]> jfr.im git - yt-dlp.git/blob - yt_dlp/YoutubeDL.py
[cleanup] Misc cleanup
[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 HAS_LEGACY as compat_has_legacy
28 from .compat import compat_os_name, compat_shlex_quote
29 from .cookies import load_cookies
30 from .downloader import FFmpegFD, get_suitable_downloader, shorten_protocol_name
31 from .downloader.rtmp import rtmpdump_version
32 from .extractor import gen_extractor_classes, get_info_extractor
33 from .extractor.openload import PhantomJSwrapper
34 from .minicurses import format_text
35 from .postprocessor import _PLUGIN_CLASSES as plugin_postprocessors
36 from .postprocessor import (
37 EmbedThumbnailPP,
38 FFmpegFixupDuplicateMoovPP,
39 FFmpegFixupDurationPP,
40 FFmpegFixupM3u8PP,
41 FFmpegFixupM4aPP,
42 FFmpegFixupStretchedPP,
43 FFmpegFixupTimestampPP,
44 FFmpegMergerPP,
45 FFmpegPostProcessor,
46 MoveFilesAfterDownloadPP,
47 get_postprocessor,
48 )
49 from .update import detect_variant
50 from .utils import (
51 DEFAULT_OUTTMPL,
52 IDENTITY,
53 LINK_TEMPLATES,
54 NO_DEFAULT,
55 NUMBER_RE,
56 OUTTMPL_TYPES,
57 POSTPROCESS_WHEN,
58 STR_FORMAT_RE_TMPL,
59 STR_FORMAT_TYPES,
60 ContentTooShortError,
61 DateRange,
62 DownloadCancelled,
63 DownloadError,
64 EntryNotInPlaylist,
65 ExistingVideoReached,
66 ExtractorError,
67 GeoRestrictedError,
68 HEADRequest,
69 ISO3166Utils,
70 LazyList,
71 MaxDownloadsReached,
72 Namespace,
73 PagedList,
74 PerRequestProxyHandler,
75 PlaylistEntries,
76 Popen,
77 PostProcessingError,
78 ReExtractInfo,
79 RejectedVideoReached,
80 SameFileError,
81 UnavailableVideoError,
82 YoutubeDLCookieProcessor,
83 YoutubeDLHandler,
84 YoutubeDLRedirectHandler,
85 age_restricted,
86 args_to_str,
87 date_from_str,
88 determine_ext,
89 determine_protocol,
90 encode_compat_str,
91 encodeFilename,
92 error_to_compat_str,
93 expand_path,
94 filter_dict,
95 float_or_none,
96 format_bytes,
97 format_decimal_suffix,
98 format_field,
99 formatSeconds,
100 get_domain,
101 int_or_none,
102 iri_to_uri,
103 join_nonempty,
104 locked_file,
105 make_dir,
106 make_HTTPS_handler,
107 merge_headers,
108 network_exceptions,
109 number_of_digits,
110 orderedSet,
111 parse_filesize,
112 preferredencoding,
113 prepend_extension,
114 register_socks_protocols,
115 remove_terminal_sequences,
116 render_table,
117 replace_extension,
118 sanitize_filename,
119 sanitize_path,
120 sanitize_url,
121 sanitized_Request,
122 std_headers,
123 str_or_none,
124 strftime_or_none,
125 subtitles_filename,
126 supports_terminal_sequences,
127 system_identifier,
128 timetuple_from_msec,
129 to_high_limit_path,
130 traverse_obj,
131 try_get,
132 url_basename,
133 variadic,
134 version_tuple,
135 windows_enable_vt_mode,
136 write_json_file,
137 write_string,
138 )
139 from .version import RELEASE_GIT_HEAD, __version__
140
141 if compat_os_name == 'nt':
142 import ctypes
143
144
145 class YoutubeDL:
146 """YoutubeDL class.
147
148 YoutubeDL objects are the ones responsible of downloading the
149 actual video file and writing it to disk if the user has requested
150 it, among some other tasks. In most cases there should be one per
151 program. As, given a video URL, the downloader doesn't know how to
152 extract all the needed information, task that InfoExtractors do, it
153 has to pass the URL to one of them.
154
155 For this, YoutubeDL objects have a method that allows
156 InfoExtractors to be registered in a given order. When it is passed
157 a URL, the YoutubeDL object handles it to the first InfoExtractor it
158 finds that reports being able to handle it. The InfoExtractor extracts
159 all the information about the video or videos the URL refers to, and
160 YoutubeDL process the extracted information, possibly using a File
161 Downloader to download the video.
162
163 YoutubeDL objects accept a lot of parameters. In order not to saturate
164 the object constructor with arguments, it receives a dictionary of
165 options instead. These options are available through the params
166 attribute for the InfoExtractors to use. The YoutubeDL also
167 registers itself as the downloader in charge for the InfoExtractors
168 that are added to it, so this is a "mutual registration".
169
170 Available options:
171
172 username: Username for authentication purposes.
173 password: Password for authentication purposes.
174 videopassword: Password for accessing a video.
175 ap_mso: Adobe Pass multiple-system operator identifier.
176 ap_username: Multiple-system operator account username.
177 ap_password: Multiple-system operator account password.
178 usenetrc: Use netrc for authentication instead.
179 verbose: Print additional info to stdout.
180 quiet: Do not print messages to stdout.
181 no_warnings: Do not print out anything for warnings.
182 forceprint: A dict with keys WHEN mapped to a list of templates to
183 print to stdout. The allowed keys are video or any of the
184 items in utils.POSTPROCESS_WHEN.
185 For compatibility, a single list is also accepted
186 print_to_file: A dict with keys WHEN (same as forceprint) mapped to
187 a list of tuples with (template, filename)
188 forcejson: Force printing info_dict as JSON.
189 dump_single_json: Force printing the info_dict of the whole playlist
190 (or video) as a single JSON line.
191 force_write_download_archive: Force writing download archive regardless
192 of 'skip_download' or 'simulate'.
193 simulate: Do not download the video files. If unset (or None),
194 simulate only if listsubtitles, listformats or list_thumbnails is used
195 format: Video format code. see "FORMAT SELECTION" for more details.
196 You can also pass a function. The function takes 'ctx' as
197 argument and returns the formats to download.
198 See "build_format_selector" for an implementation
199 allow_unplayable_formats: Allow unplayable formats to be extracted and downloaded.
200 ignore_no_formats_error: Ignore "No video formats" error. Usefull for
201 extracting metadata even if the video is not actually
202 available for download (experimental)
203 format_sort: A list of fields by which to sort the video formats.
204 See "Sorting Formats" for more details.
205 format_sort_force: Force the given format_sort. see "Sorting Formats"
206 for more details.
207 prefer_free_formats: Whether to prefer video formats with free containers
208 over non-free ones of same quality.
209 allow_multiple_video_streams: Allow multiple video streams to be merged
210 into a single file
211 allow_multiple_audio_streams: Allow multiple audio streams to be merged
212 into a single file
213 check_formats Whether to test if the formats are downloadable.
214 Can be True (check all), False (check none),
215 'selected' (check selected formats),
216 or None (check only if requested by extractor)
217 paths: Dictionary of output paths. The allowed keys are 'home'
218 'temp' and the keys of OUTTMPL_TYPES (in utils.py)
219 outtmpl: Dictionary of templates for output names. Allowed keys
220 are 'default' and the keys of OUTTMPL_TYPES (in utils.py).
221 For compatibility with youtube-dl, a single string can also be used
222 outtmpl_na_placeholder: Placeholder for unavailable meta fields.
223 restrictfilenames: Do not allow "&" and spaces in file names
224 trim_file_name: Limit length of filename (extension excluded)
225 windowsfilenames: Force the filenames to be windows compatible
226 ignoreerrors: Do not stop on download/postprocessing errors.
227 Can be 'only_download' to ignore only download errors.
228 Default is 'only_download' for CLI, but False for API
229 skip_playlist_after_errors: Number of allowed failures until the rest of
230 the playlist is skipped
231 force_generic_extractor: Force downloader to use the generic extractor
232 overwrites: Overwrite all video and metadata files if True,
233 overwrite only non-video files if None
234 and don't overwrite any file if False
235 For compatibility with youtube-dl,
236 "nooverwrites" may also be used instead
237 playlist_items: Specific indices of playlist to download.
238 playlistrandom: Download playlist items in random order.
239 lazy_playlist: Process playlist entries as they are received.
240 matchtitle: Download only matching titles.
241 rejecttitle: Reject downloads for matching titles.
242 logger: Log messages to a logging.Logger instance.
243 logtostderr: Log messages to stderr instead of stdout.
244 consoletitle: Display progress in console window's titlebar.
245 writedescription: Write the video description to a .description file
246 writeinfojson: Write the video description to a .info.json file
247 clean_infojson: Remove private fields from the infojson
248 getcomments: Extract video comments. This will not be written to disk
249 unless writeinfojson is also given
250 writeannotations: Write the video annotations to a .annotations.xml file
251 writethumbnail: Write the thumbnail image to a file
252 allow_playlist_files: Whether to write playlists' description, infojson etc
253 also to disk when using the 'write*' options
254 write_all_thumbnails: Write all thumbnail formats to files
255 writelink: Write an internet shortcut file, depending on the
256 current platform (.url/.webloc/.desktop)
257 writeurllink: Write a Windows internet shortcut file (.url)
258 writewebloclink: Write a macOS internet shortcut file (.webloc)
259 writedesktoplink: Write a Linux internet shortcut file (.desktop)
260 writesubtitles: Write the video subtitles to a file
261 writeautomaticsub: Write the automatically generated subtitles to a file
262 listsubtitles: Lists all available subtitles for the video
263 subtitlesformat: The format code for subtitles
264 subtitleslangs: List of languages of the subtitles to download (can be regex).
265 The list may contain "all" to refer to all the available
266 subtitles. The language can be prefixed with a "-" to
267 exclude it from the requested languages. Eg: ['all', '-live_chat']
268 keepvideo: Keep the video file after post-processing
269 daterange: A DateRange object, download only if the upload_date is in the range.
270 skip_download: Skip the actual download of the video file
271 cachedir: Location of the cache files in the filesystem.
272 False to disable filesystem cache.
273 noplaylist: Download single video instead of a playlist if in doubt.
274 age_limit: An integer representing the user's age in years.
275 Unsuitable videos for the given age are skipped.
276 min_views: An integer representing the minimum view count the video
277 must have in order to not be skipped.
278 Videos without view count information are always
279 downloaded. None for no limit.
280 max_views: An integer representing the maximum view count.
281 Videos that are more popular than that are not
282 downloaded.
283 Videos without view count information are always
284 downloaded. None for no limit.
285 download_archive: File name of a file where all downloads are recorded.
286 Videos already present in the file are not downloaded
287 again.
288 break_on_existing: Stop the download process after attempting to download a
289 file that is in the archive.
290 break_on_reject: Stop the download process when encountering a video that
291 has been filtered out.
292 break_per_url: Whether break_on_reject and break_on_existing
293 should act on each input URL as opposed to for the entire queue
294 cookiefile: File name or text stream from where cookies should be read and dumped to
295 cookiesfrombrowser: A tuple containing the name of the browser, the profile
296 name/pathfrom where cookies are loaded, and the name of the
297 keyring. Eg: ('chrome', ) or ('vivaldi', 'default', 'BASICTEXT')
298 legacyserverconnect: Explicitly allow HTTPS connection to servers that do not
299 support RFC 5746 secure renegotiation
300 nocheckcertificate: Do not verify SSL certificates
301 client_certificate: Path to client certificate file in PEM format. May include the private key
302 client_certificate_key: Path to private key file for client certificate
303 client_certificate_password: Password for client certificate private key, if encrypted.
304 If not provided and the key is encrypted, yt-dlp will ask interactively
305 prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
306 At the moment, this is only supported by YouTube.
307 http_headers: A dictionary of custom headers to be used for all requests
308 proxy: URL of the proxy server to use
309 geo_verification_proxy: URL of the proxy to use for IP address verification
310 on geo-restricted sites.
311 socket_timeout: Time to wait for unresponsive hosts, in seconds
312 bidi_workaround: Work around buggy terminals without bidirectional text
313 support, using fridibi
314 debug_printtraffic:Print out sent and received HTTP traffic
315 default_search: Prepend this string if an input url is not valid.
316 'auto' for elaborate guessing
317 encoding: Use this encoding instead of the system-specified.
318 extract_flat: Do not resolve URLs, return the immediate result.
319 Pass in 'in_playlist' to only show this behavior for
320 playlist items.
321 wait_for_video: If given, wait for scheduled streams to become available.
322 The value should be a tuple containing the range
323 (min_secs, max_secs) to wait between retries
324 postprocessors: A list of dictionaries, each with an entry
325 * key: The name of the postprocessor. See
326 yt_dlp/postprocessor/__init__.py for a list.
327 * when: When to run the postprocessor. Allowed values are
328 the entries of utils.POSTPROCESS_WHEN
329 Assumed to be 'post_process' if not given
330 progress_hooks: A list of functions that get called on download
331 progress, with a dictionary with the entries
332 * status: One of "downloading", "error", or "finished".
333 Check this first and ignore unknown values.
334 * info_dict: The extracted info_dict
335
336 If status is one of "downloading", or "finished", the
337 following properties may also be present:
338 * filename: The final filename (always present)
339 * tmpfilename: The filename we're currently writing to
340 * downloaded_bytes: Bytes on disk
341 * total_bytes: Size of the whole file, None if unknown
342 * total_bytes_estimate: Guess of the eventual file size,
343 None if unavailable.
344 * elapsed: The number of seconds since download started.
345 * eta: The estimated time in seconds, None if unknown
346 * speed: The download speed in bytes/second, None if
347 unknown
348 * fragment_index: The counter of the currently
349 downloaded video fragment.
350 * fragment_count: The number of fragments (= individual
351 files that will be merged)
352
353 Progress hooks are guaranteed to be called at least once
354 (with status "finished") if the download is successful.
355 postprocessor_hooks: A list of functions that get called on postprocessing
356 progress, with a dictionary with the entries
357 * status: One of "started", "processing", or "finished".
358 Check this first and ignore unknown values.
359 * postprocessor: Name of the postprocessor
360 * info_dict: The extracted info_dict
361
362 Progress hooks are guaranteed to be called at least twice
363 (with status "started" and "finished") if the processing is successful.
364 merge_output_format: Extension to use when merging formats.
365 final_ext: Expected final extension; used to detect when the file was
366 already downloaded and converted
367 fixup: Automatically correct known faults of the file.
368 One of:
369 - "never": do nothing
370 - "warn": only emit a warning
371 - "detect_or_warn": check whether we can do anything
372 about it, warn otherwise (default)
373 source_address: Client-side IP address to bind to.
374 sleep_interval_requests: Number of seconds to sleep between requests
375 during extraction
376 sleep_interval: Number of seconds to sleep before each download when
377 used alone or a lower bound of a range for randomized
378 sleep before each download (minimum possible number
379 of seconds to sleep) when used along with
380 max_sleep_interval.
381 max_sleep_interval:Upper bound of a range for randomized sleep before each
382 download (maximum possible number of seconds to sleep).
383 Must only be used along with sleep_interval.
384 Actual sleep time will be a random float from range
385 [sleep_interval; max_sleep_interval].
386 sleep_interval_subtitles: Number of seconds to sleep before each subtitle download
387 listformats: Print an overview of available video formats and exit.
388 list_thumbnails: Print a table of all thumbnails and exit.
389 match_filter: A function that gets called for every video with the signature
390 (info_dict, *, incomplete: bool) -> Optional[str]
391 For backward compatibility with youtube-dl, the signature
392 (info_dict) -> Optional[str] is also allowed.
393 - If it returns a message, the video is ignored.
394 - If it returns None, the video is downloaded.
395 - If it returns utils.NO_DEFAULT, the user is interactively
396 asked whether to download the video.
397 match_filter_func in utils.py is one example for this.
398 no_color: Do not emit color codes in output.
399 geo_bypass: Bypass geographic restriction via faking X-Forwarded-For
400 HTTP header
401 geo_bypass_country:
402 Two-letter ISO 3166-2 country code that will be used for
403 explicit geographic restriction bypassing via faking
404 X-Forwarded-For HTTP header
405 geo_bypass_ip_block:
406 IP range in CIDR notation that will be used similarly to
407 geo_bypass_country
408 external_downloader: A dictionary of protocol keys and the executable of the
409 external downloader to use for it. The allowed protocols
410 are default|http|ftp|m3u8|dash|rtsp|rtmp|mms.
411 Set the value to 'native' to use the native downloader
412 compat_opts: Compatibility options. See "Differences in default behavior".
413 The following options do not work when used through the API:
414 filename, abort-on-error, multistreams, no-live-chat, format-sort
415 no-clean-infojson, no-playlist-metafiles, no-keep-subs, no-attach-info-json.
416 Refer __init__.py for their implementation
417 progress_template: Dictionary of templates for progress outputs.
418 Allowed keys are 'download', 'postprocess',
419 'download-title' (console title) and 'postprocess-title'.
420 The template is mapped on a dictionary with keys 'progress' and 'info'
421 retry_sleep_functions: Dictionary of functions that takes the number of attempts
422 as argument and returns the time to sleep in seconds.
423 Allowed keys are 'http', 'fragment', 'file_access'
424 download_ranges: A function that gets called for every video with the signature
425 (info_dict, *, ydl) -> Iterable[Section].
426 Only the returned sections will be downloaded. Each Section contains:
427 * start_time: Start time of the section in seconds
428 * end_time: End time of the section in seconds
429 * title: Section title (Optional)
430 * index: Section number (Optional)
431
432 The following parameters are not used by YoutubeDL itself, they are used by
433 the downloader (see yt_dlp/downloader/common.py):
434 nopart, updatetime, buffersize, ratelimit, throttledratelimit, min_filesize,
435 max_filesize, test, noresizebuffer, retries, file_access_retries, fragment_retries,
436 continuedl, noprogress, xattr_set_filesize, hls_use_mpegts, http_chunk_size,
437 external_downloader_args, concurrent_fragment_downloads.
438
439 The following options are used by the post processors:
440 ffmpeg_location: Location of the ffmpeg/avconv binary; either the path
441 to the binary or its containing directory.
442 postprocessor_args: A dictionary of postprocessor/executable keys (in lower case)
443 and a list of additional command-line arguments for the
444 postprocessor/executable. The dict can also have "PP+EXE" keys
445 which are used when the given exe is used by the given PP.
446 Use 'default' as the name for arguments to passed to all PP
447 For compatibility with youtube-dl, a single list of args
448 can also be used
449
450 The following options are used by the extractors:
451 extractor_retries: Number of times to retry for known errors
452 dynamic_mpd: Whether to process dynamic DASH manifests (default: True)
453 hls_split_discontinuity: Split HLS playlists to different formats at
454 discontinuities such as ad breaks (default: False)
455 extractor_args: A dictionary of arguments to be passed to the extractors.
456 See "EXTRACTOR ARGUMENTS" for details.
457 Eg: {'youtube': {'skip': ['dash', 'hls']}}
458 mark_watched: Mark videos watched (even with --simulate). Only for YouTube
459
460 The following options are deprecated and may be removed in the future:
461
462 playliststart: - Use playlist_items
463 Playlist item to start at.
464 playlistend: - Use playlist_items
465 Playlist item to end at.
466 playlistreverse: - Use playlist_items
467 Download playlist items in reverse order.
468 forceurl: - Use forceprint
469 Force printing final URL.
470 forcetitle: - Use forceprint
471 Force printing title.
472 forceid: - Use forceprint
473 Force printing ID.
474 forcethumbnail: - Use forceprint
475 Force printing thumbnail URL.
476 forcedescription: - Use forceprint
477 Force printing description.
478 forcefilename: - Use forceprint
479 Force printing final filename.
480 forceduration: - Use forceprint
481 Force printing duration.
482 allsubtitles: - Use subtitleslangs = ['all']
483 Downloads all the subtitles of the video
484 (requires writesubtitles or writeautomaticsub)
485 include_ads: - Doesn't work
486 Download ads as well
487 call_home: - Not implemented
488 Boolean, true iff we are allowed to contact the
489 yt-dlp servers for debugging.
490 post_hooks: - Register a custom postprocessor
491 A list of functions that get called as the final step
492 for each video file, after all postprocessors have been
493 called. The filename will be passed as the only argument.
494 hls_prefer_native: - Use external_downloader = {'m3u8': 'native'} or {'m3u8': 'ffmpeg'}.
495 Use the native HLS downloader instead of ffmpeg/avconv
496 if True, otherwise use ffmpeg/avconv if False, otherwise
497 use downloader suggested by extractor if None.
