]> jfr.im git - yt-dlp.git/blob - yt_dlp/YoutubeDL.py
c34d97bba1b0a60221556124f419235633dc48ca
[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 http.cookiejar
8 import io
9 import itertools
10 import json
11 import locale
12 import operator
13 import os
14 import random
15 import re
16 import shutil
17 import string
18 import subprocess
19 import sys
20 import tempfile
21 import time
22 import tokenize
23 import traceback
24 import unicodedata
25
26 from .cache import Cache
27 from .compat import functools, urllib # isort: split
28 from .compat import compat_os_name, compat_shlex_quote, urllib_req_to_req
29 from .cookies import LenientSimpleCookie, load_cookies
30 from .downloader import FFmpegFD, get_suitable_downloader, shorten_protocol_name
31 from .downloader.rtmp import rtmpdump_version
32 from .extractor import gen_extractor_classes, get_info_extractor
33 from .extractor.common import UnsupportedURLIE
34 from .extractor.openload import PhantomJSwrapper
35 from .minicurses import format_text
36 from .networking import HEADRequest, Request, RequestDirector
37 from .networking.common import _REQUEST_HANDLERS, _RH_PREFERENCES
38 from .networking.exceptions import (
39 HTTPError,
40 NoSupportingHandlers,
41 RequestError,
42 SSLError,
43 network_exceptions,
44 )
45 from .plugins import directories as plugin_directories
46 from .postprocessor import _PLUGIN_CLASSES as plugin_pps
47 from .postprocessor import (
48 EmbedThumbnailPP,
49 FFmpegFixupDuplicateMoovPP,
50 FFmpegFixupDurationPP,
51 FFmpegFixupM3u8PP,
52 FFmpegFixupM4aPP,
53 FFmpegFixupStretchedPP,
54 FFmpegFixupTimestampPP,
55 FFmpegMergerPP,
56 FFmpegPostProcessor,
57 FFmpegVideoConvertorPP,
58 MoveFilesAfterDownloadPP,
59 get_postprocessor,
60 )
61 from .postprocessor.ffmpeg import resolve_mapping as resolve_recode_mapping
62 from .update import (
63 REPOSITORY,
64 _get_system_deprecation,
65 _make_label,
66 current_git_head,
67 detect_variant,
68 )
69 from .utils import (
70 DEFAULT_OUTTMPL,
71 IDENTITY,
72 LINK_TEMPLATES,
73 MEDIA_EXTENSIONS,
74 NO_DEFAULT,
75 NUMBER_RE,
76 OUTTMPL_TYPES,
77 POSTPROCESS_WHEN,
78 STR_FORMAT_RE_TMPL,
79 STR_FORMAT_TYPES,
80 ContentTooShortError,
81 DateRange,
82 DownloadCancelled,
83 DownloadError,
84 EntryNotInPlaylist,
85 ExistingVideoReached,
86 ExtractorError,
87 FormatSorter,
88 GeoRestrictedError,
89 ISO3166Utils,
90 LazyList,
91 MaxDownloadsReached,
92 Namespace,
93 PagedList,
94 PlaylistEntries,
95 Popen,
96 PostProcessingError,
97 ReExtractInfo,
98 RejectedVideoReached,
99 SameFileError,
100 UnavailableVideoError,
101 UserNotLive,
102 age_restricted,
103 args_to_str,
104 bug_reports_message,
105 date_from_str,
106 deprecation_warning,
107 determine_ext,
108 determine_protocol,
109 encode_compat_str,
110 encodeFilename,
111 error_to_compat_str,
112 escapeHTML,
113 expand_path,
114 extract_basic_auth,
115 filter_dict,
116 float_or_none,
117 format_bytes,
118 format_decimal_suffix,
119 format_field,
120 formatSeconds,
121 get_compatible_ext,
122 get_domain,
123 int_or_none,
124 iri_to_uri,
125 is_path_like,
126 join_nonempty,
127 locked_file,
128 make_archive_id,
129 make_dir,
130 number_of_digits,
131 orderedSet,
132 orderedSet_from_options,
133 parse_filesize,
134 preferredencoding,
135 prepend_extension,
136 remove_terminal_sequences,
137 render_table,
138 replace_extension,
139 sanitize_filename,
140 sanitize_path,
141 sanitize_url,
142 str_or_none,
143 strftime_or_none,
144 subtitles_filename,
145 supports_terminal_sequences,
146 system_identifier,
147 timetuple_from_msec,
148 to_high_limit_path,
149 traverse_obj,
150 try_call,
151 try_get,
152 url_basename,
153 variadic,
154 version_tuple,
155 windows_enable_vt_mode,
156 write_json_file,
157 write_string,
158 )
159 from .utils._utils import _YDLLogger
160 from .utils.networking import (
161 HTTPHeaderDict,
162 clean_headers,
163 clean_proxies,
164 std_headers,
165 )
166 from .version import CHANNEL, ORIGIN, RELEASE_GIT_HEAD, VARIANT, __version__
167
168 if compat_os_name == 'nt':
169 import ctypes
170
171
172 class YoutubeDL:
173 """YoutubeDL class.
174
175 YoutubeDL objects are the ones responsible of downloading the
176 actual video file and writing it to disk if the user has requested
177 it, among some other tasks. In most cases there should be one per
178 program. As, given a video URL, the downloader doesn't know how to
179 extract all the needed information, task that InfoExtractors do, it
180 has to pass the URL to one of them.
181
182 For this, YoutubeDL objects have a method that allows
183 InfoExtractors to be registered in a given order. When it is passed
184 a URL, the YoutubeDL object handles it to the first InfoExtractor it
185 finds that reports being able to handle it. The InfoExtractor extracts
186 all the information about the video or videos the URL refers to, and
187 YoutubeDL process the extracted information, possibly using a File
188 Downloader to download the video.
189
190 YoutubeDL objects accept a lot of parameters. In order not to saturate
191 the object constructor with arguments, it receives a dictionary of
192 options instead. These options are available through the params
193 attribute for the InfoExtractors to use. The YoutubeDL also
194 registers itself as the downloader in charge for the InfoExtractors
195 that are added to it, so this is a "mutual registration".
196
197 Available options:
198
199 username: Username for authentication purposes.
200 password: Password for authentication purposes.
201 videopassword: Password for accessing a video.
202 ap_mso: Adobe Pass multiple-system operator identifier.
203 ap_username: Multiple-system operator account username.
204 ap_password: Multiple-system operator account password.
205 usenetrc: Use netrc for authentication instead.
206 netrc_location: Location of the netrc file. Defaults to ~/.netrc.
207 netrc_cmd: Use a shell command to get credentials
208 verbose: Print additional info to stdout.
209 quiet: Do not print messages to stdout.
210 no_warnings: Do not print out anything for warnings.
211 forceprint: A dict with keys WHEN mapped to a list of templates to
212 print to stdout. The allowed keys are video or any of the
213 items in utils.POSTPROCESS_WHEN.
214 For compatibility, a single list is also accepted
215 print_to_file: A dict with keys WHEN (same as forceprint) mapped to
216 a list of tuples with (template, filename)
217 forcejson: Force printing info_dict as JSON.
218 dump_single_json: Force printing the info_dict of the whole playlist
219 (or video) as a single JSON line.
220 force_write_download_archive: Force writing download archive regardless
221 of 'skip_download' or 'simulate'.
222 simulate: Do not download the video files. If unset (or None),
223 simulate only if listsubtitles, listformats or list_thumbnails is used
224 format: Video format code. see "FORMAT SELECTION" for more details.
225 You can also pass a function. The function takes 'ctx' as
226 argument and returns the formats to download.
227 See "build_format_selector" for an implementation
228 allow_unplayable_formats: Allow unplayable formats to be extracted and downloaded.
229 ignore_no_formats_error: Ignore "No video formats" error. Usefull for
230 extracting metadata even if the video is not actually
231 available for download (experimental)
232 format_sort: A list of fields by which to sort the video formats.
233 See "Sorting Formats" for more details.
234 format_sort_force: Force the given format_sort. see "Sorting Formats"
235 for more details.
236 prefer_free_formats: Whether to prefer video formats with free containers
237 over non-free ones of same quality.
238 allow_multiple_video_streams: Allow multiple video streams to be merged
239 into a single file
240 allow_multiple_audio_streams: Allow multiple audio streams to be merged
241 into a single file
242 check_formats Whether to test if the formats are downloadable.
243 Can be True (check all), False (check none),
244 'selected' (check selected formats),
245 or None (check only if requested by extractor)
246 paths: Dictionary of output paths. The allowed keys are 'home'
247 'temp' and the keys of OUTTMPL_TYPES (in utils/_utils.py)
248 outtmpl: Dictionary of templates for output names. Allowed keys
249 are 'default' and the keys of OUTTMPL_TYPES (in utils/_utils.py).
250 For compatibility with youtube-dl, a single string can also be used
251 outtmpl_na_placeholder: Placeholder for unavailable meta fields.
252 restrictfilenames: Do not allow "&" and spaces in file names
253 trim_file_name: Limit length of filename (extension excluded)
254 windowsfilenames: Force the filenames to be windows compatible
255 ignoreerrors: Do not stop on download/postprocessing errors.
256 Can be 'only_download' to ignore only download errors.
257 Default is 'only_download' for CLI, but False for API
258 skip_playlist_after_errors: Number of allowed failures until the rest of
259 the playlist is skipped
260 allowed_extractors: List of regexes to match against extractor names that are allowed
261 overwrites: Overwrite all video and metadata files if True,
262 overwrite only non-video files if None
263 and don't overwrite any file if False
264 playlist_items: Specific indices of playlist to download.
265 playlistrandom: Download playlist items in random order.
266 lazy_playlist: Process playlist entries as they are received.
267 matchtitle: Download only matching titles.
268 rejecttitle: Reject downloads for matching titles.
269 logger: Log messages to a logging.Logger instance.
270 logtostderr: Print everything to stderr instead of stdout.
271 consoletitle: Display progress in console window's titlebar.
272 writedescription: Write the video description to a .description file
273 writeinfojson: Write the video description to a .info.json file
274 clean_infojson: Remove internal metadata from the infojson
275 getcomments: Extract video comments. This will not be written to disk
276 unless writeinfojson is also given
277 writeannotations: Write the video annotations to a .annotations.xml file
278 writethumbnail: Write the thumbnail image to a file
279 allow_playlist_files: Whether to write playlists' description, infojson etc
280 also to disk when using the 'write*' options
281 write_all_thumbnails: Write all thumbnail formats to files
282 writelink: Write an internet shortcut file, depending on the
283 current platform (.url/.webloc/.desktop)
284 writeurllink: Write a Windows internet shortcut file (.url)
285 writewebloclink: Write a macOS internet shortcut file (.webloc)
286 writedesktoplink: Write a Linux internet shortcut file (.desktop)
287 writesubtitles: Write the video subtitles to a file
288 writeautomaticsub: Write the automatically generated subtitles to a file
289 listsubtitles: Lists all available subtitles for the video
290 subtitlesformat: The format code for subtitles
291 subtitleslangs: List of languages of the subtitles to download (can be regex).
292 The list may contain "all" to refer to all the available
293 subtitles. The language can be prefixed with a "-" to
294 exclude it from the requested languages, e.g. ['all', '-live_chat']
295 keepvideo: Keep the video file after post-processing
296 daterange: A utils.DateRange object, download only if the upload_date is in the range.
297 skip_download: Skip the actual download of the video file
298 cachedir: Location of the cache files in the filesystem.
299 False to disable filesystem cache.
300 noplaylist: Download single video instead of a playlist if in doubt.
301 age_limit: An integer representing the user's age in years.
302 Unsuitable videos for the given age are skipped.
303 min_views: An integer representing the minimum view count the video
304 must have in order to not be skipped.
305 Videos without view count information are always
306 downloaded. None for no limit.
307 max_views: An integer representing the maximum view count.
308 Videos that are more popular than that are not
309 downloaded.
310 Videos without view count information are always
311 downloaded. None for no limit.
312 download_archive: A set, or the name of a file where all downloads are recorded.
313 Videos already present in the file are not downloaded again.
314 break_on_existing: Stop the download process after attempting to download a
315 file that is in the archive.
316 break_per_url: Whether break_on_reject and break_on_existing
317 should act on each input URL as opposed to for the entire queue
318 cookiefile: File name or text stream from where cookies should be read and dumped to
319 cookiesfrombrowser: A tuple containing the name of the browser, the profile
320 name/path from where cookies are loaded, the name of the keyring,
321 and the container name, e.g. ('chrome', ) or
322 ('vivaldi', 'default', 'BASICTEXT') or ('firefox', 'default', None, 'Meta')
323 legacyserverconnect: Explicitly allow HTTPS connection to servers that do not
324 support RFC 5746 secure renegotiation
325 nocheckcertificate: Do not verify SSL certificates
326 client_certificate: Path to client certificate file in PEM format. May include the private key
327 client_certificate_key: Path to private key file for client certificate
328 client_certificate_password: Password for client certificate private key, if encrypted.
329 If not provided and the key is encrypted, yt-dlp will ask interactively
330 prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
331 (Only supported by some extractors)
332 enable_file_urls: Enable file:// URLs. This is disabled by default for security reasons.
333 http_headers: A dictionary of custom headers to be used for all requests
334 proxy: URL of the proxy server to use
335 geo_verification_proxy: URL of the proxy to use for IP address verification
336 on geo-restricted sites.
337 socket_timeout: Time to wait for unresponsive hosts, in seconds
338 bidi_workaround: Work around buggy terminals without bidirectional text
339 support, using fridibi
340 debug_printtraffic:Print out sent and received HTTP traffic
341 default_search: Prepend this string if an input url is not valid.
342 'auto' for elaborate guessing
343 encoding: Use this encoding instead of the system-specified.
344 extract_flat: Whether to resolve and process url_results further
345 * False: Always process. Default for API
346 * True: Never process
347 * 'in_playlist': Do not process inside playlist/multi_video
348 * 'discard': Always process, but don't return the result
349 from inside playlist/multi_video
350 * 'discard_in_playlist': Same as "discard", but only for
351 playlists (not multi_video). Default for CLI
352 wait_for_video: If given, wait for scheduled streams to become available.
353 The value should be a tuple containing the range
354 (min_secs, max_secs) to wait between retries
355 postprocessors: A list of dictionaries, each with an entry
356 * key: The name of the postprocessor. See
357 yt_dlp/postprocessor/__init__.py for a list.
358 * when: When to run the postprocessor. Allowed values are
359 the entries of utils.POSTPROCESS_WHEN
360 Assumed to be 'post_process' if not given
361 progress_hooks: A list of functions that get called on download
362 progress, with a dictionary with the entries
363 * status: One of "downloading", "error", or "finished".
364 Check this first and ignore unknown values.
365 * info_dict: The extracted info_dict
366
367 If status is one of "downloading", or "finished", the
368 following properties may also be present:
369 * filename: The final filename (always present)
370 * tmpfilename: The filename we're currently writing to
371 * downloaded_bytes: Bytes on disk
372 * total_bytes: Size of the whole file, None if unknown
373 * total_bytes_estimate: Guess of the eventual file size,
374 None if unavailable.
375 * elapsed: The number of seconds since download started.
376 * eta: The estimated time in seconds, None if unknown
377 * speed: The download speed in bytes/second, None if
378 unknown
379 * fragment_index: The counter of the currently
380 downloaded video fragment.
381 * fragment_count: The number of fragments (= individual
382 files that will be merged)
383
384 Progress hooks are guaranteed to be called at least once
385 (with status "finished") if the download is successful.
386 postprocessor_hooks: A list of functions that get called on postprocessing
387 progress, with a dictionary with the entries
388 * status: One of "started", "processing", or "finished".
389 Check this first and ignore unknown values.
390 * postprocessor: Name of the postprocessor
391 * info_dict: The extracted info_dict
392
393 Progress hooks are guaranteed to be called at least twice
394 (with status "started" and "finished") if the processing is successful.
395 merge_output_format: "/" separated list of extensions to use when merging formats.
396 final_ext: Expected final extension; used to detect when the file was
397 already downloaded and converted
398 fixup: Automatically correct known faults of the file.
399 One of:
400 - "never": do nothing
401 - "warn": only emit a warning
402 - "detect_or_warn": check whether we can do anything
403 about it, warn otherwise (default)
404 source_address: Client-side IP address to bind to.
405 sleep_interval_requests: Number of seconds to sleep between requests
406 during extraction
407 sleep_interval: Number of seconds to sleep before each download when
408 used alone or a lower bound of a range for randomized
409 sleep before each download (minimum possible number
410 of seconds to sleep) when used along with
411 max_sleep_interval.
412 max_sleep_interval:Upper bound of a range for randomized sleep before each
413 download (maximum possible number of seconds to sleep).
414 Must only be used along with sleep_interval.
415 Actual sleep time will be a random float from range
416 [sleep_interval; max_sleep_interval].
417 sleep_interval_subtitles: Number of seconds to sleep before each subtitle download
418 listformats: Print an overview of available video formats and exit.
419 list_thumbnails: Print a table of all thumbnails and exit.
420 match_filter: A function that gets called for every video with the signature
421 (info_dict, *, incomplete: bool) -> Optional[str]
422 For backward compatibility with youtube-dl, the signature
423 (info_dict) -> Optional[str] is also allowed.
424 - If it returns a message, the video is ignored.
425 - If it returns None, the video is downloaded.
426 - If it returns utils.NO_DEFAULT, the user is interactively
427 asked whether to download the video.
428 - Raise utils.DownloadCancelled(msg) to abort remaining
429 downloads when a video is rejected.
430 match_filter_func in utils/_utils.py is one example for this.
431 color: A Dictionary with output stream names as keys
432 and their respective color policy as values.
433 Can also just be a single color policy,
434 in which case it applies to all outputs.
435 Valid stream names are 'stdout' and 'stderr'.
436 Valid color policies are one of 'always', 'auto', 'no_color' or 'never'.
437 geo_bypass: Bypass geographic restriction via faking X-Forwarded-For
438 HTTP header
439 geo_bypass_country:
440 Two-letter ISO 3166-2 country code that will be used for
441 explicit geographic restriction bypassing via faking
442 X-Forwarded-For HTTP header
443 geo_bypass_ip_block:
444 IP range in CIDR notation that will be used similarly to
445 geo_bypass_country
446 external_downloader: A dictionary of protocol keys and the executable of the
447 external downloader to use for it. The allowed protocols
448 are default|http|ftp|m3u8|dash|rtsp|rtmp|mms.
449 Set the value to 'native' to use the native downloader
450 compat_opts: Compatibility options. See "Differences in default behavior".
451 The following options do not work when used through the API:
452 filename, abort-on-error, multistreams, no-live-chat, format-sort
453 no-clean-infojson, no-playlist-metafiles, no-keep-subs, no-attach-info-json.
454 Refer __init__.py for their implementation
455 progress_template: Dictionary of templates for progress outputs.
456 Allowed keys are 'download', 'postprocess',
457 'download-title' (console title) and 'postprocess-title'.
458 The template is mapped on a dictionary with keys 'progress' and 'info'
459 retry_sleep_functions: Dictionary of functions that takes the number of attempts
460 as argument and returns the time to sleep in seconds.
461 Allowed keys are 'http', 'fragment', 'file_access'
462 download_ranges: A callback function that gets called for every video with
463 the signature (info_dict, ydl) -> Iterable[Section].
464 Only the returned sections will be downloaded.
465 Each Section is a dict with the following keys:
466 * start_time: Start time of the section in seconds
467 * end_time: End time of the section in seconds
468 * title: Section title (Optional)
469 * index: Section number (Optional)
470 force_keyframes_at_cuts: Re-encode the video when downloading ranges to get precise cuts
471 noprogress: Do not print the progress bar
472 live_from_start: Whether to download livestreams videos from the start
473
474 The following parameters are not used by YoutubeDL itself, they are used by
475 the downloader (see yt_dlp/downloader/common.py):
476 nopart, updatetime, buffersize, ratelimit, throttledratelimit, min_filesize,
477 max_filesize, test, noresizebuffer, retries, file_access_retries, fragment_retries,
478 continuedl, xattr_set_filesize, hls_use_mpegts, http_chunk_size,
479 external_downloader_args, concurrent_fragment_downloads.
480
481 The following options are used by the post processors:
482 ffmpeg_location: Location of the ffmpeg/avconv binary; either the path
483 to the binary or its containing directory.
484 postprocessor_args: A dictionary of postprocessor/executable keys (in lower case)
485 and a list of additional command-line arguments for the
486 postprocessor/executable. The dict can also have "PP+EXE" keys
487 which are used when the given exe is used by the given PP.
488 Use 'default' as the name for arguments to passed to all PP
489 For compatibility with youtube-dl, a single list of args
490 can also be used
491
492 The following options are used by the extractors:
493 extractor_retries: Number of times to retry for known errors (default: 3)
494 dynamic_mpd: Whether to process dynamic DASH manifests (default: True)
495 hls_split_discontinuity: Split HLS playlists to different formats at
496 discontinuities such as ad breaks (default: False)
497 extractor_args: A dictionary of arguments to be passed to the extractors.
498 See "EXTRACTOR ARGUMENTS" for details.
499 E.g. {'youtube': {'skip': ['dash', 'hls']}}
500 mark_watched: Mark videos watched (even with --simulate). Only for YouTube
501
502 The following options are deprecated and may be removed in the future:
503
504 break_on_reject: Stop the download process when encountering a video that
505 has been filtered out.
506 - `raise DownloadCancelled(msg)` in match_filter instead
507 force_generic_extractor: Force downloader to use the generic extractor
508 - Use allowed_extractors = ['generic', 'default']
509 playliststart: - Use playlist_items
510 Playlist item to start at.
511 playlistend: - Use playlist_items
512 Playlist item to end at.
513 playlistreverse: - Use playlist_items
514 Download playlist items in reverse order.
515 forceurl: - Use forceprint
516 Force printing final URL.
517 forcetitle: - Use forceprint
518 Force printing title.
519 forceid: - Use forceprint
520 Force printing ID.
521 forcethumbnail: - Use forceprint
522 Force printing thumbnail URL.
523 forcedescription: - Use forceprint
524 Force printing description.
525 forcefilename: - Use forceprint
526 Force printing final filename.
527 forceduration: - Use forceprint
528 Force printing duration.
529 allsubtitles: - Use subtitleslangs = ['all']
530 Downloads all the subtitles of the video
531 (requires writesubtitles or writeautomaticsub)
532 include_ads: - Doesn't work
533 Download ads as well
534 call_home: - Not implemented
535 Boolean, true iff we are allowed to contact the
536 yt-dlp servers for debugging.
537 post_hooks: - Register a custom postprocessor
538 A list of functions that get called as the final step
539 for each video file, after all postprocessors have been
540 called. The filename will be passed as the only argument.
541 hls_prefer_native: - Use external_downloader = {'m3u8': 'native'} or {'m3u8': 'ffmpeg'}.
542 Use the native HLS downloader instead of ffmpeg/avconv
543 if True, otherwise use ffmpeg/avconv if False, otherwise
544 use downloader suggested by extractor if None.
545 prefer_ffmpeg: - avconv support is deprecated
546 If False, use avconv instead of ffmpeg if both are available,
547 otherwise prefer ffmpeg.
548 youtube_include_dash_manifest: - Use extractor_args
549 If True (default), DASH manifests and related
550 data will be downloaded and processed by extractor.
