]> jfr.im git - yt-dlp.git/blob - yt_dlp/YoutubeDL.py
0d75640887e2a66ac4bc455c7a246b9898e15380
[yt-dlp.git] / yt_dlp / YoutubeDL.py
1 import collections
2 import contextlib
3 import datetime
4 import errno
5 import fileinput
6 import functools
7 import io
8 import itertools
9 import json
10 import locale
11 import operator
12 import os
13 import random
14 import re
15 import shutil
16 import subprocess
17 import sys
18 import tempfile
19 import time
20 import tokenize
21 import traceback
22 import unicodedata
23 import urllib.request
24 from string import ascii_letters
25
26 from .cache import Cache
27 from .compat import compat_os_name, compat_shlex_quote
28 from .cookies import load_cookies
29 from .downloader import FFmpegFD, get_suitable_downloader, shorten_protocol_name
30 from .downloader.rtmp import rtmpdump_version
31 from .extractor import gen_extractor_classes, get_info_extractor
32 from .extractor.openload import PhantomJSwrapper
33 from .minicurses import format_text
34 from .postprocessor import _PLUGIN_CLASSES as plugin_postprocessors
35 from .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 )
49 from .postprocessor.ffmpeg import resolve_mapping as resolve_recode_mapping
50 from .update import detect_variant
51 from .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_domain,
106 int_or_none,
107 iri_to_uri,
108 join_nonempty,
109 locked_file,
110 make_archive_id,
111 make_dir,
112 make_HTTPS_handler,
113 merge_headers,
114 network_exceptions,
115 number_of_digits,
116 orderedSet,
117 parse_filesize,
118 preferredencoding,
119 prepend_extension,
120 register_socks_protocols,
121 remove_terminal_sequences,
122 render_table,
123 replace_extension,
124 sanitize_filename,
125 sanitize_path,
126 sanitize_url,
127 sanitized_Request,
128 std_headers,
129 str_or_none,
130 strftime_or_none,
131 subtitles_filename,
132 supports_terminal_sequences,
133 system_identifier,
134 timetuple_from_msec,
135 to_high_limit_path,
136 traverse_obj,
137 try_get,
138 url_basename,
139 variadic,
140 version_tuple,
141 windows_enable_vt_mode,
142 write_json_file,
143 write_string,
144 )
145 from .version import RELEASE_GIT_HEAD, __version__
146
147 if compat_os_name == 'nt':
148 import ctypes
149
150
151 class YoutubeDL:
152 """YoutubeDL class.
153
154 YoutubeDL objects are the ones responsible of downloading the
155 actual video file and writing it to disk if the user has requested
156 it, among some other tasks. In most cases there should be one per
157 program. As, given a video URL, the downloader doesn't know how to
158 extract all the needed information, task that InfoExtractors do, it
159 has to pass the URL to one of them.
160
161 For this, YoutubeDL objects have a method that allows
162 InfoExtractors to be registered in a given order. When it is passed
163 a URL, the YoutubeDL object handles it to the first InfoExtractor it
164 finds that reports being able to handle it. The InfoExtractor extracts
165 all the information about the video or videos the URL refers to, and
166 YoutubeDL process the extracted information, possibly using a File
167 Downloader to download the video.
168
169 YoutubeDL objects accept a lot of parameters. In order not to saturate
170 the object constructor with arguments, it receives a dictionary of
171 options instead. These options are available through the params
172 attribute for the InfoExtractors to use. The YoutubeDL also
173 registers itself as the downloader in charge for the InfoExtractors
174 that are added to it, so this is a "mutual registration".
175
176 Available options:
177
178 username: Username for authentication purposes.
179 password: Password for authentication purposes.
180 videopassword: Password for accessing a video.
181 ap_mso: Adobe Pass multiple-system operator identifier.
182 ap_username: Multiple-system operator account username.
183 ap_password: Multiple-system operator account password.
184 usenetrc: Use netrc for authentication instead.
185 verbose: Print additional info to stdout.
186 quiet: Do not print messages to stdout.
187 no_warnings: Do not print out anything for warnings.
188 forceprint: A dict with keys WHEN mapped to a list of templates to
189 print to stdout. The allowed keys are video or any of the
190 items in utils.POSTPROCESS_WHEN.
191 For compatibility, a single list is also accepted
192 print_to_file: A dict with keys WHEN (same as forceprint) mapped to
193 a list of tuples with (template, filename)
194 forcejson: Force printing info_dict as JSON.
195 dump_single_json: Force printing the info_dict of the whole playlist
196 (or video) as a single JSON line.
197 force_write_download_archive: Force writing download archive regardless
198 of 'skip_download' or 'simulate'.
199 simulate: Do not download the video files. If unset (or None),
200 simulate only if listsubtitles, listformats or list_thumbnails is used
201 format: Video format code. see "FORMAT SELECTION" for more details.
202 You can also pass a function. The function takes 'ctx' as
203 argument and returns the formats to download.
204 See "build_format_selector" for an implementation
205 allow_unplayable_formats: Allow unplayable formats to be extracted and downloaded.
206 ignore_no_formats_error: Ignore "No video formats" error. Usefull for
207 extracting metadata even if the video is not actually
208 available for download (experimental)
209 format_sort: A list of fields by which to sort the video formats.
210 See "Sorting Formats" for more details.
211 format_sort_force: Force the given format_sort. see "Sorting Formats"
212 for more details.
213 prefer_free_formats: Whether to prefer video formats with free containers
214 over non-free ones of same quality.
215 allow_multiple_video_streams: Allow multiple video streams to be merged
216 into a single file
217 allow_multiple_audio_streams: Allow multiple audio streams to be merged
218 into a single file
219 check_formats Whether to test if the formats are downloadable.
220 Can be True (check all), False (check none),
221 'selected' (check selected formats),
222 or None (check only if requested by extractor)
223 paths: Dictionary of output paths. The allowed keys are 'home'
224 'temp' and the keys of OUTTMPL_TYPES (in utils.py)
225 outtmpl: Dictionary of templates for output names. Allowed keys
226 are 'default' and the keys of OUTTMPL_TYPES (in utils.py).
227 For compatibility with youtube-dl, a single string can also be used
228 outtmpl_na_placeholder: Placeholder for unavailable meta fields.
229 restrictfilenames: Do not allow "&" and spaces in file names
230 trim_file_name: Limit length of filename (extension excluded)
231 windowsfilenames: Force the filenames to be windows compatible
232 ignoreerrors: Do not stop on download/postprocessing errors.
233 Can be 'only_download' to ignore only download errors.
234 Default is 'only_download' for CLI, but False for API
235 skip_playlist_after_errors: Number of allowed failures until the rest of
236 the playlist is skipped
237 force_generic_extractor: Force downloader to use the generic extractor
238 overwrites: Overwrite all video and metadata files if True,
239 overwrite only non-video files if None
240 and don't overwrite any file if False
241 For compatibility with youtube-dl,
242 "nooverwrites" may also be used instead
243 playlist_items: Specific indices of playlist to download.
244 playlistrandom: Download playlist items in random order.
245 lazy_playlist: Process playlist entries as they are received.
246 matchtitle: Download only matching titles.
247 rejecttitle: Reject downloads for matching titles.
248 logger: Log messages to a logging.Logger instance.
249 logtostderr: Log messages to stderr instead of stdout.
250 consoletitle: Display progress in console window's titlebar.
251 writedescription: Write the video description to a .description file
252 writeinfojson: Write the video description to a .info.json file
253 clean_infojson: Remove private fields from the infojson
254 getcomments: Extract video comments. This will not be written to disk
255 unless writeinfojson is also given
256 writeannotations: Write the video annotations to a .annotations.xml file
257 writethumbnail: Write the thumbnail image to a file
258 allow_playlist_files: Whether to write playlists' description, infojson etc
259 also to disk when using the 'write*' options
260 write_all_thumbnails: Write all thumbnail formats to files
261 writelink: Write an internet shortcut file, depending on the
262 current platform (.url/.webloc/.desktop)
263 writeurllink: Write a Windows internet shortcut file (.url)
264 writewebloclink: Write a macOS internet shortcut file (.webloc)
265 writedesktoplink: Write a Linux internet shortcut file (.desktop)
266 writesubtitles: Write the video subtitles to a file
267 writeautomaticsub: Write the automatically generated subtitles to a file
268 listsubtitles: Lists all available subtitles for the video
269 subtitlesformat: The format code for subtitles
270 subtitleslangs: List of languages of the subtitles to download (can be regex).
271 The list may contain "all" to refer to all the available
272 subtitles. The language can be prefixed with a "-" to
273 exclude it from the requested languages. Eg: ['all', '-live_chat']
274 keepvideo: Keep the video file after post-processing
275 daterange: A DateRange object, download only if the upload_date is in the range.
276 skip_download: Skip the actual download of the video file
277 cachedir: Location of the cache files in the filesystem.
278 False to disable filesystem cache.
279 noplaylist: Download single video instead of a playlist if in doubt.
280 age_limit: An integer representing the user's age in years.
281 Unsuitable videos for the given age are skipped.
282 min_views: An integer representing the minimum view count the video
283 must have in order to not be skipped.
284 Videos without view count information are always
285 downloaded. None for no limit.
286 max_views: An integer representing the maximum view count.
287 Videos that are more popular than that are not
288 downloaded.
289 Videos without view count information are always
290 downloaded. None for no limit.
291 download_archive: File name of a file where all downloads are recorded.
292 Videos already present in the file are not downloaded
293 again.
294 break_on_existing: Stop the download process after attempting to download a
295 file that is in the archive.
296 break_on_reject: Stop the download process when encountering a video that
297 has been filtered out.
298 break_per_url: Whether break_on_reject and break_on_existing
299 should act on each input URL as opposed to for the entire queue
300 cookiefile: File name or text stream from where cookies should be read and dumped to
301 cookiesfrombrowser: A tuple containing the name of the browser, the profile
302 name/pathfrom where cookies are loaded, and the name of the
303 keyring. Eg: ('chrome', ) or ('vivaldi', 'default', 'BASICTEXT')
304 legacyserverconnect: Explicitly allow HTTPS connection to servers that do not
305 support RFC 5746 secure renegotiation
306 nocheckcertificate: Do not verify SSL certificates
307 client_certificate: Path to client certificate file in PEM format. May include the private key
308 client_certificate_key: Path to private key file for client certificate
309 client_certificate_password: Password for client certificate private key, if encrypted.
310 If not provided and the key is encrypted, yt-dlp will ask interactively
311 prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
312 (Only supported by some extractors)
313 http_headers: A dictionary of custom headers to be used for all requests
314 proxy: URL of the proxy server to use
315 geo_verification_proxy: URL of the proxy to use for IP address verification
316 on geo-restricted sites.
317 socket_timeout: Time to wait for unresponsive hosts, in seconds
318 bidi_workaround: Work around buggy terminals without bidirectional text
319 support, using fridibi
320 debug_printtraffic:Print out sent and received HTTP traffic
321 default_search: Prepend this string if an input url is not valid.
322 'auto' for elaborate guessing
323 encoding: Use this encoding instead of the system-specified.
324 extract_flat: Whether to resolve and process url_results further
325 * False: Always process (default)
326 * True: Never process
327 * 'in_playlist': Do not process inside playlist/multi_video
328 * 'discard': Always process, but don't return the result
329 from inside playlist/multi_video
330 * 'discard_in_playlist': Same as "discard", but only for
331 playlists (not multi_video)
332 wait_for_video: If given, wait for scheduled streams to become available.
333 The value should be a tuple containing the range
334 (min_secs, max_secs) to wait between retries
335 postprocessors: A list of dictionaries, each with an entry
336 * key: The name of the postprocessor. See
337 yt_dlp/postprocessor/__init__.py for a list.
338 * when: When to run the postprocessor. Allowed values are
339 the entries of utils.POSTPROCESS_WHEN
340 Assumed to be 'post_process' if not given
341 progress_hooks: A list of functions that get called on download
342 progress, with a dictionary with the entries
343 * status: One of "downloading", "error", or "finished".
344 Check this first and ignore unknown values.
345 * info_dict: The extracted info_dict
346
347 If status is one of "downloading", or "finished", the
348 following properties may also be present:
349 * filename: The final filename (always present)
350 * tmpfilename: The filename we're currently writing to
351 * downloaded_bytes: Bytes on disk
352 * total_bytes: Size of the whole file, None if unknown
353 * total_bytes_estimate: Guess of the eventual file size,
354 None if unavailable.
355 * elapsed: The number of seconds since download started.
356 * eta: The estimated time in seconds, None if unknown
357 * speed: The download speed in bytes/second, None if
358 unknown
359 * fragment_index: The counter of the currently
360 downloaded video fragment.
361 * fragment_count: The number of fragments (= individual
362 files that will be merged)
363
364 Progress hooks are guaranteed to be called at least once
365 (with status "finished") if the download is successful.
366 postprocessor_hooks: A list of functions that get called on postprocessing
367 progress, with a dictionary with the entries
368 * status: One of "started", "processing", or "finished".
369 Check this first and ignore unknown values.
370 * postprocessor: Name of the postprocessor
371 * info_dict: The extracted info_dict
372
373 Progress hooks are guaranteed to be called at least twice
374 (with status "started" and "finished") if the processing is successful.
375 merge_output_format: Extension to use when merging formats.
376 final_ext: Expected final extension; used to detect when the file was
377 already downloaded and converted
378 fixup: Automatically correct known faults of the file.
379 One of:
380 - "never": do nothing
381 - "warn": only emit a warning
382 - "detect_or_warn": check whether we can do anything
383 about it, warn otherwise (default)
384 source_address: Client-side IP address to bind to.
385 sleep_interval_requests: Number of seconds to sleep between requests
386 during extraction
387 sleep_interval: Number of seconds to sleep before each download when
388 used alone or a lower bound of a range for randomized
389 sleep before each download (minimum possible number
390 of seconds to sleep) when used along with
391 max_sleep_interval.
392 max_sleep_interval:Upper bound of a range for randomized sleep before each
393 download (maximum possible number of seconds to sleep).
394 Must only be used along with sleep_interval.
395 Actual sleep time will be a random float from range
396 [sleep_interval; max_sleep_interval].
397 sleep_interval_subtitles: Number of seconds to sleep before each subtitle download
398 listformats: Print an overview of available video formats and exit.
399 list_thumbnails: Print a table of all thumbnails and exit.
400 match_filter: A function that gets called for every video with the signature
401 (info_dict, *, incomplete: bool) -> Optional[str]
402 For backward compatibility with youtube-dl, the signature
403 (info_dict) -> Optional[str] is also allowed.
404 - If it returns a message, the video is ignored.
405 - If it returns None, the video is downloaded.
406 - If it returns utils.NO_DEFAULT, the user is interactively
407 asked whether to download the video.
408 match_filter_func in utils.py is one example for this.
409 no_color: Do not emit color codes in output.
410 geo_bypass: Bypass geographic restriction via faking X-Forwarded-For
411 HTTP header
412 geo_bypass_country:
413 Two-letter ISO 3166-2 country code that will be used for
414 explicit geographic restriction bypassing via faking
415 X-Forwarded-For HTTP header
416 geo_bypass_ip_block:
417 IP range in CIDR notation that will be used similarly to
418 geo_bypass_country
419 external_downloader: A dictionary of protocol keys and the executable of the
420 external downloader to use for it. The allowed protocols
421 are default|http|ftp|m3u8|dash|rtsp|rtmp|mms.
422 Set the value to 'native' to use the native downloader
423 compat_opts: Compatibility options. See "Differences in default behavior".
424 The following options do not work when used through the API:
425 filename, abort-on-error, multistreams, no-live-chat, format-sort
426 no-clean-infojson, no-playlist-metafiles, no-keep-subs, no-attach-info-json.
427 Refer __init__.py for their implementation
428 progress_template: Dictionary of templates for progress outputs.
429 Allowed keys are 'download', 'postprocess',
430 'download-title' (console title) and 'postprocess-title'.
431 The template is mapped on a dictionary with keys 'progress' and 'info'
432 retry_sleep_functions: Dictionary of functions that takes the number of attempts
433 as argument and returns the time to sleep in seconds.
434 Allowed keys are 'http', 'fragment', 'file_access'
435 download_ranges: A callback function that gets called for every video with
436 the signature (info_dict, ydl) -> Iterable[Section].
437 Only the returned sections will be downloaded.
438 Each Section is a dict with the following keys:
439 * start_time: Start time of the section in seconds
440 * end_time: End time of the section in seconds
441 * title: Section title (Optional)
442 * index: Section number (Optional)
443 force_keyframes_at_cuts: Re-encode the video when downloading ranges to get precise cuts
444 noprogress: Do not print the progress bar
445
446 The following parameters are not used by YoutubeDL itself, they are used by
447 the downloader (see yt_dlp/downloader/common.py):
448 nopart, updatetime, buffersize, ratelimit, throttledratelimit, min_filesize,
449 max_filesize, test, noresizebuffer, retries, file_access_retries, fragment_retries,
450 continuedl, xattr_set_filesize, hls_use_mpegts, http_chunk_size,
451 external_downloader_args, concurrent_fragment_downloads.
452
453 The following options are used by the post processors:
454 ffmpeg_location: Location of the ffmpeg/avconv binary; either the path
455 to the binary or its containing directory.
456 postprocessor_args: A dictionary of postprocessor/executable keys (in lower case)
457 and a list of additional command-line arguments for the
458 postprocessor/executable. The dict can also have "PP+EXE" keys
459 which are used when the given exe is used by the given PP.
460 Use 'default' as the name for arguments to passed to all PP
461 For compatibility with youtube-dl, a single list of args
462 can also be used
463
464 The following options are used by the extractors:
465 extractor_retries: Number of times to retry for known errors
466 dynamic_mpd: Whether to process dynamic DASH manifests (default: True)
467 hls_split_discontinuity: Split HLS playlists to different formats at
468 discontinuities such as ad breaks (default: False)
469 extractor_args: A dictionary of arguments to be passed to the extractors.
470 See "EXTRACTOR ARGUMENTS" for details.
471 Eg: {'youtube': {'skip': ['dash', 'hls']}}
472 mark_watched: Mark videos watched (even with --simulate). Only for YouTube
473
474 The following options are deprecated and may be removed in the future:
475
476 playliststart: - Use playlist_items
477 Playlist item to start at.
478 playlistend: - Use playlist_items
479 Playlist item to end at.
480 playlistreverse: - Use playlist_items
481 Download playlist items in reverse order.
482 forceurl: - Use forceprint
483 Force printing final URL.
484 forcetitle: - Use forceprint
485 Force printing title.
486 forceid: - Use forceprint
487 Force printing ID.
488 forcethumbnail: - Use forceprint
489 Force printing thumbnail URL.
490 forcedescription: - Use forceprint
491 Force printing description.
492 forcefilename: - Use forceprint
493 Force printing final filename.
494 forceduration: - Use forceprint
495 Force printing duration.
496 allsubtitles: - Use subtitleslangs = ['all']
497 Downloads all the subtitles of the video
498 (requires writesubtitles or writeautomaticsub)
499 include_ads: - Doesn't work
500 Download ads as well
501 call_home: - Not implemented
502 Boolean, true iff we are allowed to contact the
503 yt-dlp servers for debugging.
504 post_hooks: - Register a custom postprocessor
505 A list of functions that get called as the final step
506 for each video file, after all postprocessors have been
507 called. The filename will be passed as the only argument.
508 hls_prefer_native: - Use external_downloader = {'m3u8': 'native'} or {'m3u8': 'ffmpeg'}.
509 Use the native HLS downloader instead of ffmpeg/avconv
510 if True, otherwise use ffmpeg/avconv if False, otherwise
511 use downloader suggested by extractor if None.
512 prefer_ffmpeg: - avconv support is deprecated
513 If False, use avconv instead of ffmpeg if both are available,
514 otherwise prefer ffmpeg.
515 youtube_include_dash_manifest: - Use extractor_args
516 If True (default), DASH manifests and related
517 data will be downloaded and processed by extractor.
518 You can reduce network I/O by disabling it if you don't
519 care about DASH. (only for youtube)
520 youtube_include_hls_manifest: - Use extractor_args
521 If True (default), HLS manifests and related
522 data will be downloaded and processed by extractor.