498 prefer_ffmpeg: - avconv support is deprecated
499 If False, use avconv instead of ffmpeg if both are available,
500 otherwise prefer ffmpeg.
501 youtube_include_dash_manifest: - Use extractor_args
502 If True (default), DASH manifests and related
503 data will be downloaded and processed by extractor.
504 You can reduce network I/O by disabling it if you don't
505 care about DASH. (only for youtube)
506 youtube_include_hls_manifest: - Use extractor_args
507 If True (default), HLS manifests and related
508 data will be downloaded and processed by extractor.
509 You can reduce network I/O by disabling it if you don't
510 care about HLS. (only for youtube)
511 """
512
513 _NUMERIC_FIELDS = {
514 'width', 'height', 'tbr', 'abr', 'asr', 'vbr', 'fps', 'filesize', 'filesize_approx',
515 'timestamp', 'release_timestamp',
516 'duration', 'view_count', 'like_count', 'dislike_count', 'repost_count',
517 'average_rating', 'comment_count', 'age_limit',
518 'start_time', 'end_time',
519 'chapter_number', 'season_number', 'episode_number',
520 'track_number', 'disc_number', 'release_year',
521 }
522
523 _format_fields = {
524 # NB: Keep in sync with the docstring of extractor/common.py
525 'url', 'manifest_url', 'manifest_stream_number', 'ext', 'format', 'format_id', 'format_note',
526 'width', 'height', 'resolution', 'dynamic_range', 'tbr', 'abr', 'acodec', 'asr',
527 'vbr', 'fps', 'vcodec', 'container', 'filesize', 'filesize_approx',
528 'player_url', 'protocol', 'fragment_base_url', 'fragments', 'is_from_start',
529 'preference', 'language', 'language_preference', 'quality', 'source_preference',
530 'http_headers', 'stretched_ratio', 'no_resume', 'has_drm', 'downloader_options',
531 'page_url', 'app', 'play_path', 'tc_url', 'flash_version', 'rtmp_live', 'rtmp_conn', 'rtmp_protocol', 'rtmp_real_time'
532 }
533 _format_selection_exts = {
534 'audio': {'m4a', 'mp3', 'ogg', 'aac'},
535 'video': {'mp4', 'flv', 'webm', '3gp'},
536 'storyboards': {'mhtml'},
537 }
538
539 def __init__(self, params=None, auto_init=True):
540 """Create a FileDownloader object with the given options.
541 @param auto_init Whether to load the default extractors and print header (if verbose).
542 Set to 'no_verbose_header' to not print the header
543 """
544 if params is None:
545 params = {}
546 self.params = params
547 self._ies = {}
548 self._ies_instances = {}
549 self._pps = {k: [] for k in POSTPROCESS_WHEN}
550 self._printed_messages = set()
551 self._first_webpage_request = True
552 self._post_hooks = []
553 self._progress_hooks = []
554 self._postprocessor_hooks = []
555 self._download_retcode = 0
556 self._num_downloads = 0
557 self._num_videos = 0
558 self._playlist_level = 0
559 self._playlist_urls = set()
560 self.cache = Cache(self)
561
562 windows_enable_vt_mode()
563 stdout = sys.stderr if self.params.get('logtostderr') else sys.stdout
564 self._out_files = Namespace(
565 out=stdout,
566 error=sys.stderr,
567 screen=sys.stderr if self.params.get('quiet') else stdout,
568 console=None if compat_os_name == 'nt' else next(
569 filter(supports_terminal_sequences, (sys.stderr, sys.stdout)), None)
570 )
571 self._allow_colors = Namespace(**{
572 type_: not self.params.get('no_color') and supports_terminal_sequences(stream)
573 for type_, stream in self._out_files.items_ if type_ != 'console'
574 })
575
576 MIN_SUPPORTED, MIN_RECOMMENDED = (3, 6), (3, 7)
577 current_version = sys.version_info[:2]
578 if current_version < MIN_RECOMMENDED:
579 msg = 'Support for Python version %d.%d has been deprecated and will break in future versions of yt-dlp'
580 if current_version < MIN_SUPPORTED:
581 msg = 'Python version %d.%d is no longer supported'
582 self.deprecation_warning(
583 f'{msg}! Please update to Python %d.%d or above' % (*current_version, *MIN_RECOMMENDED))
584
585 if self.params.get('allow_unplayable_formats'):
586 self.report_warning(
587 f'You have asked for {self._format_err("UNPLAYABLE", self.Styles.EMPHASIS)} formats to be listed/downloaded. '
588 'This is a developer option intended for debugging. \n'
589 ' If you experience any issues while using this option, '
590 f'{self._format_err("DO NOT", self.Styles.ERROR)} open a bug report')
591
592 def check_deprecated(param, option, suggestion):
593 if self.params.get(param) is not None:
594 self.report_warning(f'{option} is deprecated. Use {suggestion} instead')
595 return True
596 return False
597
598 if check_deprecated('cn_verification_proxy', '--cn-verification-proxy', '--geo-verification-proxy'):
599 if self.params.get('geo_verification_proxy') is None:
600 self.params['geo_verification_proxy'] = self.params['cn_verification_proxy']
601
602 check_deprecated('autonumber', '--auto-number', '-o "%(autonumber)s-%(title)s.%(ext)s"')
603 check_deprecated('usetitle', '--title', '-o "%(title)s-%(id)s.%(ext)s"')
604 check_deprecated('useid', '--id', '-o "%(id)s.%(ext)s"')
605
606 for msg in self.params.get('_warnings', []):
607 self.report_warning(msg)
608 for msg in self.params.get('_deprecation_warnings', []):
609 self.deprecation_warning(msg)
610
611 self.params['compat_opts'] = set(self.params.get('compat_opts', ()))
612 if not compat_has_legacy:
613 self.params['compat_opts'].add('no-compat-legacy')
614 if 'list-formats' in self.params['compat_opts']:
615 self.params['listformats_table'] = False
616
617 if 'overwrites' not in self.params and self.params.get('nooverwrites') is not None:
618 # nooverwrites was unnecessarily changed to overwrites
619 # in 0c3d0f51778b153f65c21906031c2e091fcfb641
620 # This ensures compatibility with both keys
621 self.params['overwrites'] = not self.params['nooverwrites']
622 elif self.params.get('overwrites') is None:
623 self.params.pop('overwrites', None)
624 else:
625 self.params['nooverwrites'] = not self.params['overwrites']
626
627 self.params.setdefault('forceprint', {})
628 self.params.setdefault('print_to_file', {})
629
630 # Compatibility with older syntax
631 if not isinstance(params['forceprint'], dict):
632 self.params['forceprint'] = {'video': params['forceprint']}
633
634 if self.params.get('bidi_workaround', False):
635 try:
636 import pty
637 master, slave = pty.openpty()
638 width = shutil.get_terminal_size().columns
639 width_args = [] if width is None else ['-w', str(width)]
640 sp_kwargs = {'stdin': subprocess.PIPE, 'stdout': slave, 'stderr': self._out_files.error}
641 try:
642 self._output_process = Popen(['bidiv'] + width_args, **sp_kwargs)
643 except OSError:
644 self._output_process = Popen(['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
645 self._output_channel = os.fdopen(master, 'rb')
646 except OSError as ose:
647 if ose.errno == errno.ENOENT:
648 self.report_warning(
649 'Could not find fribidi executable, ignoring --bidi-workaround. '
650 'Make sure that fribidi is an executable file in one of the directories in your $PATH.')
651 else:
652 raise
653
654 if auto_init:
655 if auto_init != 'no_verbose_header':
656 self.print_debug_header()
657 self.add_default_info_extractors()
658
659 if (sys.platform != 'win32'
660 and sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968']
661 and not self.params.get('restrictfilenames', False)):
662 # Unicode filesystem API will throw errors (#1474, #13027)
663 self.report_warning(
664 'Assuming --restrict-filenames since file system encoding '
665 'cannot encode all characters. '
666 'Set the LC_ALL environment variable to fix this.')
667 self.params['restrictfilenames'] = True
668
669 self._parse_outtmpl()
670
671 # Creating format selector here allows us to catch syntax errors before the extraction
672 self.format_selector = (
673 self.params.get('format') if self.params.get('format') in (None, '-')
674 else self.params['format'] if callable(self.params['format'])
675 else self.build_format_selector(self.params['format']))
676
677 # Set http_headers defaults according to std_headers
678 self.params['http_headers'] = merge_headers(std_headers, self.params.get('http_headers', {}))
679
680 hooks = {
681 'post_hooks': self.add_post_hook,
682 'progress_hooks': self.add_progress_hook,
683 'postprocessor_hooks': self.add_postprocessor_hook,
684 }
685 for opt, fn in hooks.items():
686 for ph in self.params.get(opt, []):
687 fn(ph)
688
689 for pp_def_raw in self.params.get('postprocessors', []):
690 pp_def = dict(pp_def_raw)
691 when = pp_def.pop('when', 'post_process')
692 self.add_post_processor(
693 get_postprocessor(pp_def.pop('key'))(self, **pp_def),
694 when=when)
695
696 self._setup_opener()
697 register_socks_protocols()
698
699 def preload_download_archive(fn):
700 """Preload the archive, if any is specified"""
701 if fn is None:
702 return False
703 self.write_debug(f'Loading archive file {fn!r}')
704 try:
705 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
706 for line in archive_file:
707 self.archive.add(line.strip())
708 except OSError as ioe:
709 if ioe.errno != errno.ENOENT:
710 raise
711 return False
712 return True
713
714 self.archive = set()
715 preload_download_archive(self.params.get('download_archive'))
716
717 def warn_if_short_id(self, argv):
718 # short YouTube ID starting with dash?
719 idxs = [
720 i for i, a in enumerate(argv)
721 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
722 if idxs:
723 correct_argv = (
724 ['yt-dlp']
725 + [a for i, a in enumerate(argv) if i not in idxs]
726 + ['--'] + [argv[i] for i in idxs]
727 )
728 self.report_warning(
729 'Long argument string detected. '
730 'Use -- to separate parameters and URLs, like this:\n%s' %
731 args_to_str(correct_argv))
732
733 def add_info_extractor(self, ie):
734 """Add an InfoExtractor object to the end of the list."""
735 ie_key = ie.ie_key()
736 self._ies[ie_key] = ie
737 if not isinstance(ie, type):
738 self._ies_instances[ie_key] = ie
739 ie.set_downloader(self)
740
741 def _get_info_extractor_class(self, ie_key):
742 ie = self._ies.get(ie_key)
743 if ie is None:
744 ie = get_info_extractor(ie_key)
745 self.add_info_extractor(ie)
746 return ie
747
748 def get_info_extractor(self, ie_key):
749 """
750 Get an instance of an IE with name ie_key, it will try to get one from
751 the _ies list, if there's no instance it will create a new one and add
752 it to the extractor list.
753 """
754 ie = self._ies_instances.get(ie_key)
755 if ie is None:
756 ie = get_info_extractor(ie_key)()
757 self.add_info_extractor(ie)
758 return ie
759
760 def add_default_info_extractors(self):
761 """
762 Add the InfoExtractors returned by gen_extractors to the end of the list
763 """
764 for ie in gen_extractor_classes():
765 self.add_info_extractor(ie)
766
767 def add_post_processor(self, pp, when='post_process'):
768 """Add a PostProcessor object to the end of the chain."""
769 assert when in POSTPROCESS_WHEN, f'Invalid when={when}'
770 self._pps[when].append(pp)
771 pp.set_downloader(self)
772
773 def add_post_hook(self, ph):
774 """Add the post hook"""
775 self._post_hooks.append(ph)
776
777 def add_progress_hook(self, ph):
778 """Add the download progress hook"""
779 self._progress_hooks.append(ph)
780
781 def add_postprocessor_hook(self, ph):
782 """Add the postprocessing progress hook"""
783 self._postprocessor_hooks.append(ph)
784 for pps in self._pps.values():
785 for pp in pps:
786 pp.add_progress_hook(ph)
787
788 def _bidi_workaround(self, message):
789 if not hasattr(self, '_output_channel'):
790 return message
791
792 assert hasattr(self, '_output_process')
793 assert isinstance(message, str)
794 line_count = message.count('\n') + 1
795 self._output_process.stdin.write((message + '\n').encode())
796 self._output_process.stdin.flush()
797 res = ''.join(self._output_channel.readline().decode()
798 for _ in range(line_count))
799 return res[:-len('\n')]
800
801 def _write_string(self, message, out=None, only_once=False):
802 if only_once:
803 if message in self._printed_messages:
804 return
805 self._printed_messages.add(message)
806 write_string(message, out=out, encoding=self.params.get('encoding'))
807
808 def to_stdout(self, message, skip_eol=False, quiet=None):
809 """Print message to stdout"""
810 if quiet is not None:
811 self.deprecation_warning('"YoutubeDL.to_stdout" no longer accepts the argument quiet. Use "YoutubeDL.to_screen" instead')
812 if skip_eol is not False:
813 self.deprecation_warning('"YoutubeDL.to_stdout" no longer accepts the argument skip_eol. Use "YoutubeDL.to_screen" instead')
814 self._write_string(f'{self._bidi_workaround(message)}\n', self._out_files.out)
815
816 def to_screen(self, message, skip_eol=False, quiet=None):
817 """Print message to screen if not in quiet mode"""
818 if self.params.get('logger'):
819 self.params['logger'].debug(message)
820 return
821 if (self.params.get('quiet') if quiet is None else quiet) and not self.params.get('verbose'):
822 return
823 self._write_string(
824 '%s%s' % (self._bidi_workaround(message), ('' if skip_eol else '\n')),
825 self._out_files.screen)
826
827 def to_stderr(self, message, only_once=False):
828 """Print message to stderr"""
829 assert isinstance(message, str)
830 if self.params.get('logger'):
831 self.params['logger'].error(message)
832 else:
833 self._write_string(f'{self._bidi_workaround(message)}\n', self._out_files.error, only_once=only_once)
834
835 def _send_console_code(self, code):
836 if compat_os_name == 'nt' or not self._out_files.console:
837 return
838 self._write_string(code, self._out_files.console)
839
840 def to_console_title(self, message):
841 if not self.params.get('consoletitle', False):
842 return
843 message = remove_terminal_sequences(message)
844 if compat_os_name == 'nt':
845 if ctypes.windll.kernel32.GetConsoleWindow():
846 # c_wchar_p() might not be necessary if `message` is
847 # already of type unicode()
848 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
849 else:
850 self._send_console_code(f'\033]0;{message}\007')
851
852 def save_console_title(self):
853 if not self.params.get('consoletitle') or self.params.get('simulate'):
854 return
855 self._send_console_code('\033[22;0t') # Save the title on stack
856
857 def restore_console_title(self):
858 if not self.params.get('consoletitle') or self.params.get('simulate'):
859 return
860 self._send_console_code('\033[23;0t') # Restore the title from stack
861
862 def __enter__(self):
863 self.save_console_title()
864 return self
865
866 def __exit__(self, *args):
867 self.restore_console_title()
868
869 if self.params.get('cookiefile') is not None:
870 self.cookiejar.save(ignore_discard=True, ignore_expires=True)
871
872 def trouble(self, message=None, tb=None, is_error=True):
873 """Determine action to take when a download problem appears.
874
875 Depending on if the downloader has been configured to ignore
876 download errors or not, this method may throw an exception or
877 not when errors are found, after printing the message.
878
879 @param tb If given, is additional traceback information
880 @param is_error Whether to raise error according to ignorerrors
881 """
882 if message is not None:
883 self.to_stderr(message)
884 if self.params.get('verbose'):
885 if tb is None:
886 if sys.exc_info()[0]: # if .trouble has been called from an except block
887 tb = ''
888 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
889 tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
890 tb += encode_compat_str(traceback.format_exc())
891 else:
892 tb_data = traceback.format_list(traceback.extract_stack())
893 tb = ''.join(tb_data)
894 if tb:
895 self.to_stderr(tb)
896 if not is_error:
897 return
898 if not self.params.get('ignoreerrors'):
899 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
900 exc_info = sys.exc_info()[1].exc_info
901 else:
902 exc_info = sys.exc_info()
903 raise DownloadError(message, exc_info)
904 self._download_retcode = 1
905
906 Styles = Namespace(
907 HEADERS='yellow',
908 EMPHASIS='light blue',
909 FILENAME='green',
910 ID='green',
911 DELIM='blue',
912 ERROR='red',
913 WARNING='yellow',
914 SUPPRESS='light black',
915 )
916
917 def _format_text(self, handle, allow_colors, text, f, fallback=None, *, test_encoding=False):
918 text = str(text)
919 if test_encoding:
920 original_text = text
921 # handle.encoding can be None. See https://github.com/yt-dlp/yt-dlp/issues/2711
922 encoding = self.params.get('encoding') or getattr(handle, 'encoding', None) or 'ascii'
923 text = text.encode(encoding, 'ignore').decode(encoding)
924 if fallback is not None and text != original_text:
925 text = fallback
926 return format_text(text, f) if allow_colors else text if fallback is None else fallback
927
928 def _format_out(self, *args, **kwargs):
929 return self._format_text(self._out_files.out, self._allow_colors.out, *args, **kwargs)
930
931 def _format_screen(self, *args, **kwargs):
932 return self._format_text(self._out_files.screen, self._allow_colors.screen, *args, **kwargs)
933
934 def _format_err(self, *args, **kwargs):
935 return self._format_text(self._out_files.error, self._allow_colors.error, *args, **kwargs)
936
937 def report_warning(self, message, only_once=False):
938 '''
939 Print the message to stderr, it will be prefixed with 'WARNING:'
940 If stderr is a tty file the 'WARNING:' will be colored
941 '''
942 if self.params.get('logger') is not None:
943 self.params['logger'].warning(message)
944 else:
945 if self.params.get('no_warnings'):
946 return
947 self.to_stderr(f'{self._format_err("WARNING:", self.Styles.WARNING)} {message}', only_once)
948
949 def deprecation_warning(self, message):
950 if self.params.get('logger') is not None:
951 self.params['logger'].warning(f'DeprecationWarning: {message}')
952 else:
953 self.to_stderr(f'{self._format_err("DeprecationWarning:", self.Styles.ERROR)} {message}', True)
954
955 def report_error(self, message, *args, **kwargs):
956 '''
957 Do the same as trouble, but prefixes the message with 'ERROR:', colored
958 in red if stderr is a tty file.
959 '''
960 self.trouble(f'{self._format_err("ERROR:", self.Styles.ERROR)} {message}', *args, **kwargs)
961
962 def write_debug(self, message, only_once=False):
963 '''Log debug message or Print message to stderr'''
964 if not self.params.get('verbose', False):
965 return
966 message = f'[debug] {message}'
967 if self.params.get('logger'):
968 self.params['logger'].debug(message)
969 else:
970 self.to_stderr(message, only_once)
971
972 def report_file_already_downloaded(self, file_name):
973 """Report file has already been fully downloaded."""
974 try:
975 self.to_screen('[download] %s has already been downloaded' % file_name)
976 except UnicodeEncodeError:
977 self.to_screen('[download] The file has already been downloaded')
978
979 def report_file_delete(self, file_name):
980 """Report that existing file will be deleted."""
981 try:
982 self.to_screen('Deleting existing file %s' % file_name)
983 except UnicodeEncodeError:
984 self.to_screen('Deleting existing file')
985
986 def raise_no_formats(self, info, forced=False, *, msg=None):
987 has_drm = info.get('_has_drm')
988 ignored, expected = self.params.get('ignore_no_formats_error'), bool(msg)
989 msg = msg or has_drm and 'This video is DRM protected' or 'No video formats found!'
990 if forced or not ignored:
991 raise ExtractorError(msg, video_id=info['id'], ie=info['extractor'],
992 expected=has_drm or ignored or expected)
993 else:
994 self.report_warning(msg)
995
996 def parse_outtmpl(self):
997 self.deprecation_warning('"YoutubeDL.parse_outtmpl" is deprecated and may be removed in a future version')
998 self._parse_outtmpl()
999 return self.params['outtmpl']
1000
1001 def _parse_outtmpl(self):
1002 sanitize = IDENTITY
1003 if self.params.get('restrictfilenames'): # Remove spaces in the default template
1004 sanitize = lambda x: x.replace(' - ', ' ').replace(' ', '-')
1005
1006 outtmpl = self.params.setdefault('outtmpl', {})
1007 if not isinstance(outtmpl, dict):
1008 self.params['outtmpl'] = outtmpl = {'default': outtmpl}
1009 outtmpl.update({k: sanitize(v) for k, v in DEFAULT_OUTTMPL.items() if outtmpl.get(k) is None})
1010
1011 def get_output_path(self, dir_type='', filename=None):
1012 paths = self.params.get('paths', {})
1013 assert isinstance(paths, dict)
1014 path = os.path.join(
1015 expand_path(paths.get('home', '').strip()),
1016 expand_path(paths.get(dir_type, '').strip()) if dir_type else '',
1017 filename or '')
1018 return sanitize_path(path, force=self.params.get('windowsfilenames'))
1019
1020 @staticmethod
1021 def _outtmpl_expandpath(outtmpl):
1022 # expand_path translates '%%' into '%' and '$$' into '$'
1023 # correspondingly that is not what we want since we need to keep
1024 # '%%' intact for template dict substitution step. Working around
1025 # with boundary-alike separator hack.
