]> jfr.im git - yt-dlp.git/blame_incremental - yt_dlp/YoutubeDL.py
[core] Allow extractors to mark formats as potentially DRM (#7396)
[yt-dlp.git] / yt_dlp / YoutubeDL.py
... / ...
CommitLineData
1import collections
2import contextlib
3import datetime
4import errno
5import fileinput
6import functools
7import io
8import itertools
9import json
10import locale
11import operator
12import os
13import random
14import re
15import shutil
16import string
17import subprocess
18import sys
19import tempfile
20import time
21import tokenize
22import traceback
23import unicodedata
24
25from .cache import Cache
26from .compat import urllib # isort: split
27from .compat import compat_os_name, compat_shlex_quote
28from .cookies import load_cookies
29from .downloader import FFmpegFD, get_suitable_downloader, shorten_protocol_name
30from .downloader.rtmp import rtmpdump_version
31from .extractor import gen_extractor_classes, get_info_extractor
32from .extractor.common import UnsupportedURLIE
33from .extractor.openload import PhantomJSwrapper
34from .minicurses import format_text
35from .plugins import directories as plugin_directories
36from .postprocessor import _PLUGIN_CLASSES as plugin_pps
37from .postprocessor import (
38 EmbedThumbnailPP,
39 FFmpegFixupDuplicateMoovPP,
40 FFmpegFixupDurationPP,
41 FFmpegFixupM3u8PP,
42 FFmpegFixupM4aPP,
43 FFmpegFixupStretchedPP,
44 FFmpegFixupTimestampPP,
45 FFmpegMergerPP,
46 FFmpegPostProcessor,
47 FFmpegVideoConvertorPP,
48 MoveFilesAfterDownloadPP,
49 get_postprocessor,
50)
51from .postprocessor.ffmpeg import resolve_mapping as resolve_recode_mapping
52from .update import REPOSITORY, current_git_head, detect_variant
53from .utils import (
54 DEFAULT_OUTTMPL,
55 IDENTITY,
56 LINK_TEMPLATES,
57 MEDIA_EXTENSIONS,
58 NO_DEFAULT,
59 NUMBER_RE,
60 OUTTMPL_TYPES,
61 POSTPROCESS_WHEN,
62 STR_FORMAT_RE_TMPL,
63 STR_FORMAT_TYPES,
64 ContentTooShortError,
65 DateRange,
66 DownloadCancelled,
67 DownloadError,
68 EntryNotInPlaylist,
69 ExistingVideoReached,
70 ExtractorError,
71 FormatSorter,
72 GeoRestrictedError,
73 HEADRequest,
74 ISO3166Utils,
75 LazyList,
76 MaxDownloadsReached,
77 Namespace,
78 PagedList,
79 PerRequestProxyHandler,
80 PlaylistEntries,
81 Popen,
82 PostProcessingError,
83 ReExtractInfo,
84 RejectedVideoReached,
85 SameFileError,
86 UnavailableVideoError,
87 UserNotLive,
88 YoutubeDLCookieProcessor,
89 YoutubeDLHandler,
90 YoutubeDLRedirectHandler,
91 age_restricted,
92 args_to_str,
93 bug_reports_message,
94 date_from_str,
95 deprecation_warning,
96 determine_ext,
97 determine_protocol,
98 encode_compat_str,
99 encodeFilename,
100 error_to_compat_str,
101 escapeHTML,
102 expand_path,
103 filter_dict,
104 float_or_none,
105 format_bytes,
106 format_decimal_suffix,
107 format_field,
108 formatSeconds,
109 get_compatible_ext,
110 get_domain,
111 int_or_none,
112 iri_to_uri,
113 is_path_like,
114 join_nonempty,
115 locked_file,
116 make_archive_id,
117 make_dir,
118 make_HTTPS_handler,
119 merge_headers,
120 network_exceptions,
121 number_of_digits,
122 orderedSet,
123 orderedSet_from_options,
124 parse_filesize,
125 preferredencoding,
126 prepend_extension,
127 remove_terminal_sequences,
128 render_table,
129 replace_extension,
130 sanitize_filename,
131 sanitize_path,
132 sanitize_url,
133 sanitized_Request,
134 std_headers,
135 str_or_none,
136 strftime_or_none,
137 subtitles_filename,
138 supports_terminal_sequences,
139 system_identifier,
140 timetuple_from_msec,
141 to_high_limit_path,
142 traverse_obj,
143 try_call,
144 try_get,
145 url_basename,
146 variadic,
147 version_tuple,
148 windows_enable_vt_mode,
149 write_json_file,
150 write_string,
151)
152from .version import CHANNEL, RELEASE_GIT_HEAD, VARIANT, __version__
153
154if compat_os_name == 'nt':
155 import ctypes
156
157
158class YoutubeDL:
159 """YoutubeDL class.
160
161 YoutubeDL objects are the ones responsible of downloading the
162 actual video file and writing it to disk if the user has requested
163 it, among some other tasks. In most cases there should be one per
164 program. As, given a video URL, the downloader doesn't know how to
165 extract all the needed information, task that InfoExtractors do, it
166 has to pass the URL to one of them.
167
168 For this, YoutubeDL objects have a method that allows
169 InfoExtractors to be registered in a given order. When it is passed
170 a URL, the YoutubeDL object handles it to the first InfoExtractor it
171 finds that reports being able to handle it. The InfoExtractor extracts
172 all the information about the video or videos the URL refers to, and
173 YoutubeDL process the extracted information, possibly using a File
174 Downloader to download the video.
175
176 YoutubeDL objects accept a lot of parameters. In order not to saturate
177 the object constructor with arguments, it receives a dictionary of
178 options instead. These options are available through the params
179 attribute for the InfoExtractors to use. The YoutubeDL also
180 registers itself as the downloader in charge for the InfoExtractors
181 that are added to it, so this is a "mutual registration".
182
183 Available options:
184
185 username: Username for authentication purposes.
186 password: Password for authentication purposes.
187 videopassword: Password for accessing a video.
188 ap_mso: Adobe Pass multiple-system operator identifier.
189 ap_username: Multiple-system operator account username.
190 ap_password: Multiple-system operator account password.
191 usenetrc: Use netrc for authentication instead.
192 netrc_location: Location of the netrc file. Defaults to ~/.netrc.
193 netrc_cmd: Use a shell command to get credentials
194 verbose: Print additional info to stdout.
195 quiet: Do not print messages to stdout.
196 no_warnings: Do not print out anything for warnings.
197 forceprint: A dict with keys WHEN mapped to a list of templates to
198 print to stdout. The allowed keys are video or any of the
199 items in utils.POSTPROCESS_WHEN.
200 For compatibility, a single list is also accepted
201 print_to_file: A dict with keys WHEN (same as forceprint) mapped to
202 a list of tuples with (template, filename)
203 forcejson: Force printing info_dict as JSON.
204 dump_single_json: Force printing the info_dict of the whole playlist
205 (or video) as a single JSON line.
206 force_write_download_archive: Force writing download archive regardless
207 of 'skip_download' or 'simulate'.
208 simulate: Do not download the video files. If unset (or None),
209 simulate only if listsubtitles, listformats or list_thumbnails is used
210 format: Video format code. see "FORMAT SELECTION" for more details.
211 You can also pass a function. The function takes 'ctx' as
212 argument and returns the formats to download.
213 See "build_format_selector" for an implementation
214 allow_unplayable_formats: Allow unplayable formats to be extracted and downloaded.
215 ignore_no_formats_error: Ignore "No video formats" error. Usefull for
216 extracting metadata even if the video is not actually
217 available for download (experimental)
218 format_sort: A list of fields by which to sort the video formats.
219 See "Sorting Formats" for more details.
220 format_sort_force: Force the given format_sort. see "Sorting Formats"
221 for more details.
222 prefer_free_formats: Whether to prefer video formats with free containers
223 over non-free ones of same quality.
224 allow_multiple_video_streams: Allow multiple video streams to be merged
225 into a single file
226 allow_multiple_audio_streams: Allow multiple audio streams to be merged
227 into a single file
228 check_formats Whether to test if the formats are downloadable.
229 Can be True (check all), False (check none),
230 'selected' (check selected formats),
231 or None (check only if requested by extractor)
232 paths: Dictionary of output paths. The allowed keys are 'home'
233 'temp' and the keys of OUTTMPL_TYPES (in utils.py)
234 outtmpl: Dictionary of templates for output names. Allowed keys
235 are 'default' and the keys of OUTTMPL_TYPES (in utils.py).
236 For compatibility with youtube-dl, a single string can also be used
237 outtmpl_na_placeholder: Placeholder for unavailable meta fields.
238 restrictfilenames: Do not allow "&" and spaces in file names
239 trim_file_name: Limit length of filename (extension excluded)
240 windowsfilenames: Force the filenames to be windows compatible
241 ignoreerrors: Do not stop on download/postprocessing errors.
242 Can be 'only_download' to ignore only download errors.
243 Default is 'only_download' for CLI, but False for API
244 skip_playlist_after_errors: Number of allowed failures until the rest of
245 the playlist is skipped
246 allowed_extractors: List of regexes to match against extractor names that are allowed
247 overwrites: Overwrite all video and metadata files if True,
248 overwrite only non-video files if None
249 and don't overwrite any file if False
250 For compatibility with youtube-dl,
251 "nooverwrites" may also be used instead
252 playlist_items: Specific indices of playlist to download.
253 playlistrandom: Download playlist items in random order.
254 lazy_playlist: Process playlist entries as they are received.
255 matchtitle: Download only matching titles.
256 rejecttitle: Reject downloads for matching titles.
257 logger: Log messages to a logging.Logger instance.
258 logtostderr: Print everything to stderr instead of stdout.
259 consoletitle: Display progress in console window's titlebar.
260 writedescription: Write the video description to a .description file
261 writeinfojson: Write the video description to a .info.json file
262 clean_infojson: Remove internal metadata from the infojson
263 getcomments: Extract video comments. This will not be written to disk
264 unless writeinfojson is also given
265 writeannotations: Write the video annotations to a .annotations.xml file
266 writethumbnail: Write the thumbnail image to a file
267 allow_playlist_files: Whether to write playlists' description, infojson etc
268 also to disk when using the 'write*' options
269 write_all_thumbnails: Write all thumbnail formats to files
270 writelink: Write an internet shortcut file, depending on the
271 current platform (.url/.webloc/.desktop)
272 writeurllink: Write a Windows internet shortcut file (.url)
273 writewebloclink: Write a macOS internet shortcut file (.webloc)
274 writedesktoplink: Write a Linux internet shortcut file (.desktop)
275 writesubtitles: Write the video subtitles to a file
276 writeautomaticsub: Write the automatically generated subtitles to a file
277 listsubtitles: Lists all available subtitles for the video
278 subtitlesformat: The format code for subtitles
279 subtitleslangs: List of languages of the subtitles to download (can be regex).
280 The list may contain "all" to refer to all the available
281 subtitles. The language can be prefixed with a "-" to
282 exclude it from the requested languages, e.g. ['all', '-live_chat']
283 keepvideo: Keep the video file after post-processing
284 daterange: A utils.DateRange object, download only if the upload_date is in the range.
285 skip_download: Skip the actual download of the video file
286 cachedir: Location of the cache files in the filesystem.
287 False to disable filesystem cache.
288 noplaylist: Download single video instead of a playlist if in doubt.
289 age_limit: An integer representing the user's age in years.
290 Unsuitable videos for the given age are skipped.
291 min_views: An integer representing the minimum view count the video
292 must have in order to not be skipped.
293 Videos without view count information are always
294 downloaded. None for no limit.
295 max_views: An integer representing the maximum view count.
296 Videos that are more popular than that are not
297 downloaded.
298 Videos without view count information are always
299 downloaded. None for no limit.
300 download_archive: A set, or the name of a file where all downloads are recorded.
301 Videos already present in the file are not downloaded again.
302 break_on_existing: Stop the download process after attempting to download a
303 file that is in the archive.
304 break_per_url: Whether break_on_reject and break_on_existing
305 should act on each input URL as opposed to for the entire queue
306 cookiefile: File name or text stream from where cookies should be read and dumped to
307 cookiesfrombrowser: A tuple containing the name of the browser, the profile
308 name/path from where cookies are loaded, the name of the keyring,
309 and the container name, e.g. ('chrome', ) or
310 ('vivaldi', 'default', 'BASICTEXT') or ('firefox', 'default', None, 'Meta')
311 legacyserverconnect: Explicitly allow HTTPS connection to servers that do not
312 support RFC 5746 secure renegotiation
313 nocheckcertificate: Do not verify SSL certificates
314 client_certificate: Path to client certificate file in PEM format. May include the private key
315 client_certificate_key: Path to private key file for client certificate
316 client_certificate_password: Password for client certificate private key, if encrypted.
317 If not provided and the key is encrypted, yt-dlp will ask interactively
318 prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
319 (Only supported by some extractors)
320 enable_file_urls: Enable file:// URLs. This is disabled by default for security reasons.
321 http_headers: A dictionary of custom headers to be used for all requests
322 proxy: URL of the proxy server to use
323 geo_verification_proxy: URL of the proxy to use for IP address verification
324 on geo-restricted sites.
325 socket_timeout: Time to wait for unresponsive hosts, in seconds
326 bidi_workaround: Work around buggy terminals without bidirectional text
327 support, using fridibi
328 debug_printtraffic:Print out sent and received HTTP traffic
329 default_search: Prepend this string if an input url is not valid.
330 'auto' for elaborate guessing
331 encoding: Use this encoding instead of the system-specified.
332 extract_flat: Whether to resolve and process url_results further
333 * False: Always process. Default for API
334 * True: Never process
335 * 'in_playlist': Do not process inside playlist/multi_video
336 * 'discard': Always process, but don't return the result
337 from inside playlist/multi_video
338 * 'discard_in_playlist': Same as "discard", but only for
339 playlists (not multi_video). Default for CLI
340 wait_for_video: If given, wait for scheduled streams to become available.
341 The value should be a tuple containing the range
342 (min_secs, max_secs) to wait between retries
343 postprocessors: A list of dictionaries, each with an entry
344 * key: The name of the postprocessor. See
345 yt_dlp/postprocessor/__init__.py for a list.
346 * when: When to run the postprocessor. Allowed values are
347 the entries of utils.POSTPROCESS_WHEN
348 Assumed to be 'post_process' if not given
349 progress_hooks: A list of functions that get called on download
350 progress, with a dictionary with the entries
351 * status: One of "downloading", "error", or "finished".
352 Check this first and ignore unknown values.
353 * info_dict: The extracted info_dict
354
355 If status is one of "downloading", or "finished", the
356 following properties may also be present:
357 * filename: The final filename (always present)
358 * tmpfilename: The filename we're currently writing to
359 * downloaded_bytes: Bytes on disk
360 * total_bytes: Size of the whole file, None if unknown
361 * total_bytes_estimate: Guess of the eventual file size,
362 None if unavailable.
363 * elapsed: The number of seconds since download started.
364 * eta: The estimated time in seconds, None if unknown
365 * speed: The download speed in bytes/second, None if
366 unknown
367 * fragment_index: The counter of the currently
368 downloaded video fragment.
369 * fragment_count: The number of fragments (= individual
370 files that will be merged)
371
372 Progress hooks are guaranteed to be called at least once
373 (with status "finished") if the download is successful.
374 postprocessor_hooks: A list of functions that get called on postprocessing
375 progress, with a dictionary with the entries
376 * status: One of "started", "processing", or "finished".
377 Check this first and ignore unknown values.
378 * postprocessor: Name of the postprocessor
379 * info_dict: The extracted info_dict
380
381 Progress hooks are guaranteed to be called at least twice
382 (with status "started" and "finished") if the processing is successful.
383 merge_output_format: "/" separated list of extensions to use when merging formats.
384 final_ext: Expected final extension; used to detect when the file was
385 already downloaded and converted
386 fixup: Automatically correct known faults of the file.
387 One of:
388 - "never": do nothing
389 - "warn": only emit a warning
390 - "detect_or_warn": check whether we can do anything
391 about it, warn otherwise (default)
392 source_address: Client-side IP address to bind to.
393 sleep_interval_requests: Number of seconds to sleep between requests
394 during extraction
395 sleep_interval: Number of seconds to sleep before each download when
396 used alone or a lower bound of a range for randomized
397 sleep before each download (minimum possible number
398 of seconds to sleep) when used along with
399 max_sleep_interval.
400 max_sleep_interval:Upper bound of a range for randomized sleep before each
401 download (maximum possible number of seconds to sleep).
402 Must only be used along with sleep_interval.
403 Actual sleep time will be a random float from range
404 [sleep_interval; max_sleep_interval].
405 sleep_interval_subtitles: Number of seconds to sleep before each subtitle download
406 listformats: Print an overview of available video formats and exit.
407 list_thumbnails: Print a table of all thumbnails and exit.
408 match_filter: A function that gets called for every video with the signature
409 (info_dict, *, incomplete: bool) -> Optional[str]
410 For backward compatibility with youtube-dl, the signature
411 (info_dict) -> Optional[str] is also allowed.
412 - If it returns a message, the video is ignored.
413 - If it returns None, the video is downloaded.
414 - If it returns utils.NO_DEFAULT, the user is interactively
415 asked whether to download the video.
416 - Raise utils.DownloadCancelled(msg) to abort remaining
417 downloads when a video is rejected.
418 match_filter_func in utils.py is one example for this.
419 color: A Dictionary with output stream names as keys
420 and their respective color policy as values.
421 Can also just be a single color policy,
422 in which case it applies to all outputs.
423 Valid stream names are 'stdout' and 'stderr'.
424 Valid color policies are one of 'always', 'auto', 'no_color' or 'never'.
425 geo_bypass: Bypass geographic restriction via faking X-Forwarded-For
426 HTTP header
427 geo_bypass_country:
428 Two-letter ISO 3166-2 country code that will be used for
429 explicit geographic restriction bypassing via faking
430 X-Forwarded-For HTTP header
431 geo_bypass_ip_block:
432 IP range in CIDR notation that will be used similarly to
433 geo_bypass_country
434 external_downloader: A dictionary of protocol keys and the executable of the
435 external downloader to use for it. The allowed protocols
436 are default|http|ftp|m3u8|dash|rtsp|rtmp|mms.
437 Set the value to 'native' to use the native downloader
438 compat_opts: Compatibility options. See "Differences in default behavior".
439 The following options do not work when used through the API:
440 filename, abort-on-error, multistreams, no-live-chat, format-sort
441 no-clean-infojson, no-playlist-metafiles, no-keep-subs, no-attach-info-json.
442 Refer __init__.py for their implementation
443 progress_template: Dictionary of templates for progress outputs.
444 Allowed keys are 'download', 'postprocess',
445 'download-title' (console title) and 'postprocess-title'.
446 The template is mapped on a dictionary with keys 'progress' and 'info'
447 retry_sleep_functions: Dictionary of functions that takes the number of attempts
448 as argument and returns the time to sleep in seconds.
449 Allowed keys are 'http', 'fragment', 'file_access'
450 download_ranges: A callback function that gets called for every video with
451 the signature (info_dict, ydl) -> Iterable[Section].
452 Only the returned sections will be downloaded.
453 Each Section is a dict with the following keys:
454 * start_time: Start time of the section in seconds
455 * end_time: End time of the section in seconds
456 * title: Section title (Optional)
457 * index: Section number (Optional)
458 force_keyframes_at_cuts: Re-encode the video when downloading ranges to get precise cuts
459 noprogress: Do not print the progress bar
460 live_from_start: Whether to download livestreams videos from the start
461
462 The following parameters are not used by YoutubeDL itself, they are used by
463 the downloader (see yt_dlp/downloader/common.py):
464 nopart, updatetime, buffersize, ratelimit, throttledratelimit, min_filesize,
465 max_filesize, test, noresizebuffer, retries, file_access_retries, fragment_retries,
466 continuedl, xattr_set_filesize, hls_use_mpegts, http_chunk_size,
467 external_downloader_args, concurrent_fragment_downloads.
468
469 The following options are used by the post processors:
470 ffmpeg_location: Location of the ffmpeg/avconv binary; either the path
471 to the binary or its containing directory.
472 postprocessor_args: A dictionary of postprocessor/executable keys (in lower case)
473 and a list of additional command-line arguments for the
474 postprocessor/executable. The dict can also have "PP+EXE" keys
475 which are used when the given exe is used by the given PP.
476 Use 'default' as the name for arguments to passed to all PP
477 For compatibility with youtube-dl, a single list of args
478 can also be used
479
480 The following options are used by the extractors:
481 extractor_retries: Number of times to retry for known errors (default: 3)
482 dynamic_mpd: Whether to process dynamic DASH manifests (default: True)
483 hls_split_discontinuity: Split HLS playlists to different formats at
484 discontinuities such as ad breaks (default: False)
485 extractor_args: A dictionary of arguments to be passed to the extractors.
486 See "EXTRACTOR ARGUMENTS" for details.
487 E.g. {'youtube': {'skip': ['dash', 'hls']}}
488 mark_watched: Mark videos watched (even with --simulate). Only for YouTube
489
490 The following options are deprecated and may be removed in the future:
491
492 break_on_reject: Stop the download process when encountering a video that
493 has been filtered out.
494 - `raise DownloadCancelled(msg)` in match_filter instead
495 force_generic_extractor: Force downloader to use the generic extractor
496 - Use allowed_extractors = ['generic', 'default']
497 playliststart: - Use playlist_items
498 Playlist item to start at.
499 playlistend: - Use playlist_items
500 Playlist item to end at.
501 playlistreverse: - Use playlist_items
502 Download playlist items in reverse order.
503 forceurl: - Use forceprint
504 Force printing final URL.
505 forcetitle: - Use forceprint
506 Force printing title.
507 forceid: - Use forceprint
508 Force printing ID.
509 forcethumbnail: - Use forceprint
510 Force printing thumbnail URL.
511 forcedescription: - Use forceprint
512 Force printing description.
513 forcefilename: - Use forceprint
514 Force printing final filename.
515 forceduration: - Use forceprint
516 Force printing duration.
517 allsubtitles: - Use subtitleslangs = ['all']
518 Downloads all the subtitles of the video
519 (requires writesubtitles or writeautomaticsub)
520 include_ads: - Doesn't work
521 Download ads as well
522 call_home: - Not implemented
523 Boolean, true iff we are allowed to contact the
524 yt-dlp servers for debugging.