523 You can reduce network I/O by disabling it if you don't
524 care about HLS. (only for youtube)
525 """
526
527 _NUMERIC_FIELDS = {
528 'width', 'height', 'tbr', 'abr', 'asr', 'vbr', 'fps', 'filesize', 'filesize_approx',
529 'timestamp', 'release_timestamp',
530 'duration', 'view_count', 'like_count', 'dislike_count', 'repost_count',
531 'average_rating', 'comment_count', 'age_limit',
532 'start_time', 'end_time',
533 'chapter_number', 'season_number', 'episode_number',
534 'track_number', 'disc_number', 'release_year',
535 }
536
537 _format_fields = {
538 # NB: Keep in sync with the docstring of extractor/common.py
539 'url', 'manifest_url', 'manifest_stream_number', 'ext', 'format', 'format_id', 'format_note',
540 'width', 'height', 'resolution', 'dynamic_range', 'tbr', 'abr', 'acodec', 'asr',
541 'vbr', 'fps', 'vcodec', 'container', 'filesize', 'filesize_approx',
542 'player_url', 'protocol', 'fragment_base_url', 'fragments', 'is_from_start',
543 'preference', 'language', 'language_preference', 'quality', 'source_preference',
544 'http_headers', 'stretched_ratio', 'no_resume', 'has_drm', 'downloader_options',
545 'page_url', 'app', 'play_path', 'tc_url', 'flash_version', 'rtmp_live', 'rtmp_conn', 'rtmp_protocol', 'rtmp_real_time'
546 }
547 _format_selection_exts = {
548 'audio': set(MEDIA_EXTENSIONS.common_audio),
549 'video': set(MEDIA_EXTENSIONS.common_video + ('3gp', )),
550 'storyboards': set(MEDIA_EXTENSIONS.storyboards),
551 }
552
553 def __init__(self, params=None, auto_init=True):
554 """Create a FileDownloader object with the given options.
555 @param auto_init Whether to load the default extractors and print header (if verbose).
556 Set to 'no_verbose_header' to not print the header
557 """
558 if params is None:
559 params = {}
560 self.params = params
561 self._ies = {}
562 self._ies_instances = {}
563 self._pps = {k: [] for k in POSTPROCESS_WHEN}
564 self._printed_messages = set()
565 self._first_webpage_request = True
566 self._post_hooks = []
567 self._progress_hooks = []
568 self._postprocessor_hooks = []
569 self._download_retcode = 0
570 self._num_downloads = 0
571 self._num_videos = 0
572 self._playlist_level = 0
573 self._playlist_urls = set()
574 self.cache = Cache(self)
575
576 windows_enable_vt_mode()
577 stdout = sys.stderr if self.params.get('logtostderr') else sys.stdout
578 self._out_files = Namespace(
579 out=stdout,
580 error=sys.stderr,
581 screen=sys.stderr if self.params.get('quiet') else stdout,
582 console=None if compat_os_name == 'nt' else next(
583 filter(supports_terminal_sequences, (sys.stderr, sys.stdout)), None)
584 )
585 self._allow_colors = Namespace(**{
586 type_: not self.params.get('no_color') and supports_terminal_sequences(stream)
587 for type_, stream in self._out_files.items_ if type_ != 'console'
588 })
589
590 # The code is left like this to be reused for future deprecations
591 MIN_SUPPORTED, MIN_RECOMMENDED = (3, 7), (3, 7)
592 current_version = sys.version_info[:2]
593 if current_version < MIN_RECOMMENDED:
594 msg = ('Support for Python version %d.%d has been deprecated. '
595 'See https://github.com/yt-dlp/yt-dlp/issues/3764 for more details.'
596 '\n You will no longer receive updates on this version')
597 if current_version < MIN_SUPPORTED:
598 msg = 'Python version %d.%d is no longer supported'
599 self.deprecation_warning(
600 f'{msg}! Please update to Python %d.%d or above' % (*current_version, *MIN_RECOMMENDED))
601
602 if self.params.get('allow_unplayable_formats'):
603 self.report_warning(
604 f'You have asked for {self._format_err("UNPLAYABLE", self.Styles.EMPHASIS)} formats to be listed/downloaded. '
605 'This is a developer option intended for debugging. \n'
606 ' If you experience any issues while using this option, '
607 f'{self._format_err("DO NOT", self.Styles.ERROR)} open a bug report')
608
609 def check_deprecated(param, option, suggestion):
610 if self.params.get(param) is not None:
611 self.report_warning(f'{option} is deprecated. Use {suggestion} instead')
612 return True
613 return False
614
615 if check_deprecated('cn_verification_proxy', '--cn-verification-proxy', '--geo-verification-proxy'):
616 if self.params.get('geo_verification_proxy') is None:
617 self.params['geo_verification_proxy'] = self.params['cn_verification_proxy']
618
619 check_deprecated('autonumber', '--auto-number', '-o "%(autonumber)s-%(title)s.%(ext)s"')
620 check_deprecated('usetitle', '--title', '-o "%(title)s-%(id)s.%(ext)s"')
621 check_deprecated('useid', '--id', '-o "%(id)s.%(ext)s"')
622
623 for msg in self.params.get('_warnings', []):
624 self.report_warning(msg)
625 for msg in self.params.get('_deprecation_warnings', []):
626 self.deprecation_warning(msg)
627
628 self.params['compat_opts'] = set(self.params.get('compat_opts', ()))
629 if 'list-formats' in self.params['compat_opts']:
630 self.params['listformats_table'] = False
631
632 if 'overwrites' not in self.params and self.params.get('nooverwrites') is not None:
633 # nooverwrites was unnecessarily changed to overwrites
634 # in 0c3d0f51778b153f65c21906031c2e091fcfb641
635 # This ensures compatibility with both keys
636 self.params['overwrites'] = not self.params['nooverwrites']
637 elif self.params.get('overwrites') is None:
638 self.params.pop('overwrites', None)
639 else:
640 self.params['nooverwrites'] = not self.params['overwrites']
641
642 self.params.setdefault('forceprint', {})
643 self.params.setdefault('print_to_file', {})
644
645 # Compatibility with older syntax
646 if not isinstance(params['forceprint'], dict):
647 self.params['forceprint'] = {'video': params['forceprint']}
648
649 if self.params.get('bidi_workaround', False):
650 try:
651 import pty
652 master, slave = pty.openpty()
653 width = shutil.get_terminal_size().columns
654 width_args = [] if width is None else ['-w', str(width)]
655 sp_kwargs = {'stdin': subprocess.PIPE, 'stdout': slave, 'stderr': self._out_files.error}
656 try:
657 self._output_process = Popen(['bidiv'] + width_args, **sp_kwargs)
658 except OSError:
659 self._output_process = Popen(['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
660 self._output_channel = os.fdopen(master, 'rb')
661 except OSError as ose:
662 if ose.errno == errno.ENOENT:
663 self.report_warning(
664 'Could not find fribidi executable, ignoring --bidi-workaround. '
665 'Make sure that fribidi is an executable file in one of the directories in your $PATH.')
666 else:
667 raise
668
669 if auto_init:
670 if auto_init != 'no_verbose_header':
671 self.print_debug_header()
672 self.add_default_info_extractors()
673
674 if (sys.platform != 'win32'
675 and sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968']
676 and not self.params.get('restrictfilenames', False)):
677 # Unicode filesystem API will throw errors (#1474, #13027)
678 self.report_warning(
679 'Assuming --restrict-filenames since file system encoding '
680 'cannot encode all characters. '
681 'Set the LC_ALL environment variable to fix this.')
682 self.params['restrictfilenames'] = True
683
684 self._parse_outtmpl()
685
686 # Creating format selector here allows us to catch syntax errors before the extraction
687 self.format_selector = (
688 self.params.get('format') if self.params.get('format') in (None, '-')
689 else self.params['format'] if callable(self.params['format'])
690 else self.build_format_selector(self.params['format']))
691
692 # Set http_headers defaults according to std_headers
693 self.params['http_headers'] = merge_headers(std_headers, self.params.get('http_headers', {}))
694
695 hooks = {
696 'post_hooks': self.add_post_hook,
697 'progress_hooks': self.add_progress_hook,
698 'postprocessor_hooks': self.add_postprocessor_hook,
699 }
700 for opt, fn in hooks.items():
701 for ph in self.params.get(opt, []):
702 fn(ph)
703
704 for pp_def_raw in self.params.get('postprocessors', []):
705 pp_def = dict(pp_def_raw)
706 when = pp_def.pop('when', 'post_process')
707 self.add_post_processor(
708 get_postprocessor(pp_def.pop('key'))(self, **pp_def),
709 when=when)
710
711 self._setup_opener()
712 register_socks_protocols()
713
714 def preload_download_archive(fn):
715 """Preload the archive, if any is specified"""
716 if fn is None:
717 return False
718 self.write_debug(f'Loading archive file {fn!r}')
719 try:
720 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
721 for line in archive_file:
722 self.archive.add(line.strip())
723 except OSError as ioe:
724 if ioe.errno != errno.ENOENT:
725 raise
726 return False
727 return True
728
729 self.archive = set()
730 preload_download_archive(self.params.get('download_archive'))
731
732 def warn_if_short_id(self, argv):
733 # short YouTube ID starting with dash?
734 idxs = [
735 i for i, a in enumerate(argv)
736 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
737 if idxs:
738 correct_argv = (
739 ['yt-dlp']
740 + [a for i, a in enumerate(argv) if i not in idxs]
741 + ['--'] + [argv[i] for i in idxs]
742 )
743 self.report_warning(
744 'Long argument string detected. '
745 'Use -- to separate parameters and URLs, like this:\n%s' %
746 args_to_str(correct_argv))
747
748 def add_info_extractor(self, ie):
749 """Add an InfoExtractor object to the end of the list."""
750 ie_key = ie.ie_key()
751 self._ies[ie_key] = ie
752 if not isinstance(ie, type):
753 self._ies_instances[ie_key] = ie
754 ie.set_downloader(self)
755
756 def _get_info_extractor_class(self, ie_key):
757 ie = self._ies.get(ie_key)
758 if ie is None:
759 ie = get_info_extractor(ie_key)
760 self.add_info_extractor(ie)
761 return ie
762
763 def get_info_extractor(self, ie_key):
764 """
765 Get an instance of an IE with name ie_key, it will try to get one from
766 the _ies list, if there's no instance it will create a new one and add
767 it to the extractor list.
768 """
769 ie = self._ies_instances.get(ie_key)
770 if ie is None:
771 ie = get_info_extractor(ie_key)()
772 self.add_info_extractor(ie)
773 return ie
774
775 def add_default_info_extractors(self):
776 """
777 Add the InfoExtractors returned by gen_extractors to the end of the list
778 """
779 for ie in gen_extractor_classes():
780 self.add_info_extractor(ie)
781
782 def add_post_processor(self, pp, when='post_process'):
783 """Add a PostProcessor object to the end of the chain."""
784 assert when in POSTPROCESS_WHEN, f'Invalid when={when}'
785 self._pps[when].append(pp)
786 pp.set_downloader(self)
787
788 def add_post_hook(self, ph):
789 """Add the post hook"""
790 self._post_hooks.append(ph)
791
792 def add_progress_hook(self, ph):
793 """Add the download progress hook"""
794 self._progress_hooks.append(ph)
795
796 def add_postprocessor_hook(self, ph):
797 """Add the postprocessing progress hook"""
798 self._postprocessor_hooks.append(ph)
799 for pps in self._pps.values():
800 for pp in pps:
801 pp.add_progress_hook(ph)
802
803 def _bidi_workaround(self, message):
804 if not hasattr(self, '_output_channel'):
805 return message
806
807 assert hasattr(self, '_output_process')
808 assert isinstance(message, str)
809 line_count = message.count('\n') + 1
810 self._output_process.stdin.write((message + '\n').encode())
811 self._output_process.stdin.flush()
812 res = ''.join(self._output_channel.readline().decode()
813 for _ in range(line_count))
814 return res[:-len('\n')]
815
816 def _write_string(self, message, out=None, only_once=False):
817 if only_once:
818 if message in self._printed_messages:
819 return
820 self._printed_messages.add(message)
821 write_string(message, out=out, encoding=self.params.get('encoding'))
822
823 def to_stdout(self, message, skip_eol=False, quiet=None):
824 """Print message to stdout"""
825 if quiet is not None:
826 self.deprecation_warning('"YoutubeDL.to_stdout" no longer accepts the argument quiet. Use "YoutubeDL.to_screen" instead')
827 if skip_eol is not False:
828 self.deprecation_warning('"YoutubeDL.to_stdout" no longer accepts the argument skip_eol. Use "YoutubeDL.to_screen" instead')
829 self._write_string(f'{self._bidi_workaround(message)}\n', self._out_files.out)
830
831 def to_screen(self, message, skip_eol=False, quiet=None):
832 """Print message to screen if not in quiet mode"""
833 if self.params.get('logger'):
834 self.params['logger'].debug(message)
835 return
836 if (self.params.get('quiet') if quiet is None else quiet) and not self.params.get('verbose'):
837 return
838 self._write_string(
839 '%s%s' % (self._bidi_workaround(message), ('' if skip_eol else '\n')),
840 self._out_files.screen)
841
842 def to_stderr(self, message, only_once=False):
843 """Print message to stderr"""
844 assert isinstance(message, str)
845 if self.params.get('logger'):
846 self.params['logger'].error(message)
847 else:
848 self._write_string(f'{self._bidi_workaround(message)}\n', self._out_files.error, only_once=only_once)
849
850 def _send_console_code(self, code):
851 if compat_os_name == 'nt' or not self._out_files.console:
852 return
853 self._write_string(code, self._out_files.console)
854
855 def to_console_title(self, message):
856 if not self.params.get('consoletitle', False):
857 return
858 message = remove_terminal_sequences(message)
859 if compat_os_name == 'nt':
860 if ctypes.windll.kernel32.GetConsoleWindow():
861 # c_wchar_p() might not be necessary if `message` is
862 # already of type unicode()
863 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
864 else:
865 self._send_console_code(f'\033]0;{message}\007')
866
867 def save_console_title(self):
868 if not self.params.get('consoletitle') or self.params.get('simulate'):
869 return
870 self._send_console_code('\033[22;0t') # Save the title on stack
871
872 def restore_console_title(self):
873 if not self.params.get('consoletitle') or self.params.get('simulate'):
874 return
875 self._send_console_code('\033[23;0t') # Restore the title from stack
876
877 def __enter__(self):
878 self.save_console_title()
879 return self
880
881 def __exit__(self, *args):
882 self.restore_console_title()
883
884 if self.params.get('cookiefile') is not None:
885 self.cookiejar.save(ignore_discard=True, ignore_expires=True)
886
887 def trouble(self, message=None, tb=None, is_error=True):
888 """Determine action to take when a download problem appears.
889
890 Depending on if the downloader has been configured to ignore
891 download errors or not, this method may throw an exception or
892 not when errors are found, after printing the message.
893
894 @param tb If given, is additional traceback information
895 @param is_error Whether to raise error according to ignorerrors
896 """
897 if message is not None:
898 self.to_stderr(message)
899 if self.params.get('verbose'):
900 if tb is None:
901 if sys.exc_info()[0]: # if .trouble has been called from an except block
902 tb = ''
903 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
904 tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
905 tb += encode_compat_str(traceback.format_exc())
906 else:
907 tb_data = traceback.format_list(traceback.extract_stack())
908 tb = ''.join(tb_data)
909 if tb:
910 self.to_stderr(tb)
911 if not is_error:
912 return
913 if not self.params.get('ignoreerrors'):
914 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
915 exc_info = sys.exc_info()[1].exc_info
916 else:
917 exc_info = sys.exc_info()
918 raise DownloadError(message, exc_info)
919 self._download_retcode = 1
920
921 Styles = Namespace(
922 HEADERS='yellow',
923 EMPHASIS='light blue',
924 FILENAME='green',
925 ID='green',
926 DELIM='blue',
927 ERROR='red',
928 WARNING='yellow',
929 SUPPRESS='light black',
930 )
931
932 def _format_text(self, handle, allow_colors, text, f, fallback=None, *, test_encoding=False):
933 text = str(text)
934 if test_encoding:
935 original_text = text
936 # handle.encoding can be None. See https://github.com/yt-dlp/yt-dlp/issues/2711
937 encoding = self.params.get('encoding') or getattr(handle, 'encoding', None) or 'ascii'
938 text = text.encode(encoding, 'ignore').decode(encoding)
939 if fallback is not None and text != original_text:
940 text = fallback
941 return format_text(text, f) if allow_colors else text if fallback is None else fallback
942
943 def _format_out(self, *args, **kwargs):
944 return self._format_text(self._out_files.out, self._allow_colors.out, *args, **kwargs)
945
946 def _format_screen(self, *args, **kwargs):
947 return self._format_text(self._out_files.screen, self._allow_colors.screen, *args, **kwargs)
948
949 def _format_err(self, *args, **kwargs):
950 return self._format_text(self._out_files.error, self._allow_colors.error, *args, **kwargs)
951
952 def report_warning(self, message, only_once=False):
953 '''
954 Print the message to stderr, it will be prefixed with 'WARNING:'
955 If stderr is a tty file the 'WARNING:' will be colored
956 '''
957 if self.params.get('logger') is not None:
958 self.params['logger'].warning(message)
959 else:
960 if self.params.get('no_warnings'):
961 return
962 self.to_stderr(f'{self._format_err("WARNING:", self.Styles.WARNING)} {message}', only_once)
963
964 def deprecation_warning(self, message):
965 if self.params.get('logger') is not None:
966 self.params['logger'].warning(f'DeprecationWarning: {message}')
967 else:
968 self.to_stderr(f'{self._format_err("DeprecationWarning:", self.Styles.ERROR)} {message}', True)
969
970 def report_error(self, message, *args, **kwargs):
971 '''
972 Do the same as trouble, but prefixes the message with 'ERROR:', colored
973 in red if stderr is a tty file.
974 '''
975 self.trouble(f'{self._format_err("ERROR:", self.Styles.ERROR)} {message}', *args, **kwargs)
976
977 def write_debug(self, message, only_once=False):
978 '''Log debug message or Print message to stderr'''
979 if not self.params.get('verbose', False):
980 return
981 message = f'[debug] {message}'
982 if self.params.get('logger'):
983 self.params['logger'].debug(message)
984 else:
985 self.to_stderr(message, only_once)
986
987 def report_file_already_downloaded(self, file_name):
988 """Report file has already been fully downloaded."""
989 try:
990 self.to_screen('[download] %s has already been downloaded' % file_name)
991 except UnicodeEncodeError:
992 self.to_screen('[download] The file has already been downloaded')
993
994 def report_file_delete(self, file_name):
995 """Report that existing file will be deleted."""
996 try:
997 self.to_screen('Deleting existing file %s' % file_name)
998 except UnicodeEncodeError:
999 self.to_screen('Deleting existing file')
1000
1001 def raise_no_formats(self, info, forced=False, *, msg=None):
1002 has_drm = info.get('_has_drm')
1003 ignored, expected = self.params.get('ignore_no_formats_error'), bool(msg)
1004 msg = msg or has_drm and 'This video is DRM protected' or 'No video formats found!'
1005 if forced or not ignored:
1006 raise ExtractorError(msg, video_id=info['id'], ie=info['extractor'],
1007 expected=has_drm or ignored or expected)
1008 else:
1009 self.report_warning(msg)
1010
1011 def parse_outtmpl(self):
1012 self.deprecation_warning('"YoutubeDL.parse_outtmpl" is deprecated and may be removed in a future version')
1013 self._parse_outtmpl()
1014 return self.params['outtmpl']
1015
1016 def _parse_outtmpl(self):
1017 sanitize = IDENTITY
1018 if self.params.get('restrictfilenames'): # Remove spaces in the default template
1019 sanitize = lambda x: x.replace(' - ', ' ').replace(' ', '-')
1020
1021 outtmpl = self.params.setdefault('outtmpl', {})
1022 if not isinstance(outtmpl, dict):
1023 self.params['outtmpl'] = outtmpl = {'default': outtmpl}
1024 outtmpl.update({k: sanitize(v) for k, v in DEFAULT_OUTTMPL.items() if outtmpl.get(k) is None})
1025
1026 def get_output_path(self, dir_type='', filename=None):
1027 paths = self.params.get('paths', {})
1028 assert isinstance(paths, dict)
1029 path = os.path.join(
1030 expand_path(paths.get('home', '').strip()),
1031 expand_path(paths.get(dir_type, '').strip()) if dir_type else '',
1032 filename or '')
1033 return sanitize_path(path, force=self.params.get('windowsfilenames'))
1034
1035 @staticmethod
1036 def _outtmpl_expandpath(outtmpl):
1037 # expand_path translates '%%' into '%' and '$$' into '$'
1038 # correspondingly that is not what we want since we need to keep
1039 # '%%' intact for template dict substitution step. Working around
1040 # with boundary-alike separator hack.
1041 sep = ''.join([random.choice(ascii_letters) for _ in range(32)])
1042 outtmpl = outtmpl.replace('%%', f'%{sep}%').replace('$$', f'${sep}$')
1043
1044 # outtmpl should be expand_path'ed before template dict substitution
1045 # because meta fields may contain env variables we don't want to
1046 # be expanded. For example, for outtmpl "%(title)s.%(ext)s" and
1047 # title "Hello $PATH", we don't want `$PATH` to be expanded.