1026 sep = ''.join([random.choice(ascii_letters) for _ in range(32)])
1027 outtmpl = outtmpl.replace('%%', f'%{sep}%').replace('$$', f'${sep}$')
1028
1029 # outtmpl should be expand_path'ed before template dict substitution
1030 # because meta fields may contain env variables we don't want to
1031 # be expanded. For example, for outtmpl "%(title)s.%(ext)s" and
1032 # title "Hello $PATH", we don't want `$PATH` to be expanded.
1033 return expand_path(outtmpl).replace(sep, '')
1034
1035 @staticmethod
1036 def escape_outtmpl(outtmpl):
1037 ''' Escape any remaining strings like %s, %abc% etc. '''
1038 return re.sub(
1039 STR_FORMAT_RE_TMPL.format('', '(?![%(\0])'),
1040 lambda mobj: ('' if mobj.group('has_key') else '%') + mobj.group(0),
1041 outtmpl)
1042
1043 @classmethod
1044 def validate_outtmpl(cls, outtmpl):
1045 ''' @return None or Exception object '''
1046 outtmpl = re.sub(
1047 STR_FORMAT_RE_TMPL.format('[^)]*', '[ljqBUDS]'),
1048 lambda mobj: f'{mobj.group(0)[:-1]}s',
1049 cls._outtmpl_expandpath(outtmpl))
1050 try:
1051 cls.escape_outtmpl(outtmpl) % collections.defaultdict(int)
1052 return None
1053 except ValueError as err:
1054 return err
1055
1056 @staticmethod
1057 def _copy_infodict(info_dict):
1058 info_dict = dict(info_dict)
1059 info_dict.pop('__postprocessors', None)
1060 info_dict.pop('__pending_error', None)
1061 return info_dict
1062
1063 def prepare_outtmpl(self, outtmpl, info_dict, sanitize=False):
1064 """ Make the outtmpl and info_dict suitable for substitution: ydl.escape_outtmpl(outtmpl) % info_dict
1065 @param sanitize Whether to sanitize the output as a filename.
1066 For backward compatibility, a function can also be passed
1067 """
1068
1069 info_dict.setdefault('epoch', int(time.time())) # keep epoch consistent once set
1070
1071 info_dict = self._copy_infodict(info_dict)
1072 info_dict['duration_string'] = ( # %(duration>%H-%M-%S)s is wrong if duration > 24hrs
1073 formatSeconds(info_dict['duration'], '-' if sanitize else ':')
1074 if info_dict.get('duration', None) is not None
1075 else None)
1076 info_dict['autonumber'] = int(self.params.get('autonumber_start', 1) - 1 + self._num_downloads)
1077 info_dict['video_autonumber'] = self._num_videos
1078 if info_dict.get('resolution') is None:
1079 info_dict['resolution'] = self.format_resolution(info_dict, default=None)
1080
1081 # For fields playlist_index, playlist_autonumber and autonumber convert all occurrences
1082 # of %(field)s to %(field)0Nd for backward compatibility
1083 field_size_compat_map = {
1084 'playlist_index': number_of_digits(info_dict.get('__last_playlist_index') or 0),
1085 'playlist_autonumber': number_of_digits(info_dict.get('n_entries') or 0),
1086 'autonumber': self.params.get('autonumber_size') or 5,
1087 }
1088
1089 TMPL_DICT = {}
1090 EXTERNAL_FORMAT_RE = re.compile(STR_FORMAT_RE_TMPL.format('[^)]*', f'[{STR_FORMAT_TYPES}ljqBUDS]'))
1091 MATH_FUNCTIONS = {
1092 '+': float.__add__,
1093 '-': float.__sub__,
1094 }
1095 # Field is of the form key1.key2...
1096 # where keys (except first) can be string, int or slice
1097 FIELD_RE = r'\w*(?:\.(?:\w+|{num}|{num}?(?::{num}?){{1,2}}))*'.format(num=r'(?:-?\d+)')
1098 MATH_FIELD_RE = rf'(?:{FIELD_RE}|-?{NUMBER_RE})'
1099 MATH_OPERATORS_RE = r'(?:%s)' % '|'.join(map(re.escape, MATH_FUNCTIONS.keys()))
1100 INTERNAL_FORMAT_RE = re.compile(rf'''(?x)
1101 (?P<negate>-)?
1102 (?P<fields>{FIELD_RE})
1103 (?P<maths>(?:{MATH_OPERATORS_RE}{MATH_FIELD_RE})*)
1104 (?:>(?P<strf_format>.+?))?
1105 (?P<remaining>
1106 (?P<alternate>(?<!\\),[^|&)]+)?
1107 (?:&(?P<replacement>.*?))?
1108 (?:\|(?P<default>.*?))?
1109 )$''')
1110
1111 def _traverse_infodict(k):
1112 k = k.split('.')
1113 if k[0] == '':
1114 k.pop(0)
1115 return traverse_obj(info_dict, k, is_user_input=True, traverse_string=True)
1116
1117 def get_value(mdict):
1118 # Object traversal
1119 value = _traverse_infodict(mdict['fields'])
1120 # Negative
1121 if mdict['negate']:
1122 value = float_or_none(value)
1123 if value is not None:
1124 value *= -1
1125 # Do maths
1126 offset_key = mdict['maths']
1127 if offset_key:
1128 value = float_or_none(value)
1129 operator = None
1130 while offset_key:
1131 item = re.match(
1132 MATH_FIELD_RE if operator else MATH_OPERATORS_RE,
1133 offset_key).group(0)
1134 offset_key = offset_key[len(item):]
1135 if operator is None:
1136 operator = MATH_FUNCTIONS[item]
1137 continue
1138 item, multiplier = (item[1:], -1) if item[0] == '-' else (item, 1)
1139 offset = float_or_none(item)
1140 if offset is None:
1141 offset = float_or_none(_traverse_infodict(item))
1142 try:
1143 value = operator(value, multiplier * offset)
1144 except (TypeError, ZeroDivisionError):
1145 return None
1146 operator = None
1147 # Datetime formatting
1148 if mdict['strf_format']:
1149 value = strftime_or_none(value, mdict['strf_format'].replace('\\,', ','))
1150
1151 return value
1152
1153 na = self.params.get('outtmpl_na_placeholder', 'NA')
1154
1155 def filename_sanitizer(key, value, restricted=self.params.get('restrictfilenames')):
1156 return sanitize_filename(str(value), restricted=restricted, is_id=(
1157 bool(re.search(r'(^|[_.])id(\.|$)', key))
1158 if 'filename-sanitization' in self.params['compat_opts']
1159 else NO_DEFAULT))
1160
1161 sanitizer = sanitize if callable(sanitize) else filename_sanitizer
1162 sanitize = bool(sanitize)
1163
1164 def _dumpjson_default(obj):
1165 if isinstance(obj, (set, LazyList)):
1166 return list(obj)
1167 return repr(obj)
1168
1169 def create_key(outer_mobj):
1170 if not outer_mobj.group('has_key'):
1171 return outer_mobj.group(0)
1172 key = outer_mobj.group('key')
1173 mobj = re.match(INTERNAL_FORMAT_RE, key)
1174 initial_field = mobj.group('fields') if mobj else ''
1175 value, replacement, default = None, None, na
1176 while mobj:
1177 mobj = mobj.groupdict()
1178 default = mobj['default'] if mobj['default'] is not None else default
1179 value = get_value(mobj)
1180 replacement = mobj['replacement']
1181 if value is None and mobj['alternate']:
1182 mobj = re.match(INTERNAL_FORMAT_RE, mobj['remaining'][1:])
1183 else:
1184 break
1185
1186 fmt = outer_mobj.group('format')
1187 if fmt == 's' and value is not None and key in field_size_compat_map.keys():
1188 fmt = f'0{field_size_compat_map[key]:d}d'
1189
1190 value = default if value is None else value if replacement is None else replacement
1191
1192 flags = outer_mobj.group('conversion') or ''
1193 str_fmt = f'{fmt[:-1]}s'
1194 if fmt[-1] == 'l': # list
1195 delim = '\n' if '#' in flags else ', '
1196 value, fmt = delim.join(map(str, variadic(value, allowed_types=(str, bytes)))), str_fmt
1197 elif fmt[-1] == 'j': # json
1198 value, fmt = json.dumps(value, default=_dumpjson_default, indent=4 if '#' in flags else None), str_fmt
1199 elif fmt[-1] == 'q': # quoted
1200 value = map(str, variadic(value) if '#' in flags else [value])
1201 value, fmt = ' '.join(map(compat_shlex_quote, value)), str_fmt
1202 elif fmt[-1] == 'B': # bytes
1203 value = f'%{str_fmt}'.encode() % str(value).encode()
1204 value, fmt = value.decode('utf-8', 'ignore'), 's'
1205 elif fmt[-1] == 'U': # unicode normalized
1206 value, fmt = unicodedata.normalize(
1207 # "+" = compatibility equivalence, "#" = NFD
1208 'NF%s%s' % ('K' if '+' in flags else '', 'D' if '#' in flags else 'C'),
1209 value), str_fmt
1210 elif fmt[-1] == 'D': # decimal suffix
1211 num_fmt, fmt = fmt[:-1].replace('#', ''), 's'
1212 value = format_decimal_suffix(value, f'%{num_fmt}f%s' if num_fmt else '%d%s',
1213 factor=1024 if '#' in flags else 1000)
1214 elif fmt[-1] == 'S': # filename sanitization
1215 value, fmt = filename_sanitizer(initial_field, value, restricted='#' in flags), str_fmt
1216 elif fmt[-1] == 'c':
1217 if value:
1218 value = str(value)[0]
1219 else:
1220 fmt = str_fmt
1221 elif fmt[-1] not in 'rs': # numeric
1222 value = float_or_none(value)
1223 if value is None:
1224 value, fmt = default, 's'
1225
1226 if sanitize:
1227 if fmt[-1] == 'r':
1228 # If value is an object, sanitize might convert it to a string
1229 # So we convert it to repr first
1230 value, fmt = repr(value), str_fmt
1231 if fmt[-1] in 'csr':
1232 value = sanitizer(initial_field, value)
1233
1234 key = '%s\0%s' % (key.replace('%', '%\0'), outer_mobj.group('format'))
1235 TMPL_DICT[key] = value
1236 return '{prefix}%({key}){fmt}'.format(key=key, fmt=fmt, prefix=outer_mobj.group('prefix'))
1237
1238 return EXTERNAL_FORMAT_RE.sub(create_key, outtmpl), TMPL_DICT
1239
1240 def evaluate_outtmpl(self, outtmpl, info_dict, *args, **kwargs):
1241 outtmpl, info_dict = self.prepare_outtmpl(outtmpl, info_dict, *args, **kwargs)
1242 return self.escape_outtmpl(outtmpl) % info_dict
1243
1244 def _prepare_filename(self, info_dict, *, outtmpl=None, tmpl_type=None):
1245 assert None in (outtmpl, tmpl_type), 'outtmpl and tmpl_type are mutually exclusive'
1246 if outtmpl is None:
1247 outtmpl = self.params['outtmpl'].get(tmpl_type or 'default', self.params['outtmpl']['default'])
1248 try:
1249 outtmpl = self._outtmpl_expandpath(outtmpl)
1250 filename = self.evaluate_outtmpl(outtmpl, info_dict, True)
1251 if not filename:
1252 return None
1253
1254 if tmpl_type in ('', 'temp'):
1255 final_ext, ext = self.params.get('final_ext'), info_dict.get('ext')
1256 if final_ext and ext and final_ext != ext and filename.endswith(f'.{final_ext}'):
1257 filename = replace_extension(filename, ext, final_ext)
1258 elif tmpl_type:
1259 force_ext = OUTTMPL_TYPES[tmpl_type]
1260 if force_ext:
1261 filename = replace_extension(filename, force_ext, info_dict.get('ext'))
1262
1263 # https://github.com/blackjack4494/youtube-dlc/issues/85
1264 trim_file_name = self.params.get('trim_file_name', False)
1265 if trim_file_name:
1266 no_ext, *ext = filename.rsplit('.', 2)
1267 filename = join_nonempty(no_ext[:trim_file_name], *ext, delim='.')
1268
1269 return filename
1270 except ValueError as err:
1271 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
1272 return None
1273
1274 def prepare_filename(self, info_dict, dir_type='', *, outtmpl=None, warn=False):
1275 """Generate the output filename"""
1276 if outtmpl:
1277 assert not dir_type, 'outtmpl and dir_type are mutually exclusive'
1278 dir_type = None
1279 filename = self._prepare_filename(info_dict, tmpl_type=dir_type, outtmpl=outtmpl)
1280 if not filename and dir_type not in ('', 'temp'):
1281 return ''
1282
1283 if warn:
1284 if not self.params.get('paths'):
1285 pass
1286 elif filename == '-':
1287 self.report_warning('--paths is ignored when an outputting to stdout', only_once=True)
1288 elif os.path.isabs(filename):
1289 self.report_warning('--paths is ignored since an absolute path is given in output template', only_once=True)
1290 if filename == '-' or not filename:
1291 return filename
1292
1293 return self.get_output_path(dir_type, filename)
1294
1295 def _match_entry(self, info_dict, incomplete=False, silent=False):
1296 """ Returns None if the file should be downloaded """
1297
1298 video_title = info_dict.get('title', info_dict.get('id', 'video'))
1299
1300 def check_filter():
1301 if 'title' in info_dict:
1302 # This can happen when we're just evaluating the playlist
1303 title = info_dict['title']
1304 matchtitle = self.params.get('matchtitle', False)
1305 if matchtitle:
1306 if not re.search(matchtitle, title, re.IGNORECASE):
1307 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
1308 rejecttitle = self.params.get('rejecttitle', False)
1309 if rejecttitle:
1310 if re.search(rejecttitle, title, re.IGNORECASE):
1311 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
1312 date = info_dict.get('upload_date')
1313 if date is not None:
1314 dateRange = self.params.get('daterange', DateRange())
1315 if date not in dateRange:
1316 return f'{date_from_str(date).isoformat()} upload date is not in range {dateRange}'
1317 view_count = info_dict.get('view_count')
1318 if view_count is not None:
1319 min_views = self.params.get('min_views')
1320 if min_views is not None and view_count < min_views:
1321 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
1322 max_views = self.params.get('max_views')
1323 if max_views is not None and view_count > max_views:
1324 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
1325 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
1326 return 'Skipping "%s" because it is age restricted' % video_title
1327
1328 match_filter = self.params.get('match_filter')
1329 if match_filter is not None:
1330 try:
1331 ret = match_filter(info_dict, incomplete=incomplete)
1332 except TypeError:
1333 # For backward compatibility
1334 ret = None if incomplete else match_filter(info_dict)
1335 if ret is NO_DEFAULT:
1336 while True:
1337 filename = self._format_screen(self.prepare_filename(info_dict), self.Styles.FILENAME)
1338 reply = input(self._format_screen(
1339 f'Download "{filename}"? (Y/n): ', self.Styles.EMPHASIS)).lower().strip()
1340 if reply in {'y', ''}:
1341 return None
1342 elif reply == 'n':
1343 return f'Skipping {video_title}'
1344 elif ret is not None:
1345 return ret
1346 return None
1347
1348 if self.in_download_archive(info_dict):
1349 reason = '%s has already been recorded in the archive' % video_title
1350 break_opt, break_err = 'break_on_existing', ExistingVideoReached
1351 else:
1352 reason = check_filter()
1353 break_opt, break_err = 'break_on_reject', RejectedVideoReached
1354 if reason is not None:
1355 if not silent:
1356 self.to_screen('[download] ' + reason)
1357 if self.params.get(break_opt, False):
1358 raise break_err()
1359 return reason
1360
1361 @staticmethod
1362 def add_extra_info(info_dict, extra_info):
1363 '''Set the keys from extra_info in info dict if they are missing'''
1364 for key, value in extra_info.items():
1365 info_dict.setdefault(key, value)
1366
1367 def extract_info(self, url, download=True, ie_key=None, extra_info=None,
1368 process=True, force_generic_extractor=False):
1369 """
1370 Return a list with a dictionary for each video extracted.
1371
1372 Arguments:
1373 url -- URL to extract
1374
1375 Keyword arguments:
1376 download -- whether to download videos during extraction
1377 ie_key -- extractor key hint
1378 extra_info -- dictionary containing the extra values to add to each result
1379 process -- whether to resolve all unresolved references (URLs, playlist items),
1380 must be True for download to work.
1381 force_generic_extractor -- force using the generic extractor
1382 """
1383
1384 if extra_info is None:
1385 extra_info = {}
1386
1387 if not ie_key and force_generic_extractor:
1388 ie_key = 'Generic'
1389
1390 if ie_key:
1391 ies = {ie_key: self._get_info_extractor_class(ie_key)}
1392 else:
1393 ies = self._ies
1394
1395 for ie_key, ie in ies.items():
1396 if not ie.suitable(url):
1397 continue
1398
1399 if not ie.working():
1400 self.report_warning('The program functionality for this site has been marked as broken, '
1401 'and will probably not work.')
1402
1403 temp_id = ie.get_temp_id(url)
1404 if temp_id is not None and self.in_download_archive({'id': temp_id, 'ie_key': ie_key}):
1405 self.to_screen(f'[{ie_key}] {temp_id}: has already been recorded in the archive')
1406 if self.params.get('break_on_existing', False):
1407 raise ExistingVideoReached()
1408 break
1409 return self.__extract_info(url, self.get_info_extractor(ie_key), download, extra_info, process)
1410 else:
1411 self.report_error('no suitable InfoExtractor for URL %s' % url)
1412
1413 def _handle_extraction_exceptions(func):
1414 @functools.wraps(func)
1415 def wrapper(self, *args, **kwargs):
1416 while True:
1417 try:
1418 return func(self, *args, **kwargs)
1419 except (DownloadCancelled, LazyList.IndexError, PagedList.IndexError):
1420 raise
1421 except ReExtractInfo as e:
1422 if e.expected:
1423 self.to_screen(f'{e}; Re-extracting data')
1424 else:
1425 self.to_stderr('\r')
1426 self.report_warning(f'{e}; Re-extracting data')
1427 continue
1428 except GeoRestrictedError as e:
1429 msg = e.msg
1430 if e.countries:
1431 msg += '\nThis video is available in %s.' % ', '.join(
1432 map(ISO3166Utils.short2full, e.countries))
1433 msg += '\nYou might want to use a VPN or a proxy server (with --proxy) to workaround.'
