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