1048 return expand_path(outtmpl).replace(sep, '')
1049
1050 @staticmethod
1051 def escape_outtmpl(outtmpl):
1052 ''' Escape any remaining strings like %s, %abc% etc. '''
1053 return re.sub(
1054 STR_FORMAT_RE_TMPL.format('', '(?![%(\0])'),
1055 lambda mobj: ('' if mobj.group('has_key') else '%') + mobj.group(0),
1056 outtmpl)
1057
1058 @classmethod
1059 def validate_outtmpl(cls, outtmpl):
1060 ''' @return None or Exception object '''
1061 outtmpl = re.sub(
1062 STR_FORMAT_RE_TMPL.format('[^)]*', '[ljhqBUDS]'),
1063 lambda mobj: f'{mobj.group(0)[:-1]}s',
1064 cls._outtmpl_expandpath(outtmpl))
1065 try:
1066 cls.escape_outtmpl(outtmpl) % collections.defaultdict(int)
1067 return None
1068 except ValueError as err:
1069 return err
1070
1071 @staticmethod
1072 def _copy_infodict(info_dict):
1073 info_dict = dict(info_dict)
1074 info_dict.pop('__postprocessors', None)
1075 info_dict.pop('__pending_error', None)
1076 return info_dict
1077
1078 def prepare_outtmpl(self, outtmpl, info_dict, sanitize=False):
1079 """ Make the outtmpl and info_dict suitable for substitution: ydl.escape_outtmpl(outtmpl) % info_dict
1080 @param sanitize Whether to sanitize the output as a filename.
1081 For backward compatibility, a function can also be passed
1082 """
1083
1084 info_dict.setdefault('epoch', int(time.time())) # keep epoch consistent once set
1085
1086 info_dict = self._copy_infodict(info_dict)
1087 info_dict['duration_string'] = ( # %(duration>%H-%M-%S)s is wrong if duration > 24hrs
1088 formatSeconds(info_dict['duration'], '-' if sanitize else ':')
1089 if info_dict.get('duration', None) is not None
1090 else None)
1091 info_dict['autonumber'] = int(self.params.get('autonumber_start', 1) - 1 + self._num_downloads)
1092 info_dict['video_autonumber'] = self._num_videos
1093 if info_dict.get('resolution') is None:
1094 info_dict['resolution'] = self.format_resolution(info_dict, default=None)
1095
1096 # For fields playlist_index, playlist_autonumber and autonumber convert all occurrences
1097 # of %(field)s to %(field)0Nd for backward compatibility
1098 field_size_compat_map = {
1099 'playlist_index': number_of_digits(info_dict.get('__last_playlist_index') or 0),
1100 'playlist_autonumber': number_of_digits(info_dict.get('n_entries') or 0),
1101 'autonumber': self.params.get('autonumber_size') or 5,
1102 }
1103
1104 TMPL_DICT = {}
1105 EXTERNAL_FORMAT_RE = re.compile(STR_FORMAT_RE_TMPL.format('[^)]*', f'[{STR_FORMAT_TYPES}ljhqBUDS]'))
1106 MATH_FUNCTIONS = {
1107 '+': float.__add__,
1108 '-': float.__sub__,
1109 }
1110 # Field is of the form key1.key2...
1111 # where keys (except first) can be string, int or slice
1112 FIELD_RE = r'\w*(?:\.(?:\w+|{num}|{num}?(?::{num}?){{1,2}}))*'.format(num=r'(?:-?\d+)')
1113 MATH_FIELD_RE = rf'(?:{FIELD_RE}|-?{NUMBER_RE})'
1114 MATH_OPERATORS_RE = r'(?:%s)' % '|'.join(map(re.escape, MATH_FUNCTIONS.keys()))
1115 INTERNAL_FORMAT_RE = re.compile(rf'''(?x)
1116 (?P<negate>-)?
1117 (?P<fields>{FIELD_RE})
1118 (?P<maths>(?:{MATH_OPERATORS_RE}{MATH_FIELD_RE})*)
1119 (?:>(?P<strf_format>.+?))?
1120 (?P<remaining>
1121 (?P<alternate>(?<!\\),[^|&)]+)?
1122 (?:&(?P<replacement>.*?))?
1123 (?:\|(?P<default>.*?))?
1124 )$''')
1125
1126 def _traverse_infodict(k):
1127 k = k.split('.')
1128 if k[0] == '':
1129 k.pop(0)
1130 return traverse_obj(info_dict, k, is_user_input=True, traverse_string=True)
1131
1132 def get_value(mdict):
1133 # Object traversal
1134 value = _traverse_infodict(mdict['fields'])
1135 # Negative
1136 if mdict['negate']:
1137 value = float_or_none(value)
1138 if value is not None:
1139 value *= -1
1140 # Do maths
1141 offset_key = mdict['maths']
1142 if offset_key:
1143 value = float_or_none(value)
1144 operator = None
1145 while offset_key:
1146 item = re.match(
1147 MATH_FIELD_RE if operator else MATH_OPERATORS_RE,
1148 offset_key).group(0)
1149 offset_key = offset_key[len(item):]
1150 if operator is None:
1151 operator = MATH_FUNCTIONS[item]
1152 continue
1153 item, multiplier = (item[1:], -1) if item[0] == '-' else (item, 1)
1154 offset = float_or_none(item)
1155 if offset is None:
1156 offset = float_or_none(_traverse_infodict(item))
1157 try:
1158 value = operator(value, multiplier * offset)
1159 except (TypeError, ZeroDivisionError):
1160 return None
1161 operator = None
1162 # Datetime formatting
1163 if mdict['strf_format']:
1164 value = strftime_or_none(value, mdict['strf_format'].replace('\\,', ','))
1165
1166 # XXX: Workaround for https://github.com/yt-dlp/yt-dlp/issues/4485
1167 if sanitize and value == '':
1168 value = None
1169 return value
1170
1171 na = self.params.get('outtmpl_na_placeholder', 'NA')
1172
1173 def filename_sanitizer(key, value, restricted=self.params.get('restrictfilenames')):
1174 return sanitize_filename(str(value), restricted=restricted, is_id=(
1175 bool(re.search(r'(^|[_.])id(\.|$)', key))
1176 if 'filename-sanitization' in self.params['compat_opts']
1177 else NO_DEFAULT))
1178
1179 sanitizer = sanitize if callable(sanitize) else filename_sanitizer
1180 sanitize = bool(sanitize)
1181
1182 def _dumpjson_default(obj):
1183 if isinstance(obj, (set, LazyList)):
1184 return list(obj)
1185 return repr(obj)
1186
1187 def create_key(outer_mobj):
1188 if not outer_mobj.group('has_key'):
1189 return outer_mobj.group(0)
1190 key = outer_mobj.group('key')
1191 mobj = re.match(INTERNAL_FORMAT_RE, key)
1192 initial_field = mobj.group('fields') if mobj else ''
1193 value, replacement, default = None, None, na
1194 while mobj:
1195 mobj = mobj.groupdict()
1196 default = mobj['default'] if mobj['default'] is not None else default
1197 value = get_value(mobj)
1198 replacement = mobj['replacement']
1199 if value is None and mobj['alternate']:
1200 mobj = re.match(INTERNAL_FORMAT_RE, mobj['remaining'][1:])
1201 else:
1202 break
1203
1204 fmt = outer_mobj.group('format')
1205 if fmt == 's' and value is not None and key in field_size_compat_map.keys():
1206 fmt = f'0{field_size_compat_map[key]:d}d'
1207
1208 value = default if value is None else value if replacement is None else replacement
1209
1210 flags = outer_mobj.group('conversion') or ''
1211 str_fmt = f'{fmt[:-1]}s'
1212 if fmt[-1] == 'l': # list
1213 delim = '\n' if '#' in flags else ', '
1214 value, fmt = delim.join(map(str, variadic(value, allowed_types=(str, bytes)))), str_fmt
1215 elif fmt[-1] == 'j': # json
1216 value, fmt = json.dumps(value, default=_dumpjson_default, indent=4 if '#' in flags else None), str_fmt
1217 elif fmt[-1] == 'h': # html
1218 value, fmt = escapeHTML(value), str_fmt
1219 elif fmt[-1] == 'q': # quoted
1220 value = map(str, variadic(value) if '#' in flags else [value])
1221 value, fmt = ' '.join(map(compat_shlex_quote, value)), str_fmt
1222 elif fmt[-1] == 'B': # bytes
1223 value = f'%{str_fmt}'.encode() % str(value).encode()
1224 value, fmt = value.decode('utf-8', 'ignore'), 's'
1225 elif fmt[-1] == 'U': # unicode normalized
1226 value, fmt = unicodedata.normalize(
1227 # "+" = compatibility equivalence, "#" = NFD
1228 'NF%s%s' % ('K' if '+' in flags else '', 'D' if '#' in flags else 'C'),
1229 value), str_fmt
1230 elif fmt[-1] == 'D': # decimal suffix
1231 num_fmt, fmt = fmt[:-1].replace('#', ''), 's'
1232 value = format_decimal_suffix(value, f'%{num_fmt}f%s' if num_fmt else '%d%s',
1233 factor=1024 if '#' in flags else 1000)
1234 elif fmt[-1] == 'S': # filename sanitization
1235 value, fmt = filename_sanitizer(initial_field, value, restricted='#' in flags), str_fmt
1236 elif fmt[-1] == 'c':
1237 if value:
1238 value = str(value)[0]
1239 else:
1240 fmt = str_fmt
1241 elif fmt[-1] not in 'rs': # numeric
1242 value = float_or_none(value)
1243 if value is None:
1244 value, fmt = default, 's'
1245
1246 if sanitize:
1247 if fmt[-1] == 'r':
1248 # If value is an object, sanitize might convert it to a string
1249 # So we convert it to repr first
1250 value, fmt = repr(value), str_fmt
1251 if fmt[-1] in 'csr':
1252 value = sanitizer(initial_field, value)
1253
1254 key = '%s\0%s' % (key.replace('%', '%\0'), outer_mobj.group('format'))
1255 TMPL_DICT[key] = value
1256 return '{prefix}%({key}){fmt}'.format(key=key, fmt=fmt, prefix=outer_mobj.group('prefix'))
1257
1258 return EXTERNAL_FORMAT_RE.sub(create_key, outtmpl), TMPL_DICT
1259
1260 def evaluate_outtmpl(self, outtmpl, info_dict, *args, **kwargs):
1261 outtmpl, info_dict = self.prepare_outtmpl(outtmpl, info_dict, *args, **kwargs)
1262 return self.escape_outtmpl(outtmpl) % info_dict
1263
1264 def _prepare_filename(self, info_dict, *, outtmpl=None, tmpl_type=None):
1265 assert None in (outtmpl, tmpl_type), 'outtmpl and tmpl_type are mutually exclusive'
1266 if outtmpl is None:
1267 outtmpl = self.params['outtmpl'].get(tmpl_type or 'default', self.params['outtmpl']['default'])
1268 try:
1269 outtmpl = self._outtmpl_expandpath(outtmpl)
1270 filename = self.evaluate_outtmpl(outtmpl, info_dict, True)
1271 if not filename:
1272 return None
1273
1274 if tmpl_type in ('', 'temp'):
1275 final_ext, ext = self.params.get('final_ext'), info_dict.get('ext')
1276 if final_ext and ext and final_ext != ext and filename.endswith(f'.{final_ext}'):
1277 filename = replace_extension(filename, ext, final_ext)
1278 elif tmpl_type:
1279 force_ext = OUTTMPL_TYPES[tmpl_type]
1280 if force_ext:
1281 filename = replace_extension(filename, force_ext, info_dict.get('ext'))
1282
1283 # https://github.com/blackjack4494/youtube-dlc/issues/85
1284 trim_file_name = self.params.get('trim_file_name', False)
1285 if trim_file_name:
1286 no_ext, *ext = filename.rsplit('.', 2)
1287 filename = join_nonempty(no_ext[:trim_file_name], *ext, delim='.')
1288
1289 return filename
1290 except ValueError as err:
1291 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
1292 return None
1293
1294 def prepare_filename(self, info_dict, dir_type='', *, outtmpl=None, warn=False):
1295 """Generate the output filename"""
1296 if outtmpl:
1297 assert not dir_type, 'outtmpl and dir_type are mutually exclusive'
1298 dir_type = None
1299 filename = self._prepare_filename(info_dict, tmpl_type=dir_type, outtmpl=outtmpl)
1300 if not filename and dir_type not in ('', 'temp'):
1301 return ''
1302
1303 if warn:
1304 if not self.params.get('paths'):
1305 pass
1306 elif filename == '-':
1307 self.report_warning('--paths is ignored when an outputting to stdout', only_once=True)
1308 elif os.path.isabs(filename):
1309 self.report_warning('--paths is ignored since an absolute path is given in output template', only_once=True)
1310 if filename == '-' or not filename:
1311 return filename
1312
1313 return self.get_output_path(dir_type, filename)
1314
1315 def _match_entry(self, info_dict, incomplete=False, silent=False):
1316 """ Returns None if the file should be downloaded """
1317
1318 video_title = info_dict.get('title', info_dict.get('id', 'entry'))
1319
1320 def check_filter():
1321 if 'title' in info_dict:
1322 # This can happen when we're just evaluating the playlist
1323 title = info_dict['title']
1324 matchtitle = self.params.get('matchtitle', False)
1325 if matchtitle:
1326 if not re.search(matchtitle, title, re.IGNORECASE):
1327 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
1328 rejecttitle = self.params.get('rejecttitle', False)
1329 if rejecttitle:
1330 if re.search(rejecttitle, title, re.IGNORECASE):
1331 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
1332 date = info_dict.get('upload_date')
1333 if date is not None:
1334 dateRange = self.params.get('daterange', DateRange())
1335 if date not in dateRange:
1336 return f'{date_from_str(date).isoformat()} upload date is not in range {dateRange}'
1337 view_count = info_dict.get('view_count')
1338 if view_count is not None:
1339 min_views = self.params.get('min_views')
1340 if min_views is not None and view_count < min_views:
1341 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
1342 max_views = self.params.get('max_views')
1343 if max_views is not None and view_count > max_views:
1344 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
1345 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
1346 return 'Skipping "%s" because it is age restricted' % video_title
1347
1348 match_filter = self.params.get('match_filter')
1349 if match_filter is not None:
1350 try:
1351 ret = match_filter(info_dict, incomplete=incomplete)
1352 except TypeError:
1353 # For backward compatibility
1354 ret = None if incomplete else match_filter(info_dict)
1355 if ret is NO_DEFAULT:
1356 while True:
1357 filename = self._format_screen(self.prepare_filename(info_dict), self.Styles.FILENAME)
1358 reply = input(self._format_screen(
1359 f'Download "{filename}"? (Y/n): ', self.Styles.EMPHASIS)).lower().strip()
1360 if reply in {'y', ''}:
1361 return None
1362 elif reply == 'n':
1363 return f'Skipping {video_title}'
1364 elif ret is not None:
1365 return ret
1366 return None
1367
1368 if self.in_download_archive(info_dict):
1369 reason = '%s has already been recorded in the archive' % video_title
1370 break_opt, break_err = 'break_on_existing', ExistingVideoReached
1371 else:
1372 reason = check_filter()
1373 break_opt, break_err = 'break_on_reject', RejectedVideoReached
1374 if reason is not None:
1375 if not silent:
1376 self.to_screen('[download] ' + reason)
1377 if self.params.get(break_opt, False):
1378 raise break_err()
1379 return reason
1380
1381 @staticmethod
1382 def add_extra_info(info_dict, extra_info):
1383 '''Set the keys from extra_info in info dict if they are missing'''
1384 for key, value in extra_info.items():
1385 info_dict.setdefault(key, value)
1386
1387 def extract_info(self, url, download=True, ie_key=None, extra_info=None,
1388 process=True, force_generic_extractor=False):
1389 """
1390 Return a list with a dictionary for each video extracted.
1391
1392 Arguments:
1393 url -- URL to extract
1394
1395 Keyword arguments:
1396 download -- whether to download videos during extraction
1397 ie_key -- extractor key hint
1398 extra_info -- dictionary containing the extra values to add to each result
1399 process -- whether to resolve all unresolved references (URLs, playlist items),
1400 must be True for download to work.
1401 force_generic_extractor -- force using the generic extractor
1402 """
1403
1404 if extra_info is None:
1405 extra_info = {}
1406
1407 if not ie_key and force_generic_extractor:
1408 ie_key = 'Generic'
1409
1410 if ie_key:
1411 ies = {ie_key: self._get_info_extractor_class(ie_key)}
1412 else:
1413 ies = self._ies
1414
1415 for ie_key, ie in ies.items():
1416 if not ie.suitable(url):
1417 continue
1418
1419 if not ie.working():
1420 self.report_warning('The program functionality for this site has been marked as broken, '
1421 'and will probably not work.')
1422
1423 temp_id = ie.get_temp_id(url)
1424 if temp_id is not None and self.in_download_archive({'id': temp_id, 'ie_key': ie_key}):
1425 self.to_screen(f'[{ie_key}] {temp_id}: has already been recorded in the archive')
1426 if self.params.get('break_on_existing', False):
1427 raise ExistingVideoReached()
1428 break
1429 return self.__extract_info(url, self.get_info_extractor(ie_key), download, extra_info, process)
1430 else:
1431 self.report_error('no suitable InfoExtractor for URL %s' % url)
1432
1433 def _handle_extraction_exceptions(func):
1434 @functools.wraps(func)
1435 def wrapper(self, *args, **kwargs):
1436 while True:
1437 try:
1438 return func(self, *args, **kwargs)
1439 except (DownloadCancelled, LazyList.IndexError, PagedList.IndexError):
1440 raise
1441 except ReExtractInfo as e:
1442 if e.expected:
1443 self.to_screen(f'{e}; Re-extracting data')
1444 else:
1445 self.to_stderr('\r')
1446 self.report_warning(f'{e}; Re-extracting data')
1447 continue
1448 except GeoRestrictedError as e:
1449 msg = e.msg
1450 if e.countries:
1451 msg += '\nThis video is available in %s.' % ', '.join(
1452 map(ISO3166Utils.short2full, e.countries))
1453 msg += '\nYou might want to use a VPN or a proxy server (with --proxy) to workaround.'
1454 self.report_error(msg)
1455 except ExtractorError as e: # An error we somewhat expected
1456 self.report_error(str(e), e.format_traceback())
1457 except Exception as e:
1458 if self.params.get('ignoreerrors'):
1459 self.report_error(str(e), tb=encode_compat_str(traceback.format_exc()))
1460 else:
1461 raise
1462 break
1463 return wrapper
1464
1465 def _wait_for_video(self, ie_result={}):
1466 if (not self.params.get('wait_for_video')
1467 or ie_result.get('_type', 'video') != 'video'
1468 or ie_result.get('formats') or ie_result.get('url')):
1469 return
1470
1471 format_dur = lambda dur: '%02d:%02d:%02d' % timetuple_from_msec(dur * 1000)[:-1]
1472 last_msg = ''
1473
1474 def progress(msg):
1475 nonlocal last_msg
1476 full_msg = f'{msg}\n'
1477 if not self.params.get('noprogress'):
1478 full_msg = msg + ' ' * (len(last_msg) - len(msg)) + '\r'
1479 elif last_msg:
1480 return
1481 self.to_screen(full_msg, skip_eol=True)
1482 last_msg = msg
1483
1484 min_wait, max_wait = self.params.get('wait_for_video')
1485 diff = try_get(ie_result, lambda x: x['release_timestamp'] - time.time())
1486 if diff is None and ie_result.get('live_status') == 'is_upcoming':
1487 diff = round(random.uniform(min_wait, max_wait) if (max_wait and min_wait) else (max_wait or min_wait), 0)
1488 self.report_warning('Release time of video is not known')
1489 elif ie_result and (diff or 0) <= 0:
1490 self.report_warning('Video should already be available according to extracted info')
1491 diff = min(max(diff or 0, min_wait or 0), max_wait or float('inf'))
1492 self.to_screen(f'[wait] Waiting for {format_dur(diff)} - Press Ctrl+C to try now')
1493
1494 wait_till = time.time() + diff
1495 try:
1496 while True:
1497 diff = wait_till - time.time()
1498 if diff <= 0:
1499 progress('')
1500 raise ReExtractInfo('[wait] Wait period ended', expected=True)
1501 progress(f'[wait] Remaining time until next attempt: {self._format_screen(format_dur(diff), self.Styles.EMPHASIS)}')
1502 time.sleep(1)
1503 except KeyboardInterrupt:
1504 progress('')
1505 raise ReExtractInfo('[wait] Interrupted by user', expected=True)
1506 except BaseException as e:
1507 if not isinstance(e, ReExtractInfo):
1508 self.to_screen('')
1509 raise
1510
1511 @_handle_extraction_exceptions
1512 def __extract_info(self, url, ie, download, extra_info, process):
1513 try:
1514 ie_result = ie.extract(url)
1515 except UserNotLive as e:
1516 if process:
1517 if self.params.get('wait_for_video'):
1518 self.report_warning(e)
1519 self._wait_for_video()
1520 raise
1521 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
1522 self.report_warning(f'Extractor {ie.IE_NAME} returned nothing{bug_reports_message()}')
1523 return
1524 if isinstance(ie_result, list):
1525 # Backwards compatibility: old IE result format
1526 ie_result = {
1527 '_type': 'compat_list',
1528 'entries': ie_result,
1529 }
1530 if extra_info.get('original_url'):
1531 ie_result.setdefault('original_url', extra_info['original_url'])
1532 self.add_default_extra_info(ie_result, ie, url)
1533 if process:
1534 self._wait_for_video(ie_result)
1535 return self.process_ie_result(ie_result, download, extra_info)
1536 else:
1537 return ie_result
1538
1539 def add_default_extra_info(self, ie_result, ie, url):
1540 if url is not None:
1541 self.add_extra_info(ie_result, {
1542 'webpage_url': url,
1543 'original_url': url,
1544 })
1545 webpage_url = ie_result.get('webpage_url')
1546 if webpage_url:
1547 self.add_extra_info(ie_result, {
1548 'webpage_url_basename': url_basename(webpage_url),
1549 'webpage_url_domain': get_domain(webpage_url),
1550 })
1551 if ie is not None:
1552 self.add_extra_info(ie_result, {
1553 'extractor': ie.IE_NAME,
1554 'extractor_key': ie.ie_key(),
1555 })
1556
1557 def process_ie_result(self, ie_result, download=True, extra_info=None):
1558 """
1559 Take the result of the ie(may be modified) and resolve all unresolved
1560 references (URLs, playlist items).