1434 self.report_error(msg)
1435 except ExtractorError as e: # An error we somewhat expected
1436 self.report_error(str(e), e.format_traceback())
1437 except Exception as e:
1438 if self.params.get('ignoreerrors'):
1439 self.report_error(str(e), tb=encode_compat_str(traceback.format_exc()))
1440 else:
1441 raise
1442 break
1443 return wrapper
1444
1445 def _wait_for_video(self, ie_result):
1446 if (not self.params.get('wait_for_video')
1447 or ie_result.get('_type', 'video') != 'video'
1448 or ie_result.get('formats') or ie_result.get('url')):
1449 return
1450
1451 format_dur = lambda dur: '%02d:%02d:%02d' % timetuple_from_msec(dur * 1000)[:-1]
1452 last_msg = ''
1453
1454 def progress(msg):
1455 nonlocal last_msg
1456 self.to_screen(msg + ' ' * (len(last_msg) - len(msg)) + '\r', skip_eol=True)
1457 last_msg = msg
1458
1459 min_wait, max_wait = self.params.get('wait_for_video')
1460 diff = try_get(ie_result, lambda x: x['release_timestamp'] - time.time())
1461 if diff is None and ie_result.get('live_status') == 'is_upcoming':
1462 diff = round(random.uniform(min_wait, max_wait) if (max_wait and min_wait) else (max_wait or min_wait), 0)
1463 self.report_warning('Release time of video is not known')
1464 elif (diff or 0) <= 0:
1465 self.report_warning('Video should already be available according to extracted info')
1466 diff = min(max(diff or 0, min_wait or 0), max_wait or float('inf'))
1467 self.to_screen(f'[wait] Waiting for {format_dur(diff)} - Press Ctrl+C to try now')
1468
1469 wait_till = time.time() + diff
1470 try:
1471 while True:
1472 diff = wait_till - time.time()
1473 if diff <= 0:
1474 progress('')
1475 raise ReExtractInfo('[wait] Wait period ended', expected=True)
1476 progress(f'[wait] Remaining time until next attempt: {self._format_screen(format_dur(diff), self.Styles.EMPHASIS)}')
1477 time.sleep(1)
1478 except KeyboardInterrupt:
1479 progress('')
1480 raise ReExtractInfo('[wait] Interrupted by user', expected=True)
1481 except BaseException as e:
1482 if not isinstance(e, ReExtractInfo):
1483 self.to_screen('')
1484 raise
1485
1486 @_handle_extraction_exceptions
1487 def __extract_info(self, url, ie, download, extra_info, process):
1488 ie_result = ie.extract(url)
1489 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
1490 return
1491 if isinstance(ie_result, list):
1492 # Backwards compatibility: old IE result format
1493 ie_result = {
1494 '_type': 'compat_list',
1495 'entries': ie_result,
1496 }
1497 if extra_info.get('original_url'):
1498 ie_result.setdefault('original_url', extra_info['original_url'])
1499 self.add_default_extra_info(ie_result, ie, url)
1500 if process:
1501 self._wait_for_video(ie_result)
1502 return self.process_ie_result(ie_result, download, extra_info)
1503 else:
1504 return ie_result
1505
1506 def add_default_extra_info(self, ie_result, ie, url):
1507 if url is not None:
1508 self.add_extra_info(ie_result, {
1509 'webpage_url': url,
1510 'original_url': url,
1511 })
1512 webpage_url = ie_result.get('webpage_url')
1513 if webpage_url:
1514 self.add_extra_info(ie_result, {
1515 'webpage_url_basename': url_basename(webpage_url),
1516 'webpage_url_domain': get_domain(webpage_url),
1517 })
1518 if ie is not None:
1519 self.add_extra_info(ie_result, {
1520 'extractor': ie.IE_NAME,
1521 'extractor_key': ie.ie_key(),
1522 })
1523
1524 def process_ie_result(self, ie_result, download=True, extra_info=None):
1525 """
1526 Take the result of the ie(may be modified) and resolve all unresolved
1527 references (URLs, playlist items).
1528
1529 It will also download the videos if 'download'.
1530 Returns the resolved ie_result.
1531 """
1532 if extra_info is None:
1533 extra_info = {}
1534 result_type = ie_result.get('_type', 'video')
1535
1536 if result_type in ('url', 'url_transparent'):
1537 ie_result['url'] = sanitize_url(ie_result['url'])
1538 if ie_result.get('original_url'):
1539 extra_info.setdefault('original_url', ie_result['original_url'])
1540
1541 extract_flat = self.params.get('extract_flat', False)
1542 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info)
1543 or extract_flat is True):
1544 info_copy = ie_result.copy()
1545 ie = try_get(ie_result.get('ie_key'), self.get_info_extractor)
1546 if ie and not ie_result.get('id'):
1547 info_copy['id'] = ie.get_temp_id(ie_result['url'])
1548 self.add_default_extra_info(info_copy, ie, ie_result['url'])
1549 self.add_extra_info(info_copy, extra_info)
1550 info_copy, _ = self.pre_process(info_copy)
1551 self.__forced_printings(info_copy, self.prepare_filename(info_copy), incomplete=True)
1552 self._raise_pending_errors(info_copy)
1553 if self.params.get('force_write_download_archive', False):
1554 self.record_download_archive(info_copy)
1555 return ie_result
1556
1557 if result_type == 'video':
1558 self.add_extra_info(ie_result, extra_info)
1559 ie_result = self.process_video_result(ie_result, download=download)
1560 self._raise_pending_errors(ie_result)
1561 additional_urls = (ie_result or {}).get('additional_urls')
1562 if additional_urls:
1563 # TODO: Improve MetadataParserPP to allow setting a list
1564 if isinstance(additional_urls, str):
1565 additional_urls = [additional_urls]
1566 self.to_screen(
1567 '[info] %s: %d additional URL(s) requested' % (ie_result['id'], len(additional_urls)))
1568 self.write_debug('Additional URLs: "%s"' % '", "'.join(additional_urls))
1569 ie_result['additional_entries'] = [
1570 self.extract_info(
1571 url, download, extra_info=extra_info,
1572 force_generic_extractor=self.params.get('force_generic_extractor'))
1573 for url in additional_urls
1574 ]
1575 return ie_result
1576 elif result_type == 'url':
1577 # We have to add extra_info to the results because it may be
1578 # contained in a playlist
1579 return self.extract_info(
1580 ie_result['url'], download,
1581 ie_key=ie_result.get('ie_key'),
1582 extra_info=extra_info)
1583 elif result_type == 'url_transparent':
1584 # Use the information from the embedding page
1585 info = self.extract_info(
1586 ie_result['url'], ie_key=ie_result.get('ie_key'),
1587 extra_info=extra_info, download=False, process=False)
1588
1589 # extract_info may return None when ignoreerrors is enabled and
1590 # extraction failed with an error, don't crash and return early
1591 # in this case
1592 if not info:
1593 return info
1594
1595 exempted_fields = {'_type', 'url', 'ie_key'}
1596 if not ie_result.get('section_end') and ie_result.get('section_start') is None:
1597 # For video clips, the id etc of the clip extractor should be used
1598 exempted_fields |= {'id', 'extractor', 'extractor_key'}
1599
1600 new_result = info.copy()
1601 new_result.update(filter_dict(ie_result, lambda k, v: v is not None and k not in exempted_fields))
1602
1603 # Extracted info may not be a video result (i.e.
1604 # info.get('_type', 'video') != video) but rather an url or
1605 # url_transparent. In such cases outer metadata (from ie_result)
1606 # should be propagated to inner one (info). For this to happen
1607 # _type of info should be overridden with url_transparent. This
1608 # fixes issue from https://github.com/ytdl-org/youtube-dl/pull/11163.
1609 if new_result.get('_type') == 'url':
1610 new_result['_type'] = 'url_transparent'
1611
1612 return self.process_ie_result(
1613 new_result, download=download, extra_info=extra_info)
1614 elif result_type in ('playlist', 'multi_video'):
1615 # Protect from infinite recursion due to recursively nested playlists
1616 # (see https://github.com/ytdl-org/youtube-dl/issues/27833)
1617 webpage_url = ie_result['webpage_url']
1618 if webpage_url in self._playlist_urls:
1619 self.to_screen(
1620 '[download] Skipping already downloaded playlist: %s'
1621 % ie_result.get('title') or ie_result.get('id'))
1622 return
1623
1624 self._playlist_level += 1
1625 self._playlist_urls.add(webpage_url)
1626 self._fill_common_fields(ie_result, False)
1627 self._sanitize_thumbnails(ie_result)
1628 try:
1629 return self.__process_playlist(ie_result, download)
1630 finally:
1631 self._playlist_level -= 1
1632 if not self._playlist_level:
1633 self._playlist_urls.clear()
1634 elif result_type == 'compat_list':
1635 self.report_warning(
1636 'Extractor %s returned a compat_list result. '
1637 'It needs to be updated.' % ie_result.get('extractor'))
1638
1639 def _fixup(r):
1640 self.add_extra_info(r, {
1641 'extractor': ie_result['extractor'],
1642 'webpage_url': ie_result['webpage_url'],
1643 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1644 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1645 'extractor_key': ie_result['extractor_key'],
1646 })
1647 return r
1648 ie_result['entries'] = [
1649 self.process_ie_result(_fixup(r), download, extra_info)
1650 for r in ie_result['entries']
1651 ]
1652 return ie_result
1653 else:
1654 raise Exception('Invalid result type: %s' % result_type)
1655
1656 def _ensure_dir_exists(self, path):
1657 return make_dir(path, self.report_error)
1658
1659 @staticmethod
1660 def _playlist_infodict(ie_result, **kwargs):
1661 return {
1662 **ie_result,
1663 'playlist': ie_result.get('title') or ie_result.get('id'),
1664 'playlist_id': ie_result.get('id'),
1665 'playlist_title': ie_result.get('title'),
1666 'playlist_uploader': ie_result.get('uploader'),
1667 'playlist_uploader_id': ie_result.get('uploader_id'),
1668 'playlist_index': 0,
1669 **kwargs,
1670 }
1671
1672 def __process_playlist(self, ie_result, download):
1673 """Process each entry in the playlist"""
1674 title = ie_result.get('title') or ie_result.get('id') or '<Untitled>'
1675 self.to_screen(f'[download] Downloading playlist: {title}')
1676
1677 all_entries = PlaylistEntries(self, ie_result)
1678 entries = orderedSet(all_entries.get_requested_items(), lazy=True)
1679
1680 lazy = self.params.get('lazy_playlist')
1681 if lazy:
1682 resolved_entries, n_entries = [], 'N/A'
1683 ie_result['requested_entries'], ie_result['entries'] = None, None
1684 else:
1685 entries = resolved_entries = list(entries)
1686 n_entries = len(resolved_entries)
1687 ie_result['requested_entries'], ie_result['entries'] = tuple(zip(*resolved_entries)) or ([], [])
1688 if not ie_result.get('playlist_count'):
1689 # Better to do this after potentially exhausting entries
1690 ie_result['playlist_count'] = all_entries.get_full_count()
1691
1692 _infojson_written = False
1693 write_playlist_files = self.params.get('allow_playlist_files', True)
1694 if write_playlist_files and self.params.get('list_thumbnails'):
1695 self.list_thumbnails(ie_result)
1696 if write_playlist_files and not self.params.get('simulate'):
1697 ie_copy = self._playlist_infodict(ie_result, n_entries=int_or_none(n_entries))
1698 _infojson_written = self._write_info_json(
1699 'playlist', ie_result, self.prepare_filename(ie_copy, 'pl_infojson'))
1700 if _infojson_written is None:
1701 return
1702 if self._write_description('playlist', ie_result,
1703 self.prepare_filename(ie_copy, 'pl_description')) is None:
1704 return
1705 # TODO: This should be passed to ThumbnailsConvertor if necessary
1706 self._write_thumbnails('playlist', ie_copy, self.prepare_filename(ie_copy, 'pl_thumbnail'))
1707
1708 if lazy:
1709 if self.params.get('playlistreverse') or self.params.get('playlistrandom'):
1710 self.report_warning('playlistreverse and playlistrandom are not supported with lazy_playlist', only_once=True)
1711 elif self.params.get('playlistreverse'):
1712 entries.reverse()
1713 elif self.params.get('playlistrandom'):
1714 random.shuffle(entries)
1715
1716 self.to_screen(f'[{ie_result["extractor"]}] Playlist {title}: Downloading {n_entries} videos'
1717 f'{format_field(ie_result, "playlist_count", " of %s")}')
1718
1719 failures = 0
1720 max_failures = self.params.get('skip_playlist_after_errors') or float('inf')
1721 for i, (playlist_index, entry) in enumerate(entries):
1722 if lazy:
1723 resolved_entries.append((playlist_index, entry))
1724
1725 # TODO: Add auto-generated fields
1726 if not entry or self._match_entry(entry, incomplete=True) is not None:
1727 continue
1728
1729 self.to_screen('[download] Downloading video %s of %s' % (
1730 self._format_screen(i + 1, self.Styles.ID), self._format_screen(n_entries, self.Styles.EMPHASIS)))
1731
1732 entry['__x_forwarded_for_ip'] = ie_result.get('__x_forwarded_for_ip')
1733 if not lazy and 'playlist-index' in self.params.get('compat_opts', []):
1734 playlist_index = ie_result['requested_entries'][i]
1735
1736 entry_result = self.__process_iterable_entry(entry, download, {
1737 'n_entries': int_or_none(n_entries),
1738 '__last_playlist_index': max(ie_result['requested_entries'] or (0, 0)),
1739 'playlist_count': ie_result.get('playlist_count'),
1740 'playlist_index': playlist_index,
1741 'playlist_autonumber': i + 1,
1742 'playlist': title,
1743 'playlist_id': ie_result.get('id'),
1744 'playlist_title': ie_result.get('title'),
1745 'playlist_uploader': ie_result.get('uploader'),
1746 'playlist_uploader_id': ie_result.get('uploader_id'),
1747 'extractor': ie_result['extractor'],
1748 'webpage_url': ie_result['webpage_url'],
1749 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1750 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1751 'extractor_key': ie_result['extractor_key'],
1752 })
1753 if not entry_result:
1754 failures += 1
1755 if failures >= max_failures:
1756 self.report_error(
1757 f'Skipping the remaining entries in playlist "{title}" since {failures} items failed extraction')
1758 break
1759 resolved_entries[i] = (playlist_index, entry_result)
1760
1761 # Update with processed data
1762 ie_result['requested_entries'], ie_result['entries'] = tuple(zip(*resolved_entries)) or ([], [])
1763
1764 # Write the updated info to json
1765 if _infojson_written is True and self._write_info_json(
1766 'updated playlist', ie_result,
1767 self.prepare_filename(ie_copy, 'pl_infojson'), overwrite=True) is None:
1768 return
1769
1770 ie_result = self.run_all_pps('playlist', ie_result)
1771 self.to_screen(f'[download] Finished downloading playlist: {title}')
1772 return ie_result
1773
1774 @_handle_extraction_exceptions
1775 def __process_iterable_entry(self, entry, download, extra_info):
1776 return self.process_ie_result(
1777 entry, download=download, extra_info=extra_info)
1778
1779 def _build_format_filter(self, filter_spec):
1780 " Returns a function to filter the formats according to the filter_spec "
1781
1782 OPERATORS = {
1783 '<': operator.lt,
1784 '<=': operator.le,
1785 '>': operator.gt,
1786 '>=': operator.ge,
1787 '=': operator.eq,
1788 '!=': operator.ne,
1789 }
1790 operator_rex = re.compile(r'''(?x)\s*
1791 (?P<key>width|height|tbr|abr|vbr|asr|filesize|filesize_approx|fps)\s*
1792 (?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
1793 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)\s*
1794 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
1795 m = operator_rex.fullmatch(filter_spec)
1796 if m:
1797 try:
1798 comparison_value = int(m.group('value'))
1799 except ValueError:
1800 comparison_value = parse_filesize(m.group('value'))
1801 if comparison_value is None:
1802 comparison_value = parse_filesize(m.group('value') + 'B')
1803 if comparison_value is None:
1804 raise ValueError(
1805 'Invalid value %r in format specification %r' % (
1806 m.group('value'), filter_spec))
1807 op = OPERATORS[m.group('op')]
1808
1809 if not m:
1810 STR_OPERATORS = {
1811 '=': operator.eq,
1812 '^=': lambda attr, value: attr.startswith(value),
1813 '$=': lambda attr, value: attr.endswith(value),
1814 '*=': lambda attr, value: value in attr,
1815 '~=': lambda attr, value: value.search(attr) is not None
1816 }
1817 str_operator_rex = re.compile(r'''(?x)\s*
1818 (?P<key>[a-zA-Z0-9._-]+)\s*
1819 (?P<negation>!\s*)?(?P<op>%s)\s*(?P<none_inclusive>\?\s*)?
1820 (?P<quote>["'])?