525 post_hooks: - Register a custom postprocessor
526 A list of functions that get called as the final step
527 for each video file, after all postprocessors have been
528 called. The filename will be passed as the only argument.
529 hls_prefer_native: - Use external_downloader = {'m3u8': 'native'} or {'m3u8': 'ffmpeg'}.
530 Use the native HLS downloader instead of ffmpeg/avconv
531 if True, otherwise use ffmpeg/avconv if False, otherwise
532 use downloader suggested by extractor if None.
533 prefer_ffmpeg: - avconv support is deprecated
534 If False, use avconv instead of ffmpeg if both are available,
535 otherwise prefer ffmpeg.
536 youtube_include_dash_manifest: - Use extractor_args
537 If True (default), DASH manifests and related
538 data will be downloaded and processed by extractor.
539 You can reduce network I/O by disabling it if you don't
540 care about DASH. (only for youtube)
541 youtube_include_hls_manifest: - Use extractor_args
542 If True (default), HLS manifests and related
543 data will be downloaded and processed by extractor.
544 You can reduce network I/O by disabling it if you don't
545 care about HLS. (only for youtube)
546 no_color: Same as `color='no_color'`
547 """
548
549 _NUMERIC_FIELDS = {
550 'width', 'height', 'asr', 'audio_channels', 'fps',
551 'tbr', 'abr', 'vbr', 'filesize', 'filesize_approx',
552 'timestamp', 'release_timestamp',
553 'duration', 'view_count', 'like_count', 'dislike_count', 'repost_count',
554 'average_rating', 'comment_count', 'age_limit',
555 'start_time', 'end_time',
556 'chapter_number', 'season_number', 'episode_number',
557 'track_number', 'disc_number', 'release_year',
558 }
559
560 _format_fields = {
561 # NB: Keep in sync with the docstring of extractor/common.py
562 'url', 'manifest_url', 'manifest_stream_number', 'ext', 'format', 'format_id', 'format_note',
563 'width', 'height', 'aspect_ratio', 'resolution', 'dynamic_range', 'tbr', 'abr', 'acodec', 'asr', 'audio_channels',
564 'vbr', 'fps', 'vcodec', 'container', 'filesize', 'filesize_approx', 'rows', 'columns',
565 'player_url', 'protocol', 'fragment_base_url', 'fragments', 'is_from_start',
566 'preference', 'language', 'language_preference', 'quality', 'source_preference',
567 'http_headers', 'stretched_ratio', 'no_resume', 'has_drm', 'extra_param_to_segment_url', 'hls_aes', 'downloader_options',
568 'page_url', 'app', 'play_path', 'tc_url', 'flash_version', 'rtmp_live', 'rtmp_conn', 'rtmp_protocol', 'rtmp_real_time'
569 }
570 _format_selection_exts = {
571 'audio': set(MEDIA_EXTENSIONS.common_audio),
572 'video': set(MEDIA_EXTENSIONS.common_video + ('3gp', )),
573 'storyboards': set(MEDIA_EXTENSIONS.storyboards),
574 }
575
576 def __init__(self, params=None, auto_init=True):
577 """Create a FileDownloader object with the given options.
578 @param auto_init Whether to load the default extractors and print header (if verbose).
579 Set to 'no_verbose_header' to not print the header
580 """
581 if params is None:
582 params = {}
583 self.params = params
584 self._ies = {}
585 self._ies_instances = {}
586 self._pps = {k: [] for k in POSTPROCESS_WHEN}
587 self._printed_messages = set()
588 self._first_webpage_request = True
589 self._post_hooks = []
590 self._progress_hooks = []
591 self._postprocessor_hooks = []
592 self._download_retcode = 0
593 self._num_downloads = 0
594 self._num_videos = 0
595 self._playlist_level = 0
596 self._playlist_urls = set()
597 self.cache = Cache(self)
598
599 stdout = sys.stderr if self.params.get('logtostderr') else sys.stdout
600 self._out_files = Namespace(
601 out=stdout,
602 error=sys.stderr,
603 screen=sys.stderr if self.params.get('quiet') else stdout,
604 console=None if compat_os_name == 'nt' else next(
605 filter(supports_terminal_sequences, (sys.stderr, sys.stdout)), None)
606 )
607
608 try:
609 windows_enable_vt_mode()
610 except Exception as e:
611 self.write_debug(f'Failed to enable VT mode: {e}')
612
613 if self.params.get('no_color'):
614 if self.params.get('color') is not None:
615 self.report_warning('Overwriting params from "color" with "no_color"')
616 self.params['color'] = 'no_color'
617
618 term_allow_color = os.environ.get('TERM', '').lower() != 'dumb'
619
620 def process_color_policy(stream):
621 stream_name = {sys.stdout: 'stdout', sys.stderr: 'stderr'}[stream]
622 policy = traverse_obj(self.params, ('color', (stream_name, None), {str}), get_all=False)
623 if policy in ('auto', None):
624 return term_allow_color and supports_terminal_sequences(stream)
625 assert policy in ('always', 'never', 'no_color')
626 return {'always': True, 'never': False}.get(policy, policy)
627
628 self._allow_colors = Namespace(**{
629 name: process_color_policy(stream)
630 for name, stream in self._out_files.items_ if name != 'console'
631 })
632
633 # The code is left like this to be reused for future deprecations
634 MIN_SUPPORTED, MIN_RECOMMENDED = (3, 7), (3, 7)
635 current_version = sys.version_info[:2]
636 if current_version < MIN_RECOMMENDED:
637 msg = ('Support for Python version %d.%d has been deprecated. '
638 'See https://github.com/yt-dlp/yt-dlp/issues/3764 for more details.'
639 '\n You will no longer receive updates on this version')
640 if current_version < MIN_SUPPORTED:
641 msg = 'Python version %d.%d is no longer supported'
642 self.deprecated_feature(
643 f'{msg}! Please update to Python %d.%d or above' % (*current_version, *MIN_RECOMMENDED))
644
645 if self.params.get('allow_unplayable_formats'):
646 self.report_warning(
647 f'You have asked for {self._format_err("UNPLAYABLE", self.Styles.EMPHASIS)} formats to be listed/downloaded. '
648 'This is a developer option intended for debugging. \n'
649 ' If you experience any issues while using this option, '
650 f'{self._format_err("DO NOT", self.Styles.ERROR)} open a bug report')
651
652 if self.params.get('bidi_workaround', False):
653 try:
654 import pty
655 master, slave = pty.openpty()
656 width = shutil.get_terminal_size().columns
657 width_args = [] if width is None else ['-w', str(width)]
658 sp_kwargs = {'stdin': subprocess.PIPE, 'stdout': slave, 'stderr': self._out_files.error}
659 try:
660 self._output_process = Popen(['bidiv'] + width_args, **sp_kwargs)
661 except OSError:
662 self._output_process = Popen(['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
663 self._output_channel = os.fdopen(master, 'rb')
664 except OSError as ose:
665 if ose.errno == errno.ENOENT:
666 self.report_warning(
667 'Could not find fribidi executable, ignoring --bidi-workaround. '
668 'Make sure that fribidi is an executable file in one of the directories in your $PATH.')
669 else:
670 raise
671
672 self.params['compat_opts'] = set(self.params.get('compat_opts', ()))
673 if auto_init and auto_init != 'no_verbose_header':
674 self.print_debug_header()
675
676 def check_deprecated(param, option, suggestion):
677 if self.params.get(param) is not None:
678 self.report_warning(f'{option} is deprecated. Use {suggestion} instead')
679 return True
680 return False
681
682 if check_deprecated('cn_verification_proxy', '--cn-verification-proxy', '--geo-verification-proxy'):
683 if self.params.get('geo_verification_proxy') is None:
684 self.params['geo_verification_proxy'] = self.params['cn_verification_proxy']
685
686 check_deprecated('autonumber', '--auto-number', '-o "%(autonumber)s-%(title)s.%(ext)s"')
687 check_deprecated('usetitle', '--title', '-o "%(title)s-%(id)s.%(ext)s"')
688 check_deprecated('useid', '--id', '-o "%(id)s.%(ext)s"')
689
690 for msg in self.params.get('_warnings', []):
691 self.report_warning(msg)
692 for msg in self.params.get('_deprecation_warnings', []):
693 self.deprecated_feature(msg)
694
695 if 'list-formats' in self.params['compat_opts']:
696 self.params['listformats_table'] = False
697
698 if 'overwrites' not in self.params and self.params.get('nooverwrites') is not None:
699 # nooverwrites was unnecessarily changed to overwrites
700 # in 0c3d0f51778b153f65c21906031c2e091fcfb641
701 # This ensures compatibility with both keys
702 self.params['overwrites'] = not self.params['nooverwrites']
703 elif self.params.get('overwrites') is None:
704 self.params.pop('overwrites', None)
705 else:
706 self.params['nooverwrites'] = not self.params['overwrites']
707
708 if self.params.get('simulate') is None and any((
709 self.params.get('list_thumbnails'),
710 self.params.get('listformats'),
711 self.params.get('listsubtitles'),
712 )):
713 self.params['simulate'] = 'list_only'
714
715 self.params.setdefault('forceprint', {})
716 self.params.setdefault('print_to_file', {})
717
718 # Compatibility with older syntax
719 if not isinstance(params['forceprint'], dict):
720 self.params['forceprint'] = {'video': params['forceprint']}
721
722 if auto_init:
723 self.add_default_info_extractors()
724
725 if (sys.platform != 'win32'
726 and sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968']
727 and not self.params.get('restrictfilenames', False)):
728 # Unicode filesystem API will throw errors (#1474, #13027)
729 self.report_warning(
730 'Assuming --restrict-filenames since file system encoding '
731 'cannot encode all characters. '
732 'Set the LC_ALL environment variable to fix this.')
733 self.params['restrictfilenames'] = True
734
735 self._parse_outtmpl()
736
737 # Creating format selector here allows us to catch syntax errors before the extraction
738 self.format_selector = (
739 self.params.get('format') if self.params.get('format') in (None, '-')
740 else self.params['format'] if callable(self.params['format'])
741 else self.build_format_selector(self.params['format']))
742
743 # Set http_headers defaults according to std_headers
744 self.params['http_headers'] = merge_headers(std_headers, self.params.get('http_headers', {}))
745
746 hooks = {
747 'post_hooks': self.add_post_hook,
748 'progress_hooks': self.add_progress_hook,
749 'postprocessor_hooks': self.add_postprocessor_hook,
750 }
751 for opt, fn in hooks.items():
752 for ph in self.params.get(opt, []):
753 fn(ph)
754
755 for pp_def_raw in self.params.get('postprocessors', []):
756 pp_def = dict(pp_def_raw)
757 when = pp_def.pop('when', 'post_process')
758 self.add_post_processor(
759 get_postprocessor(pp_def.pop('key'))(self, **pp_def),
760 when=when)
761
762 self._setup_opener()
763
764 def preload_download_archive(fn):
765 """Preload the archive, if any is specified"""
766 archive = set()
767 if fn is None:
768 return archive
769 elif not is_path_like(fn):
770 return fn
771
772 self.write_debug(f'Loading archive file {fn!r}')
773 try:
774 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
775 for line in archive_file:
776 archive.add(line.strip())
777 except OSError as ioe:
778 if ioe.errno != errno.ENOENT:
779 raise
780 return archive
781
782 self.archive = preload_download_archive(self.params.get('download_archive'))
783
784 def warn_if_short_id(self, argv):
785 # short YouTube ID starting with dash?
786 idxs = [
787 i for i, a in enumerate(argv)
788 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
789 if idxs:
790 correct_argv = (
791 ['yt-dlp']
792 + [a for i, a in enumerate(argv) if i not in idxs]
793 + ['--'] + [argv[i] for i in idxs]
794 )
795 self.report_warning(
796 'Long argument string detected. '
797 'Use -- to separate parameters and URLs, like this:\n%s' %
798 args_to_str(correct_argv))
799
800 def add_info_extractor(self, ie):
801 """Add an InfoExtractor object to the end of the list."""
802 ie_key = ie.ie_key()
803 self._ies[ie_key] = ie
804 if not isinstance(ie, type):
805 self._ies_instances[ie_key] = ie
806 ie.set_downloader(self)
807
808 def get_info_extractor(self, ie_key):
809 """
810 Get an instance of an IE with name ie_key, it will try to get one from
811 the _ies list, if there's no instance it will create a new one and add
812 it to the extractor list.
813 """
814 ie = self._ies_instances.get(ie_key)
815 if ie is None:
816 ie = get_info_extractor(ie_key)()
817 self.add_info_extractor(ie)
818 return ie
819
820 def add_default_info_extractors(self):
821 """
822 Add the InfoExtractors returned by gen_extractors to the end of the list
823 """
824 all_ies = {ie.IE_NAME.lower(): ie for ie in gen_extractor_classes()}
825 all_ies['end'] = UnsupportedURLIE()
826 try:
827 ie_names = orderedSet_from_options(
828 self.params.get('allowed_extractors', ['default']), {
829 'all': list(all_ies),
830 'default': [name for name, ie in all_ies.items() if ie._ENABLED],
831 }, use_regex=True)
832 except re.error as e:
833 raise ValueError(f'Wrong regex for allowed_extractors: {e.pattern}')
834 for name in ie_names:
835 self.add_info_extractor(all_ies[name])
836 self.write_debug(f'Loaded {len(ie_names)} extractors')
837
838 def add_post_processor(self, pp, when='post_process'):
839 """Add a PostProcessor object to the end of the chain."""
840 assert when in POSTPROCESS_WHEN, f'Invalid when={when}'
841 self._pps[when].append(pp)
842 pp.set_downloader(self)
843
844 def add_post_hook(self, ph):
845 """Add the post hook"""
846 self._post_hooks.append(ph)
847
848 def add_progress_hook(self, ph):
849 """Add the download progress hook"""
850 self._progress_hooks.append(ph)
851
852 def add_postprocessor_hook(self, ph):
853 """Add the postprocessing progress hook"""
854 self._postprocessor_hooks.append(ph)
855 for pps in self._pps.values():
856 for pp in pps:
857 pp.add_progress_hook(ph)
858
859 def _bidi_workaround(self, message):
860 if not hasattr(self, '_output_channel'):
861 return message
862
863 assert hasattr(self, '_output_process')
864 assert isinstance(message, str)
865 line_count = message.count('\n') + 1
866 self._output_process.stdin.write((message + '\n').encode())
867 self._output_process.stdin.flush()
868 res = ''.join(self._output_channel.readline().decode()
869 for _ in range(line_count))
870 return res[:-len('\n')]
871
872 def _write_string(self, message, out=None, only_once=False):
873 if only_once:
874 if message in self._printed_messages:
875 return
876 self._printed_messages.add(message)
877 write_string(message, out=out, encoding=self.params.get('encoding'))
878
879 def to_stdout(self, message, skip_eol=False, quiet=None):
880 """Print message to stdout"""
881 if quiet is not None:
882 self.deprecation_warning('"YoutubeDL.to_stdout" no longer accepts the argument quiet. '
883 'Use "YoutubeDL.to_screen" instead')
884 if skip_eol is not False:
885 self.deprecation_warning('"YoutubeDL.to_stdout" no longer accepts the argument skip_eol. '
886 'Use "YoutubeDL.to_screen" instead')
887 self._write_string(f'{self._bidi_workaround(message)}\n', self._out_files.out)
888
889 def to_screen(self, message, skip_eol=False, quiet=None, only_once=False):
890 """Print message to screen if not in quiet mode"""
891 if self.params.get('logger'):
892 self.params['logger'].debug(message)
893 return
894 if (self.params.get('quiet') if quiet is None else quiet) and not self.params.get('verbose'):
895 return
896 self._write_string(
897 '%s%s' % (self._bidi_workaround(message), ('' if skip_eol else '\n')),
898 self._out_files.screen, only_once=only_once)
899
900 def to_stderr(self, message, only_once=False):
901 """Print message to stderr"""
902 assert isinstance(message, str)
903 if self.params.get('logger'):
904 self.params['logger'].error(message)
905 else:
906 self._write_string(f'{self._bidi_workaround(message)}\n', self._out_files.error, only_once=only_once)
907
908 def _send_console_code(self, code):
909 if compat_os_name == 'nt' or not self._out_files.console:
910 return
911 self._write_string(code, self._out_files.console)
912
913 def to_console_title(self, message):
914 if not self.params.get('consoletitle', False):
915 return
916 message = remove_terminal_sequences(message)
917 if compat_os_name == 'nt':
918 if ctypes.windll.kernel32.GetConsoleWindow():
919 # c_wchar_p() might not be necessary if `message` is
920 # already of type unicode()
921 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
922 else:
923 self._send_console_code(f'\033]0;{message}\007')
924
925 def save_console_title(self):
926 if not self.params.get('consoletitle') or self.params.get('simulate'):
927 return
928 self._send_console_code('\033[22;0t') # Save the title on stack
929
930 def restore_console_title(self):
931 if not self.params.get('consoletitle') or self.params.get('simulate'):
932 return
933 self._send_console_code('\033[23;0t') # Restore the title from stack
934
935 def __enter__(self):
936 self.save_console_title()
937 return self
938
939 def __exit__(self, *args):
940 self.restore_console_title()
941
942 if self.params.get('cookiefile') is not None:
943 self.cookiejar.save(ignore_discard=True, ignore_expires=True)
944
945 def trouble(self, message=None, tb=None, is_error=True):
946 """Determine action to take when a download problem appears.
947
948 Depending on if the downloader has been configured to ignore
949 download errors or not, this method may throw an exception or
950 not when errors are found, after printing the message.
951
952 @param tb If given, is additional traceback information
953 @param is_error Whether to raise error according to ignorerrors
954 """
955 if message is not None:
956 self.to_stderr(message)
957 if self.params.get('verbose'):
958 if tb is None:
959 if sys.exc_info()[0]: # if .trouble has been called from an except block
960 tb = ''
961 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
962 tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
963 tb += encode_compat_str(traceback.format_exc())
964 else:
965 tb_data = traceback.format_list(traceback.extract_stack())
966 tb = ''.join(tb_data)
967 if tb:
968 self.to_stderr(tb)
969 if not is_error:
970 return
971 if not self.params.get('ignoreerrors'):
972 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
973 exc_info = sys.exc_info()[1].exc_info
974 else:
975 exc_info = sys.exc_info()
976 raise DownloadError(message, exc_info)
977 self._download_retcode = 1
978
979 Styles = Namespace(
980 HEADERS='yellow',
981 EMPHASIS='light blue',
982 FILENAME='green',
983 ID='green',
984 DELIM='blue',
985 ERROR='red',
986 BAD_FORMAT='light red',
987 WARNING='yellow',
988 SUPPRESS='light black',
989 )
990
991 def _format_text(self, handle, allow_colors, text, f, fallback=None, *, test_encoding=False):
992 text = str(text)
993 if test_encoding:
994 original_text = text
995 # handle.encoding can be None. See https://github.com/yt-dlp/yt-dlp/issues/2711
996 encoding = self.params.get('encoding') or getattr(handle, 'encoding', None) or 'ascii'
997 text = text.encode(encoding, 'ignore').decode(encoding)
998 if fallback is not None and text != original_text:
999 text = fallback
1000 return format_text(text, f) if allow_colors is True else text if fallback is None else fallback
1001
1002 def _format_out(self, *args, **kwargs):
1003 return self._format_text(self._out_files.out, self._allow_colors.out, *args, **kwargs)
1004
1005 def _format_screen(self, *args, **kwargs):
1006 return self._format_text(self._out_files.screen, self._allow_colors.screen, *args, **kwargs)
1007
1008 def _format_err(self, *args, **kwargs):
1009 return self._format_text(self._out_files.error, self._allow_colors.error, *args, **kwargs)
1010
1011 def report_warning(self, message, only_once=False):
1012 '''
1013 Print the message to stderr, it will be prefixed with 'WARNING:'
1014 If stderr is a tty file the 'WARNING:' will be colored
1015 '''
1016 if self.params.get('logger') is not None:
1017 self.params['logger'].warning(message)
1018 else:
1019 if self.params.get('no_warnings'):
1020 return
1021 self.to_stderr(f'{self._format_err("WARNING:", self.Styles.WARNING)} {message}', only_once)
1022
1023 def deprecation_warning(self, message, *, stacklevel=0):
1024 deprecation_warning(
1025 message, stacklevel=stacklevel + 1, printer=self.report_error, is_error=False)
1026
1027 def deprecated_feature(self, message):
1028 if self.params.get('logger') is not None:
1029 self.params['logger'].warning(f'Deprecated Feature: {message}')
1030 self.to_stderr(f'{self._format_err("Deprecated Feature:", self.Styles.ERROR)} {message}', True)
1031
1032 def report_error(self, message, *args, **kwargs):
1033 '''
1034 Do the same as trouble, but prefixes the message with 'ERROR:', colored
1035 in red if stderr is a tty file.
1036 '''
1037 self.trouble(f'{self._format_err("ERROR:", self.Styles.ERROR)} {message}', *args, **kwargs)
1038
1039 def write_debug(self, message, only_once=False):
1040 '''Log debug message or Print message to stderr'''
1041 if not self.params.get('verbose', False):
1042 return
1043 message = f'[debug] {message}'
1044 if self.params.get('logger'):
1045 self.params['logger'].debug(message)
1046 else:
1047 self.to_stderr(message, only_once)
1048
1049 def report_file_already_downloaded(self, file_name):
1050 """Report file has already been fully downloaded."""
1051 try:
1052 self.to_screen('[download] %s has already been downloaded' % file_name)
1053 except UnicodeEncodeError:
1054 self.to_screen('[download] The file has already been downloaded')
1055
1056 def report_file_delete(self, file_name):
1057 """Report that existing file will be deleted."""
1058 try:
1059 self.to_screen('Deleting existing file %s' % file_name)
1060 except UnicodeEncodeError:
1061 self.to_screen('Deleting existing file')
1062
1063 def raise_no_formats(self, info, forced=False, *, msg=None):
1064 has_drm = info.get('_has_drm')
1065 ignored, expected = self.params.get('ignore_no_formats_error'), bool(msg)
1066 msg = msg or has_drm and 'This video is DRM protected' or 'No video formats found!'