1561
1562 It will also download the videos if 'download'.
1563 Returns the resolved ie_result.
1564 """
1565 if extra_info is None:
1566 extra_info = {}
1567 result_type = ie_result.get('_type', 'video')
1568
1569 if result_type in ('url', 'url_transparent'):
1570 ie_result['url'] = sanitize_url(
1571 ie_result['url'], scheme='http' if self.params.get('prefer_insecure') else 'https')
1572 if ie_result.get('original_url'):
1573 extra_info.setdefault('original_url', ie_result['original_url'])
1574
1575 extract_flat = self.params.get('extract_flat', False)
1576 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info)
1577 or extract_flat is True):
1578 info_copy = ie_result.copy()
1579 ie = try_get(ie_result.get('ie_key'), self.get_info_extractor)
1580 if ie and not ie_result.get('id'):
1581 info_copy['id'] = ie.get_temp_id(ie_result['url'])
1582 self.add_default_extra_info(info_copy, ie, ie_result['url'])
1583 self.add_extra_info(info_copy, extra_info)
1584 info_copy, _ = self.pre_process(info_copy)
1585 self.__forced_printings(info_copy, self.prepare_filename(info_copy), incomplete=True)
1586 self._raise_pending_errors(info_copy)
1587 if self.params.get('force_write_download_archive', False):
1588 self.record_download_archive(info_copy)
1589 return ie_result
1590
1591 if result_type == 'video':
1592 self.add_extra_info(ie_result, extra_info)
1593 ie_result = self.process_video_result(ie_result, download=download)
1594 self._raise_pending_errors(ie_result)
1595 additional_urls = (ie_result or {}).get('additional_urls')
1596 if additional_urls:
1597 # TODO: Improve MetadataParserPP to allow setting a list
1598 if isinstance(additional_urls, str):
1599 additional_urls = [additional_urls]
1600 self.to_screen(
1601 '[info] %s: %d additional URL(s) requested' % (ie_result['id'], len(additional_urls)))
1602 self.write_debug('Additional URLs: "%s"' % '", "'.join(additional_urls))
1603 ie_result['additional_entries'] = [
1604 self.extract_info(
1605 url, download, extra_info=extra_info,
1606 force_generic_extractor=self.params.get('force_generic_extractor'))
1607 for url in additional_urls
1608 ]
1609 return ie_result
1610 elif result_type == 'url':
1611 # We have to add extra_info to the results because it may be
1612 # contained in a playlist
1613 return self.extract_info(
1614 ie_result['url'], download,
1615 ie_key=ie_result.get('ie_key'),
1616 extra_info=extra_info)
1617 elif result_type == 'url_transparent':
1618 # Use the information from the embedding page
1619 info = self.extract_info(
1620 ie_result['url'], ie_key=ie_result.get('ie_key'),
1621 extra_info=extra_info, download=False, process=False)
1622
1623 # extract_info may return None when ignoreerrors is enabled and
1624 # extraction failed with an error, don't crash and return early
1625 # in this case
1626 if not info:
1627 return info
1628
1629 exempted_fields = {'_type', 'url', 'ie_key'}
1630 if not ie_result.get('section_end') and ie_result.get('section_start') is None:
1631 # For video clips, the id etc of the clip extractor should be used
1632 exempted_fields |= {'id', 'extractor', 'extractor_key'}
1633
1634 new_result = info.copy()
1635 new_result.update(filter_dict(ie_result, lambda k, v: v is not None and k not in exempted_fields))
1636
1637 # Extracted info may not be a video result (i.e.
1638 # info.get('_type', 'video') != video) but rather an url or
1639 # url_transparent. In such cases outer metadata (from ie_result)
1640 # should be propagated to inner one (info). For this to happen
1641 # _type of info should be overridden with url_transparent. This
1642 # fixes issue from https://github.com/ytdl-org/youtube-dl/pull/11163.
1643 if new_result.get('_type') == 'url':
1644 new_result['_type'] = 'url_transparent'
1645
1646 return self.process_ie_result(
1647 new_result, download=download, extra_info=extra_info)
1648 elif result_type in ('playlist', 'multi_video'):
1649 # Protect from infinite recursion due to recursively nested playlists
1650 # (see https://github.com/ytdl-org/youtube-dl/issues/27833)
1651 webpage_url = ie_result['webpage_url']
1652 if webpage_url in self._playlist_urls:
1653 self.to_screen(
1654 '[download] Skipping already downloaded playlist: %s'
1655 % ie_result.get('title') or ie_result.get('id'))
1656 return
1657
1658 self._playlist_level += 1
1659 self._playlist_urls.add(webpage_url)
1660 self._fill_common_fields(ie_result, False)
1661 self._sanitize_thumbnails(ie_result)
1662 try:
1663 return self.__process_playlist(ie_result, download)
1664 finally:
1665 self._playlist_level -= 1
1666 if not self._playlist_level:
1667 self._playlist_urls.clear()
1668 elif result_type == 'compat_list':
1669 self.report_warning(
1670 'Extractor %s returned a compat_list result. '
1671 'It needs to be updated.' % ie_result.get('extractor'))
1672
1673 def _fixup(r):
1674 self.add_extra_info(r, {
1675 'extractor': ie_result['extractor'],
1676 'webpage_url': ie_result['webpage_url'],
1677 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1678 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1679 'extractor_key': ie_result['extractor_key'],
1680 })
1681 return r
1682 ie_result['entries'] = [
1683 self.process_ie_result(_fixup(r), download, extra_info)
1684 for r in ie_result['entries']
1685 ]
1686 return ie_result
1687 else:
1688 raise Exception('Invalid result type: %s' % result_type)
1689
1690 def _ensure_dir_exists(self, path):
1691 return make_dir(path, self.report_error)
1692
1693 @staticmethod
1694 def _playlist_infodict(ie_result, strict=False, **kwargs):
1695 info = {
1696 'playlist_count': ie_result.get('playlist_count'),
1697 'playlist': ie_result.get('title') or ie_result.get('id'),
1698 'playlist_id': ie_result.get('id'),
1699 'playlist_title': ie_result.get('title'),
1700 'playlist_uploader': ie_result.get('uploader'),
1701 'playlist_uploader_id': ie_result.get('uploader_id'),
1702 **kwargs,
1703 }
1704 if strict:
1705 return info
1706 return {
1707 **info,
1708 'playlist_index': 0,
1709 '__last_playlist_index': max(ie_result['requested_entries'] or (0, 0)),
1710 'extractor': ie_result['extractor'],
1711 'webpage_url': ie_result['webpage_url'],
1712 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1713 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1714 'extractor_key': ie_result['extractor_key'],
1715 }
1716
1717 def __process_playlist(self, ie_result, download):
1718 """Process each entry in the playlist"""
1719 assert ie_result['_type'] in ('playlist', 'multi_video')
1720
1721 common_info = self._playlist_infodict(ie_result, strict=True)
1722 title = common_info.get('playlist') or '<Untitled>'
1723 if self._match_entry(common_info, incomplete=True) is not None:
1724 return
1725 self.to_screen(f'[download] Downloading {ie_result["_type"]}: {title}')
1726
1727 all_entries = PlaylistEntries(self, ie_result)
1728 entries = orderedSet(all_entries.get_requested_items(), lazy=True)
1729
1730 lazy = self.params.get('lazy_playlist')
1731 if lazy:
1732 resolved_entries, n_entries = [], 'N/A'
1733 ie_result['requested_entries'], ie_result['entries'] = None, None
1734 else:
1735 entries = resolved_entries = list(entries)
1736 n_entries = len(resolved_entries)
1737 ie_result['requested_entries'], ie_result['entries'] = tuple(zip(*resolved_entries)) or ([], [])
1738 if not ie_result.get('playlist_count'):
1739 # Better to do this after potentially exhausting entries
1740 ie_result['playlist_count'] = all_entries.get_full_count()
1741
1742 extra = self._playlist_infodict(ie_result, n_entries=int_or_none(n_entries))
1743 ie_copy = collections.ChainMap(ie_result, extra)
1744
1745 _infojson_written = False
1746 write_playlist_files = self.params.get('allow_playlist_files', True)
1747 if write_playlist_files and self.params.get('list_thumbnails'):
1748 self.list_thumbnails(ie_result)
1749 if write_playlist_files and not self.params.get('simulate'):
1750 _infojson_written = self._write_info_json(
1751 'playlist', ie_result, self.prepare_filename(ie_copy, 'pl_infojson'))
1752 if _infojson_written is None:
1753 return
1754 if self._write_description('playlist', ie_result,
1755 self.prepare_filename(ie_copy, 'pl_description')) is None:
1756 return
1757 # TODO: This should be passed to ThumbnailsConvertor if necessary
1758 self._write_thumbnails('playlist', ie_result, self.prepare_filename(ie_copy, 'pl_thumbnail'))
1759
1760 if lazy:
1761 if self.params.get('playlistreverse') or self.params.get('playlistrandom'):
1762 self.report_warning('playlistreverse and playlistrandom are not supported with lazy_playlist', only_once=True)
1763 elif self.params.get('playlistreverse'):
1764 entries.reverse()
1765 elif self.params.get('playlistrandom'):
1766 random.shuffle(entries)
1767
1768 self.to_screen(f'[{ie_result["extractor"]}] Playlist {title}: Downloading {n_entries} videos'
1769 f'{format_field(ie_result, "playlist_count", " of %s")}')
1770
1771 keep_resolved_entries = self.params.get('extract_flat') != 'discard'
1772 if self.params.get('extract_flat') == 'discard_in_playlist':
1773 keep_resolved_entries = ie_result['_type'] != 'playlist'
1774 if keep_resolved_entries:
1775 self.write_debug('The information of all playlist entries will be held in memory')
1776
1777 failures = 0
1778 max_failures = self.params.get('skip_playlist_after_errors') or float('inf')
1779 for i, (playlist_index, entry) in enumerate(entries):
1780 if lazy:
1781 resolved_entries.append((playlist_index, entry))
1782 if not entry:
1783 continue
1784
1785 entry['__x_forwarded_for_ip'] = ie_result.get('__x_forwarded_for_ip')
1786 if not lazy and 'playlist-index' in self.params.get('compat_opts', []):
1787 playlist_index = ie_result['requested_entries'][i]
1788
1789 entry_copy = collections.ChainMap(entry, {
1790 **common_info,
1791 'n_entries': int_or_none(n_entries),
1792 'playlist_index': playlist_index,
1793 'playlist_autonumber': i + 1,
1794 })
1795
1796 if self._match_entry(entry_copy, incomplete=True) is not None:
1797 continue
1798
1799 self.to_screen('[download] Downloading video %s of %s' % (
1800 self._format_screen(i + 1, self.Styles.ID), self._format_screen(n_entries, self.Styles.EMPHASIS)))
1801
1802 extra.update({
1803 'playlist_index': playlist_index,
1804 'playlist_autonumber': i + 1,
1805 })
1806 entry_result = self.__process_iterable_entry(entry, download, extra)
1807 if not entry_result:
1808 failures += 1
1809 if failures >= max_failures:
1810 self.report_error(
1811 f'Skipping the remaining entries in playlist "{title}" since {failures} items failed extraction')
1812 break
1813 if keep_resolved_entries:
1814 resolved_entries[i] = (playlist_index, entry_result)
1815
1816 # Update with processed data
1817 ie_result['requested_entries'], ie_result['entries'] = tuple(zip(*resolved_entries)) or ([], [])
1818
1819 # Write the updated info to json
1820 if _infojson_written is True and self._write_info_json(
1821 'updated playlist', ie_result,
1822 self.prepare_filename(ie_copy, 'pl_infojson'), overwrite=True) is None:
1823 return
1824
1825 ie_result = self.run_all_pps('playlist', ie_result)
1826 self.to_screen(f'[download] Finished downloading playlist: {title}')
1827 return ie_result
1828
1829 @_handle_extraction_exceptions
1830 def __process_iterable_entry(self, entry, download, extra_info):
1831 return self.process_ie_result(
1832 entry, download=download, extra_info=extra_info)
1833
1834 def _build_format_filter(self, filter_spec):
1835 " Returns a function to filter the formats according to the filter_spec "
1836
1837 OPERATORS = {
1838 '<': operator.lt,
1839 '<=': operator.le,
1840 '>': operator.gt,
1841 '>=': operator.ge,
1842 '=': operator.eq,
1843 '!=': operator.ne,
1844 }
1845 operator_rex = re.compile(r'''(?x)\s*
1846 (?P<key>width|height|tbr|abr|vbr|asr|filesize|filesize_approx|fps)\s*
1847 (?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
1848 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)\s*
1849 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
1850 m = operator_rex.fullmatch(filter_spec)
1851 if m:
1852 try:
1853 comparison_value = int(m.group('value'))
1854 except ValueError:
1855 comparison_value = parse_filesize(m.group('value'))
1856 if comparison_value is None:
1857 comparison_value = parse_filesize(m.group('value') + 'B')
1858 if comparison_value is None:
1859 raise ValueError(
1860 'Invalid value %r in format specification %r' % (
1861 m.group('value'), filter_spec))
1862 op = OPERATORS[m.group('op')]
1863
1864 if not m:
1865 STR_OPERATORS = {
1866 '=': operator.eq,
1867 '^=': lambda attr, value: attr.startswith(value),
1868 '$=': lambda attr, value: attr.endswith(value),
1869 '*=': lambda attr, value: value in attr,
1870 '~=': lambda attr, value: value.search(attr) is not None
1871 }
1872 str_operator_rex = re.compile(r'''(?x)\s*
1873 (?P<key>[a-zA-Z0-9._-]+)\s*
1874 (?P<negation>!\s*)?(?P<op>%s)\s*(?P<none_inclusive>\?\s*)?
1875 (?P<quote>["'])?