551 You can reduce network I/O by disabling it if you don't
552 care about DASH. (only for youtube)
553 youtube_include_hls_manifest: - Use extractor_args
554 If True (default), HLS manifests and related
555 data will be downloaded and processed by extractor.
556 You can reduce network I/O by disabling it if you don't
557 care about HLS. (only for youtube)
558 no_color: Same as `color='no_color'`
559 no_overwrites: Same as `overwrites=False`
560 """
561
562 _NUMERIC_FIELDS = {
563 'width', 'height', 'asr', 'audio_channels', 'fps',
564 'tbr', 'abr', 'vbr', 'filesize', 'filesize_approx',
565 'timestamp', 'release_timestamp',
566 'duration', 'view_count', 'like_count', 'dislike_count', 'repost_count',
567 'average_rating', 'comment_count', 'age_limit',
568 'start_time', 'end_time',
569 'chapter_number', 'season_number', 'episode_number',
570 'track_number', 'disc_number', 'release_year',
571 }
572
573 _format_fields = {
574 # NB: Keep in sync with the docstring of extractor/common.py
575 'url', 'manifest_url', 'manifest_stream_number', 'ext', 'format', 'format_id', 'format_note',
576 'width', 'height', 'aspect_ratio', 'resolution', 'dynamic_range', 'tbr', 'abr', 'acodec', 'asr', 'audio_channels',
577 'vbr', 'fps', 'vcodec', 'container', 'filesize', 'filesize_approx', 'rows', 'columns',
578 'player_url', 'protocol', 'fragment_base_url', 'fragments', 'is_from_start', 'is_dash_periods', 'request_data',
579 'preference', 'language', 'language_preference', 'quality', 'source_preference', 'cookies',
580 'http_headers', 'stretched_ratio', 'no_resume', 'has_drm', 'extra_param_to_segment_url', 'hls_aes', 'downloader_options',
581 'page_url', 'app', 'play_path', 'tc_url', 'flash_version', 'rtmp_live', 'rtmp_conn', 'rtmp_protocol', 'rtmp_real_time'
582 }
583 _deprecated_multivalue_fields = {
584 'album_artist': 'album_artists',
585 'artist': 'artists',
586 'composer': 'composers',
587 'creator': 'creators',
588 'genre': 'genres',
589 }
590 _format_selection_exts = {
591 'audio': set(MEDIA_EXTENSIONS.common_audio),
592 'video': set(MEDIA_EXTENSIONS.common_video + ('3gp', )),
593 'storyboards': set(MEDIA_EXTENSIONS.storyboards),
594 }
595
596 def __init__(self, params=None, auto_init=True):
597 """Create a FileDownloader object with the given options.
598 @param auto_init Whether to load the default extractors and print header (if verbose).
599 Set to 'no_verbose_header' to not print the header
600 """
601 if params is None:
602 params = {}
603 self.params = params
604 self._ies = {}
605 self._ies_instances = {}
606 self._pps = {k: [] for k in POSTPROCESS_WHEN}
607 self._printed_messages = set()
608 self._first_webpage_request = True
609 self._post_hooks = []
610 self._progress_hooks = []
611 self._postprocessor_hooks = []
612 self._download_retcode = 0
613 self._num_downloads = 0
614 self._num_videos = 0
615 self._playlist_level = 0
616 self._playlist_urls = set()
617 self.cache = Cache(self)
618 self.__header_cookies = []
619
620 stdout = sys.stderr if self.params.get('logtostderr') else sys.stdout
621 self._out_files = Namespace(
622 out=stdout,
623 error=sys.stderr,
624 screen=sys.stderr if self.params.get('quiet') else stdout,
625 console=None if compat_os_name == 'nt' else next(
626 filter(supports_terminal_sequences, (sys.stderr, sys.stdout)), None)
627 )
628
629 try:
630 windows_enable_vt_mode()
631 except Exception as e:
632 self.write_debug(f'Failed to enable VT mode: {e}')
633
634 if self.params.get('no_color'):
635 if self.params.get('color') is not None:
636 self.params.setdefault('_warnings', []).append(
637 'Overwriting params from "color" with "no_color"')
638 self.params['color'] = 'no_color'
639
640 term_allow_color = os.getenv('TERM', '').lower() != 'dumb'
641 no_color = bool(os.getenv('NO_COLOR'))
642
643 def process_color_policy(stream):
644 stream_name = {sys.stdout: 'stdout', sys.stderr: 'stderr'}[stream]
645 policy = traverse_obj(self.params, ('color', (stream_name, None), {str}), get_all=False)
646 if policy in ('auto', None):
647 if term_allow_color and supports_terminal_sequences(stream):
648 return 'no_color' if no_color else True
649 return False
650 assert policy in ('always', 'never', 'no_color'), policy
651 return {'always': True, 'never': False}.get(policy, policy)
652
653 self._allow_colors = Namespace(**{
654 name: process_color_policy(stream)
655 for name, stream in self._out_files.items_ if name != 'console'
656 })
657
658 system_deprecation = _get_system_deprecation()
659 if system_deprecation:
660 self.deprecated_feature(system_deprecation.replace('\n', '\n '))
661
662 if self.params.get('allow_unplayable_formats'):
663 self.report_warning(
664 f'You have asked for {self._format_err("UNPLAYABLE", self.Styles.EMPHASIS)} formats to be listed/downloaded. '
665 'This is a developer option intended for debugging. \n'
666 ' If you experience any issues while using this option, '
667 f'{self._format_err("DO NOT", self.Styles.ERROR)} open a bug report')
668
669 if self.params.get('bidi_workaround', False):
670 try:
671 import pty
672 master, slave = pty.openpty()
673 width = shutil.get_terminal_size().columns
674 width_args = [] if width is None else ['-w', str(width)]
675 sp_kwargs = {'stdin': subprocess.PIPE, 'stdout': slave, 'stderr': self._out_files.error}
676 try:
677 self._output_process = Popen(['bidiv'] + width_args, **sp_kwargs)
678 except OSError:
679 self._output_process = Popen(['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
680 self._output_channel = os.fdopen(master, 'rb')
681 except OSError as ose:
682 if ose.errno == errno.ENOENT:
683 self.report_warning(
684 'Could not find fribidi executable, ignoring --bidi-workaround. '
685 'Make sure that fribidi is an executable file in one of the directories in your $PATH.')
686 else:
687 raise
688
689 self.params['compat_opts'] = set(self.params.get('compat_opts', ()))
690 self.params['http_headers'] = HTTPHeaderDict(std_headers, self.params.get('http_headers'))
691 self._load_cookies(self.params['http_headers'].get('Cookie')) # compat
692 self.params['http_headers'].pop('Cookie', None)
693
694 if auto_init and auto_init != 'no_verbose_header':
695 self.print_debug_header()
696
697 def check_deprecated(param, option, suggestion):
698 if self.params.get(param) is not None:
699 self.report_warning(f'{option} is deprecated. Use {suggestion} instead')
700 return True
701 return False
702
703 if check_deprecated('cn_verification_proxy', '--cn-verification-proxy', '--geo-verification-proxy'):
704 if self.params.get('geo_verification_proxy') is None:
705 self.params['geo_verification_proxy'] = self.params['cn_verification_proxy']
706
707 check_deprecated('autonumber', '--auto-number', '-o "%(autonumber)s-%(title)s.%(ext)s"')
708 check_deprecated('usetitle', '--title', '-o "%(title)s-%(id)s.%(ext)s"')
709 check_deprecated('useid', '--id', '-o "%(id)s.%(ext)s"')
710
711 for msg in self.params.get('_warnings', []):
712 self.report_warning(msg)
713 for msg in self.params.get('_deprecation_warnings', []):
714 self.deprecated_feature(msg)
715
716 if 'list-formats' in self.params['compat_opts']:
717 self.params['listformats_table'] = False
718
719 if 'overwrites' not in self.params and self.params.get('nooverwrites') is not None:
720 # nooverwrites was unnecessarily changed to overwrites
721 # in 0c3d0f51778b153f65c21906031c2e091fcfb641
722 # This ensures compatibility with both keys
723 self.params['overwrites'] = not self.params['nooverwrites']
724 elif self.params.get('overwrites') is None:
725 self.params.pop('overwrites', None)
726 else:
727 self.params['nooverwrites'] = not self.params['overwrites']
728
729 if self.params.get('simulate') is None and any((
730 self.params.get('list_thumbnails'),
731 self.params.get('listformats'),
732 self.params.get('listsubtitles'),
733 )):
734 self.params['simulate'] = 'list_only'
735
736 self.params.setdefault('forceprint', {})
737 self.params.setdefault('print_to_file', {})
738
739 # Compatibility with older syntax
740 if not isinstance(params['forceprint'], dict):
741 self.params['forceprint'] = {'video': params['forceprint']}
742
743 if auto_init:
744 self.add_default_info_extractors()
745
746 if (sys.platform != 'win32'
747 and sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968']
748 and not self.params.get('restrictfilenames', False)):
749 # Unicode filesystem API will throw errors (#1474, #13027)
750 self.report_warning(
751 'Assuming --restrict-filenames since file system encoding '
752 'cannot encode all characters. '
753 'Set the LC_ALL environment variable to fix this.')
754 self.params['restrictfilenames'] = True
755
756 self._parse_outtmpl()
757
758 # Creating format selector here allows us to catch syntax errors before the extraction
759 self.format_selector = (
760 self.params.get('format') if self.params.get('format') in (None, '-')
761 else self.params['format'] if callable(self.params['format'])
762 else self.build_format_selector(self.params['format']))
763
764 hooks = {
765 'post_hooks': self.add_post_hook,
766 'progress_hooks': self.add_progress_hook,
767 'postprocessor_hooks': self.add_postprocessor_hook,
768 }
769 for opt, fn in hooks.items():
770 for ph in self.params.get(opt, []):
771 fn(ph)
772
773 for pp_def_raw in self.params.get('postprocessors', []):
774 pp_def = dict(pp_def_raw)
775 when = pp_def.pop('when', 'post_process')
776 self.add_post_processor(
777 get_postprocessor(pp_def.pop('key'))(self, **pp_def),
778 when=when)
779
780 def preload_download_archive(fn):
781 """Preload the archive, if any is specified"""
782 archive = set()
783 if fn is None:
784 return archive
785 elif not is_path_like(fn):
786 return fn
787
788 self.write_debug(f'Loading archive file {fn!r}')
789 try:
790 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
791 for line in archive_file:
792 archive.add(line.strip())
793 except OSError as ioe:
794 if ioe.errno != errno.ENOENT:
795 raise
796 return archive
797
798 self.archive = preload_download_archive(self.params.get('download_archive'))
799
800 def warn_if_short_id(self, argv):
801 # short YouTube ID starting with dash?
802 idxs = [
803 i for i, a in enumerate(argv)
804 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
805 if idxs:
806 correct_argv = (
807 ['yt-dlp']
808 + [a for i, a in enumerate(argv) if i not in idxs]
809 + ['--'] + [argv[i] for i in idxs]
810 )
811 self.report_warning(
812 'Long argument string detected. '
813 'Use -- to separate parameters and URLs, like this:\n%s' %
814 args_to_str(correct_argv))
815
816 def add_info_extractor(self, ie):
817 """Add an InfoExtractor object to the end of the list."""
818 ie_key = ie.ie_key()
819 self._ies[ie_key] = ie
820 if not isinstance(ie, type):
821 self._ies_instances[ie_key] = ie
822 ie.set_downloader(self)
823
824 def get_info_extractor(self, ie_key):
825 """
826 Get an instance of an IE with name ie_key, it will try to get one from
827 the _ies list, if there's no instance it will create a new one and add
828 it to the extractor list.
829 """
830 ie = self._ies_instances.get(ie_key)
831 if ie is None:
832 ie = get_info_extractor(ie_key)()
833 self.add_info_extractor(ie)
834 return ie
835
836 def add_default_info_extractors(self):
837 """
838 Add the InfoExtractors returned by gen_extractors to the end of the list
839 """
840 all_ies = {ie.IE_NAME.lower(): ie for ie in gen_extractor_classes()}
841 all_ies['end'] = UnsupportedURLIE()
842 try:
843 ie_names = orderedSet_from_options(
844 self.params.get('allowed_extractors', ['default']), {
845 'all': list(all_ies),
846 'default': [name for name, ie in all_ies.items() if ie._ENABLED],
847 }, use_regex=True)
848 except re.error as e:
849 raise ValueError(f'Wrong regex for allowed_extractors: {e.pattern}')
850 for name in ie_names:
851 self.add_info_extractor(all_ies[name])
852 self.write_debug(f'Loaded {len(ie_names)} extractors')
853
854 def add_post_processor(self, pp, when='post_process'):
855 """Add a PostProcessor object to the end of the chain."""
856 assert when in POSTPROCESS_WHEN, f'Invalid when={when}'
857 self._pps[when].append(pp)
858 pp.set_downloader(self)
859
860 def add_post_hook(self, ph):
861 """Add the post hook"""
862 self._post_hooks.append(ph)
863
864 def add_progress_hook(self, ph):
865 """Add the download progress hook"""
866 self._progress_hooks.append(ph)
867
868 def add_postprocessor_hook(self, ph):
869 """Add the postprocessing progress hook"""
870 self._postprocessor_hooks.append(ph)
871 for pps in self._pps.values():
872 for pp in pps:
873 pp.add_progress_hook(ph)
874
875 def _bidi_workaround(self, message):
876 if not hasattr(self, '_output_channel'):
877 return message
878
879 assert hasattr(self, '_output_process')
880 assert isinstance(message, str)
881 line_count = message.count('\n') + 1
882 self._output_process.stdin.write((message + '\n').encode())
883 self._output_process.stdin.flush()
884 res = ''.join(self._output_channel.readline().decode()
885 for _ in range(line_count))
886 return res[:-len('\n')]
887
888 def _write_string(self, message, out=None, only_once=False):
889 if only_once:
890 if message in self._printed_messages:
891 return
892 self._printed_messages.add(message)
893 write_string(message, out=out, encoding=self.params.get('encoding'))
894
895 def to_stdout(self, message, skip_eol=False, quiet=None):
896 """Print message to stdout"""
897 if quiet is not None:
898 self.deprecation_warning('"YoutubeDL.to_stdout" no longer accepts the argument quiet. '
899 'Use "YoutubeDL.to_screen" instead')
900 if skip_eol is not False:
901 self.deprecation_warning('"YoutubeDL.to_stdout" no longer accepts the argument skip_eol. '
902 'Use "YoutubeDL.to_screen" instead')
903 self._write_string(f'{self._bidi_workaround(message)}\n', self._out_files.out)
904
905 def to_screen(self, message, skip_eol=False, quiet=None, only_once=False):
906 """Print message to screen if not in quiet mode"""
907 if self.params.get('logger'):
908 self.params['logger'].debug(message)
909 return
910 if (self.params.get('quiet') if quiet is None else quiet) and not self.params.get('verbose'):
911 return
912 self._write_string(
913 '%s%s' % (self._bidi_workaround(message), ('' if skip_eol else '\n')),
914 self._out_files.screen, only_once=only_once)
915
916 def to_stderr(self, message, only_once=False):
917 """Print message to stderr"""
918 assert isinstance(message, str)
919 if self.params.get('logger'):
920 self.params['logger'].error(message)
921 else:
922 self._write_string(f'{self._bidi_workaround(message)}\n', self._out_files.error, only_once=only_once)
923
924 def _send_console_code(self, code):
925 if compat_os_name == 'nt' or not self._out_files.console:
926 return
927 self._write_string(code, self._out_files.console)
928
929 def to_console_title(self, message):
930 if not self.params.get('consoletitle', False):
931 return
932 message = remove_terminal_sequences(message)
933 if compat_os_name == 'nt':
934 if ctypes.windll.kernel32.GetConsoleWindow():
935 # c_wchar_p() might not be necessary if `message` is
936 # already of type unicode()
937 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
938 else:
939 self._send_console_code(f'\033]0;{message}\007')
940
941 def save_console_title(self):
942 if not self.params.get('consoletitle') or self.params.get('simulate'):
943 return
944 self._send_console_code('\033[22;0t') # Save the title on stack
945
946 def restore_console_title(self):
947 if not self.params.get('consoletitle') or self.params.get('simulate'):
948 return
949 self._send_console_code('\033[23;0t') # Restore the title from stack
950
951 def __enter__(self):
952 self.save_console_title()
953 return self
954
955 def save_cookies(self):
956 if self.params.get('cookiefile') is not None:
957 self.cookiejar.save()
958
959 def __exit__(self, *args):
960 self.restore_console_title()
961 self.close()
962
963 def close(self):
964 self.save_cookies()
965 if '_request_director' in self.__dict__:
966 self._request_director.close()
967 del self._request_director
968
969 def trouble(self, message=None, tb=None, is_error=True):
970 """Determine action to take when a download problem appears.
971
972 Depending on if the downloader has been configured to ignore
973 download errors or not, this method may throw an exception or
974 not when errors are found, after printing the message.
975
976 @param tb If given, is additional traceback information
977 @param is_error Whether to raise error according to ignorerrors
978 """
979 if message is not None:
980 self.to_stderr(message)
981 if self.params.get('verbose'):
982 if tb is None:
983 if sys.exc_info()[0]: # if .trouble has been called from an except block
984 tb = ''
985 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
986 tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
987 tb += encode_compat_str(traceback.format_exc())
988 else:
989 tb_data = traceback.format_list(traceback.extract_stack())
990 tb = ''.join(tb_data)
991 if tb:
992 self.to_stderr(tb)
993 if not is_error:
994 return
995 if not self.params.get('ignoreerrors'):
996 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
997 exc_info = sys.exc_info()[1].exc_info
998 else:
999 exc_info = sys.exc_info()
1000 raise DownloadError(message, exc_info)
1001 self._download_retcode = 1
1002
1003 Styles = Namespace(
1004 HEADERS='yellow',
1005 EMPHASIS='light blue',
1006 FILENAME='green',
1007 ID='green',
1008 DELIM='blue',
1009 ERROR='red',
1010 BAD_FORMAT='light red',
1011 WARNING='yellow',
1012 SUPPRESS='light black',
1013 )
1014
1015 def _format_text(self, handle, allow_colors, text, f, fallback=None, *, test_encoding=False):
1016 text = str(text)
1017 if test_encoding:
1018 original_text = text
1019 # handle.encoding can be None. See https://github.com/yt-dlp/yt-dlp/issues/2711
1020 encoding = self.params.get('encoding') or getattr(handle, 'encoding', None) or 'ascii'
1021 text = text.encode(encoding, 'ignore').decode(encoding)
1022 if fallback is not None and text != original_text:
1023 text = fallback
1024 return format_text(text, f) if allow_colors is True else text if fallback is None else fallback
1025
1026 def _format_out(self, *args, **kwargs):
1027 return self._format_text(self._out_files.out, self._allow_colors.out, *args, **kwargs)
1028
1029 def _format_screen(self, *args, **kwargs):
1030 return self._format_text(self._out_files.screen, self._allow_colors.screen, *args, **kwargs)
1031
1032 def _format_err(self, *args, **kwargs):
1033 return self._format_text(self._out_files.error, self._allow_colors.error, *args, **kwargs)
1034
1035 def report_warning(self, message, only_once=False):
1036 '''
1037 Print the message to stderr, it will be prefixed with 'WARNING:'
1038 If stderr is a tty file the 'WARNING:' will be colored
1039 '''
1040 if self.params.get('logger') is not None:
1041 self.params['logger'].warning(message)
1042 else:
1043 if self.params.get('no_warnings'):
1044 return
1045 self.to_stderr(f'{self._format_err("WARNING:", self.Styles.WARNING)} {message}', only_once)
1046
1047 def deprecation_warning(self, message, *, stacklevel=0):
1048 deprecation_warning(
1049 message, stacklevel=stacklevel + 1, printer=self.report_error, is_error=False)
1050
1051 def deprecated_feature(self, message):
1052 if self.params.get('logger') is not None:
1053 self.params['logger'].warning(f'Deprecated Feature: {message}')
1054 self.to_stderr(f'{self._format_err("Deprecated Feature:", self.Styles.ERROR)} {message}', True)
1055
1056 def report_error(self, message, *args, **kwargs):
1057 '''
1058 Do the same as trouble, but prefixes the message with 'ERROR:', colored
1059 in red if stderr is a tty file.
1060 '''
1061 self.trouble(f'{self._format_err("ERROR:", self.Styles.ERROR)} {message}', *args, **kwargs)
1062
1063 def write_debug(self, message, only_once=False):
1064 '''Log debug message or Print message to stderr'''
1065 if not self.params.get('verbose', False):
1066 return
1067 message = f'[debug] {message}'
1068 if self.params.get('logger'):
1069 self.params['logger'].debug(message)
1070 else:
1071 self.to_stderr(message, only_once)
1072
1073 def report_file_already_downloaded(self, file_name):
1074 """Report file has already been fully downloaded."""
1075 try:
1076 self.to_screen('[download] %s has already been downloaded' % file_name)
1077 except UnicodeEncodeError:
1078 self.to_screen('[download] The file has already been downloaded')
1079
1080 def report_file_delete(self, file_name):
1081 """Report that existing file will be deleted."""
1082 try:
1083 self.to_screen('Deleting existing file %s' % file_name)
1084 except UnicodeEncodeError:
1085 self.to_screen('Deleting existing file')
1086
1087 def raise_no_formats(self, info, forced=False, *, msg=None):
1088 has_drm = info.get('_has_drm')
1089 ignored, expected = self.params.get('ignore_no_formats_error'), bool(msg)
1090 msg = msg or has_drm and 'This video is DRM protected' or 'No video formats found!'
1091 if forced or not ignored:
1092 raise ExtractorError(msg, video_id=info['id'], ie=info['extractor'],
1093 expected=has_drm or ignored or expected)
1094 else:
1095 self.report_warning(msg)
1096
1097 def parse_outtmpl(self):
1098 self.deprecation_warning('"YoutubeDL.parse_outtmpl" is deprecated and may be removed in a future version')
1099 self._parse_outtmpl()
1100 return self.params['outtmpl']
1101
1102 def _parse_outtmpl(self):
1103 sanitize = IDENTITY
1104 if self.params.get('restrictfilenames'): # Remove spaces in the default template
1105 sanitize = lambda x: x.replace(' - ', ' ').replace(' ', '-')
1106
1107 outtmpl = self.params.setdefault('outtmpl', {})
1108 if not isinstance(outtmpl, dict):
1109 self.params['outtmpl'] = outtmpl = {'default': outtmpl}
1110 outtmpl.update({k: sanitize(v) for k, v in DEFAULT_OUTTMPL.items() if outtmpl.get(k) is None})
1111
1112 def get_output_path(self, dir_type='', filename=None):
1113 paths = self.params.get('paths', {})
1114 assert isinstance(paths, dict), '"paths" parameter must be a dictionary'
1115 path = os.path.join(
1116 expand_path(paths.get('home', '').strip()),
1117 expand_path(paths.get(dir_type, '').strip()) if dir_type else '',
1118 filename or '')
1119 return sanitize_path(path, force=self.params.get('windowsfilenames'))
1120
1121 @staticmethod
1122 def _outtmpl_expandpath(outtmpl):
1123 # expand_path translates '%%' into '%' and '$$' into '$'
1124 # correspondingly that is not what we want since we need to keep
1125 # '%%' intact for template dict substitution step. Working around
1126 # with boundary-alike separator hack.
