffmpeg' if FFmpegFD.can_merge_formats(info_dict, self.params)
3327 else 'but the formats are incompatible for simultaneous download' if merger.available
3328 else 'but ffmpeg is not installed')
3329 self.report_warning(
3330 f'You have requested downloading multiple formats to stdout {reason}. '
3331 'The formats will be streamed one after the other')
3332 fname = temp_filename
3333 for f in info_dict['requested_formats']:
3334 new_info = dict(info_dict)
3335 del new_info['requested_formats']
3336 new_info.update(f)
3337 if temp_filename != '-':
3338 fname = prepend_extension(
3339 correct_ext(temp_filename, new_info['ext']),
3340 'f%s' % f['format_id'], new_info['ext'])
3341 if not self._ensure_dir_exists(fname):
3342 return
3343 f['filepath'] = fname
3344 downloaded.append(fname)
3345 partial_success, real_download = self.dl(fname, new_info)
3346 info_dict['__real_download'] = info_dict['__real_download'] or real_download
3347 success = success and partial_success
3348
3349 if downloaded and merger.available and not self.params.get('allow_unplayable_formats'):
3350 info_dict['__postprocessors'].append(merger)
3351 info_dict['__files_to_merge'] = downloaded
3352 # Even if there were no downloads, it is being merged only now
3353 info_dict['__real_download'] = True
3354 else:
3355 for file in downloaded:
3356 files_to_move[file] = None
3357 else:
3358 # Just a single file
3359 dl_filename = existing_video_file(full_filename, temp_filename)
3360 if dl_filename is None or dl_filename == temp_filename:
3361 # dl_filename == temp_filename could mean that the file was partially downloaded with --no-part.
3362 # So we should try to resume the download
3363 success, real_download = self.dl(temp_filename, info_dict)
3364 info_dict['__real_download'] = real_download
3365 else:
3366 self.report_file_already_downloaded(dl_filename)
3367
3368 dl_filename = dl_filename or temp_filename
3369 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
3370
3371 except network_exceptions as err:
3372 self.report_error('unable to download video data: %s' % error_to_compat_str(err))
3373 return
3374 except OSError as err:
3375 raise UnavailableVideoError(err)
3376 except (ContentTooShortError, ) as err:
3377 self.report_error(f'content too short (expected {err.expected} bytes and served {err.downloaded})')
3378 return
3379
3380 self._raise_pending_errors(info_dict)
3381 if success and full_filename != '-':
3382
3383 def fixup():
3384 do_fixup = True
3385 fixup_policy = self.params.get('fixup')
3386 vid = info_dict['id']
3387
3388 if fixup_policy in ('ignore', 'never'):
3389 return
3390 elif fixup_policy == 'warn':
3391 do_fixup = 'warn'
3392 elif fixup_policy != 'force':
3393 assert fixup_policy in ('detect_or_warn', None)
3394 if not info_dict.get('__real_download'):
3395 do_fixup = False
3396
3397 def ffmpeg_fixup(cndn, msg, cls):
3398 if not (do_fixup and cndn):
3399 return
3400 elif do_fixup == 'warn':
3401 self.report_warning(f'{vid}: {msg}')
3402 return
3403 pp = cls(self)
3404 if pp.available:
3405 info_dict['__postprocessors'].append(pp)
3406 else:
3407 self.report_warning(f'{vid}: {msg}. Install ffmpeg to fix this automatically')
3408
3409 stretched_ratio = info_dict.get('stretched_ratio')
3410 ffmpeg_fixup(stretched_ratio not in (1, None),
3411 f'Non-uniform pixel ratio {stretched_ratio}',
3412 FFmpegFixupStretchedPP)
3413
3414 downloader = get_suitable_downloader(info_dict, self.params) if 'protocol' in info_dict else None
3415 downloader = downloader.FD_NAME if downloader else None
3416
3417 ext = info_dict.get('ext')
3418 postprocessed_by_ffmpeg = info_dict.get('requested_formats') or any((
3419 isinstance(pp, FFmpegVideoConvertorPP)
3420 and resolve_recode_mapping(ext, pp.mapping)[0] not in (ext, None)
3421 ) for pp in self._pps['post_process'])
3422
3423 if not postprocessed_by_ffmpeg:
3424 ffmpeg_fixup(ext == 'm4a' and info_dict.get('container') == 'm4a_dash',
3425 'writing DASH m4a. Only some players support this container',
3426 FFmpegFixupM4aPP)
3427 ffmpeg_fixup(downloader == 'hlsnative' and not self.params.get('hls_use_mpegts')
3428 or info_dict.get('is_live') and self.params.get('hls_use_mpegts') is None,
3429 'Possible MPEG-TS in MP4 container or malformed AAC timestamps',
3430 FFmpegFixupM3u8PP)
3431 ffmpeg_fixup(info_dict.get('is_live') and downloader == 'dashsegments',