1876 (?P<value>(?(quote)(?:(?!(?P=quote))[^\\]|\\.)+|[\w.-]+))
1877 (?(quote)(?P=quote))\s*
1878 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
1879 m = str_operator_rex.fullmatch(filter_spec)
1880 if m:
1881 if m.group('op') == '~=':
1882 comparison_value = re.compile(m.group('value'))
1883 else:
1884 comparison_value = re.sub(r'''\\([\\"'])''', r'\1', m.group('value'))
1885 str_op = STR_OPERATORS[m.group('op')]
1886 if m.group('negation'):
1887 op = lambda attr, value: not str_op(attr, value)
1888 else:
1889 op = str_op
1890
1891 if not m:
1892 raise SyntaxError('Invalid filter specification %r' % filter_spec)
1893
1894 def _filter(f):
1895 actual_value = f.get(m.group('key'))
1896 if actual_value is None:
1897 return m.group('none_inclusive')
1898 return op(actual_value, comparison_value)
1899 return _filter
1900
1901 def _check_formats(self, formats):
1902 for f in formats:
1903 self.to_screen('[info] Testing format %s' % f['format_id'])
1904 path = self.get_output_path('temp')
1905 if not self._ensure_dir_exists(f'{path}/'):
1906 continue
1907 temp_file = tempfile.NamedTemporaryFile(suffix='.tmp', delete=False, dir=path or None)
1908 temp_file.close()
1909 try:
1910 success, _ = self.dl(temp_file.name, f, test=True)
1911 except (DownloadError, OSError, ValueError) + network_exceptions:
1912 success = False
1913 finally:
1914 if os.path.exists(temp_file.name):
1915 try:
1916 os.remove(temp_file.name)
1917 except OSError:
1918 self.report_warning('Unable to delete temporary file "%s"' % temp_file.name)
1919 if success:
1920 yield f
1921 else:
1922 self.to_screen('[info] Unable to download format %s. Skipping...' % f['format_id'])
1923
1924 def _default_format_spec(self, info_dict, download=True):
1925
1926 def can_merge():
1927 merger = FFmpegMergerPP(self)
1928 return merger.available and merger.can_merge()
1929
1930 prefer_best = (
1931 not self.params.get('simulate')
1932 and download
1933 and (
1934 not can_merge()
1935 or info_dict.get('is_live') and not self.params.get('live_from_start')
1936 or self.params['outtmpl']['default'] == '-'))
1937 compat = (
1938 prefer_best
1939 or self.params.get('allow_multiple_audio_streams', False)
1940 or 'format-spec' in self.params['compat_opts'])
1941
1942 return (
1943 'best/bestvideo+bestaudio' if prefer_best
1944 else 'bestvideo*+bestaudio/best' if not compat
1945 else 'bestvideo+bestaudio/best')
1946
1947 def build_format_selector(self, format_spec):
1948 def syntax_error(note, start):
1949 message = (
1950 'Invalid format specification: '
1951 '{}\n\t{}\n\t{}^'.format(note, format_spec, ' ' * start[1]))
1952 return SyntaxError(message)
1953
1954 PICKFIRST = 'PICKFIRST'
1955 MERGE = 'MERGE'
1956 SINGLE = 'SINGLE'
1957 GROUP = 'GROUP'
1958 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
1959
1960 allow_multiple_streams = {'audio': self.params.get('allow_multiple_audio_streams', False),
1961 'video': self.params.get('allow_multiple_video_streams', False)}
1962
1963 check_formats = self.params.get('check_formats') == 'selected'
1964
1965 def _parse_filter(tokens):
1966 filter_parts = []
1967 for type, string, start, _, _ in tokens:
1968 if type == tokenize.OP and string == ']':
1969 return ''.join(filter_parts)
1970 else:
1971 filter_parts.append(string)
1972
1973 def _remove_unused_ops(tokens):
1974 # Remove operators that we don't use and join them with the surrounding strings
1975 # for example: 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
1976 ALLOWED_OPS = ('/', '+', ',', '(', ')')
1977 last_string, last_start, last_end, last_line = None, None, None, None
1978 for type, string, start, end, line in tokens:
1979 if type == tokenize.OP and string == '[':
1980 if last_string:
1981 yield tokenize.NAME, last_string, last_start, last_end, last_line
1982 last_string = None
1983 yield type, string, start, end, line
1984 # everything inside brackets will be handled by _parse_filter
1985 for type, string, start, end, line in tokens:
1986 yield type, string, start, end, line
1987 if type == tokenize.OP and string == ']':
1988 break
1989 elif type == tokenize.OP and string in ALLOWED_OPS:
1990 if last_string:
1991 yield tokenize.NAME, last_string, last_start, last_end, last_line
1992 last_string = None
1993 yield type, string, start, end, line
1994 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
1995 if not last_string:
1996 last_string = string
1997 last_start = start
1998 last_end = end
1999 else:
2000 last_string += string
2001 if last_string:
2002 yield tokenize.NAME, last_string, last_start, last_end, last_line
2003
2004 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
2005 selectors = []
2006 current_selector = None
2007 for type, string, start, _, _ in tokens:
2008 # ENCODING is only defined in python 3.x
2009 if type == getattr(tokenize, 'ENCODING', None):
2010 continue
2011 elif type in [tokenize.NAME, tokenize.NUMBER]:
2012 current_selector = FormatSelector(SINGLE, string, [])
2013 elif type == tokenize.OP:
2014 if string == ')':
2015 if not inside_group:
2016 # ')' will be handled by the parentheses group
2017 tokens.restore_last_token()
2018 break
2019 elif inside_merge and string in ['/', ',']:
2020 tokens.restore_last_token()
2021 break
2022 elif inside_choice and string == ',':
2023 tokens.restore_last_token()
2024 break
2025 elif string == ',':
2026 if not current_selector:
2027 raise syntax_error('"," must follow a format selector', start)
2028 selectors.append(current_selector)
2029 current_selector = None
2030 elif string == '/':
2031 if not current_selector:
2032 raise syntax_error('"/" must follow a format selector', start)
2033 first_choice = current_selector
2034 second_choice = _parse_format_selection(tokens, inside_choice=True)
2035 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
2036 elif string == '[':
2037 if not current_selector:
2038 current_selector = FormatSelector(SINGLE, 'best', [])
2039 format_filter = _parse_filter(tokens)
2040 current_selector.filters.append(format_filter)
2041 elif string == '(':
2042 if current_selector:
2043 raise syntax_error('Unexpected "("', start)
2044 group = _parse_format_selection(tokens, inside_group=True)
2045 current_selector = FormatSelector(GROUP, group, [])
2046 elif string == '+':
2047 if not current_selector:
2048 raise syntax_error('Unexpected "+"', start)
2049 selector_1 = current_selector
2050 selector_2 = _parse_format_selection(tokens, inside_merge=True)
2051 if not selector_2:
2052 raise syntax_error('Expected a selector', start)
2053 current_selector = FormatSelector(MERGE, (selector_1, selector_2), [])
2054 else:
2055 raise syntax_error(f'Operator not recognized: "{string}"', start)
2056 elif type == tokenize.ENDMARKER:
2057 break
2058 if current_selector:
2059 selectors.append(current_selector)
2060 return selectors
2061
2062 def _merge(formats_pair):
2063 format_1, format_2 = formats_pair
2064
2065 formats_info = []
2066 formats_info.extend(format_1.get('requested_formats', (format_1,)))
2067 formats_info.extend(format_2.get('requested_formats', (format_2,)))
2068
2069 if not allow_multiple_streams['video'] or not allow_multiple_streams['audio']:
2070 get_no_more = {'video': False, 'audio': False}
2071 for (i, fmt_info) in enumerate(formats_info):
2072 if fmt_info.get('acodec') == fmt_info.get('vcodec') == 'none':
2073 formats_info.pop(i)
2074 continue
2075 for aud_vid in ['audio', 'video']:
2076 if not allow_multiple_streams[aud_vid] and fmt_info.get(aud_vid[0] + 'codec') != 'none':
2077 if get_no_more[aud_vid]:
2078 formats_info.pop(i)
2079 break
2080 get_no_more[aud_vid] = True
2081
2082 if len(formats_info) == 1:
2083 return formats_info[0]
2084
2085 video_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('vcodec') != 'none']
2086 audio_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('acodec') != 'none']
2087
2088 the_only_video = video_fmts[0] if len(video_fmts) == 1 else None
2089 the_only_audio = audio_fmts[0] if len(audio_fmts) == 1 else None
2090
2091 output_ext = self.params.get('merge_output_format')
2092 if not output_ext:
2093 if the_only_video:
2094 output_ext = the_only_video['ext']
2095 elif the_only_audio and not video_fmts:
2096 output_ext = the_only_audio['ext']
2097 else:
2098 output_ext = 'mkv'
2099
2100 filtered = lambda *keys: filter(None, (traverse_obj(fmt, *keys) for fmt in formats_info))
2101
2102 new_dict = {
2103 'requested_formats': formats_info,
2104 'format': '+'.join(filtered('format')),
2105 'format_id': '+'.join(filtered('format_id')),
2106 'ext': output_ext,
2107 'protocol': '+'.join(map(determine_protocol, formats_info)),
2108 'language': '+'.join(orderedSet(filtered('language'))) or None,
2109 'format_note': '+'.join(orderedSet(filtered('format_note'))) or None,
2110 'filesize_approx': sum(filtered('filesize', 'filesize_approx')) or None,
2111 'tbr': sum(filtered('tbr', 'vbr', 'abr')),
2112 }
2113
2114 if the_only_video:
2115 new_dict.update({
2116 'width': the_only_video.get('width'),
2117 'height': the_only_video.get('height'),
2118 'resolution': the_only_video.get('resolution') or self.format_resolution(the_only_video),
2119 'fps': the_only_video.get('fps'),
2120 'dynamic_range': the_only_video.get('dynamic_range'),
2121 'vcodec': the_only_video.get('vcodec'),
2122 'vbr': the_only_video.get('vbr'),
2123 'stretched_ratio': the_only_video.get('stretched_ratio'),
2124 })
2125
2126 if the_only_audio:
2127 new_dict.update({
2128 'acodec': the_only_audio.get('acodec'),
2129 'abr': the_only_audio.get('abr'),
2130 'asr': the_only_audio.get('asr'),
2131 })
2132
2133 return new_dict
2134
2135 def _check_formats(formats):
2136 if not check_formats:
2137 yield from formats
2138 return
2139 yield from self._check_formats(formats)
2140
2141 def _build_selector_function(selector):
2142 if isinstance(selector, list): # ,
2143 fs = [_build_selector_function(s) for s in selector]
2144
2145 def selector_function(ctx):
2146 for f in fs:
2147 yield from f(ctx)
2148 return selector_function
2149
2150 elif selector.type == GROUP: # ()
2151 selector_function = _build_selector_function(selector.selector)
2152
2153 elif selector.type == PICKFIRST: # /
2154 fs = [_build_selector_function(s) for s in selector.selector]
2155
2156 def selector_function(ctx):
2157 for f in fs:
2158 picked_formats = list(f(ctx))
2159 if picked_formats:
2160 return picked_formats
2161 return []
2162
2163 elif selector.type == MERGE: # +
2164 selector_1, selector_2 = map(_build_selector_function, selector.selector)
2165
2166 def selector_function(ctx):
2167 for pair in itertools.product(selector_1(ctx), selector_2(ctx)):
2168 yield _merge(pair)
2169
2170 elif selector.type == SINGLE: # atom
2171 format_spec = selector.selector or 'best'
2172
2173 # TODO: Add allvideo, allaudio etc by generalizing the code with best/worst selector
2174 if format_spec == 'all':
2175 def selector_function(ctx):
2176 yield from _check_formats(ctx['formats'][::-1])
2177 elif format_spec == 'mergeall':
2178 def selector_function(ctx):
2179 formats = list(_check_formats(
2180 f for f in ctx['formats'] if f.get('vcodec') != 'none' or f.get('acodec') != 'none'))
2181 if not formats:
2182 return
2183 merged_format = formats[-1]
2184 for f in formats[-2::-1]:
2185 merged_format = _merge((merged_format, f))
2186 yield merged_format
2187
2188 else:
2189 format_fallback, seperate_fallback, format_reverse, format_idx = False, None, True, 1
2190 mobj = re.match(
2191 r'(?P<bw>best|worst|b|w)(?P<type>video|audio|v|a)?(?P<mod>\*)?(?:\.(?P<n>[1-9]\d*))?$',
2192 format_spec)
2193 if mobj is not None:
2194 format_idx = int_or_none(mobj.group('n'), default=1)
2195 format_reverse = mobj.group('bw')[0] == 'b'
2196 format_type = (mobj.group('type') or [None])[0]
2197 not_format_type = {'v': 'a', 'a': 'v'}.get(format_type)
2198 format_modified = mobj.group('mod') is not None
2199
2200 format_fallback = not format_type and not format_modified # for b, w
2201 _filter_f = (
2202 (lambda f: f.get('%scodec' % format_type) != 'none')
2203 if format_type and format_modified # bv*, ba*, wv*, wa*
2204 else (lambda f: f.get('%scodec' % not_format_type) == 'none')
2205 if format_type # bv, ba, wv, wa
2206 else (lambda f: f.get('vcodec') != 'none' and f.get('acodec') != 'none')
2207 if not format_modified # b, w
2208 else lambda f: True) # b*, w*
2209 filter_f = lambda f: _filter_f(f) and (
2210 f.get('vcodec') != 'none' or f.get('acodec') != 'none')
2211 else:
2212 if format_spec in self._format_selection_exts['audio']:
2213 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none'
2214 elif format_spec in self._format_selection_exts['video']:
2215 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none' and f.get('vcodec') != 'none'
2216 seperate_fallback = lambda f: f.get('ext') == format_spec and f.get('vcodec') != 'none'
2217 elif format_spec in self._format_selection_exts['storyboards']:
2218 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') == 'none' and f.get('vcodec') == 'none'
2219 else:
2220 filter_f = lambda f: f.get('format_id') == format_spec # id
2221
2222 def selector_function(ctx):
2223 formats = list(ctx['formats'])
2224 matches = list(filter(filter_f, formats)) if filter_f is not None else formats
2225 if not matches:
2226 if format_fallback and ctx['incomplete_formats']:
2227 # for extractors with incomplete formats (audio only (soundcloud)
2228 # or video only (imgur)) best/worst will fallback to
2229 # best/worst {video,audio}-only format
2230 matches = formats
2231 elif seperate_fallback and not ctx['has_merged_format']:
2232 # for compatibility with youtube-dl when there is no pre-merged format
2233 matches = list(filter(seperate_fallback, formats))
2234 matches = LazyList(_check_formats(matches[::-1 if format_reverse else 1]))
2235 try:
2236 yield matches[format_idx - 1]
2237 except LazyList.IndexError:
2238 return
2239
2240 filters = [self._build_format_filter(f) for f in selector.filters]
2241
2242 def final_selector(ctx):
2243 ctx_copy = dict(ctx)
2244 for _filter in filters:
2245 ctx_copy['formats'] = list(filter(_filter, ctx_copy['formats']))
2246 return selector_function(ctx_copy)
2247 return final_selector
2248
2249 stream = io.BytesIO(format_spec.encode())
2250 try:
2251 tokens = list(_remove_unused_ops(tokenize.tokenize(stream.readline)))
2252 except tokenize.TokenError:
2253 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
2254
2255 class TokenIterator:
2256 def __init__(self, tokens):
2257 self.tokens = tokens
2258 self.counter = 0
2259
2260 def __iter__(self):
2261 return self
2262
2263 def __next__(self):
2264 if self.counter >= len(self.tokens):
2265 raise StopIteration()
2266 value = self.tokens[self.counter]
2267 self.counter += 1
2268 return value
2269
2270 next = __next__
2271
2272 def restore_last_token(self):
2273 self.counter -= 1
2274
2275 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
2276 return _build_selector_function(parsed_selector)
2277
2278 def _calc_headers(self, info_dict):
2279 res = merge_headers(self.params['http_headers'], info_dict.get('http_headers') or {})
2280
2281 cookies = self._calc_cookies(info_dict['url'])
2282 if cookies:
2283 res['Cookie'] = cookies
2284
2285 if 'X-Forwarded-For' not in res:
2286 x_forwarded_for_ip = info_dict.get('__x_forwarded_for_ip')
2287 if x_forwarded_for_ip:
2288 res['X-Forwarded-For'] = x_forwarded_for_ip
2289
2290 return res
2291
2292 def _calc_cookies(self, url):
2293 pr = sanitized_Request(url)
2294 self.cookiejar.add_cookie_header(pr)
2295 return pr.get_header('Cookie')
2296
2297 def _sort_thumbnails(self, thumbnails):
2298 thumbnails.sort(key=lambda t: (
2299 t.get('preference') if t.get('preference') is not None else -1,
2300 t.get('width') if t.get('width') is not None else -1,
2301 t.get('height') if t.get('height') is not None else -1,
2302 t.get('id') if t.get('id') is not None else '',
2303 t.get('url')))
2304
2305 def _sanitize_thumbnails(self, info_dict):
2306 thumbnails = info_dict.get('thumbnails')
2307 if thumbnails is None:
2308 thumbnail = info_dict.get('thumbnail')
2309 if thumbnail:
2310 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
2311 if not thumbnails:
2312 return
2313
2314 def check_thumbnails(thumbnails):
2315 for t in thumbnails:
2316 self.to_screen(f'[info] Testing thumbnail {t["id"]}')
2317 try:
2318 self.urlopen(HEADRequest(t['url']))
2319 except network_exceptions as err:
2320 self.to_screen(f'[info] Unable to connect to thumbnail {t["id"]} URL {t["url"]!r} - {err}. Skipping...')
2321 continue
2322 yield t
2323
2324 self._sort_thumbnails(thumbnails)
2325 for i, t in enumerate(thumbnails):
2326 if t.get('id') is None:
2327 t['id'] = '%d' % i
2328 if t.get('width') and t.get('height'):
2329 t['resolution'] = '%dx%d' % (t['width'], t['height'])
2330 t['url'] = sanitize_url(t['url'])
2331
2332 if self.params.get('check_formats') is True:
2333 info_dict['thumbnails'] = LazyList(check_thumbnails(thumbnails[::-1]), reverse=True)
2334 else:
2335 info_dict['thumbnails'] = thumbnails
2336
2337 def _fill_common_fields(self, info_dict, is_video=True):
2338 # TODO: move sanitization here
2339 if is_video:
2340 # playlists are allowed to lack "title"
2341 title = info_dict.get('title', NO_DEFAULT)
2342 if title is NO_DEFAULT:
2343 raise ExtractorError('Missing "title" field in extractor result',
2344 video_id=info_dict['id'], ie=info_dict['extractor'])
2345 info_dict['fulltitle'] = title
2346 if not title:
2347 if title == '':
2348 self.write_debug('Extractor gave empty title. Creating a generic title')
2349 else:
2350 self.report_warning('Extractor failed to obtain "title". Creating a generic title instead')
2351 info_dict['title'] = f'{info_dict["extractor"].replace(":", "-")} video #{info_dict["id"]}'
2352
2353 if info_dict.get('duration') is not None:
2354 info_dict['duration_string'] = formatSeconds(info_dict['duration'])
2355
2356 for ts_key, date_key in (
2357 ('timestamp', 'upload_date'),
2358 ('release_timestamp', 'release_date'),
2359 ('modified_timestamp', 'modified_date'),
2360 ):
2361 if info_dict.get(date_key) is None and info_dict.get(ts_key) is not None:
2362 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
2363 # see http://bugs.python.org/issue1646728)
2364 with contextlib.suppress(ValueError, OverflowError, OSError):
2365 upload_date = datetime.datetime.utcfromtimestamp(info_dict[ts_key])
2366 info_dict[date_key] = upload_date.strftime('%Y%m%d')
2367
2368 live_keys = ('is_live', 'was_live')
2369 live_status = info_dict.get('live_status')
2370 if live_status is None:
2371 for key in live_keys:
2372 if info_dict.get(key) is False:
2373 continue
2374 if info_dict.get(key):
2375 live_status = key
2376 break
2377 if all(info_dict.get(key) is False for key in live_keys):
2378 live_status = 'not_live'
2379 if live_status:
2380 info_dict['live_status'] = live_status
2381 for key in live_keys:
2382 if info_dict.get(key) is None:
2383 info_dict[key] = (live_status == key)
2384
2385 # Auto generate title fields corresponding to the *_number fields when missing
2386 # in order to always have clean titles. This is very common for TV series.
2387 for field in ('chapter', 'season', 'episode'):
2388 if info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
2389 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
2390
2391 def _raise_pending_errors(self, info):
2392 err = info.pop('__pending_error', None)
2393 if err:
2394 self.report_error(err, tb=False)
2395
2396 def process_video_result(self, info_dict, download=True):
2397 assert info_dict.get('_type', 'video') == 'video'
2398 self._num_videos += 1
2399
2400 if 'id' not in info_dict:
2401 raise ExtractorError('Missing "id" field in extractor result', ie=info_dict['extractor'])
2402 elif not info_dict.get('id'):
2403 raise ExtractorError('Extractor failed to obtain "id"', ie=info_dict['extractor'])
2404
2405 def report_force_conversion(field, field_not, conversion):
2406 self.report_warning(
2407 '"%s" field is not %s - forcing %s conversion, there is an error in extractor'
2408 % (field, field_not, conversion))
2409
2410 def sanitize_string_field(info, string_field):
2411 field = info.get(string_field)
2412 if field is None or isinstance(field, str):
2413 return
2414 report_force_conversion(string_field, 'a string', 'string')
2415 info[string_field] = str(field)
2416
2417 def sanitize_numeric_fields(info):
2418 for numeric_field in self._NUMERIC_FIELDS:
2419 field = info.get(numeric_field)
2420 if field is None or isinstance(field, (int, float)):
2421 continue
2422 report_force_conversion(numeric_field, 'numeric', 'int')
2423 info[numeric_field] = int_or_none(field)
2424
2425 sanitize_string_field(info_dict, 'id')
2426 sanitize_numeric_fields(info_dict)
2427 if info_dict.get('section_end') and info_dict.get('section_start') is not None:
2428 info_dict['duration'] = round(info_dict['section_end'] - info_dict['section_start'], 3)
2429 if (info_dict.get('duration') or 0) <= 0 and info_dict.pop('duration', None):
2430 self.report_warning('"duration" field is negative, there is an error in extractor')
2431
2432 chapters = info_dict.get('chapters') or []
2433 if chapters and chapters[0].get('start_time'):
2434 chapters.insert(0, {'start_time': 0})
2435
2436 dummy_chapter = {'end_time': 0, 'start_time': info_dict.get('duration')}
2437 for idx, (prev, current, next_) in enumerate(zip(
2438 (dummy_chapter, *chapters), chapters, (*chapters[1:], dummy_chapter)), 1):
2439 if current.get('start_time') is None:
2440 current['start_time'] = prev.get('end_time')
2441 if not current.get('end_time'):
2442 current['end_time'] = next_.get('start_time')
2443 if not current.get('title'):
2444 current['title'] = f'<Untitled Chapter {idx}>'
2445
2446 if 'playlist' not in info_dict:
2447 # It isn't part of a playlist
2448 info_dict['playlist'] = None
2449 info_dict['playlist_index'] = None
2450
2451 self._sanitize_thumbnails(info_dict)
2452
2453 thumbnail = info_dict.get('thumbnail')
2454 thumbnails = info_dict.get('thumbnails')
2455 if thumbnail:
2456 info_dict['thumbnail'] = sanitize_url(thumbnail)
2457 elif thumbnails:
2458 info_dict['thumbnail'] = thumbnails[-1]['url']
2459
2460 if info_dict.get('display_id') is None and 'id' in info_dict:
2461 info_dict['display_id'] = info_dict['id']
2462
2463 self._fill_common_fields(info_dict)
2464
2465 for cc_kind in ('subtitles', 'automatic_captions'):
2466 cc = info_dict.get(cc_kind)
2467 if cc:
2468 for _, subtitle in cc.items():
2469 for subtitle_format in subtitle:
2470 if subtitle_format.get('url'):
2471 subtitle_format['url'] = sanitize_url(subtitle_format['url'])
2472 if subtitle_format.get('ext') is None:
2473 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
2474
2475 automatic_captions = info_dict.get('automatic_captions')
2476 subtitles = info_dict.get('subtitles')
2477
2478 info_dict['requested_subtitles'] = self.process_subtitles(
2479 info_dict['id'], subtitles, automatic_captions)
2480
2481 if info_dict.get('formats') is None:
2482 # There's only one format available
2483 formats = [info_dict]
2484 else:
2485 formats = info_dict['formats']
2486
2487 # or None ensures --clean-infojson removes it
2488 info_dict['_has_drm'] = any(f.get('has_drm') for f in formats) or None
2489 if not self.params.get('allow_unplayable_formats'):
2490 formats = [f for f in formats if not f.get('has_drm')]
2491 if info_dict['_has_drm'] and formats and all(
2492 f.get('acodec') == f.get('vcodec') == 'none' for f in formats):
2493 self.report_warning(
2494 'This video is DRM protected and only images are available for download. '
2495 'Use --list-formats to see them')
2496
2497 get_from_start = not info_dict.get('is_live') or bool(self.params.get('live_from_start'))
2498 if not get_from_start:
2499 info_dict['title'] += ' ' + datetime.datetime.now().strftime('%Y-%m-%d %H:%M')
2500 if info_dict.get('is_live') and formats:
2501 formats = [f for f in formats if bool(f.get('is_from_start')) == get_from_start]
2502 if get_from_start and not formats:
2503 self.raise_no_formats(info_dict, msg=(
2504 '--live-from-start is passed, but there are no formats that can be downloaded from the start. '
2505 'If you want to download from the current time, use --no-live-from-start'))
2506
2507 if not formats:
2508 self.raise_no_formats(info_dict)
2509
2510 def is_wellformed(f):
2511 url = f.get('url')
2512 if not url:
2513 self.report_warning(
2514 '"url" field is missing or empty - skipping format, '
2515 'there is an error in extractor')
2516 return False
2517 if isinstance(url, bytes):
2518 sanitize_string_field(f, 'url')
2519 return True
2520
2521 # Filter out malformed formats for better extraction robustness
2522 formats = list(filter(is_wellformed, formats))
2523
2524 formats_dict = {}
2525
2526 # We check that all the formats have the format and format_id fields
2527 for i, format in enumerate(formats):
2528 sanitize_string_field(format, 'format_id')
2529 sanitize_numeric_fields(format)
2530 format['url'] = sanitize_url(format['url'])
2531 if not format.get('format_id'):
2532 format['format_id'] = str(i)
2533 else:
2534 # Sanitize format_id from characters used in format selector expression
2535 format['format_id'] = re.sub(r'[\s,/+\[\]()]', '_', format['format_id'])
2536 format_id = format['format_id']
2537 if format_id not in formats_dict:
2538 formats_dict[format_id] = []
2539 formats_dict[format_id].append(format)
2540
2541 # Make sure all formats have unique format_id
2542 common_exts = set(itertools.chain(*self._format_selection_exts.values()))
2543 for format_id, ambiguous_formats in formats_dict.items():
2544 ambigious_id = len(ambiguous_formats) > 1
2545 for i, format in enumerate(ambiguous_formats):
2546 if ambigious_id:
2547 format['format_id'] = '%s-%d' % (format_id, i)
2548 if format.get('ext') is None:
2549 format['ext'] = determine_ext(format['url']).lower()
2550 # Ensure there is no conflict between id and ext in format selection
2551 # See https://github.com/yt-dlp/yt-dlp/issues/1282
2552 if format['format_id'] != format['ext'] and format['format_id'] in common_exts:
2553 format['format_id'] = 'f%s' % format['format_id']
2554
2555 for i, format in enumerate(formats):
2556 if format.get('format') is None:
2557 format['format'] = '{id} - {res}{note}'.format(
2558 id=format['format_id'],
2559 res=self.format_resolution(format),
2560 note=format_field(format, 'format_note', ' (%s)'),
2561 )
2562 if format.get('protocol') is None:
2563 format['protocol'] = determine_protocol(format)
2564 if format.get('resolution') is None:
2565 format['resolution'] = self.format_resolution(format, default=None)
2566 if format.get('dynamic_range') is None and format.get('vcodec') != 'none':
2567 format['dynamic_range'] = 'SDR'
2568 if (info_dict.get('duration') and format.get('tbr')
2569 and not format.get('filesize') and not format.get('filesize_approx')):
2570 format['filesize_approx'] = int(info_dict['duration'] * format['tbr'] * (1024 / 8))
2571
2572 # Add HTTP headers, so that external programs can use them from the
2573 # json output
2574 full_format_info = info_dict.copy()
2575 full_format_info.update(format)
2576 format['http_headers'] = self._calc_headers(full_format_info)
2577 # Remove private housekeeping stuff
2578 if '__x_forwarded_for_ip' in info_dict:
2579 del info_dict['__x_forwarded_for_ip']
2580
2581 if self.params.get('check_formats') is True:
2582 formats = LazyList(self._check_formats(formats[::-1]), reverse=True)
2583
2584 if not formats or formats[0] is not info_dict:
2585 # only set the 'formats' fields if the original info_dict list them
2586 # otherwise we end up with a circular reference, the first (and unique)
2587 # element in the 'formats' field in info_dict is info_dict itself,
2588 # which can't be exported to json
2589 info_dict['formats'] = formats
2590
2591 info_dict, _ = self.pre_process(info_dict)
2592
2593 if self._match_entry(info_dict, incomplete=self._format_fields) is not None:
2594 return info_dict
2595
2596 self.post_extract(info_dict)
2597 info_dict, _ = self.pre_process(info_dict, 'after_filter')
2598
2599 # The pre-processors may have modified the formats
2600 formats = info_dict.get('formats', [info_dict])
2601
2602 list_only = self.params.get('simulate') is None and (
2603 self.params.get('list_thumbnails') or self.params.get('listformats') or self.params.get('listsubtitles'))
2604 interactive_format_selection = not list_only and self.format_selector == '-'
2605 if self.params.get('list_thumbnails'):
2606 self.list_thumbnails(info_dict)
2607 if self.params.get('listsubtitles'):
2608 if 'automatic_captions' in info_dict:
2609 self.list_subtitles(
2610 info_dict['id'], automatic_captions, 'automatic captions')
2611 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
2612 if self.params.get('listformats') or interactive_format_selection:
2613 self.list_formats(info_dict)
2614 if list_only:
2615 # Without this printing, -F --print-json will not work
2616 self.__forced_printings(info_dict, self.prepare_filename(info_dict), incomplete=True)
2617 return info_dict
2618
2619 format_selector = self.format_selector
2620 if format_selector is None:
2621 req_format = self._default_format_spec(info_dict, download=download)
2622 self.write_debug('Default format spec: %s' % req_format)
2623 format_selector = self.build_format_selector(req_format)
2624
2625 while True:
2626 if interactive_format_selection:
2627 req_format = input(
2628 self._format_screen('\nEnter format selector: ', self.Styles.EMPHASIS))
2629 try:
2630 format_selector = self.build_format_selector(req_format)
2631 except SyntaxError as err:
2632 self.report_error(err, tb=False, is_error=False)
2633 continue
2634
2635 formats_to_download = list(format_selector({
2636 'formats': formats,
2637 'has_merged_format': any('none' not in (f.get('acodec'), f.get('vcodec')) for f in formats),
2638 'incomplete_formats': (
2639 # All formats are video-only or
2640 all(f.get('vcodec') != 'none' and f.get('acodec') == 'none' for f in formats)
2641 # all formats are audio-only
2642 or all(f.get('vcodec') == 'none' and f.get('acodec') != 'none' for f in formats)),
2643 }))
2644 if interactive_format_selection and not formats_to_download:
2645 self.report_error('Requested format is not available', tb=False, is_error=False)
2646 continue
2647 break
2648
2649 if not formats_to_download:
2650 if not self.params.get('ignore_no_formats_error'):
2651 raise ExtractorError(
2652 'Requested format is not available. Use --list-formats for a list of available formats',
2653 expected=True, video_id=info_dict['id'], ie=info_dict['extractor'])
2654 self.report_warning('Requested format is not available')
2655 # Process what we can, even without any available formats.
