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