3432 'Possible duplicate MOOV atoms', FFmpegFixupDuplicateMoovPP)
3433
3434 ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed timestamps detected', FFmpegFixupTimestampPP)
3435 ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed duration detected', FFmpegFixupDurationPP)
3436
3437 fixup()
3438 try:
3439 replace_info_dict(self.post_process(dl_filename, info_dict, files_to_move))
3440 except PostProcessingError as err:
3441 self.report_error('Postprocessing: %s' % str(err))
3442 return
3443 try:
3444 for ph in self._post_hooks:
3445 ph(info_dict['filepath'])
3446 except Exception as err:
3447 self.report_error('post hooks: %s' % str(err))
3448 return
3449 info_dict['__write_download_archive'] = True
3450
3451 assert info_dict is original_infodict # Make sure the info_dict was modified in-place
3452 if self.params.get('force_write_download_archive'):
3453 info_dict['__write_download_archive'] = True
3454 check_max_downloads()
3455
3456 def __download_wrapper(self, func):
3457 @functools.wraps(func)
3458 def wrapper(*args, **kwargs):
3459 try:
3460 res = func(*args, **kwargs)
3461 except UnavailableVideoError as e:
3462 self.report_error(e)
3463 except DownloadCancelled as e:
3464 self.to_screen(f'[info] {e}')
3465 if not self.params.get('break_per_url'):
3466 raise
3467 self._num_downloads = 0
3468 else:
3469 if self.params.get('dump_single_json', False):
3470 self.post_extract(res)
3471 self.to_stdout(json.dumps(self.sanitize_info(res)))
3472 return wrapper
3473
3474 def download(self, url_list):
3475 """Download a given list of URLs."""
3476 url_list = variadic(url_list) # Passing a single URL is a common mistake
3477 outtmpl = self.params['outtmpl']['default']
3478 if (len(url_list) > 1
3479 and outtmpl != '-'
3480 and '%' not in outtmpl
3481 and self.params.get('max_downloads') != 1):
3482 raise SameFileError(outtmpl)
3483
3484 for url in url_list:
3485 self.__download_wrapper(self.extract_info)(
3486 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
3487
3488 return self._download_retcode
3489
3490 def download_with_info_file(self, info_filename):
3491 with contextlib.closing(fileinput.FileInput(
3492 [info_filename], mode='r',
3493 openhook=fileinput.hook_encoded('utf-8'))) as f:
3494 # FileInput doesn't have a read method, we can't call json.load
3495 infos = [self.sanitize_info(info, self.params.get('clean_infojson', True))
3496 for info in variadic(json.loads('\n'.join(f)))]
3497 for info in infos:
3498 self._load_cookies(info.get('cookies'), from_headers=False)
3499 self._load_cookies(traverse_obj(info.get('http_headers'), 'Cookie', casesense=False)) # compat
3500 try:
3501 self.__download_wrapper(self.process_ie_result)(info, download=True)
3502 except (DownloadError, EntryNotInPlaylist, ReExtractInfo) as e:
3503 if not isinstance(e, EntryNotInPlaylist):
3504 self.to_stderr('\r')
3505 webpage_url = info.get('webpage_url')
3506 if webpage_url is None:
3507 raise
3508 self.report_warning(f'The info failed to download: {e}; trying with URL {webpage_url}')
3509 self.download([webpage_url])
3510 return self._download_retcode
3511
3512 @staticmethod
3513 def sanitize_info(info_dict, remove_private_keys=False):
3514 ''' Sanitize the infodict for converting to json '''
3515 if info_dict is None:
3516 return info_dict
3517 info_dict.setdefault('epoch', int(time.time()))
3518 info_dict.setdefault('_type', 'video')
3519 info_dict.setdefault('_version', {
3520 'version': __version__,
3521 'current_git_head': current_git_head(),
3522 'release_git_head': RELEASE_GIT_HEAD,
3523 'repository': REPOSITORY,
3524 })
3525
3526 if remove_private_keys:
3527 reject = lambda k, v: v is None or k.startswith('__') or k in {
3528 'requested_downloads', 'requested_formats', 'requested_subtitles', 'requested_entries',
3529 'entries', 'filepath', '_filename', 'filename', 'infojson_filename', 'original_url',
3530 'playlist_autonumber', '_format_sort_fields',
3531 }
3532 else:
3533 reject = lambda k, v: False
3534
3535 def filter_fn(obj):
3536 if isinstance(obj, dict):
3537 return {k: filter_fn(v) for k, v in obj.items() if not reject(k, v)}
3538 elif isinstance(obj, (list, tuple, set, LazyList)):
3539 return list(map(filter_fn, obj))
3540 elif obj is None or isinstance(obj, (str, int, float, bool)):
3541 return obj
3542 else:
3543 return repr(obj)
3544
3545 return filter_fn(info_dict)
3546
3547 @staticmethod