2656 formats_to_download = [{}]
2657
2658 requested_ranges = self.params.get('download_ranges')
2659 if requested_ranges:
2660 requested_ranges = tuple(requested_ranges(info_dict, self))
2661
2662 best_format, downloaded_formats = formats_to_download[-1], []
2663 if download:
2664 if best_format:
2665 def to_screen(*msg):
2666 self.to_screen(f'[info] {info_dict["id"]}: {" ".join(", ".join(variadic(m)) for m in msg)}')
2667
2668 to_screen(f'Downloading {len(formats_to_download)} format(s):',
2669 (f['format_id'] for f in formats_to_download))
2670 if requested_ranges:
2671 to_screen(f'Downloading {len(requested_ranges)} time ranges:',
2672 (f'{int(c["start_time"])}-{int(c["end_time"])}' for c in requested_ranges))
2673 max_downloads_reached = False
2674
2675 for fmt, chapter in itertools.product(formats_to_download, requested_ranges or [{}]):
2676 new_info = self._copy_infodict(info_dict)
2677 new_info.update(fmt)
2678 offset, duration = info_dict.get('section_start') or 0, info_dict.get('duration') or float('inf')
2679 if chapter or offset:
2680 new_info.update({
2681 'section_start': offset + chapter.get('start_time', 0),
2682 'section_end': offset + min(chapter.get('end_time', duration), duration),
2683 'section_title': chapter.get('title'),
2684 'section_number': chapter.get('index'),
2685 })
2686 downloaded_formats.append(new_info)
2687 try:
2688 self.process_info(new_info)
2689 except MaxDownloadsReached:
2690 max_downloads_reached = True
2691 self._raise_pending_errors(new_info)
2692 # Remove copied info
2693 for key, val in tuple(new_info.items()):
2694 if info_dict.get(key) == val:
2695 new_info.pop(key)
2696 if max_downloads_reached:
2697 break
2698
2699 write_archive = {f.get('__write_download_archive', False) for f in downloaded_formats}
2700 assert write_archive.issubset({True, False, 'ignore'})
2701 if True in write_archive and False not in write_archive:
2702 self.record_download_archive(info_dict)
2703
2704 info_dict['requested_downloads'] = downloaded_formats
2705 info_dict = self.run_all_pps('after_video', info_dict)
2706 if max_downloads_reached:
2707 raise MaxDownloadsReached()
2708
2709 # We update the info dict with the selected best quality format (backwards compatibility)
2710 info_dict.update(best_format)
2711 return info_dict
2712
2713 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
2714 """Select the requested subtitles and their format"""
2715 available_subs, normal_sub_langs = {}, []
2716 if normal_subtitles and self.params.get('writesubtitles'):
2717 available_subs.update(normal_subtitles)
2718 normal_sub_langs = tuple(normal_subtitles.keys())
2719 if automatic_captions and self.params.get('writeautomaticsub'):
2720 for lang, cap_info in automatic_captions.items():
2721 if lang not in available_subs:
2722 available_subs[lang] = cap_info
2723
2724 if (not self.params.get('writesubtitles') and not
2725 self.params.get('writeautomaticsub') or not
2726 available_subs):
2727 return None
2728
2729 all_sub_langs = tuple(available_subs.keys())
2730 if self.params.get('allsubtitles', False):
2731 requested_langs = all_sub_langs
2732 elif self.params.get('subtitleslangs', False):
2733 # A list is used so that the order of languages will be the same as
2734 # given in subtitleslangs. See https://github.com/yt-dlp/yt-dlp/issues/1041
2735 requested_langs = []
2736 for lang_re in self.params.get('subtitleslangs'):
2737 discard = lang_re[0] == '-'
2738 if discard:
2739 lang_re = lang_re[1:]
2740 if lang_re == 'all':
2741 if discard:
2742 requested_langs = []
2743 else:
2744 requested_langs.extend(all_sub_langs)
2745 continue
2746 current_langs = filter(re.compile(lang_re + '$').match, all_sub_langs)
2747 if discard:
2748 for lang in current_langs:
2749 while lang in requested_langs:
2750 requested_langs.remove(lang)
2751 else:
2752 requested_langs.extend(current_langs)
2753 requested_langs = orderedSet(requested_langs)
2754 elif normal_sub_langs:
2755 requested_langs = ['en'] if 'en' in normal_sub_langs else normal_sub_langs[:1]
2756 else:
2757 requested_langs = ['en'] if 'en' in all_sub_langs else all_sub_langs[:1]
2758 if requested_langs:
2759 self.write_debug('Downloading subtitles: %s' % ', '.join(requested_langs))
2760
2761 formats_query = self.params.get('subtitlesformat', 'best')
2762 formats_preference = formats_query.split('/') if formats_query else []
2763 subs = {}
2764 for lang in requested_langs:
2765 formats = available_subs.get(lang)
2766 if formats is None:
2767 self.report_warning(f'{lang} subtitles not available for {video_id}')
2768 continue
2769 for ext in formats_preference:
2770 if ext == 'best':
2771 f = formats[-1]
2772 break
2773 matches = list(filter(lambda f: f['ext'] == ext, formats))
2774 if matches:
2775 f = matches[-1]
2776 break
2777 else:
2778 f = formats[-1]
2779 self.report_warning(
2780 'No subtitle format found matching "%s" for language %s, '
2781 'using %s' % (formats_query, lang, f['ext']))
2782 subs[lang] = f
2783 return subs
2784
2785 def _forceprint(self, key, info_dict):
2786 if info_dict is None:
2787 return
2788 info_copy = info_dict.copy()
2789 info_copy['formats_table'] = self.render_formats_table(info_dict)
2790 info_copy['thumbnails_table'] = self.render_thumbnails_table(info_dict)
2791 info_copy['subtitles_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('subtitles'))
2792 info_copy['automatic_captions_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('automatic_captions'))
2793
2794 def format_tmpl(tmpl):
2795 mobj = re.match(r'\w+(=?)$', tmpl)
2796 if mobj and mobj.group(1):
2797 return f'{tmpl[:-1]} = %({tmpl[:-1]})r'
2798 elif mobj:
2799 return f'%({tmpl})s'
2800 return tmpl
2801
2802 for tmpl in self.params['forceprint'].get(key, []):
2803 self.to_stdout(self.evaluate_outtmpl(format_tmpl(tmpl), info_copy))
2804
2805 for tmpl, file_tmpl in self.params['print_to_file'].get(key, []):
2806 filename = self.prepare_filename(info_dict, outtmpl=file_tmpl)
2807 tmpl = format_tmpl(tmpl)
2808 self.to_screen(f'[info] Writing {tmpl!r} to: {filename}')
2809 if self._ensure_dir_exists(filename):
2810 with open(filename, 'a', encoding='utf-8') as f:
2811 f.write(self.evaluate_outtmpl(tmpl, info_copy) + '\n')
2812
2813 def __forced_printings(self, info_dict, filename, incomplete):
2814 def print_mandatory(field, actual_field=None):
2815 if actual_field is None:
2816 actual_field = field
2817 if (self.params.get('force%s' % field, False)
2818 and (not incomplete or info_dict.get(actual_field) is not None)):
2819 self.to_stdout(info_dict[actual_field])
2820
2821 def print_optional(field):
2822 if (self.params.get('force%s' % field, False)
2823 and info_dict.get(field) is not None):
2824 self.to_stdout(info_dict[field])
2825
2826 info_dict = info_dict.copy()
2827 if filename is not None:
2828 info_dict['filename'] = filename
2829 if info_dict.get('requested_formats') is not None:
2830 # For RTMP URLs, also include the playpath
2831 info_dict['urls'] = '\n'.join(f['url'] + f.get('play_path', '') for f in info_dict['requested_formats'])
2832 elif info_dict.get('url'):
2833 info_dict['urls'] = info_dict['url'] + info_dict.get('play_path', '')
2834
2835 if (self.params.get('forcejson')
2836 or self.params['forceprint'].get('video')
2837 or self.params['print_to_file'].get('video')):
2838 self.post_extract(info_dict)
2839 self._forceprint('video', info_dict)
2840
2841 print_mandatory('title')
2842 print_mandatory('id')
2843 print_mandatory('url', 'urls')
2844 print_optional('thumbnail')
2845 print_optional('description')
2846 print_optional('filename')
2847 if self.params.get('forceduration') and info_dict.get('duration') is not None:
2848 self.to_stdout(formatSeconds(info_dict['duration']))
2849 print_mandatory('format')
2850
2851 if self.params.get('forcejson'):
2852 self.to_stdout(json.dumps(self.sanitize_info(info_dict)))
2853
2854 def dl(self, name, info, subtitle=False, test=False):
2855 if not info.get('url'):
2856 self.raise_no_formats(info, True)
2857
2858 if test:
2859 verbose = self.params.get('verbose')
2860 params = {
2861 'test': True,
2862 'quiet': self.params.get('quiet') or not verbose,
2863 'verbose': verbose,
2864 'noprogress': not verbose,
2865 'nopart': True,
2866 'skip_unavailable_fragments': False,
2867 'keep_fragments': False,
2868 'overwrites': True,
2869 '_no_ytdl_file': True,
2870 }
2871 else:
2872 params = self.params
2873 fd = get_suitable_downloader(info, params, to_stdout=(name == '-'))(self, params)
2874 if not test:
2875 for ph in self._progress_hooks:
2876 fd.add_progress_hook(ph)
2877 urls = '", "'.join(
2878 (f['url'].split(',')[0] + ',<data>' if f['url'].startswith('data:') else f['url'])
2879 for f in info.get('requested_formats', []) or [info])
2880 self.write_debug(f'Invoking {fd.FD_NAME} downloader on "{urls}"')
2881
2882 # Note: Ideally info should be a deep-copied so that hooks cannot modify it.
2883 # But it may contain objects that are not deep-copyable
2884 new_info = self._copy_infodict(info)
2885 if new_info.get('http_headers') is None:
2886 new_info['http_headers'] = self._calc_headers(new_info)
2887 return fd.download(name, new_info, subtitle)
2888
2889 def existing_file(self, filepaths, *, default_overwrite=True):
2890 existing_files = list(filter(os.path.exists, orderedSet(filepaths)))
2891 if existing_files and not self.params.get('overwrites', default_overwrite):
2892 return existing_files[0]
2893
2894 for file in existing_files:
2895 self.report_file_delete(file)
2896 os.remove(file)
2897 return None
2898
2899 def process_info(self, info_dict):
2900 """Process a single resolved IE result. (Modifies it in-place)"""
2901
2902 assert info_dict.get('_type', 'video') == 'video'
2903 original_infodict = info_dict
2904
2905 if 'format' not in info_dict and 'ext' in info_dict:
2906 info_dict['format'] = info_dict['ext']
2907
2908 # This is mostly just for backward compatibility of process_info
2909 # As a side-effect, this allows for format-specific filters
2910 if self._match_entry(info_dict) is not None:
2911 info_dict['__write_download_archive'] = 'ignore'
2912 return
2913
2914 # Does nothing under normal operation - for backward compatibility of process_info
2915 self.post_extract(info_dict)
2916 self._num_downloads += 1
2917
2918 # info_dict['_filename'] needs to be set for backward compatibility
2919 info_dict['_filename'] = full_filename = self.prepare_filename(info_dict, warn=True)
2920 temp_filename = self.prepare_filename(info_dict, 'temp')
2921 files_to_move = {}
2922
2923 # Forced printings
2924 self.__forced_printings(info_dict, full_filename, incomplete=('format' not in info_dict))
2925
2926 def check_max_downloads():
2927 if self._num_downloads >= float(self.params.get('max_downloads') or 'inf'):
2928 raise MaxDownloadsReached()
2929
2930 if self.params.get('simulate'):
2931 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
2932 check_max_downloads()
2933 return
2934
2935 if full_filename is None:
2936 return
2937 if not self._ensure_dir_exists(encodeFilename(full_filename)):
2938 return
2939 if not self._ensure_dir_exists(encodeFilename(temp_filename)):
2940 return
2941
2942 if self._write_description('video', info_dict,
2943 self.prepare_filename(info_dict, 'description')) is None:
2944 return
2945
2946 sub_files = self._write_subtitles(info_dict, temp_filename)
2947 if sub_files is None:
2948 return
2949 files_to_move.update(dict(sub_files))
2950
2951 thumb_files = self._write_thumbnails(
2952 'video', info_dict, temp_filename, self.prepare_filename(info_dict, 'thumbnail'))
2953 if thumb_files is None:
2954 return
2955 files_to_move.update(dict(thumb_files))
2956
2957 infofn = self.prepare_filename(info_dict, 'infojson')
2958 _infojson_written = self._write_info_json('video', info_dict, infofn)
2959 if _infojson_written:
2960 info_dict['infojson_filename'] = infofn
2961 # For backward compatibility, even though it was a private field
2962 info_dict['__infojson_filename'] = infofn
2963 elif _infojson_written is None:
2964 return
2965
2966 # Note: Annotations are deprecated
2967 annofn = None
2968 if self.params.get('writeannotations', False):
2969 annofn = self.prepare_filename(info_dict, 'annotation')
2970 if annofn:
2971 if not self._ensure_dir_exists(encodeFilename(annofn)):
2972 return
2973 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(annofn)):
2974 self.to_screen('[info] Video annotations are already present')
2975 elif not info_dict.get('annotations'):
2976 self.report_warning('There are no annotations to write.')
2977 else:
2978 try:
2979 self.to_screen('[info] Writing video annotations to: ' + annofn)
2980 with open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
2981 annofile.write(info_dict['annotations'])
2982 except (KeyError, TypeError):
2983 self.report_warning('There are no annotations to write.')