1127 sep = ''.join(random.choices(string.ascii_letters, k=32))
1128 outtmpl = outtmpl.replace('%%', f'%{sep}%').replace('$$', f'${sep}$')
1129
1130 # outtmpl should be expand_path'ed before template dict substitution
1131 # because meta fields may contain env variables we don't want to
1132 # be expanded. E.g. for outtmpl "%(title)s.%(ext)s" and
1133 # title "Hello $PATH", we don't want `$PATH` to be expanded.
1134 return expand_path(outtmpl).replace(sep, '')
1135
1136 @staticmethod
1137 def escape_outtmpl(outtmpl):
1138 ''' Escape any remaining strings like %s, %abc% etc. '''
1139 return re.sub(
1140 STR_FORMAT_RE_TMPL.format('', '(?![%(\0])'),
1141 lambda mobj: ('' if mobj.group('has_key') else '%') + mobj.group(0),
1142 outtmpl)
1143
1144 @classmethod
1145 def validate_outtmpl(cls, outtmpl):
1146 ''' @return None or Exception object '''
1147 outtmpl = re.sub(
1148 STR_FORMAT_RE_TMPL.format('[^)]*', '[ljhqBUDS]'),
1149 lambda mobj: f'{mobj.group(0)[:-1]}s',
1150 cls._outtmpl_expandpath(outtmpl))
1151 try:
1152 cls.escape_outtmpl(outtmpl) % collections.defaultdict(int)
1153 return None
1154 except ValueError as err:
1155 return err
1156
1157 @staticmethod
1158 def _copy_infodict(info_dict):
1159 info_dict = dict(info_dict)
1160 info_dict.pop('__postprocessors', None)
1161 info_dict.pop('__pending_error', None)
1162 return info_dict
1163
1164 def prepare_outtmpl(self, outtmpl, info_dict, sanitize=False):
1165 """ Make the outtmpl and info_dict suitable for substitution: ydl.escape_outtmpl(outtmpl) % info_dict
1166 @param sanitize Whether to sanitize the output as a filename.
1167 For backward compatibility, a function can also be passed
1168 """
1169
1170 info_dict.setdefault('epoch', int(time.time())) # keep epoch consistent once set
1171
1172 info_dict = self._copy_infodict(info_dict)
1173 info_dict['duration_string'] = ( # %(duration>%H-%M-%S)s is wrong if duration > 24hrs
1174 formatSeconds(info_dict['duration'], '-' if sanitize else ':')
1175 if info_dict.get('duration', None) is not None
1176 else None)
1177 info_dict['autonumber'] = int(self.params.get('autonumber_start', 1) - 1 + self._num_downloads)
1178 info_dict['video_autonumber'] = self._num_videos
1179 if info_dict.get('resolution') is None:
1180 info_dict['resolution'] = self.format_resolution(info_dict, default=None)
1181
1182 # For fields playlist_index, playlist_autonumber and autonumber convert all occurrences
1183 # of %(field)s to %(field)0Nd for backward compatibility
1184 field_size_compat_map = {
1185 'playlist_index': number_of_digits(info_dict.get('__last_playlist_index') or 0),
1186 'playlist_autonumber': number_of_digits(info_dict.get('n_entries') or 0),
1187 'autonumber': self.params.get('autonumber_size') or 5,
1188 }
1189
1190 TMPL_DICT = {}
1191 EXTERNAL_FORMAT_RE = re.compile(STR_FORMAT_RE_TMPL.format('[^)]*', f'[{STR_FORMAT_TYPES}ljhqBUDS]'))
1192 MATH_FUNCTIONS = {
1193 '+': float.__add__,
1194 '-': float.__sub__,
1195 '*': float.__mul__,
1196 }
1197 # Field is of the form key1.key2...
1198 # where keys (except first) can be string, int, slice or "{field, ...}"
1199 FIELD_INNER_RE = r'(?:\w+|%(num)s|%(num)s?(?::%(num)s?){1,2})' % {'num': r'(?:-?\d+)'}
1200 FIELD_RE = r'\w*(?:\.(?:%(inner)s|{%(field)s(?:,%(field)s)*}))*' % {
1201 'inner': FIELD_INNER_RE,
1202 'field': rf'\w*(?:\.{FIELD_INNER_RE})*'
1203 }
1204 MATH_FIELD_RE = rf'(?:{FIELD_RE}|-?{NUMBER_RE})'
1205 MATH_OPERATORS_RE = r'(?:%s)' % '|'.join(map(re.escape, MATH_FUNCTIONS.keys()))
1206 INTERNAL_FORMAT_RE = re.compile(rf'''(?xs)
1207 (?P<negate>-)?
1208 (?P<fields>{FIELD_RE})
1209 (?P<maths>(?:{MATH_OPERATORS_RE}{MATH_FIELD_RE})*)
1210 (?:>(?P<strf_format>.+?))?
1211 (?P<remaining>
1212 (?P<alternate>(?<!\\),[^|&)]+)?
1213 (?:&(?P<replacement>.*?))?
1214 (?:\|(?P<default>.*?))?
1215 )$''')
1216
1217 def _from_user_input(field):
1218 if field == ':':
1219 return ...
1220 elif ':' in field:
1221 return slice(*map(int_or_none, field.split(':')))
1222 elif int_or_none(field) is not None:
1223 return int(field)
1224 return field
1225
1226 def _traverse_infodict(fields):
1227 fields = [f for x in re.split(r'\.({.+?})\.?', fields)
1228 for f in ([x] if x.startswith('{') else x.split('.'))]
1229 for i in (0, -1):
1230 if fields and not fields[i]:
1231 fields.pop(i)
1232
1233 for i, f in enumerate(fields):
1234 if not f.startswith('{'):
1235 fields[i] = _from_user_input(f)
1236 continue
1237 assert f.endswith('}'), f'No closing brace for {f} in {fields}'
1238 fields[i] = {k: list(map(_from_user_input, k.split('.'))) for k in f[1:-1].split(',')}
1239
1240 return traverse_obj(info_dict, fields, traverse_string=True)
1241
1242 def get_value(mdict):
1243 # Object traversal
1244 value = _traverse_infodict(mdict['fields'])
1245 # Negative
1246 if mdict['negate']:
1247 value = float_or_none(value)
1248 if value is not None:
1249 value *= -1
1250 # Do maths
1251 offset_key = mdict['maths']
1252 if offset_key:
1253 value = float_or_none(value)
1254 operator = None
1255 while offset_key:
1256 item = re.match(
1257 MATH_FIELD_RE if operator else MATH_OPERATORS_RE,
1258 offset_key).group(0)
1259 offset_key = offset_key[len(item):]
1260 if operator is None:
1261 operator = MATH_FUNCTIONS[item]
1262 continue
1263 item, multiplier = (item[1:], -1) if item[0] == '-' else (item, 1)
1264 offset = float_or_none(item)
1265 if offset is None:
1266 offset = float_or_none(_traverse_infodict(item))
1267 try:
1268 value = operator(value, multiplier * offset)
1269 except (TypeError, ZeroDivisionError):
1270 return None
1271 operator = None
1272 # Datetime formatting
1273 if mdict['strf_format']:
1274 value = strftime_or_none(value, mdict['strf_format'].replace('\\,', ','))
1275
1276 # XXX: Workaround for https://github.com/yt-dlp/yt-dlp/issues/4485
1277 if sanitize and value == '':
1278 value = None
1279 return value
1280
1281 na = self.params.get('outtmpl_na_placeholder', 'NA')
1282
1283 def filename_sanitizer(key, value, restricted=self.params.get('restrictfilenames')):
1284 return sanitize_filename(str(value), restricted=restricted, is_id=(
1285 bool(re.search(r'(^|[_.])id(\.|$)', key))
1286 if 'filename-sanitization' in self.params['compat_opts']
1287 else NO_DEFAULT))
1288
1289 sanitizer = sanitize if callable(sanitize) else filename_sanitizer
1290 sanitize = bool(sanitize)
1291
1292 def _dumpjson_default(obj):
1293 if isinstance(obj, (set, LazyList)):
1294 return list(obj)
1295 return repr(obj)
1296
1297 class _ReplacementFormatter(string.Formatter):
1298 def get_field(self, field_name, args, kwargs):
1299 if field_name.isdigit():
1300 return args[0], -1
1301 raise ValueError('Unsupported field')
1302
1303 replacement_formatter = _ReplacementFormatter()
1304
1305 def create_key(outer_mobj):
1306 if not outer_mobj.group('has_key'):
1307 return outer_mobj.group(0)
1308 key = outer_mobj.group('key')
1309 mobj = re.match(INTERNAL_FORMAT_RE, key)
1310 value, replacement, default, last_field = None, None, na, ''
1311 while mobj:
1312 mobj = mobj.groupdict()
1313 default = mobj['default'] if mobj['default'] is not None else default
1314 value = get_value(mobj)
1315 last_field, replacement = mobj['fields'], mobj['replacement']
1316 if value is None and mobj['alternate']:
1317 mobj = re.match(INTERNAL_FORMAT_RE, mobj['remaining'][1:])
1318 else:
1319 break
1320
1321 if None not in (value, replacement):
1322 try:
1323 value = replacement_formatter.format(replacement, value)
1324 except ValueError:
1325 value, default = None, na
1326
1327 fmt = outer_mobj.group('format')
1328 if fmt == 's' and last_field in field_size_compat_map.keys() and isinstance(value, int):
1329 fmt = f'0{field_size_compat_map[last_field]:d}d'
1330
1331 flags = outer_mobj.group('conversion') or ''
1332 str_fmt = f'{fmt[:-1]}s'
1333 if value is None:
1334 value, fmt = default, 's'
1335 elif fmt[-1] == 'l': # list
1336 delim = '\n' if '#' in flags else ', '
1337 value, fmt = delim.join(map(str, variadic(value, allowed_types=(str, bytes)))), str_fmt
1338 elif fmt[-1] == 'j': # json
1339 value, fmt = json.dumps(
1340 value, default=_dumpjson_default,
1341 indent=4 if '#' in flags else None, ensure_ascii='+' not in flags), str_fmt
1342 elif fmt[-1] == 'h': # html
1343 value, fmt = escapeHTML(str(value)), str_fmt
1344 elif fmt[-1] == 'q': # quoted
1345 value = map(str, variadic(value) if '#' in flags else [value])
1346 value, fmt = ' '.join(map(compat_shlex_quote, value)), str_fmt
1347 elif fmt[-1] == 'B': # bytes
1348 value = f'%{str_fmt}'.encode() % str(value).encode()
1349 value, fmt = value.decode('utf-8', 'ignore'), 's'
1350 elif fmt[-1] == 'U': # unicode normalized
1351 value, fmt = unicodedata.normalize(
1352 # "+" = compatibility equivalence, "#" = NFD
1353 'NF%s%s' % ('K' if '+' in flags else '', 'D' if '#' in flags else 'C'),
1354 value), str_fmt
1355 elif fmt[-1] == 'D': # decimal suffix
1356 num_fmt, fmt = fmt[:-1].replace('#', ''), 's'
1357 value = format_decimal_suffix(value, f'%{num_fmt}f%s' if num_fmt else '%d%s',
1358 factor=1024 if '#' in flags else 1000)
1359 elif fmt[-1] == 'S': # filename sanitization
1360 value, fmt = filename_sanitizer(last_field, value, restricted='#' in flags), str_fmt
1361 elif fmt[-1] == 'c':
1362 if value:
1363 value = str(value)[0]
1364 else:
1365 fmt = str_fmt
1366 elif fmt[-1] not in 'rsa': # numeric
1367 value = float_or_none(value)
1368 if value is None:
1369 value, fmt = default, 's'
1370
1371 if sanitize:
1372 # If value is an object, sanitize might convert it to a string
1373 # So we convert it to repr first
1374 if fmt[-1] == 'r':
1375 value, fmt = repr(value), str_fmt
1376 elif fmt[-1] == 'a':
1377 value, fmt = ascii(value), str_fmt
1378 if fmt[-1] in 'csra':
1379 value = sanitizer(last_field, value)
1380
1381 key = '%s\0%s' % (key.replace('%', '%\0'), outer_mobj.group('format'))
1382 TMPL_DICT[key] = value
1383 return '{prefix}%({key}){fmt}'.format(key=key, fmt=fmt, prefix=outer_mobj.group('prefix'))
1384
1385 return EXTERNAL_FORMAT_RE.sub(create_key, outtmpl), TMPL_DICT
1386
1387 def evaluate_outtmpl(self, outtmpl, info_dict, *args, **kwargs):
1388 outtmpl, info_dict = self.prepare_outtmpl(outtmpl, info_dict, *args, **kwargs)
1389 return self.escape_outtmpl(outtmpl) % info_dict
1390
1391 def _prepare_filename(self, info_dict, *, outtmpl=None, tmpl_type=None):
1392 assert None in (outtmpl, tmpl_type), 'outtmpl and tmpl_type are mutually exclusive'
1393 if outtmpl is None:
1394 outtmpl = self.params['outtmpl'].get(tmpl_type or 'default', self.params['outtmpl']['default'])
1395 try:
1396 outtmpl = self._outtmpl_expandpath(outtmpl)
1397 filename = self.evaluate_outtmpl(outtmpl, info_dict, True)
1398 if not filename:
1399 return None
1400
1401 if tmpl_type in ('', 'temp'):
1402 final_ext, ext = self.params.get('final_ext'), info_dict.get('ext')
1403 if final_ext and ext and final_ext != ext and filename.endswith(f'.{final_ext}'):
1404 filename = replace_extension(filename, ext, final_ext)
1405 elif tmpl_type:
1406 force_ext = OUTTMPL_TYPES[tmpl_type]
1407 if force_ext:
1408 filename = replace_extension(filename, force_ext, info_dict.get('ext'))
1409
1410 # https://github.com/blackjack4494/youtube-dlc/issues/85
1411 trim_file_name = self.params.get('trim_file_name', False)
1412 if trim_file_name:
1413 no_ext, *ext = filename.rsplit('.', 2)
1414 filename = join_nonempty(no_ext[:trim_file_name], *ext, delim='.')
1415
1416 return filename
1417 except ValueError as err:
1418 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
1419 return None
1420
1421 def prepare_filename(self, info_dict, dir_type='', *, outtmpl=None, warn=False):
1422 """Generate the output filename"""
1423 if outtmpl:
1424 assert not dir_type, 'outtmpl and dir_type are mutually exclusive'
1425 dir_type = None
1426 filename = self._prepare_filename(info_dict, tmpl_type=dir_type, outtmpl=outtmpl)
1427 if not filename and dir_type not in ('', 'temp'):
1428 return ''
1429
1430 if warn:
1431 if not self.params.get('paths'):
1432 pass
1433 elif filename == '-':
1434 self.report_warning('--paths is ignored when an outputting to stdout', only_once=True)
1435 elif os.path.isabs(filename):
1436 self.report_warning('--paths is ignored since an absolute path is given in output template', only_once=True)
1437 if filename == '-' or not filename:
1438 return filename
1439
1440 return self.get_output_path(dir_type, filename)
1441
1442 def _match_entry(self, info_dict, incomplete=False, silent=False):
1443 """Returns None if the file should be downloaded"""
1444 _type = 'video' if 'playlist-match-filter' in self.params['compat_opts'] else info_dict.get('_type', 'video')
1445 assert incomplete or _type == 'video', 'Only video result can be considered complete'
1446
1447 video_title = info_dict.get('title', info_dict.get('id', 'entry'))
1448
1449 def check_filter():
1450 if _type in ('playlist', 'multi_video'):
1451 return
1452 elif _type in ('url', 'url_transparent') and not try_call(
1453 lambda: self.get_info_extractor(info_dict['ie_key']).is_single_video(info_dict['url'])):
1454 return
1455
1456 if 'title' in info_dict:
1457 # This can happen when we're just evaluating the playlist
1458 title = info_dict['title']
1459 matchtitle = self.params.get('matchtitle', False)
1460 if matchtitle:
1461 if not re.search(matchtitle, title, re.IGNORECASE):
1462 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
1463 rejecttitle = self.params.get('rejecttitle', False)
1464 if rejecttitle:
1465 if re.search(rejecttitle, title, re.IGNORECASE):
1466 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
1467
1468 date = info_dict.get('upload_date')
1469 if date is not None:
1470 dateRange = self.params.get('daterange', DateRange())
1471 if date not in dateRange:
1472 return f'{date_from_str(date).isoformat()} upload date is not in range {dateRange}'
1473 view_count = info_dict.get('view_count')
1474 if view_count is not None:
1475 min_views = self.params.get('min_views')
1476 if min_views is not None and view_count < min_views:
1477 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
1478 max_views = self.params.get('max_views')
1479 if max_views is not None and view_count > max_views:
1480 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
1481 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
1482 return 'Skipping "%s" because it is age restricted' % video_title
1483
1484 match_filter = self.params.get('match_filter')
1485 if match_filter is None:
1486 return None
1487
1488 cancelled = None
1489 try:
1490 try:
1491 ret = match_filter(info_dict, incomplete=incomplete)
1492 except TypeError:
1493 # For backward compatibility
1494 ret = None if incomplete else match_filter(info_dict)
1495 except DownloadCancelled as err:
1496 if err.msg is not NO_DEFAULT:
1497 raise
1498 ret, cancelled = err.msg, err
1499
1500 if ret is NO_DEFAULT:
1501 while True:
1502 filename = self._format_screen(self.prepare_filename(info_dict), self.Styles.FILENAME)
1503 reply = input(self._format_screen(
1504 f'Download "{filename}"? (Y/n): ', self.Styles.EMPHASIS)).lower().strip()
1505 if reply in {'y', ''}:
1506 return None
1507 elif reply == 'n':
1508 if cancelled:
1509 raise type(cancelled)(f'Skipping {video_title}')
1510 return f'Skipping {video_title}'
1511 return ret
1512
1513 if self.in_download_archive(info_dict):
1514 reason = ''.join((
1515 format_field(info_dict, 'id', f'{self._format_screen("%s", self.Styles.ID)}: '),
1516 format_field(info_dict, 'title', f'{self._format_screen("%s", self.Styles.EMPHASIS)} '),
1517 'has already been recorded in the archive'))
1518 break_opt, break_err = 'break_on_existing', ExistingVideoReached
1519 else:
1520 try:
1521 reason = check_filter()
1522 except DownloadCancelled as e:
1523 reason, break_opt, break_err = e.msg, 'match_filter', type(e)
1524 else:
1525 break_opt, break_err = 'break_on_reject', RejectedVideoReached
1526 if reason is not None:
1527 if not silent:
1528 self.to_screen('[download] ' + reason)
1529 if self.params.get(break_opt, False):
1530 raise break_err()
1531 return reason
1532
1533 @staticmethod
1534 def add_extra_info(info_dict, extra_info):
1535 '''Set the keys from extra_info in info dict if they are missing'''
1536 for key, value in extra_info.items():
1537 info_dict.setdefault(key, value)
1538
1539 def extract_info(self, url, download=True, ie_key=None, extra_info=None,
1540 process=True, force_generic_extractor=False):
1541 """
1542 Extract and return the information dictionary of the URL
1543
1544 Arguments:
1545 @param url URL to extract
1546
1547 Keyword arguments:
1548 @param download Whether to download videos
1549 @param process Whether to resolve all unresolved references (URLs, playlist items).
1550 Must be True for download to work
1551 @param ie_key Use only the extractor with this key
1552
1553 @param extra_info Dictionary containing the extra values to add to the info (For internal use only)
1554 @force_generic_extractor Force using the generic extractor (Deprecated; use ie_key='Generic')
1555 """
1556
1557 if extra_info is None:
1558 extra_info = {}
1559
1560 if not ie_key and force_generic_extractor:
1561 ie_key = 'Generic'
1562
1563 if ie_key:
1564 ies = {ie_key: self._ies[ie_key]} if ie_key in self._ies else {}
1565 else:
1566 ies = self._ies
1567
1568 for key, ie in ies.items():
1569 if not ie.suitable(url):
1570 continue
1571
1572 if not ie.working():
1573 self.report_warning('The program functionality for this site has been marked as broken, '
1574 'and will probably not work.')
1575
1576 temp_id = ie.get_temp_id(url)
1577 if temp_id is not None and self.in_download_archive({'id': temp_id, 'ie_key': key}):
1578 self.to_screen(f'[download] {self._format_screen(temp_id, self.Styles.ID)}: '
1579 'has already been recorded in the archive')
1580 if self.params.get('break_on_existing', False):
1581 raise ExistingVideoReached()
1582 break
1583 return self.__extract_info(url, self.get_info_extractor(key), download, extra_info, process)
1584 else:
1585 extractors_restricted = self.params.get('allowed_extractors') not in (None, ['default'])
1586 self.report_error(f'No suitable extractor{format_field(ie_key, None, " (%s)")} found for URL {url}',
1587 tb=False if extractors_restricted else None)
1588
1589 def _handle_extraction_exceptions(func):
1590 @functools.wraps(func)
1591 def wrapper(self, *args, **kwargs):
1592 while True:
1593 try:
1594 return func(self, *args, **kwargs)
1595 except (DownloadCancelled, LazyList.IndexError, PagedList.IndexError):
1596 raise
1597 except ReExtractInfo as e:
1598 if e.expected:
1599 self.to_screen(f'{e}; Re-extracting data')
1600 else:
1601 self.to_stderr('\r')
1602 self.report_warning(f'{e}; Re-extracting data')
1603 continue
1604 except GeoRestrictedError as e:
1605 msg = e.msg
1606 if e.countries:
1607 msg += '\nThis video is available in %s.' % ', '.join(
1608 map(ISO3166Utils.short2full, e.countries))
1609 msg += '\nYou might want to use a VPN or a proxy server (with --proxy) to workaround.'