1821 (?P<value>(?(quote)(?:(?!(?P=quote))[^\\]|\\.)+|[\w.-]+))
1822 (?(quote)(?P=quote))\s*
1823 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
1824 m = str_operator_rex.fullmatch(filter_spec)
1825 if m:
1826 if m.group('op') == '~=':
1827 comparison_value = re.compile(m.group('value'))
1828 else:
1829 comparison_value = re.sub(r'''\\([\\"'])''', r'\1', m.group('value'))
1830 str_op = STR_OPERATORS[m.group('op')]
1831 if m.group('negation'):
1832 op = lambda attr, value: not str_op(attr, value)
1833 else:
1834 op = str_op
1835
1836 if not m:
1837 raise SyntaxError('Invalid filter specification %r' % filter_spec)
1838
1839 def _filter(f):
1840 actual_value = f.get(m.group('key'))
1841 if actual_value is None:
1842 return m.group('none_inclusive')
1843 return op(actual_value, comparison_value)
1844 return _filter
1845
1846 def _check_formats(self, formats):
1847 for f in formats:
1848 self.to_screen('[info] Testing format %s' % f['format_id'])
1849 path = self.get_output_path('temp')
1850 if not self._ensure_dir_exists(f'{path}/'):
1851 continue
1852 temp_file = tempfile.NamedTemporaryFile(suffix='.tmp', delete=False, dir=path or None)
1853 temp_file.close()
1854 try:
1855 success, _ = self.dl(temp_file.name, f, test=True)
1856 except (DownloadError, OSError, ValueError) + network_exceptions:
1857 success = False
1858 finally:
1859 if os.path.exists(temp_file.name):
1860 try:
1861 os.remove(temp_file.name)
1862 except OSError:
1863 self.report_warning('Unable to delete temporary file "%s"' % temp_file.name)
1864 if success:
1865 yield f
1866 else:
1867 self.to_screen('[info] Unable to download format %s. Skipping...' % f['format_id'])
1868
1869 def _default_format_spec(self, info_dict, download=True):
1870
1871 def can_merge():
1872 merger = FFmpegMergerPP(self)
1873 return merger.available and merger.can_merge()
1874
1875 prefer_best = (
1876 not self.params.get('simulate')
1877 and download
1878 and (
1879 not can_merge()
1880 or info_dict.get('is_live') and not self.params.get('live_from_start')
1881 or self.params['outtmpl']['default'] == '-'))
1882 compat = (
1883 prefer_best
1884 or self.params.get('allow_multiple_audio_streams', False)
1885 or 'format-spec' in self.params['compat_opts'])
1886
1887 return (
1888 'best/bestvideo+bestaudio' if prefer_best
1889 else 'bestvideo*+bestaudio/best' if not compat
1890 else 'bestvideo+bestaudio/best')
1891
1892 def build_format_selector(self, format_spec):
1893 def syntax_error(note, start):
1894 message = (
1895 'Invalid format specification: '
1896 '{}\n\t{}\n\t{}^'.format(note, format_spec, ' ' * start[1]))
1897 return SyntaxError(message)
1898
1899 PICKFIRST = 'PICKFIRST'
1900 MERGE = 'MERGE'
1901 SINGLE = 'SINGLE'
1902 GROUP = 'GROUP'
1903 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
1904
1905 allow_multiple_streams = {'audio': self.params.get('allow_multiple_audio_streams', False),
1906 'video': self.params.get('allow_multiple_video_streams', False)}
1907
1908 check_formats = self.params.get('check_formats') == 'selected'
1909
1910 def _parse_filter(tokens):
1911 filter_parts = []
1912 for type, string, start, _, _ in tokens:
1913 if type == tokenize.OP and string == ']':
1914 return ''.join(filter_parts)
1915 else:
1916 filter_parts.append(string)
1917
1918 def _remove_unused_ops(tokens):
1919 # Remove operators that we don't use and join them with the surrounding strings
1920 # for example: 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
1921 ALLOWED_OPS = ('/', '+', ',', '(', ')')
1922 last_string, last_start, last_end, last_line = None, None, None, None
1923 for type, string, start, end, line in tokens:
1924 if type == tokenize.OP and string == '[':
1925 if last_string:
1926 yield tokenize.NAME, last_string, last_start, last_end, last_line
1927 last_string = None
1928 yield type, string, start, end, line
1929 # everything inside brackets will be handled by _parse_filter
1930 for type, string, start, end, line in tokens:
1931 yield type, string, start, end, line
1932 if type == tokenize.OP and string == ']':
1933 break
1934 elif type == tokenize.OP and string in ALLOWED_OPS:
1935 if last_string:
1936 yield tokenize.NAME, last_string, last_start, last_end, last_line
1937 last_string = None
1938 yield type, string, start, end, line
1939 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
1940 if not last_string:
1941 last_string = string
1942 last_start = start
1943 last_end = end
1944 else:
1945 last_string += string
1946 if last_string:
1947 yield tokenize.NAME, last_string, last_start, last_end, last_line
1948
1949 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
1950 selectors = []
1951 current_selector = None
1952 for type, string, start, _, _ in tokens:
1953 # ENCODING is only defined in python 3.x
1954 if type == getattr(tokenize, 'ENCODING', None):
1955 continue
1956 elif type in [tokenize.NAME, tokenize.NUMBER]:
1957 current_selector = FormatSelector(SINGLE, string, [])
1958 elif type == tokenize.OP:
1959 if string == ')':
1960 if not inside_group:
1961 # ')' will be handled by the parentheses group
1962 tokens.restore_last_token()
1963 break
1964 elif inside_merge and string in ['/', ',']:
1965 tokens.restore_last_token()
1966 break
1967 elif inside_choice and string == ',':
1968 tokens.restore_last_token()
1969 break
1970 elif string == ',':
1971 if not current_selector:
1972 raise syntax_error('"," must follow a format selector', start)
1973 selectors.append(current_selector)
1974 current_selector = None
1975 elif string == '/':
1976 if not current_selector:
1977 raise syntax_error('"/" must follow a format selector', start)
1978 first_choice = current_selector
1979 second_choice = _parse_format_selection(tokens, inside_choice=True)
1980 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
1981 elif string == '[':
1982 if not current_selector:
1983 current_selector = FormatSelector(SINGLE, 'best', [])
1984 format_filter = _parse_filter(tokens)
1985 current_selector.filters.append(format_filter)
1986 elif string == '(':
1987 if current_selector:
1988 raise syntax_error('Unexpected "("', start)
1989 group = _parse_format_selection(tokens, inside_group=True)
1990 current_selector = FormatSelector(GROUP, group, [])
1991 elif string == '+':
1992 if not current_selector:
1993 raise syntax_error('Unexpected "+"', start)
1994 selector_1 = current_selector
1995 selector_2 = _parse_format_selection(tokens, inside_merge=True)
1996 if not selector_2:
1997 raise syntax_error('Expected a selector', start)
1998 current_selector = FormatSelector(MERGE, (selector_1, selector_2), [])
1999 else:
2000 raise syntax_error(f'Operator not recognized: "{string}"', start)
2001 elif type == tokenize.ENDMARKER:
2002 break
2003 if current_selector:
2004 selectors.append(current_selector)
2005 return selectors
2006
2007 def _merge(formats_pair):
2008 format_1, format_2 = formats_pair
2009
2010 formats_info = []
2011 formats_info.extend(format_1.get('requested_formats', (format_1,)))
2012 formats_info.extend(format_2.get('requested_formats', (format_2,)))
2013
2014 if not allow_multiple_streams['video'] or not allow_multiple_streams['audio']:
2015 get_no_more = {'video': False, 'audio': False}
2016 for (i, fmt_info) in enumerate(formats_info):
2017 if fmt_info.get('acodec') == fmt_info.get('vcodec') == 'none':
2018 formats_info.pop(i)
2019 continue
2020 for aud_vid in ['audio', 'video']:
2021 if not allow_multiple_streams[aud_vid] and fmt_info.get(aud_vid[0] + 'codec') != 'none':
2022 if get_no_more[aud_vid]:
2023 formats_info.pop(i)
2024 break
2025 get_no_more[aud_vid] = True
2026
2027 if len(formats_info) == 1:
2028 return formats_info[0]
2029
2030 video_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('vcodec') != 'none']
2031 audio_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('acodec') != 'none']
2032
2033 the_only_video = video_fmts[0] if len(video_fmts) == 1 else None
2034 the_only_audio = audio_fmts[0] if len(audio_fmts) == 1 else None
2035
2036 output_ext = self.params.get('merge_output_format')
2037 if not output_ext:
2038 if the_only_video:
2039 output_ext = the_only_video['ext']
2040 elif the_only_audio and not video_fmts:
2041 output_ext = the_only_audio['ext']
2042 else:
2043 output_ext = 'mkv'
2044
2045 filtered = lambda *keys: filter(None, (traverse_obj(fmt, *keys) for fmt in formats_info))
2046
2047 new_dict = {
2048 'requested_formats': formats_info,
2049 'format': '+'.join(filtered('format')),
2050 'format_id': '+'.join(filtered('format_id')),
2051 'ext': output_ext,
2052 'protocol': '+'.join(map(determine_protocol, formats_info)),
2053 'language': '+'.join(orderedSet(filtered('language'))) or None,
2054 'format_note': '+'.join(orderedSet(filtered('format_note'))) or None,
2055 'filesize_approx': sum(filtered('filesize', 'filesize_approx')) or None,
2056 'tbr': sum(filtered('tbr', 'vbr', 'abr')),
2057 }
2058
2059 if the_only_video:
2060 new_dict.update({
2061 'width': the_only_video.get('width'),
2062 'height': the_only_video.get('height'),
2063 'resolution': the_only_video.get('resolution') or self.format_resolution(the_only_video),
2064 'fps': the_only_video.get('fps'),
2065 'dynamic_range': the_only_video.get('dynamic_range'),
2066 'vcodec': the_only_video.get('vcodec'),
2067 'vbr': the_only_video.get('vbr'),
2068 'stretched_ratio': the_only_video.get('stretched_ratio'),
2069 })
2070
2071 if the_only_audio:
2072 new_dict.update({
2073 'acodec': the_only_audio.get('acodec'),
2074 'abr': the_only_audio.get('abr'),
2075 'asr': the_only_audio.get('asr'),
2076 })
2077
2078 return new_dict
2079
2080 def _check_formats(formats):
2081 if not check_formats:
2082 yield from formats
2083 return
2084 yield from self._check_formats(formats)
2085
2086 def _build_selector_function(selector):
2087 if isinstance(selector, list): # ,
2088 fs = [_build_selector_function(s) for s in selector]
2089
2090 def selector_function(ctx):
2091 for f in fs:
2092 yield from f(ctx)
2093 return selector_function
2094
2095 elif selector.type == GROUP: # ()
2096 selector_function = _build_selector_function(selector.selector)
2097
2098 elif selector.type == PICKFIRST: # /
2099 fs = [_build_selector_function(s) for s in selector.selector]
2100
2101 def selector_function(ctx):
2102 for f in fs:
2103 picked_formats = list(f(ctx))
2104 if picked_formats:
2105 return picked_formats
2106 return []
2107
2108 elif selector.type == MERGE: # +
2109 selector_1, selector_2 = map(_build_selector_function, selector.selector)
2110
2111 def selector_function(ctx):
2112 for pair in itertools.product(selector_1(ctx), selector_2(ctx)):
2113 yield _merge(pair)
2114
2115 elif selector.type == SINGLE: # atom
2116 format_spec = selector.selector or 'best'
2117
2118 # TODO: Add allvideo, allaudio etc by generalizing the code with best/worst selector
2119 if format_spec == 'all':
2120 def selector_function(ctx):
2121 yield from _check_formats(ctx['formats'][::-1])
2122 elif format_spec == 'mergeall':
2123 def selector_function(ctx):
2124 formats = list(_check_formats(
2125 f for f in ctx['formats'] if f.get('vcodec') != 'none' or f.get('acodec') != 'none'))
2126 if not formats:
2127 return
2128 merged_format = formats[-1]
2129 for f in formats[-2::-1]:
2130 merged_format = _merge((merged_format, f))
2131 yield merged_format
2132
2133 else:
2134 format_fallback, seperate_fallback, format_reverse, format_idx = False, None, True, 1
2135 mobj = re.match(
2136 r'(?P<bw>best|worst|b|w)(?P<type>video|audio|v|a)?(?P<mod>\*)?(?:\.(?P<n>[1-9]\d*))?$',
2137 format_spec)
2138 if mobj is not None:
2139 format_idx = int_or_none(mobj.group('n'), default=1)
2140 format_reverse = mobj.group('bw')[0] == 'b'
2141 format_type = (mobj.group('type') or [None])[0]
2142 not_format_type = {'v': 'a', 'a': 'v'}.get(format_type)
2143 format_modified = mobj.group('mod') is not None
2144
2145 format_fallback = not format_type and not format_modified # for b, w
2146 _filter_f = (
2147 (lambda f: f.get('%scodec' % format_type) != 'none')
2148 if format_type and format_modified # bv*, ba*, wv*, wa*
2149 else (lambda f: f.get('%scodec' % not_format_type) == 'none')
2150 if format_type # bv, ba, wv, wa
2151 else (lambda f: f.get('vcodec') != 'none' and f.get('acodec') != 'none')
2152 if not format_modified # b, w
2153 else lambda f: True) # b*, w*
2154 filter_f = lambda f: _filter_f(f) and (
2155 f.get('vcodec') != 'none' or f.get('acodec') != 'none')
2156 else:
2157 if format_spec in self._format_selection_exts['audio']:
2158 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none'
2159 elif format_spec in self._format_selection_exts['video']:
2160 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none' and f.get('vcodec') != 'none'
2161 seperate_fallback = lambda f: f.get('ext') == format_spec and f.get('vcodec') != 'none'
2162 elif format_spec in self._format_selection_exts['storyboards']:
2163 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') == 'none' and f.get('vcodec') == 'none'
2164 else:
2165 filter_f = lambda f: f.get('format_id') == format_spec # id
2166
2167 def selector_function(ctx):
2168 formats = list(ctx['formats'])
2169 matches = list(filter(filter_f, formats)) if filter_f is not None else formats
2170 if not matches:
2171 if format_fallback and ctx['incomplete_formats']:
2172 # for extractors with incomplete formats (audio only (soundcloud)
2173 # or video only (imgur)) best/worst will fallback to
2174 # best/worst {video,audio}-only format
2175 matches = formats
2176 elif seperate_fallback and not ctx['has_merged_format']:
2177 # for compatibility with youtube-dl when there is no pre-merged format
2178 matches = list(filter(seperate_fallback, formats))
2179 matches = LazyList(_check_formats(matches[::-1 if format_reverse else 1]))
2180 try:
2181 yield matches[format_idx - 1]
2182 except LazyList.IndexError:
2183 return
2184
2185 filters = [self._build_format_filter(f) for f in selector.filters]
2186
2187 def final_selector(ctx):
2188 ctx_copy = dict(ctx)
2189 for _filter in filters:
2190 ctx_copy['formats'] = list(filter(_filter, ctx_copy['formats']))
2191 return selector_function(ctx_copy)
2192 return final_selector
2193
2194 stream = io.BytesIO(format_spec.encode())
2195 try:
2196 tokens = list(_remove_unused_ops(tokenize.tokenize(stream.readline)))
2197 except tokenize.TokenError:
2198 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
2199
2200 class TokenIterator:
2201 def __init__(self, tokens):
2202 self.tokens = tokens
2203 self.counter = 0
2204
2205 def __iter__(self):
2206 return self
2207
2208 def __next__(self):
2209 if self.counter >= len(self.tokens):
2210 raise StopIteration()
2211 value = self.tokens[self.counter]
2212 self.counter += 1
2213 return value
2214
2215 next = __next__
2216
2217 def restore_last_token(self):
2218 self.counter -= 1
2219
2220 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
2221 return _build_selector_function(parsed_selector)
2222
2223 def _calc_headers(self, info_dict):
2224 res = merge_headers(self.params['http_headers'], info_dict.get('http_headers') or {})
2225
2226 cookies = self._calc_cookies(info_dict['url'])
2227 if cookies:
2228 res['Cookie'] = cookies
2229
2230 if 'X-Forwarded-For' not in res:
2231 x_forwarded_for_ip = info_dict.get('__x_forwarded_for_ip')
2232 if x_forwarded_for_ip:
2233 res['X-Forwarded-For'] = x_forwarded_for_ip
2234
2235 return res
2236
2237 def _calc_cookies(self, url):
2238 pr = sanitized_Request(url)
2239 self.cookiejar.add_cookie_header(pr)
2240 return pr.get_header('Cookie')
2241
2242 def _sort_thumbnails(self, thumbnails):
2243 thumbnails.sort(key=lambda t: (
2244 t.get('preference') if t.get('preference') is not None else -1,
2245 t.get('width') if t.get('width') is not None else -1,
2246 t.get('height') if t.get('height') is not None else -1,
2247 t.get('id') if t.get('id') is not None else '',
2248 t.get('url')))
2249
2250 def _sanitize_thumbnails(self, info_dict):
2251 thumbnails = info_dict.get('thumbnails')
2252 if thumbnails is None:
2253 thumbnail = info_dict.get('thumbnail')
2254 if thumbnail:
2255 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
2256 if not thumbnails:
2257 return
2258
2259 def check_thumbnails(thumbnails):
2260 for t in thumbnails:
2261 self.to_screen(f'[info] Testing thumbnail {t["id"]}')
2262 try:
2263 self.urlopen(HEADRequest(t['url']))
2264 except network_exceptions as err:
2265 self.to_screen(f'[info] Unable to connect to thumbnail {t["id"]} URL {t["url"]!r} - {err}. Skipping...')
2266 continue
2267 yield t
2268
2269 self._sort_thumbnails(thumbnails)
2270 for i, t in enumerate(thumbnails):
2271 if t.get('id') is None:
2272 t['id'] = '%d' % i
2273 if t.get('width') and t.get('height'):
2274 t['resolution'] = '%dx%d' % (t['width'], t['height'])
2275 t['url'] = sanitize_url(t['url'])
2276
2277 if self.params.get('check_formats') is True:
2278 info_dict['thumbnails'] = LazyList(check_thumbnails(thumbnails[::-1]), reverse=True)
2279 else:
2280 info_dict['thumbnails'] = thumbnails
2281
2282 def _fill_common_fields(self, info_dict, is_video=True):
2283 # TODO: move sanitization here
2284 if is_video:
2285 # playlists are allowed to lack "title"
2286 title = info_dict.get('title', NO_DEFAULT)
2287 if title is NO_DEFAULT:
2288 raise ExtractorError('Missing "title" field in extractor result',
2289 video_id=info_dict['id'], ie=info_dict['extractor'])
2290 info_dict['fulltitle'] = title
2291 if not title:
2292 if title == '':
2293 self.write_debug('Extractor gave empty title. Creating a generic title')
2294 else:
2295 self.report_warning('Extractor failed to obtain "title". Creating a generic title instead')
2296 info_dict['title'] = f'{info_dict["extractor"].replace(":", "-")} video #{info_dict["id"]}'
2297
2298 if info_dict.get('duration') is not None:
2299 info_dict['duration_string'] = formatSeconds(info_dict['duration'])
2300
2301 for ts_key, date_key in (
2302 ('timestamp', 'upload_date'),
2303 ('release_timestamp', 'release_date'),
2304 ('modified_timestamp', 'modified_date'),
2305 ):
2306 if info_dict.get(date_key) is None and info_dict.get(ts_key) is not None:
2307 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
2308 # see http://bugs.python.org/issue1646728)
2309 with contextlib.suppress(ValueError, OverflowError, OSError):
2310 upload_date = datetime.datetime.utcfromtimestamp(info_dict[ts_key])
2311 info_dict[date_key] = upload_date.strftime('%Y%m%d')
2312
2313 live_keys = ('is_live', 'was_live')
2314 live_status = info_dict.get('live_status')
2315 if live_status is None:
2316 for key in live_keys:
2317 if info_dict.get(key) is False:
2318 continue
2319 if info_dict.get(key):
2320 live_status = key
2321 break
2322 if all(info_dict.get(key) is False for key in live_keys):
2323 live_status = 'not_live'
2324 if live_status:
2325 info_dict['live_status'] = live_status
2326 for key in live_keys:
2327 if info_dict.get(key) is None:
2328 info_dict[key] = (live_status == key)
2329
2330 # Auto generate title fields corresponding to the *_number fields when missing
2331 # in order to always have clean titles. This is very common for TV series.