3548 def filter_requested_info(info_dict, actually_filter=True):
3549 ''' Alias of sanitize_info for backward compatibility '''
3550 return YoutubeDL.sanitize_info(info_dict, actually_filter)
3551
3552 def _delete_downloaded_files(self, *files_to_delete, info={}, msg=None):
3553 for filename in set(filter(None, files_to_delete)):
3554 if msg:
3555 self.to_screen(msg % filename)
3556 try:
3557 os.remove(filename)
3558 except OSError:
3559 self.report_warning(f'Unable to delete file {filename}')
3560 if filename in info.get('__files_to_move', []): # NB: Delete even if None
3561 del info['__files_to_move'][filename]
3562
3563 @staticmethod
3564 def post_extract(info_dict):
3565 def actual_post_extract(info_dict):
3566 if info_dict.get('_type') in ('playlist', 'multi_video'):
3567 for video_dict in info_dict.get('entries', {}):
3568 actual_post_extract(video_dict or {})
3569 return
3570
3571 post_extractor = info_dict.pop('__post_extractor', None) or (lambda: {})
3572 info_dict.update(post_extractor())
3573
3574 actual_post_extract(info_dict or {})
3575
3576 def run_pp(self, pp, infodict):
3577 files_to_delete = []
3578 if '__files_to_move' not in infodict:
3579 infodict['__files_to_move'] = {}
3580 try:
3581 files_to_delete, infodict = pp.run(infodict)
3582 except PostProcessingError as e:
3583 # Must be True and not 'only_download'
3584 if self.params.get('ignoreerrors') is True:
3585 self.report_error(e)
3586 return infodict
3587 raise
3588
3589 if not files_to_delete:
3590 return infodict
3591 if self.params.get('keepvideo', False):
3592 for f in files_to_delete:
3593 infodict['__files_to_move'].setdefault(f, '')
3594 else:
3595 self._delete_downloaded_files(
3596 *files_to_delete, info=infodict, msg='Deleting original file %s (pass -k to keep)')
3597 return infodict
3598
3599 def run_all_pps(self, key, info, *, additional_pps=None):
3600 if key != 'video':
3601 self._forceprint(key, info)
3602 for pp in (additional_pps or []) + self._pps[key]:
3603 info = self.run_pp(pp, info)
3604 return info
3605
3606 def pre_process(self, ie_info, key='pre_process', files_to_move=None):
3607 info = dict(ie_info)
3608 info['__files_to_move'] = files_to_move or {}
3609 try:
3610 info = self.run_all_pps(key, info)
3611 except PostProcessingError as err:
3612 msg = f'Preprocessing: {err}'
3613 info.setdefault('__pending_error', msg)
3614 self.report_error(msg, is_error=False)
3615 return info, info.pop('__files_to_move', None)
3616
3617 def post_process(self, filename, info, files_to_move=None):
3618 """Run all the postprocessors on the given file."""
3619 info['filepath'] = filename
3620 info['__files_to_move'] = files_to_move or {}
3621 info = self.run_all_pps('post_process', info, additional_pps=info.get('__postprocessors'))
3622 info = self.run_pp(MoveFilesAfterDownloadPP(self), info)
3623 del info['__files_to_move']
3624 return self.run_all_pps('after_move', info)
3625
3626 def _make_archive_id(self, info_dict):
3627 video_id = info_dict.get('id')
3628 if not video_id:
3629 return
3630 # Future-proof against any change in case
3631 # and backwards compatibility with prior versions
3632 extractor = info_dict.get('extractor_key') or info_dict.get('ie_key') # key in a playlist
3633 if extractor is None:
3634 url = str_or_none(info_dict.get('url'))
3635 if not url:
3636 return
3637 # Try to find matching extractor for the URL and take its ie_key
3638 for ie_key, ie in self._ies.items():
3639 if ie.suitable(url):
3640 extractor = ie_key
3641 break
3642 else:
3643 return
3644 return make_archive_id(extractor, video_id)
3645
3646 def in_download_archive(self, info_dict):
3647 if not self.archive:
3648 return False
3649
3650 vid_ids = [self._make_archive_id(info_dict)]
3651 vid_ids.extend(info_dict.get('_old_archive_ids') or [])
3652 return any(id_ in self.archive for id_ in vid_ids)
3653
3654 def record_download_archive(self, info_dict):
3655 fn = self.params.get('download_archive')
3656 if fn is None:
3657 return
3658 vid_id = self._make_archive_id(info_dict)
3659 assert vid_id
3660
3661 self.write_debug(f'Adding to archive: {vid_id}')
3662 if is_path_like(fn):
3663 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
3664 archive_file.write(vid_id + '\n')
3665 self.archive.add(vid_id)
3666
3667 @staticmethod
3668 def format_resolution(format, default='unknown'):
3669 if format.get('vcodec') == 'none' and format.get('acodec') != 'none':