1610 self.report_error(msg)
1611 except ExtractorError as e: # An error we somewhat expected
1612 self.report_error(str(e), e.format_traceback())
1613 except Exception as e:
1614 if self.params.get('ignoreerrors'):
1615 self.report_error(str(e), tb=encode_compat_str(traceback.format_exc()))
1616 else:
1617 raise
1618 break
1619 return wrapper
1620
1621 def _wait_for_video(self, ie_result={}):
1622 if (not self.params.get('wait_for_video')
1623 or ie_result.get('_type', 'video') != 'video'
1624 or ie_result.get('formats') or ie_result.get('url')):
1625 return
1626
1627 format_dur = lambda dur: '%02d:%02d:%02d' % timetuple_from_msec(dur * 1000)[:-1]
1628 last_msg = ''
1629
1630 def progress(msg):
1631 nonlocal last_msg
1632 full_msg = f'{msg}\n'
1633 if not self.params.get('noprogress'):
1634 full_msg = msg + ' ' * (len(last_msg) - len(msg)) + '\r'
1635 elif last_msg:
1636 return
1637 self.to_screen(full_msg, skip_eol=True)
1638 last_msg = msg
1639
1640 min_wait, max_wait = self.params.get('wait_for_video')
1641 diff = try_get(ie_result, lambda x: x['release_timestamp'] - time.time())
1642 if diff is None and ie_result.get('live_status') == 'is_upcoming':
1643 diff = round(random.uniform(min_wait, max_wait) if (max_wait and min_wait) else (max_wait or min_wait), 0)
1644 self.report_warning('Release time of video is not known')
1645 elif ie_result and (diff or 0) <= 0:
1646 self.report_warning('Video should already be available according to extracted info')
1647 diff = min(max(diff or 0, min_wait or 0), max_wait or float('inf'))
1648 self.to_screen(f'[wait] Waiting for {format_dur(diff)} - Press Ctrl+C to try now')
1649
1650 wait_till = time.time() + diff
1651 try:
1652 while True:
1653 diff = wait_till - time.time()
1654 if diff <= 0:
1655 progress('')
1656 raise ReExtractInfo('[wait] Wait period ended', expected=True)
1657 progress(f'[wait] Remaining time until next attempt: {self._format_screen(format_dur(diff), self.Styles.EMPHASIS)}')
1658 time.sleep(1)
1659 except KeyboardInterrupt:
1660 progress('')
1661 raise ReExtractInfo('[wait] Interrupted by user', expected=True)
1662 except BaseException as e:
1663 if not isinstance(e, ReExtractInfo):
1664 self.to_screen('')
1665 raise
1666
1667 def _load_cookies(self, data, *, autoscope=True):
1668 """Loads cookies from a `Cookie` header
1669
1670 This tries to work around the security vulnerability of passing cookies to every domain.
1671 See: https://github.com/yt-dlp/yt-dlp/security/advisories/GHSA-v8mc-9377-rwjj
1672
1673 @param data The Cookie header as string to load the cookies from
1674 @param autoscope If `False`, scope cookies using Set-Cookie syntax and error for cookie without domains
1675 If `True`, save cookies for later to be stored in the jar with a limited scope
1676 If a URL, save cookies in the jar with the domain of the URL
1677 """
1678 for cookie in LenientSimpleCookie(data).values():
1679 if autoscope and any(cookie.values()):
1680 raise ValueError('Invalid syntax in Cookie Header')
1681
1682 domain = cookie.get('domain') or ''
1683 expiry = cookie.get('expires')
1684 if expiry == '': # 0 is valid
1685 expiry = None
1686 prepared_cookie = http.cookiejar.Cookie(
1687 cookie.get('version') or 0, cookie.key, cookie.value, None, False,
1688 domain, True, True, cookie.get('path') or '', bool(cookie.get('path')),
1689 cookie.get('secure') or False, expiry, False, None, None, {})
1690
1691 if domain:
1692 self.cookiejar.set_cookie(prepared_cookie)
1693 elif autoscope is True:
1694 self.deprecated_feature(
1695 'Passing cookies as a header is a potential security risk; '
1696 'they will be scoped to the domain of the downloaded urls. '
1697 'Please consider loading cookies from a file or browser instead.')
1698 self.__header_cookies.append(prepared_cookie)
1699 elif autoscope:
1700 self.report_warning(
1701 'The extractor result contains an unscoped cookie as an HTTP header. '
1702 f'If you are using yt-dlp with an input URL{bug_reports_message(before=",")}',
1703 only_once=True)
1704 self._apply_header_cookies(autoscope, [prepared_cookie])
1705 else:
1706 self.report_error('Unscoped cookies are not allowed; please specify some sort of scoping',
1707 tb=False, is_error=False)
1708
1709 def _apply_header_cookies(self, url, cookies=None):
1710 """Applies stray header cookies to the provided url
1711
1712 This loads header cookies and scopes them to the domain provided in `url`.
1713 While this is not ideal, it helps reduce the risk of them being sent
1714 to an unintended destination while mostly maintaining compatibility.
1715 """
1716 parsed = urllib.parse.urlparse(url)
1717 if not parsed.hostname:
1718 return
1719
1720 for cookie in map(copy.copy, cookies or self.__header_cookies):
1721 cookie.domain = f'.{parsed.hostname}'
1722 self.cookiejar.set_cookie(cookie)
1723
1724 @_handle_extraction_exceptions
1725 def __extract_info(self, url, ie, download, extra_info, process):
1726 self._apply_header_cookies(url)
1727
1728 try:
1729 ie_result = ie.extract(url)
1730 except UserNotLive as e:
1731 if process:
1732 if self.params.get('wait_for_video'):
1733 self.report_warning(e)
1734 self._wait_for_video()
1735 raise
1736 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
1737 self.report_warning(f'Extractor {ie.IE_NAME} returned nothing{bug_reports_message()}')
1738 return
1739 if isinstance(ie_result, list):
1740 # Backwards compatibility: old IE result format
1741 ie_result = {
1742 '_type': 'compat_list',
1743 'entries': ie_result,
1744 }
1745 if extra_info.get('original_url'):
1746 ie_result.setdefault('original_url', extra_info['original_url'])
1747 self.add_default_extra_info(ie_result, ie, url)
1748 if process:
1749 self._wait_for_video(ie_result)
1750 return self.process_ie_result(ie_result, download, extra_info)
1751 else:
1752 return ie_result
1753
1754 def add_default_extra_info(self, ie_result, ie, url):
1755 if url is not None:
1756 self.add_extra_info(ie_result, {
1757 'webpage_url': url,
1758 'original_url': url,
1759 })
1760 webpage_url = ie_result.get('webpage_url')
1761 if webpage_url:
1762 self.add_extra_info(ie_result, {
1763 'webpage_url_basename': url_basename(webpage_url),
1764 'webpage_url_domain': get_domain(webpage_url),
1765 })
1766 if ie is not None:
1767 self.add_extra_info(ie_result, {
1768 'extractor': ie.IE_NAME,
1769 'extractor_key': ie.ie_key(),
1770 })
1771
1772 def process_ie_result(self, ie_result, download=True, extra_info=None):
1773 """
1774 Take the result of the ie(may be modified) and resolve all unresolved
1775 references (URLs, playlist items).
1776
1777 It will also download the videos if 'download'.
1778 Returns the resolved ie_result.
1779 """
1780 if extra_info is None:
1781 extra_info = {}
1782 result_type = ie_result.get('_type', 'video')
1783
1784 if result_type in ('url', 'url_transparent'):
1785 ie_result['url'] = sanitize_url(
1786 ie_result['url'], scheme='http' if self.params.get('prefer_insecure') else 'https')
1787 if ie_result.get('original_url') and not extra_info.get('original_url'):
1788 extra_info = {'original_url': ie_result['original_url'], **extra_info}
1789
1790 extract_flat = self.params.get('extract_flat', False)
1791 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info)
1792 or extract_flat is True):
1793 info_copy = ie_result.copy()
1794 ie = try_get(ie_result.get('ie_key'), self.get_info_extractor)
1795 if ie and not ie_result.get('id'):
1796 info_copy['id'] = ie.get_temp_id(ie_result['url'])
1797 self.add_default_extra_info(info_copy, ie, ie_result['url'])
1798 self.add_extra_info(info_copy, extra_info)
1799 info_copy, _ = self.pre_process(info_copy)
1800 self._fill_common_fields(info_copy, False)
1801 self.__forced_printings(info_copy)
1802 self._raise_pending_errors(info_copy)
1803 if self.params.get('force_write_download_archive', False):
1804 self.record_download_archive(info_copy)
1805 return ie_result
1806
1807 if result_type == 'video':
1808 self.add_extra_info(ie_result, extra_info)
1809 ie_result = self.process_video_result(ie_result, download=download)
1810 self._raise_pending_errors(ie_result)
1811 additional_urls = (ie_result or {}).get('additional_urls')
1812 if additional_urls:
1813 # TODO: Improve MetadataParserPP to allow setting a list
1814 if isinstance(additional_urls, str):
1815 additional_urls = [additional_urls]
1816 self.to_screen(
1817 '[info] %s: %d additional URL(s) requested' % (ie_result['id'], len(additional_urls)))
1818 self.write_debug('Additional URLs: "%s"' % '", "'.join(additional_urls))
1819 ie_result['additional_entries'] = [
1820 self.extract_info(
1821 url, download, extra_info=extra_info,
1822 force_generic_extractor=self.params.get('force_generic_extractor'))
1823 for url in additional_urls
1824 ]
1825 return ie_result
1826 elif result_type == 'url':
1827 # We have to add extra_info to the results because it may be
1828 # contained in a playlist
1829 return self.extract_info(
1830 ie_result['url'], download,
1831 ie_key=ie_result.get('ie_key'),
1832 extra_info=extra_info)
1833 elif result_type == 'url_transparent':
1834 # Use the information from the embedding page
1835 info = self.extract_info(
1836 ie_result['url'], ie_key=ie_result.get('ie_key'),
1837 extra_info=extra_info, download=False, process=False)
1838
1839 # extract_info may return None when ignoreerrors is enabled and
1840 # extraction failed with an error, don't crash and return early
1841 # in this case
1842 if not info:
1843 return info
1844
1845 exempted_fields = {'_type', 'url', 'ie_key'}
1846 if not ie_result.get('section_end') and ie_result.get('section_start') is None:
1847 # For video clips, the id etc of the clip extractor should be used
1848 exempted_fields |= {'id', 'extractor', 'extractor_key'}
1849
1850 new_result = info.copy()
1851 new_result.update(filter_dict(ie_result, lambda k, v: v is not None and k not in exempted_fields))
1852
1853 # Extracted info may not be a video result (i.e.
1854 # info.get('_type', 'video') != video) but rather an url or
1855 # url_transparent. In such cases outer metadata (from ie_result)
1856 # should be propagated to inner one (info). For this to happen
1857 # _type of info should be overridden with url_transparent. This
1858 # fixes issue from https://github.com/ytdl-org/youtube-dl/pull/11163.
1859 if new_result.get('_type') == 'url':
1860 new_result['_type'] = 'url_transparent'
1861
1862 return self.process_ie_result(
1863 new_result, download=download, extra_info=extra_info)
1864 elif result_type in ('playlist', 'multi_video'):
1865 # Protect from infinite recursion due to recursively nested playlists
1866 # (see https://github.com/ytdl-org/youtube-dl/issues/27833)
1867 webpage_url = ie_result.get('webpage_url') # Playlists maynot have webpage_url
1868 if webpage_url and webpage_url in self._playlist_urls:
1869 self.to_screen(
1870 '[download] Skipping already downloaded playlist: %s'
1871 % ie_result.get('title') or ie_result.get('id'))
1872 return
1873
1874 self._playlist_level += 1
1875 self._playlist_urls.add(webpage_url)
1876 self._fill_common_fields(ie_result, False)
1877 self._sanitize_thumbnails(ie_result)
1878 try:
1879 return self.__process_playlist(ie_result, download)
1880 finally:
1881 self._playlist_level -= 1
1882 if not self._playlist_level:
1883 self._playlist_urls.clear()
1884 elif result_type == 'compat_list':
1885 self.report_warning(
1886 'Extractor %s returned a compat_list result. '
1887 'It needs to be updated.' % ie_result.get('extractor'))
1888
1889 def _fixup(r):
1890 self.add_extra_info(r, {
1891 'extractor': ie_result['extractor'],
1892 'webpage_url': ie_result['webpage_url'],
1893 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1894 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1895 'extractor_key': ie_result['extractor_key'],
1896 })
1897 return r
1898 ie_result['entries'] = [
1899 self.process_ie_result(_fixup(r), download, extra_info)
1900 for r in ie_result['entries']
1901 ]
1902 return ie_result
1903 else:
1904 raise Exception('Invalid result type: %s' % result_type)
1905
1906 def _ensure_dir_exists(self, path):
1907 return make_dir(path, self.report_error)
1908
1909 @staticmethod
1910 def _playlist_infodict(ie_result, strict=False, **kwargs):
1911 info = {
1912 'playlist_count': ie_result.get('playlist_count'),
1913 'playlist': ie_result.get('title') or ie_result.get('id'),
1914 'playlist_id': ie_result.get('id'),
1915 'playlist_title': ie_result.get('title'),
1916 'playlist_uploader': ie_result.get('uploader'),
1917 'playlist_uploader_id': ie_result.get('uploader_id'),
1918 **kwargs,
1919 }
1920 if strict:
1921 return info
1922 if ie_result.get('webpage_url'):
1923 info.update({
1924 'webpage_url': ie_result['webpage_url'],
1925 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1926 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1927 })
1928 return {
1929 **info,
1930 'playlist_index': 0,
1931 '__last_playlist_index': max(ie_result.get('requested_entries') or (0, 0)),
1932 'extractor': ie_result['extractor'],
1933 'extractor_key': ie_result['extractor_key'],
1934 }
1935
1936 def __process_playlist(self, ie_result, download):
1937 """Process each entry in the playlist"""
1938 assert ie_result['_type'] in ('playlist', 'multi_video')
1939
1940 common_info = self._playlist_infodict(ie_result, strict=True)
1941 title = common_info.get('playlist') or '<Untitled>'
1942 if self._match_entry(common_info, incomplete=True) is not None:
1943 return
1944 self.to_screen(f'[download] Downloading {ie_result["_type"]}: {title}')
1945
1946 all_entries = PlaylistEntries(self, ie_result)
1947 entries = orderedSet(all_entries.get_requested_items(), lazy=True)
1948
1949 lazy = self.params.get('lazy_playlist')
1950 if lazy:
1951 resolved_entries, n_entries = [], 'N/A'
1952 ie_result['requested_entries'], ie_result['entries'] = None, None
1953 else:
1954 entries = resolved_entries = list(entries)
1955 n_entries = len(resolved_entries)
1956 ie_result['requested_entries'], ie_result['entries'] = tuple(zip(*resolved_entries)) or ([], [])
1957 if not ie_result.get('playlist_count'):
1958 # Better to do this after potentially exhausting entries
1959 ie_result['playlist_count'] = all_entries.get_full_count()
1960
1961 extra = self._playlist_infodict(ie_result, n_entries=int_or_none(n_entries))
1962 ie_copy = collections.ChainMap(ie_result, extra)
1963
1964 _infojson_written = False
1965 write_playlist_files = self.params.get('allow_playlist_files', True)
1966 if write_playlist_files and self.params.get('list_thumbnails'):
1967 self.list_thumbnails(ie_result)
1968 if write_playlist_files and not self.params.get('simulate'):
1969 _infojson_written = self._write_info_json(
1970 'playlist', ie_result, self.prepare_filename(ie_copy, 'pl_infojson'))
1971 if _infojson_written is None:
1972 return
1973 if self._write_description('playlist', ie_result,
1974 self.prepare_filename(ie_copy, 'pl_description')) is None:
1975 return
1976 # TODO: This should be passed to ThumbnailsConvertor if necessary
1977 self._write_thumbnails('playlist', ie_result, self.prepare_filename(ie_copy, 'pl_thumbnail'))
1978
1979 if lazy:
1980 if self.params.get('playlistreverse') or self.params.get('playlistrandom'):
1981 self.report_warning('playlistreverse and playlistrandom are not supported with lazy_playlist', only_once=True)
1982 elif self.params.get('playlistreverse'):
1983 entries.reverse()
1984 elif self.params.get('playlistrandom'):
1985 random.shuffle(entries)
1986
1987 self.to_screen(f'[{ie_result["extractor"]}] Playlist {title}: Downloading {n_entries} items'
1988 f'{format_field(ie_result, "playlist_count", " of %s")}')
1989
1990 keep_resolved_entries = self.params.get('extract_flat') != 'discard'
1991 if self.params.get('extract_flat') == 'discard_in_playlist':
1992 keep_resolved_entries = ie_result['_type'] != 'playlist'
1993 if keep_resolved_entries:
1994 self.write_debug('The information of all playlist entries will be held in memory')
1995
1996 failures = 0
1997 max_failures = self.params.get('skip_playlist_after_errors') or float('inf')
1998 for i, (playlist_index, entry) in enumerate(entries):
1999 if lazy:
2000 resolved_entries.append((playlist_index, entry))
2001 if not entry:
2002 continue
2003
2004 entry['__x_forwarded_for_ip'] = ie_result.get('__x_forwarded_for_ip')
2005 if not lazy and 'playlist-index' in self.params['compat_opts']:
2006 playlist_index = ie_result['requested_entries'][i]
2007
2008 entry_copy = collections.ChainMap(entry, {
2009 **common_info,
2010 'n_entries': int_or_none(n_entries),
2011 'playlist_index': playlist_index,
2012 'playlist_autonumber': i + 1,
2013 })
2014
2015 if self._match_entry(entry_copy, incomplete=True) is not None:
2016 # For compatabilty with youtube-dl. See https://github.com/yt-dlp/yt-dlp/issues/4369
2017 resolved_entries[i] = (playlist_index, NO_DEFAULT)
2018 continue
2019
2020 self.to_screen('[download] Downloading item %s of %s' % (
2021 self._format_screen(i + 1, self.Styles.ID), self._format_screen(n_entries, self.Styles.EMPHASIS)))
2022
2023 entry_result = self.__process_iterable_entry(entry, download, collections.ChainMap({
2024 'playlist_index': playlist_index,
2025 'playlist_autonumber': i + 1,
2026 }, extra))
2027 if not entry_result:
2028 failures += 1
2029 if failures >= max_failures:
2030 self.report_error(
2031 f'Skipping the remaining entries in playlist "{title}" since {failures} items failed extraction')
2032 break
2033 if keep_resolved_entries:
2034 resolved_entries[i] = (playlist_index, entry_result)
2035
2036 # Update with processed data
2037 ie_result['entries'] = [e for _, e in resolved_entries if e is not NO_DEFAULT]
2038 ie_result['requested_entries'] = [i for i, e in resolved_entries if e is not NO_DEFAULT]
2039 if ie_result['requested_entries'] == try_call(lambda: list(range(1, ie_result['playlist_count'] + 1))):
2040 # Do not set for full playlist
2041 ie_result.pop('requested_entries')
2042
2043 # Write the updated info to json
2044 if _infojson_written is True and self._write_info_json(
2045 'updated playlist', ie_result,
2046 self.prepare_filename(ie_copy, 'pl_infojson'), overwrite=True) is None:
2047 return
2048
2049 ie_result = self.run_all_pps('playlist', ie_result)
2050 self.to_screen(f'[download] Finished downloading playlist: {title}')
2051 return ie_result
2052
2053 @_handle_extraction_exceptions
2054 def __process_iterable_entry(self, entry, download, extra_info):
2055 return self.process_ie_result(
2056 entry, download=download, extra_info=extra_info)
2057
2058 def _build_format_filter(self, filter_spec):
2059 " Returns a function to filter the formats according to the filter_spec "
2060
2061 OPERATORS = {
2062 '<': operator.lt,
2063 '<=': operator.le,
2064 '>': operator.gt,
2065 '>=': operator.ge,
2066 '=': operator.eq,
2067 '!=': operator.ne,
2068 }
2069 operator_rex = re.compile(r'''(?x)\s*
2070 (?P<key>[\w.-]+)\s*
2071 (?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
2072 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)\s*
2073 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
2074 m = operator_rex.fullmatch(filter_spec)
2075 if m:
2076 try:
2077 comparison_value = int(m.group('value'))
2078 except ValueError:
2079 comparison_value = parse_filesize(m.group('value'))
2080 if comparison_value is None:
2081 comparison_value = parse_filesize(m.group('value') + 'B')
2082 if comparison_value is None:
2083 raise ValueError(
2084 'Invalid value %r in format specification %r' % (
2085 m.group('value'), filter_spec))
2086 op = OPERATORS[m.group('op')]
2087
2088 if not m:
2089 STR_OPERATORS = {
2090 '=': operator.eq,
2091 '^=': lambda attr, value: attr.startswith(value),
2092 '$=': lambda attr, value: attr.endswith(value),
2093 '*=': lambda attr, value: value in attr,
2094 '~=': lambda attr, value: value.search(attr) is not None
2095 }
2096 str_operator_rex = re.compile(r'''(?x)\s*
2097 (?P<key>[a-zA-Z0-9._-]+)\s*
2098 (?P<negation>!\s*)?(?P<op>%s)\s*(?P<none_inclusive>\?\s*)?
2099 (?P<quote>["'])?
2100 (?P<value>(?(quote)(?:(?!(?P=quote))[^\\]|\\.)+|[\w.-]+))
2101 (?(quote)(?P=quote))\s*
2102 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
2103 m = str_operator_rex.fullmatch(filter_spec)
2104 if m:
2105 if m.group('op') == '~=':
2106 comparison_value = re.compile(m.group('value'))
2107 else:
2108 comparison_value = re.sub(r'''\\([\\"'])''', r'\1', m.group('value'))
2109 str_op = STR_OPERATORS[m.group('op')]
2110 if m.group('negation'):
2111 op = lambda attr, value: not str_op(attr, value)
2112 else:
2113 op = str_op
2114
2115 if not m:
2116 raise SyntaxError('Invalid filter specification %r' % filter_spec)
2117
2118 def _filter(f):
2119 actual_value = f.get(m.group('key'))
2120 if actual_value is None:
2121 return m.group('none_inclusive')
2122 return op(actual_value, comparison_value)
2123 return _filter
2124
2125 def _check_formats(self, formats):
2126 for f in formats:
2127 self.to_screen('[info] Testing format %s' % f['format_id'])
2128 path = self.get_output_path('temp')
2129 if not self._ensure_dir_exists(f'{path}/'):
2130 continue
2131 temp_file = tempfile.NamedTemporaryFile(suffix='.tmp', delete=False, dir=path or None)
2132 temp_file.close()
2133 try:
2134 success, _ = self.dl(temp_file.name, f, test=True)
2135 except (DownloadError, OSError, ValueError) + network_exceptions:
2136 success = False
2137 finally:
2138 if os.path.exists(temp_file.name):
2139 try:
2140 os.remove(temp_file.name)
2141 except OSError:
2142 self.report_warning('Unable to delete temporary file "%s"' % temp_file.name)
2143 if success:
2144 yield f
2145 else:
2146 self.to_screen('[info] Unable to download format %s. Skipping...' % f['format_id'])
2147
2148 def _default_format_spec(self, info_dict, download=True):
2149
2150 def can_merge():
2151 merger = FFmpegMergerPP(self)
2152 return merger.available and merger.can_merge()
2153
2154 prefer_best = (
2155 not self.params.get('simulate')
2156 and download
2157 and (
2158 not can_merge()
2159 or info_dict.get('is_live') and not self.params.get('live_from_start')
2160 or self.params['outtmpl']['default'] == '-'))
2161 compat = (
2162 prefer_best
2163 or self.params.get('allow_multiple_audio_streams', False)
2164 or 'format-spec' in self.params['compat_opts'])
2165
2166 return (
2167 'best/bestvideo+bestaudio' if prefer_best
2168 else 'bestvideo*+bestaudio/best' if not compat
2169 else 'bestvideo+bestaudio/best')
2170
2171 def build_format_selector(self, format_spec):
2172 def syntax_error(note, start):
2173 message = (
2174 'Invalid format specification: '
2175 '{}\n\t{}\n\t{}^'.format(note, format_spec, ' ' * start[1]))
2176 return SyntaxError(message)
2177
2178 PICKFIRST = 'PICKFIRST'
2179 MERGE = 'MERGE'
2180 SINGLE = 'SINGLE'
2181 GROUP = 'GROUP'
2182 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
2183
2184 allow_multiple_streams = {'audio': self.params.get('allow_multiple_audio_streams', False),
2185 'video': self.params.get('allow_multiple_video_streams', False)}
2186
2187 def _parse_filter(tokens):
2188 filter_parts = []
2189 for type, string_, start, _, _ in tokens:
2190 if type == tokenize.OP and string_ == ']':
2191 return ''.join(filter_parts)
2192 else:
2193 filter_parts.append(string_)
2194
2195 def _remove_unused_ops(tokens):
2196 # Remove operators that we don't use and join them with the surrounding strings.