1067 if forced or not ignored:
1068 raise ExtractorError(msg, video_id=info['id'], ie=info['extractor'],
1069 expected=has_drm or ignored or expected)
1070 else:
1071 self.report_warning(msg)
1072
1073 def parse_outtmpl(self):
1074 self.deprecation_warning('"YoutubeDL.parse_outtmpl" is deprecated and may be removed in a future version')
1075 self._parse_outtmpl()
1076 return self.params['outtmpl']
1077
1078 def _parse_outtmpl(self):
1079 sanitize = IDENTITY
1080 if self.params.get('restrictfilenames'): # Remove spaces in the default template
1081 sanitize = lambda x: x.replace(' - ', ' ').replace(' ', '-')
1082
1083 outtmpl = self.params.setdefault('outtmpl', {})
1084 if not isinstance(outtmpl, dict):
1085 self.params['outtmpl'] = outtmpl = {'default': outtmpl}
1086 outtmpl.update({k: sanitize(v) for k, v in DEFAULT_OUTTMPL.items() if outtmpl.get(k) is None})
1087
1088 def get_output_path(self, dir_type='', filename=None):
1089 paths = self.params.get('paths', {})
1090 assert isinstance(paths, dict), '"paths" parameter must be a dictionary'
1091 path = os.path.join(
1092 expand_path(paths.get('home', '').strip()),
1093 expand_path(paths.get(dir_type, '').strip()) if dir_type else '',
1094 filename or '')
1095 return sanitize_path(path, force=self.params.get('windowsfilenames'))
1096
1097 @staticmethod
1098 def _outtmpl_expandpath(outtmpl):
1099 # expand_path translates '%%' into '%' and '$$' into '$'
1100 # correspondingly that is not what we want since we need to keep
1101 # '%%' intact for template dict substitution step. Working around
1102 # with boundary-alike separator hack.
1103 sep = ''.join(random.choices(string.ascii_letters, k=32))
1104 outtmpl = outtmpl.replace('%%', f'%{sep}%').replace('$$', f'${sep}$')
1105
1106 # outtmpl should be expand_path'ed before template dict substitution
1107 # because meta fields may contain env variables we don't want to
1108 # be expanded. E.g. for outtmpl "%(title)s.%(ext)s" and
1109 # title "Hello $PATH", we don't want `$PATH` to be expanded.
1110 return expand_path(outtmpl).replace(sep, '')
1111
1112 @staticmethod
1113 def escape_outtmpl(outtmpl):
1114 ''' Escape any remaining strings like %s, %abc% etc. '''
1115 return re.sub(
1116 STR_FORMAT_RE_TMPL.format('', '(?![%(\0])'),
1117 lambda mobj: ('' if mobj.group('has_key') else '%') + mobj.group(0),
1118 outtmpl)
1119
1120 @classmethod
1121 def validate_outtmpl(cls, outtmpl):
1122 ''' @return None or Exception object '''
1123 outtmpl = re.sub(
1124 STR_FORMAT_RE_TMPL.format('[^)]*', '[ljhqBUDS]'),
1125 lambda mobj: f'{mobj.group(0)[:-1]}s',
1126 cls._outtmpl_expandpath(outtmpl))
1127 try:
1128 cls.escape_outtmpl(outtmpl) % collections.defaultdict(int)
1129 return None
1130 except ValueError as err:
1131 return err
1132
1133 @staticmethod
1134 def _copy_infodict(info_dict):
1135 info_dict = dict(info_dict)
1136 info_dict.pop('__postprocessors', None)
1137 info_dict.pop('__pending_error', None)
1138 return info_dict
1139
1140 def prepare_outtmpl(self, outtmpl, info_dict, sanitize=False):
1141 """ Make the outtmpl and info_dict suitable for substitution: ydl.escape_outtmpl(outtmpl) % info_dict
1142 @param sanitize Whether to sanitize the output as a filename.
1143 For backward compatibility, a function can also be passed
1144 """
1145
1146 info_dict.setdefault('epoch', int(time.time())) # keep epoch consistent once set
1147
1148 info_dict = self._copy_infodict(info_dict)
1149 info_dict['duration_string'] = ( # %(duration>%H-%M-%S)s is wrong if duration > 24hrs
1150 formatSeconds(info_dict['duration'], '-' if sanitize else ':')
1151 if info_dict.get('duration', None) is not None
1152 else None)
1153 info_dict['autonumber'] = int(self.params.get('autonumber_start', 1) - 1 + self._num_downloads)
1154 info_dict['video_autonumber'] = self._num_videos
1155 if info_dict.get('resolution') is None:
1156 info_dict['resolution'] = self.format_resolution(info_dict, default=None)
1157
1158 # For fields playlist_index, playlist_autonumber and autonumber convert all occurrences
1159 # of %(field)s to %(field)0Nd for backward compatibility
1160 field_size_compat_map = {
1161 'playlist_index': number_of_digits(info_dict.get('__last_playlist_index') or 0),
1162 'playlist_autonumber': number_of_digits(info_dict.get('n_entries') or 0),
1163 'autonumber': self.params.get('autonumber_size') or 5,
1164 }
1165
1166 TMPL_DICT = {}
1167 EXTERNAL_FORMAT_RE = re.compile(STR_FORMAT_RE_TMPL.format('[^)]*', f'[{STR_FORMAT_TYPES}ljhqBUDS]'))
1168 MATH_FUNCTIONS = {
1169 '+': float.__add__,
1170 '-': float.__sub__,
1171 }
1172 # Field is of the form key1.key2...
1173 # where keys (except first) can be string, int, slice or "{field, ...}"
1174 FIELD_INNER_RE = r'(?:\w+|%(num)s|%(num)s?(?::%(num)s?){1,2})' % {'num': r'(?:-?\d+)'}
1175 FIELD_RE = r'\w*(?:\.(?:%(inner)s|{%(field)s(?:,%(field)s)*}))*' % {
1176 'inner': FIELD_INNER_RE,
1177 'field': rf'\w*(?:\.{FIELD_INNER_RE})*'
1178 }
1179 MATH_FIELD_RE = rf'(?:{FIELD_RE}|-?{NUMBER_RE})'
1180 MATH_OPERATORS_RE = r'(?:%s)' % '|'.join(map(re.escape, MATH_FUNCTIONS.keys()))
1181 INTERNAL_FORMAT_RE = re.compile(rf'''(?xs)
1182 (?P<negate>-)?
1183 (?P<fields>{FIELD_RE})
1184 (?P<maths>(?:{MATH_OPERATORS_RE}{MATH_FIELD_RE})*)
1185 (?:>(?P<strf_format>.+?))?
1186 (?P<remaining>
1187 (?P<alternate>(?<!\\),[^|&)]+)?
1188 (?:&(?P<replacement>.*?))?
1189 (?:\|(?P<default>.*?))?
1190 )$''')
1191
1192 def _traverse_infodict(fields):
1193 fields = [f for x in re.split(r'\.({.+?})\.?', fields)
1194 for f in ([x] if x.startswith('{') else x.split('.'))]
1195 for i in (0, -1):
1196 if fields and not fields[i]:
1197 fields.pop(i)
1198
1199 for i, f in enumerate(fields):
1200 if not f.startswith('{'):
1201 continue
1202 assert f.endswith('}'), f'No closing brace for {f} in {fields}'
1203 fields[i] = {k: k.split('.') for k in f[1:-1].split(',')}
1204
1205 return traverse_obj(info_dict, fields, is_user_input=True, traverse_string=True)
1206
1207 def get_value(mdict):
1208 # Object traversal
1209 value = _traverse_infodict(mdict['fields'])
1210 # Negative
1211 if mdict['negate']:
1212 value = float_or_none(value)
1213 if value is not None:
1214 value *= -1
1215 # Do maths
1216 offset_key = mdict['maths']
1217 if offset_key:
1218 value = float_or_none(value)
1219 operator = None
1220 while offset_key:
1221 item = re.match(
1222 MATH_FIELD_RE if operator else MATH_OPERATORS_RE,
1223 offset_key).group(0)
1224 offset_key = offset_key[len(item):]
1225 if operator is None:
1226 operator = MATH_FUNCTIONS[item]
1227 continue
1228 item, multiplier = (item[1:], -1) if item[0] == '-' else (item, 1)
1229 offset = float_or_none(item)
1230 if offset is None:
1231 offset = float_or_none(_traverse_infodict(item))
1232 try:
1233 value = operator(value, multiplier * offset)
1234 except (TypeError, ZeroDivisionError):
1235 return None
1236 operator = None
1237 # Datetime formatting
1238 if mdict['strf_format']:
1239 value = strftime_or_none(value, mdict['strf_format'].replace('\\,', ','))
1240
1241 # XXX: Workaround for https://github.com/yt-dlp/yt-dlp/issues/4485
1242 if sanitize and value == '':
1243 value = None
1244 return value
1245
1246 na = self.params.get('outtmpl_na_placeholder', 'NA')
1247
1248 def filename_sanitizer(key, value, restricted=self.params.get('restrictfilenames')):
1249 return sanitize_filename(str(value), restricted=restricted, is_id=(
1250 bool(re.search(r'(^|[_.])id(\.|$)', key))
1251 if 'filename-sanitization' in self.params['compat_opts']
1252 else NO_DEFAULT))
1253
1254 sanitizer = sanitize if callable(sanitize) else filename_sanitizer
1255 sanitize = bool(sanitize)
1256
1257 def _dumpjson_default(obj):
1258 if isinstance(obj, (set, LazyList)):
1259 return list(obj)
1260 return repr(obj)
1261
1262 class _ReplacementFormatter(string.Formatter):
1263 def get_field(self, field_name, args, kwargs):
1264 if field_name.isdigit():
1265 return args[0], -1
1266 raise ValueError('Unsupported field')
1267
1268 replacement_formatter = _ReplacementFormatter()
1269
1270 def create_key(outer_mobj):
1271 if not outer_mobj.group('has_key'):
1272 return outer_mobj.group(0)
1273 key = outer_mobj.group('key')
1274 mobj = re.match(INTERNAL_FORMAT_RE, key)
1275 value, replacement, default, last_field = None, None, na, ''
1276 while mobj:
1277 mobj = mobj.groupdict()
1278 default = mobj['default'] if mobj['default'] is not None else default
1279 value = get_value(mobj)
1280 last_field, replacement = mobj['fields'], mobj['replacement']
1281 if value is None and mobj['alternate']:
1282 mobj = re.match(INTERNAL_FORMAT_RE, mobj['remaining'][1:])
1283 else:
1284 break
1285
1286 fmt = outer_mobj.group('format')
1287 if fmt == 's' and value is not None and last_field in field_size_compat_map.keys():
1288 fmt = f'0{field_size_compat_map[last_field]:d}d'
1289
1290 if None not in (value, replacement):
1291 try:
1292 value = replacement_formatter.format(replacement, value)
1293 except ValueError:
1294 value, default = None, na
1295
1296 flags = outer_mobj.group('conversion') or ''
1297 str_fmt = f'{fmt[:-1]}s'
1298 if value is None:
1299 value, fmt = default, 's'
1300 elif fmt[-1] == 'l': # list
1301 delim = '\n' if '#' in flags else ', '
1302 value, fmt = delim.join(map(str, variadic(value, allowed_types=(str, bytes)))), str_fmt
1303 elif fmt[-1] == 'j': # json
1304 value, fmt = json.dumps(
1305 value, default=_dumpjson_default,
1306 indent=4 if '#' in flags else None, ensure_ascii='+' not in flags), str_fmt
1307 elif fmt[-1] == 'h': # html
1308 value, fmt = escapeHTML(str(value)), str_fmt
1309 elif fmt[-1] == 'q': # quoted
1310 value = map(str, variadic(value) if '#' in flags else [value])
1311 value, fmt = ' '.join(map(compat_shlex_quote, value)), str_fmt
1312 elif fmt[-1] == 'B': # bytes
1313 value = f'%{str_fmt}'.encode() % str(value).encode()
1314 value, fmt = value.decode('utf-8', 'ignore'), 's'
1315 elif fmt[-1] == 'U': # unicode normalized
1316 value, fmt = unicodedata.normalize(
1317 # "+" = compatibility equivalence, "#" = NFD
1318 'NF%s%s' % ('K' if '+' in flags else '', 'D' if '#' in flags else 'C'),
1319 value), str_fmt
1320 elif fmt[-1] == 'D': # decimal suffix
1321 num_fmt, fmt = fmt[:-1].replace('#', ''), 's'
1322 value = format_decimal_suffix(value, f'%{num_fmt}f%s' if num_fmt else '%d%s',
1323 factor=1024 if '#' in flags else 1000)
1324 elif fmt[-1] == 'S': # filename sanitization
1325 value, fmt = filename_sanitizer(last_field, value, restricted='#' in flags), str_fmt
1326 elif fmt[-1] == 'c':
1327 if value:
1328 value = str(value)[0]
1329 else:
1330 fmt = str_fmt
1331 elif fmt[-1] not in 'rsa': # numeric
1332 value = float_or_none(value)
1333 if value is None:
1334 value, fmt = default, 's'
1335
1336 if sanitize:
1337 # If value is an object, sanitize might convert it to a string
1338 # So we convert it to repr first
1339 if fmt[-1] == 'r':
1340 value, fmt = repr(value), str_fmt
1341 elif fmt[-1] == 'a':
1342 value, fmt = ascii(value), str_fmt
1343 if fmt[-1] in 'csra':
1344 value = sanitizer(last_field, value)
1345
1346 key = '%s\0%s' % (key.replace('%', '%\0'), outer_mobj.group('format'))
1347 TMPL_DICT[key] = value
1348 return '{prefix}%({key}){fmt}'.format(key=key, fmt=fmt, prefix=outer_mobj.group('prefix'))
1349
1350 return EXTERNAL_FORMAT_RE.sub(create_key, outtmpl), TMPL_DICT
1351
1352 def evaluate_outtmpl(self, outtmpl, info_dict, *args, **kwargs):
1353 outtmpl, info_dict = self.prepare_outtmpl(outtmpl, info_dict, *args, **kwargs)
1354 return self.escape_outtmpl(outtmpl) % info_dict
1355
1356 def _prepare_filename(self, info_dict, *, outtmpl=None, tmpl_type=None):
1357 assert None in (outtmpl, tmpl_type), 'outtmpl and tmpl_type are mutually exclusive'
1358 if outtmpl is None:
1359 outtmpl = self.params['outtmpl'].get(tmpl_type or 'default', self.params['outtmpl']['default'])
1360 try:
1361 outtmpl = self._outtmpl_expandpath(outtmpl)
1362 filename = self.evaluate_outtmpl(outtmpl, info_dict, True)
1363 if not filename:
1364 return None
1365
1366 if tmpl_type in ('', 'temp'):
1367 final_ext, ext = self.params.get('final_ext'), info_dict.get('ext')
1368 if final_ext and ext and final_ext != ext and filename.endswith(f'.{final_ext}'):
1369 filename = replace_extension(filename, ext, final_ext)
1370 elif tmpl_type:
1371 force_ext = OUTTMPL_TYPES[tmpl_type]
1372 if force_ext:
1373 filename = replace_extension(filename, force_ext, info_dict.get('ext'))
1374
1375 # https://github.com/blackjack4494/youtube-dlc/issues/85
1376 trim_file_name = self.params.get('trim_file_name', False)
1377 if trim_file_name:
1378 no_ext, *ext = filename.rsplit('.', 2)
1379 filename = join_nonempty(no_ext[:trim_file_name], *ext, delim='.')
1380
1381 return filename
1382 except ValueError as err:
1383 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
1384 return None
1385
1386 def prepare_filename(self, info_dict, dir_type='', *, outtmpl=None, warn=False):
1387 """Generate the output filename"""
1388 if outtmpl:
1389 assert not dir_type, 'outtmpl and dir_type are mutually exclusive'
1390 dir_type = None
1391 filename = self._prepare_filename(info_dict, tmpl_type=dir_type, outtmpl=outtmpl)
1392 if not filename and dir_type not in ('', 'temp'):
1393 return ''
1394
1395 if warn:
1396 if not self.params.get('paths'):
1397 pass
1398 elif filename == '-':
1399 self.report_warning('--paths is ignored when an outputting to stdout', only_once=True)
1400 elif os.path.isabs(filename):
1401 self.report_warning('--paths is ignored since an absolute path is given in output template', only_once=True)
1402 if filename == '-' or not filename:
1403 return filename
1404
1405 return self.get_output_path(dir_type, filename)
1406
1407 def _match_entry(self, info_dict, incomplete=False, silent=False):
1408 """Returns None if the file should be downloaded"""
1409 _type = 'video' if 'playlist-match-filter' in self.params['compat_opts'] else info_dict.get('_type', 'video')
1410 assert incomplete or _type == 'video', 'Only video result can be considered complete'
1411
1412 video_title = info_dict.get('title', info_dict.get('id', 'entry'))
1413
1414 def check_filter():
1415 if _type in ('playlist', 'multi_video'):
1416 return
1417 elif _type in ('url', 'url_transparent') and not try_call(
1418 lambda: self.get_info_extractor(info_dict['ie_key']).is_single_video(info_dict['url'])):
1419 return
1420
1421 if 'title' in info_dict:
1422 # This can happen when we're just evaluating the playlist
1423 title = info_dict['title']
1424 matchtitle = self.params.get('matchtitle', False)
1425 if matchtitle:
1426 if not re.search(matchtitle, title, re.IGNORECASE):
1427 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
1428 rejecttitle = self.params.get('rejecttitle', False)
1429 if rejecttitle:
1430 if re.search(rejecttitle, title, re.IGNORECASE):
1431 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
1432
1433 date = info_dict.get('upload_date')
1434 if date is not None:
1435 dateRange = self.params.get('daterange', DateRange())
1436 if date not in dateRange:
1437 return f'{date_from_str(date).isoformat()} upload date is not in range {dateRange}'
1438 view_count = info_dict.get('view_count')
1439 if view_count is not None:
1440 min_views = self.params.get('min_views')
1441 if min_views is not None and view_count < min_views:
1442 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
1443 max_views = self.params.get('max_views')
1444 if max_views is not None and view_count > max_views:
1445 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
1446 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
1447 return 'Skipping "%s" because it is age restricted' % video_title
1448
1449 match_filter = self.params.get('match_filter')
1450 if match_filter is None:
1451 return None
1452
1453 cancelled = None
1454 try:
1455 try:
1456 ret = match_filter(info_dict, incomplete=incomplete)
1457 except TypeError:
1458 # For backward compatibility
1459 ret = None if incomplete else match_filter(info_dict)
1460 except DownloadCancelled as err:
1461 if err.msg is not NO_DEFAULT:
1462 raise
1463 ret, cancelled = err.msg, err
1464
1465 if ret is NO_DEFAULT:
1466 while True:
1467 filename = self._format_screen(self.prepare_filename(info_dict), self.Styles.FILENAME)
1468 reply = input(self._format_screen(
1469 f'Download "{filename}"? (Y/n): ', self.Styles.EMPHASIS)).lower().strip()
1470 if reply in {'y', ''}:
1471 return None
1472 elif reply == 'n':
1473 if cancelled:
1474 raise type(cancelled)(f'Skipping {video_title}')
1475 return f'Skipping {video_title}'
1476 return ret
1477
1478 if self.in_download_archive(info_dict):
1479 reason = '%s has already been recorded in the archive' % video_title
1480 break_opt, break_err = 'break_on_existing', ExistingVideoReached
1481 else:
1482 try:
1483 reason = check_filter()
1484 except DownloadCancelled as e:
1485 reason, break_opt, break_err = e.msg, 'match_filter', type(e)
1486 else:
1487 break_opt, break_err = 'break_on_reject', RejectedVideoReached
1488 if reason is not None:
1489 if not silent:
1490 self.to_screen('[download] ' + reason)
1491 if self.params.get(break_opt, False):
1492 raise break_err()
1493 return reason
1494
1495 @staticmethod
1496 def add_extra_info(info_dict, extra_info):
1497 '''Set the keys from extra_info in info dict if they are missing'''
1498 for key, value in extra_info.items():
1499 info_dict.setdefault(key, value)
1500
1501 def extract_info(self, url, download=True, ie_key=None, extra_info=None,
1502 process=True, force_generic_extractor=False):
1503 """
1504 Extract and return the information dictionary of the URL
1505
1506 Arguments:
1507 @param url URL to extract
1508
1509 Keyword arguments:
1510 @param download Whether to download videos
1511 @param process Whether to resolve all unresolved references (URLs, playlist items).
1512 Must be True for download to work
1513 @param ie_key Use only the extractor with this key
1514
1515 @param extra_info Dictionary containing the extra values to add to the info (For internal use only)
1516 @force_generic_extractor Force using the generic extractor (Deprecated; use ie_key='Generic')
1517 """
1518
1519 if extra_info is None:
1520 extra_info = {}
1521
1522 if not ie_key and force_generic_extractor:
1523 ie_key = 'Generic'
1524
1525 if ie_key:
1526 ies = {ie_key: self._ies[ie_key]} if ie_key in self._ies else {}
1527 else:
1528 ies = self._ies
1529
1530 for key, ie in ies.items():
1531 if not ie.suitable(url):
1532 continue
1533
1534 if not ie.working():
1535 self.report_warning('The program functionality for this site has been marked as broken, '
1536 'and will probably not work.')
1537
1538 temp_id = ie.get_temp_id(url)
1539 if temp_id is not None and self.in_download_archive({'id': temp_id, 'ie_key': key}):
1540 self.to_screen(f'[{key}] {temp_id}: has already been recorded in the archive')
1541 if self.params.get('break_on_existing', False):
1542 raise ExistingVideoReached()
1543 break
1544 return self.__extract_info(url, self.get_info_extractor(key), download, extra_info, process)
1545 else:
1546 extractors_restricted = self.params.get('allowed_extractors') not in (None, ['default'])
1547 self.report_error(f'No suitable extractor{format_field(ie_key, None, " (%s)")} found for URL {url}',
1548 tb=False if extractors_restricted else None)
1549
1550 def _handle_extraction_exceptions(func):
1551 @functools.wraps(func)
1552 def wrapper(self, *args, **kwargs):
1553 while True:
1554 try:
1555 return func(self, *args, **kwargs)
1556 except (DownloadCancelled, LazyList.IndexError, PagedList.IndexError):
1557 raise
1558 except ReExtractInfo as e:
1559 if e.expected:
1560 self.to_screen(f'{e}; Re-extracting data')
1561 else:
1562 self.to_stderr('\r')
1563 self.report_warning(f'{e}; Re-extracting data')
1564 continue
1565 except GeoRestrictedError as e:
1566 msg = e.msg
1567 if e.countries:
1568 msg += '\nThis video is available in %s.' % ', '.join(
1569 map(ISO3166Utils.short2full, e.countries))
1570 msg += '\nYou might want to use a VPN or a proxy server (with --proxy) to workaround.'