3670 return 'audio only'
3671 if format.get('resolution') is not None:
3672 return format['resolution']
3673 if format.get('width') and format.get('height'):
3674 return '%dx%d' % (format['width'], format['height'])
3675 elif format.get('height'):
3676 return '%sp' % format['height']
3677 elif format.get('width'):
3678 return '%dx?' % format['width']
3679 return default
3680
3681 def _list_format_headers(self, *headers):
3682 if self.params.get('listformats_table', True) is not False:
3683 return [self._format_out(header, self.Styles.HEADERS) for header in headers]
3684 return headers
3685
3686 def _format_note(self, fdict):
3687 res = ''
3688 if fdict.get('ext') in ['f4f', 'f4m']:
3689 res += '(unsupported)'
3690 if fdict.get('language'):
3691 if res:
3692 res += ' '
3693 res += '[%s]' % fdict['language']
3694 if fdict.get('format_note') is not None:
3695 if res:
3696 res += ' '
3697 res += fdict['format_note']
3698 if fdict.get('tbr') is not None:
3699 if res:
3700 res += ', '
3701 res += '%4dk' % fdict['tbr']
3702 if fdict.get('container') is not None:
3703 if res:
3704 res += ', '
3705 res += '%s container' % fdict['container']
3706 if (fdict.get('vcodec') is not None
3707 and fdict.get('vcodec') != 'none'):
3708 if res:
3709 res += ', '
3710 res += fdict['vcodec']
3711 if fdict.get('vbr') is not None:
3712 res += '@'
3713 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
3714 res += 'video@'
3715 if fdict.get('vbr') is not None:
3716 res += '%4dk' % fdict['vbr']
3717 if fdict.get('fps') is not None:
3718 if res:
3719 res += ', '
3720 res += '%sfps' % fdict['fps']
3721 if fdict.get('acodec') is not None:
3722 if res:
3723 res += ', '
3724 if fdict['acodec'] == 'none':
3725 res += 'video only'
3726 else:
3727 res += '%-5s' % fdict['acodec']
3728 elif fdict.get('abr') is not None:
3729 if res:
3730 res += ', '
3731 res += 'audio'
3732 if fdict.get('abr') is not None:
3733 res += '@%3dk' % fdict['abr']
3734 if fdict.get('asr') is not None:
3735 res += ' (%5dHz)' % fdict['asr']
3736 if fdict.get('filesize') is not None:
3737 if res:
3738 res += ', '
3739 res += format_bytes(fdict['filesize'])
3740 elif fdict.get('filesize_approx') is not None:
3741 if res:
3742 res += ', '
3743 res += '~' + format_bytes(fdict['filesize_approx'])
3744 return res
3745
3746 def _get_formats(self, info_dict):
3747 if info_dict.get('formats') is None:
3748 if info_dict.get('url') and info_dict.get('_type', 'video') == 'video':
3749 return [info_dict]
3750 return []
3751 return info_dict['formats']
3752
3753 def render_formats_table(self, info_dict):
3754 formats = self._get_formats(info_dict)
3755 if not formats:
3756 return
3757 if not self.params.get('listformats_table', True) is not False:
3758 table = [
3759 [
3760 format_field(f, 'format_id'),
3761 format_field(f, 'ext'),
3762 self.format_resolution(f),
3763 self._format_note(f)
3764 ] for f in formats if (f.get('preference') or 0) >= -1000]
3765 return render_table(['format code', 'extension', 'resolution', 'note'], table, extra_gap=1)
3766
3767 def simplified_codec(f, field):
3768 assert field in ('acodec', 'vcodec')
3769 codec = f.get(field)
3770 if not codec:
3771 return 'unknown'
3772 elif codec != 'none':
3773 return '.'.join(codec.split('.')[:4])
3774
3775 if field == 'vcodec' and f.get('acodec') == 'none':
3776 return 'images'
3777 elif field == 'acodec' and f.get('vcodec') == 'none':
3778 return ''
3779 return self._format_out('audio only' if field == 'vcodec' else 'video only',
3780 self.Styles.SUPPRESS)
3781
3782 delim = self._format_out('\u2502', self.Styles.DELIM, '|', test_encoding=True)
3783 table = [
3784 [
3785 self._format_out(format_field(f, 'format_id'), self.Styles.ID),
3786 format_field(f, 'ext'),
3787 format_field(f, func=self.format_resolution, ignore=('audio only', 'images')),
3788 format_field(f, 'fps', '\t%d', func=round),
3789 format_field(f, 'dynamic_range', '%s', ignore=(None, 'SDR')).replace('HDR', ''),
3790 format_field(f, 'audio_channels', '\t%s'),
3791 delim, (
3792 format_field(f, 'filesize', ' \t%s', func=format_bytes)
3793 or format_field(f, 'filesize_approx', '≈\t%s', func=format_bytes)
3794 or format_field(try_call(lambda: format_bytes(int(info_dict['duration'] * f['tbr'] * (1024 / 8)))),
3795 None, self._format_out('~\t%s', self.Styles.SUPPRESS))),
3796 format_field(f, 'tbr', '\t%dk', func=round),