2197 # E.g. 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
2198 ALLOWED_OPS = ('/', '+', ',', '(', ')')
2199 last_string, last_start, last_end, last_line = None, None, None, None
2200 for type, string_, start, end, line in tokens:
2201 if type == tokenize.OP and string_ == '[':
2202 if last_string:
2203 yield tokenize.NAME, last_string, last_start, last_end, last_line
2204 last_string = None
2205 yield type, string_, start, end, line
2206 # everything inside brackets will be handled by _parse_filter
2207 for type, string_, start, end, line in tokens:
2208 yield type, string_, start, end, line
2209 if type == tokenize.OP and string_ == ']':
2210 break
2211 elif type == tokenize.OP and string_ in ALLOWED_OPS:
2212 if last_string:
2213 yield tokenize.NAME, last_string, last_start, last_end, last_line
2214 last_string = None
2215 yield type, string_, start, end, line
2216 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
2217 if not last_string:
2218 last_string = string_
2219 last_start = start
2220 last_end = end
2221 else:
2222 last_string += string_
2223 if last_string:
2224 yield tokenize.NAME, last_string, last_start, last_end, last_line
2225
2226 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
2227 selectors = []
2228 current_selector = None
2229 for type, string_, start, _, _ in tokens:
2230 # ENCODING is only defined in Python 3.x
2231 if type == getattr(tokenize, 'ENCODING', None):
2232 continue
2233 elif type in [tokenize.NAME, tokenize.NUMBER]:
2234 current_selector = FormatSelector(SINGLE, string_, [])
2235 elif type == tokenize.OP:
2236 if string_ == ')':
2237 if not inside_group:
2238 # ')' will be handled by the parentheses group
2239 tokens.restore_last_token()
2240 break
2241 elif inside_merge and string_ in ['/', ',']:
2242 tokens.restore_last_token()
2243 break
2244 elif inside_choice and string_ == ',':
2245 tokens.restore_last_token()
2246 break
2247 elif string_ == ',':
2248 if not current_selector:
2249 raise syntax_error('"," must follow a format selector', start)
2250 selectors.append(current_selector)
2251 current_selector = None
2252 elif string_ == '/':
2253 if not current_selector:
2254 raise syntax_error('"/" must follow a format selector', start)
2255 first_choice = current_selector
2256 second_choice = _parse_format_selection(tokens, inside_choice=True)
2257 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
2258 elif string_ == '[':
2259 if not current_selector:
2260 current_selector = FormatSelector(SINGLE, 'best', [])
2261 format_filter = _parse_filter(tokens)
2262 current_selector.filters.append(format_filter)
2263 elif string_ == '(':
2264 if current_selector:
2265 raise syntax_error('Unexpected "("', start)
2266 group = _parse_format_selection(tokens, inside_group=True)
2267 current_selector = FormatSelector(GROUP, group, [])
2268 elif string_ == '+':
2269 if not current_selector:
2270 raise syntax_error('Unexpected "+"', start)
2271 selector_1 = current_selector
2272 selector_2 = _parse_format_selection(tokens, inside_merge=True)
2273 if not selector_2:
2274 raise syntax_error('Expected a selector', start)
2275 current_selector = FormatSelector(MERGE, (selector_1, selector_2), [])
2276 else:
2277 raise syntax_error(f'Operator not recognized: "{string_}"', start)
2278 elif type == tokenize.ENDMARKER:
2279 break
2280 if current_selector:
2281 selectors.append(current_selector)
2282 return selectors
2283
2284 def _merge(formats_pair):
2285 format_1, format_2 = formats_pair
2286
2287 formats_info = []
2288 formats_info.extend(format_1.get('requested_formats', (format_1,)))
2289 formats_info.extend(format_2.get('requested_formats', (format_2,)))
2290
2291 if not allow_multiple_streams['video'] or not allow_multiple_streams['audio']:
2292 get_no_more = {'video': False, 'audio': False}
2293 for (i, fmt_info) in enumerate(formats_info):
2294 if fmt_info.get('acodec') == fmt_info.get('vcodec') == 'none':
2295 formats_info.pop(i)
2296 continue
2297 for aud_vid in ['audio', 'video']:
2298 if not allow_multiple_streams[aud_vid] and fmt_info.get(aud_vid[0] + 'codec') != 'none':
2299 if get_no_more[aud_vid]:
2300 formats_info.pop(i)
2301 break
2302 get_no_more[aud_vid] = True
2303
2304 if len(formats_info) == 1:
2305 return formats_info[0]
2306
2307 video_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('vcodec') != 'none']
2308 audio_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('acodec') != 'none']
2309
2310 the_only_video = video_fmts[0] if len(video_fmts) == 1 else None
2311 the_only_audio = audio_fmts[0] if len(audio_fmts) == 1 else None
2312
2313 output_ext = get_compatible_ext(
2314 vcodecs=[f.get('vcodec') for f in video_fmts],
2315 acodecs=[f.get('acodec') for f in audio_fmts],
2316 vexts=[f['ext'] for f in video_fmts],
2317 aexts=[f['ext'] for f in audio_fmts],
2318 preferences=(try_call(lambda: self.params['merge_output_format'].split('/'))
2319 or self.params.get('prefer_free_formats') and ('webm', 'mkv')))
2320
2321 filtered = lambda *keys: filter(None, (traverse_obj(fmt, *keys) for fmt in formats_info))
2322
2323 new_dict = {
2324 'requested_formats': formats_info,
2325 'format': '+'.join(filtered('format')),
2326 'format_id': '+'.join(filtered('format_id')),
2327 'ext': output_ext,
2328 'protocol': '+'.join(map(determine_protocol, formats_info)),
2329 'language': '+'.join(orderedSet(filtered('language'))) or None,
2330 'format_note': '+'.join(orderedSet(filtered('format_note'))) or None,
2331 'filesize_approx': sum(filtered('filesize', 'filesize_approx')) or None,
2332 'tbr': sum(filtered('tbr', 'vbr', 'abr')),
2333 }
2334
2335 if the_only_video:
2336 new_dict.update({
2337 'width': the_only_video.get('width'),
2338 'height': the_only_video.get('height'),
2339 'resolution': the_only_video.get('resolution') or self.format_resolution(the_only_video),
2340 'fps': the_only_video.get('fps'),
2341 'dynamic_range': the_only_video.get('dynamic_range'),
2342 'vcodec': the_only_video.get('vcodec'),
2343 'vbr': the_only_video.get('vbr'),
2344 'stretched_ratio': the_only_video.get('stretched_ratio'),
2345 'aspect_ratio': the_only_video.get('aspect_ratio'),
2346 })
2347
2348 if the_only_audio:
2349 new_dict.update({
2350 'acodec': the_only_audio.get('acodec'),
2351 'abr': the_only_audio.get('abr'),
2352 'asr': the_only_audio.get('asr'),
2353 'audio_channels': the_only_audio.get('audio_channels')
2354 })
2355
2356 return new_dict
2357
2358 def _check_formats(formats):
2359 if self.params.get('check_formats') == 'selected':
2360 yield from self._check_formats(formats)
2361 return
2362 elif (self.params.get('check_formats') is not None
2363 or self.params.get('allow_unplayable_formats')):
2364 yield from formats
2365 return
2366
2367 for f in formats:
2368 if f.get('has_drm') or f.get('__needs_testing'):
2369 yield from self._check_formats([f])
2370 else:
2371 yield f
2372
2373 def _build_selector_function(selector):
2374 if isinstance(selector, list): # ,
2375 fs = [_build_selector_function(s) for s in selector]
2376
2377 def selector_function(ctx):
2378 for f in fs:
2379 yield from f(ctx)
2380 return selector_function
2381
2382 elif selector.type == GROUP: # ()
2383 selector_function = _build_selector_function(selector.selector)
2384
2385 elif selector.type == PICKFIRST: # /
2386 fs = [_build_selector_function(s) for s in selector.selector]
2387
2388 def selector_function(ctx):
2389 for f in fs:
2390 picked_formats = list(f(ctx))
2391 if picked_formats:
2392 return picked_formats
2393 return []
2394
2395 elif selector.type == MERGE: # +
2396 selector_1, selector_2 = map(_build_selector_function, selector.selector)
2397
2398 def selector_function(ctx):
2399 for pair in itertools.product(selector_1(ctx), selector_2(ctx)):
2400 yield _merge(pair)
2401
2402 elif selector.type == SINGLE: # atom
2403 format_spec = selector.selector or 'best'
2404
2405 # TODO: Add allvideo, allaudio etc by generalizing the code with best/worst selector
2406 if format_spec == 'all':
2407 def selector_function(ctx):
2408 yield from _check_formats(ctx['formats'][::-1])
2409 elif format_spec == 'mergeall':
2410 def selector_function(ctx):
2411 formats = list(_check_formats(
2412 f for f in ctx['formats'] if f.get('vcodec') != 'none' or f.get('acodec') != 'none'))
2413 if not formats:
2414 return
2415 merged_format = formats[-1]
2416 for f in formats[-2::-1]:
2417 merged_format = _merge((merged_format, f))
2418 yield merged_format
2419
2420 else:
2421 format_fallback, seperate_fallback, format_reverse, format_idx = False, None, True, 1
2422 mobj = re.match(
2423 r'(?P<bw>best|worst|b|w)(?P<type>video|audio|v|a)?(?P<mod>\*)?(?:\.(?P<n>[1-9]\d*))?$',
2424 format_spec)
2425 if mobj is not None:
2426 format_idx = int_or_none(mobj.group('n'), default=1)
2427 format_reverse = mobj.group('bw')[0] == 'b'
2428 format_type = (mobj.group('type') or [None])[0]
2429 not_format_type = {'v': 'a', 'a': 'v'}.get(format_type)
2430 format_modified = mobj.group('mod') is not None
2431
2432 format_fallback = not format_type and not format_modified # for b, w
2433 _filter_f = (
2434 (lambda f: f.get('%scodec' % format_type) != 'none')
2435 if format_type and format_modified # bv*, ba*, wv*, wa*
2436 else (lambda f: f.get('%scodec' % not_format_type) == 'none')
2437 if format_type # bv, ba, wv, wa
2438 else (lambda f: f.get('vcodec') != 'none' and f.get('acodec') != 'none')
2439 if not format_modified # b, w
2440 else lambda f: True) # b*, w*
2441 filter_f = lambda f: _filter_f(f) and (
2442 f.get('vcodec') != 'none' or f.get('acodec') != 'none')
2443 else:
2444 if format_spec in self._format_selection_exts['audio']:
2445 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none'
2446 elif format_spec in self._format_selection_exts['video']:
2447 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none' and f.get('vcodec') != 'none'
2448 seperate_fallback = lambda f: f.get('ext') == format_spec and f.get('vcodec') != 'none'
2449 elif format_spec in self._format_selection_exts['storyboards']:
2450 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') == 'none' and f.get('vcodec') == 'none'
2451 else:
2452 filter_f = lambda f: f.get('format_id') == format_spec # id
2453
2454 def selector_function(ctx):
2455 formats = list(ctx['formats'])
2456 matches = list(filter(filter_f, formats)) if filter_f is not None else formats
2457 if not matches:
2458 if format_fallback and ctx['incomplete_formats']:
2459 # for extractors with incomplete formats (audio only (soundcloud)
2460 # or video only (imgur)) best/worst will fallback to
2461 # best/worst {video,audio}-only format
2462 matches = list(filter(lambda f: f.get('vcodec') != 'none' or f.get('acodec') != 'none', formats))
2463 elif seperate_fallback and not ctx['has_merged_format']:
2464 # for compatibility with youtube-dl when there is no pre-merged format
2465 matches = list(filter(seperate_fallback, formats))
2466 matches = LazyList(_check_formats(matches[::-1 if format_reverse else 1]))
2467 try:
2468 yield matches[format_idx - 1]
2469 except LazyList.IndexError:
2470 return
2471
2472 filters = [self._build_format_filter(f) for f in selector.filters]
2473
2474 def final_selector(ctx):
2475 ctx_copy = dict(ctx)
2476 for _filter in filters:
2477 ctx_copy['formats'] = list(filter(_filter, ctx_copy['formats']))
2478 return selector_function(ctx_copy)
2479 return final_selector
2480
2481 # HACK: Python 3.12 changed the underlying parser, rendering '7_a' invalid
2482 # Prefix numbers with random letters to avoid it being classified as a number
2483 # See: https://github.com/yt-dlp/yt-dlp/pulls/8797
2484 # TODO: Implement parser not reliant on tokenize.tokenize
2485 prefix = ''.join(random.choices(string.ascii_letters, k=32))
2486 stream = io.BytesIO(re.sub(r'\d[_\d]*', rf'{prefix}\g<0>', format_spec).encode())
2487 try:
2488 tokens = list(_remove_unused_ops(
2489 token._replace(string=token.string.replace(prefix, ''))
2490 for token in tokenize.tokenize(stream.readline)))
2491 except tokenize.TokenError:
2492 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
2493
2494 class TokenIterator:
2495 def __init__(self, tokens):
2496 self.tokens = tokens
2497 self.counter = 0
2498
2499 def __iter__(self):
2500 return self
2501
2502 def __next__(self):
2503 if self.counter >= len(self.tokens):
2504 raise StopIteration()
2505 value = self.tokens[self.counter]
2506 self.counter += 1
2507 return value
2508
2509 next = __next__
2510
2511 def restore_last_token(self):
2512 self.counter -= 1
2513
2514 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
2515 return _build_selector_function(parsed_selector)
2516
2517 def _calc_headers(self, info_dict, load_cookies=False):
2518 res = HTTPHeaderDict(self.params['http_headers'], info_dict.get('http_headers'))
2519 clean_headers(res)
2520
2521 if load_cookies: # For --load-info-json
2522 self._load_cookies(res.get('Cookie'), autoscope=info_dict['url']) # compat
2523 self._load_cookies(info_dict.get('cookies'), autoscope=False)
2524 # The `Cookie` header is removed to prevent leaks and unscoped cookies.
2525 # See: https://github.com/yt-dlp/yt-dlp/security/advisories/GHSA-v8mc-9377-rwjj
2526 res.pop('Cookie', None)
2527 cookies = self.cookiejar.get_cookies_for_url(info_dict['url'])
2528 if cookies:
2529 encoder = LenientSimpleCookie()
2530 values = []
2531 for cookie in cookies:
2532 _, value = encoder.value_encode(cookie.value)
2533 values.append(f'{cookie.name}={value}')
2534 if cookie.domain:
2535 values.append(f'Domain={cookie.domain}')
2536 if cookie.path:
2537 values.append(f'Path={cookie.path}')
2538 if cookie.secure:
2539 values.append('Secure')
2540 if cookie.expires:
2541 values.append(f'Expires={cookie.expires}')
2542 if cookie.version:
2543 values.append(f'Version={cookie.version}')
2544 info_dict['cookies'] = '; '.join(values)
2545
2546 if 'X-Forwarded-For' not in res:
2547 x_forwarded_for_ip = info_dict.get('__x_forwarded_for_ip')
2548 if x_forwarded_for_ip:
2549 res['X-Forwarded-For'] = x_forwarded_for_ip
2550
2551 return res
2552
2553 def _calc_cookies(self, url):
2554 self.deprecation_warning('"YoutubeDL._calc_cookies" is deprecated and may be removed in a future version')
2555 return self.cookiejar.get_cookie_header(url)
2556
2557 def _sort_thumbnails(self, thumbnails):
2558 thumbnails.sort(key=lambda t: (
2559 t.get('preference') if t.get('preference') is not None else -1,
2560 t.get('width') if t.get('width') is not None else -1,
2561 t.get('height') if t.get('height') is not None else -1,
2562 t.get('id') if t.get('id') is not None else '',
2563 t.get('url')))
2564
2565 def _sanitize_thumbnails(self, info_dict):
2566 thumbnails = info_dict.get('thumbnails')
2567 if thumbnails is None:
2568 thumbnail = info_dict.get('thumbnail')
2569 if thumbnail:
2570 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
2571 if not thumbnails:
2572 return
2573
2574 def check_thumbnails(thumbnails):
2575 for t in thumbnails:
2576 self.to_screen(f'[info] Testing thumbnail {t["id"]}')
2577 try:
2578 self.urlopen(HEADRequest(t['url']))
2579 except network_exceptions as err:
2580 self.to_screen(f'[info] Unable to connect to thumbnail {t["id"]} URL {t["url"]!r} - {err}. Skipping...')
2581 continue
2582 yield t
2583
2584 self._sort_thumbnails(thumbnails)
2585 for i, t in enumerate(thumbnails):
2586 if t.get('id') is None:
2587 t['id'] = '%d' % i
2588 if t.get('width') and t.get('height'):
2589 t['resolution'] = '%dx%d' % (t['width'], t['height'])
2590 t['url'] = sanitize_url(t['url'])
2591
2592 if self.params.get('check_formats') is True:
2593 info_dict['thumbnails'] = LazyList(check_thumbnails(thumbnails[::-1]), reverse=True)
2594 else:
2595 info_dict['thumbnails'] = thumbnails
2596
2597 def _fill_common_fields(self, info_dict, final=True):
2598 # TODO: move sanitization here
2599 if final:
2600 title = info_dict['fulltitle'] = info_dict.get('title')
2601 if not title:
2602 if title == '':
2603 self.write_debug('Extractor gave empty title. Creating a generic title')
2604 else:
2605 self.report_warning('Extractor failed to obtain "title". Creating a generic title instead')
2606 info_dict['title'] = f'{info_dict["extractor"].replace(":", "-")} video #{info_dict["id"]}'
2607
2608 if info_dict.get('duration') is not None:
2609 info_dict['duration_string'] = formatSeconds(info_dict['duration'])
2610
2611 for ts_key, date_key in (
2612 ('timestamp', 'upload_date'),
2613 ('release_timestamp', 'release_date'),
2614 ('modified_timestamp', 'modified_date'),
2615 ):
2616 if info_dict.get(date_key) is None and info_dict.get(ts_key) is not None:
2617 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
2618 # see http://bugs.python.org/issue1646728)
2619 with contextlib.suppress(ValueError, OverflowError, OSError):
2620 upload_date = datetime.datetime.fromtimestamp(info_dict[ts_key], datetime.timezone.utc)
2621 info_dict[date_key] = upload_date.strftime('%Y%m%d')
2622
2623 if not info_dict.get('release_year'):
2624 info_dict['release_year'] = traverse_obj(info_dict, ('release_date', {lambda x: int(x[:4])}))
2625
2626 live_keys = ('is_live', 'was_live')
2627 live_status = info_dict.get('live_status')
2628 if live_status is None:
2629 for key in live_keys:
2630 if info_dict.get(key) is False:
2631 continue
2632 if info_dict.get(key):
2633 live_status = key
2634 break
2635 if all(info_dict.get(key) is False for key in live_keys):
2636 live_status = 'not_live'
2637 if live_status:
2638 info_dict['live_status'] = live_status
2639 for key in live_keys:
2640 if info_dict.get(key) is None:
2641 info_dict[key] = (live_status == key)
2642 if live_status == 'post_live':
2643 info_dict['was_live'] = True
2644
2645 # Auto generate title fields corresponding to the *_number fields when missing
2646 # in order to always have clean titles. This is very common for TV series.