2332 for field in ('chapter', 'season', 'episode'):
2333 if info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
2334 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
2335
2336 def _raise_pending_errors(self, info):
2337 err = info.pop('__pending_error', None)
2338 if err:
2339 self.report_error(err, tb=False)
2340
2341 def process_video_result(self, info_dict, download=True):
2342 assert info_dict.get('_type', 'video') == 'video'
2343 self._num_videos += 1
2344
2345 if 'id' not in info_dict:
2346 raise ExtractorError('Missing "id" field in extractor result', ie=info_dict['extractor'])
2347 elif not info_dict.get('id'):
2348 raise ExtractorError('Extractor failed to obtain "id"', ie=info_dict['extractor'])
2349
2350 def report_force_conversion(field, field_not, conversion):
2351 self.report_warning(
2352 '"%s" field is not %s - forcing %s conversion, there is an error in extractor'
2353 % (field, field_not, conversion))
2354
2355 def sanitize_string_field(info, string_field):
2356 field = info.get(string_field)
2357 if field is None or isinstance(field, str):
2358 return
2359 report_force_conversion(string_field, 'a string', 'string')
2360 info[string_field] = str(field)
2361
2362 def sanitize_numeric_fields(info):
2363 for numeric_field in self._NUMERIC_FIELDS:
2364 field = info.get(numeric_field)
2365 if field is None or isinstance(field, (int, float)):
2366 continue
2367 report_force_conversion(numeric_field, 'numeric', 'int')
2368 info[numeric_field] = int_or_none(field)
2369
2370 sanitize_string_field(info_dict, 'id')
2371 sanitize_numeric_fields(info_dict)
2372 if info_dict.get('section_end') and info_dict.get('section_start') is not None:
2373 info_dict['duration'] = round(info_dict['section_end'] - info_dict['section_start'], 3)
2374 if (info_dict.get('duration') or 0) <= 0 and info_dict.pop('duration', None):
2375 self.report_warning('"duration" field is negative, there is an error in extractor')
2376
2377 chapters = info_dict.get('chapters') or []
2378 dummy_chapter = {'end_time': 0, 'start_time': info_dict.get('duration')}
2379 for prev, current, next_ in zip(
2380 (dummy_chapter, *chapters), chapters, (*chapters[1:], dummy_chapter)):
2381 if current.get('start_time') is None:
2382 current['start_time'] = prev.get('end_time')
2383 if not current.get('end_time'):
2384 current['end_time'] = next_.get('start_time')
2385
2386 if 'playlist' not in info_dict:
2387 # It isn't part of a playlist
2388 info_dict['playlist'] = None
2389 info_dict['playlist_index'] = None
2390
2391 self._sanitize_thumbnails(info_dict)
2392
2393 thumbnail = info_dict.get('thumbnail')
2394 thumbnails = info_dict.get('thumbnails')
2395 if thumbnail:
2396 info_dict['thumbnail'] = sanitize_url(thumbnail)
2397 elif thumbnails:
2398 info_dict['thumbnail'] = thumbnails[-1]['url']
2399
2400 if info_dict.get('display_id') is None and 'id' in info_dict:
2401 info_dict['display_id'] = info_dict['id']
2402
2403 self._fill_common_fields(info_dict)
2404
2405 for cc_kind in ('subtitles', 'automatic_captions'):
2406 cc = info_dict.get(cc_kind)
2407 if cc:
2408 for _, subtitle in cc.items():
2409 for subtitle_format in subtitle:
2410 if subtitle_format.get('url'):
2411 subtitle_format['url'] = sanitize_url(subtitle_format['url'])
2412 if subtitle_format.get('ext') is None:
2413 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
2414
2415 automatic_captions = info_dict.get('automatic_captions')
2416 subtitles = info_dict.get('subtitles')
2417
2418 info_dict['requested_subtitles'] = self.process_subtitles(
2419 info_dict['id'], subtitles, automatic_captions)
2420
2421 if info_dict.get('formats') is None:
2422 # There's only one format available
2423 formats = [info_dict]
2424 else:
2425 formats = info_dict['formats']
2426
2427 # or None ensures --clean-infojson removes it
2428 info_dict['_has_drm'] = any(f.get('has_drm') for f in formats) or None
2429 if not self.params.get('allow_unplayable_formats'):
2430 formats = [f for f in formats if not f.get('has_drm')]
2431 if info_dict['_has_drm'] and all(
2432 f.get('acodec') == f.get('vcodec') == 'none' for f in formats):
2433 self.report_warning(
2434 'This video is DRM protected and only images are available for download. '
2435 'Use --list-formats to see them')
2436
2437 get_from_start = not info_dict.get('is_live') or bool(self.params.get('live_from_start'))
2438 if not get_from_start:
2439 info_dict['title'] += ' ' + datetime.datetime.now().strftime('%Y-%m-%d %H:%M')
2440 if info_dict.get('is_live') and formats:
2441 formats = [f for f in formats if bool(f.get('is_from_start')) == get_from_start]
2442 if get_from_start and not formats:
2443 self.raise_no_formats(info_dict, msg=(
2444 '--live-from-start is passed, but there are no formats that can be downloaded from the start. '
2445 'If you want to download from the current time, use --no-live-from-start'))
2446
2447 if not formats:
2448 self.raise_no_formats(info_dict)
2449
2450 def is_wellformed(f):
2451 url = f.get('url')
2452 if not url:
2453 self.report_warning(
2454 '"url" field is missing or empty - skipping format, '
2455 'there is an error in extractor')
2456 return False
2457 if isinstance(url, bytes):
2458 sanitize_string_field(f, 'url')
2459 return True
2460
2461 # Filter out malformed formats for better extraction robustness
2462 formats = list(filter(is_wellformed, formats))
2463
2464 formats_dict = {}
2465
2466 # We check that all the formats have the format and format_id fields
2467 for i, format in enumerate(formats):
2468 sanitize_string_field(format, 'format_id')
2469 sanitize_numeric_fields(format)
2470 format['url'] = sanitize_url(format['url'])
2471 if not format.get('format_id'):
2472 format['format_id'] = str(i)
2473 else:
2474 # Sanitize format_id from characters used in format selector expression
2475 format['format_id'] = re.sub(r'[\s,/+\[\]()]', '_', format['format_id'])
2476 format_id = format['format_id']
2477 if format_id not in formats_dict:
2478 formats_dict[format_id] = []
2479 formats_dict[format_id].append(format)
2480
2481 # Make sure all formats have unique format_id
2482 common_exts = set(itertools.chain(*self._format_selection_exts.values()))
2483 for format_id, ambiguous_formats in formats_dict.items():
2484 ambigious_id = len(ambiguous_formats) > 1
2485 for i, format in enumerate(ambiguous_formats):
2486 if ambigious_id:
2487 format['format_id'] = '%s-%d' % (format_id, i)
2488 if format.get('ext') is None:
2489 format['ext'] = determine_ext(format['url']).lower()
2490 # Ensure there is no conflict between id and ext in format selection
2491 # See https://github.com/yt-dlp/yt-dlp/issues/1282
2492 if format['format_id'] != format['ext'] and format['format_id'] in common_exts:
2493 format['format_id'] = 'f%s' % format['format_id']
2494
2495 for i, format in enumerate(formats):
2496 if format.get('format') is None:
2497 format['format'] = '{id} - {res}{note}'.format(
2498 id=format['format_id'],
2499 res=self.format_resolution(format),
2500 note=format_field(format, 'format_note', ' (%s)'),
2501 )
2502 if format.get('protocol') is None:
2503 format['protocol'] = determine_protocol(format)
2504 if format.get('resolution') is None:
2505 format['resolution'] = self.format_resolution(format, default=None)
2506 if format.get('dynamic_range') is None and format.get('vcodec') != 'none':
2507 format['dynamic_range'] = 'SDR'
2508 if (info_dict.get('duration') and format.get('tbr')
2509 and not format.get('filesize') and not format.get('filesize_approx')):
2510 format['filesize_approx'] = int(info_dict['duration'] * format['tbr'] * (1024 / 8))
2511
2512 # Add HTTP headers, so that external programs can use them from the
2513 # json output
2514 full_format_info = info_dict.copy()
2515 full_format_info.update(format)
2516 format['http_headers'] = self._calc_headers(full_format_info)
2517 # Remove private housekeeping stuff
2518 if '__x_forwarded_for_ip' in info_dict:
2519 del info_dict['__x_forwarded_for_ip']
2520
2521 if self.params.get('check_formats') is True:
2522 formats = LazyList(self._check_formats(formats[::-1]), reverse=True)
2523
2524 if not formats or formats[0] is not info_dict:
2525 # only set the 'formats' fields if the original info_dict list them
2526 # otherwise we end up with a circular reference, the first (and unique)
2527 # element in the 'formats' field in info_dict is info_dict itself,
2528 # which can't be exported to json
2529 info_dict['formats'] = formats
2530
2531 info_dict, _ = self.pre_process(info_dict)
2532
2533 if self._match_entry(info_dict, incomplete=self._format_fields) is not None:
2534 return info_dict
2535
2536 self.post_extract(info_dict)
2537 info_dict, _ = self.pre_process(info_dict, 'after_filter')
2538
2539 # The pre-processors may have modified the formats
2540 formats = info_dict.get('formats', [info_dict])
2541
2542 list_only = self.params.get('simulate') is None and (
2543 self.params.get('list_thumbnails') or self.params.get('listformats') or self.params.get('listsubtitles'))
2544 interactive_format_selection = not list_only and self.format_selector == '-'
2545 if self.params.get('list_thumbnails'):
2546 self.list_thumbnails(info_dict)
2547 if self.params.get('listsubtitles'):
2548 if 'automatic_captions' in info_dict:
2549 self.list_subtitles(
2550 info_dict['id'], automatic_captions, 'automatic captions')
2551 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
2552 if self.params.get('listformats') or interactive_format_selection:
2553 self.list_formats(info_dict)
2554 if list_only:
2555 # Without this printing, -F --print-json will not work
2556 self.__forced_printings(info_dict, self.prepare_filename(info_dict), incomplete=True)
2557 return info_dict
2558
2559 format_selector = self.format_selector
2560 if format_selector is None:
2561 req_format = self._default_format_spec(info_dict, download=download)
2562 self.write_debug('Default format spec: %s' % req_format)
2563 format_selector = self.build_format_selector(req_format)
2564
2565 while True:
2566 if interactive_format_selection:
2567 req_format = input(
2568 self._format_screen('\nEnter format selector: ', self.Styles.EMPHASIS))
2569 try:
2570 format_selector = self.build_format_selector(req_format)
2571 except SyntaxError as err:
2572 self.report_error(err, tb=False, is_error=False)
2573 continue
2574
2575 formats_to_download = list(format_selector({
2576 'formats': formats,
2577 'has_merged_format': any('none' not in (f.get('acodec'), f.get('vcodec')) for f in formats),
2578 'incomplete_formats': (
2579 # All formats are video-only or
2580 all(f.get('vcodec') != 'none' and f.get('acodec') == 'none' for f in formats)
2581 # all formats are audio-only
2582 or all(f.get('vcodec') == 'none' and f.get('acodec') != 'none' for f in formats)),
2583 }))
2584 if interactive_format_selection and not formats_to_download:
2585 self.report_error('Requested format is not available', tb=False, is_error=False)
2586 continue
2587 break
2588
2589 if not formats_to_download:
2590 if not self.params.get('ignore_no_formats_error'):
2591 raise ExtractorError(
2592 'Requested format is not available. Use --list-formats for a list of available formats',
2593 expected=True, video_id=info_dict['id'], ie=info_dict['extractor'])
2594 self.report_warning('Requested format is not available')
2595 # Process what we can, even without any available formats.
2596 formats_to_download = [{}]
2597
2598 requested_ranges = self.params.get('download_ranges')
2599 if requested_ranges:
2600 requested_ranges = tuple(requested_ranges(info_dict, self))
2601
2602 best_format, downloaded_formats = formats_to_download[-1], []
2603 if download:
2604 if best_format:
2605 def to_screen(*msg):
2606 self.to_screen(f'[info] {info_dict["id"]}: {" ".join(", ".join(variadic(m)) for m in msg)}')
2607
2608 to_screen(f'Downloading {len(formats_to_download)} format(s):',
2609 (f['format_id'] for f in formats_to_download))
2610 if requested_ranges:
2611 to_screen(f'Downloading {len(requested_ranges)} time ranges:',
2612 (f'{int(c["start_time"])}-{int(c["end_time"])}' for c in requested_ranges))
2613 max_downloads_reached = False
2614
2615 for fmt, chapter in itertools.product(formats_to_download, requested_ranges or [{}]):
2616 new_info = self._copy_infodict(info_dict)
2617 new_info.update(fmt)
2618 offset, duration = info_dict.get('section_start') or 0, info_dict.get('duration') or float('inf')
2619 if chapter or offset:
2620 new_info.update({
2621 'section_start': offset + chapter.get('start_time', 0),
2622 'section_end': offset + min(chapter.get('end_time', duration), duration),
2623 'section_title': chapter.get('title'),
2624 'section_number': chapter.get('index'),
2625 })
2626 downloaded_formats.append(new_info)
2627 try:
2628 self.process_info(new_info)
2629 except MaxDownloadsReached:
2630 max_downloads_reached = True
2631 self._raise_pending_errors(new_info)
2632 # Remove copied info
2633 for key, val in tuple(new_info.items()):
2634 if info_dict.get(key) == val:
2635 new_info.pop(key)
2636 if max_downloads_reached:
2637 break
2638
2639 write_archive = {f.get('__write_download_archive', False) for f in downloaded_formats}
2640 assert write_archive.issubset({True, False, 'ignore'})
2641 if True in write_archive and False not in write_archive:
2642 self.record_download_archive(info_dict)
2643
2644 info_dict['requested_downloads'] = downloaded_formats
2645 info_dict = self.run_all_pps('after_video', info_dict)
2646 if max_downloads_reached:
2647 raise MaxDownloadsReached()
2648
2649 # We update the info dict with the selected best quality format (backwards compatibility)
2650 info_dict.update(best_format)
2651 return info_dict
2652
2653 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
2654 """Select the requested subtitles and their format"""
2655 available_subs, normal_sub_langs = {}, []
2656 if normal_subtitles and self.params.get('writesubtitles'):
2657 available_subs.update(normal_subtitles)
2658 normal_sub_langs = tuple(normal_subtitles.keys())
2659 if automatic_captions and self.params.get('writeautomaticsub'):
2660 for lang, cap_info in automatic_captions.items():
2661 if lang not in available_subs:
2662 available_subs[lang] = cap_info
2663
2664 if (not self.params.get('writesubtitles') and not
2665 self.params.get('writeautomaticsub') or not
2666 available_subs):
2667 return None
2668
2669 all_sub_langs = tuple(available_subs.keys())
2670 if self.params.get('allsubtitles', False):
2671 requested_langs = all_sub_langs
2672 elif self.params.get('subtitleslangs', False):
2673 # A list is used so that the order of languages will be the same as
2674 # given in subtitleslangs. See https://github.com/yt-dlp/yt-dlp/issues/1041
2675 requested_langs = []
2676 for lang_re in self.params.get('subtitleslangs'):
2677 discard = lang_re[0] == '-'
2678 if discard:
2679 lang_re = lang_re[1:]
2680 if lang_re == 'all':
2681 if discard:
2682 requested_langs = []
2683 else:
2684 requested_langs.extend(all_sub_langs)
2685 continue
2686 current_langs = filter(re.compile(lang_re + '$').match, all_sub_langs)
2687 if discard:
2688 for lang in current_langs:
2689 while lang in requested_langs:
2690 requested_langs.remove(lang)
2691 else:
2692 requested_langs.extend(current_langs)
2693 requested_langs = orderedSet(requested_langs)
2694 elif normal_sub_langs:
2695 requested_langs = ['en'] if 'en' in normal_sub_langs else normal_sub_langs[:1]
2696 else:
2697 requested_langs = ['en'] if 'en' in all_sub_langs else all_sub_langs[:1]
2698 if requested_langs:
2699 self.write_debug('Downloading subtitles: %s' % ', '.join(requested_langs))
2700
2701 formats_query = self.params.get('subtitlesformat', 'best')
2702 formats_preference = formats_query.split('/') if formats_query else []
2703 subs = {}
2704 for lang in requested_langs:
2705 formats = available_subs.get(lang)
2706 if formats is None:
2707 self.report_warning(f'{lang} subtitles not available for {video_id}')
2708 continue
2709 for ext in formats_preference:
2710 if ext == 'best':
2711 f = formats[-1]
2712 break
2713 matches = list(filter(lambda f: f['ext'] == ext, formats))
2714 if matches:
2715 f = matches[-1]
2716 break
2717 else:
2718 f = formats[-1]
2719 self.report_warning(
2720 'No subtitle format found matching "%s" for language %s, '
2721 'using %s' % (formats_query, lang, f['ext']))
2722 subs[lang] = f
2723 return subs
2724
2725 def _forceprint(self, key, info_dict):
2726 if info_dict is None:
2727 return
2728 info_copy = info_dict.copy()
2729 info_copy['formats_table'] = self.render_formats_table(info_dict)
2730 info_copy['thumbnails_table'] = self.render_thumbnails_table(info_dict)
2731 info_copy['subtitles_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('subtitles'))
2732 info_copy['automatic_captions_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('automatic_captions'))
2733
2734 def format_tmpl(tmpl):
2735 mobj = re.match(r'\w+(=?)$', tmpl)
2736 if mobj and mobj.group(1):
2737 return f'{tmpl[:-1]} = %({tmpl[:-1]})r'
2738 elif mobj:
2739 return f'%({tmpl})s'
2740 return tmpl
2741
2742 for tmpl in self.params['forceprint'].get(key, []):
2743 self.to_stdout(self.evaluate_outtmpl(format_tmpl(tmpl), info_copy))
2744
2745 for tmpl, file_tmpl in self.params['print_to_file'].get(key, []):
2746 filename = self.prepare_filename(info_dict, outtmpl=file_tmpl)
2747 tmpl = format_tmpl(tmpl)
2748 self.to_screen(f'[info] Writing {tmpl!r} to: {filename}')
2749 if self._ensure_dir_exists(filename):
2750 with open(filename, 'a', encoding='utf-8') as f:
2751 f.write(self.evaluate_outtmpl(tmpl, info_copy) + '\n')
2752
2753 def __forced_printings(self, info_dict, filename, incomplete):
2754 def print_mandatory(field, actual_field=None):
2755 if actual_field is None:
2756 actual_field = field
2757 if (self.params.get('force%s' % field, False)
2758 and (not incomplete or info_dict.get(actual_field) is not None)):
2759 self.to_stdout(info_dict[actual_field])
2760
2761 def print_optional(field):
2762 if (self.params.get('force%s' % field, False)
2763 and info_dict.get(field) is not None):
2764 self.to_stdout(info_dict[field])
2765
2766 info_dict = info_dict.copy()
2767 if filename is not None:
2768 info_dict['filename'] = filename
2769 if info_dict.get('requested_formats') is not None:
2770 # For RTMP URLs, also include the playpath
2771 info_dict['urls'] = '\n'.join(f['url'] + f.get('play_path', '') for f in info_dict['requested_formats'])
2772 elif info_dict.get('url'):
2773 info_dict['urls'] = info_dict['url'] + info_dict.get('play_path', '')
2774
2775 if (self.params.get('forcejson')
2776 or self.params['forceprint'].get('video')
2777 or self.params['print_to_file'].get('video')):
2778 self.post_extract(info_dict)
2779 self._forceprint('video', info_dict)
2780
2781 print_mandatory('title')
2782 print_mandatory('id')
2783 print_mandatory('url', 'urls')
2784 print_optional('thumbnail')
2785 print_optional('description')
2786 print_optional('filename')
2787 if self.params.get('forceduration') and info_dict.get('duration') is not None:
2788 self.to_stdout(formatSeconds(info_dict['duration']))
2789 print_mandatory('format')
2790
2791 if self.params.get('forcejson'):
2792 self.to_stdout(json.dumps(self.sanitize_info(info_dict)))
2793
2794 def dl(self, name, info, subtitle=False, test=False):
2795 if not info.get('url'):
2796 self.raise_no_formats(info, True)
2797
2798 if test:
2799 verbose = self.params.get('verbose')
2800 params = {
2801 'test': True,
2802 'quiet': self.params.get('quiet') or not verbose,
2803 'verbose': verbose,
2804 'noprogress': not verbose,
2805 'nopart': True,
2806 'skip_unavailable_fragments': False,
2807 'keep_fragments': False,
2808 'overwrites': True,
2809 '_no_ytdl_file': True,
2810 }
2811 else:
2812 params = self.params
2813 fd = get_suitable_downloader(info, params, to_stdout=(name == '-'))(self, params)
2814 if not test:
2815 for ph in self._progress_hooks:
2816 fd.add_progress_hook(ph)
2817 urls = '", "'.join(
2818 (f['url'].split(',')[0] + ',<data>' if f['url'].startswith('data:') else f['url'])
2819 for f in info.get('requested_formats', []) or [info])
2820 self.write_debug(f'Invoking {fd.FD_NAME} downloader on "{urls}"')
2821
2822 # Note: Ideally info should be a deep-copied so that hooks cannot modify it.
2823 # But it may contain objects that are not deep-copyable
2824 new_info = self._copy_infodict(info)
2825 if new_info.get('http_headers') is None:
2826 new_info['http_headers'] = self._calc_headers(new_info)
2827 return fd.download(name, new_info, subtitle)
2828
2829 def existing_file(self, filepaths, *, default_overwrite=True):
2830 existing_files = list(filter(os.path.exists, orderedSet(filepaths)))
2831 if existing_files and not self.params.get('overwrites', default_overwrite):
2832 return existing_files[0]
2833
2834 for file in existing_files:
2835 self.report_file_delete(file)
2836 os.remove(file)
2837 return None
2838
2839 def process_info(self, info_dict):
2840 """Process a single resolved IE result. (Modifies it in-place)"""
2841
2842 assert info_dict.get('_type', 'video') == 'video'
2843 original_infodict = info_dict
2844
2845 if 'format' not in info_dict and 'ext' in info_dict:
2846 info_dict['format'] = info_dict['ext']
2847
2848 # This is mostly just for backward compatibility of process_info
2849 # As a side-effect, this allows for format-specific filters
2850 if self._match_entry(info_dict) is not None:
2851 info_dict['__write_download_archive'] = 'ignore'
2852 return
2853
2854 # Does nothing under normal operation - for backward compatibility of process_info
2855 self.post_extract(info_dict)
2856 self._num_downloads += 1
2857
2858 # info_dict['_filename'] needs to be set for backward compatibility
2859 info_dict['_filename'] = full_filename = self.prepare_filename(info_dict, warn=True)
2860 temp_filename = self.prepare_filename(info_dict, 'temp')
2861 files_to_move = {}
2862
2863 # Forced printings
2864 self.__forced_printings(info_dict, full_filename, incomplete=('format' not in info_dict))
2865
2866 def check_max_downloads():
2867 if self._num_downloads >= float(self.params.get('max_downloads') or 'inf'):
2868 raise MaxDownloadsReached()
2869
2870 if self.params.get('simulate'):
2871 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
2872 check_max_downloads()
2873 return
2874
2875 if full_filename is None:
2876 return
2877 if not self._ensure_dir_exists(encodeFilename(full_filename)):
2878 return
2879 if not self._ensure_dir_exists(encodeFilename(temp_filename)):
2880 return
2881
2882 if self._write_description('video', info_dict,
2883 self.prepare_filename(info_dict, 'description')) is None:
2884 return
2885
2886 sub_files = self._write_subtitles(info_dict, temp_filename)
2887 if sub_files is None:
2888 return
2889 files_to_move.update(dict(sub_files))
2890
2891 thumb_files = self._write_thumbnails(
2892 'video', info_dict, temp_filename, self.prepare_filename(info_dict, 'thumbnail'))
2893 if thumb_files is None:
2894 return
2895 files_to_move.update(dict(thumb_files))
2896
2897 infofn = self.prepare_filename(info_dict, 'infojson')
2898 _infojson_written = self._write_info_json('video', info_dict, infofn)
2899 if _infojson_written:
2900 info_dict['infojson_filename'] = infofn
2901 # For backward compatibility, even though it was a private field
2902 info_dict['__infojson_filename'] = infofn
2903 elif _infojson_written is None:
2904 return
2905
2906 # Note: Annotations are deprecated
2907 annofn = None
2908 if self.params.get('writeannotations', False):
2909 annofn = self.prepare_filename(info_dict, 'annotation')
2910 if annofn:
2911 if not self._ensure_dir_exists(encodeFilename(annofn)):
2912 return
2913 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(annofn)):
2914 self.to_screen('[info] Video annotations are already present')
2915 elif not info_dict.get('annotations'):
2916 self.report_warning('There are no annotations to write.')
2917 else:
2918 try:
2919 self.to_screen('[info] Writing video annotations to: ' + annofn)
2920 with open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
2921 annofile.write(info_dict['annotations'])
2922 except (KeyError, TypeError):
2923 self.report_warning('There are no annotations to write.')