3797 shorten_protocol_name(f.get('protocol', '')),
3798 delim,
3799 simplified_codec(f, 'vcodec'),
3800 format_field(f, 'vbr', '\t%dk', func=round),
3801 simplified_codec(f, 'acodec'),
3802 format_field(f, 'abr', '\t%dk', func=round),
3803 format_field(f, 'asr', '\t%s', func=format_decimal_suffix),
3804 join_nonempty(format_field(f, 'language', '[%s]'), join_nonempty(
3805 self._format_out('UNSUPPORTED', self.Styles.BAD_FORMAT) if f.get('ext') in ('f4f', 'f4m') else None,
3806 (self._format_out('Maybe DRM', self.Styles.WARNING) if f.get('has_drm') == 'maybe'
3807 else self._format_out('DRM', self.Styles.BAD_FORMAT) if f.get('has_drm') else None),
3808 format_field(f, 'format_note'),
3809 format_field(f, 'container', ignore=(None, f.get('ext'))),
3810 delim=', '), delim=' '),
3811 ] for f in formats if f.get('preference') is None or f['preference'] >= -1000]
3812 header_line = self._list_format_headers(
3813 'ID', 'EXT', 'RESOLUTION', '\tFPS', 'HDR', 'CH', delim, '\tFILESIZE', '\tTBR', 'PROTO',
3814 delim, 'VCODEC', '\tVBR', 'ACODEC', '\tABR', '\tASR', 'MORE INFO')
3815
3816 return render_table(
3817 header_line, table, hide_empty=True,
3818 delim=self._format_out('\u2500', self.Styles.DELIM, '-', test_encoding=True))
3819
3820 def render_thumbnails_table(self, info_dict):
3821 thumbnails = list(info_dict.get('thumbnails') or [])
3822 if not thumbnails:
3823 return None
3824 return render_table(
3825 self._list_format_headers('ID', 'Width', 'Height', 'URL'),
3826 [[t.get('id'), t.get('width') or 'unknown', t.get('height') or 'unknown', t['url']] for t in thumbnails])
3827
3828 def render_subtitles_table(self, video_id, subtitles):
3829 def _row(lang, formats):
3830 exts, names = zip(*((f['ext'], f.get('name') or 'unknown') for f in reversed(formats)))
3831 if len(set(names)) == 1:
3832 names = [] if names[0] == 'unknown' else names[:1]
3833 return [lang, ', '.join(names), ', '.join(exts)]
3834
3835 if not subtitles:
3836 return None
3837 return render_table(
3838 self._list_format_headers('Language', 'Name', 'Formats'),
3839 [_row(lang, formats) for lang, formats in subtitles.items()],
3840 hide_empty=True)
3841
3842 def __list_table(self, video_id, name, func, *args):
3843 table = func(*args)
3844 if not table:
3845 self.to_screen(f'{video_id} has no {name}')
3846 return
3847 self.to_screen(f'[info] Available {name} for {video_id}:')
3848 self.to_stdout(table)
3849
3850 def list_formats(self, info_dict):
3851 self.__list_table(info_dict['id'], 'formats', self.render_formats_table, info_dict)
3852
3853 def list_thumbnails(self, info_dict):
3854 self.__list_table(info_dict['id'], 'thumbnails', self.render_thumbnails_table, info_dict)
3855
3856 def list_subtitles(self, video_id, subtitles, name='subtitles'):
3857 self.__list_table(video_id, name, self.render_subtitles_table, video_id, subtitles)
3858
3859 def urlopen(self, req):
3860 """ Start an HTTP download """
3861 if isinstance(req, str):
3862 req = sanitized_Request(req)
3863 return self._opener.open(req, timeout=self._socket_timeout)
3864
3865 def print_debug_header(self):
3866 if not self.params.get('verbose'):
3867 return
3868
3869 from . import _IN_CLI # Must be delayed import
3870
3871 # These imports can be slow. So import them only as needed
3872 from .extractor.extractors import _LAZY_LOADER
3873 from .extractor.extractors import (
3874 _PLUGIN_CLASSES as plugin_ies,
3875 _PLUGIN_OVERRIDES as plugin_ie_overrides
3876 )
3877
3878 def get_encoding(stream):
3879 ret = str(getattr(stream, 'encoding', 'missing (%s)' % type(stream).__name__))
3880 additional_info = []
3881 if os.environ.get('TERM', '').lower() == 'dumb':
3882 additional_info.append('dumb')
3883 if not supports_terminal_sequences(stream):
3884 from .utils import WINDOWS_VT_MODE # Must be imported locally
3885 additional_info.append('No VT' if WINDOWS_VT_MODE is False else 'No ANSI')
3886 if additional_info:
3887 ret = f'{ret} ({",".join(additional_info)})'
3888 return ret
3889
3890 encoding_str = 'Encodings: locale %s, fs %s, pref %s, %s' % (
3891 locale.getpreferredencoding(),
3892 sys.getfilesystemencoding(),
3893 self.get_encoding(),
3894 ', '.join(
3895 f'{key} {get_encoding(stream)}' for key, stream in self._out_files.items_
3896 if stream is not None and key != 'console')
3897 )
3898
3899 logger = self.params.get('logger')
3900 if logger:
3901 write_debug = lambda msg: logger.debug(f'[debug] {msg}')
3902 write_debug(encoding_str)
3903 else:
3904 write_string(f'[debug] {encoding_str}\n', encoding=None)
3905 write_debug = lambda msg: self._write_string(f'[debug] {msg}\n')
3906
3907 source = detect_variant()
3908 if VARIANT not in (None, 'pip'):
3909 source += '*'
3910 klass = type(self)
3911 write_debug(join_nonempty(
3912 f'{"yt-dlp" if REPOSITORY == "yt-dlp/yt-dlp" else REPOSITORY} version',
3913 f'{CHANNEL}@{__version__}',
3914 f'[{RELEASE_GIT_HEAD[:9]}]' if RELEASE_GIT_HEAD else '',
3915 '' if source == 'unknown' else f'({source})',
3916 '' if _IN_CLI else 'API' if klass == YoutubeDL else f'API:{self.__module__}.{klass.__qualname__}',
3917 delim=' '))
3918
3919 if not _IN_CLI:
3920 write_debug(f'params: {self.params}')
3921
3922 if not _LAZY_LOADER:
3923 if os.environ.get('YTDLP_NO_LAZY_EXTRACTORS'):
3924 write_debug('Lazy loading extractors is forcibly disabled')
3925 else:
3926 write_debug('Lazy loading extractors is disabled')
3927 if self.params['compat_opts']:
3928 write_debug('Compatibility options: %s' % ', '.join(self.params['compat_opts']))
3929
3930 if current_git_head():
3931 write_debug(f'Git HEAD: {current_git_head()}')
3932 write_debug(system_identifier())
3933
3934 exe_versions, ffmpeg_features = FFmpegPostProcessor.get_versions_and_features(self)
3935 ffmpeg_features = {key for key, val in ffmpeg_features.items() if val}
3936 if ffmpeg_features:
3937 exe_versions['ffmpeg'] += ' (%s)' % ','.join(sorted(ffmpeg_features))
3938
3939 exe_versions['rtmpdump'] = rtmpdump_version()
3940 exe_versions['phantomjs'] = PhantomJSwrapper._version()
3941 exe_str = ', '.join(
3942 f'{exe} {v}' for exe, v in sorted(exe_versions.items()) if v
3943 ) or 'none'
3944 write_debug('exe versions: %s' % exe_str)
3945
3946 from .compat.compat_utils import get_package_info
3947 from .dependencies import available_dependencies
3948
3949 write_debug('Optional libraries: %s' % (', '.join(sorted({
3950 join_nonempty(*get_package_info(m)) for m in available_dependencies.values()
3951 })) or 'none'))
3952
3953 self._setup_opener()
3954 proxy_map = {}
3955 for handler in self._opener.handlers:
3956 if hasattr(handler, 'proxies'):
3957 proxy_map.update(handler.proxies)
3958 write_debug(f'Proxy map: {proxy_map}')
3959
3960 for plugin_type, plugins in {'Extractor': plugin_ies, 'Post-Processor': plugin_pps}.items():
3961 display_list = ['%s%s' % (
3962 klass.__name__, '' if klass.__name__ == name else f' as {name}')
3963 for name, klass in plugins.items()]
3964 if plugin_type == 'Extractor':
3965 display_list.extend(f'{plugins[-1].IE_NAME.partition("+")[2]} ({parent.__name__})'
3966 for parent, plugins in plugin_ie_overrides.items())
3967 if not display_list:
3968 continue
3969 write_debug(f'{plugin_type} Plugins: {", ".join(sorted(display_list))}')
3970
3971 plugin_dirs = plugin_directories()
3972 if plugin_dirs:
3973 write_debug(f'Plugin directories: {plugin_dirs}')
3974
3975 # Not implemented
3976 if False and self.params.get('call_home'):
3977 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode()
3978 write_debug('Public IP address: %s' % ipaddr)
3979 latest_version = self.urlopen(
3980 'https://yt-dl.org/latest/version').read().decode()
3981 if version_tuple(latest_version) > version_tuple(__version__):
3982 self.report_warning(
3983 'You are using an outdated version (newest version: %s)! '
3984 'See https://yt-dl.org/update if you need help updating.' %
3985 latest_version)
3986
3987 def _setup_opener(self):
3988 if hasattr(self, '_opener'):
3989 return
3990 timeout_val = self.params.get('socket_timeout')
3991 self._socket_timeout = 20 if timeout_val is None else float(timeout_val)
3992
3993 opts_cookiesfrombrowser = self.params.get('cookiesfrombrowser')
3994 opts_cookiefile = self.params.get('cookiefile')
3995 opts_proxy = self.params.get('proxy')
3996
3997 self.cookiejar = load_cookies(opts_cookiefile, opts_cookiesfrombrowser, self)
3998
3999 cookie_processor = YoutubeDLCookieProcessor(self.cookiejar)
4000 if opts_proxy is not None:
4001 if opts_proxy == '':
4002 proxies = {}
4003 else:
4004 proxies = {'http': opts_proxy, 'https': opts_proxy}
4005 else:
4006 proxies = urllib.request.getproxies()
4007 # Set HTTPS proxy to HTTP one if given (https://github.com/ytdl-org/youtube-dl/issues/805)
4008 if 'http' in proxies and 'https' not in proxies:
4009 proxies['https'] = proxies['http']
4010 proxy_handler = PerRequestProxyHandler(proxies)
4011
4012 debuglevel = 1 if self.params.get('debug_printtraffic') else 0
4013 https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
4014 ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
4015 redirect_handler = YoutubeDLRedirectHandler()
4016 data_handler = urllib.request.DataHandler()
4017
4018 # When passing our own FileHandler instance, build_opener won't add the
4019 # default FileHandler and allows us to disable the file protocol, which
4020 # can be used for malicious purposes (see
4021 # https://github.com/ytdl-org/youtube-dl/issues/8227)
4022 file_handler = urllib.request.FileHandler()
4023
4024 if not self.params.get('enable_file_urls'):
4025 def file_open(*args, **kwargs):
4026 raise urllib.error.URLError(
4027 'file:// URLs are explicitly disabled in yt-dlp for security reasons. '
4028 'Use --enable-file-urls to enable at your own risk.')
4029 file_handler.file_open = file_open
4030
4031 opener = urllib.request.build_opener(
4032 proxy_handler, https_handler, cookie_processor, ydlh, redirect_handler, data_handler, file_handler)
4033
4034 # Delete the default user-agent header, which would otherwise apply in
4035 # cases where our custom HTTP handler doesn't come into play
4036 # (See https://github.com/ytdl-org/youtube-dl/issues/1309 for details)
4037 opener.addheaders = []
4038 self._opener = opener
4039
4040 def encode(self, s):
4041 if isinstance(s, bytes):
4042 return s # Already encoded
4043
4044 try:
4045 return s.encode(self.get_encoding())
4046 except UnicodeEncodeError as err:
4047 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
4048 raise
4049
4050 def get_encoding(self):
4051 encoding = self.params.get('encoding')
4052 if encoding is None:
4053 encoding = preferredencoding()
4054 return encoding
4055
4056 def _write_info_json(self, label, ie_result, infofn, overwrite=None):
4057 ''' Write infojson and returns True = written, 'exists' = Already exists, False = skip, None = error '''
4058 if overwrite is None:
4059 overwrite = self.params.get('overwrites', True)
4060 if not self.params.get('writeinfojson'):
4061 return False
4062 elif not infofn:
4063 self.write_debug(f'Skipping writing {label} infojson')
4064 return False
4065 elif not self._ensure_dir_exists(infofn):
4066 return None
4067 elif not overwrite and os.path.exists(infofn):
4068 self.to_screen(f'[info] {label.title()} metadata is already present')
4069 return 'exists'
4070
4071 self.to_screen(f'[info] Writing {label} metadata as JSON to: {infofn}')
4072 try:
4073 write_json_file(self.sanitize_info(ie_result, self.params.get('clean_infojson', True)), infofn)
4074 return True
4075 except OSError:
4076 self.report_error(f'Cannot write {label} metadata to JSON file {infofn}')
4077 return None
4078
4079 def _write_description(self, label, ie_result, descfn):
4080 ''' Write description and returns True = written, False = skip, None = error '''
4081 if not self.params.get('writedescription'):
4082 return False
4083 elif not descfn:
4084 self.write_debug(f'Skipping writing {label} description')
4085 return False
4086 elif not self._ensure_dir_exists(descfn):
4087 return None
4088 elif not self.params.get('overwrites', True) and os.path.exists(descfn):
4089 self.to_screen(f'[info] {label.title()} description is already present')
4090 elif ie_result.get('description') is None:
4091 self.to_screen(f'[info] There\'s no {label} description to write')
4092 return False
4093 else:
4094 try:
4095 self.to_screen(f'[info] Writing {label} description to: {descfn}')
4096 with open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
4097 descfile.write(ie_result['description'])
4098 except OSError:
4099 self.report_error(f'Cannot write {label} description file {descfn}')
4100 return None
4101 return True
4102
4103 def _write_subtitles(self, info_dict, filename):
4104 ''' Write subtitles to file and return list of (sub_filename, final_sub_filename); or None if error'''
4105 ret = []
4106 subtitles = info_dict.get('requested_subtitles')
4107 if not (self.params.get('writesubtitles') or self.params.get('writeautomaticsub')):
4108 # subtitles download errors are already managed as troubles in relevant IE
4109 # that way it will silently go on when used with unsupporting IE
4110 return ret
4111 elif not subtitles:
4112 self.to_screen('[info] There are no subtitles for the requested languages')
4113 return ret
4114 sub_filename_base = self.prepare_filename(info_dict, 'subtitle')
4115 if not sub_filename_base:
4116 self.to_screen('[info] Skipping writing video subtitles')
4117 return ret
4118
4119 for sub_lang, sub_info in subtitles.items():
4120 sub_format = sub_info['ext']
4121 sub_filename = subtitles_filename(filename, sub_lang, sub_format, info_dict.get('ext'))
4122 sub_filename_final = subtitles_filename(sub_filename_base, sub_lang, sub_format, info_dict.get('ext'))
4123 existing_sub = self.existing_file((sub_filename_final, sub_filename))
4124 if existing_sub:
4125 self.to_screen(f'[info] Video subtitle {sub_lang}.{sub_format} is already present')
4126 sub_info['filepath'] = existing_sub
4127 ret.append((existing_sub, sub_filename_final))
4128 continue
4129
4130 self.to_screen(f'[info] Writing video subtitles to: {sub_filename}')
4131 if sub_info.get('data') is not None:
4132 try:
4133 # Use newline='' to prevent conversion of newline characters
4134 # See https://github.com/ytdl-org/youtube-dl/issues/10268
4135 with open(sub_filename, 'w', encoding='utf-8', newline='') as subfile:
4136 subfile.write(sub_info['data'])
4137 sub_info['filepath'] = sub_filename
4138 ret.append((sub_filename, sub_filename_final))
4139 continue
4140 except OSError:
4141 self.report_error(f'Cannot write video subtitles file {sub_filename}')
4142 return None
4143
4144 try:
4145 sub_copy = sub_info.copy()
4146 sub_copy.setdefault('http_headers', info_dict.get('http_headers'))
4147 self.dl(sub_filename, sub_copy, subtitle=True)
4148 sub_info['filepath'] = sub_filename
4149 ret.append((sub_filename, sub_filename_final))
4150 except (DownloadError, ExtractorError, IOError, OSError, ValueError) + network_exceptions as err:
4151 msg = f'Unable to download video subtitles for {sub_lang!r}: {err}'
4152 if self.params.get('ignoreerrors') is not True: # False or 'only_download'
4153 if not self.params.get('ignoreerrors'):
4154 self.report_error(msg)
4155 raise DownloadError(msg)
4156 self.report_warning(msg)
4157 return ret
4158
4159 def _write_thumbnails(self, label, info_dict, filename, thumb_filename_base=None):
4160 ''' Write thumbnails to file and return list of (thumb_filename, final_thumb_filename) '''
4161 write_all = self.params.get('write_all_thumbnails', False)
4162 thumbnails, ret = [], []
4163 if write_all or self.params.get('writethumbnail', False):
4164 thumbnails = info_dict.get('thumbnails') or []
4165 if not thumbnails:
4166 self.to_screen(f'[info] There are no {label} thumbnails to download')
4167 return ret
4168 multiple = write_all and len(thumbnails) > 1
4169
4170 if thumb_filename_base is None:
4171 thumb_filename_base = filename
4172 if thumbnails and not thumb_filename_base:
4173 self.write_debug(f'Skipping writing {label} thumbnail')
4174 return ret
4175
4176 for idx, t in list(enumerate(thumbnails))[::-1]:
4177 thumb_ext = (f'{t["id"]}.' if multiple else '') + determine_ext(t['url'], 'jpg')
4178 thumb_display_id = f'{label} thumbnail {t["id"]}'
4179 thumb_filename = replace_extension(filename, thumb_ext, info_dict.get('ext'))
4180 thumb_filename_final = replace_extension(thumb_filename_base, thumb_ext, info_dict.get('ext'))
4181
4182 existing_thumb = self.existing_file((thumb_filename_final, thumb_filename))
4183 if existing_thumb:
4184 self.to_screen('[info] %s is already present' % (
4185 thumb_display_id if multiple else f'{label} thumbnail').capitalize())
4186 t['filepath'] = existing_thumb
4187 ret.append((existing_thumb, thumb_filename_final))
4188 else:
4189 self.to_screen(f'[info] Downloading {thumb_display_id} ...')
4190 try:
4191 uf = self.urlopen(sanitized_Request(t['url'], headers=t.get('http_headers', {})))
4192 self.to_screen(f'[info] Writing {thumb_display_id} to: {thumb_filename}')
4193 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
4194 shutil.copyfileobj(uf, thumbf)
4195 ret.append((thumb_filename, thumb_filename_final))
4196 t['filepath'] = thumb_filename
4197 except network_exceptions as err:
4198 if isinstance(err, urllib.error.HTTPError) and err.code == 404:
4199 self.to_screen(f'[info] {thumb_display_id.title()} does not exist')
4200 else:
4201 self.report_warning(f'Unable to download {thumb_display_id}: {err}')
4202 thumbnails.pop(idx)
4203 if ret and not write_all:
4204 break
4205 return ret