2647 for field in ('chapter', 'season', 'episode'):
2648 if final and info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
2649 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
2650
2651 for old_key, new_key in self._deprecated_multivalue_fields.items():
2652 if new_key in info_dict and old_key in info_dict:
2653 if '_version' not in info_dict: # HACK: Do not warn when using --load-info-json
2654 self.deprecation_warning(f'Do not return {old_key!r} when {new_key!r} is present')
2655 elif old_value := info_dict.get(old_key):
2656 info_dict[new_key] = old_value.split(', ')
2657 elif new_value := info_dict.get(new_key):
2658 info_dict[old_key] = ', '.join(v.replace(',', '\N{FULLWIDTH COMMA}') for v in new_value)
2659
2660 def _raise_pending_errors(self, info):
2661 err = info.pop('__pending_error', None)
2662 if err:
2663 self.report_error(err, tb=False)
2664
2665 def sort_formats(self, info_dict):
2666 formats = self._get_formats(info_dict)
2667 formats.sort(key=FormatSorter(
2668 self, info_dict.get('_format_sort_fields') or []).calculate_preference)
2669
2670 def process_video_result(self, info_dict, download=True):
2671 assert info_dict.get('_type', 'video') == 'video'
2672 self._num_videos += 1
2673
2674 if 'id' not in info_dict:
2675 raise ExtractorError('Missing "id" field in extractor result', ie=info_dict['extractor'])
2676 elif not info_dict.get('id'):
2677 raise ExtractorError('Extractor failed to obtain "id"', ie=info_dict['extractor'])
2678
2679 def report_force_conversion(field, field_not, conversion):
2680 self.report_warning(
2681 '"%s" field is not %s - forcing %s conversion, there is an error in extractor'
2682 % (field, field_not, conversion))
2683
2684 def sanitize_string_field(info, string_field):
2685 field = info.get(string_field)
2686 if field is None or isinstance(field, str):
2687 return
2688 report_force_conversion(string_field, 'a string', 'string')
2689 info[string_field] = str(field)
2690
2691 def sanitize_numeric_fields(info):
2692 for numeric_field in self._NUMERIC_FIELDS:
2693 field = info.get(numeric_field)
2694 if field is None or isinstance(field, (int, float)):
2695 continue
2696 report_force_conversion(numeric_field, 'numeric', 'int')
2697 info[numeric_field] = int_or_none(field)
2698
2699 sanitize_string_field(info_dict, 'id')
2700 sanitize_numeric_fields(info_dict)
2701 if info_dict.get('section_end') and info_dict.get('section_start') is not None:
2702 info_dict['duration'] = round(info_dict['section_end'] - info_dict['section_start'], 3)
2703 if (info_dict.get('duration') or 0) <= 0 and info_dict.pop('duration', None):
2704 self.report_warning('"duration" field is negative, there is an error in extractor')
2705
2706 chapters = info_dict.get('chapters') or []
2707 if chapters and chapters[0].get('start_time'):
2708 chapters.insert(0, {'start_time': 0})
2709
2710 dummy_chapter = {'end_time': 0, 'start_time': info_dict.get('duration')}
2711 for idx, (prev, current, next_) in enumerate(zip(
2712 (dummy_chapter, *chapters), chapters, (*chapters[1:], dummy_chapter)), 1):
2713 if current.get('start_time') is None:
2714 current['start_time'] = prev.get('end_time')
2715 if not current.get('end_time'):
2716 current['end_time'] = next_.get('start_time')
2717 if not current.get('title'):
2718 current['title'] = f'<Untitled Chapter {idx}>'
2719
2720 if 'playlist' not in info_dict:
2721 # It isn't part of a playlist
2722 info_dict['playlist'] = None
2723 info_dict['playlist_index'] = None
2724
2725 self._sanitize_thumbnails(info_dict)
2726
2727 thumbnail = info_dict.get('thumbnail')
2728 thumbnails = info_dict.get('thumbnails')
2729 if thumbnail:
2730 info_dict['thumbnail'] = sanitize_url(thumbnail)
2731 elif thumbnails:
2732 info_dict['thumbnail'] = thumbnails[-1]['url']
2733
2734 if info_dict.get('display_id') is None and 'id' in info_dict:
2735 info_dict['display_id'] = info_dict['id']
2736
2737 self._fill_common_fields(info_dict)
2738
2739 for cc_kind in ('subtitles', 'automatic_captions'):
2740 cc = info_dict.get(cc_kind)
2741 if cc:
2742 for _, subtitle in cc.items():
2743 for subtitle_format in subtitle:
2744 if subtitle_format.get('url'):
2745 subtitle_format['url'] = sanitize_url(subtitle_format['url'])
2746 if subtitle_format.get('ext') is None:
2747 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
2748
2749 automatic_captions = info_dict.get('automatic_captions')
2750 subtitles = info_dict.get('subtitles')
2751
2752 info_dict['requested_subtitles'] = self.process_subtitles(
2753 info_dict['id'], subtitles, automatic_captions)
2754
2755 formats = self._get_formats(info_dict)
2756
2757 # Backward compatibility with InfoExtractor._sort_formats
2758 field_preference = (formats or [{}])[0].pop('__sort_fields', None)
2759 if field_preference:
2760 info_dict['_format_sort_fields'] = field_preference
2761
2762 info_dict['_has_drm'] = any( # or None ensures --clean-infojson removes it
2763 f.get('has_drm') and f['has_drm'] != 'maybe' for f in formats) or None
2764 if not self.params.get('allow_unplayable_formats'):
2765 formats = [f for f in formats if not f.get('has_drm') or f['has_drm'] == 'maybe']
2766
2767 if formats and all(f.get('acodec') == f.get('vcodec') == 'none' for f in formats):
2768 self.report_warning(
2769 f'{"This video is DRM protected and " if info_dict["_has_drm"] else ""}'
2770 'only images are available for download. Use --list-formats to see them'.capitalize())
2771
2772 get_from_start = not info_dict.get('is_live') or bool(self.params.get('live_from_start'))
2773 if not get_from_start:
2774 info_dict['title'] += ' ' + datetime.datetime.now().strftime('%Y-%m-%d %H:%M')
2775 if info_dict.get('is_live') and formats:
2776 formats = [f for f in formats if bool(f.get('is_from_start')) == get_from_start]
2777 if get_from_start and not formats:
2778 self.raise_no_formats(info_dict, msg=(
2779 '--live-from-start is passed, but there are no formats that can be downloaded from the start. '
2780 'If you want to download from the current time, use --no-live-from-start'))
2781
2782 def is_wellformed(f):
2783 url = f.get('url')
2784 if not url:
2785 self.report_warning(
2786 '"url" field is missing or empty - skipping format, '
2787 'there is an error in extractor')
2788 return False
2789 if isinstance(url, bytes):
2790 sanitize_string_field(f, 'url')
2791 return True
2792
2793 # Filter out malformed formats for better extraction robustness
2794 formats = list(filter(is_wellformed, formats or []))
2795
2796 if not formats:
2797 self.raise_no_formats(info_dict)
2798
2799 for format in formats:
2800 sanitize_string_field(format, 'format_id')
2801 sanitize_numeric_fields(format)
2802 format['url'] = sanitize_url(format['url'])
2803 if format.get('ext') is None:
2804 format['ext'] = determine_ext(format['url']).lower()
2805 if format.get('protocol') is None:
2806 format['protocol'] = determine_protocol(format)
2807 if format.get('resolution') is None:
2808 format['resolution'] = self.format_resolution(format, default=None)
2809 if format.get('dynamic_range') is None and format.get('vcodec') != 'none':
2810 format['dynamic_range'] = 'SDR'
2811 if format.get('aspect_ratio') is None:
2812 format['aspect_ratio'] = try_call(lambda: round(format['width'] / format['height'], 2))
2813 # For fragmented formats, "tbr" is often max bitrate and not average
2814 if (('manifest-filesize-approx' in self.params['compat_opts'] or not format.get('manifest_url'))
2815 and info_dict.get('duration') and format.get('tbr')
2816 and not format.get('filesize') and not format.get('filesize_approx')):
2817 format['filesize_approx'] = int(info_dict['duration'] * format['tbr'] * (1024 / 8))
2818 format['http_headers'] = self._calc_headers(collections.ChainMap(format, info_dict), load_cookies=True)
2819
2820 # Safeguard against old/insecure infojson when using --load-info-json
2821 if info_dict.get('http_headers'):
2822 info_dict['http_headers'] = HTTPHeaderDict(info_dict['http_headers'])
2823 info_dict['http_headers'].pop('Cookie', None)
2824
2825 # This is copied to http_headers by the above _calc_headers and can now be removed
2826 if '__x_forwarded_for_ip' in info_dict:
2827 del info_dict['__x_forwarded_for_ip']
2828
2829 self.sort_formats({
2830 'formats': formats,
2831 '_format_sort_fields': info_dict.get('_format_sort_fields')
2832 })
2833
2834 # Sanitize and group by format_id
2835 formats_dict = {}
2836 for i, format in enumerate(formats):
2837 if not format.get('format_id'):
2838 format['format_id'] = str(i)
2839 else:
2840 # Sanitize format_id from characters used in format selector expression
2841 format['format_id'] = re.sub(r'[\s,/+\[\]()]', '_', format['format_id'])
2842 formats_dict.setdefault(format['format_id'], []).append(format)
2843
2844 # Make sure all formats have unique format_id
2845 common_exts = set(itertools.chain(*self._format_selection_exts.values()))
2846 for format_id, ambiguous_formats in formats_dict.items():
2847 ambigious_id = len(ambiguous_formats) > 1
2848 for i, format in enumerate(ambiguous_formats):
2849 if ambigious_id:
2850 format['format_id'] = '%s-%d' % (format_id, i)
2851 # Ensure there is no conflict between id and ext in format selection
2852 # See https://github.com/yt-dlp/yt-dlp/issues/1282
2853 if format['format_id'] != format['ext'] and format['format_id'] in common_exts:
2854 format['format_id'] = 'f%s' % format['format_id']
2855
2856 if format.get('format') is None:
2857 format['format'] = '{id} - {res}{note}'.format(
2858 id=format['format_id'],
2859 res=self.format_resolution(format),
2860 note=format_field(format, 'format_note', ' (%s)'),
2861 )
2862
2863 if self.params.get('check_formats') is True:
2864 formats = LazyList(self._check_formats(formats[::-1]), reverse=True)
2865
2866 if not formats or formats[0] is not info_dict:
2867 # only set the 'formats' fields if the original info_dict list them
2868 # otherwise we end up with a circular reference, the first (and unique)
2869 # element in the 'formats' field in info_dict is info_dict itself,
2870 # which can't be exported to json
2871 info_dict['formats'] = formats
2872
2873 info_dict, _ = self.pre_process(info_dict)
2874
2875 if self._match_entry(info_dict, incomplete=self._format_fields) is not None:
2876 return info_dict
2877
2878 self.post_extract(info_dict)
2879 info_dict, _ = self.pre_process(info_dict, 'after_filter')
2880
2881 # The pre-processors may have modified the formats
2882 formats = self._get_formats(info_dict)
2883
2884 list_only = self.params.get('simulate') == 'list_only'
2885 interactive_format_selection = not list_only and self.format_selector == '-'
2886 if self.params.get('list_thumbnails'):
2887 self.list_thumbnails(info_dict)
2888 if self.params.get('listsubtitles'):
2889 if 'automatic_captions' in info_dict:
2890 self.list_subtitles(
2891 info_dict['id'], automatic_captions, 'automatic captions')
2892 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
2893 if self.params.get('listformats') or interactive_format_selection:
2894 self.list_formats(info_dict)
2895 if list_only:
2896 # Without this printing, -F --print-json will not work
2897 self.__forced_printings(info_dict)
2898 return info_dict
2899
2900 format_selector = self.format_selector
2901 while True:
2902 if interactive_format_selection:
2903 req_format = input(self._format_screen('\nEnter format selector ', self.Styles.EMPHASIS)
2904 + '(Press ENTER for default, or Ctrl+C to quit)'
2905 + self._format_screen(': ', self.Styles.EMPHASIS))
2906 try:
2907 format_selector = self.build_format_selector(req_format) if req_format else None
2908 except SyntaxError as err:
2909 self.report_error(err, tb=False, is_error=False)
2910 continue
2911
2912 if format_selector is None:
2913 req_format = self._default_format_spec(info_dict, download=download)
2914 self.write_debug(f'Default format spec: {req_format}')
2915 format_selector = self.build_format_selector(req_format)
2916
2917 formats_to_download = list(format_selector({
2918 'formats': formats,
2919 'has_merged_format': any('none' not in (f.get('acodec'), f.get('vcodec')) for f in formats),
2920 'incomplete_formats': (all(f.get('vcodec') == 'none' for f in formats) # No formats with video
2921 or all(f.get('acodec') == 'none' for f in formats)), # OR, No formats with audio
2922 }))
2923 if interactive_format_selection and not formats_to_download:
2924 self.report_error('Requested format is not available', tb=False, is_error=False)
2925 continue
2926 break
2927
2928 if not formats_to_download:
2929 if not self.params.get('ignore_no_formats_error'):
2930 raise ExtractorError(
2931 'Requested format is not available. Use --list-formats for a list of available formats',
2932 expected=True, video_id=info_dict['id'], ie=info_dict['extractor'])
2933 self.report_warning('Requested format is not available')
2934 # Process what we can, even without any available formats.
2935 formats_to_download = [{}]
2936
2937 requested_ranges = tuple(self.params.get('download_ranges', lambda *_: [{}])(info_dict, self))
2938 best_format, downloaded_formats = formats_to_download[-1], []
2939 if download:
2940 if best_format and requested_ranges:
2941 def to_screen(*msg):
2942 self.to_screen(f'[info] {info_dict["id"]}: {" ".join(", ".join(variadic(m)) for m in msg)}')
2943
2944 to_screen(f'Downloading {len(formats_to_download)} format(s):',
2945 (f['format_id'] for f in formats_to_download))
2946 if requested_ranges != ({}, ):
2947 to_screen(f'Downloading {len(requested_ranges)} time ranges:',
2948 (f'{c["start_time"]:.1f}-{c["end_time"]:.1f}' for c in requested_ranges))
2949 max_downloads_reached = False
2950
2951 for fmt, chapter in itertools.product(formats_to_download, requested_ranges):
2952 new_info = self._copy_infodict(info_dict)
2953 new_info.update(fmt)
2954 offset, duration = info_dict.get('section_start') or 0, info_dict.get('duration') or float('inf')
2955 end_time = offset + min(chapter.get('end_time', duration), duration)
2956 # duration may not be accurate. So allow deviations <1sec
2957 if end_time == float('inf') or end_time > offset + duration + 1:
2958 end_time = None
2959 if chapter or offset:
2960 new_info.update({
2961 'section_start': offset + chapter.get('start_time', 0),
2962 'section_end': end_time,
2963 'section_title': chapter.get('title'),
2964 'section_number': chapter.get('index'),
2965 })
2966 downloaded_formats.append(new_info)
2967 try:
2968 self.process_info(new_info)
2969 except MaxDownloadsReached:
2970 max_downloads_reached = True
2971 self._raise_pending_errors(new_info)
2972 # Remove copied info
2973 for key, val in tuple(new_info.items()):
2974 if info_dict.get(key) == val:
2975 new_info.pop(key)
2976 if max_downloads_reached:
2977 break
2978
2979 write_archive = {f.get('__write_download_archive', False) for f in downloaded_formats}
2980 assert write_archive.issubset({True, False, 'ignore'})
2981 if True in write_archive and False not in write_archive:
2982 self.record_download_archive(info_dict)
2983
2984 info_dict['requested_downloads'] = downloaded_formats
2985 info_dict = self.run_all_pps('after_video', info_dict)
2986 if max_downloads_reached:
2987 raise MaxDownloadsReached()
2988
2989 # We update the info dict with the selected best quality format (backwards compatibility)
2990 info_dict.update(best_format)
2991 return info_dict
2992
2993 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
2994 """Select the requested subtitles and their format"""
2995 available_subs, normal_sub_langs = {}, []
2996 if normal_subtitles and self.params.get('writesubtitles'):
2997 available_subs.update(normal_subtitles)
2998 normal_sub_langs = tuple(normal_subtitles.keys())
2999 if automatic_captions and self.params.get('writeautomaticsub'):
3000 for lang, cap_info in automatic_captions.items():
3001 if lang not in available_subs:
3002 available_subs[lang] = cap_info
3003
3004 if not available_subs or (
3005 not self.params.get('writesubtitles')
3006 and not self.params.get('writeautomaticsub')):
3007 return None
3008
3009 all_sub_langs = tuple(available_subs.keys())
3010 if self.params.get('allsubtitles', False):
3011 requested_langs = all_sub_langs
3012 elif self.params.get('subtitleslangs', False):
3013 try:
3014 requested_langs = orderedSet_from_options(
3015 self.params.get('subtitleslangs'), {'all': all_sub_langs}, use_regex=True)
3016 except re.error as e:
3017 raise ValueError(f'Wrong regex for subtitlelangs: {e.pattern}')
3018 else:
3019 requested_langs = LazyList(itertools.chain(
3020 ['en'] if 'en' in normal_sub_langs else [],
3021 filter(lambda f: f.startswith('en'), normal_sub_langs),
3022 ['en'] if 'en' in all_sub_langs else [],
3023 filter(lambda f: f.startswith('en'), all_sub_langs),
3024 normal_sub_langs, all_sub_langs,
3025 ))[:1]
3026 if requested_langs:
3027 self.to_screen(f'[info] {video_id}: Downloading subtitles: {", ".join(requested_langs)}')
3028
3029 formats_query = self.params.get('subtitlesformat', 'best')
3030 formats_preference = formats_query.split('/') if formats_query else []
3031 subs = {}
3032 for lang in requested_langs:
3033 formats = available_subs.get(lang)
3034 if formats is None:
3035 self.report_warning(f'{lang} subtitles not available for {video_id}')
3036 continue
3037 for ext in formats_preference:
3038 if ext == 'best':
3039 f = formats[-1]
3040 break
3041 matches = list(filter(lambda f: f['ext'] == ext, formats))
3042 if matches:
3043 f = matches[-1]
3044 break
3045 else:
3046 f = formats[-1]
3047 self.report_warning(
3048 'No subtitle format found matching "%s" for language %s, '
3049 'using %s' % (formats_query, lang, f['ext']))
3050 subs[lang] = f
3051 return subs
3052
3053 def _forceprint(self, key, info_dict):
3054 if info_dict is None:
3055 return
3056 info_copy = info_dict.copy()
3057 info_copy.setdefault('filename', self.prepare_filename(info_dict))
3058 if info_dict.get('requested_formats') is not None:
3059 # For RTMP URLs, also include the playpath
3060 info_copy['urls'] = '\n'.join(f['url'] + f.get('play_path', '') for f in info_dict['requested_formats'])
3061 elif info_dict.get('url'):
3062 info_copy['urls'] = info_dict['url'] + info_dict.get('play_path', '')
3063 info_copy['formats_table'] = self.render_formats_table(info_dict)
3064 info_copy['thumbnails_table'] = self.render_thumbnails_table(info_dict)
3065 info_copy['subtitles_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('subtitles'))
3066 info_copy['automatic_captions_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('automatic_captions'))
3067
3068 def format_tmpl(tmpl):
3069 mobj = re.fullmatch(r'([\w.:,]|-\d|(?P<dict>{([\w.:,]|-\d)+}))+=?', tmpl)
3070 if not mobj:
3071 return tmpl
3072
3073 fmt = '%({})s'
3074 if tmpl.startswith('{'):
3075 tmpl, fmt = f'.{tmpl}', '%({})j'
3076 if tmpl.endswith('='):
3077 tmpl, fmt = tmpl[:-1], '{0} = %({0})#j'
3078 return '\n'.join(map(fmt.format, [tmpl] if mobj.group('dict') else tmpl.split(',')))
3079
3080 for tmpl in self.params['forceprint'].get(key, []):
3081 self.to_stdout(self.evaluate_outtmpl(format_tmpl(tmpl), info_copy))
3082
3083 for tmpl, file_tmpl in self.params['print_to_file'].get(key, []):
3084 filename = self.prepare_filename(info_dict, outtmpl=file_tmpl)
3085 tmpl = format_tmpl(tmpl)
3086 self.to_screen(f'[info] Writing {tmpl!r} to: {filename}')
3087 if self._ensure_dir_exists(filename):
3088 with open(filename, 'a', encoding='utf-8', newline='') as f:
3089 f.write(self.evaluate_outtmpl(tmpl, info_copy) + os.linesep)
3090
3091 return info_copy
3092
3093 def __forced_printings(self, info_dict, filename=None, incomplete=True):
3094 if (self.params.get('forcejson')
3095 or self.params['forceprint'].get('video')
3096 or self.params['print_to_file'].get('video')):
3097 self.post_extract(info_dict)
3098 if filename:
3099 info_dict['filename'] = filename
3100 info_copy = self._forceprint('video', info_dict)
3101
3102 def print_field(field, actual_field=None, optional=False):
3103 if actual_field is None:
3104 actual_field = field
3105 if self.params.get(f'force{field}') and (
3106 info_copy.get(field) is not None or (not optional and not incomplete)):
3107 self.to_stdout(info_copy[actual_field])
3108
3109 print_field('title')
3110 print_field('id')
3111 print_field('url', 'urls')
3112 print_field('thumbnail', optional=True)
3113 print_field('description', optional=True)
3114 print_field('filename')
3115 if self.params.get('forceduration') and info_copy.get('duration') is not None:
3116 self.to_stdout(formatSeconds(info_copy['duration']))
3117 print_field('format')
3118
3119 if self.params.get('forcejson'):
3120 self.to_stdout(json.dumps(self.sanitize_info(info_dict)))
3121
3122 def dl(self, name, info, subtitle=False, test=False):
3123 if not info.get('url'):
3124 self.raise_no_formats(info, True)
3125
3126 if test:
3127 verbose = self.params.get('verbose')
3128 params = {
3129 'test': True,
3130 'quiet': self.params.get('quiet') or not verbose,
3131 'verbose': verbose,
3132 'noprogress': not verbose,
3133 'nopart': True,
3134 'skip_unavailable_fragments': False,
3135 'keep_fragments': False,
3136 'overwrites': True,
3137 '_no_ytdl_file': True,
3138 }
3139 else:
3140 params = self.params
3141 fd = get_suitable_downloader(info, params, to_stdout=(name == '-'))(self, params)
3142 if not test:
3143 for ph in self._progress_hooks:
3144 fd.add_progress_hook(ph)
3145 urls = '", "'.join(
3146 (f['url'].split(',')[0] + ',<data>' if f['url'].startswith('data:') else f['url'])
3147 for f in info.get('requested_formats', []) or [info])
3148 self.write_debug(f'Invoking {fd.FD_NAME} downloader on "{urls}"')
3149
3150 # Note: Ideally info should be a deep-copied so that hooks cannot modify it.
3151 # But it may contain objects that are not deep-copyable
3152 new_info = self._copy_infodict(info)
3153 if new_info.get('http_headers') is None:
3154 new_info['http_headers'] = self._calc_headers(new_info)
3155 return fd.download(name, new_info, subtitle)
3156
3157 def existing_file(self, filepaths, *, default_overwrite=True):
3158 existing_files = list(filter(os.path.exists, orderedSet(filepaths)))
3159 if existing_files and not self.params.get('overwrites', default_overwrite):
3160 return existing_files[0]
3161
3162 for file in existing_files:
3163 self.report_file_delete(file)
3164 os.remove(file)
3165 return None
3166
3167 def process_info(self, info_dict):
3168 """Process a single resolved IE result. (Modifies it in-place)"""
3169
3170 assert info_dict.get('_type', 'video') == 'video'
3171 original_infodict = info_dict
3172
3173 if 'format' not in info_dict and 'ext' in info_dict:
3174 info_dict['format'] = info_dict['ext']
3175
3176 if self._match_entry(info_dict) is not None:
3177 info_dict['__write_download_archive'] = 'ignore'
3178 return
3179
3180 # Does nothing under normal operation - for backward compatibility of process_info
3181 self.post_extract(info_dict)
3182
3183 def replace_info_dict(new_info):
3184 nonlocal info_dict
3185 if new_info == info_dict:
3186 return
3187 info_dict.clear()
3188 info_dict.update(new_info)
3189
3190 new_info, _ = self.pre_process(info_dict, 'video')
3191 replace_info_dict(new_info)
3192 self._num_downloads += 1
3193
3194 # info_dict['_filename'] needs to be set for backward compatibility
3195 info_dict['_filename'] = full_filename = self.prepare_filename(info_dict, warn=True)
3196 temp_filename = self.prepare_filename(info_dict, 'temp')
3197 files_to_move = {}
3198
3199 # Forced printings
3200 self.__forced_printings(info_dict, full_filename, incomplete=('format' not in info_dict))
3201
3202 def check_max_downloads():
3203 if self._num_downloads >= float(self.params.get('max_downloads') or 'inf'):
3204 raise MaxDownloadsReached()
3205
3206 if self.params.get('simulate'):
3207 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
3208 check_max_downloads()
3209 return
3210
3211 if full_filename is None:
3212 return
3213 if not self._ensure_dir_exists(encodeFilename(full_filename)):
3214 return
3215 if not self._ensure_dir_exists(encodeFilename(temp_filename)):
3216 return
3217
3218 if self._write_description('video', info_dict,
3219 self.prepare_filename(info_dict, 'description')) is None:
3220 return
3221
3222 sub_files = self._write_subtitles(info_dict, temp_filename)
3223 if sub_files is None:
3224 return
3225 files_to_move.update(dict(sub_files))
3226
3227 thumb_files = self._write_thumbnails(
3228 'video', info_dict, temp_filename, self.prepare_filename(info_dict, 'thumbnail'))
3229 if thumb_files is None:
3230 return
3231 files_to_move.update(dict(thumb_files))
3232
3233 infofn = self.prepare_filename(info_dict, 'infojson')
3234 _infojson_written = self._write_info_json('video', info_dict, infofn)
3235 if _infojson_written:
3236 info_dict['infojson_filename'] = infofn
3237 # For backward compatibility, even though it was a private field
3238 info_dict['__infojson_filename'] = infofn
3239 elif _infojson_written is None:
3240 return
3241
3242 # Note: Annotations are deprecated
3243 annofn = None
3244 if self.params.get('writeannotations', False):
3245 annofn = self.prepare_filename(info_dict, 'annotation')
3246 if annofn:
3247 if not self._ensure_dir_exists(encodeFilename(annofn)):
3248 return
3249 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(annofn)):
3250 self.to_screen('[info] Video annotations are already present')
3251 elif not info_dict.get('annotations'):
3252 self.report_warning('There are no annotations to write.')
3253 else:
3254 try:
3255 self.to_screen('[info] Writing video annotations to: ' + annofn)
3256 with open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
3257 annofile.write(info_dict['annotations'])
3258 except (KeyError, TypeError):
3259 self.report_warning('There are no annotations to write.')