2924 except OSError:
2925 self.report_error('Cannot write annotations file: ' + annofn)
2926 return
2927
2928 # Write internet shortcut files
2929 def _write_link_file(link_type):
2930 url = try_get(info_dict['webpage_url'], iri_to_uri)
2931 if not url:
2932 self.report_warning(
2933 f'Cannot write internet shortcut file because the actual URL of "{info_dict["webpage_url"]}" is unknown')
2934 return True
2935 linkfn = replace_extension(self.prepare_filename(info_dict, 'link'), link_type, info_dict.get('ext'))
2936 if not self._ensure_dir_exists(encodeFilename(linkfn)):
2937 return False
2938 if self.params.get('overwrites', True) and os.path.exists(encodeFilename(linkfn)):
2939 self.to_screen(f'[info] Internet shortcut (.{link_type}) is already present')
2940 return True
2941 try:
2942 self.to_screen(f'[info] Writing internet shortcut (.{link_type}) to: {linkfn}')
2943 with open(encodeFilename(to_high_limit_path(linkfn)), 'w', encoding='utf-8',
2944 newline='\r\n' if link_type == 'url' else '\n') as linkfile:
2945 template_vars = {'url': url}
2946 if link_type == 'desktop':
2947 template_vars['filename'] = linkfn[:-(len(link_type) + 1)]
2948 linkfile.write(LINK_TEMPLATES[link_type] % template_vars)
2949 except OSError:
2950 self.report_error(f'Cannot write internet shortcut {linkfn}')
2951 return False
2952 return True
2953
2954 write_links = {
2955 'url': self.params.get('writeurllink'),
2956 'webloc': self.params.get('writewebloclink'),
2957 'desktop': self.params.get('writedesktoplink'),
2958 }
2959 if self.params.get('writelink'):
2960 link_type = ('webloc' if sys.platform == 'darwin'
2961 else 'desktop' if sys.platform.startswith('linux')
2962 else 'url')
2963 write_links[link_type] = True
2964
2965 if any(should_write and not _write_link_file(link_type)
2966 for link_type, should_write in write_links.items()):
2967 return
2968
2969 def replace_info_dict(new_info):
2970 nonlocal info_dict
2971 if new_info == info_dict:
2972 return
2973 info_dict.clear()
2974 info_dict.update(new_info)
2975
2976 new_info, files_to_move = self.pre_process(info_dict, 'before_dl', files_to_move)
2977 replace_info_dict(new_info)
2978
2979 if self.params.get('skip_download'):
2980 info_dict['filepath'] = temp_filename
2981 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
2982 info_dict['__files_to_move'] = files_to_move
2983 replace_info_dict(self.run_pp(MoveFilesAfterDownloadPP(self, False), info_dict))
2984 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
2985 else:
2986 # Download
2987 info_dict.setdefault('__postprocessors', [])
2988 try:
2989
2990 def existing_video_file(*filepaths):
2991 ext = info_dict.get('ext')
2992 converted = lambda file: replace_extension(file, self.params.get('final_ext') or ext, ext)
2993 file = self.existing_file(itertools.chain(*zip(map(converted, filepaths), filepaths)),
2994 default_overwrite=False)
2995 if file:
2996 info_dict['ext'] = os.path.splitext(file)[1][1:]
2997 return file
2998
2999 fd, success = None, True
3000 if info_dict.get('protocol') or info_dict.get('url'):
3001 fd = get_suitable_downloader(info_dict, self.params, to_stdout=temp_filename == '-')
3002 if fd is not FFmpegFD and (
3003 info_dict.get('section_start') or info_dict.get('section_end')):
3004 msg = ('This format cannot be partially downloaded' if FFmpegFD.available()
3005 else 'You have requested downloading the video partially, but ffmpeg is not installed')
3006 self.report_error(f'{msg}. Aborting')
3007 return
3008
3009 if info_dict.get('requested_formats') is not None:
3010
3011 def compatible_formats(formats):
3012 # TODO: some formats actually allow this (mkv, webm, ogg, mp4), but not all of them.
3013 video_formats = [format for format in formats if format.get('vcodec') != 'none']
3014 audio_formats = [format for format in formats if format.get('acodec') != 'none']
3015 if len(video_formats) > 2 or len(audio_formats) > 2:
3016 return False
3017
3018 # Check extension
3019 exts = {format.get('ext') for format in formats}
3020 COMPATIBLE_EXTS = (
3021 {'mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'ismv', 'isma'},
3022 {'webm'},
3023 )
3024 for ext_sets in COMPATIBLE_EXTS:
3025 if ext_sets.issuperset(exts):
3026 return True
3027 # TODO: Check acodec/vcodec
3028 return False
3029
3030 requested_formats = info_dict['requested_formats']
3031 old_ext = info_dict['ext']
3032 if self.params.get('merge_output_format') is None:
3033 if not compatible_formats(requested_formats):
3034 info_dict['ext'] = 'mkv'
3035 self.report_warning(
3036 'Requested formats are incompatible for merge and will be merged into mkv')
3037 if (info_dict['ext'] == 'webm'
3038 and info_dict.get('thumbnails')
3039 # check with type instead of pp_key, __name__, or isinstance
3040 # since we dont want any custom PPs to trigger this
3041 and any(type(pp) == EmbedThumbnailPP for pp in self._pps['post_process'])): # noqa: E721
3042 info_dict['ext'] = 'mkv'
3043 self.report_warning(
3044 'webm doesn\'t support embedding a thumbnail, mkv will be used')
3045 new_ext = info_dict['ext']
3046
3047 def correct_ext(filename, ext=new_ext):
3048 if filename == '-':
3049 return filename
3050 filename_real_ext = os.path.splitext(filename)[1][1:]
3051 filename_wo_ext = (
3052 os.path.splitext(filename)[0]
3053 if filename_real_ext in (old_ext, new_ext)
3054 else filename)
3055 return f'{filename_wo_ext}.{ext}'
3056
3057 # Ensure filename always has a correct extension for successful merge
3058 full_filename = correct_ext(full_filename)
3059 temp_filename = correct_ext(temp_filename)
3060 dl_filename = existing_video_file(full_filename, temp_filename)
3061 info_dict['__real_download'] = False
3062
3063 merger = FFmpegMergerPP(self)
3064 downloaded = []
3065 if dl_filename is not None:
3066 self.report_file_already_downloaded(dl_filename)
3067 elif fd:
3068 for f in requested_formats if fd != FFmpegFD else []:
3069 f['filepath'] = fname = prepend_extension(
3070 correct_ext(temp_filename, info_dict['ext']),
3071 'f%s' % f['format_id'], info_dict['ext'])
3072 downloaded.append(fname)
3073 info_dict['url'] = '\n'.join(f['url'] for f in requested_formats)
3074 success, real_download = self.dl(temp_filename, info_dict)
3075 info_dict['__real_download'] = real_download
3076 else:
3077 if self.params.get('allow_unplayable_formats'):
3078 self.report_warning(
3079 'You have requested merging of multiple formats '
3080 'while also allowing unplayable formats to be downloaded. '
3081 'The formats won\'t be merged to prevent data corruption.')
3082 elif not merger.available:
3083 msg = 'You have requested merging of multiple formats but ffmpeg is not installed'
3084 if not self.params.get('ignoreerrors'):
3085 self.report_error(f'{msg}. Aborting due to --abort-on-error')
3086 return
3087 self.report_warning(f'{msg}. The formats won\'t be merged')
3088
3089 if temp_filename == '-':
3090 reason = ('using a downloader other than ffmpeg' if FFmpegFD.can_merge_formats(info_dict, self.params)
3091 else 'but the formats are incompatible for simultaneous download' if merger.available
3092 else 'but ffmpeg is not installed')
3093 self.report_warning(
3094 f'You have requested downloading multiple formats to stdout {reason}. '
3095 'The formats will be streamed one after the other')
3096 fname = temp_filename
3097 for f in requested_formats:
3098 new_info = dict(info_dict)
3099 del new_info['requested_formats']
3100 new_info.update(f)
3101 if temp_filename != '-':
3102 fname = prepend_extension(
3103 correct_ext(temp_filename, new_info['ext']),
3104 'f%s' % f['format_id'], new_info['ext'])
3105 if not self._ensure_dir_exists(fname):
3106 return
3107 f['filepath'] = fname
3108 downloaded.append(fname)
3109 partial_success, real_download = self.dl(fname, new_info)
3110 info_dict['__real_download'] = info_dict['__real_download'] or real_download
3111 success = success and partial_success
3112
3113 if downloaded and merger.available and not self.params.get('allow_unplayable_formats'):
3114 info_dict['__postprocessors'].append(merger)
3115 info_dict['__files_to_merge'] = downloaded
3116 # Even if there were no downloads, it is being merged only now
3117 info_dict['__real_download'] = True
3118 else:
3119 for file in downloaded:
3120 files_to_move[file] = None
3121 else:
3122 # Just a single file
3123 dl_filename = existing_video_file(full_filename, temp_filename)
3124 if dl_filename is None or dl_filename == temp_filename:
3125 # dl_filename == temp_filename could mean that the file was partially downloaded with --no-part.
3126 # So we should try to resume the download
3127 success, real_download = self.dl(temp_filename, info_dict)
3128 info_dict['__real_download'] = real_download
3129 else:
3130 self.report_file_already_downloaded(dl_filename)
3131
3132 dl_filename = dl_filename or temp_filename
3133 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
3134
3135 except network_exceptions as err:
3136 self.report_error('unable to download video data: %s' % error_to_compat_str(err))
3137 return
3138 except OSError as err:
3139 raise UnavailableVideoError(err)
3140 except (ContentTooShortError, ) as err:
3141 self.report_error(f'content too short (expected {err.expected} bytes and served {err.downloaded})')
3142 return
3143
3144 self._raise_pending_errors(info_dict)
3145 if success and full_filename != '-':
3146
3147 def fixup():
3148 do_fixup = True
3149 fixup_policy = self.params.get('fixup')
3150 vid = info_dict['id']
3151
3152 if fixup_policy in ('ignore', 'never'):
3153 return
3154 elif fixup_policy == 'warn':
3155 do_fixup = 'warn'
3156 elif fixup_policy != 'force':
3157 assert fixup_policy in ('detect_or_warn', None)
3158 if not info_dict.get('__real_download'):
3159 do_fixup = False
3160
3161 def ffmpeg_fixup(cndn, msg, cls):
3162 if not (do_fixup and cndn):
3163 return
3164 elif do_fixup == 'warn':
3165 self.report_warning(f'{vid}: {msg}')
3166 return
3167 pp = cls(self)
3168 if pp.available:
3169 info_dict['__postprocessors'].append(pp)
3170 else:
3171 self.report_warning(f'{vid}: {msg}. Install ffmpeg to fix this automatically')
3172
3173 stretched_ratio = info_dict.get('stretched_ratio')
3174 ffmpeg_fixup(
3175 stretched_ratio not in (1, None),
3176 f'Non-uniform pixel ratio {stretched_ratio}',
3177 FFmpegFixupStretchedPP)
3178
3179 ffmpeg_fixup(
3180 (info_dict.get('requested_formats') is None
3181 and info_dict.get('container') == 'm4a_dash'
3182 and info_dict.get('ext') == 'm4a'),
3183 'writing DASH m4a. Only some players support this container',
3184 FFmpegFixupM4aPP)
3185
3186 downloader = get_suitable_downloader(info_dict, self.params) if 'protocol' in info_dict else None
3187 downloader = downloader.FD_NAME if downloader else None
3188
3189 if info_dict.get('requested_formats') is None: # Not necessary if doing merger
3190 ffmpeg_fixup(downloader == 'hlsnative' and not self.params.get('hls_use_mpegts')
3191 or info_dict.get('is_live') and self.params.get('hls_use_mpegts') is None,
3192 'Possible MPEG-TS in MP4 container or malformed AAC timestamps',
3193 FFmpegFixupM3u8PP)
3194 ffmpeg_fixup(info_dict.get('is_live') and downloader == 'DashSegmentsFD',
3195 'Possible duplicate MOOV atoms', FFmpegFixupDuplicateMoovPP)
3196
3197 ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed timestamps detected', FFmpegFixupTimestampPP)
3198 ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed duration detected', FFmpegFixupDurationPP)
3199
3200 fixup()
3201 try:
3202 replace_info_dict(self.post_process(dl_filename, info_dict, files_to_move))
3203 except PostProcessingError as err:
3204 self.report_error('Postprocessing: %s' % str(err))
3205 return
3206 try:
3207 for ph in self._post_hooks:
3208 ph(info_dict['filepath'])
3209 except Exception as err:
3210 self.report_error('post hooks: %s' % str(err))
3211 return
3212 info_dict['__write_download_archive'] = True
3213
3214 assert info_dict is original_infodict # Make sure the info_dict was modified in-place
3215 if self.params.get('force_write_download_archive'):
3216 info_dict['__write_download_archive'] = True
3217 check_max_downloads()
3218
3219 def __download_wrapper(self, func):
3220 @functools.wraps(func)
3221 def wrapper(*args, **kwargs):
3222 try:
3223 res = func(*args, **kwargs)
3224 except UnavailableVideoError as e:
3225 self.report_error(e)
3226 except DownloadCancelled as e:
3227 self.to_screen(f'[info] {e}')
3228 if not self.params.get('break_per_url'):
3229 raise
3230 else:
3231 if self.params.get('dump_single_json', False):
3232 self.post_extract(res)
3233 self.to_stdout(json.dumps(self.sanitize_info(res)))
3234 return wrapper
3235
3236 def download(self, url_list):
3237 """Download a given list of URLs."""
3238 url_list = variadic(url_list) # Passing a single URL is a common mistake
3239 outtmpl = self.params['outtmpl']['default']
3240 if (len(url_list) > 1
3241 and outtmpl != '-'
3242 and '%' not in outtmpl
3243 and self.params.get('max_downloads') != 1):
3244 raise SameFileError(outtmpl)
3245
3246 for url in url_list:
3247 self.__download_wrapper(self.extract_info)(
3248 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
3249
3250 return self._download_retcode
3251
3252 def download_with_info_file(self, info_filename):
3253 with contextlib.closing(fileinput.FileInput(
3254 [info_filename], mode='r',
3255 openhook=fileinput.hook_encoded('utf-8'))) as f:
3256 # FileInput doesn't have a read method, we can't call json.load
3257 info = self.sanitize_info(json.loads('\n'.join(f)), self.params.get('clean_infojson', True))
3258 try:
3259 self.__download_wrapper(self.process_ie_result)(info, download=True)
3260 except (DownloadError, EntryNotInPlaylist, ReExtractInfo) as e:
3261 if not isinstance(e, EntryNotInPlaylist):
3262 self.to_stderr('\r')
3263 webpage_url = info.get('webpage_url')
3264 if webpage_url is not None:
3265 self.report_warning(f'The info failed to download: {e}; trying with URL {webpage_url}')
3266 return self.download([webpage_url])
3267 else:
3268 raise
3269 return self._download_retcode
3270
3271 @staticmethod
3272 def sanitize_info(info_dict, remove_private_keys=False):
3273 ''' Sanitize the infodict for converting to json '''
3274 if info_dict is None:
3275 return info_dict
3276 info_dict.setdefault('epoch', int(time.time()))
3277 info_dict.setdefault('_type', 'video')
3278
3279 if remove_private_keys:
3280 reject = lambda k, v: v is None or k.startswith('__') or k in {
3281 'requested_downloads', 'requested_formats', 'requested_subtitles', 'requested_entries',
3282 'entries', 'filepath', '_filename', 'infojson_filename', 'original_url', 'playlist_autonumber',
3283 }
3284 else:
3285 reject = lambda k, v: False
3286
3287 def filter_fn(obj):
3288 if isinstance(obj, dict):
3289 return {k: filter_fn(v) for k, v in obj.items() if not reject(k, v)}
3290 elif isinstance(obj, (list, tuple, set, LazyList)):
3291 return list(map(filter_fn, obj))
3292 elif obj is None or isinstance(obj, (str, int, float, bool)):
3293 return obj
3294 else:
3295 return repr(obj)
3296
3297 return filter_fn(info_dict)
3298
3299 @staticmethod
3300 def filter_requested_info(info_dict, actually_filter=True):
3301 ''' Alias of sanitize_info for backward compatibility '''
3302 return YoutubeDL.sanitize_info(info_dict, actually_filter)
3303
3304 def _delete_downloaded_files(self, *files_to_delete, info={}, msg=None):
3305 for filename in set(filter(None, files_to_delete)):
3306 if msg:
3307 self.to_screen(msg % filename)
3308 try:
3309 os.remove(filename)
3310 except OSError:
3311 self.report_warning(f'Unable to delete file {filename}')
3312 if filename in info.get('__files_to_move', []): # NB: Delete even if None
3313 del info['__files_to_move'][filename]
3314
3315 @staticmethod
3316 def post_extract(info_dict):
3317 def actual_post_extract(info_dict):
3318 if info_dict.get('_type') in ('playlist', 'multi_video'):
3319 for video_dict in info_dict.get('entries', {}):
3320 actual_post_extract(video_dict or {})
3321 return
3322
3323 post_extractor = info_dict.pop('__post_extractor', None) or (lambda: {})
3324 info_dict.update(post_extractor())
3325
3326 actual_post_extract(info_dict or {})
3327
3328 def run_pp(self, pp, infodict):
3329 files_to_delete = []
3330 if '__files_to_move' not in infodict:
3331 infodict['__files_to_move'] = {}
3332 try:
3333 files_to_delete, infodict = pp.run(infodict)
3334 except PostProcessingError as e:
3335 # Must be True and not 'only_download'
3336 if self.params.get('ignoreerrors') is True:
3337 self.report_error(e)
3338 return infodict
3339 raise
3340
3341 if not files_to_delete:
3342 return infodict
3343 if self.params.get('keepvideo', False):
3344 for f in files_to_delete:
3345 infodict['__files_to_move'].setdefault(f, '')
3346 else:
3347 self._delete_downloaded_files(
3348 *files_to_delete, info=infodict, msg='Deleting original file %s (pass -k to keep)')
3349 return infodict
3350
3351 def run_all_pps(self, key, info, *, additional_pps=None):
3352 self._forceprint(key, info)
3353 for pp in (additional_pps or []) + self._pps[key]:
3354 info = self.run_pp(pp, info)
3355 return info
3356
3357 def pre_process(self, ie_info, key='pre_process', files_to_move=None):
3358 info = dict(ie_info)
3359 info['__files_to_move'] = files_to_move or {}
3360 try:
3361 info = self.run_all_pps(key, info)
3362 except PostProcessingError as err:
3363 msg = f'Preprocessing: {err}'
3364 info.setdefault('__pending_error', msg)
3365 self.report_error(msg, is_error=False)
3366 return info, info.pop('__files_to_move', None)
3367
3368 def post_process(self, filename, info, files_to_move=None):
3369 """Run all the postprocessors on the given file."""