1571 self.report_error(msg)
1572 except ExtractorError as e: # An error we somewhat expected
1573 self.report_error(str(e), e.format_traceback())
1574 except Exception as e:
1575 if self.params.get('ignoreerrors'):
1576 self.report_error(str(e), tb=encode_compat_str(traceback.format_exc()))
1577 else:
1578 raise
1579 break
1580 return wrapper
1581
1582 def _wait_for_video(self, ie_result={}):
1583 if (not self.params.get('wait_for_video')
1584 or ie_result.get('_type', 'video') != 'video'
1585 or ie_result.get('formats') or ie_result.get('url')):
1586 return
1587
1588 format_dur = lambda dur: '%02d:%02d:%02d' % timetuple_from_msec(dur * 1000)[:-1]
1589 last_msg = ''
1590
1591 def progress(msg):
1592 nonlocal last_msg
1593 full_msg = f'{msg}\n'
1594 if not self.params.get('noprogress'):
1595 full_msg = msg + ' ' * (len(last_msg) - len(msg)) + '\r'
1596 elif last_msg:
1597 return
1598 self.to_screen(full_msg, skip_eol=True)
1599 last_msg = msg
1600
1601 min_wait, max_wait = self.params.get('wait_for_video')
1602 diff = try_get(ie_result, lambda x: x['release_timestamp'] - time.time())
1603 if diff is None and ie_result.get('live_status') == 'is_upcoming':
1604 diff = round(random.uniform(min_wait, max_wait) if (max_wait and min_wait) else (max_wait or min_wait), 0)
1605 self.report_warning('Release time of video is not known')
1606 elif ie_result and (diff or 0) <= 0:
1607 self.report_warning('Video should already be available according to extracted info')
1608 diff = min(max(diff or 0, min_wait or 0), max_wait or float('inf'))
1609 self.to_screen(f'[wait] Waiting for {format_dur(diff)} - Press Ctrl+C to try now')
1610
1611 wait_till = time.time() + diff
1612 try:
1613 while True:
1614 diff = wait_till - time.time()
1615 if diff <= 0:
1616 progress('')
1617 raise ReExtractInfo('[wait] Wait period ended', expected=True)
1618 progress(f'[wait] Remaining time until next attempt: {self._format_screen(format_dur(diff), self.Styles.EMPHASIS)}')
1619 time.sleep(1)
1620 except KeyboardInterrupt:
1621 progress('')
1622 raise ReExtractInfo('[wait] Interrupted by user', expected=True)
1623 except BaseException as e:
1624 if not isinstance(e, ReExtractInfo):
1625 self.to_screen('')
1626 raise
1627
1628 @_handle_extraction_exceptions
1629 def __extract_info(self, url, ie, download, extra_info, process):
1630 try:
1631 ie_result = ie.extract(url)
1632 except UserNotLive as e:
1633 if process:
1634 if self.params.get('wait_for_video'):
1635 self.report_warning(e)
1636 self._wait_for_video()
1637 raise
1638 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
1639 self.report_warning(f'Extractor {ie.IE_NAME} returned nothing{bug_reports_message()}')
1640 return
1641 if isinstance(ie_result, list):
1642 # Backwards compatibility: old IE result format
1643 ie_result = {
1644 '_type': 'compat_list',
1645 'entries': ie_result,
1646 }
1647 if extra_info.get('original_url'):
1648 ie_result.setdefault('original_url', extra_info['original_url'])
1649 self.add_default_extra_info(ie_result, ie, url)
1650 if process:
1651 self._wait_for_video(ie_result)
1652 return self.process_ie_result(ie_result, download, extra_info)
1653 else:
1654 return ie_result
1655
1656 def add_default_extra_info(self, ie_result, ie, url):
1657 if url is not None:
1658 self.add_extra_info(ie_result, {
1659 'webpage_url': url,
1660 'original_url': url,
1661 })
1662 webpage_url = ie_result.get('webpage_url')
1663 if webpage_url:
1664 self.add_extra_info(ie_result, {
1665 'webpage_url_basename': url_basename(webpage_url),
1666 'webpage_url_domain': get_domain(webpage_url),
1667 })
1668 if ie is not None:
1669 self.add_extra_info(ie_result, {
1670 'extractor': ie.IE_NAME,
1671 'extractor_key': ie.ie_key(),
1672 })
1673
1674 def process_ie_result(self, ie_result, download=True, extra_info=None):
1675 """
1676 Take the result of the ie(may be modified) and resolve all unresolved
1677 references (URLs, playlist items).
1678
1679 It will also download the videos if 'download'.
1680 Returns the resolved ie_result.
1681 """
1682 if extra_info is None:
1683 extra_info = {}
1684 result_type = ie_result.get('_type', 'video')
1685
1686 if result_type in ('url', 'url_transparent'):
1687 ie_result['url'] = sanitize_url(
1688 ie_result['url'], scheme='http' if self.params.get('prefer_insecure') else 'https')
1689 if ie_result.get('original_url') and not extra_info.get('original_url'):
1690 extra_info = {'original_url': ie_result['original_url'], **extra_info}
1691
1692 extract_flat = self.params.get('extract_flat', False)
1693 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info)
1694 or extract_flat is True):
1695 info_copy = ie_result.copy()
1696 ie = try_get(ie_result.get('ie_key'), self.get_info_extractor)
1697 if ie and not ie_result.get('id'):
1698 info_copy['id'] = ie.get_temp_id(ie_result['url'])
1699 self.add_default_extra_info(info_copy, ie, ie_result['url'])
1700 self.add_extra_info(info_copy, extra_info)
1701 info_copy, _ = self.pre_process(info_copy)
1702 self._fill_common_fields(info_copy, False)
1703 self.__forced_printings(info_copy)
1704 self._raise_pending_errors(info_copy)
1705 if self.params.get('force_write_download_archive', False):
1706 self.record_download_archive(info_copy)
1707 return ie_result
1708
1709 if result_type == 'video':
1710 self.add_extra_info(ie_result, extra_info)
1711 ie_result = self.process_video_result(ie_result, download=download)
1712 self._raise_pending_errors(ie_result)
1713 additional_urls = (ie_result or {}).get('additional_urls')
1714 if additional_urls:
1715 # TODO: Improve MetadataParserPP to allow setting a list
1716 if isinstance(additional_urls, str):
1717 additional_urls = [additional_urls]
1718 self.to_screen(
1719 '[info] %s: %d additional URL(s) requested' % (ie_result['id'], len(additional_urls)))
1720 self.write_debug('Additional URLs: "%s"' % '", "'.join(additional_urls))
1721 ie_result['additional_entries'] = [
1722 self.extract_info(
1723 url, download, extra_info=extra_info,
1724 force_generic_extractor=self.params.get('force_generic_extractor'))
1725 for url in additional_urls
1726 ]
1727 return ie_result
1728 elif result_type == 'url':
1729 # We have to add extra_info to the results because it may be
1730 # contained in a playlist
1731 return self.extract_info(
1732 ie_result['url'], download,
1733 ie_key=ie_result.get('ie_key'),
1734 extra_info=extra_info)
1735 elif result_type == 'url_transparent':
1736 # Use the information from the embedding page
1737 info = self.extract_info(
1738 ie_result['url'], ie_key=ie_result.get('ie_key'),
1739 extra_info=extra_info, download=False, process=False)
1740
1741 # extract_info may return None when ignoreerrors is enabled and
1742 # extraction failed with an error, don't crash and return early
1743 # in this case
1744 if not info:
1745 return info
1746
1747 exempted_fields = {'_type', 'url', 'ie_key'}
1748 if not ie_result.get('section_end') and ie_result.get('section_start') is None:
1749 # For video clips, the id etc of the clip extractor should be used
1750 exempted_fields |= {'id', 'extractor', 'extractor_key'}
1751
1752 new_result = info.copy()
1753 new_result.update(filter_dict(ie_result, lambda k, v: v is not None and k not in exempted_fields))
1754
1755 # Extracted info may not be a video result (i.e.
1756 # info.get('_type', 'video') != video) but rather an url or
1757 # url_transparent. In such cases outer metadata (from ie_result)
1758 # should be propagated to inner one (info). For this to happen
1759 # _type of info should be overridden with url_transparent. This
1760 # fixes issue from https://github.com/ytdl-org/youtube-dl/pull/11163.
1761 if new_result.get('_type') == 'url':
1762 new_result['_type'] = 'url_transparent'
1763
1764 return self.process_ie_result(
1765 new_result, download=download, extra_info=extra_info)
1766 elif result_type in ('playlist', 'multi_video'):
1767 # Protect from infinite recursion due to recursively nested playlists
1768 # (see https://github.com/ytdl-org/youtube-dl/issues/27833)
1769 webpage_url = ie_result.get('webpage_url') # Playlists maynot have webpage_url
1770 if webpage_url and webpage_url in self._playlist_urls:
1771 self.to_screen(
1772 '[download] Skipping already downloaded playlist: %s'
1773 % ie_result.get('title') or ie_result.get('id'))
1774 return
1775
1776 self._playlist_level += 1
1777 self._playlist_urls.add(webpage_url)
1778 self._fill_common_fields(ie_result, False)
1779 self._sanitize_thumbnails(ie_result)
1780 try:
1781 return self.__process_playlist(ie_result, download)
1782 finally:
1783 self._playlist_level -= 1
1784 if not self._playlist_level:
1785 self._playlist_urls.clear()
1786 elif result_type == 'compat_list':
1787 self.report_warning(
1788 'Extractor %s returned a compat_list result. '
1789 'It needs to be updated.' % ie_result.get('extractor'))
1790
1791 def _fixup(r):
1792 self.add_extra_info(r, {
1793 'extractor': ie_result['extractor'],
1794 'webpage_url': ie_result['webpage_url'],
1795 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1796 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1797 'extractor_key': ie_result['extractor_key'],
1798 })
1799 return r
1800 ie_result['entries'] = [
1801 self.process_ie_result(_fixup(r), download, extra_info)
1802 for r in ie_result['entries']
1803 ]
1804 return ie_result
1805 else:
1806 raise Exception('Invalid result type: %s' % result_type)
1807
1808 def _ensure_dir_exists(self, path):
1809 return make_dir(path, self.report_error)
1810
1811 @staticmethod
1812 def _playlist_infodict(ie_result, strict=False, **kwargs):
1813 info = {
1814 'playlist_count': ie_result.get('playlist_count'),
1815 'playlist': ie_result.get('title') or ie_result.get('id'),
1816 'playlist_id': ie_result.get('id'),
1817 'playlist_title': ie_result.get('title'),
1818 'playlist_uploader': ie_result.get('uploader'),
1819 'playlist_uploader_id': ie_result.get('uploader_id'),
1820 **kwargs,
1821 }
1822 if strict:
1823 return info
1824 if ie_result.get('webpage_url'):
1825 info.update({
1826 'webpage_url': ie_result['webpage_url'],
1827 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1828 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1829 })
1830 return {
1831 **info,
1832 'playlist_index': 0,
1833 '__last_playlist_index': max(ie_result.get('requested_entries') or (0, 0)),
1834 'extractor': ie_result['extractor'],
1835 'extractor_key': ie_result['extractor_key'],
1836 }
1837
1838 def __process_playlist(self, ie_result, download):
1839 """Process each entry in the playlist"""
1840 assert ie_result['_type'] in ('playlist', 'multi_video')
1841
1842 common_info = self._playlist_infodict(ie_result, strict=True)
1843 title = common_info.get('playlist') or '<Untitled>'
1844 if self._match_entry(common_info, incomplete=True) is not None:
1845 return
1846 self.to_screen(f'[download] Downloading {ie_result["_type"]}: {title}')
1847
1848 all_entries = PlaylistEntries(self, ie_result)
1849 entries = orderedSet(all_entries.get_requested_items(), lazy=True)
1850
1851 lazy = self.params.get('lazy_playlist')
1852 if lazy:
1853 resolved_entries, n_entries = [], 'N/A'
1854 ie_result['requested_entries'], ie_result['entries'] = None, None
1855 else:
1856 entries = resolved_entries = list(entries)
1857 n_entries = len(resolved_entries)
1858 ie_result['requested_entries'], ie_result['entries'] = tuple(zip(*resolved_entries)) or ([], [])
1859 if not ie_result.get('playlist_count'):
1860 # Better to do this after potentially exhausting entries
1861 ie_result['playlist_count'] = all_entries.get_full_count()
1862
1863 extra = self._playlist_infodict(ie_result, n_entries=int_or_none(n_entries))
1864 ie_copy = collections.ChainMap(ie_result, extra)
1865
1866 _infojson_written = False
1867 write_playlist_files = self.params.get('allow_playlist_files', True)
1868 if write_playlist_files and self.params.get('list_thumbnails'):
1869 self.list_thumbnails(ie_result)
1870 if write_playlist_files and not self.params.get('simulate'):
1871 _infojson_written = self._write_info_json(
1872 'playlist', ie_result, self.prepare_filename(ie_copy, 'pl_infojson'))
1873 if _infojson_written is None:
1874 return
1875 if self._write_description('playlist', ie_result,
1876 self.prepare_filename(ie_copy, 'pl_description')) is None:
1877 return
1878 # TODO: This should be passed to ThumbnailsConvertor if necessary
1879 self._write_thumbnails('playlist', ie_result, self.prepare_filename(ie_copy, 'pl_thumbnail'))
1880
1881 if lazy:
1882 if self.params.get('playlistreverse') or self.params.get('playlistrandom'):
1883 self.report_warning('playlistreverse and playlistrandom are not supported with lazy_playlist', only_once=True)
1884 elif self.params.get('playlistreverse'):
1885 entries.reverse()
1886 elif self.params.get('playlistrandom'):
1887 random.shuffle(entries)
1888
1889 self.to_screen(f'[{ie_result["extractor"]}] Playlist {title}: Downloading {n_entries} items'
1890 f'{format_field(ie_result, "playlist_count", " of %s")}')
1891
1892 keep_resolved_entries = self.params.get('extract_flat') != 'discard'
1893 if self.params.get('extract_flat') == 'discard_in_playlist':
1894 keep_resolved_entries = ie_result['_type'] != 'playlist'
1895 if keep_resolved_entries:
1896 self.write_debug('The information of all playlist entries will be held in memory')
1897
1898 failures = 0
1899 max_failures = self.params.get('skip_playlist_after_errors') or float('inf')
1900 for i, (playlist_index, entry) in enumerate(entries):
1901 if lazy:
1902 resolved_entries.append((playlist_index, entry))
1903 if not entry:
1904 continue
1905
1906 entry['__x_forwarded_for_ip'] = ie_result.get('__x_forwarded_for_ip')
1907 if not lazy and 'playlist-index' in self.params['compat_opts']:
1908 playlist_index = ie_result['requested_entries'][i]
1909
1910 entry_copy = collections.ChainMap(entry, {
1911 **common_info,
1912 'n_entries': int_or_none(n_entries),
1913 'playlist_index': playlist_index,
1914 'playlist_autonumber': i + 1,
1915 })
1916
1917 if self._match_entry(entry_copy, incomplete=True) is not None:
1918 # For compatabilty with youtube-dl. See https://github.com/yt-dlp/yt-dlp/issues/4369
1919 resolved_entries[i] = (playlist_index, NO_DEFAULT)
1920 continue
1921
1922 self.to_screen('[download] Downloading item %s of %s' % (
1923 self._format_screen(i + 1, self.Styles.ID), self._format_screen(n_entries, self.Styles.EMPHASIS)))
1924
1925 entry_result = self.__process_iterable_entry(entry, download, collections.ChainMap({
1926 'playlist_index': playlist_index,
1927 'playlist_autonumber': i + 1,
1928 }, extra))
1929 if not entry_result:
1930 failures += 1
1931 if failures >= max_failures:
1932 self.report_error(
1933 f'Skipping the remaining entries in playlist "{title}" since {failures} items failed extraction')
1934 break
1935 if keep_resolved_entries:
1936 resolved_entries[i] = (playlist_index, entry_result)
1937
1938 # Update with processed data
1939 ie_result['entries'] = [e for _, e in resolved_entries if e is not NO_DEFAULT]
1940 ie_result['requested_entries'] = [i for i, e in resolved_entries if e is not NO_DEFAULT]
1941 if ie_result['requested_entries'] == try_call(lambda: list(range(1, ie_result['playlist_count'] + 1))):
1942 # Do not set for full playlist
1943 ie_result.pop('requested_entries')
1944
1945 # Write the updated info to json
1946 if _infojson_written is True and self._write_info_json(
1947 'updated playlist', ie_result,
1948 self.prepare_filename(ie_copy, 'pl_infojson'), overwrite=True) is None:
1949 return
1950
1951 ie_result = self.run_all_pps('playlist', ie_result)
1952 self.to_screen(f'[download] Finished downloading playlist: {title}')
1953 return ie_result
1954
1955 @_handle_extraction_exceptions
1956 def __process_iterable_entry(self, entry, download, extra_info):
1957 return self.process_ie_result(
1958 entry, download=download, extra_info=extra_info)
1959
1960 def _build_format_filter(self, filter_spec):
1961 " Returns a function to filter the formats according to the filter_spec "
1962
1963 OPERATORS = {
1964 '<': operator.lt,
1965 '<=': operator.le,
1966 '>': operator.gt,
1967 '>=': operator.ge,
1968 '=': operator.eq,
1969 '!=': operator.ne,
1970 }
1971 operator_rex = re.compile(r'''(?x)\s*
1972 (?P<key>[\w.-]+)\s*
1973 (?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
1974 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)\s*
1975 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
1976 m = operator_rex.fullmatch(filter_spec)
1977 if m:
1978 try:
1979 comparison_value = int(m.group('value'))
1980 except ValueError:
1981 comparison_value = parse_filesize(m.group('value'))
1982 if comparison_value is None:
1983 comparison_value = parse_filesize(m.group('value') + 'B')
1984 if comparison_value is None:
1985 raise ValueError(
1986 'Invalid value %r in format specification %r' % (
1987 m.group('value'), filter_spec))
1988 op = OPERATORS[m.group('op')]
1989
1990 if not m:
1991 STR_OPERATORS = {
1992 '=': operator.eq,
1993 '^=': lambda attr, value: attr.startswith(value),
1994 '$=': lambda attr, value: attr.endswith(value),
1995 '*=': lambda attr, value: value in attr,
1996 '~=': lambda attr, value: value.search(attr) is not None
1997 }
1998 str_operator_rex = re.compile(r'''(?x)\s*
1999 (?P<key>[a-zA-Z0-9._-]+)\s*
2000 (?P<negation>!\s*)?(?P<op>%s)\s*(?P<none_inclusive>\?\s*)?
2001 (?P<quote>["'])?
2002 (?P<value>(?(quote)(?:(?!(?P=quote))[^\\]|\\.)+|[\w.-]+))
2003 (?(quote)(?P=quote))\s*
2004 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
2005 m = str_operator_rex.fullmatch(filter_spec)
2006 if m:
2007 if m.group('op') == '~=':
2008 comparison_value = re.compile(m.group('value'))
2009 else:
2010 comparison_value = re.sub(r'''\\([\\"'])''', r'\1', m.group('value'))
2011 str_op = STR_OPERATORS[m.group('op')]
2012 if m.group('negation'):
2013 op = lambda attr, value: not str_op(attr, value)
2014 else:
2015 op = str_op
2016
2017 if not m:
2018 raise SyntaxError('Invalid filter specification %r' % filter_spec)
2019
2020 def _filter(f):
2021 actual_value = f.get(m.group('key'))
2022 if actual_value is None:
2023 return m.group('none_inclusive')
2024 return op(actual_value, comparison_value)
2025 return _filter
2026
2027 def _check_formats(self, formats):
2028 for f in formats:
2029 self.to_screen('[info] Testing format %s' % f['format_id'])
2030 path = self.get_output_path('temp')
2031 if not self._ensure_dir_exists(f'{path}/'):
2032 continue
2033 temp_file = tempfile.NamedTemporaryFile(suffix='.tmp', delete=False, dir=path or None)
2034 temp_file.close()
2035 try:
2036 success, _ = self.dl(temp_file.name, f, test=True)
2037 except (DownloadError, OSError, ValueError) + network_exceptions:
2038 success = False
2039 finally:
2040 if os.path.exists(temp_file.name):
2041 try:
2042 os.remove(temp_file.name)
2043 except OSError:
2044 self.report_warning('Unable to delete temporary file "%s"' % temp_file.name)
2045 if success:
2046 yield f
2047 else:
2048 self.to_screen('[info] Unable to download format %s. Skipping...' % f['format_id'])
2049
2050 def _default_format_spec(self, info_dict, download=True):
2051
2052 def can_merge():
2053 merger = FFmpegMergerPP(self)
2054 return merger.available and merger.can_merge()
2055
2056 prefer_best = (
2057 not self.params.get('simulate')
2058 and download
2059 and (
2060 not can_merge()
2061 or info_dict.get('is_live') and not self.params.get('live_from_start')
2062 or self.params['outtmpl']['default'] == '-'))
2063 compat = (
2064 prefer_best
2065 or self.params.get('allow_multiple_audio_streams', False)
2066 or 'format-spec' in self.params['compat_opts'])
2067
2068 return (
2069 'best/bestvideo+bestaudio' if prefer_best
2070 else 'bestvideo*+bestaudio/best' if not compat
2071 else 'bestvideo+bestaudio/best')
2072
2073 def build_format_selector(self, format_spec):
2074 def syntax_error(note, start):
2075 message = (
2076 'Invalid format specification: '
2077 '{}\n\t{}\n\t{}^'.format(note, format_spec, ' ' * start[1]))
2078 return SyntaxError(message)
2079
2080 PICKFIRST = 'PICKFIRST'
2081 MERGE = 'MERGE'
2082 SINGLE = 'SINGLE'
2083 GROUP = 'GROUP'
2084 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
2085
2086 allow_multiple_streams = {'audio': self.params.get('allow_multiple_audio_streams', False),
2087 'video': self.params.get('allow_multiple_video_streams', False)}
2088
2089 def _parse_filter(tokens):
2090 filter_parts = []
2091 for type, string_, start, _, _ in tokens:
2092 if type == tokenize.OP and string_ == ']':
2093 return ''.join(filter_parts)
2094 else:
2095 filter_parts.append(string_)
2096
2097 def _remove_unused_ops(tokens):
2098 # Remove operators that we don't use and join them with the surrounding strings.