3260 except OSError:
3261 self.report_error('Cannot write annotations file: ' + annofn)
3262 return
3263
3264 # Write internet shortcut files
3265 def _write_link_file(link_type):
3266 url = try_get(info_dict['webpage_url'], iri_to_uri)
3267 if not url:
3268 self.report_warning(
3269 f'Cannot write internet shortcut file because the actual URL of "{info_dict["webpage_url"]}" is unknown')
3270 return True
3271 linkfn = replace_extension(self.prepare_filename(info_dict, 'link'), link_type, info_dict.get('ext'))
3272 if not self._ensure_dir_exists(encodeFilename(linkfn)):
3273 return False
3274 if self.params.get('overwrites', True) and os.path.exists(encodeFilename(linkfn)):
3275 self.to_screen(f'[info] Internet shortcut (.{link_type}) is already present')
3276 return True
3277 try:
3278 self.to_screen(f'[info] Writing internet shortcut (.{link_type}) to: {linkfn}')
3279 with open(encodeFilename(to_high_limit_path(linkfn)), 'w', encoding='utf-8',
3280 newline='\r\n' if link_type == 'url' else '\n') as linkfile:
3281 template_vars = {'url': url}
3282 if link_type == 'desktop':
3283 template_vars['filename'] = linkfn[:-(len(link_type) + 1)]
3284 linkfile.write(LINK_TEMPLATES[link_type] % template_vars)
3285 except OSError:
3286 self.report_error(f'Cannot write internet shortcut {linkfn}')
3287 return False
3288 return True
3289
3290 write_links = {
3291 'url': self.params.get('writeurllink'),
3292 'webloc': self.params.get('writewebloclink'),
3293 'desktop': self.params.get('writedesktoplink'),
3294 }
3295 if self.params.get('writelink'):
3296 link_type = ('webloc' if sys.platform == 'darwin'
3297 else 'desktop' if sys.platform.startswith('linux')
3298 else 'url')
3299 write_links[link_type] = True
3300
3301 if any(should_write and not _write_link_file(link_type)
3302 for link_type, should_write in write_links.items()):
3303 return
3304
3305 new_info, files_to_move = self.pre_process(info_dict, 'before_dl', files_to_move)
3306 replace_info_dict(new_info)
3307
3308 if self.params.get('skip_download'):
3309 info_dict['filepath'] = temp_filename
3310 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
3311 info_dict['__files_to_move'] = files_to_move
3312 replace_info_dict(self.run_pp(MoveFilesAfterDownloadPP(self, False), info_dict))
3313 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
3314 else:
3315 # Download
3316 info_dict.setdefault('__postprocessors', [])
3317 try:
3318
3319 def existing_video_file(*filepaths):
3320 ext = info_dict.get('ext')
3321 converted = lambda file: replace_extension(file, self.params.get('final_ext') or ext, ext)
3322 file = self.existing_file(itertools.chain(*zip(map(converted, filepaths), filepaths)),
3323 default_overwrite=False)
3324 if file:
3325 info_dict['ext'] = os.path.splitext(file)[1][1:]
3326 return file
3327
3328 fd, success = None, True
3329 if info_dict.get('protocol') or info_dict.get('url'):
3330 fd = get_suitable_downloader(info_dict, self.params, to_stdout=temp_filename == '-')
3331 if fd != FFmpegFD and 'no-direct-merge' not in self.params['compat_opts'] and (
3332 info_dict.get('section_start') or info_dict.get('section_end')):
3333 msg = ('This format cannot be partially downloaded' if FFmpegFD.available()
3334 else 'You have requested downloading the video partially, but ffmpeg is not installed')
3335 self.report_error(f'{msg}. Aborting')
3336 return
3337
3338 if info_dict.get('requested_formats') is not None:
3339 old_ext = info_dict['ext']
3340 if self.params.get('merge_output_format') is None:
3341 if (info_dict['ext'] == 'webm'
3342 and info_dict.get('thumbnails')
3343 # check with type instead of pp_key, __name__, or isinstance
3344 # since we dont want any custom PPs to trigger this
3345 and any(type(pp) == EmbedThumbnailPP for pp in self._pps['post_process'])): # noqa: E721
3346 info_dict['ext'] = 'mkv'
3347 self.report_warning(
3348 'webm doesn\'t support embedding a thumbnail, mkv will be used')
3349 new_ext = info_dict['ext']
3350
3351 def correct_ext(filename, ext=new_ext):
3352 if filename == '-':
3353 return filename
3354 filename_real_ext = os.path.splitext(filename)[1][1:]
3355 filename_wo_ext = (
3356 os.path.splitext(filename)[0]
3357 if filename_real_ext in (old_ext, new_ext)
3358 else filename)
3359 return f'{filename_wo_ext}.{ext}'
3360
3361 # Ensure filename always has a correct extension for successful merge
3362 full_filename = correct_ext(full_filename)
3363 temp_filename = correct_ext(temp_filename)
3364 dl_filename = existing_video_file(full_filename, temp_filename)
3365
3366 info_dict['__real_download'] = False
3367 # NOTE: Copy so that original format dicts are not modified
3368 info_dict['requested_formats'] = list(map(dict, info_dict['requested_formats']))
3369
3370 merger = FFmpegMergerPP(self)
3371 downloaded = []
3372 if dl_filename is not None:
3373 self.report_file_already_downloaded(dl_filename)
3374 elif fd:
3375 for f in info_dict['requested_formats'] if fd != FFmpegFD else []:
3376 f['filepath'] = fname = prepend_extension(
3377 correct_ext(temp_filename, info_dict['ext']),
3378 'f%s' % f['format_id'], info_dict['ext'])
3379 downloaded.append(fname)
3380 info_dict['url'] = '\n'.join(f['url'] for f in info_dict['requested_formats'])
3381 success, real_download = self.dl(temp_filename, info_dict)
3382 info_dict['__real_download'] = real_download
3383 else:
3384 if self.params.get('allow_unplayable_formats'):
3385 self.report_warning(
3386 'You have requested merging of multiple formats '
3387 'while also allowing unplayable formats to be downloaded. '
3388 'The formats won\'t be merged to prevent data corruption.')
3389 elif not merger.available:
3390 msg = 'You have requested merging of multiple formats but ffmpeg is not installed'
3391 if not self.params.get('ignoreerrors'):
3392 self.report_error(f'{msg}. Aborting due to --abort-on-error')
3393 return
3394 self.report_warning(f'{msg}. The formats won\'t be merged')
3395
3396 if temp_filename == '-':
3397 reason = ('using a downloader other than ffmpeg' if FFmpegFD.can_merge_formats(info_dict, self.params)
3398 else 'but the formats are incompatible for simultaneous download' if merger.available
3399 else 'but ffmpeg is not installed')
3400 self.report_warning(
3401 f'You have requested downloading multiple formats to stdout {reason}. '
3402 'The formats will be streamed one after the other')
3403 fname = temp_filename
3404 for f in info_dict['requested_formats']:
3405 new_info = dict(info_dict)
3406 del new_info['requested_formats']
3407 new_info.update(f)
3408 if temp_filename != '-':
3409 fname = prepend_extension(
3410 correct_ext(temp_filename, new_info['ext']),
3411 'f%s' % f['format_id'], new_info['ext'])
3412 if not self._ensure_dir_exists(fname):
3413 return
3414 f['filepath'] = fname
3415 downloaded.append(fname)
3416 partial_success, real_download = self.dl(fname, new_info)
3417 info_dict['__real_download'] = info_dict['__real_download'] or real_download
3418 success = success and partial_success
3419
3420 if downloaded and merger.available and not self.params.get('allow_unplayable_formats'):
3421 info_dict['__postprocessors'].append(merger)
3422 info_dict['__files_to_merge'] = downloaded
3423 # Even if there were no downloads, it is being merged only now
3424 info_dict['__real_download'] = True
3425 else:
3426 for file in downloaded:
3427 files_to_move[file] = None
3428 else:
3429 # Just a single file
3430 dl_filename = existing_video_file(full_filename, temp_filename)
3431 if dl_filename is None or dl_filename == temp_filename:
3432 # dl_filename == temp_filename could mean that the file was partially downloaded with --no-part.
3433 # So we should try to resume the download
3434 success, real_download = self.dl(temp_filename, info_dict)
3435 info_dict['__real_download'] = real_download
3436 else:
3437 self.report_file_already_downloaded(dl_filename)
3438
3439 dl_filename = dl_filename or temp_filename
3440 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
3441
3442 except network_exceptions as err:
3443 self.report_error('unable to download video data: %s' % error_to_compat_str(err))
3444 return
3445 except OSError as err:
3446 raise UnavailableVideoError(err)
3447 except (ContentTooShortError, ) as err:
3448 self.report_error(f'content too short (expected {err.expected} bytes and served {err.downloaded})')
3449 return
3450
3451 self._raise_pending_errors(info_dict)
3452 if success and full_filename != '-':
3453
3454 def fixup():
3455 do_fixup = True
3456 fixup_policy = self.params.get('fixup')
3457 vid = info_dict['id']
3458
3459 if fixup_policy in ('ignore', 'never'):
3460 return
3461 elif fixup_policy == 'warn':
3462 do_fixup = 'warn'
3463 elif fixup_policy != 'force':
3464 assert fixup_policy in ('detect_or_warn', None)
3465 if not info_dict.get('__real_download'):
3466 do_fixup = False
3467
3468 def ffmpeg_fixup(cndn, msg, cls):
3469 if not (do_fixup and cndn):
3470 return
3471 elif do_fixup == 'warn':
3472 self.report_warning(f'{vid}: {msg}')
3473 return
3474 pp = cls(self)
3475 if pp.available:
3476 info_dict['__postprocessors'].append(pp)
3477 else:
3478 self.report_warning(f'{vid}: {msg}. Install ffmpeg to fix this automatically')
3479
3480 stretched_ratio = info_dict.get('stretched_ratio')
3481 ffmpeg_fixup(stretched_ratio not in (1, None),
3482 f'Non-uniform pixel ratio {stretched_ratio}',
3483 FFmpegFixupStretchedPP)
3484
3485 downloader = get_suitable_downloader(info_dict, self.params) if 'protocol' in info_dict else None
3486 downloader = downloader.FD_NAME if downloader else None
3487
3488 ext = info_dict.get('ext')
3489 postprocessed_by_ffmpeg = info_dict.get('requested_formats') or any((
3490 isinstance(pp, FFmpegVideoConvertorPP)
3491 and resolve_recode_mapping(ext, pp.mapping)[0] not in (ext, None)
3492 ) for pp in self._pps['post_process'])
3493
3494 if not postprocessed_by_ffmpeg:
3495 ffmpeg_fixup(fd != FFmpegFD and ext == 'm4a'
3496 and info_dict.get('container') == 'm4a_dash',
3497 'writing DASH m4a. Only some players support this container',
3498 FFmpegFixupM4aPP)
3499 ffmpeg_fixup(downloader == 'hlsnative' and not self.params.get('hls_use_mpegts')
3500 or info_dict.get('is_live') and self.params.get('hls_use_mpegts') is None,
3501 'Possible MPEG-TS in MP4 container or malformed AAC timestamps',
3502 FFmpegFixupM3u8PP)
3503 ffmpeg_fixup(downloader == 'dashsegments'
3504 and (info_dict.get('is_live') or info_dict.get('is_dash_periods')),
3505 'Possible duplicate MOOV atoms', FFmpegFixupDuplicateMoovPP)
3506
3507 ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed timestamps detected', FFmpegFixupTimestampPP)
3508 ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed duration detected', FFmpegFixupDurationPP)
3509
3510 fixup()
3511 try:
3512 replace_info_dict(self.post_process(dl_filename, info_dict, files_to_move))
3513 except PostProcessingError as err:
3514 self.report_error('Postprocessing: %s' % str(err))
3515 return
3516 try:
3517 for ph in self._post_hooks:
3518 ph(info_dict['filepath'])
3519 except Exception as err:
3520 self.report_error('post hooks: %s' % str(err))
3521 return
3522 info_dict['__write_download_archive'] = True
3523
3524 assert info_dict is original_infodict # Make sure the info_dict was modified in-place
3525 if self.params.get('force_write_download_archive'):
3526 info_dict['__write_download_archive'] = True
3527 check_max_downloads()
3528
3529 def __download_wrapper(self, func):
3530 @functools.wraps(func)
3531 def wrapper(*args, **kwargs):
3532 try:
3533 res = func(*args, **kwargs)
3534 except UnavailableVideoError as e:
3535 self.report_error(e)
3536 except DownloadCancelled as e:
3537 self.to_screen(f'[info] {e}')
3538 if not self.params.get('break_per_url'):
3539 raise
3540 self._num_downloads = 0
3541 else:
3542 if self.params.get('dump_single_json', False):
3543 self.post_extract(res)
3544 self.to_stdout(json.dumps(self.sanitize_info(res)))
3545 return wrapper
3546
3547 def download(self, url_list):
3548 """Download a given list of URLs."""
3549 url_list = variadic(url_list) # Passing a single URL is a common mistake
3550 outtmpl = self.params['outtmpl']['default']
3551 if (len(url_list) > 1
3552 and outtmpl != '-'
3553 and '%' not in outtmpl
3554 and self.params.get('max_downloads') != 1):
3555 raise SameFileError(outtmpl)
3556
3557 for url in url_list:
3558 self.__download_wrapper(self.extract_info)(
3559 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
3560
3561 return self._download_retcode
3562
3563 def download_with_info_file(self, info_filename):
3564 with contextlib.closing(fileinput.FileInput(
3565 [info_filename], mode='r',
3566 openhook=fileinput.hook_encoded('utf-8'))) as f:
3567 # FileInput doesn't have a read method, we can't call json.load
3568 infos = [self.sanitize_info(info, self.params.get('clean_infojson', True))
3569 for info in variadic(json.loads('\n'.join(f)))]
3570 for info in infos:
3571 try:
3572 self.__download_wrapper(self.process_ie_result)(info, download=True)
3573 except (DownloadError, EntryNotInPlaylist, ReExtractInfo) as e:
3574 if not isinstance(e, EntryNotInPlaylist):
3575 self.to_stderr('\r')
3576 webpage_url = info.get('webpage_url')
3577 if webpage_url is None:
3578 raise
3579 self.report_warning(f'The info failed to download: {e}; trying with URL {webpage_url}')
3580 self.download([webpage_url])
3581 except ExtractorError as e:
3582 self.report_error(e)
3583 return self._download_retcode
3584
3585 @staticmethod
3586 def sanitize_info(info_dict, remove_private_keys=False):
3587 ''' Sanitize the infodict for converting to json '''
3588 if info_dict is None:
3589 return info_dict
3590 info_dict.setdefault('epoch', int(time.time()))
3591 info_dict.setdefault('_type', 'video')
3592 info_dict.setdefault('_version', {
3593 'version': __version__,
3594 'current_git_head': current_git_head(),
3595 'release_git_head': RELEASE_GIT_HEAD,
3596 'repository': ORIGIN,
3597 })
3598
3599 if remove_private_keys:
3600 reject = lambda k, v: v is None or k.startswith('__') or k in {
3601 'requested_downloads', 'requested_formats', 'requested_subtitles', 'requested_entries',
3602 'entries', 'filepath', '_filename', 'filename', 'infojson_filename', 'original_url',
3603 'playlist_autonumber',
3604 }
3605 else:
3606 reject = lambda k, v: False
3607
3608 def filter_fn(obj):
3609 if isinstance(obj, dict):
3610 return {k: filter_fn(v) for k, v in obj.items() if not reject(k, v)}
3611 elif isinstance(obj, (list, tuple, set, LazyList)):
3612 return list(map(filter_fn, obj))
3613 elif obj is None or isinstance(obj, (str, int, float, bool)):
3614 return obj
3615 else:
3616 return repr(obj)
3617
3618 return filter_fn(info_dict)
3619
3620 @staticmethod
3621 def filter_requested_info(info_dict, actually_filter=True):
3622 ''' Alias of sanitize_info for backward compatibility '''
3623 return YoutubeDL.sanitize_info(info_dict, actually_filter)
3624
3625 def _delete_downloaded_files(self, *files_to_delete, info={}, msg=None):
3626 for filename in set(filter(None, files_to_delete)):
3627 if msg:
3628 self.to_screen(msg % filename)
3629 try:
3630 os.remove(filename)
3631 except OSError:
3632 self.report_warning(f'Unable to delete file {filename}')
3633 if filename in info.get('__files_to_move', []): # NB: Delete even if None
3634 del info['__files_to_move'][filename]
3635
3636 @staticmethod
3637 def post_extract(info_dict):
3638 def actual_post_extract(info_dict):
3639 if info_dict.get('_type') in ('playlist', 'multi_video'):
3640 for video_dict in info_dict.get('entries', {}):
3641 actual_post_extract(video_dict or {})
3642 return
3643
3644 post_extractor = info_dict.pop('__post_extractor', None) or (lambda: {})
3645 info_dict.update(post_extractor())
3646
3647 actual_post_extract(info_dict or {})
3648
3649 def run_pp(self, pp, infodict):
3650 files_to_delete = []
3651 if '__files_to_move' not in infodict:
3652 infodict['__files_to_move'] = {}
3653 try:
3654 files_to_delete, infodict = pp.run(infodict)
3655 except PostProcessingError as e:
3656 # Must be True and not 'only_download'
3657 if self.params.get('ignoreerrors') is True:
3658 self.report_error(e)
3659 return infodict
3660 raise
3661
3662 if not files_to_delete:
3663 return infodict
3664 if self.params.get('keepvideo', False):
3665 for f in files_to_delete:
3666 infodict['__files_to_move'].setdefault(f, '')
3667 else:
3668 self._delete_downloaded_files(
3669 *files_to_delete, info=infodict, msg='Deleting original file %s (pass -k to keep)')
3670 return infodict
3671
3672 def run_all_pps(self, key, info, *, additional_pps=None):
3673 if key != 'video':
3674 self._forceprint(key, info)
3675 for pp in (additional_pps or []) + self._pps[key]:
3676 info = self.run_pp(pp, info)
3677 return info
3678
3679 def pre_process(self, ie_info, key='pre_process', files_to_move=None):
3680 info = dict(ie_info)
3681 info['__files_to_move'] = files_to_move or {}
3682 try:
3683 info = self.run_all_pps(key, info)
3684 except PostProcessingError as err:
3685 msg = f'Preprocessing: {err}'
3686 info.setdefault('__pending_error', msg)
3687 self.report_error(msg, is_error=False)
3688 return info, info.pop('__files_to_move', None)
3689
3690 def post_process(self, filename, info, files_to_move=None):
3691 """Run all the postprocessors on the given file."""