3370 info['filepath'] = filename
3371 info['__files_to_move'] = files_to_move or {}
3372 info = self.run_all_pps('post_process', info, additional_pps=info.get('__postprocessors'))
3373 info = self.run_pp(MoveFilesAfterDownloadPP(self), info)
3374 del info['__files_to_move']
3375 return self.run_all_pps('after_move', info)
3376
3377 def _make_archive_id(self, info_dict):
3378 video_id = info_dict.get('id')
3379 if not video_id:
3380 return
3381 # Future-proof against any change in case
3382 # and backwards compatibility with prior versions
3383 extractor = info_dict.get('extractor_key') or info_dict.get('ie_key') # key in a playlist
3384 if extractor is None:
3385 url = str_or_none(info_dict.get('url'))
3386 if not url:
3387 return
3388 # Try to find matching extractor for the URL and take its ie_key
3389 for ie_key, ie in self._ies.items():
3390 if ie.suitable(url):
3391 extractor = ie_key
3392 break
3393 else:
3394 return
3395 return f'{extractor.lower()} {video_id}'
3396
3397 def in_download_archive(self, info_dict):
3398 fn = self.params.get('download_archive')
3399 if fn is None:
3400 return False
3401
3402 vid_id = self._make_archive_id(info_dict)
3403 if not vid_id:
3404 return False # Incomplete video information
3405
3406 return vid_id in self.archive
3407
3408 def record_download_archive(self, info_dict):
3409 fn = self.params.get('download_archive')
3410 if fn is None:
3411 return
3412 vid_id = self._make_archive_id(info_dict)
3413 assert vid_id
3414 self.write_debug(f'Adding to archive: {vid_id}')
3415 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
3416 archive_file.write(vid_id + '\n')
3417 self.archive.add(vid_id)
3418
3419 @staticmethod
3420 def format_resolution(format, default='unknown'):
3421 if format.get('vcodec') == 'none' and format.get('acodec') != 'none':
3422 return 'audio only'
3423 if format.get('resolution') is not None:
3424 return format['resolution']
3425 if format.get('width') and format.get('height'):
3426 return '%dx%d' % (format['width'], format['height'])
3427 elif format.get('height'):
3428 return '%sp' % format['height']
3429 elif format.get('width'):
3430 return '%dx?' % format['width']
3431 return default
3432
3433 def _list_format_headers(self, *headers):
3434 if self.params.get('listformats_table', True) is not False:
3435 return [self._format_out(header, self.Styles.HEADERS) for header in headers]
3436 return headers
3437
3438 def _format_note(self, fdict):
3439 res = ''
3440 if fdict.get('ext') in ['f4f', 'f4m']:
3441 res += '(unsupported)'
3442 if fdict.get('language'):
3443 if res:
3444 res += ' '
3445 res += '[%s]' % fdict['language']
3446 if fdict.get('format_note') is not None:
3447 if res:
3448 res += ' '
3449 res += fdict['format_note']
3450 if fdict.get('tbr') is not None:
3451 if res:
3452 res += ', '
3453 res += '%4dk' % fdict['tbr']
3454 if fdict.get('container') is not None:
3455 if res:
3456 res += ', '
3457 res += '%s container' % fdict['container']
3458 if (fdict.get('vcodec') is not None
3459 and fdict.get('vcodec') != 'none'):
3460 if res:
3461 res += ', '
3462 res += fdict['vcodec']
3463 if fdict.get('vbr') is not None:
3464 res += '@'
3465 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
3466 res += 'video@'
3467 if fdict.get('vbr') is not None:
3468 res += '%4dk' % fdict['vbr']
3469 if fdict.get('fps') is not None:
3470 if res:
3471 res += ', '
3472 res += '%sfps' % fdict['fps']
3473 if fdict.get('acodec') is not None:
3474 if res:
3475 res += ', '
3476 if fdict['acodec'] == 'none':
3477 res += 'video only'
3478 else:
3479 res += '%-5s' % fdict['acodec']
3480 elif fdict.get('abr') is not None:
3481 if res:
3482 res += ', '
3483 res += 'audio'
3484 if fdict.get('abr') is not None:
3485 res += '@%3dk' % fdict['abr']
3486 if fdict.get('asr') is not None:
3487 res += ' (%5dHz)' % fdict['asr']
3488 if fdict.get('filesize') is not None:
3489 if res:
3490 res += ', '
3491 res += format_bytes(fdict['filesize'])
3492 elif fdict.get('filesize_approx') is not None:
3493 if res:
3494 res += ', '
3495 res += '~' + format_bytes(fdict['filesize_approx'])
3496 return res
3497
3498 def render_formats_table(self, info_dict):
3499 if not info_dict.get('formats') and not info_dict.get('url'):
3500 return None
3501
3502 formats = info_dict.get('formats', [info_dict])
3503 if not self.params.get('listformats_table', True) is not False:
3504 table = [
3505 [
3506 format_field(f, 'format_id'),
3507 format_field(f, 'ext'),
3508 self.format_resolution(f),
3509 self._format_note(f)
3510 ] for f in formats if f.get('preference') is None or f['preference'] >= -1000]
3511 return render_table(['format code', 'extension', 'resolution', 'note'], table, extra_gap=1)
3512
3513 delim = self._format_out('\u2502', self.Styles.DELIM, '|', test_encoding=True)
3514 table = [
3515 [
3516 self._format_out(format_field(f, 'format_id'), self.Styles.ID),
3517 format_field(f, 'ext'),
3518 format_field(f, func=self.format_resolution, ignore=('audio only', 'images')),
3519 format_field(f, 'fps', '\t%d'),
3520 format_field(f, 'dynamic_range', '%s', ignore=(None, 'SDR')).replace('HDR', ''),
3521 delim,
3522 format_field(f, 'filesize', ' \t%s', func=format_bytes) + format_field(f, 'filesize_approx', '~\t%s', func=format_bytes),
3523 format_field(f, 'tbr', '\t%dk'),
3524 shorten_protocol_name(f.get('protocol', '')),
3525 delim,
3526 format_field(f, 'vcodec', default='unknown').replace(
3527 'none', 'images' if f.get('acodec') == 'none'
3528 else self._format_out('audio only', self.Styles.SUPPRESS)),
3529 format_field(f, 'vbr', '\t%dk'),
3530 format_field(f, 'acodec', default='unknown').replace(
3531 'none', '' if f.get('vcodec') == 'none'
3532 else self._format_out('video only', self.Styles.SUPPRESS)),
3533 format_field(f, 'abr', '\t%dk'),
3534 format_field(f, 'asr', '\t%s', func=format_decimal_suffix),
3535 join_nonempty(
3536 self._format_out('UNSUPPORTED', 'light red') if f.get('ext') in ('f4f', 'f4m') else None,
3537 format_field(f, 'language', '[%s]'),
3538 join_nonempty(format_field(f, 'format_note'),
3539 format_field(f, 'container', ignore=(None, f.get('ext'))),
3540 delim=', '),
3541 delim=' '),
3542 ] for f in formats if f.get('preference') is None or f['preference'] >= -1000]
3543 header_line = self._list_format_headers(
3544 'ID', 'EXT', 'RESOLUTION', '\tFPS', 'HDR', delim, '\tFILESIZE', '\tTBR', 'PROTO',
3545 delim, 'VCODEC', '\tVBR', 'ACODEC', '\tABR', '\tASR', 'MORE INFO')
3546
3547 return render_table(
3548 header_line, table, hide_empty=True,
3549 delim=self._format_out('\u2500', self.Styles.DELIM, '-', test_encoding=True))
3550
3551 def render_thumbnails_table(self, info_dict):
3552 thumbnails = list(info_dict.get('thumbnails') or [])
3553 if not thumbnails:
3554 return None
3555 return render_table(
3556 self._list_format_headers('ID', 'Width', 'Height', 'URL'),
3557 [[t.get('id'), t.get('width', 'unknown'), t.get('height', 'unknown'), t['url']] for t in thumbnails])
3558
3559 def render_subtitles_table(self, video_id, subtitles):
3560 def _row(lang, formats):
3561 exts, names = zip(*((f['ext'], f.get('name') or 'unknown') for f in reversed(formats)))
3562 if len(set(names)) == 1:
3563 names = [] if names[0] == 'unknown' else names[:1]
3564 return [lang, ', '.join(names), ', '.join(exts)]
3565
3566 if not subtitles:
3567 return None
3568 return render_table(
3569 self._list_format_headers('Language', 'Name', 'Formats'),
3570 [_row(lang, formats) for lang, formats in subtitles.items()],
3571 hide_empty=True)
3572
3573 def __list_table(self, video_id, name, func, *args):
3574 table = func(*args)
3575 if not table:
3576 self.to_screen(f'{video_id} has no {name}')
3577 return
3578 self.to_screen(f'[info] Available {name} for {video_id}:')
3579 self.to_stdout(table)
3580
3581 def list_formats(self, info_dict):
3582 self.__list_table(info_dict['id'], 'formats', self.render_formats_table, info_dict)
3583
3584 def list_thumbnails(self, info_dict):
3585 self.__list_table(info_dict['id'], 'thumbnails', self.render_thumbnails_table, info_dict)
3586
3587 def list_subtitles(self, video_id, subtitles, name='subtitles'):
3588 self.__list_table(video_id, name, self.render_subtitles_table, video_id, subtitles)
3589
3590 def urlopen(self, req):
3591 """ Start an HTTP download """
3592 if isinstance(req, str):
3593 req = sanitized_Request(req)
3594 return self._opener.open(req, timeout=self._socket_timeout)
3595
3596 def print_debug_header(self):
3597 if not self.params.get('verbose'):
3598 return
3599
3600 # These imports can be slow. So import them only as needed
3601 from .extractor.extractors import _LAZY_LOADER
3602 from .extractor.extractors import _PLUGIN_CLASSES as plugin_extractors
3603
3604 def get_encoding(stream):
3605 ret = str(getattr(stream, 'encoding', 'missing (%s)' % type(stream).__name__))
3606 if not supports_terminal_sequences(stream):
3607 from .utils import WINDOWS_VT_MODE # Must be imported locally
3608 ret += ' (No VT)' if WINDOWS_VT_MODE is False else ' (No ANSI)'
3609 return ret
3610
3611 encoding_str = 'Encodings: locale %s, fs %s, pref %s, %s' % (
3612 locale.getpreferredencoding(),
3613 sys.getfilesystemencoding(),
3614 self.get_encoding(),
3615 ', '.join(
3616 f'{key} {get_encoding(stream)}' for key, stream in self._out_files.items_
3617 if stream is not None and key != 'console')
3618 )
3619
3620 logger = self.params.get('logger')
3621 if logger:
3622 write_debug = lambda msg: logger.debug(f'[debug] {msg}')
3623 write_debug(encoding_str)
3624 else:
3625 write_string(f'[debug] {encoding_str}\n', encoding=None)
3626 write_debug = lambda msg: self._write_string(f'[debug] {msg}\n')
3627
3628 source = detect_variant()
3629 write_debug(join_nonempty(
3630 'yt-dlp version', __version__,
3631 f'[{RELEASE_GIT_HEAD}]' if RELEASE_GIT_HEAD else '',
3632 '' if source == 'unknown' else f'({source})',
3633 delim=' '))
3634 if not _LAZY_LOADER:
3635 if os.environ.get('YTDLP_NO_LAZY_EXTRACTORS'):
3636 write_debug('Lazy loading extractors is forcibly disabled')
3637 else:
3638 write_debug('Lazy loading extractors is disabled')
3639 if plugin_extractors or plugin_postprocessors:
3640 write_debug('Plugins: %s' % [
3641 '%s%s' % (klass.__name__, '' if klass.__name__ == name else f' as {name}')
3642 for name, klass in itertools.chain(plugin_extractors.items(), plugin_postprocessors.items())])
3643 if self.params['compat_opts']:
3644 write_debug('Compatibility options: %s' % ', '.join(self.params['compat_opts']))
3645
3646 if source == 'source':
3647 try:
3648 stdout, _, _ = Popen.run(
3649 ['git', 'rev-parse', '--short', 'HEAD'],
3650 text=True, cwd=os.path.dirname(os.path.abspath(__file__)),
3651 stdout=subprocess.PIPE, stderr=subprocess.PIPE)
3652 if re.fullmatch('[0-9a-f]+', stdout.strip()):
3653 write_debug(f'Git HEAD: {stdout.strip()}')
3654 except Exception:
3655 with contextlib.suppress(Exception):
3656 sys.exc_clear()
3657
3658 write_debug(system_identifier())
3659
3660 exe_versions, ffmpeg_features = FFmpegPostProcessor.get_versions_and_features(self)
3661 ffmpeg_features = {key for key, val in ffmpeg_features.items() if val}
3662 if ffmpeg_features:
3663 exe_versions['ffmpeg'] += ' (%s)' % ','.join(sorted(ffmpeg_features))
3664
3665 exe_versions['rtmpdump'] = rtmpdump_version()
3666 exe_versions['phantomjs'] = PhantomJSwrapper._version()
3667 exe_str = ', '.join(
3668 f'{exe} {v}' for exe, v in sorted(exe_versions.items()) if v
3669 ) or 'none'
3670 write_debug('exe versions: %s' % exe_str)
3671
3672 from .compat.compat_utils import get_package_info
3673 from .dependencies import available_dependencies
3674
3675 write_debug('Optional libraries: %s' % (', '.join(sorted({
3676 join_nonempty(*get_package_info(m)) for m in available_dependencies.values()
3677 })) or 'none'))
3678
3679 self._setup_opener()
3680 proxy_map = {}
3681 for handler in self._opener.handlers:
3682 if hasattr(handler, 'proxies'):
3683 proxy_map.update(handler.proxies)
3684 write_debug(f'Proxy map: {proxy_map}')
3685
3686 # Not implemented
3687 if False and self.params.get('call_home'):
3688 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode()
3689 write_debug('Public IP address: %s' % ipaddr)
3690 latest_version = self.urlopen(
3691 'https://yt-dl.org/latest/version').read().decode()
3692 if version_tuple(latest_version) > version_tuple(__version__):
3693 self.report_warning(
3694 'You are using an outdated version (newest version: %s)! '
3695 'See https://yt-dl.org/update if you need help updating.' %
3696 latest_version)
3697
3698 def _setup_opener(self):
3699 if hasattr(self, '_opener'):
3700 return
3701 timeout_val = self.params.get('socket_timeout')
3702 self._socket_timeout = 20 if timeout_val is None else float(timeout_val)
3703
3704 opts_cookiesfrombrowser = self.params.get('cookiesfrombrowser')
3705 opts_cookiefile = self.params.get('cookiefile')
3706 opts_proxy = self.params.get('proxy')
3707
3708 self.cookiejar = load_cookies(opts_cookiefile, opts_cookiesfrombrowser, self)
3709
3710 cookie_processor = YoutubeDLCookieProcessor(self.cookiejar)
3711 if opts_proxy is not None:
3712 if opts_proxy == '':
3713 proxies = {}
3714 else:
3715 proxies = {'http': opts_proxy, 'https': opts_proxy}
3716 else:
3717 proxies = urllib.request.getproxies()
3718 # Set HTTPS proxy to HTTP one if given (https://github.com/ytdl-org/youtube-dl/issues/805)
3719 if 'http' in proxies and 'https' not in proxies:
3720 proxies['https'] = proxies['http']
3721 proxy_handler = PerRequestProxyHandler(proxies)
3722
3723 debuglevel = 1 if self.params.get('debug_printtraffic') else 0
3724 https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
3725 ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
3726 redirect_handler = YoutubeDLRedirectHandler()
3727 data_handler = urllib.request.DataHandler()
3728
3729 # When passing our own FileHandler instance, build_opener won't add the
3730 # default FileHandler and allows us to disable the file protocol, which
3731 # can be used for malicious purposes (see
3732 # https://github.com/ytdl-org/youtube-dl/issues/8227)
3733 file_handler = urllib.request.FileHandler()
3734
3735 def file_open(*args, **kwargs):
3736 raise urllib.error.URLError('file:// scheme is explicitly disabled in yt-dlp for security reasons')
3737 file_handler.file_open = file_open
3738
3739 opener = urllib.request.build_opener(
3740 proxy_handler, https_handler, cookie_processor, ydlh, redirect_handler, data_handler, file_handler)
3741
3742 # Delete the default user-agent header, which would otherwise apply in
3743 # cases where our custom HTTP handler doesn't come into play
3744 # (See https://github.com/ytdl-org/youtube-dl/issues/1309 for details)
3745 opener.addheaders = []
3746 self._opener = opener
3747
3748 def encode(self, s):
3749 if isinstance(s, bytes):
3750 return s # Already encoded
3751
3752 try:
3753 return s.encode(self.get_encoding())
3754 except UnicodeEncodeError as err:
3755 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
3756 raise
3757
3758 def get_encoding(self):
3759 encoding = self.params.get('encoding')
3760 if encoding is None:
3761 encoding = preferredencoding()
3762 return encoding
3763
3764 def _write_info_json(self, label, ie_result, infofn, overwrite=None):
3765 ''' Write infojson and returns True = written, 'exists' = Already exists, False = skip, None = error '''
3766 if overwrite is None:
3767 overwrite = self.params.get('overwrites', True)
3768 if not self.params.get('writeinfojson'):
3769 return False
3770 elif not infofn:
3771 self.write_debug(f'Skipping writing {label} infojson')
3772 return False
3773 elif not self._ensure_dir_exists(infofn):
3774 return None
3775 elif not overwrite and os.path.exists(infofn):
3776 self.to_screen(f'[info] {label.title()} metadata is already present')
3777 return 'exists'
3778
3779 self.to_screen(f'[info] Writing {label} metadata as JSON to: {infofn}')
3780 try:
3781 write_json_file(self.sanitize_info(ie_result, self.params.get('clean_infojson', True)), infofn)
3782 return True
3783 except OSError:
3784 self.report_error(f'Cannot write {label} metadata to JSON file {infofn}')
3785 return None
3786
3787 def _write_description(self, label, ie_result, descfn):
3788 ''' Write description and returns True = written, False = skip, None = error '''
3789 if not self.params.get('writedescription'):
3790 return False
3791 elif not descfn:
3792 self.write_debug(f'Skipping writing {label} description')
3793 return False
3794 elif not self._ensure_dir_exists(descfn):
3795 return None
3796 elif not self.params.get('overwrites', True) and os.path.exists(descfn):
3797 self.to_screen(f'[info] {label.title()} description is already present')
3798 elif ie_result.get('description') is None:
3799 self.report_warning(f'There\'s no {label} description to write')
3800 return False
3801 else:
3802 try:
3803 self.to_screen(f'[info] Writing {label} description to: {descfn}')
3804 with open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
3805 descfile.write(ie_result['description'])
3806 except OSError:
3807 self.report_error(f'Cannot write {label} description file {descfn}')
3808 return None
3809 return True
3810
3811 def _write_subtitles(self, info_dict, filename):
3812 ''' Write subtitles to file and return list of (sub_filename, final_sub_filename); or None if error'''
3813 ret = []
3814 subtitles = info_dict.get('requested_subtitles')
3815 if not subtitles or not (self.params.get('writesubtitles') or self.params.get('writeautomaticsub')):
3816 # subtitles download errors are already managed as troubles in relevant IE
3817 # that way it will silently go on when used with unsupporting IE
3818 return ret
3819
3820 sub_filename_base = self.prepare_filename(info_dict, 'subtitle')
3821 if not sub_filename_base:
3822 self.to_screen('[info] Skipping writing video subtitles')
3823 return ret
3824 for sub_lang, sub_info in subtitles.items():
3825 sub_format = sub_info['ext']
3826 sub_filename = subtitles_filename(filename, sub_lang, sub_format, info_dict.get('ext'))
3827 sub_filename_final = subtitles_filename(sub_filename_base, sub_lang, sub_format, info_dict.get('ext'))
3828 existing_sub = self.existing_file((sub_filename_final, sub_filename))
3829 if existing_sub:
3830 self.to_screen(f'[info] Video subtitle {sub_lang}.{sub_format} is already present')
3831 sub_info['filepath'] = existing_sub
3832 ret.append((existing_sub, sub_filename_final))
3833 continue
3834
3835 self.to_screen(f'[info] Writing video subtitles to: {sub_filename}')
3836 if sub_info.get('data') is not None:
3837 try:
3838 # Use newline='' to prevent conversion of newline characters
3839 # See https://github.com/ytdl-org/youtube-dl/issues/10268
3840 with open(sub_filename, 'w', encoding='utf-8', newline='') as subfile:
3841 subfile.write(sub_info['data'])
3842 sub_info['filepath'] = sub_filename
3843 ret.append((sub_filename, sub_filename_final))
3844 continue
3845 except OSError:
3846 self.report_error(f'Cannot write video subtitles file {sub_filename}')
3847 return None
3848
3849 try:
3850 sub_copy = sub_info.copy()
3851 sub_copy.setdefault('http_headers', info_dict.get('http_headers'))
3852 self.dl(sub_filename, sub_copy, subtitle=True)
3853 sub_info['filepath'] = sub_filename
3854 ret.append((sub_filename, sub_filename_final))
3855 except (DownloadError, ExtractorError, IOError, OSError, ValueError) + network_exceptions as err:
3856 msg = f'Unable to download video subtitles for {sub_lang!r}: {err}'
3857 if self.params.get('ignoreerrors') is not True: # False or 'only_download'
3858 if not self.params.get('ignoreerrors'):
3859 self.report_error(msg)
3860 raise DownloadError(msg)
3861 self.report_warning(msg)
3862 return ret
3863
3864 def _write_thumbnails(self, label, info_dict, filename, thumb_filename_base=None):
3865 ''' Write thumbnails to file and return list of (thumb_filename, final_thumb_filename) '''
3866 write_all = self.params.get('write_all_thumbnails', False)
3867 thumbnails, ret = [], []
3868 if write_all or self.params.get('writethumbnail', False):
3869 thumbnails = info_dict.get('thumbnails') or []
3870 multiple = write_all and len(thumbnails) > 1
3871
3872 if thumb_filename_base is None:
3873 thumb_filename_base = filename
3874 if thumbnails and not thumb_filename_base:
3875 self.write_debug(f'Skipping writing {label} thumbnail')
3876 return ret
3877
3878 for idx, t in list(enumerate(thumbnails))[::-1]:
3879 thumb_ext = (f'{t["id"]}.' if multiple else '') + determine_ext(t['url'], 'jpg')
3880 thumb_display_id = f'{label} thumbnail {t["id"]}'
3881 thumb_filename = replace_extension(filename, thumb_ext, info_dict.get('ext'))
3882 thumb_filename_final = replace_extension(thumb_filename_base, thumb_ext, info_dict.get('ext'))
3883
3884 existing_thumb = self.existing_file((thumb_filename_final, thumb_filename))
3885 if existing_thumb:
3886 self.to_screen('[info] %s is already present' % (
3887 thumb_display_id if multiple else f'{label} thumbnail').capitalize())
3888 t['filepath'] = existing_thumb
3889 ret.append((existing_thumb, thumb_filename_final))
3890 else:
3891 self.to_screen(f'[info] Downloading {thumb_display_id} ...')
3892 try:
3893 uf = self.urlopen(sanitized_Request(t['url'], headers=t.get('http_headers', {})))
3894 self.to_screen(f'[info] Writing {thumb_display_id} to: {thumb_filename}')
3895 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
3896 shutil.copyfileobj(uf, thumbf)
3897 ret.append((thumb_filename, thumb_filename_final))
3898 t['filepath'] = thumb_filename
3899 except network_exceptions as err:
3900 thumbnails.pop(idx)
3901 self.report_warning(f'Unable to download {thumb_display_id}: {err}')
3902 if ret and not write_all:
3903 break
3904 return ret