2099 # E.g. 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
2100 ALLOWED_OPS = ('/', '+', ',', '(', ')')
2101 last_string, last_start, last_end, last_line = None, None, None, None
2102 for type, string_, start, end, line in tokens:
2103 if type == tokenize.OP and string_ == '[':
2104 if last_string:
2105 yield tokenize.NAME, last_string, last_start, last_end, last_line
2106 last_string = None
2107 yield type, string_, start, end, line
2108 # everything inside brackets will be handled by _parse_filter
2109 for type, string_, start, end, line in tokens:
2110 yield type, string_, start, end, line
2111 if type == tokenize.OP and string_ == ']':
2112 break
2113 elif type == tokenize.OP and string_ in ALLOWED_OPS:
2114 if last_string:
2115 yield tokenize.NAME, last_string, last_start, last_end, last_line
2116 last_string = None
2117 yield type, string_, start, end, line
2118 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
2119 if not last_string:
2120 last_string = string_
2121 last_start = start
2122 last_end = end
2123 else:
2124 last_string += string_
2125 if last_string:
2126 yield tokenize.NAME, last_string, last_start, last_end, last_line
2127
2128 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
2129 selectors = []
2130 current_selector = None
2131 for type, string_, start, _, _ in tokens:
2132 # ENCODING is only defined in python 3.x
2133 if type == getattr(tokenize, 'ENCODING', None):
2134 continue
2135 elif type in [tokenize.NAME, tokenize.NUMBER]:
2136 current_selector = FormatSelector(SINGLE, string_, [])
2137 elif type == tokenize.OP:
2138 if string_ == ')':
2139 if not inside_group:
2140 # ')' will be handled by the parentheses group
2141 tokens.restore_last_token()
2142 break
2143 elif inside_merge and string_ in ['/', ',']:
2144 tokens.restore_last_token()
2145 break
2146 elif inside_choice and string_ == ',':
2147 tokens.restore_last_token()
2148 break
2149 elif string_ == ',':
2150 if not current_selector:
2151 raise syntax_error('"," must follow a format selector', start)
2152 selectors.append(current_selector)
2153 current_selector = None
2154 elif string_ == '/':
2155 if not current_selector:
2156 raise syntax_error('"/" must follow a format selector', start)
2157 first_choice = current_selector
2158 second_choice = _parse_format_selection(tokens, inside_choice=True)
2159 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
2160 elif string_ == '[':
2161 if not current_selector:
2162 current_selector = FormatSelector(SINGLE, 'best', [])
2163 format_filter = _parse_filter(tokens)
2164 current_selector.filters.append(format_filter)
2165 elif string_ == '(':
2166 if current_selector:
2167 raise syntax_error('Unexpected "("', start)
2168 group = _parse_format_selection(tokens, inside_group=True)
2169 current_selector = FormatSelector(GROUP, group, [])
2170 elif string_ == '+':
2171 if not current_selector:
2172 raise syntax_error('Unexpected "+"', start)
2173 selector_1 = current_selector
2174 selector_2 = _parse_format_selection(tokens, inside_merge=True)
2175 if not selector_2:
2176 raise syntax_error('Expected a selector', start)
2177 current_selector = FormatSelector(MERGE, (selector_1, selector_2), [])
2178 else:
2179 raise syntax_error(f'Operator not recognized: "{string_}"', start)
2180 elif type == tokenize.ENDMARKER:
2181 break
2182 if current_selector:
2183 selectors.append(current_selector)
2184 return selectors
2185
2186 def _merge(formats_pair):
2187 format_1, format_2 = formats_pair
2188
2189 formats_info = []
2190 formats_info.extend(format_1.get('requested_formats', (format_1,)))
2191 formats_info.extend(format_2.get('requested_formats', (format_2,)))
2192
2193 if not allow_multiple_streams['video'] or not allow_multiple_streams['audio']:
2194 get_no_more = {'video': False, 'audio': False}
2195 for (i, fmt_info) in enumerate(formats_info):
2196 if fmt_info.get('acodec') == fmt_info.get('vcodec') == 'none':
2197 formats_info.pop(i)
2198 continue
2199 for aud_vid in ['audio', 'video']:
2200 if not allow_multiple_streams[aud_vid] and fmt_info.get(aud_vid[0] + 'codec') != 'none':
2201 if get_no_more[aud_vid]:
2202 formats_info.pop(i)
2203 break
2204 get_no_more[aud_vid] = True
2205
2206 if len(formats_info) == 1:
2207 return formats_info[0]
2208
2209 video_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('vcodec') != 'none']
2210 audio_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('acodec') != 'none']
2211
2212 the_only_video = video_fmts[0] if len(video_fmts) == 1 else None
2213 the_only_audio = audio_fmts[0] if len(audio_fmts) == 1 else None
2214
2215 output_ext = get_compatible_ext(
2216 vcodecs=[f.get('vcodec') for f in video_fmts],
2217 acodecs=[f.get('acodec') for f in audio_fmts],
2218 vexts=[f['ext'] for f in video_fmts],
2219 aexts=[f['ext'] for f in audio_fmts],
2220 preferences=(try_call(lambda: self.params['merge_output_format'].split('/'))
2221 or self.params.get('prefer_free_formats') and ('webm', 'mkv')))
2222
2223 filtered = lambda *keys: filter(None, (traverse_obj(fmt, *keys) for fmt in formats_info))
2224
2225 new_dict = {
2226 'requested_formats': formats_info,
2227 'format': '+'.join(filtered('format')),
2228 'format_id': '+'.join(filtered('format_id')),
2229 'ext': output_ext,
2230 'protocol': '+'.join(map(determine_protocol, formats_info)),
2231 'language': '+'.join(orderedSet(filtered('language'))) or None,
2232 'format_note': '+'.join(orderedSet(filtered('format_note'))) or None,
2233 'filesize_approx': sum(filtered('filesize', 'filesize_approx')) or None,
2234 'tbr': sum(filtered('tbr', 'vbr', 'abr')),
2235 }
2236
2237 if the_only_video:
2238 new_dict.update({
2239 'width': the_only_video.get('width'),
2240 'height': the_only_video.get('height'),
2241 'resolution': the_only_video.get('resolution') or self.format_resolution(the_only_video),
2242 'fps': the_only_video.get('fps'),
2243 'dynamic_range': the_only_video.get('dynamic_range'),
2244 'vcodec': the_only_video.get('vcodec'),
2245 'vbr': the_only_video.get('vbr'),
2246 'stretched_ratio': the_only_video.get('stretched_ratio'),
2247 'aspect_ratio': the_only_video.get('aspect_ratio'),
2248 })
2249
2250 if the_only_audio:
2251 new_dict.update({
2252 'acodec': the_only_audio.get('acodec'),
2253 'abr': the_only_audio.get('abr'),
2254 'asr': the_only_audio.get('asr'),
2255 'audio_channels': the_only_audio.get('audio_channels')
2256 })
2257
2258 return new_dict
2259
2260 def _check_formats(formats):
2261 if (self.params.get('check_formats') is not None
2262 or self.params.get('allow_unplayable_formats')):
2263 yield from formats
2264 return
2265 elif self.params.get('check_formats') == 'selected':
2266 yield from self._check_formats(formats)
2267 return
2268
2269 for f in formats:
2270 if f.get('has_drm'):
2271 yield from self._check_formats([f])
2272 else:
2273 yield f
2274
2275 def _build_selector_function(selector):
2276 if isinstance(selector, list): # ,
2277 fs = [_build_selector_function(s) for s in selector]
2278
2279 def selector_function(ctx):
2280 for f in fs:
2281 yield from f(ctx)
2282 return selector_function
2283
2284 elif selector.type == GROUP: # ()
2285 selector_function = _build_selector_function(selector.selector)
2286
2287 elif selector.type == PICKFIRST: # /
2288 fs = [_build_selector_function(s) for s in selector.selector]
2289
2290 def selector_function(ctx):
2291 for f in fs:
2292 picked_formats = list(f(ctx))
2293 if picked_formats:
2294 return picked_formats
2295 return []
2296
2297 elif selector.type == MERGE: # +
2298 selector_1, selector_2 = map(_build_selector_function, selector.selector)
2299
2300 def selector_function(ctx):
2301 for pair in itertools.product(selector_1(ctx), selector_2(ctx)):
2302 yield _merge(pair)
2303
2304 elif selector.type == SINGLE: # atom
2305 format_spec = selector.selector or 'best'
2306
2307 # TODO: Add allvideo, allaudio etc by generalizing the code with best/worst selector
2308 if format_spec == 'all':
2309 def selector_function(ctx):
2310 yield from _check_formats(ctx['formats'][::-1])
2311 elif format_spec == 'mergeall':
2312 def selector_function(ctx):
2313 formats = list(_check_formats(
2314 f for f in ctx['formats'] if f.get('vcodec') != 'none' or f.get('acodec') != 'none'))
2315 if not formats:
2316 return
2317 merged_format = formats[-1]
2318 for f in formats[-2::-1]:
2319 merged_format = _merge((merged_format, f))
2320 yield merged_format
2321
2322 else:
2323 format_fallback, seperate_fallback, format_reverse, format_idx = False, None, True, 1
2324 mobj = re.match(
2325 r'(?P<bw>best|worst|b|w)(?P<type>video|audio|v|a)?(?P<mod>\*)?(?:\.(?P<n>[1-9]\d*))?$',
2326 format_spec)
2327 if mobj is not None:
2328 format_idx = int_or_none(mobj.group('n'), default=1)
2329 format_reverse = mobj.group('bw')[0] == 'b'
2330 format_type = (mobj.group('type') or [None])[0]
2331 not_format_type = {'v': 'a', 'a': 'v'}.get(format_type)
2332 format_modified = mobj.group('mod') is not None
2333
2334 format_fallback = not format_type and not format_modified # for b, w
2335 _filter_f = (
2336 (lambda f: f.get('%scodec' % format_type) != 'none')
2337 if format_type and format_modified # bv*, ba*, wv*, wa*
2338 else (lambda f: f.get('%scodec' % not_format_type) == 'none')
2339 if format_type # bv, ba, wv, wa
2340 else (lambda f: f.get('vcodec') != 'none' and f.get('acodec') != 'none')
2341 if not format_modified # b, w
2342 else lambda f: True) # b*, w*
2343 filter_f = lambda f: _filter_f(f) and (
2344 f.get('vcodec') != 'none' or f.get('acodec') != 'none')
2345 else:
2346 if format_spec in self._format_selection_exts['audio']:
2347 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none'
2348 elif format_spec in self._format_selection_exts['video']:
2349 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none' and f.get('vcodec') != 'none'
2350 seperate_fallback = lambda f: f.get('ext') == format_spec and f.get('vcodec') != 'none'
2351 elif format_spec in self._format_selection_exts['storyboards']:
2352 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') == 'none' and f.get('vcodec') == 'none'
2353 else:
2354 filter_f = lambda f: f.get('format_id') == format_spec # id
2355
2356 def selector_function(ctx):
2357 formats = list(ctx['formats'])
2358 matches = list(filter(filter_f, formats)) if filter_f is not None else formats
2359 if not matches:
2360 if format_fallback and ctx['incomplete_formats']:
2361 # for extractors with incomplete formats (audio only (soundcloud)
2362 # or video only (imgur)) best/worst will fallback to
2363 # best/worst {video,audio}-only format
2364 matches = formats
2365 elif seperate_fallback and not ctx['has_merged_format']:
2366 # for compatibility with youtube-dl when there is no pre-merged format
2367 matches = list(filter(seperate_fallback, formats))
2368 matches = LazyList(_check_formats(matches[::-1 if format_reverse else 1]))
2369 try:
2370 yield matches[format_idx - 1]
2371 except LazyList.IndexError:
2372 return
2373
2374 filters = [self._build_format_filter(f) for f in selector.filters]
2375
2376 def final_selector(ctx):
2377 ctx_copy = dict(ctx)
2378 for _filter in filters:
2379 ctx_copy['formats'] = list(filter(_filter, ctx_copy['formats']))
2380 return selector_function(ctx_copy)
2381 return final_selector
2382
2383 stream = io.BytesIO(format_spec.encode())
2384 try:
2385 tokens = list(_remove_unused_ops(tokenize.tokenize(stream.readline)))
2386 except tokenize.TokenError:
2387 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
2388
2389 class TokenIterator:
2390 def __init__(self, tokens):
2391 self.tokens = tokens
2392 self.counter = 0
2393
2394 def __iter__(self):
2395 return self
2396
2397 def __next__(self):
2398 if self.counter >= len(self.tokens):
2399 raise StopIteration()
2400 value = self.tokens[self.counter]
2401 self.counter += 1
2402 return value
2403
2404 next = __next__
2405
2406 def restore_last_token(self):
2407 self.counter -= 1
2408
2409 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
2410 return _build_selector_function(parsed_selector)
2411
2412 def _calc_headers(self, info_dict):
2413 res = merge_headers(self.params['http_headers'], info_dict.get('http_headers') or {})
2414 if 'Youtubedl-No-Compression' in res: # deprecated
2415 res.pop('Youtubedl-No-Compression', None)
2416 res['Accept-Encoding'] = 'identity'
2417 cookies = self.cookiejar.get_cookie_header(info_dict['url'])
2418 if cookies:
2419 res['Cookie'] = cookies
2420
2421 if 'X-Forwarded-For' not in res:
2422 x_forwarded_for_ip = info_dict.get('__x_forwarded_for_ip')
2423 if x_forwarded_for_ip:
2424 res['X-Forwarded-For'] = x_forwarded_for_ip
2425
2426 return res
2427
2428 def _calc_cookies(self, url):
2429 self.deprecation_warning('"YoutubeDL._calc_cookies" is deprecated and may be removed in a future version')
2430 return self.cookiejar.get_cookie_header(url)
2431
2432 def _sort_thumbnails(self, thumbnails):
2433 thumbnails.sort(key=lambda t: (
2434 t.get('preference') if t.get('preference') is not None else -1,
2435 t.get('width') if t.get('width') is not None else -1,
2436 t.get('height') if t.get('height') is not None else -1,
2437 t.get('id') if t.get('id') is not None else '',
2438 t.get('url')))
2439
2440 def _sanitize_thumbnails(self, info_dict):
2441 thumbnails = info_dict.get('thumbnails')
2442 if thumbnails is None:
2443 thumbnail = info_dict.get('thumbnail')
2444 if thumbnail:
2445 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
2446 if not thumbnails:
2447 return
2448
2449 def check_thumbnails(thumbnails):
2450 for t in thumbnails:
2451 self.to_screen(f'[info] Testing thumbnail {t["id"]}')
2452 try:
2453 self.urlopen(HEADRequest(t['url']))
2454 except network_exceptions as err:
2455 self.to_screen(f'[info] Unable to connect to thumbnail {t["id"]} URL {t["url"]!r} - {err}. Skipping...')
2456 continue
2457 yield t
2458
2459 self._sort_thumbnails(thumbnails)
2460 for i, t in enumerate(thumbnails):
2461 if t.get('id') is None:
2462 t['id'] = '%d' % i
2463 if t.get('width') and t.get('height'):
2464 t['resolution'] = '%dx%d' % (t['width'], t['height'])
2465 t['url'] = sanitize_url(t['url'])
2466
2467 if self.params.get('check_formats') is True:
2468 info_dict['thumbnails'] = LazyList(check_thumbnails(thumbnails[::-1]), reverse=True)
2469 else:
2470 info_dict['thumbnails'] = thumbnails
2471
2472 def _fill_common_fields(self, info_dict, final=True):
2473 # TODO: move sanitization here
2474 if final:
2475 title = info_dict['fulltitle'] = info_dict.get('title')
2476 if not title:
2477 if title == '':
2478 self.write_debug('Extractor gave empty title. Creating a generic title')
2479 else:
2480 self.report_warning('Extractor failed to obtain "title". Creating a generic title instead')
2481 info_dict['title'] = f'{info_dict["extractor"].replace(":", "-")} video #{info_dict["id"]}'
2482
2483 if info_dict.get('duration') is not None:
2484 info_dict['duration_string'] = formatSeconds(info_dict['duration'])
2485
2486 for ts_key, date_key in (
2487 ('timestamp', 'upload_date'),
2488 ('release_timestamp', 'release_date'),
2489 ('modified_timestamp', 'modified_date'),
2490 ):
2491 if info_dict.get(date_key) is None and info_dict.get(ts_key) is not None:
2492 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
2493 # see http://bugs.python.org/issue1646728)
2494 with contextlib.suppress(ValueError, OverflowError, OSError):
2495 upload_date = datetime.datetime.utcfromtimestamp(info_dict[ts_key])
2496 info_dict[date_key] = upload_date.strftime('%Y%m%d')
2497
2498 live_keys = ('is_live', 'was_live')
2499 live_status = info_dict.get('live_status')
2500 if live_status is None:
2501 for key in live_keys:
2502 if info_dict.get(key) is False:
2503 continue
2504 if info_dict.get(key):
2505 live_status = key
2506 break
2507 if all(info_dict.get(key) is False for key in live_keys):
2508 live_status = 'not_live'
2509 if live_status:
2510 info_dict['live_status'] = live_status
2511 for key in live_keys:
2512 if info_dict.get(key) is None:
2513 info_dict[key] = (live_status == key)
2514 if live_status == 'post_live':
2515 info_dict['was_live'] = True
2516
2517 # Auto generate title fields corresponding to the *_number fields when missing
2518 # in order to always have clean titles. This is very common for TV series.