3692 info['filepath'] = filename
3693 info['__files_to_move'] = files_to_move or {}
3694 info = self.run_all_pps('post_process', info, additional_pps=info.get('__postprocessors'))
3695 info = self.run_pp(MoveFilesAfterDownloadPP(self), info)
3696 del info['__files_to_move']
3697 return self.run_all_pps('after_move', info)
3698
3699 def _make_archive_id(self, info_dict):
3700 video_id = info_dict.get('id')
3701 if not video_id:
3702 return
3703 # Future-proof against any change in case
3704 # and backwards compatibility with prior versions
3705 extractor = info_dict.get('extractor_key') or info_dict.get('ie_key') # key in a playlist
3706 if extractor is None:
3707 url = str_or_none(info_dict.get('url'))
3708 if not url:
3709 return
3710 # Try to find matching extractor for the URL and take its ie_key
3711 for ie_key, ie in self._ies.items():
3712 if ie.suitable(url):
3713 extractor = ie_key
3714 break
3715 else:
3716 return
3717 return make_archive_id(extractor, video_id)
3718
3719 def in_download_archive(self, info_dict):
3720 if not self.archive:
3721 return False
3722
3723 vid_ids = [self._make_archive_id(info_dict)]
3724 vid_ids.extend(info_dict.get('_old_archive_ids') or [])
3725 return any(id_ in self.archive for id_ in vid_ids)
3726
3727 def record_download_archive(self, info_dict):
3728 fn = self.params.get('download_archive')
3729 if fn is None:
3730 return
3731 vid_id = self._make_archive_id(info_dict)
3732 assert vid_id
3733
3734 self.write_debug(f'Adding to archive: {vid_id}')
3735 if is_path_like(fn):
3736 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
3737 archive_file.write(vid_id + '\n')
3738 self.archive.add(vid_id)
3739
3740 @staticmethod
3741 def format_resolution(format, default='unknown'):
3742 if format.get('vcodec') == 'none' and format.get('acodec') != 'none':
3743 return 'audio only'
3744 if format.get('resolution') is not None:
3745 return format['resolution']
3746 if format.get('width') and format.get('height'):
3747 return '%dx%d' % (format['width'], format['height'])
3748 elif format.get('height'):
3749 return '%sp' % format['height']
3750 elif format.get('width'):
3751 return '%dx?' % format['width']
3752 return default
3753
3754 def _list_format_headers(self, *headers):
3755 if self.params.get('listformats_table', True) is not False:
3756 return [self._format_out(header, self.Styles.HEADERS) for header in headers]
3757 return headers
3758
3759 def _format_note(self, fdict):
3760 res = ''
3761 if fdict.get('ext') in ['f4f', 'f4m']:
3762 res += '(unsupported)'
3763 if fdict.get('language'):
3764 if res:
3765 res += ' '
3766 res += '[%s]' % fdict['language']
3767 if fdict.get('format_note') is not None:
3768 if res:
3769 res += ' '
3770 res += fdict['format_note']
3771 if fdict.get('tbr') is not None:
3772 if res:
3773 res += ', '
3774 res += '%4dk' % fdict['tbr']
3775 if fdict.get('container') is not None:
3776 if res:
3777 res += ', '
3778 res += '%s container' % fdict['container']
3779 if (fdict.get('vcodec') is not None
3780 and fdict.get('vcodec') != 'none'):
3781 if res:
3782 res += ', '
3783 res += fdict['vcodec']
3784 if fdict.get('vbr') is not None:
3785 res += '@'
3786 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
3787 res += 'video@'
3788 if fdict.get('vbr') is not None:
3789 res += '%4dk' % fdict['vbr']
3790 if fdict.get('fps') is not None:
3791 if res:
3792 res += ', '
3793 res += '%sfps' % fdict['fps']
3794 if fdict.get('acodec') is not None:
3795 if res:
3796 res += ', '
3797 if fdict['acodec'] == 'none':
3798 res += 'video only'
3799 else:
3800 res += '%-5s' % fdict['acodec']
3801 elif fdict.get('abr') is not None:
3802 if res:
3803 res += ', '
3804 res += 'audio'
3805 if fdict.get('abr') is not None:
3806 res += '@%3dk' % fdict['abr']
3807 if fdict.get('asr') is not None:
3808 res += ' (%5dHz)' % fdict['asr']
3809 if fdict.get('filesize') is not None:
3810 if res:
3811 res += ', '
3812 res += format_bytes(fdict['filesize'])
3813 elif fdict.get('filesize_approx') is not None:
3814 if res:
3815 res += ', '
3816 res += '~' + format_bytes(fdict['filesize_approx'])
3817 return res
3818
3819 def _get_formats(self, info_dict):
3820 if info_dict.get('formats') is None:
3821 if info_dict.get('url') and info_dict.get('_type', 'video') == 'video':
3822 return [info_dict]
3823 return []
3824 return info_dict['formats']
3825
3826 def render_formats_table(self, info_dict):
3827 formats = self._get_formats(info_dict)
3828 if not formats:
3829 return
3830 if not self.params.get('listformats_table', True) is not False:
3831 table = [
3832 [
3833 format_field(f, 'format_id'),
3834 format_field(f, 'ext'),
3835 self.format_resolution(f),
3836 self._format_note(f)
3837 ] for f in formats if (f.get('preference') or 0) >= -1000]
3838 return render_table(['format code', 'extension', 'resolution', 'note'], table, extra_gap=1)
3839
3840 def simplified_codec(f, field):
3841 assert field in ('acodec', 'vcodec')
3842 codec = f.get(field)
3843 if not codec:
3844 return 'unknown'
3845 elif codec != 'none':
3846 return '.'.join(codec.split('.')[:4])
3847
3848 if field == 'vcodec' and f.get('acodec') == 'none':
3849 return 'images'
3850 elif field == 'acodec' and f.get('vcodec') == 'none':
3851 return ''
3852 return self._format_out('audio only' if field == 'vcodec' else 'video only',
3853 self.Styles.SUPPRESS)
3854
3855 delim = self._format_out('\u2502', self.Styles.DELIM, '|', test_encoding=True)
3856 table = [
3857 [
3858 self._format_out(format_field(f, 'format_id'), self.Styles.ID),
3859 format_field(f, 'ext'),
3860 format_field(f, func=self.format_resolution, ignore=('audio only', 'images')),
3861 format_field(f, 'fps', '\t%d', func=round),
3862 format_field(f, 'dynamic_range', '%s', ignore=(None, 'SDR')).replace('HDR', ''),
3863 format_field(f, 'audio_channels', '\t%s'),
3864 delim, (
3865 format_field(f, 'filesize', ' \t%s', func=format_bytes)
3866 or format_field(f, 'filesize_approx', '≈\t%s', func=format_bytes)
3867 or format_field(try_call(lambda: format_bytes(int(info_dict['duration'] * f['tbr'] * (1024 / 8)))),
3868 None, self._format_out('~\t%s', self.Styles.SUPPRESS))),
3869 format_field(f, 'tbr', '\t%dk', func=round),
3870 shorten_protocol_name(f.get('protocol', '')),
3871 delim,
3872 simplified_codec(f, 'vcodec'),
3873 format_field(f, 'vbr', '\t%dk', func=round),
3874 simplified_codec(f, 'acodec'),
3875 format_field(f, 'abr', '\t%dk', func=round),
3876 format_field(f, 'asr', '\t%s', func=format_decimal_suffix),
3877 join_nonempty(format_field(f, 'language', '[%s]'), join_nonempty(
3878 self._format_out('UNSUPPORTED', self.Styles.BAD_FORMAT) if f.get('ext') in ('f4f', 'f4m') else None,
3879 (self._format_out('Maybe DRM', self.Styles.WARNING) if f.get('has_drm') == 'maybe'
3880 else self._format_out('DRM', self.Styles.BAD_FORMAT) if f.get('has_drm') else None),
3881 format_field(f, 'format_note'),
3882 format_field(f, 'container', ignore=(None, f.get('ext'))),
3883 delim=', '), delim=' '),
3884 ] for f in formats if f.get('preference') is None or f['preference'] >= -1000]
3885 header_line = self._list_format_headers(
3886 'ID', 'EXT', 'RESOLUTION', '\tFPS', 'HDR', 'CH', delim, '\tFILESIZE', '\tTBR', 'PROTO',
3887 delim, 'VCODEC', '\tVBR', 'ACODEC', '\tABR', '\tASR', 'MORE INFO')
3888
3889 return render_table(
3890 header_line, table, hide_empty=True,
3891 delim=self._format_out('\u2500', self.Styles.DELIM, '-', test_encoding=True))
3892
3893 def render_thumbnails_table(self, info_dict):
3894 thumbnails = list(info_dict.get('thumbnails') or [])
3895 if not thumbnails:
3896 return None
3897 return render_table(
3898 self._list_format_headers('ID', 'Width', 'Height', 'URL'),
3899 [[t.get('id'), t.get('width') or 'unknown', t.get('height') or 'unknown', t['url']] for t in thumbnails])
3900
3901 def render_subtitles_table(self, video_id, subtitles):
3902 def _row(lang, formats):
3903 exts, names = zip(*((f['ext'], f.get('name') or 'unknown') for f in reversed(formats)))
3904 if len(set(names)) == 1:
3905 names = [] if names[0] == 'unknown' else names[:1]
3906 return [lang, ', '.join(names), ', '.join(exts)]
3907
3908 if not subtitles:
3909 return None
3910 return render_table(
3911 self._list_format_headers('Language', 'Name', 'Formats'),
3912 [_row(lang, formats) for lang, formats in subtitles.items()],
3913 hide_empty=True)
3914
3915 def __list_table(self, video_id, name, func, *args):
3916 table = func(*args)
3917 if not table:
3918 self.to_screen(f'{video_id} has no {name}')
3919 return
3920 self.to_screen(f'[info] Available {name} for {video_id}:')
3921 self.to_stdout(table)
3922
3923 def list_formats(self, info_dict):
3924 self.__list_table(info_dict['id'], 'formats', self.render_formats_table, info_dict)
3925
3926 def list_thumbnails(self, info_dict):
3927 self.__list_table(info_dict['id'], 'thumbnails', self.render_thumbnails_table, info_dict)
3928
3929 def list_subtitles(self, video_id, subtitles, name='subtitles'):
3930 self.__list_table(video_id, name, self.render_subtitles_table, video_id, subtitles)
3931
3932 def print_debug_header(self):
3933 if not self.params.get('verbose'):
3934 return
3935
3936 from . import _IN_CLI # Must be delayed import
3937
3938 # These imports can be slow. So import them only as needed
3939 from .extractor.extractors import _LAZY_LOADER
3940 from .extractor.extractors import (
3941 _PLUGIN_CLASSES as plugin_ies,
3942 _PLUGIN_OVERRIDES as plugin_ie_overrides
3943 )
3944
3945 def get_encoding(stream):
3946 ret = str(getattr(stream, 'encoding', 'missing (%s)' % type(stream).__name__))
3947 additional_info = []
3948 if os.environ.get('TERM', '').lower() == 'dumb':
3949 additional_info.append('dumb')
3950 if not supports_terminal_sequences(stream):
3951 from .utils import WINDOWS_VT_MODE # Must be imported locally
3952 additional_info.append('No VT' if WINDOWS_VT_MODE is False else 'No ANSI')
3953 if additional_info:
3954 ret = f'{ret} ({",".join(additional_info)})'
3955 return ret
3956
3957 encoding_str = 'Encodings: locale %s, fs %s, pref %s, %s' % (
3958 locale.getpreferredencoding(),
3959 sys.getfilesystemencoding(),
3960 self.get_encoding(),
3961 ', '.join(
3962 f'{key} {get_encoding(stream)}' for key, stream in self._out_files.items_
3963 if stream is not None and key != 'console')
3964 )
3965
3966 logger = self.params.get('logger')
3967 if logger:
3968 write_debug = lambda msg: logger.debug(f'[debug] {msg}')
3969 write_debug(encoding_str)
3970 else:
3971 write_string(f'[debug] {encoding_str}\n', encoding=None)
3972 write_debug = lambda msg: self._write_string(f'[debug] {msg}\n')
3973
3974 source = detect_variant()
3975 if VARIANT not in (None, 'pip'):
3976 source += '*'
3977 klass = type(self)
3978 write_debug(join_nonempty(
3979 f'{REPOSITORY.rpartition("/")[2]} version',
3980 _make_label(ORIGIN, CHANNEL.partition('@')[2] or __version__, __version__),
3981 f'[{RELEASE_GIT_HEAD[:9]}]' if RELEASE_GIT_HEAD else '',
3982 '' if source == 'unknown' else f'({source})',
3983 '' if _IN_CLI else 'API' if klass == YoutubeDL else f'API:{self.__module__}.{klass.__qualname__}',
3984 delim=' '))
3985
3986 if not _IN_CLI:
3987 write_debug(f'params: {self.params}')
3988
3989 if not _LAZY_LOADER:
3990 if os.environ.get('YTDLP_NO_LAZY_EXTRACTORS'):
3991 write_debug('Lazy loading extractors is forcibly disabled')
3992 else:
3993 write_debug('Lazy loading extractors is disabled')
3994 if self.params['compat_opts']:
3995 write_debug('Compatibility options: %s' % ', '.join(self.params['compat_opts']))
3996
3997 if current_git_head():
3998 write_debug(f'Git HEAD: {current_git_head()}')
3999 write_debug(system_identifier())
4000
4001 exe_versions, ffmpeg_features = FFmpegPostProcessor.get_versions_and_features(self)
4002 ffmpeg_features = {key for key, val in ffmpeg_features.items() if val}
4003 if ffmpeg_features:
4004 exe_versions['ffmpeg'] += ' (%s)' % ','.join(sorted(ffmpeg_features))
4005
4006 exe_versions['rtmpdump'] = rtmpdump_version()
4007 exe_versions['phantomjs'] = PhantomJSwrapper._version()
4008 exe_str = ', '.join(
4009 f'{exe} {v}' for exe, v in sorted(exe_versions.items()) if v
4010 ) or 'none'
4011 write_debug('exe versions: %s' % exe_str)
4012
4013 from .compat.compat_utils import get_package_info
4014 from .dependencies import available_dependencies
4015
4016 write_debug('Optional libraries: %s' % (', '.join(sorted({
4017 join_nonempty(*get_package_info(m)) for m in available_dependencies.values()
4018 })) or 'none'))
4019
4020 write_debug(f'Proxy map: {self.proxies}')
4021 write_debug(f'Request Handlers: {", ".join(rh.RH_NAME for rh in self._request_director.handlers.values())}')
4022 for plugin_type, plugins in {'Extractor': plugin_ies, 'Post-Processor': plugin_pps}.items():
4023 display_list = ['%s%s' % (
4024 klass.__name__, '' if klass.__name__ == name else f' as {name}')
4025 for name, klass in plugins.items()]
4026 if plugin_type == 'Extractor':
4027 display_list.extend(f'{plugins[-1].IE_NAME.partition("+")[2]} ({parent.__name__})'
4028 for parent, plugins in plugin_ie_overrides.items())
4029 if not display_list:
4030 continue
4031 write_debug(f'{plugin_type} Plugins: {", ".join(sorted(display_list))}')
4032
4033 plugin_dirs = plugin_directories()
4034 if plugin_dirs:
4035 write_debug(f'Plugin directories: {plugin_dirs}')
4036
4037 # Not implemented
4038 if False and self.params.get('call_home'):
4039 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode()
4040 write_debug('Public IP address: %s' % ipaddr)
4041 latest_version = self.urlopen(
4042 'https://yt-dl.org/latest/version').read().decode()
4043 if version_tuple(latest_version) > version_tuple(__version__):
4044 self.report_warning(
4045 'You are using an outdated version (newest version: %s)! '
4046 'See https://yt-dl.org/update if you need help updating.' %
4047 latest_version)
4048
4049 @functools.cached_property
4050 def proxies(self):
4051 """Global proxy configuration"""
4052 opts_proxy = self.params.get('proxy')
4053 if opts_proxy is not None:
4054 if opts_proxy == '':
4055 opts_proxy = '__noproxy__'
4056 proxies = {'all': opts_proxy}
4057 else:
4058 proxies = urllib.request.getproxies()
4059 # compat. Set HTTPS_PROXY to __noproxy__ to revert
4060 if 'http' in proxies and 'https' not in proxies:
4061 proxies['https'] = proxies['http']
4062
4063 return proxies
4064
4065 @functools.cached_property
4066 def cookiejar(self):
4067 """Global cookiejar instance"""
4068 return load_cookies(
4069 self.params.get('cookiefile'), self.params.get('cookiesfrombrowser'), self)
4070
4071 @property
4072 def _opener(self):
4073 """
4074 Get a urllib OpenerDirector from the Urllib handler (deprecated).
4075 """
4076 self.deprecation_warning('YoutubeDL._opener is deprecated, use YoutubeDL.urlopen()')
4077 handler = self._request_director.handlers['Urllib']
4078 return handler._get_instance(cookiejar=self.cookiejar, proxies=self.proxies)
4079
4080 def urlopen(self, req):
4081 """ Start an HTTP download """
4082 if isinstance(req, str):
4083 req = Request(req)
4084 elif isinstance(req, urllib.request.Request):
4085 self.deprecation_warning(
4086 'Passing a urllib.request.Request object to YoutubeDL.urlopen() is deprecated. '
4087 'Use yt_dlp.networking.common.Request instead.')
4088 req = urllib_req_to_req(req)
4089 assert isinstance(req, Request)
4090
4091 # compat: Assume user:pass url params are basic auth
4092 url, basic_auth_header = extract_basic_auth(req.url)
4093 if basic_auth_header:
4094 req.headers['Authorization'] = basic_auth_header
4095 req.url = sanitize_url(url)
4096
4097 clean_proxies(proxies=req.proxies, headers=req.headers)
4098 clean_headers(req.headers)
4099
4100 try:
4101 return self._request_director.send(req)
4102 except NoSupportingHandlers as e:
4103 for ue in e.unsupported_errors:
4104 # FIXME: This depends on the order of errors.
4105 if not (ue.handler and ue.msg):
4106 continue
4107 if ue.handler.RH_KEY == 'Urllib' and 'unsupported url scheme: "file"' in ue.msg.lower():
4108 raise RequestError(
4109 'file:// URLs are disabled by default in yt-dlp for security reasons. '
4110 'Use --enable-file-urls to enable at your own risk.', cause=ue) from ue
4111 if 'unsupported proxy type: "https"' in ue.msg.lower():
4112 raise RequestError(
4113 'To use an HTTPS proxy for this request, one of the following dependencies needs to be installed: requests')
4114
4115 elif (
4116 re.match(r'unsupported url scheme: "wss?"', ue.msg.lower())
4117 and 'websockets' not in self._request_director.handlers
4118 ):
4119 raise RequestError(
4120 'This request requires WebSocket support. '
4121 'Ensure one of the following dependencies are installed: websockets',
4122 cause=ue) from ue
4123 raise
4124 except SSLError as e:
4125 if 'UNSAFE_LEGACY_RENEGOTIATION_DISABLED' in str(e):
4126 raise RequestError('UNSAFE_LEGACY_RENEGOTIATION_DISABLED: Try using --legacy-server-connect', cause=e) from e
4127 elif 'SSLV3_ALERT_HANDSHAKE_FAILURE' in str(e):
4128 raise RequestError(
4129 'SSLV3_ALERT_HANDSHAKE_FAILURE: The server may not support the current cipher list. '
4130 'Try using --legacy-server-connect', cause=e) from e
4131 raise
4132
4133 def build_request_director(self, handlers, preferences=None):
4134 logger = _YDLLogger(self)
4135 headers = self.params['http_headers'].copy()
4136 proxies = self.proxies.copy()
4137 clean_headers(headers)
4138 clean_proxies(proxies, headers)
4139
4140 director = RequestDirector(logger=logger, verbose=self.params.get('debug_printtraffic'))
4141 for handler in handlers:
4142 director.add_handler(handler(
4143 logger=logger,
4144 headers=headers,
4145 cookiejar=self.cookiejar,
4146 proxies=proxies,
4147 prefer_system_certs='no-certifi' in self.params['compat_opts'],
4148 verify=not self.params.get('nocheckcertificate'),
4149 **traverse_obj(self.params, {
4150 'verbose': 'debug_printtraffic',
4151 'source_address': 'source_address',
4152 'timeout': 'socket_timeout',
4153 'legacy_ssl_support': 'legacyserverconnect',
4154 'enable_file_urls': 'enable_file_urls',
4155 'client_cert': {
4156 'client_certificate': 'client_certificate',
4157 'client_certificate_key': 'client_certificate_key',
4158 'client_certificate_password': 'client_certificate_password',
4159 },
4160 }),
4161 ))
4162 director.preferences.update(preferences or [])
4163 if 'prefer-legacy-http-handler' in self.params['compat_opts']:
4164 director.preferences.add(lambda rh, _: 500 if rh.RH_KEY == 'Urllib' else 0)
4165 return director
4166
4167 @functools.cached_property
4168 def _request_director(self):
4169 return self.build_request_director(_REQUEST_HANDLERS.values(), _RH_PREFERENCES)
4170
4171 def encode(self, s):
4172 if isinstance(s, bytes):
4173 return s # Already encoded
4174
4175 try:
4176 return s.encode(self.get_encoding())
4177 except UnicodeEncodeError as err:
4178 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
4179 raise
4180
4181 def get_encoding(self):
4182 encoding = self.params.get('encoding')
4183 if encoding is None:
4184 encoding = preferredencoding()
4185 return encoding
4186
4187 def _write_info_json(self, label, ie_result, infofn, overwrite=None):
4188 ''' Write infojson and returns True = written, 'exists' = Already exists, False = skip, None = error '''
4189 if overwrite is None:
4190 overwrite = self.params.get('overwrites', True)
4191 if not self.params.get('writeinfojson'):
4192 return False
4193 elif not infofn:
4194 self.write_debug(f'Skipping writing {label} infojson')
4195 return False
4196 elif not self._ensure_dir_exists(infofn):
4197 return None
4198 elif not overwrite and os.path.exists(infofn):
4199 self.to_screen(f'[info] {label.title()} metadata is already present')
4200 return 'exists'
4201
4202 self.to_screen(f'[info] Writing {label} metadata as JSON to: {infofn}')
4203 try:
4204 write_json_file(self.sanitize_info(ie_result, self.params.get('clean_infojson', True)), infofn)
4205 return True
4206 except OSError:
4207 self.report_error(f'Cannot write {label} metadata to JSON file {infofn}')
4208 return None
4209
4210 def _write_description(self, label, ie_result, descfn):
4211 ''' Write description and returns True = written, False = skip, None = error '''
4212 if not self.params.get('writedescription'):
4213 return False
4214 elif not descfn:
4215 self.write_debug(f'Skipping writing {label} description')
4216 return False
4217 elif not self._ensure_dir_exists(descfn):
4218 return None
4219 elif not self.params.get('overwrites', True) and os.path.exists(descfn):
4220 self.to_screen(f'[info] {label.title()} description is already present')
4221 elif ie_result.get('description') is None:
4222 self.to_screen(f'[info] There\'s no {label} description to write')
4223 return False
4224 else:
4225 try:
4226 self.to_screen(f'[info] Writing {label} description to: {descfn}')
4227 with open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
4228 descfile.write(ie_result['description'])
4229 except OSError:
4230 self.report_error(f'Cannot write {label} description file {descfn}')
4231 return None
4232 return True
4233
4234 def _write_subtitles(self, info_dict, filename):
4235 ''' Write subtitles to file and return list of (sub_filename, final_sub_filename); or None if error'''
4236 ret = []
4237 subtitles = info_dict.get('requested_subtitles')
4238 if not (self.params.get('writesubtitles') or self.params.get('writeautomaticsub')):
4239 # subtitles download errors are already managed as troubles in relevant IE
4240 # that way it will silently go on when used with unsupporting IE
4241 return ret
4242 elif not subtitles:
4243 self.to_screen('[info] There are no subtitles for the requested languages')
4244 return ret
4245 sub_filename_base = self.prepare_filename(info_dict, 'subtitle')
4246 if not sub_filename_base:
4247 self.to_screen('[info] Skipping writing video subtitles')
4248 return ret
4249
4250 for sub_lang, sub_info in subtitles.items():
4251 sub_format = sub_info['ext']
4252 sub_filename = subtitles_filename(filename, sub_lang, sub_format, info_dict.get('ext'))
4253 sub_filename_final = subtitles_filename(sub_filename_base, sub_lang, sub_format, info_dict.get('ext'))
4254 existing_sub = self.existing_file((sub_filename_final, sub_filename))
4255 if existing_sub:
4256 self.to_screen(f'[info] Video subtitle {sub_lang}.{sub_format} is already present')
4257 sub_info['filepath'] = existing_sub
4258 ret.append((existing_sub, sub_filename_final))
4259 continue
4260
4261 self.to_screen(f'[info] Writing video subtitles to: {sub_filename}')
4262 if sub_info.get('data') is not None:
4263 try:
4264 # Use newline='' to prevent conversion of newline characters
4265 # See https://github.com/ytdl-org/youtube-dl/issues/10268
4266 with open(sub_filename, 'w', encoding='utf-8', newline='') as subfile:
4267 subfile.write(sub_info['data'])
4268 sub_info['filepath'] = sub_filename
4269 ret.append((sub_filename, sub_filename_final))
4270 continue
4271 except OSError:
4272 self.report_error(f'Cannot write video subtitles file {sub_filename}')
4273 return None
4274
4275 try:
4276 sub_copy = sub_info.copy()
4277 sub_copy.setdefault('http_headers', info_dict.get('http_headers'))
4278 self.dl(sub_filename, sub_copy, subtitle=True)
4279 sub_info['filepath'] = sub_filename
4280 ret.append((sub_filename, sub_filename_final))
4281 except (DownloadError, ExtractorError, IOError, OSError, ValueError) + network_exceptions as err:
4282 msg = f'Unable to download video subtitles for {sub_lang!r}: {err}'
4283 if self.params.get('ignoreerrors') is not True: # False or 'only_download'
4284 if not self.params.get('ignoreerrors'):
4285 self.report_error(msg)
4286 raise DownloadError(msg)
4287 self.report_warning(msg)
4288 return ret
4289
4290 def _write_thumbnails(self, label, info_dict, filename, thumb_filename_base=None):
4291 ''' Write thumbnails to file and return list of (thumb_filename, final_thumb_filename); or None if error '''
4292 write_all = self.params.get('write_all_thumbnails', False)
4293 thumbnails, ret = [], []
4294 if write_all or self.params.get('writethumbnail', False):
4295 thumbnails = info_dict.get('thumbnails') or []
4296 if not thumbnails:
4297 self.to_screen(f'[info] There are no {label} thumbnails to download')
4298 return ret
4299 multiple = write_all and len(thumbnails) > 1
4300
4301 if thumb_filename_base is None:
4302 thumb_filename_base = filename
4303 if thumbnails and not thumb_filename_base:
4304 self.write_debug(f'Skipping writing {label} thumbnail')
4305 return ret
4306
4307 if thumbnails and not self._ensure_dir_exists(filename):
4308 return None
4309
4310 for idx, t in list(enumerate(thumbnails))[::-1]:
4311 thumb_ext = (f'{t["id"]}.' if multiple else '') + determine_ext(t['url'], 'jpg')
4312 thumb_display_id = f'{label} thumbnail {t["id"]}'
4313 thumb_filename = replace_extension(filename, thumb_ext, info_dict.get('ext'))
4314 thumb_filename_final = replace_extension(thumb_filename_base, thumb_ext, info_dict.get('ext'))
4315
4316 existing_thumb = self.existing_file((thumb_filename_final, thumb_filename))
4317 if existing_thumb:
4318 self.to_screen('[info] %s is already present' % (
4319 thumb_display_id if multiple else f'{label} thumbnail').capitalize())
4320 t['filepath'] = existing_thumb
4321 ret.append((existing_thumb, thumb_filename_final))
4322 else:
4323 self.to_screen(f'[info] Downloading {thumb_display_id} ...')
4324 try:
4325 uf = self.urlopen(Request(t['url'], headers=t.get('http_headers', {})))
4326 self.to_screen(f'[info] Writing {thumb_display_id} to: {thumb_filename}')
4327 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
4328 shutil.copyfileobj(uf, thumbf)
4329 ret.append((thumb_filename, thumb_filename_final))
4330 t['filepath'] = thumb_filename
4331 except network_exceptions as err:
4332 if isinstance(err, HTTPError) and err.status == 404:
4333 self.to_screen(f'[info] {thumb_display_id.title()} does not exist')
4334 else:
4335 self.report_warning(f'Unable to download {thumb_display_id}: {err}')
4336 thumbnails.pop(idx)
4337 if ret and not write_all:
4338 break
4339 return ret