2984 except OSError:
2985 self.report_error('Cannot write annotations file: ' + annofn)
2986 return
2987
2988 # Write internet shortcut files
2989 def _write_link_file(link_type):
2990 url = try_get(info_dict['webpage_url'], iri_to_uri)
2991 if not url:
2992 self.report_warning(
2993 f'Cannot write internet shortcut file because the actual URL of "{info_dict["webpage_url"]}" is unknown')
2994 return True
2995 linkfn = replace_extension(self.prepare_filename(info_dict, 'link'), link_type, info_dict.get('ext'))
2996 if not self._ensure_dir_exists(encodeFilename(linkfn)):
2997 return False
2998 if self.params.get('overwrites', True) and os.path.exists(encodeFilename(linkfn)):
2999 self.to_screen(f'[info] Internet shortcut (.{link_type}) is already present')
3000 return True
3001 try:
3002 self.to_screen(f'[info] Writing internet shortcut (.{link_type}) to: {linkfn}')
3003 with open(encodeFilename(to_high_limit_path(linkfn)), 'w', encoding='utf-8',
3004 newline='\r\n' if link_type == 'url' else '\n') as linkfile:
3005 template_vars = {'url': url}
3006 if link_type == 'desktop':
3007 template_vars['filename'] = linkfn[:-(len(link_type) + 1)]
3008 linkfile.write(LINK_TEMPLATES[link_type] % template_vars)
3009 except OSError:
3010 self.report_error(f'Cannot write internet shortcut {linkfn}')
3011 return False
3012 return True
3013
3014 write_links = {
3015 'url': self.params.get('writeurllink'),
3016 'webloc': self.params.get('writewebloclink'),
3017 'desktop': self.params.get('writedesktoplink'),
3018 }
3019 if self.params.get('writelink'):
3020 link_type = ('webloc' if sys.platform == 'darwin'
3021 else 'desktop' if sys.platform.startswith('linux')
3022 else 'url')
3023 write_links[link_type] = True
3024
3025 if any(should_write and not _write_link_file(link_type)
3026 for link_type, should_write in write_links.items()):
3027 return
3028
3029 def replace_info_dict(new_info):
3030 nonlocal info_dict
3031 if new_info == info_dict:
3032 return
3033 info_dict.clear()
3034 info_dict.update(new_info)
3035
3036 new_info, files_to_move = self.pre_process(info_dict, 'before_dl', files_to_move)
3037 replace_info_dict(new_info)
3038
3039 if self.params.get('skip_download'):
3040 info_dict['filepath'] = temp_filename
3041 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
3042 info_dict['__files_to_move'] = files_to_move
3043 replace_info_dict(self.run_pp(MoveFilesAfterDownloadPP(self, False), info_dict))
3044 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
3045 else:
3046 # Download
3047 info_dict.setdefault('__postprocessors', [])
3048 try:
3049
3050 def existing_video_file(*filepaths):
3051 ext = info_dict.get('ext')
3052 converted = lambda file: replace_extension(file, self.params.get('final_ext') or ext, ext)
3053 file = self.existing_file(itertools.chain(*zip(map(converted, filepaths), filepaths)),
3054 default_overwrite=False)
3055 if file:
3056 info_dict['ext'] = os.path.splitext(file)[1][1:]
3057 return file
3058
3059 fd, success = None, True
3060 if info_dict.get('protocol') or info_dict.get('url'):
3061 fd = get_suitable_downloader(info_dict, self.params, to_stdout=temp_filename == '-')
3062 if fd is not FFmpegFD and (
3063 info_dict.get('section_start') or info_dict.get('section_end')):
3064 msg = ('This format cannot be partially downloaded' if FFmpegFD.available()
3065 else 'You have requested downloading the video partially, but ffmpeg is not installed')
3066 self.report_error(f'{msg}. Aborting')
3067 return
3068
3069 if info_dict.get('requested_formats') is not None:
3070
3071 def compatible_formats(formats):
3072 # TODO: some formats actually allow this (mkv, webm, ogg, mp4), but not all of them.
3073 video_formats = [format for format in formats if format.get('vcodec') != 'none']
3074 audio_formats = [format for format in formats if format.get('acodec') != 'none']
3075 if len(video_formats) > 2 or len(audio_formats) > 2:
3076 return False
3077
3078 # Check extension
3079 exts = {format.get('ext') for format in formats}
3080 COMPATIBLE_EXTS = (
3081 {'mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'ismv', 'isma'},
3082 {'webm'},
3083 )
3084 for ext_sets in COMPATIBLE_EXTS:
3085 if ext_sets.issuperset(exts):
3086 return True
3087 # TODO: Check acodec/vcodec
3088 return False
3089
3090 requested_formats = info_dict['requested_formats']
3091 old_ext = info_dict['ext']
3092 if self.params.get('merge_output_format') is None:
3093 if not compatible_formats(requested_formats):
3094 info_dict['ext'] = 'mkv'
3095 self.report_warning(
3096 'Requested formats are incompatible for merge and will be merged into mkv')
3097 if (info_dict['ext'] == 'webm'
3098 and info_dict.get('thumbnails')
3099 # check with type instead of pp_key, __name__, or isinstance
3100 # since we dont want any custom PPs to trigger this
3101 and any(type(pp) == EmbedThumbnailPP for pp in self._pps['post_process'])): # noqa: E721
3102 info_dict['ext'] = 'mkv'
3103 self.report_warning(
3104 'webm doesn\'t support embedding a thumbnail, mkv will be used')
3105 new_ext = info_dict['ext']
3106
3107 def correct_ext(filename, ext=new_ext):
3108 if filename == '-':
3109 return filename
3110 filename_real_ext = os.path.splitext(filename)[1][1:]
3111 filename_wo_ext = (
3112 os.path.splitext(filename)[0]
3113 if filename_real_ext in (old_ext, new_ext)
3114 else filename)
3115 return f'{filename_wo_ext}.{ext}'
3116
3117 # Ensure filename always has a correct extension for successful merge
3118 full_filename = correct_ext(full_filename)
3119 temp_filename = correct_ext(temp_filename)
3120 dl_filename = existing_video_file(full_filename, temp_filename)
3121 info_dict['__real_download'] = False
3122
3123 merger = FFmpegMergerPP(self)
3124 downloaded = []
3125 if dl_filename is not None:
3126 self.report_file_already_downloaded(dl_filename)
3127 elif fd:
3128 for f in requested_formats if fd != FFmpegFD else []:
3129 f['filepath'] = fname = prepend_extension(
3130 correct_ext(temp_filename, info_dict['ext']),
3131 'f%s' % f['format_id'], info_dict['ext'])
3132 downloaded.append(fname)
3133 info_dict['url'] = '\n'.join(f['url'] for f in requested_formats)
3134 success, real_download = self.dl(temp_filename, info_dict)
3135 info_dict['__real_download'] = real_download
3136 else:
3137 if self.params.get('allow_unplayable_formats'):
3138 self.report_warning(
3139 'You have requested merging of multiple formats '
3140 'while also allowing unplayable formats to be downloaded. '
3141 'The formats won\'t be merged to prevent data corruption.')
3142 elif not merger.available:
3143 msg = 'You have requested merging of multiple formats but ffmpeg is not installed'
3144 if not self.params.get('ignoreerrors'):
3145 self.report_error(f'{msg}. Aborting due to --abort-on-error')
3146 return
3147 self.report_warning(f'{msg}. The formats won\'t be merged')
3148
3149 if temp_filename == '-':
3150 reason = ('using a downloader other than ffmpeg' if FFmpegFD.can_merge_formats(info_dict, self.params)
3151 else 'but the formats are incompatible for simultaneous download' if merger.available
3152 else 'but ffmpeg is not installed')
3153 self.report_warning(
3154 f'You have requested downloading multiple formats to stdout {reason}. '
3155 'The formats will be streamed one after the other')
3156 fname = temp_filename
3157 for f in requested_formats:
3158 new_info = dict(info_dict)
3159 del new_info['requested_formats']
3160 new_info.update(f)
3161 if temp_filename != '-':
3162 fname = prepend_extension(
3163 correct_ext(temp_filename, new_info['ext']),
3164 'f%s' % f['format_id'], new_info['ext'])
3165 if not self._ensure_dir_exists(fname):
3166 return
3167 f['filepath'] = fname
3168 downloaded.append(fname)
3169 partial_success, real_download = self.dl(fname, new_info)
3170 info_dict['__real_download'] = info_dict['__real_download'] or real_download
3171 success = success and partial_success
3172
3173 if downloaded and merger.available and not self.params.get('allow_unplayable_formats'):
3174 info_dict['__postprocessors'].append(merger)
3175 info_dict['__files_to_merge'] = downloaded
3176 # Even if there were no downloads, it is being merged only now
3177 info_dict['__real_download'] = True
3178 else:
3179 for file in downloaded:
3180 files_to_move[file] = None
3181 else:
3182 # Just a single file
3183 dl_filename = existing_video_file(full_filename, temp_filename)
3184 if dl_filename is None or dl_filename == temp_filename:
3185 # dl_filename == temp_filename could mean that the file was partially downloaded with --no-part.
3186 # So we should try to resume the download
3187 success, real_download = self.dl(temp_filename, info_dict)
3188 info_dict['__real_download'] = real_download
3189 else:
3190 self.report_file_already_downloaded(dl_filename)
3191
3192 dl_filename = dl_filename or temp_filename
3193 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
3194
3195 except network_exceptions as err:
3196 self.report_error('unable to download video data: %s' % error_to_compat_str(err))
3197 return
3198 except OSError as err:
3199 raise UnavailableVideoError(err)
3200 except (ContentTooShortError, ) as err:
3201 self.report_error(f'content too short (expected {err.expected} bytes and served {err.downloaded})')
3202 return
3203
3204 self._raise_pending_errors(info_dict)
3205 if success and full_filename != '-':
3206
3207 def fixup():
3208 do_fixup = True
3209 fixup_policy = self.params.get('fixup')
3210 vid = info_dict['id']
3211
3212 if fixup_policy in ('ignore', 'never'):
3213 return
3214 elif fixup_policy == 'warn':
3215 do_fixup = 'warn'
3216 elif fixup_policy != 'force':
3217 assert fixup_policy in ('detect_or_warn', None)
3218 if not info_dict.get('__real_download'):
3219 do_fixup = False
3220
3221 def ffmpeg_fixup(cndn, msg, cls):
3222 if not (do_fixup and cndn):
3223 return
3224 elif do_fixup == 'warn':
3225 self.report_warning(f'{vid}: {msg}')
3226 return
3227 pp = cls(self)
3228 if pp.available:
3229 info_dict['__postprocessors'].append(pp)
3230 else:
3231 self.report_warning(f'{vid}: {msg}. Install ffmpeg to fix this automatically')
3232
3233 stretched_ratio = info_dict.get('stretched_ratio')
3234 ffmpeg_fixup(stretched_ratio not in (1, None),
3235 f'Non-uniform pixel ratio {stretched_ratio}',
3236 FFmpegFixupStretchedPP)
3237
3238 downloader = get_suitable_downloader(info_dict, self.params) if 'protocol' in info_dict else None
3239 downloader = downloader.FD_NAME if downloader else None
3240
3241 ext = info_dict.get('ext')
3242 postprocessed_by_ffmpeg = info_dict.get('requested_formats') or any((
3243 isinstance(pp, FFmpegVideoConvertorPP)
3244 and resolve_recode_mapping(ext, pp.mapping)[0] not in (ext, None)
3245 ) for pp in self._pps['post_process'])
3246
3247 if not postprocessed_by_ffmpeg:
3248 ffmpeg_fixup(ext == 'm4a' and info_dict.get('container') == 'm4a_dash',
3249 'writing DASH m4a. Only some players support this container',
3250 FFmpegFixupM4aPP)
3251 ffmpeg_fixup(downloader == 'hlsnative' and not self.params.get('hls_use_mpegts')
3252 or info_dict.get('is_live') and self.params.get('hls_use_mpegts') is None,
3253 'Possible MPEG-TS in MP4 container or malformed AAC timestamps',
3254 FFmpegFixupM3u8PP)
3255 ffmpeg_fixup(info_dict.get('is_live') and downloader == 'DashSegmentsFD',
3256 'Possible duplicate MOOV atoms', FFmpegFixupDuplicateMoovPP)
3257
3258 ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed timestamps detected', FFmpegFixupTimestampPP)
3259 ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed duration detected', FFmpegFixupDurationPP)
3260
3261 fixup()
3262 try:
3263 replace_info_dict(self.post_process(dl_filename, info_dict, files_to_move))
3264 except PostProcessingError as err:
3265 self.report_error('Postprocessing: %s' % str(err))
3266 return
3267 try:
3268 for ph in self._post_hooks:
3269 ph(info_dict['filepath'])
3270 except Exception as err:
3271 self.report_error('post hooks: %s' % str(err))
3272 return
3273 info_dict['__write_download_archive'] = True
3274
3275 assert info_dict is original_infodict # Make sure the info_dict was modified in-place
3276 if self.params.get('force_write_download_archive'):
3277 info_dict['__write_download_archive'] = True
3278 check_max_downloads()
3279
3280 def __download_wrapper(self, func):
3281 @functools.wraps(func)
3282 def wrapper(*args, **kwargs):
3283 try:
3284 res = func(*args, **kwargs)
3285 except UnavailableVideoError as e:
3286 self.report_error(e)
3287 except DownloadCancelled as e:
3288 self.to_screen(f'[info] {e}')
3289 if not self.params.get('break_per_url'):
3290 raise
3291 else:
3292 if self.params.get('dump_single_json', False):
3293 self.post_extract(res)
3294 self.to_stdout(json.dumps(self.sanitize_info(res)))
3295 return wrapper
3296
3297 def download(self, url_list):
3298 """Download a given list of URLs."""
3299 url_list = variadic(url_list) # Passing a single URL is a common mistake
3300 outtmpl = self.params['outtmpl']['default']
3301 if (len(url_list) > 1
3302 and outtmpl != '-'
3303 and '%' not in outtmpl
3304 and self.params.get('max_downloads') != 1):
3305 raise SameFileError(outtmpl)
3306
3307 for url in url_list:
3308 self.__download_wrapper(self.extract_info)(
3309 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
3310
3311 return self._download_retcode
3312
3313 def download_with_info_file(self, info_filename):
3314 with contextlib.closing(fileinput.FileInput(
3315 [info_filename], mode='r',
3316 openhook=fileinput.hook_encoded('utf-8'))) as f:
3317 # FileInput doesn't have a read method, we can't call json.load
3318 info = self.sanitize_info(json.loads('\n'.join(f)), self.params.get('clean_infojson', True))
3319 try:
3320 self.__download_wrapper(self.process_ie_result)(info, download=True)
3321 except (DownloadError, EntryNotInPlaylist, ReExtractInfo) as e:
3322 if not isinstance(e, EntryNotInPlaylist):
3323 self.to_stderr('\r')
3324 webpage_url = info.get('webpage_url')
3325 if webpage_url is not None:
3326 self.report_warning(f'The info failed to download: {e}; trying with URL {webpage_url}')
3327 return self.download([webpage_url])
3328 else:
3329 raise
3330 return self._download_retcode
3331
3332 @staticmethod
3333 def sanitize_info(info_dict, remove_private_keys=False):
3334 ''' Sanitize the infodict for converting to json '''
3335 if info_dict is None:
3336 return info_dict
3337 info_dict.setdefault('epoch', int(time.time()))
3338 info_dict.setdefault('_type', 'video')
3339
3340 if remove_private_keys:
3341 reject = lambda k, v: v is None or k.startswith('__') or k in {
3342 'requested_downloads', 'requested_formats', 'requested_subtitles', 'requested_entries',
3343 'entries', 'filepath', '_filename', 'infojson_filename', 'original_url', 'playlist_autonumber',
3344 }
3345 else:
3346 reject = lambda k, v: False
3347
3348 def filter_fn(obj):
3349 if isinstance(obj, dict):
3350 return {k: filter_fn(v) for k, v in obj.items() if not reject(k, v)}
3351 elif isinstance(obj, (list, tuple, set, LazyList)):
3352 return list(map(filter_fn, obj))
3353 elif obj is None or isinstance(obj, (str, int, float, bool)):
3354 return obj
3355 else:
3356 return repr(obj)
3357
3358 return filter_fn(info_dict)
3359
3360 @staticmethod
3361 def filter_requested_info(info_dict, actually_filter=True):
3362 ''' Alias of sanitize_info for backward compatibility '''
3363 return YoutubeDL.sanitize_info(info_dict, actually_filter)
3364
3365 def _delete_downloaded_files(self, *files_to_delete, info={}, msg=None):
3366 for filename in set(filter(None, files_to_delete)):
3367 if msg:
3368 self.to_screen(msg % filename)
3369 try:
3370 os.remove(filename)
3371 except OSError:
3372 self.report_warning(f'Unable to delete file {filename}')
3373 if filename in info.get('__files_to_move', []): # NB: Delete even if None
3374 del info['__files_to_move'][filename]
3375
3376 @staticmethod
3377 def post_extract(info_dict):
3378 def actual_post_extract(info_dict):
3379 if info_dict.get('_type') in ('playlist', 'multi_video'):
3380 for video_dict in info_dict.get('entries', {}):
3381 actual_post_extract(video_dict or {})
3382 return
3383
3384 post_extractor = info_dict.pop('__post_extractor', None) or (lambda: {})
3385 info_dict.update(post_extractor())
3386
3387 actual_post_extract(info_dict or {})
3388
3389 def run_pp(self, pp, infodict):
3390 files_to_delete = []
3391 if '__files_to_move' not in infodict:
3392 infodict['__files_to_move'] = {}
3393 try:
3394 files_to_delete, infodict = pp.run(infodict)
3395 except PostProcessingError as e:
3396 # Must be True and not 'only_download'
3397 if self.params.get('ignoreerrors') is True:
3398 self.report_error(e)
3399 return infodict
3400 raise
3401
3402 if not files_to_delete:
3403 return infodict
3404 if self.params.get('keepvideo', False):
3405 for f in files_to_delete:
3406 infodict['__files_to_move'].setdefault(f, '')
3407 else:
3408 self._delete_downloaded_files(
3409 *files_to_delete, info=infodict, msg='Deleting original file %s (pass -k to keep)')
3410 return infodict
3411
3412 def run_all_pps(self, key, info, *, additional_pps=None):
3413 self._forceprint(key, info)
3414 for pp in (additional_pps or []) + self._pps[key]:
3415 info = self.run_pp(pp, info)
3416 return info
3417
3418 def pre_process(self, ie_info, key='pre_process', files_to_move=None):
3419 info = dict(ie_info)
3420 info['__files_to_move'] = files_to_move or {}
3421 try:
3422 info = self.run_all_pps(key, info)
3423 except PostProcessingError as err:
3424 msg = f'Preprocessing: {err}'
3425 info.setdefault('__pending_error', msg)
3426 self.report_error(msg, is_error=False)
3427 return info, info.pop('__files_to_move', None)
3428
3429 def post_process(self, filename, info, files_to_move=None):
3430 """Run all the postprocessors on the given file."""