2519 for field in ('chapter', 'season', 'episode'):
2520 if final and info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
2521 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
2522
2523 def _raise_pending_errors(self, info):
2524 err = info.pop('__pending_error', None)
2525 if err:
2526 self.report_error(err, tb=False)
2527
2528 def sort_formats(self, info_dict):
2529 formats = self._get_formats(info_dict)
2530 formats.sort(key=FormatSorter(
2531 self, info_dict.get('_format_sort_fields') or []).calculate_preference)
2532
2533 def process_video_result(self, info_dict, download=True):
2534 assert info_dict.get('_type', 'video') == 'video'
2535 self._num_videos += 1
2536
2537 if 'id' not in info_dict:
2538 raise ExtractorError('Missing "id" field in extractor result', ie=info_dict['extractor'])
2539 elif not info_dict.get('id'):
2540 raise ExtractorError('Extractor failed to obtain "id"', ie=info_dict['extractor'])
2541
2542 def report_force_conversion(field, field_not, conversion):
2543 self.report_warning(
2544 '"%s" field is not %s - forcing %s conversion, there is an error in extractor'
2545 % (field, field_not, conversion))
2546
2547 def sanitize_string_field(info, string_field):
2548 field = info.get(string_field)
2549 if field is None or isinstance(field, str):
2550 return
2551 report_force_conversion(string_field, 'a string', 'string')
2552 info[string_field] = str(field)
2553
2554 def sanitize_numeric_fields(info):
2555 for numeric_field in self._NUMERIC_FIELDS:
2556 field = info.get(numeric_field)
2557 if field is None or isinstance(field, (int, float)):
2558 continue
2559 report_force_conversion(numeric_field, 'numeric', 'int')
2560 info[numeric_field] = int_or_none(field)
2561
2562 sanitize_string_field(info_dict, 'id')
2563 sanitize_numeric_fields(info_dict)
2564 if info_dict.get('section_end') and info_dict.get('section_start') is not None:
2565 info_dict['duration'] = round(info_dict['section_end'] - info_dict['section_start'], 3)
2566 if (info_dict.get('duration') or 0) <= 0 and info_dict.pop('duration', None):
2567 self.report_warning('"duration" field is negative, there is an error in extractor')
2568
2569 chapters = info_dict.get('chapters') or []
2570 if chapters and chapters[0].get('start_time'):
2571 chapters.insert(0, {'start_time': 0})
2572
2573 dummy_chapter = {'end_time': 0, 'start_time': info_dict.get('duration')}
2574 for idx, (prev, current, next_) in enumerate(zip(
2575 (dummy_chapter, *chapters), chapters, (*chapters[1:], dummy_chapter)), 1):
2576 if current.get('start_time') is None:
2577 current['start_time'] = prev.get('end_time')
2578 if not current.get('end_time'):
2579 current['end_time'] = next_.get('start_time')
2580 if not current.get('title'):
2581 current['title'] = f'<Untitled Chapter {idx}>'
2582
2583 if 'playlist' not in info_dict:
2584 # It isn't part of a playlist
2585 info_dict['playlist'] = None
2586 info_dict['playlist_index'] = None
2587
2588 self._sanitize_thumbnails(info_dict)
2589
2590 thumbnail = info_dict.get('thumbnail')
2591 thumbnails = info_dict.get('thumbnails')
2592 if thumbnail:
2593 info_dict['thumbnail'] = sanitize_url(thumbnail)
2594 elif thumbnails:
2595 info_dict['thumbnail'] = thumbnails[-1]['url']
2596
2597 if info_dict.get('display_id') is None and 'id' in info_dict:
2598 info_dict['display_id'] = info_dict['id']
2599
2600 self._fill_common_fields(info_dict)
2601
2602 for cc_kind in ('subtitles', 'automatic_captions'):
2603 cc = info_dict.get(cc_kind)
2604 if cc:
2605 for _, subtitle in cc.items():
2606 for subtitle_format in subtitle:
2607 if subtitle_format.get('url'):
2608 subtitle_format['url'] = sanitize_url(subtitle_format['url'])
2609 if subtitle_format.get('ext') is None:
2610 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
2611
2612 automatic_captions = info_dict.get('automatic_captions')
2613 subtitles = info_dict.get('subtitles')
2614
2615 info_dict['requested_subtitles'] = self.process_subtitles(
2616 info_dict['id'], subtitles, automatic_captions)
2617
2618 formats = self._get_formats(info_dict)
2619
2620 # Backward compatibility with InfoExtractor._sort_formats
2621 field_preference = (formats or [{}])[0].pop('__sort_fields', None)
2622 if field_preference:
2623 info_dict['_format_sort_fields'] = field_preference
2624
2625 info_dict['_has_drm'] = any( # or None ensures --clean-infojson removes it
2626 f.get('has_drm') and f['has_drm'] != 'maybe' for f in formats) or None
2627 if not self.params.get('allow_unplayable_formats'):
2628 formats = [f for f in formats if not f.get('has_drm') or f['has_drm'] == 'maybe']
2629
2630 if formats and all(f.get('acodec') == f.get('vcodec') == 'none' for f in formats):
2631 self.report_warning(
2632 f'{"This video is DRM protected and " if info_dict["_has_drm"] else ""}'
2633 'only images are available for download. Use --list-formats to see them'.capitalize())
2634
2635 get_from_start = not info_dict.get('is_live') or bool(self.params.get('live_from_start'))
2636 if not get_from_start:
2637 info_dict['title'] += ' ' + datetime.datetime.now().strftime('%Y-%m-%d %H:%M')
2638 if info_dict.get('is_live') and formats:
2639 formats = [f for f in formats if bool(f.get('is_from_start')) == get_from_start]
2640 if get_from_start and not formats:
2641 self.raise_no_formats(info_dict, msg=(
2642 '--live-from-start is passed, but there are no formats that can be downloaded from the start. '
2643 'If you want to download from the current time, use --no-live-from-start'))
2644
2645 def is_wellformed(f):
2646 url = f.get('url')
2647 if not url:
2648 self.report_warning(
2649 '"url" field is missing or empty - skipping format, '
2650 'there is an error in extractor')
2651 return False
2652 if isinstance(url, bytes):
2653 sanitize_string_field(f, 'url')
2654 return True
2655
2656 # Filter out malformed formats for better extraction robustness
2657 formats = list(filter(is_wellformed, formats or []))
2658
2659 if not formats:
2660 self.raise_no_formats(info_dict)
2661
2662 for format in formats:
2663 sanitize_string_field(format, 'format_id')
2664 sanitize_numeric_fields(format)
2665 format['url'] = sanitize_url(format['url'])
2666 if format.get('ext') is None:
2667 format['ext'] = determine_ext(format['url']).lower()
2668 if format.get('protocol') is None:
2669 format['protocol'] = determine_protocol(format)
2670 if format.get('resolution') is None:
2671 format['resolution'] = self.format_resolution(format, default=None)
2672 if format.get('dynamic_range') is None and format.get('vcodec') != 'none':
2673 format['dynamic_range'] = 'SDR'
2674 if format.get('aspect_ratio') is None:
2675 format['aspect_ratio'] = try_call(lambda: round(format['width'] / format['height'], 2))
2676 if (not format.get('manifest_url') # For fragmented formats, "tbr" is often max bitrate and not average
2677 and info_dict.get('duration') and format.get('tbr')
2678 and not format.get('filesize') and not format.get('filesize_approx')):
2679 format['filesize_approx'] = int(info_dict['duration'] * format['tbr'] * (1024 / 8))
2680 format['http_headers'] = self._calc_headers(collections.ChainMap(format, info_dict))
2681
2682 # This is copied to http_headers by the above _calc_headers and can now be removed
2683 if '__x_forwarded_for_ip' in info_dict:
2684 del info_dict['__x_forwarded_for_ip']
2685
2686 self.sort_formats({
2687 'formats': formats,
2688 '_format_sort_fields': info_dict.get('_format_sort_fields')
2689 })
2690
2691 # Sanitize and group by format_id
2692 formats_dict = {}
2693 for i, format in enumerate(formats):
2694 if not format.get('format_id'):
2695 format['format_id'] = str(i)
2696 else:
2697 # Sanitize format_id from characters used in format selector expression
2698 format['format_id'] = re.sub(r'[\s,/+\[\]()]', '_', format['format_id'])
2699 formats_dict.setdefault(format['format_id'], []).append(format)
2700
2701 # Make sure all formats have unique format_id
2702 common_exts = set(itertools.chain(*self._format_selection_exts.values()))
2703 for format_id, ambiguous_formats in formats_dict.items():
2704 ambigious_id = len(ambiguous_formats) > 1
2705 for i, format in enumerate(ambiguous_formats):
2706 if ambigious_id:
2707 format['format_id'] = '%s-%d' % (format_id, i)
2708 # Ensure there is no conflict between id and ext in format selection
2709 # See https://github.com/yt-dlp/yt-dlp/issues/1282
2710 if format['format_id'] != format['ext'] and format['format_id'] in common_exts:
2711 format['format_id'] = 'f%s' % format['format_id']
2712
2713 if format.get('format') is None:
2714 format['format'] = '{id} - {res}{note}'.format(
2715 id=format['format_id'],
2716 res=self.format_resolution(format),
2717 note=format_field(format, 'format_note', ' (%s)'),
2718 )
2719
2720 if self.params.get('check_formats') is True:
2721 formats = LazyList(self._check_formats(formats[::-1]), reverse=True)
2722
2723 if not formats or formats[0] is not info_dict:
2724 # only set the 'formats' fields if the original info_dict list them
2725 # otherwise we end up with a circular reference, the first (and unique)
2726 # element in the 'formats' field in info_dict is info_dict itself,
2727 # which can't be exported to json
2728 info_dict['formats'] = formats
2729
2730 info_dict, _ = self.pre_process(info_dict)
2731
2732 if self._match_entry(info_dict, incomplete=self._format_fields) is not None:
2733 return info_dict
2734
2735 self.post_extract(info_dict)
2736 info_dict, _ = self.pre_process(info_dict, 'after_filter')
2737
2738 # The pre-processors may have modified the formats
2739 formats = self._get_formats(info_dict)
2740
2741 list_only = self.params.get('simulate') == 'list_only'
2742 interactive_format_selection = not list_only and self.format_selector == '-'
2743 if self.params.get('list_thumbnails'):
2744 self.list_thumbnails(info_dict)
2745 if self.params.get('listsubtitles'):
2746 if 'automatic_captions' in info_dict:
2747 self.list_subtitles(
2748 info_dict['id'], automatic_captions, 'automatic captions')
2749 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
2750 if self.params.get('listformats') or interactive_format_selection:
2751 self.list_formats(info_dict)
2752 if list_only:
2753 # Without this printing, -F --print-json will not work
2754 self.__forced_printings(info_dict)
2755 return info_dict
2756
2757 format_selector = self.format_selector
2758 while True:
2759 if interactive_format_selection:
2760 req_format = input(self._format_screen('\nEnter format selector ', self.Styles.EMPHASIS)
2761 + '(Press ENTER for default, or Ctrl+C to quit)'
2762 + self._format_screen(': ', self.Styles.EMPHASIS))
2763 try:
2764 format_selector = self.build_format_selector(req_format) if req_format else None
2765 except SyntaxError as err:
2766 self.report_error(err, tb=False, is_error=False)
2767 continue
2768
2769 if format_selector is None:
2770 req_format = self._default_format_spec(info_dict, download=download)
2771 self.write_debug(f'Default format spec: {req_format}')
2772 format_selector = self.build_format_selector(req_format)
2773
2774 formats_to_download = list(format_selector({
2775 'formats': formats,
2776 'has_merged_format': any('none' not in (f.get('acodec'), f.get('vcodec')) for f in formats),
2777 'incomplete_formats': (all(f.get('vcodec') == 'none' for f in formats) # No formats with video
2778 or all(f.get('acodec') == 'none' for f in formats)), # OR, No formats with audio
2779 }))
2780 if interactive_format_selection and not formats_to_download:
2781 self.report_error('Requested format is not available', tb=False, is_error=False)
2782 continue
2783 break
2784
2785 if not formats_to_download:
2786 if not self.params.get('ignore_no_formats_error'):
2787 raise ExtractorError(
2788 'Requested format is not available. Use --list-formats for a list of available formats',
2789 expected=True, video_id=info_dict['id'], ie=info_dict['extractor'])
2790 self.report_warning('Requested format is not available')
2791 # Process what we can, even without any available formats.
2792 formats_to_download = [{}]
2793
2794 requested_ranges = tuple(self.params.get('download_ranges', lambda *_: [{}])(info_dict, self))
2795 best_format, downloaded_formats = formats_to_download[-1], []
2796 if download:
2797 if best_format and requested_ranges:
2798 def to_screen(*msg):
2799 self.to_screen(f'[info] {info_dict["id"]}: {" ".join(", ".join(variadic(m)) for m in msg)}')
2800
2801 to_screen(f'Downloading {len(formats_to_download)} format(s):',
2802 (f['format_id'] for f in formats_to_download))
2803 if requested_ranges != ({}, ):
2804 to_screen(f'Downloading {len(requested_ranges)} time ranges:',
2805 (f'{c["start_time"]:.1f}-{c["end_time"]:.1f}' for c in requested_ranges))
2806 max_downloads_reached = False
2807
2808 for fmt, chapter in itertools.product(formats_to_download, requested_ranges):
2809 new_info = self._copy_infodict(info_dict)
2810 new_info.update(fmt)
2811 offset, duration = info_dict.get('section_start') or 0, info_dict.get('duration') or float('inf')
2812 end_time = offset + min(chapter.get('end_time', duration), duration)
2813 # duration may not be accurate. So allow deviations <1sec
2814 if end_time == float('inf') or end_time > offset + duration + 1:
2815 end_time = None
2816 if chapter or offset:
2817 new_info.update({
2818 'section_start': offset + chapter.get('start_time', 0),
2819 'section_end': end_time,
2820 'section_title': chapter.get('title'),
2821 'section_number': chapter.get('index'),
2822 })
2823 downloaded_formats.append(new_info)
2824 try:
2825 self.process_info(new_info)
2826 except MaxDownloadsReached:
2827 max_downloads_reached = True
2828 self._raise_pending_errors(new_info)
2829 # Remove copied info
2830 for key, val in tuple(new_info.items()):
2831 if info_dict.get(key) == val:
2832 new_info.pop(key)
2833 if max_downloads_reached:
2834 break
2835
2836 write_archive = {f.get('__write_download_archive', False) for f in downloaded_formats}
2837 assert write_archive.issubset({True, False, 'ignore'})
2838 if True in write_archive and False not in write_archive:
2839 self.record_download_archive(info_dict)
2840
2841 info_dict['requested_downloads'] = downloaded_formats
2842 info_dict = self.run_all_pps('after_video', info_dict)
2843 if max_downloads_reached:
2844 raise MaxDownloadsReached()
2845
2846 # We update the info dict with the selected best quality format (backwards compatibility)
2847 info_dict.update(best_format)
2848 return info_dict
2849
2850 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
2851 """Select the requested subtitles and their format"""
2852 available_subs, normal_sub_langs = {}, []
2853 if normal_subtitles and self.params.get('writesubtitles'):
2854 available_subs.update(normal_subtitles)
2855 normal_sub_langs = tuple(normal_subtitles.keys())
2856 if automatic_captions and self.params.get('writeautomaticsub'):
2857 for lang, cap_info in automatic_captions.items():
2858 if lang not in available_subs:
2859 available_subs[lang] = cap_info
2860
2861 if not available_subs or (
2862 not self.params.get('writesubtitles')
2863 and not self.params.get('writeautomaticsub')):
2864 return None
2865
2866 all_sub_langs = tuple(available_subs.keys())
2867 if self.params.get('allsubtitles', False):
2868 requested_langs = all_sub_langs
2869 elif self.params.get('subtitleslangs', False):
2870 try:
2871 requested_langs = orderedSet_from_options(
2872 self.params.get('subtitleslangs'), {'all': all_sub_langs}, use_regex=True)
2873 except re.error as e:
2874 raise ValueError(f'Wrong regex for subtitlelangs: {e.pattern}')
2875 else:
2876 requested_langs = LazyList(itertools.chain(
2877 ['en'] if 'en' in normal_sub_langs else [],
2878 filter(lambda f: f.startswith('en'), normal_sub_langs),
2879 ['en'] if 'en' in all_sub_langs else [],
2880 filter(lambda f: f.startswith('en'), all_sub_langs),
2881 normal_sub_langs, all_sub_langs,
2882 ))[:1]
2883 if requested_langs:
2884 self.to_screen(f'[info] {video_id}: Downloading subtitles: {", ".join(requested_langs)}')
2885
2886 formats_query = self.params.get('subtitlesformat', 'best')
2887 formats_preference = formats_query.split('/') if formats_query else []
2888 subs = {}
2889 for lang in requested_langs:
2890 formats = available_subs.get(lang)
2891 if formats is None:
2892 self.report_warning(f'{lang} subtitles not available for {video_id}')
2893 continue
2894 for ext in formats_preference:
2895 if ext == 'best':
2896 f = formats[-1]
2897 break
2898 matches = list(filter(lambda f: f['ext'] == ext, formats))
2899 if matches:
2900 f = matches[-1]
2901 break
2902 else:
2903 f = formats[-1]
2904 self.report_warning(
2905 'No subtitle format found matching "%s" for language %s, '
2906 'using %s' % (formats_query, lang, f['ext']))
2907 subs[lang] = f
2908 return subs
2909
2910 def _forceprint(self, key, info_dict):
2911 if info_dict is None:
2912 return
2913 info_copy = info_dict.copy()
2914 info_copy.setdefault('filename', self.prepare_filename(info_dict))
2915 if info_dict.get('requested_formats') is not None:
2916 # For RTMP URLs, also include the playpath
2917 info_copy['urls'] = '\n'.join(f['url'] + f.get('play_path', '') for f in info_dict['requested_formats'])
2918 elif info_dict.get('url'):
2919 info_copy['urls'] = info_dict['url'] + info_dict.get('play_path', '')
2920 info_copy['formats_table'] = self.render_formats_table(info_dict)
2921 info_copy['thumbnails_table'] = self.render_thumbnails_table(info_dict)
2922 info_copy['subtitles_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('subtitles'))
2923 info_copy['automatic_captions_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('automatic_captions'))
2924
2925 def format_tmpl(tmpl):
2926 mobj = re.fullmatch(r'([\w.:,]|-\d|(?P<dict>{([\w.:,]|-\d)+}))+=?', tmpl)
2927 if not mobj:
2928 return tmpl
2929
2930 fmt = '%({})s'
2931 if tmpl.startswith('{'):
2932 tmpl, fmt = f'.{tmpl}', '%({})j'
2933 if tmpl.endswith('='):
2934 tmpl, fmt = tmpl[:-1], '{0} = %({0})#j'
2935 return '\n'.join(map(fmt.format, [tmpl] if mobj.group('dict') else tmpl.split(',')))
2936
2937 for tmpl in self.params['forceprint'].get(key, []):
2938 self.to_stdout(self.evaluate_outtmpl(format_tmpl(tmpl), info_copy))
2939
2940 for tmpl, file_tmpl in self.params['print_to_file'].get(key, []):
2941 filename = self.prepare_filename(info_dict, outtmpl=file_tmpl)
2942 tmpl = format_tmpl(tmpl)
2943 self.to_screen(f'[info] Writing {tmpl!r} to: {filename}')
2944 if self._ensure_dir_exists(filename):
2945 with open(filename, 'a', encoding='utf-8', newline='') as f:
2946 f.write(self.evaluate_outtmpl(tmpl, info_copy) + os.linesep)
2947
2948 return info_copy
2949
2950 def __forced_printings(self, info_dict, filename=None, incomplete=True):
2951 if (self.params.get('forcejson')
2952 or self.params['forceprint'].get('video')
2953 or self.params['print_to_file'].get('video')):
2954 self.post_extract(info_dict)
2955 if filename:
2956 info_dict['filename'] = filename
2957 info_copy = self._forceprint('video', info_dict)
2958
2959 def print_field(field, actual_field=None, optional=False):
2960 if actual_field is None:
2961 actual_field = field
2962 if self.params.get(f'force{field}') and (
2963 info_copy.get(field) is not None or (not optional and not incomplete)):
2964 self.to_stdout(info_copy[actual_field])
2965
2966 print_field('title')
2967 print_field('id')
2968 print_field('url', 'urls')
2969 print_field('thumbnail', optional=True)
2970 print_field('description', optional=True)
2971 print_field('filename')
2972 if self.params.get('forceduration') and info_copy.get('duration') is not None:
2973 self.to_stdout(formatSeconds(info_copy['duration']))
2974 print_field('format')
2975
2976 if self.params.get('forcejson'):
2977 self.to_stdout(json.dumps(self.sanitize_info(info_dict)))
2978
2979 def dl(self, name, info, subtitle=False, test=False):
2980 if not info.get('url'):
2981 self.raise_no_formats(info, True)
2982
2983 if test:
2984 verbose = self.params.get('verbose')
2985 params = {
2986 'test': True,
2987 'quiet': self.params.get('quiet') or not verbose,
2988 'verbose': verbose,
2989 'noprogress': not verbose,
2990 'nopart': True,
2991 'skip_unavailable_fragments': False,
2992 'keep_fragments': False,
2993 'overwrites': True,
2994 '_no_ytdl_file': True,
2995 }
2996 else:
2997 params = self.params
2998 fd = get_suitable_downloader(info, params, to_stdout=(name == '-'))(self, params)
2999 if not test:
3000 for ph in self._progress_hooks:
3001 fd.add_progress_hook(ph)
3002 urls = '", "'.join(
3003 (f['url'].split(',')[0] + ',<data>' if f['url'].startswith('data:') else f['url'])
3004 for f in info.get('requested_formats', []) or [info])
3005 self.write_debug(f'Invoking {fd.FD_NAME} downloader on "{urls}"')
3006
3007 # Note: Ideally info should be a deep-copied so that hooks cannot modify it.
3008 # But it may contain objects that are not deep-copyable
3009 new_info = self._copy_infodict(info)
3010 if new_info.get('http_headers') is None:
3011 new_info['http_headers'] = self._calc_headers(new_info)
3012 return fd.download(name, new_info, subtitle)
3013
3014 def existing_file(self, filepaths, *, default_overwrite=True):
3015 existing_files = list(filter(os.path.exists, orderedSet(filepaths)))
3016 if existing_files and not self.params.get('overwrites', default_overwrite):
3017 return existing_files[0]
3018
3019 for file in existing_files:
3020 self.report_file_delete(file)
3021 os.remove(file)
3022 return None
3023
3024 def process_info(self, info_dict):
3025 """Process a single resolved IE result. (Modifies it in-place)"""
3026
3027 assert info_dict.get('_type', 'video') == 'video'
3028 original_infodict = info_dict
3029
3030 if 'format' not in info_dict and 'ext' in info_dict:
3031 info_dict['format'] = info_dict['ext']
3032
3033 if self._match_entry(info_dict) is not None:
3034 info_dict['__write_download_archive'] = 'ignore'
3035 return
3036
3037 # Does nothing under normal operation - for backward compatibility of process_info
3038 self.post_extract(info_dict)
3039
3040 def replace_info_dict(new_info):
3041 nonlocal info_dict
3042 if new_info == info_dict:
3043 return
3044 info_dict.clear()
3045 info_dict.update(new_info)
3046
3047 new_info, _ = self.pre_process(info_dict, 'video')
3048 replace_info_dict(new_info)
3049 self._num_downloads += 1
3050
3051 # info_dict['_filename'] needs to be set for backward compatibility
3052 info_dict['_filename'] = full_filename = self.prepare_filename(info_dict, warn=True)
3053 temp_filename = self.prepare_filename(info_dict, 'temp')
3054 files_to_move = {}
3055
3056 # Forced printings
3057 self.__forced_printings(info_dict, full_filename, incomplete=('format' not in info_dict))
3058
3059 def check_max_downloads():
3060 if self._num_downloads >= float(self.params.get('max_downloads') or 'inf'):
3061 raise MaxDownloadsReached()
3062
3063 if self.params.get('simulate'):
3064 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
3065 check_max_downloads()
3066 return
3067
3068 if full_filename is None:
3069 return
3070 if not self._ensure_dir_exists(encodeFilename(full_filename)):
3071 return
3072 if not self._ensure_dir_exists(encodeFilename(temp_filename)):
3073 return
3074
3075 if self._write_description('video', info_dict,
3076 self.prepare_filename(info_dict, 'description')) is None:
3077 return
3078
3079 sub_files = self._write_subtitles(info_dict, temp_filename)
3080 if sub_files is None:
3081 return
3082 files_to_move.update(dict(sub_files))
3083
3084 thumb_files = self._write_thumbnails(
3085 'video', info_dict, temp_filename, self.prepare_filename(info_dict, 'thumbnail'))
3086 if thumb_files is None:
3087 return
3088 files_to_move.update(dict(thumb_files))
3089
3090 infofn = self.prepare_filename(info_dict, 'infojson')
3091 _infojson_written = self._write_info_json('video', info_dict, infofn)
3092 if _infojson_written:
3093 info_dict['infojson_filename'] = infofn
3094 # For backward compatibility, even though it was a private field
3095 info_dict['__infojson_filename'] = infofn
3096 elif _infojson_written is None:
3097 return
3098
3099 # Note: Annotations are deprecated
3100 annofn = None
3101 if self.params.get('writeannotations', False):
3102 annofn = self.prepare_filename(info_dict, 'annotation')
3103 if annofn:
3104 if not self._ensure_dir_exists(encodeFilename(annofn)):
3105 return
3106 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(annofn)):
3107 self.to_screen('[info] Video annotations are already present')
3108 elif not info_dict.get('annotations'):
3109 self.report_warning('There are no annotations to write.')
3110 else:
3111 try:
3112 self.to_screen('[info] Writing video annotations to: ' + annofn)
3113 with open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
3114 annofile.write(info_dict['annotations'])
3115 except (KeyError, TypeError):
3116 self.report_warning('There are no annotations to write.')