3431 info['filepath'] = filename
3432 info['__files_to_move'] = files_to_move or {}
3433 info = self.run_all_pps('post_process', info, additional_pps=info.get('__postprocessors'))
3434 info = self.run_pp(MoveFilesAfterDownloadPP(self), info)
3435 del info['__files_to_move']
3436 return self.run_all_pps('after_move', info)
3437
3438 def _make_archive_id(self, info_dict):
3439 video_id = info_dict.get('id')
3440 if not video_id:
3441 return
3442 # Future-proof against any change in case
3443 # and backwards compatibility with prior versions
3444 extractor = info_dict.get('extractor_key') or info_dict.get('ie_key') # key in a playlist
3445 if extractor is None:
3446 url = str_or_none(info_dict.get('url'))
3447 if not url:
3448 return
3449 # Try to find matching extractor for the URL and take its ie_key
3450 for ie_key, ie in self._ies.items():
3451 if ie.suitable(url):
3452 extractor = ie_key
3453 break
3454 else:
3455 return
3456 return make_archive_id(extractor, video_id)
3457
3458 def in_download_archive(self, info_dict):
3459 fn = self.params.get('download_archive')
3460 if fn is None:
3461 return False
3462
3463 vid_ids = [self._make_archive_id(info_dict)]
3464 vid_ids.extend(info_dict.get('_old_archive_ids', []))
3465 return any(id_ in self.archive for id_ in vid_ids)
3466
3467 def record_download_archive(self, info_dict):
3468 fn = self.params.get('download_archive')
3469 if fn is None:
3470 return
3471 vid_id = self._make_archive_id(info_dict)
3472 assert vid_id
3473 self.write_debug(f'Adding to archive: {vid_id}')
3474 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
3475 archive_file.write(vid_id + '\n')
3476 self.archive.add(vid_id)
3477
3478 @staticmethod
3479 def format_resolution(format, default='unknown'):
3480 if format.get('vcodec') == 'none' and format.get('acodec') != 'none':
3481 return 'audio only'
3482 if format.get('resolution') is not None:
3483 return format['resolution']
3484 if format.get('width') and format.get('height'):
3485 return '%dx%d' % (format['width'], format['height'])
3486 elif format.get('height'):
3487 return '%sp' % format['height']
3488 elif format.get('width'):
3489 return '%dx?' % format['width']
3490 return default
3491
3492 def _list_format_headers(self, *headers):
3493 if self.params.get('listformats_table', True) is not False:
3494 return [self._format_out(header, self.Styles.HEADERS) for header in headers]
3495 return headers
3496
3497 def _format_note(self, fdict):
3498 res = ''
3499 if fdict.get('ext') in ['f4f', 'f4m']:
3500 res += '(unsupported)'
3501 if fdict.get('language'):
3502 if res:
3503 res += ' '
3504 res += '[%s]' % fdict['language']
3505 if fdict.get('format_note') is not None:
3506 if res:
3507 res += ' '
3508 res += fdict['format_note']
3509 if fdict.get('tbr') is not None:
3510 if res:
3511 res += ', '
3512 res += '%4dk' % fdict['tbr']
3513 if fdict.get('container') is not None:
3514 if res:
3515 res += ', '
3516 res += '%s container' % fdict['container']
3517 if (fdict.get('vcodec') is not None
3518 and fdict.get('vcodec') != 'none'):
3519 if res:
3520 res += ', '
3521 res += fdict['vcodec']
3522 if fdict.get('vbr') is not None:
3523 res += '@'
3524 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
3525 res += 'video@'
3526 if fdict.get('vbr') is not None:
3527 res += '%4dk' % fdict['vbr']
3528 if fdict.get('fps') is not None:
3529 if res:
3530 res += ', '
3531 res += '%sfps' % fdict['fps']
3532 if fdict.get('acodec') is not None:
3533 if res:
3534 res += ', '
3535 if fdict['acodec'] == 'none':
3536 res += 'video only'
3537 else:
3538 res += '%-5s' % fdict['acodec']
3539 elif fdict.get('abr') is not None:
3540 if res:
3541 res += ', '
3542 res += 'audio'
3543 if fdict.get('abr') is not None:
3544 res += '@%3dk' % fdict['abr']
3545 if fdict.get('asr') is not None:
3546 res += ' (%5dHz)' % fdict['asr']
3547 if fdict.get('filesize') is not None:
3548 if res:
3549 res += ', '
3550 res += format_bytes(fdict['filesize'])
3551 elif fdict.get('filesize_approx') is not None:
3552 if res:
3553 res += ', '
3554 res += '~' + format_bytes(fdict['filesize_approx'])
3555 return res
3556
3557 def render_formats_table(self, info_dict):
3558 if not info_dict.get('formats') and not info_dict.get('url'):
3559 return None
3560
3561 formats = info_dict.get('formats', [info_dict])
3562 if not self.params.get('listformats_table', True) is not False:
3563 table = [
3564 [
3565 format_field(f, 'format_id'),
3566 format_field(f, 'ext'),
3567 self.format_resolution(f),
3568 self._format_note(f)
3569 ] for f in formats if f.get('preference') is None or f['preference'] >= -1000]
3570 return render_table(['format code', 'extension', 'resolution', 'note'], table, extra_gap=1)
3571
3572 def simplified_codec(f, field):
3573 assert field in ('acodec', 'vcodec')
3574 codec = f.get(field, 'unknown')
3575 if not codec:
3576 return 'unknown'
3577 elif codec != 'none':
3578 return '.'.join(codec.split('.')[:4])
3579
3580 if field == 'vcodec' and f.get('acodec') == 'none':
3581 return 'images'
3582 elif field == 'acodec' and f.get('vcodec') == 'none':
3583 return ''
3584 return self._format_out('audio only' if field == 'vcodec' else 'video only',
3585 self.Styles.SUPPRESS)
3586
3587 delim = self._format_out('\u2502', self.Styles.DELIM, '|', test_encoding=True)
3588 table = [
3589 [
3590 self._format_out(format_field(f, 'format_id'), self.Styles.ID),
3591 format_field(f, 'ext'),
3592 format_field(f, func=self.format_resolution, ignore=('audio only', 'images')),
3593 format_field(f, 'fps', '\t%d', func=round),
3594 format_field(f, 'dynamic_range', '%s', ignore=(None, 'SDR')).replace('HDR', ''),
3595 delim,
3596 format_field(f, 'filesize', ' \t%s', func=format_bytes) + format_field(f, 'filesize_approx', '~\t%s', func=format_bytes),
3597 format_field(f, 'tbr', '\t%dk', func=round),
3598 shorten_protocol_name(f.get('protocol', '')),
3599 delim,
3600 simplified_codec(f, 'vcodec'),
3601 format_field(f, 'vbr', '\t%dk', func=round),
3602 simplified_codec(f, 'acodec'),
3603 format_field(f, 'abr', '\t%dk', func=round),
3604 format_field(f, 'asr', '\t%s', func=format_decimal_suffix),
3605 join_nonempty(
3606 self._format_out('UNSUPPORTED', 'light red') if f.get('ext') in ('f4f', 'f4m') else None,
3607 format_field(f, 'language', '[%s]'),
3608 join_nonempty(format_field(f, 'format_note'),
3609 format_field(f, 'container', ignore=(None, f.get('ext'))),
3610 delim=', '),
3611 delim=' '),
3612 ] for f in formats if f.get('preference') is None or f['preference'] >= -1000]
3613 header_line = self._list_format_headers(
3614 'ID', 'EXT', 'RESOLUTION', '\tFPS', 'HDR', delim, '\tFILESIZE', '\tTBR', 'PROTO',
3615 delim, 'VCODEC', '\tVBR', 'ACODEC', '\tABR', '\tASR', 'MORE INFO')
3616
3617 return render_table(
3618 header_line, table, hide_empty=True,
3619 delim=self._format_out('\u2500', self.Styles.DELIM, '-', test_encoding=True))
3620
3621 def render_thumbnails_table(self, info_dict):
3622 thumbnails = list(info_dict.get('thumbnails') or [])
3623 if not thumbnails:
3624 return None
3625 return render_table(
3626 self._list_format_headers('ID', 'Width', 'Height', 'URL'),
3627 [[t.get('id'), t.get('width', 'unknown'), t.get('height', 'unknown'), t['url']] for t in thumbnails])
3628
3629 def render_subtitles_table(self, video_id, subtitles):
3630 def _row(lang, formats):
3631 exts, names = zip(*((f['ext'], f.get('name') or 'unknown') for f in reversed(formats)))
3632 if len(set(names)) == 1:
3633 names = [] if names[0] == 'unknown' else names[:1]
3634 return [lang, ', '.join(names), ', '.join(exts)]
3635
3636 if not subtitles:
3637 return None
3638 return render_table(
3639 self._list_format_headers('Language', 'Name', 'Formats'),
3640 [_row(lang, formats) for lang, formats in subtitles.items()],
3641 hide_empty=True)
3642
3643 def __list_table(self, video_id, name, func, *args):
3644 table = func(*args)
3645 if not table:
3646 self.to_screen(f'{video_id} has no {name}')
3647 return
3648 self.to_screen(f'[info] Available {name} for {video_id}:')
3649 self.to_stdout(table)
3650
3651 def list_formats(self, info_dict):
3652 self.__list_table(info_dict['id'], 'formats', self.render_formats_table, info_dict)
3653
3654 def list_thumbnails(self, info_dict):
3655 self.__list_table(info_dict['id'], 'thumbnails', self.render_thumbnails_table, info_dict)
3656
3657 def list_subtitles(self, video_id, subtitles, name='subtitles'):
3658 self.__list_table(video_id, name, self.render_subtitles_table, video_id, subtitles)
3659
3660 def urlopen(self, req):
3661 """ Start an HTTP download """
3662 if isinstance(req, str):
3663 req = sanitized_Request(req)
3664 return self._opener.open(req, timeout=self._socket_timeout)
3665
3666 def print_debug_header(self):
3667 if not self.params.get('verbose'):
3668 return
3669
3670 # These imports can be slow. So import them only as needed
3671 from .extractor.extractors import _LAZY_LOADER
3672 from .extractor.extractors import _PLUGIN_CLASSES as plugin_extractors
3673
3674 def get_encoding(stream):
3675 ret = str(getattr(stream, 'encoding', 'missing (%s)' % type(stream).__name__))
3676 if not supports_terminal_sequences(stream):
3677 from .utils import WINDOWS_VT_MODE # Must be imported locally
3678 ret += ' (No VT)' if WINDOWS_VT_MODE is False else ' (No ANSI)'
3679 return ret
3680
3681 encoding_str = 'Encodings: locale %s, fs %s, pref %s, %s' % (
3682 locale.getpreferredencoding(),
3683 sys.getfilesystemencoding(),
3684 self.get_encoding(),
3685 ', '.join(
3686 f'{key} {get_encoding(stream)}' for key, stream in self._out_files.items_
3687 if stream is not None and key != 'console')
3688 )
3689
3690 logger = self.params.get('logger')
3691 if logger:
3692 write_debug = lambda msg: logger.debug(f'[debug] {msg}')
3693 write_debug(encoding_str)
3694 else:
3695 write_string(f'[debug] {encoding_str}\n', encoding=None)
3696 write_debug = lambda msg: self._write_string(f'[debug] {msg}\n')
3697
3698 source = detect_variant()
3699 write_debug(join_nonempty(
3700 'yt-dlp version', __version__,
3701 f'[{RELEASE_GIT_HEAD}]' if RELEASE_GIT_HEAD else '',
3702 '' if source == 'unknown' else f'({source})',
3703 delim=' '))
3704 if not _LAZY_LOADER:
3705 if os.environ.get('YTDLP_NO_LAZY_EXTRACTORS'):
3706 write_debug('Lazy loading extractors is forcibly disabled')
3707 else:
3708 write_debug('Lazy loading extractors is disabled')
3709 if plugin_extractors or plugin_postprocessors:
3710 write_debug('Plugins: %s' % [
3711 '%s%s' % (klass.__name__, '' if klass.__name__ == name else f' as {name}')
3712 for name, klass in itertools.chain(plugin_extractors.items(), plugin_postprocessors.items())])
3713 if self.params['compat_opts']:
3714 write_debug('Compatibility options: %s' % ', '.join(self.params['compat_opts']))
3715
3716 if source == 'source':
3717 try:
3718 stdout, _, _ = Popen.run(
3719 ['git', 'rev-parse', '--short', 'HEAD'],
3720 text=True, cwd=os.path.dirname(os.path.abspath(__file__)),
3721 stdout=subprocess.PIPE, stderr=subprocess.PIPE)
3722 if re.fullmatch('[0-9a-f]+', stdout.strip()):
3723 write_debug(f'Git HEAD: {stdout.strip()}')
3724 except Exception:
3725 with contextlib.suppress(Exception):
3726 sys.exc_clear()
3727
3728 write_debug(system_identifier())
3729
3730 exe_versions, ffmpeg_features = FFmpegPostProcessor.get_versions_and_features(self)
3731 ffmpeg_features = {key for key, val in ffmpeg_features.items() if val}
3732 if ffmpeg_features:
3733 exe_versions['ffmpeg'] += ' (%s)' % ','.join(sorted(ffmpeg_features))
3734
3735 exe_versions['rtmpdump'] = rtmpdump_version()
3736 exe_versions['phantomjs'] = PhantomJSwrapper._version()
3737 exe_str = ', '.join(
3738 f'{exe} {v}' for exe, v in sorted(exe_versions.items()) if v
3739 ) or 'none'
3740 write_debug('exe versions: %s' % exe_str)
3741
3742 from .compat.compat_utils import get_package_info
3743 from .dependencies import available_dependencies
3744
3745 write_debug('Optional libraries: %s' % (', '.join(sorted({
3746 join_nonempty(*get_package_info(m)) for m in available_dependencies.values()
3747 })) or 'none'))
3748
3749 self._setup_opener()
3750 proxy_map = {}
3751 for handler in self._opener.handlers:
3752 if hasattr(handler, 'proxies'):
3753 proxy_map.update(handler.proxies)
3754 write_debug(f'Proxy map: {proxy_map}')
3755
3756 # Not implemented
3757 if False and self.params.get('call_home'):
3758 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode()
3759 write_debug('Public IP address: %s' % ipaddr)
3760 latest_version = self.urlopen(
3761 'https://yt-dl.org/latest/version').read().decode()
3762 if version_tuple(latest_version) > version_tuple(__version__):
3763 self.report_warning(
3764 'You are using an outdated version (newest version: %s)! '
3765 'See https://yt-dl.org/update if you need help updating.' %
3766 latest_version)
3767
3768 def _setup_opener(self):
3769 if hasattr(self, '_opener'):
3770 return
3771 timeout_val = self.params.get('socket_timeout')
3772 self._socket_timeout = 20 if timeout_val is None else float(timeout_val)
3773
3774 opts_cookiesfrombrowser = self.params.get('cookiesfrombrowser')
3775 opts_cookiefile = self.params.get('cookiefile')
3776 opts_proxy = self.params.get('proxy')
3777
3778 self.cookiejar = load_cookies(opts_cookiefile, opts_cookiesfrombrowser, self)
3779
3780 cookie_processor = YoutubeDLCookieProcessor(self.cookiejar)
3781 if opts_proxy is not None:
3782 if opts_proxy == '':
3783 proxies = {}
3784 else:
3785 proxies = {'http': opts_proxy, 'https': opts_proxy}
3786 else:
3787 proxies = urllib.request.getproxies()
3788 # Set HTTPS proxy to HTTP one if given (https://github.com/ytdl-org/youtube-dl/issues/805)
3789 if 'http' in proxies and 'https' not in proxies:
3790 proxies['https'] = proxies['http']
3791 proxy_handler = PerRequestProxyHandler(proxies)
3792
3793 debuglevel = 1 if self.params.get('debug_printtraffic') else 0
3794 https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
3795 ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
3796 redirect_handler = YoutubeDLRedirectHandler()
3797 data_handler = urllib.request.DataHandler()
3798
3799 # When passing our own FileHandler instance, build_opener won't add the
3800 # default FileHandler and allows us to disable the file protocol, which
3801 # can be used for malicious purposes (see
3802 # https://github.com/ytdl-org/youtube-dl/issues/8227)
3803 file_handler = urllib.request.FileHandler()
3804
3805 def file_open(*args, **kwargs):
3806 raise urllib.error.URLError('file:// scheme is explicitly disabled in yt-dlp for security reasons')
3807 file_handler.file_open = file_open
3808
3809 opener = urllib.request.build_opener(
3810 proxy_handler, https_handler, cookie_processor, ydlh, redirect_handler, data_handler, file_handler)
3811
3812 # Delete the default user-agent header, which would otherwise apply in
3813 # cases where our custom HTTP handler doesn't come into play
3814 # (See https://github.com/ytdl-org/youtube-dl/issues/1309 for details)
3815 opener.addheaders = []
3816 self._opener = opener
3817
3818 def encode(self, s):
3819 if isinstance(s, bytes):
3820 return s # Already encoded
3821
3822 try:
3823 return s.encode(self.get_encoding())
3824 except UnicodeEncodeError as err:
3825 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
3826 raise
3827
3828 def get_encoding(self):
3829 encoding = self.params.get('encoding')
3830 if encoding is None:
3831 encoding = preferredencoding()
3832 return encoding
3833
3834 def _write_info_json(self, label, ie_result, infofn, overwrite=None):
3835 ''' Write infojson and returns True = written, 'exists' = Already exists, False = skip, None = error '''
3836 if overwrite is None:
3837 overwrite = self.params.get('overwrites', True)
3838 if not self.params.get('writeinfojson'):
3839 return False
3840 elif not infofn:
3841 self.write_debug(f'Skipping writing {label} infojson')
3842 return False
3843 elif not self._ensure_dir_exists(infofn):
3844 return None
3845 elif not overwrite and os.path.exists(infofn):
3846 self.to_screen(f'[info] {label.title()} metadata is already present')
3847 return 'exists'
3848
3849 self.to_screen(f'[info] Writing {label} metadata as JSON to: {infofn}')
3850 try:
3851 write_json_file(self.sanitize_info(ie_result, self.params.get('clean_infojson', True)), infofn)
3852 return True
3853 except OSError:
3854 self.report_error(f'Cannot write {label} metadata to JSON file {infofn}')
3855 return None
3856
3857 def _write_description(self, label, ie_result, descfn):
3858 ''' Write description and returns True = written, False = skip, None = error '''
3859 if not self.params.get('writedescription'):
3860 return False
3861 elif not descfn:
3862 self.write_debug(f'Skipping writing {label} description')
3863 return False
3864 elif not self._ensure_dir_exists(descfn):
3865 return None
3866 elif not self.params.get('overwrites', True) and os.path.exists(descfn):
3867 self.to_screen(f'[info] {label.title()} description is already present')
3868 elif ie_result.get('description') is None:
3869 self.report_warning(f'There\'s no {label} description to write')
3870 return False
3871 else:
3872 try:
3873 self.to_screen(f'[info] Writing {label} description to: {descfn}')
3874 with open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
3875 descfile.write(ie_result['description'])
3876 except OSError:
3877 self.report_error(f'Cannot write {label} description file {descfn}')
3878 return None
3879 return True
3880
3881 def _write_subtitles(self, info_dict, filename):
3882 ''' Write subtitles to file and return list of (sub_filename, final_sub_filename); or None if error'''
3883 ret = []
3884 subtitles = info_dict.get('requested_subtitles')
3885 if not subtitles or not (self.params.get('writesubtitles') or self.params.get('writeautomaticsub')):
3886 # subtitles download errors are already managed as troubles in relevant IE
3887 # that way it will silently go on when used with unsupporting IE
3888 return ret
3889
3890 sub_filename_base = self.prepare_filename(info_dict, 'subtitle')
3891 if not sub_filename_base:
3892 self.to_screen('[info] Skipping writing video subtitles')
3893 return ret
3894 for sub_lang, sub_info in subtitles.items():
3895 sub_format = sub_info['ext']
3896 sub_filename = subtitles_filename(filename, sub_lang, sub_format, info_dict.get('ext'))
3897 sub_filename_final = subtitles_filename(sub_filename_base, sub_lang, sub_format, info_dict.get('ext'))
3898 existing_sub = self.existing_file((sub_filename_final, sub_filename))
3899 if existing_sub:
3900 self.to_screen(f'[info] Video subtitle {sub_lang}.{sub_format} is already present')
3901 sub_info['filepath'] = existing_sub
3902 ret.append((existing_sub, sub_filename_final))
3903 continue
3904
3905 self.to_screen(f'[info] Writing video subtitles to: {sub_filename}')
3906 if sub_info.get('data') is not None:
3907 try:
3908 # Use newline='' to prevent conversion of newline characters
3909 # See https://github.com/ytdl-org/youtube-dl/issues/10268
3910 with open(sub_filename, 'w', encoding='utf-8', newline='') as subfile:
3911 subfile.write(sub_info['data'])
3912 sub_info['filepath'] = sub_filename
3913 ret.append((sub_filename, sub_filename_final))
3914 continue
3915 except OSError:
3916 self.report_error(f'Cannot write video subtitles file {sub_filename}')
3917 return None
3918
3919 try:
3920 sub_copy = sub_info.copy()
3921 sub_copy.setdefault('http_headers', info_dict.get('http_headers'))
3922 self.dl(sub_filename, sub_copy, subtitle=True)
3923 sub_info['filepath'] = sub_filename
3924 ret.append((sub_filename, sub_filename_final))
3925 except (DownloadError, ExtractorError, IOError, OSError, ValueError) + network_exceptions as err:
3926 msg = f'Unable to download video subtitles for {sub_lang!r}: {err}'
3927 if self.params.get('ignoreerrors') is not True: # False or 'only_download'
3928 if not self.params.get('ignoreerrors'):
3929 self.report_error(msg)
3930 raise DownloadError(msg)
3931 self.report_warning(msg)
3932 return ret
3933
3934 def _write_thumbnails(self, label, info_dict, filename, thumb_filename_base=None):
3935 ''' Write thumbnails to file and return list of (thumb_filename, final_thumb_filename) '''
3936 write_all = self.params.get('write_all_thumbnails', False)
3937 thumbnails, ret = [], []
3938 if write_all or self.params.get('writethumbnail', False):
3939 thumbnails = info_dict.get('thumbnails') or []
3940 multiple = write_all and len(thumbnails) > 1
3941
3942 if thumb_filename_base is None:
3943 thumb_filename_base = filename
3944 if thumbnails and not thumb_filename_base:
3945 self.write_debug(f'Skipping writing {label} thumbnail')
3946 return ret
3947
3948 for idx, t in list(enumerate(thumbnails))[::-1]:
3949 thumb_ext = (f'{t["id"]}.' if multiple else '') + determine_ext(t['url'], 'jpg')
3950 thumb_display_id = f'{label} thumbnail {t["id"]}'
3951 thumb_filename = replace_extension(filename, thumb_ext, info_dict.get('ext'))
3952 thumb_filename_final = replace_extension(thumb_filename_base, thumb_ext, info_dict.get('ext'))
3953
3954 existing_thumb = self.existing_file((thumb_filename_final, thumb_filename))
3955 if existing_thumb:
3956 self.to_screen('[info] %s is already present' % (
3957 thumb_display_id if multiple else f'{label} thumbnail').capitalize())
3958 t['filepath'] = existing_thumb
3959 ret.append((existing_thumb, thumb_filename_final))
3960 else:
3961 self.to_screen(f'[info] Downloading {thumb_display_id} ...')
3962 try:
3963 uf = self.urlopen(sanitized_Request(t['url'], headers=t.get('http_headers', {})))
3964 self.to_screen(f'[info] Writing {thumb_display_id} to: {thumb_filename}')
3965 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
3966 shutil.copyfileobj(uf, thumbf)
3967 ret.append((thumb_filename, thumb_filename_final))
3968 t['filepath'] = thumb_filename
3969 except network_exceptions as err:
3970 thumbnails.pop(idx)
3971 self.report_warning(f'Unable to download {thumb_display_id}: {err}')
3972 if ret and not write_all:
3973 break
3974 return ret