3117 except OSError:
3118 self.report_error('Cannot write annotations file: ' + annofn)
3119 return
3120
3121 # Write internet shortcut files
3122 def _write_link_file(link_type):
3123 url = try_get(info_dict['webpage_url'], iri_to_uri)
3124 if not url:
3125 self.report_warning(
3126 f'Cannot write internet shortcut file because the actual URL of "{info_dict["webpage_url"]}" is unknown')
3127 return True
3128 linkfn = replace_extension(self.prepare_filename(info_dict, 'link'), link_type, info_dict.get('ext'))
3129 if not self._ensure_dir_exists(encodeFilename(linkfn)):
3130 return False
3131 if self.params.get('overwrites', True) and os.path.exists(encodeFilename(linkfn)):
3132 self.to_screen(f'[info] Internet shortcut (.{link_type}) is already present')
3133 return True
3134 try:
3135 self.to_screen(f'[info] Writing internet shortcut (.{link_type}) to: {linkfn}')
3136 with open(encodeFilename(to_high_limit_path(linkfn)), 'w', encoding='utf-8',
3137 newline='\r\n' if link_type == 'url' else '\n') as linkfile:
3138 template_vars = {'url': url}
3139 if link_type == 'desktop':
3140 template_vars['filename'] = linkfn[:-(len(link_type) + 1)]
3141 linkfile.write(LINK_TEMPLATES[link_type] % template_vars)
3142 except OSError:
3143 self.report_error(f'Cannot write internet shortcut {linkfn}')
3144 return False
3145 return True
3146
3147 write_links = {
3148 'url': self.params.get('writeurllink'),
3149 'webloc': self.params.get('writewebloclink'),
3150 'desktop': self.params.get('writedesktoplink'),
3151 }
3152 if self.params.get('writelink'):
3153 link_type = ('webloc' if sys.platform == 'darwin'
3154 else 'desktop' if sys.platform.startswith('linux')
3155 else 'url')
3156 write_links[link_type] = True
3157
3158 if any(should_write and not _write_link_file(link_type)
3159 for link_type, should_write in write_links.items()):
3160 return
3161
3162 new_info, files_to_move = self.pre_process(info_dict, 'before_dl', files_to_move)
3163 replace_info_dict(new_info)
3164
3165 if self.params.get('skip_download'):
3166 info_dict['filepath'] = temp_filename
3167 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
3168 info_dict['__files_to_move'] = files_to_move
3169 replace_info_dict(self.run_pp(MoveFilesAfterDownloadPP(self, False), info_dict))
3170 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
3171 else:
3172 # Download
3173 info_dict.setdefault('__postprocessors', [])
3174 try:
3175
3176 def existing_video_file(*filepaths):
3177 ext = info_dict.get('ext')
3178 converted = lambda file: replace_extension(file, self.params.get('final_ext') or ext, ext)
3179 file = self.existing_file(itertools.chain(*zip(map(converted, filepaths), filepaths)),
3180 default_overwrite=False)
3181 if file:
3182 info_dict['ext'] = os.path.splitext(file)[1][1:]
3183 return file
3184
3185 fd, success = None, True
3186 if info_dict.get('protocol') or info_dict.get('url'):
3187 fd = get_suitable_downloader(info_dict, self.params, to_stdout=temp_filename == '-')
3188 if fd is not FFmpegFD and 'no-direct-merge' not in self.params['compat_opts'] and (
3189 info_dict.get('section_start') or info_dict.get('section_end')):
3190 msg = ('This format cannot be partially downloaded' if FFmpegFD.available()
3191 else 'You have requested downloading the video partially, but ffmpeg is not installed')
3192 self.report_error(f'{msg}. Aborting')
3193 return
3194
3195 if info_dict.get('requested_formats') is not None:
3196 old_ext = info_dict['ext']
3197 if self.params.get('merge_output_format') is None:
3198 if (info_dict['ext'] == 'webm'
3199 and info_dict.get('thumbnails')
3200 # check with type instead of pp_key, __name__, or isinstance
3201 # since we dont want any custom PPs to trigger this
3202 and any(type(pp) == EmbedThumbnailPP for pp in self._pps['post_process'])): # noqa: E721
3203 info_dict['ext'] = 'mkv'
3204 self.report_warning(
3205 'webm doesn\'t support embedding a thumbnail, mkv will be used')
3206 new_ext = info_dict['ext']
3207
3208 def correct_ext(filename, ext=new_ext):
3209 if filename == '-':
3210 return filename
3211 filename_real_ext = os.path.splitext(filename)[1][1:]
3212 filename_wo_ext = (
3213 os.path.splitext(filename)[0]
3214 if filename_real_ext in (old_ext, new_ext)
3215 else filename)
3216 return f'{filename_wo_ext}.{ext}'
3217
3218 # Ensure filename always has a correct extension for successful merge
3219 full_filename = correct_ext(full_filename)
3220 temp_filename = correct_ext(temp_filename)
3221 dl_filename = existing_video_file(full_filename, temp_filename)
3222
3223 info_dict['__real_download'] = False
3224 # NOTE: Copy so that original format dicts are not modified
3225 info_dict['requested_formats'] = list(map(dict, info_dict['requested_formats']))
3226
3227 merger = FFmpegMergerPP(self)
3228 downloaded = []
3229 if dl_filename is not None:
3230 self.report_file_already_downloaded(dl_filename)
3231 elif fd:
3232 for f in info_dict['requested_formats'] if fd != FFmpegFD else []:
3233 f['filepath'] = fname = prepend_extension(
3234 correct_ext(temp_filename, info_dict['ext']),
3235 'f%s' % f['format_id'], info_dict['ext'])
3236 downloaded.append(fname)
3237 info_dict['url'] = '\n'.join(f['url'] for f in info_dict['requested_formats'])
3238 success, real_download = self.dl(temp_filename, info_dict)
3239 info_dict['__real_download'] = real_download
3240 else:
3241 if self.params.get('allow_unplayable_formats'):
3242 self.report_warning(
3243 'You have requested merging of multiple formats '
3244 'while also allowing unplayable formats to be downloaded. '
3245 'The formats won\'t be merged to prevent data corruption.')
3246 elif not merger.available:
3247 msg = 'You have requested merging of multiple formats but ffmpeg is not installed'
3248 if not self.params.get('ignoreerrors'):
3249 self.report_error(f'{msg}. Aborting due to --abort-on-error')
3250 return
3251 self.report_warning(f'{msg}. The formats won\'t be merged')
3252
3253 if temp_filename == '-':
3254 reason = ('using a downloader other than ffmpeg' if FFmpegFD.can_merge_formats(info_dict, self.params)
3255 else 'but the formats are incompatible for simultaneous download' if merger.available
3256 else 'but ffmpeg is not installed')
3257 self.report_warning(
3258 f'You have requested downloading multiple formats to stdout {reason}. '
3259 'The formats will be streamed one after the other')
3260 fname = temp_filename
3261 for f in info_dict['requested_formats']:
3262 new_info = dict(info_dict)
3263 del new_info['requested_formats']
3264 new_info.update(f)
3265 if temp_filename != '-':
3266 fname = prepend_extension(
3267 correct_ext(temp_filename, new_info['ext']),
3268 'f%s' % f['format_id'], new_info['ext'])
3269 if not self._ensure_dir_exists(fname):
3270 return
3271 f['filepath'] = fname
3272 downloaded.append(fname)
3273 partial_success, real_download = self.dl(fname, new_info)
3274 info_dict['__real_download'] = info_dict['__real_download'] or real_download
3275 success = success and partial_success
3276
3277 if downloaded and merger.available and not self.params.get('allow_unplayable_formats'):
3278 info_dict['__postprocessors'].append(merger)
3279 info_dict['__files_to_merge'] = downloaded
3280 # Even if there were no downloads, it is being merged only now
3281 info_dict['__real_download'] = True
3282 else:
3283 for file in downloaded:
3284 files_to_move[file] = None
3285 else:
3286 # Just a single file
3287 dl_filename = existing_video_file(full_filename, temp_filename)
3288 if dl_filename is None or dl_filename == temp_filename:
3289 # dl_filename == temp_filename could mean that the file was partially downloaded with --no-part.
3290 # So we should try to resume the download
3291 success, real_download = self.dl(temp_filename, info_dict)
3292 info_dict['__real_download'] = real_download
3293 else:
3294 self.report_file_already_downloaded(dl_filename)
3295
3296 dl_filename = dl_filename or temp_filename
3297 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
3298
3299 except network_exceptions as err:
3300 self.report_error('unable to download video data: %s' % error_to_compat_str(err))
3301 return
3302 except OSError as err:
3303 raise UnavailableVideoError(err)
3304 except (ContentTooShortError, ) as err:
3305 self.report_error(f'content too short (expected {err.expected} bytes and served {err.downloaded})')
3306 return
3307
3308 self._raise_pending_errors(info_dict)
3309 if success and full_filename != '-':
3310
3311 def fixup():
3312 do_fixup = True
3313 fixup_policy = self.params.get('fixup')
3314 vid = info_dict['id']
3315
3316 if fixup_policy in ('ignore', 'never'):
3317 return
3318 elif fixup_policy == 'warn':
3319 do_fixup = 'warn'
3320 elif fixup_policy != 'force':
3321 assert fixup_policy in ('detect_or_warn', None)
3322 if not info_dict.get('__real_download'):
3323 do_fixup = False
3324
3325 def ffmpeg_fixup(cndn, msg, cls):
3326 if not (do_fixup and cndn):
3327 return
3328 elif do_fixup == 'warn':
3329 self.report_warning(f'{vid}: {msg}')
3330 return
3331 pp = cls(self)
3332 if pp.available:
3333 info_dict['__postprocessors'].append(pp)
3334 else:
3335 self.report_warning(f'{vid}: {msg}. Install ffmpeg to fix this automatically')
3336
3337 stretched_ratio = info_dict.get('stretched_ratio')
3338 ffmpeg_fixup(stretched_ratio not in (1, None),
3339 f'Non-uniform pixel ratio {stretched_ratio}',
3340 FFmpegFixupStretchedPP)
3341
3342 downloader = get_suitable_downloader(info_dict, self.params) if 'protocol' in info_dict else None
3343 downloader = downloader.FD_NAME if downloader else None
3344
3345 ext = info_dict.get('ext')
3346 postprocessed_by_ffmpeg = info_dict.get('requested_formats') or any((
3347 isinstance(pp, FFmpegVideoConvertorPP)
3348 and resolve_recode_mapping(ext, pp.mapping)[0] not in (ext, None)
3349 ) for pp in self._pps['post_process'])
3350
3351 if not postprocessed_by_ffmpeg:
3352 ffmpeg_fixup(ext == 'm4a' and info_dict.get('container') == 'm4a_dash',
3353 'writing DASH m4a. Only some players support this container',
3354 FFmpegFixupM4aPP)
3355 ffmpeg_fixup(downloader == 'hlsnative' and not self.params.get('hls_use_mpegts')
3356 or info_dict.get('is_live') and self.params.get('hls_use_mpegts') is None,
3357 'Possible MPEG-TS in MP4 container or malformed AAC timestamps',
3358 FFmpegFixupM3u8PP)
3359 ffmpeg_fixup(info_dict.get('is_live') and downloader == 'dashsegments',
3360 'Possible duplicate MOOV atoms', FFmpegFixupDuplicateMoovPP)
3361
3362 ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed timestamps detected', FFmpegFixupTimestampPP)
3363 ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed duration detected', FFmpegFixupDurationPP)
3364
3365 fixup()
3366 try:
3367 replace_info_dict(self.post_process(dl_filename, info_dict, files_to_move))
3368 except PostProcessingError as err:
3369 self.report_error('Postprocessing: %s' % str(err))
3370 return
3371 try:
3372 for ph in self._post_hooks:
3373 ph(info_dict['filepath'])
3374 except Exception as err:
3375 self.report_error('post hooks: %s' % str(err))
3376 return
3377 info_dict['__write_download_archive'] = True
3378
3379 assert info_dict is original_infodict # Make sure the info_dict was modified in-place
3380 if self.params.get('force_write_download_archive'):
3381 info_dict['__write_download_archive'] = True
3382 check_max_downloads()
3383
3384 def __download_wrapper(self, func):
3385 @functools.wraps(func)
3386 def wrapper(*args, **kwargs):
3387 try:
3388 res = func(*args, **kwargs)
3389 except UnavailableVideoError as e:
3390 self.report_error(e)
3391 except DownloadCancelled as e:
3392 self.to_screen(f'[info] {e}')
3393 if not self.params.get('break_per_url'):
3394 raise
3395 self._num_downloads = 0
3396 else:
3397 if self.params.get('dump_single_json', False):
3398 self.post_extract(res)
3399 self.to_stdout(json.dumps(self.sanitize_info(res)))
3400 return wrapper
3401
3402 def download(self, url_list):
3403 """Download a given list of URLs."""
3404 url_list = variadic(url_list) # Passing a single URL is a common mistake
3405 outtmpl = self.params['outtmpl']['default']
3406 if (len(url_list) > 1
3407 and outtmpl != '-'
3408 and '%' not in outtmpl
3409 and self.params.get('max_downloads') != 1):
3410 raise SameFileError(outtmpl)
3411
3412 for url in url_list:
3413 self.__download_wrapper(self.extract_info)(
3414 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
3415
3416 return self._download_retcode
3417
3418 def download_with_info_file(self, info_filename):
3419 with contextlib.closing(fileinput.FileInput(
3420 [info_filename], mode='r',
3421 openhook=fileinput.hook_encoded('utf-8'))) as f:
3422 # FileInput doesn't have a read method, we can't call json.load
3423 infos = [self.sanitize_info(info, self.params.get('clean_infojson', True))
3424 for info in variadic(json.loads('\n'.join(f)))]
3425 for info in infos:
3426 try:
3427 self.__download_wrapper(self.process_ie_result)(info, download=True)
3428 except (DownloadError, EntryNotInPlaylist, ReExtractInfo) as e:
3429 if not isinstance(e, EntryNotInPlaylist):
3430 self.to_stderr('\r')
3431 webpage_url = info.get('webpage_url')
3432 if webpage_url is None:
3433 raise
3434 self.report_warning(f'The info failed to download: {e}; trying with URL {webpage_url}')
3435 self.download([webpage_url])
3436 return self._download_retcode
3437
3438 @staticmethod
3439 def sanitize_info(info_dict, remove_private_keys=False):
3440 ''' Sanitize the infodict for converting to json '''
3441 if info_dict is None:
3442 return info_dict
3443 info_dict.setdefault('epoch', int(time.time()))
3444 info_dict.setdefault('_type', 'video')
3445 info_dict.setdefault('_version', {
3446 'version': __version__,
3447 'current_git_head': current_git_head(),
3448 'release_git_head': RELEASE_GIT_HEAD,
3449 'repository': REPOSITORY,
3450 })
3451
3452 if remove_private_keys:
3453 reject = lambda k, v: v is None or k.startswith('__') or k in {
3454 'requested_downloads', 'requested_formats', 'requested_subtitles', 'requested_entries',
3455 'entries', 'filepath', '_filename', 'filename', 'infojson_filename', 'original_url',
3456 'playlist_autonumber', '_format_sort_fields',
3457 }
3458 else:
3459 reject = lambda k, v: False
3460
3461 def filter_fn(obj):
3462 if isinstance(obj, dict):
3463 return {k: filter_fn(v) for k, v in obj.items() if not reject(k, v)}
3464 elif isinstance(obj, (list, tuple, set, LazyList)):
3465 return list(map(filter_fn, obj))
3466 elif obj is None or isinstance(obj, (str, int, float, bool)):
3467 return obj
3468 else:
3469 return repr(obj)
3470
3471 return filter_fn(info_dict)
3472
3473 @staticmethod
3474 def filter_requested_info(info_dict, actually_filter=True):
3475 ''' Alias of sanitize_info for backward compatibility '''
3476 return YoutubeDL.sanitize_info(info_dict, actually_filter)
3477
3478 def _delete_downloaded_files(self, *files_to_delete, info={}, msg=None):
3479 for filename in set(filter(None, files_to_delete)):
3480 if msg:
3481 self.to_screen(msg % filename)
3482 try:
3483 os.remove(filename)
3484 except OSError:
3485 self.report_warning(f'Unable to delete file {filename}')
3486 if filename in info.get('__files_to_move', []): # NB: Delete even if None
3487 del info['__files_to_move'][filename]
3488
3489 @staticmethod
3490 def post_extract(info_dict):
3491 def actual_post_extract(info_dict):
3492 if info_dict.get('_type') in ('playlist', 'multi_video'):
3493 for video_dict in info_dict.get('entries', {}):
3494 actual_post_extract(video_dict or {})
3495 return
3496
3497 post_extractor = info_dict.pop('__post_extractor', None) or (lambda: {})
3498 info_dict.update(post_extractor())
3499
3500 actual_post_extract(info_dict or {})
3501
3502 def run_pp(self, pp, infodict):
3503 files_to_delete = []
3504 if '__files_to_move' not in infodict:
3505 infodict['__files_to_move'] = {}
3506 try:
3507 files_to_delete, infodict = pp.run(infodict)
3508 except PostProcessingError as e:
3509 # Must be True and not 'only_download'
3510 if self.params.get('ignoreerrors') is True:
3511 self.report_error(e)
3512 return infodict
3513 raise
3514
3515 if not files_to_delete:
3516 return infodict
3517 if self.params.get('keepvideo', False):
3518 for f in files_to_delete:
3519 infodict['__files_to_move'].setdefault(f, '')
3520 else:
3521 self._delete_downloaded_files(
3522 *files_to_delete, info=infodict, msg='Deleting original file %s (pass -k to keep)')
3523 return infodict
3524
3525 def run_all_pps(self, key, info, *, additional_pps=None):
3526 if key != 'video':
3527 self._forceprint(key, info)
3528 for pp in (additional_pps or []) + self._pps[key]:
3529 info = self.run_pp(pp, info)
3530 return info
3531
3532 def pre_process(self, ie_info, key='pre_process', files_to_move=None):
3533 info = dict(ie_info)
3534 info['__files_to_move'] = files_to_move or {}
3535 try:
3536 info = self.run_all_pps(key, info)
3537 except PostProcessingError as err:
3538 msg = f'Preprocessing: {err}'
3539 info.setdefault('__pending_error', msg)
3540 self.report_error(msg, is_error=False)
3541 return info, info.pop('__files_to_move', None)
3542
3543 def post_process(self, filename, info, files_to_move=None):
3544 """Run all the postprocessors on the given file."""
3545 info['filepath'] = filename
3546 info['__files_to_move'] = files_to_move or {}
3547 info = self.run_all_pps('post_process', info, additional_pps=info.get('__postprocessors'))
3548 info = self.run_pp(MoveFilesAfterDownloadPP(self), info)
3549 del info['__files_to_move']
3550 return self.run_all_pps('after_move', info)
3551
3552 def _make_archive_id(self, info_dict):
3553 video_id = info_dict.get('id')
3554 if not video_id:
3555 return
3556 # Future-proof against any change in case
3557 # and backwards compatibility with prior versions
3558 extractor = info_dict.get('extractor_key') or info_dict.get('ie_key') # key in a playlist
3559 if extractor is None:
3560 url = str_or_none(info_dict.get('url'))
3561 if not url:
3562 return
3563 # Try to find matching extractor for the URL and take its ie_key
3564 for ie_key, ie in self._ies.items():
3565 if ie.suitable(url):
3566 extractor = ie_key
3567 break
3568 else:
3569 return
3570 return make_archive_id(extractor, video_id)
3571
3572 def in_download_archive(self, info_dict):
3573 if not self.archive:
3574 return False
3575
3576 vid_ids = [self._make_archive_id(info_dict)]
3577 vid_ids.extend(info_dict.get('_old_archive_ids') or [])
3578 return any(id_ in self.archive for id_ in vid_ids)
3579
3580 def record_download_archive(self, info_dict):
3581 fn = self.params.get('download_archive')
3582 if fn is None:
3583 return
3584 vid_id = self._make_archive_id(info_dict)
3585 assert vid_id
3586
3587 self.write_debug(f'Adding to archive: {vid_id}')
3588 if is_path_like(fn):
3589 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
3590 archive_file.write(vid_id + '\n')
3591 self.archive.add(vid_id)
3592
3593 @staticmethod
3594 def format_resolution(format, default='unknown'):
3595 if format.get('vcodec') == 'none' and format.get('acodec') != 'none':
3596 return 'audio only'
3597 if format.get('resolution') is not None:
3598 return format['resolution']
3599 if format.get('width') and format.get('height'):
3600 return '%dx%d' % (format['width'], format['height'])
3601 elif format.get('height'):
3602 return '%sp' % format['height']
3603 elif format.get('width'):
3604 return '%dx?' % format['width']
3605 return default
3606
3607 def _list_format_headers(self, *headers):
3608 if self.params.get('listformats_table', True) is not False:
3609 return [self._format_out(header, self.Styles.HEADERS) for header in headers]
3610 return headers
3611
3612 def _format_note(self, fdict):
3613 res = ''
3614 if fdict.get('ext') in ['f4f', 'f4m']:
3615 res += '(unsupported)'
3616 if fdict.get('language'):
3617 if res:
3618 res += ' '
3619 res += '[%s]' % fdict['language']
3620 if fdict.get('format_note') is not None:
3621 if res:
3622 res += ' '
3623 res += fdict['format_note']
3624 if fdict.get('tbr') is not None:
3625 if res:
3626 res += ', '
3627 res += '%4dk' % fdict['tbr']
3628 if fdict.get('container') is not None:
3629 if res:
3630 res += ', '
3631 res += '%s container' % fdict['container']
3632 if (fdict.get('vcodec') is not None
3633 and fdict.get('vcodec') != 'none'):
3634 if res:
3635 res += ', '
3636 res += fdict['vcodec']
3637 if fdict.get('vbr') is not None:
3638 res += '@'
3639 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
3640 res += 'video@'
3641 if fdict.get('vbr') is not None:
3642 res += '%4dk' % fdict['vbr']
3643 if fdict.get('fps') is not None:
3644 if res:
3645 res += ', '
3646 res += '%sfps' % fdict['fps']
3647 if fdict.get('acodec') is not None:
3648 if res:
3649 res += ', '
3650 if fdict['acodec'] == 'none':
3651 res += 'video only'
3652 else:
3653 res += '%-5s' % fdict['acodec']
3654 elif fdict.get('abr') is not None:
3655 if res:
3656 res += ', '
3657 res += 'audio'
3658 if fdict.get('abr') is not None:
3659 res += '@%3dk' % fdict['abr']
3660 if fdict.get('asr') is not None:
3661 res += ' (%5dHz)' % fdict['asr']
3662 if fdict.get('filesize') is not None:
3663 if res:
3664 res += ', '
3665 res += format_bytes(fdict['filesize'])
3666 elif fdict.get('filesize_approx') is not None:
3667 if res:
3668 res += ', '
3669 res += '~' + format_bytes(fdict['filesize_approx'])
3670 return res
3671
3672 def _get_formats(self, info_dict):
3673 if info_dict.get('formats') is None:
3674 if info_dict.get('url') and info_dict.get('_type', 'video') == 'video':
3675 return [info_dict]
3676 return []
3677 return info_dict['formats']
3678
3679 def render_formats_table(self, info_dict):
3680 formats = self._get_formats(info_dict)
3681 if not formats:
3682 return
3683 if not self.params.get('listformats_table', True) is not False:
3684 table = [
3685 [
3686 format_field(f, 'format_id'),
3687 format_field(f, 'ext'),
3688 self.format_resolution(f),
3689 self._format_note(f)
3690 ] for f in formats if (f.get('preference') or 0) >= -1000]
3691 return render_table(['format code', 'extension', 'resolution', 'note'], table, extra_gap=1)
3692
3693 def simplified_codec(f, field):
3694 assert field in ('acodec', 'vcodec')
3695 codec = f.get(field)
3696 if not codec:
3697 return 'unknown'
3698 elif codec != 'none':
3699 return '.'.join(codec.split('.')[:4])
3700
3701 if field == 'vcodec' and f.get('acodec') == 'none':
3702 return 'images'
3703 elif field == 'acodec' and f.get('vcodec') == 'none':
3704 return ''
3705 return self._format_out('audio only' if field == 'vcodec' else 'video only',
3706 self.Styles.SUPPRESS)
3707
3708 delim = self._format_out('\u2502', self.Styles.DELIM, '|', test_encoding=True)
3709 table = [
3710 [
3711 self._format_out(format_field(f, 'format_id'), self.Styles.ID),
3712 format_field(f, 'ext'),
3713 format_field(f, func=self.format_resolution, ignore=('audio only', 'images')),
3714 format_field(f, 'fps', '\t%d', func=round),
3715 format_field(f, 'dynamic_range', '%s', ignore=(None, 'SDR')).replace('HDR', ''),
3716 format_field(f, 'audio_channels', '\t%s'),
3717 delim, (
3718 format_field(f, 'filesize', ' \t%s', func=format_bytes)
3719 or format_field(f, 'filesize_approx', '≈\t%s', func=format_bytes)
3720 or format_field(try_call(lambda: format_bytes(int(info_dict['duration'] * f['tbr'] * (1024 / 8)))),
3721 None, self._format_out('~\t%s', self.Styles.SUPPRESS))),
3722 format_field(f, 'tbr', '\t%dk', func=round),
3723 shorten_protocol_name(f.get('protocol', '')),
3724 delim,
3725 simplified_codec(f, 'vcodec'),
3726 format_field(f, 'vbr', '\t%dk', func=round),
3727 simplified_codec(f, 'acodec'),
3728 format_field(f, 'abr', '\t%dk', func=round),
3729 format_field(f, 'asr', '\t%s', func=format_decimal_suffix),
3730 join_nonempty(format_field(f, 'language', '[%s]'), join_nonempty(
3731 self._format_out('UNSUPPORTED', self.Styles.BAD_FORMAT) if f.get('ext') in ('f4f', 'f4m') else None,
3732 (self._format_out('Maybe DRM', self.Styles.WARNING) if f.get('has_drm') == 'maybe'
3733 else self._format_out('DRM', self.Styles.BAD_FORMAT) if f.get('has_drm') else None),
3734 format_field(f, 'format_note'),
3735 format_field(f, 'container', ignore=(None, f.get('ext'))),
3736 delim=', '), delim=' '),
3737 ] for f in formats if f.get('preference') is None or f['preference'] >= -1000]
3738 header_line = self._list_format_headers(
3739 'ID', 'EXT', 'RESOLUTION', '\tFPS', 'HDR', 'CH', delim, '\tFILESIZE', '\tTBR', 'PROTO',
3740 delim, 'VCODEC', '\tVBR', 'ACODEC', '\tABR', '\tASR', 'MORE INFO')
3741
3742 return render_table(
3743 header_line, table, hide_empty=True,
3744 delim=self._format_out('\u2500', self.Styles.DELIM, '-', test_encoding=True))
3745
3746 def render_thumbnails_table(self, info_dict):
3747 thumbnails = list(info_dict.get('thumbnails') or [])
3748 if not thumbnails:
3749 return None
3750 return render_table(
3751 self._list_format_headers('ID', 'Width', 'Height', 'URL'),
3752 [[t.get('id'), t.get('width') or 'unknown', t.get('height') or 'unknown', t['url']] for t in thumbnails])
3753
3754 def render_subtitles_table(self, video_id, subtitles):
3755 def _row(lang, formats):
3756 exts, names = zip(*((f['ext'], f.get('name') or 'unknown') for f in reversed(formats)))
3757 if len(set(names)) == 1:
3758 names = [] if names[0] == 'unknown' else names[:1]
3759 return [lang, ', '.join(names), ', '.join(exts)]
3760
3761 if not subtitles:
3762 return None
3763 return render_table(
3764 self._list_format_headers('Language', 'Name', 'Formats'),
3765 [_row(lang, formats) for lang, formats in subtitles.items()],
3766 hide_empty=True)
3767
3768 def __list_table(self, video_id, name, func, *args):
3769 table = func(*args)
3770 if not table:
3771 self.to_screen(f'{video_id} has no {name}')
3772 return
3773 self.to_screen(f'[info] Available {name} for {video_id}:')
3774 self.to_stdout(table)
3775
3776 def list_formats(self, info_dict):
3777 self.__list_table(info_dict['id'], 'formats', self.render_formats_table, info_dict)
3778
3779 def list_thumbnails(self, info_dict):
3780 self.__list_table(info_dict['id'], 'thumbnails', self.render_thumbnails_table, info_dict)
3781
3782 def list_subtitles(self, video_id, subtitles, name='subtitles'):
3783 self.__list_table(video_id, name, self.render_subtitles_table, video_id, subtitles)
3784
3785 def urlopen(self, req):
3786 """ Start an HTTP download """
3787 if isinstance(req, str):
3788 req = sanitized_Request(req)
3789 return self._opener.open(req, timeout=self._socket_timeout)
3790
3791 def print_debug_header(self):
3792 if not self.params.get('verbose'):
3793 return
3794
3795 from . import _IN_CLI # Must be delayed import
3796
3797 # These imports can be slow. So import them only as needed
3798 from .extractor.extractors import _LAZY_LOADER
3799 from .extractor.extractors import (
3800 _PLUGIN_CLASSES as plugin_ies,
3801 _PLUGIN_OVERRIDES as plugin_ie_overrides
3802 )
3803
3804 def get_encoding(stream):
3805 ret = str(getattr(stream, 'encoding', 'missing (%s)' % type(stream).__name__))
3806 additional_info = []
3807 if os.environ.get('TERM', '').lower() == 'dumb':
3808 additional_info.append('dumb')
3809 if not supports_terminal_sequences(stream):
3810 from .utils import WINDOWS_VT_MODE # Must be imported locally
3811 additional_info.append('No VT' if WINDOWS_VT_MODE is False else 'No ANSI')
3812 if additional_info:
3813 ret = f'{ret} ({",".join(additional_info)})'
3814 return ret
3815
3816 encoding_str = 'Encodings: locale %s, fs %s, pref %s, %s' % (
3817 locale.getpreferredencoding(),
3818 sys.getfilesystemencoding(),
3819 self.get_encoding(),
3820 ', '.join(
3821 f'{key} {get_encoding(stream)}' for key, stream in self._out_files.items_
3822 if stream is not None and key != 'console')
3823 )
3824
3825 logger = self.params.get('logger')
3826 if logger:
3827 write_debug = lambda msg: logger.debug(f'[debug] {msg}')
3828 write_debug(encoding_str)
3829 else:
3830 write_string(f'[debug] {encoding_str}\n', encoding=None)
3831 write_debug = lambda msg: self._write_string(f'[debug] {msg}\n')
3832
3833 source = detect_variant()
3834 if VARIANT not in (None, 'pip'):
3835 source += '*'
3836 klass = type(self)
3837 write_debug(join_nonempty(
3838 f'{"yt-dlp" if REPOSITORY == "yt-dlp/yt-dlp" else REPOSITORY} version',
3839 f'{CHANNEL}@{__version__}',
3840 f'[{RELEASE_GIT_HEAD[:9]}]' if RELEASE_GIT_HEAD else '',
3841 '' if source == 'unknown' else f'({source})',
3842 '' if _IN_CLI else 'API' if klass == YoutubeDL else f'API:{self.__module__}.{klass.__qualname__}',
3843 delim=' '))
3844
3845 if not _IN_CLI:
3846 write_debug(f'params: {self.params}')
3847
3848 if not _LAZY_LOADER:
3849 if os.environ.get('YTDLP_NO_LAZY_EXTRACTORS'):
3850 write_debug('Lazy loading extractors is forcibly disabled')
3851 else:
3852 write_debug('Lazy loading extractors is disabled')
3853 if self.params['compat_opts']:
3854 write_debug('Compatibility options: %s' % ', '.join(self.params['compat_opts']))
3855
3856 if current_git_head():
3857 write_debug(f'Git HEAD: {current_git_head()}')
3858 write_debug(system_identifier())
3859
3860 exe_versions, ffmpeg_features = FFmpegPostProcessor.get_versions_and_features(self)
3861 ffmpeg_features = {key for key, val in ffmpeg_features.items() if val}
3862 if ffmpeg_features:
3863 exe_versions['ffmpeg'] += ' (%s)' % ','.join(sorted(ffmpeg_features))
3864
3865 exe_versions['rtmpdump'] = rtmpdump_version()
3866 exe_versions['phantomjs'] = PhantomJSwrapper._version()
3867 exe_str = ', '.join(
3868 f'{exe} {v}' for exe, v in sorted(exe_versions.items()) if v
3869 ) or 'none'
3870 write_debug('exe versions: %s' % exe_str)
3871
3872 from .compat.compat_utils import get_package_info
3873 from .dependencies import available_dependencies
3874
3875 write_debug('Optional libraries: %s' % (', '.join(sorted({
3876 join_nonempty(*get_package_info(m)) for m in available_dependencies.values()
3877 })) or 'none'))
3878
3879 self._setup_opener()
3880 proxy_map = {}
3881 for handler in self._opener.handlers:
3882 if hasattr(handler, 'proxies'):
3883 proxy_map.update(handler.proxies)
3884 write_debug(f'Proxy map: {proxy_map}')
3885
3886 for plugin_type, plugins in {'Extractor': plugin_ies, 'Post-Processor': plugin_pps}.items():
3887 display_list = ['%s%s' % (
3888 klass.__name__, '' if klass.__name__ == name else f' as {name}')
3889 for name, klass in plugins.items()]
3890 if plugin_type == 'Extractor':
3891 display_list.extend(f'{plugins[-1].IE_NAME.partition("+")[2]} ({parent.__name__})'
3892 for parent, plugins in plugin_ie_overrides.items())
3893 if not display_list:
3894 continue
3895 write_debug(f'{plugin_type} Plugins: {", ".join(sorted(display_list))}')
3896
3897 plugin_dirs = plugin_directories()
3898 if plugin_dirs:
3899 write_debug(f'Plugin directories: {plugin_dirs}')
3900
3901 # Not implemented
3902 if False and self.params.get('call_home'):
3903 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode()
3904 write_debug('Public IP address: %s' % ipaddr)
3905 latest_version = self.urlopen(
3906 'https://yt-dl.org/latest/version').read().decode()
3907 if version_tuple(latest_version) > version_tuple(__version__):
3908 self.report_warning(
3909 'You are using an outdated version (newest version: %s)! '
3910 'See https://yt-dl.org/update if you need help updating.' %
3911 latest_version)
3912
3913 def _setup_opener(self):
3914 if hasattr(self, '_opener'):
3915 return
3916 timeout_val = self.params.get('socket_timeout')
3917 self._socket_timeout = 20 if timeout_val is None else float(timeout_val)
3918
3919 opts_cookiesfrombrowser = self.params.get('cookiesfrombrowser')
3920 opts_cookiefile = self.params.get('cookiefile')
3921 opts_proxy = self.params.get('proxy')
3922
3923 self.cookiejar = load_cookies(opts_cookiefile, opts_cookiesfrombrowser, self)
3924
3925 cookie_processor = YoutubeDLCookieProcessor(self.cookiejar)
3926 if opts_proxy is not None:
3927 if opts_proxy == '':
3928 proxies = {}
3929 else:
3930 proxies = {'http': opts_proxy, 'https': opts_proxy}
3931 else:
3932 proxies = urllib.request.getproxies()
3933 # Set HTTPS proxy to HTTP one if given (https://github.com/ytdl-org/youtube-dl/issues/805)
3934 if 'http' in proxies and 'https' not in proxies:
3935 proxies['https'] = proxies['http']
3936 proxy_handler = PerRequestProxyHandler(proxies)
3937
3938 debuglevel = 1 if self.params.get('debug_printtraffic') else 0
3939 https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
3940 ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
3941 redirect_handler = YoutubeDLRedirectHandler()
3942 data_handler = urllib.request.DataHandler()
3943
3944 # When passing our own FileHandler instance, build_opener won't add the
3945 # default FileHandler and allows us to disable the file protocol, which
3946 # can be used for malicious purposes (see
3947 # https://github.com/ytdl-org/youtube-dl/issues/8227)
3948 file_handler = urllib.request.FileHandler()
3949
3950 if not self.params.get('enable_file_urls'):
3951 def file_open(*args, **kwargs):
3952 raise urllib.error.URLError(
3953 'file:// URLs are explicitly disabled in yt-dlp for security reasons. '
3954 'Use --enable-file-urls to enable at your own risk.')
3955 file_handler.file_open = file_open
3956
3957 opener = urllib.request.build_opener(
3958 proxy_handler, https_handler, cookie_processor, ydlh, redirect_handler, data_handler, file_handler)
3959
3960 # Delete the default user-agent header, which would otherwise apply in
3961 # cases where our custom HTTP handler doesn't come into play
3962 # (See https://github.com/ytdl-org/youtube-dl/issues/1309 for details)
3963 opener.addheaders = []
3964 self._opener = opener
3965
3966 def encode(self, s):
3967 if isinstance(s, bytes):
3968 return s # Already encoded
3969
3970 try:
3971 return s.encode(self.get_encoding())
3972 except UnicodeEncodeError as err:
3973 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
3974 raise
3975
3976 def get_encoding(self):
3977 encoding = self.params.get('encoding')
3978 if encoding is None:
3979 encoding = preferredencoding()
3980 return encoding
3981
3982 def _write_info_json(self, label, ie_result, infofn, overwrite=None):
3983 ''' Write infojson and returns True = written, 'exists' = Already exists, False = skip, None = error '''
3984 if overwrite is None:
3985 overwrite = self.params.get('overwrites', True)
3986 if not self.params.get('writeinfojson'):
3987 return False
3988 elif not infofn:
3989 self.write_debug(f'Skipping writing {label} infojson')
3990 return False
3991 elif not self._ensure_dir_exists(infofn):
3992 return None
3993 elif not overwrite and os.path.exists(infofn):
3994 self.to_screen(f'[info] {label.title()} metadata is already present')
3995 return 'exists'
3996
3997 self.to_screen(f'[info] Writing {label} metadata as JSON to: {infofn}')
3998 try:
3999 write_json_file(self.sanitize_info(ie_result, self.params.get('clean_infojson', True)), infofn)
4000 return True
4001 except OSError:
4002 self.report_error(f'Cannot write {label} metadata to JSON file {infofn}')
4003 return None
4004
4005 def _write_description(self, label, ie_result, descfn):
4006 ''' Write description and returns True = written, False = skip, None = error '''
4007 if not self.params.get('writedescription'):
4008 return False
4009 elif not descfn:
4010 self.write_debug(f'Skipping writing {label} description')
4011 return False
4012 elif not self._ensure_dir_exists(descfn):
4013 return None
4014 elif not self.params.get('overwrites', True) and os.path.exists(descfn):
4015 self.to_screen(f'[info] {label.title()} description is already present')
4016 elif ie_result.get('description') is None:
4017 self.to_screen(f'[info] There\'s no {label} description to write')
4018 return False
4019 else:
4020 try:
4021 self.to_screen(f'[info] Writing {label} description to: {descfn}')
4022 with open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
4023 descfile.write(ie_result['description'])
4024 except OSError:
4025 self.report_error(f'Cannot write {label} description file {descfn}')
4026 return None
4027 return True
4028
4029 def _write_subtitles(self, info_dict, filename):
4030 ''' Write subtitles to file and return list of (sub_filename, final_sub_filename); or None if error'''
4031 ret = []
4032 subtitles = info_dict.get('requested_subtitles')
4033 if not (self.params.get('writesubtitles') or self.params.get('writeautomaticsub')):
4034 # subtitles download errors are already managed as troubles in relevant IE
4035 # that way it will silently go on when used with unsupporting IE
4036 return ret
4037 elif not subtitles:
4038 self.to_screen('[info] There are no subtitles for the requested languages')
4039 return ret
4040 sub_filename_base = self.prepare_filename(info_dict, 'subtitle')
4041 if not sub_filename_base:
4042 self.to_screen('[info] Skipping writing video subtitles')
4043 return ret
4044
4045 for sub_lang, sub_info in subtitles.items():
4046 sub_format = sub_info['ext']
4047 sub_filename = subtitles_filename(filename, sub_lang, sub_format, info_dict.get('ext'))
4048 sub_filename_final = subtitles_filename(sub_filename_base, sub_lang, sub_format, info_dict.get('ext'))
4049 existing_sub = self.existing_file((sub_filename_final, sub_filename))
4050 if existing_sub:
4051 self.to_screen(f'[info] Video subtitle {sub_lang}.{sub_format} is already present')
4052 sub_info['filepath'] = existing_sub
4053 ret.append((existing_sub, sub_filename_final))
4054 continue
4055
4056 self.to_screen(f'[info] Writing video subtitles to: {sub_filename}')
4057 if sub_info.get('data') is not None:
4058 try:
4059 # Use newline='' to prevent conversion of newline characters
4060 # See https://github.com/ytdl-org/youtube-dl/issues/10268
4061 with open(sub_filename, 'w', encoding='utf-8', newline='') as subfile:
4062 subfile.write(sub_info['data'])
4063 sub_info['filepath'] = sub_filename
4064 ret.append((sub_filename, sub_filename_final))
4065 continue
4066 except OSError:
4067 self.report_error(f'Cannot write video subtitles file {sub_filename}')
4068 return None
4069
4070 try:
4071 sub_copy = sub_info.copy()
4072 sub_copy.setdefault('http_headers', info_dict.get('http_headers'))
4073 self.dl(sub_filename, sub_copy, subtitle=True)
4074 sub_info['filepath'] = sub_filename
4075 ret.append((sub_filename, sub_filename_final))
4076 except (DownloadError, ExtractorError, IOError, OSError, ValueError) + network_exceptions as err:
4077 msg = f'Unable to download video subtitles for {sub_lang!r}: {err}'
4078 if self.params.get('ignoreerrors') is not True: # False or 'only_download'
4079 if not self.params.get('ignoreerrors'):
4080 self.report_error(msg)
4081 raise DownloadError(msg)
4082 self.report_warning(msg)
4083 return ret
4084
4085 def _write_thumbnails(self, label, info_dict, filename, thumb_filename_base=None):
4086 ''' Write thumbnails to file and return list of (thumb_filename, final_thumb_filename) '''
4087 write_all = self.params.get('write_all_thumbnails', False)
4088 thumbnails, ret = [], []
4089 if write_all or self.params.get('writethumbnail', False):
4090 thumbnails = info_dict.get('thumbnails') or []
4091 if not thumbnails:
4092 self.to_screen(f'[info] There are no {label} thumbnails to download')
4093 return ret
4094 multiple = write_all and len(thumbnails) > 1
4095
4096 if thumb_filename_base is None:
4097 thumb_filename_base = filename
4098 if thumbnails and not thumb_filename_base:
4099 self.write_debug(f'Skipping writing {label} thumbnail')
4100 return ret
4101
4102 for idx, t in list(enumerate(thumbnails))[::-1]:
4103 thumb_ext = (f'{t["id"]}.' if multiple else '') + determine_ext(t['url'], 'jpg')
4104 thumb_display_id = f'{label} thumbnail {t["id"]}'
4105 thumb_filename = replace_extension(filename, thumb_ext, info_dict.get('ext'))
4106 thumb_filename_final = replace_extension(thumb_filename_base, thumb_ext, info_dict.get('ext'))
4107
4108 existing_thumb = self.existing_file((thumb_filename_final, thumb_filename))
4109 if existing_thumb:
4110 self.to_screen('[info] %s is already present' % (
4111 thumb_display_id if multiple else f'{label} thumbnail').capitalize())
4112 t['filepath'] = existing_thumb
4113 ret.append((existing_thumb, thumb_filename_final))
4114 else:
4115 self.to_screen(f'[info] Downloading {thumb_display_id} ...')
4116 try:
4117 uf = self.urlopen(sanitized_Request(t['url'], headers=t.get('http_headers', {})))
4118 self.to_screen(f'[info] Writing {thumb_display_id} to: {thumb_filename}')
4119 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
4120 shutil.copyfileobj(uf, thumbf)
4121 ret.append((thumb_filename, thumb_filename_final))
4122 t['filepath'] = thumb_filename
4123 except network_exceptions as err:
4124 if isinstance(err, urllib.error.HTTPError) and err.code == 404:
4125 self.to_screen(f'[info] {thumb_display_id.title()} does not exist')
4126 else:
4127 self.report_warning(f'Unable to download {thumb_display_id}: {err}')
4128 thumbnails.pop(idx)
4129 if ret and not write_all:
4130 break
4131 return ret