]> jfr.im git - yt-dlp.git/blob - yt_dlp/YoutubeDL.py
[utils] Fix bug in 0b9c08b47bb5e95c21b067044ace4e824d19a9c2
[yt-dlp.git] / yt_dlp / YoutubeDL.py
1 #!/usr/bin/env python3
2 import collections
3 import contextlib
4 import datetime
5 import errno
6 import fileinput
7 import functools
8 import io
9 import itertools
10 import json
11 import locale
12 import operator
13 import os
14 import platform
15 import random
16 import re
17 import shutil
18 import subprocess
19 import sys
20 import tempfile
21 import time
22 import tokenize
23 import traceback
24 import unicodedata
25 import urllib.request
26 from string import ascii_letters
27
28 from .cache import Cache
29 from .compat import (
30 compat_get_terminal_size,
31 compat_os_name,
32 compat_shlex_quote,
33 compat_str,
34 compat_urllib_error,
35 compat_urllib_request,
36 )
37 from .cookies import load_cookies
38 from .downloader import FFmpegFD, get_suitable_downloader, shorten_protocol_name
39 from .downloader.rtmp import rtmpdump_version
40 from .extractor import _LAZY_LOADER
41 from .extractor import _PLUGIN_CLASSES as plugin_extractors
42 from .extractor import gen_extractor_classes, get_info_extractor
43 from .extractor.openload import PhantomJSwrapper
44 from .minicurses import format_text
45 from .postprocessor import _PLUGIN_CLASSES as plugin_postprocessors
46 from .postprocessor import (
47 EmbedThumbnailPP,
48 FFmpegFixupDuplicateMoovPP,
49 FFmpegFixupDurationPP,
50 FFmpegFixupM3u8PP,
51 FFmpegFixupM4aPP,
52 FFmpegFixupStretchedPP,
53 FFmpegFixupTimestampPP,
54 FFmpegMergerPP,
55 FFmpegPostProcessor,
56 MoveFilesAfterDownloadPP,
57 get_postprocessor,
58 )
59 from .update import detect_variant
60 from .utils import (
61 DEFAULT_OUTTMPL,
62 LINK_TEMPLATES,
63 NO_DEFAULT,
64 NUMBER_RE,
65 OUTTMPL_TYPES,
66 POSTPROCESS_WHEN,
67 STR_FORMAT_RE_TMPL,
68 STR_FORMAT_TYPES,
69 ContentTooShortError,
70 DateRange,
71 DownloadCancelled,
72 DownloadError,
73 EntryNotInPlaylist,
74 ExistingVideoReached,
75 ExtractorError,
76 GeoRestrictedError,
77 HEADRequest,
78 InAdvancePagedList,
79 ISO3166Utils,
80 LazyList,
81 MaxDownloadsReached,
82 Namespace,
83 PagedList,
84 PerRequestProxyHandler,
85 Popen,
86 PostProcessingError,
87 ReExtractInfo,
88 RejectedVideoReached,
89 SameFileError,
90 UnavailableVideoError,
91 YoutubeDLCookieProcessor,
92 YoutubeDLHandler,
93 YoutubeDLRedirectHandler,
94 age_restricted,
95 args_to_str,
96 date_from_str,
97 determine_ext,
98 determine_protocol,
99 encode_compat_str,
100 encodeFilename,
101 error_to_compat_str,
102 expand_path,
103 filter_dict,
104 float_or_none,
105 format_bytes,
106 format_decimal_suffix,
107 format_field,
108 formatSeconds,
109 get_domain,
110 int_or_none,
111 iri_to_uri,
112 join_nonempty,
113 locked_file,
114 make_dir,
115 make_HTTPS_handler,
116 merge_headers,
117 network_exceptions,
118 number_of_digits,
119 orderedSet,
120 parse_filesize,
121 platform_name,
122 preferredencoding,
123 prepend_extension,
124 register_socks_protocols,
125 remove_terminal_sequences,
126 render_table,
127 replace_extension,
128 sanitize_filename,
129 sanitize_path,
130 sanitize_url,
131 sanitized_Request,
132 std_headers,
133 str_or_none,
134 strftime_or_none,
135 subtitles_filename,
136 supports_terminal_sequences,
137 timetuple_from_msec,
138 to_high_limit_path,
139 traverse_obj,
140 try_get,
141 url_basename,
142 variadic,
143 version_tuple,
144 windows_enable_vt_mode,
145 write_json_file,
146 write_string,
147 )
148 from .version import RELEASE_GIT_HEAD, __version__
149
150 if compat_os_name == 'nt':
151 import ctypes
152
153
154 class YoutubeDL:
155 """YoutubeDL class.
156
157 YoutubeDL objects are the ones responsible of downloading the
158 actual video file and writing it to disk if the user has requested
159 it, among some other tasks. In most cases there should be one per
160 program. As, given a video URL, the downloader doesn't know how to
161 extract all the needed information, task that InfoExtractors do, it
162 has to pass the URL to one of them.
163
164 For this, YoutubeDL objects have a method that allows
165 InfoExtractors to be registered in a given order. When it is passed
166 a URL, the YoutubeDL object handles it to the first InfoExtractor it
167 finds that reports being able to handle it. The InfoExtractor extracts
168 all the information about the video or videos the URL refers to, and
169 YoutubeDL process the extracted information, possibly using a File
170 Downloader to download the video.
171
172 YoutubeDL objects accept a lot of parameters. In order not to saturate
173 the object constructor with arguments, it receives a dictionary of
174 options instead. These options are available through the params
175 attribute for the InfoExtractors to use. The YoutubeDL also
176 registers itself as the downloader in charge for the InfoExtractors
177 that are added to it, so this is a "mutual registration".
178
179 Available options:
180
181 username: Username for authentication purposes.
182 password: Password for authentication purposes.
183 videopassword: Password for accessing a video.
184 ap_mso: Adobe Pass multiple-system operator identifier.
185 ap_username: Multiple-system operator account username.
186 ap_password: Multiple-system operator account password.
187 usenetrc: Use netrc for authentication instead.
188 verbose: Print additional info to stdout.
189 quiet: Do not print messages to stdout.
190 no_warnings: Do not print out anything for warnings.
191 forceprint: A dict with keys WHEN mapped to a list of templates to
192 print to stdout. The allowed keys are video or any of the
193 items in utils.POSTPROCESS_WHEN.
194 For compatibility, a single list is also accepted
195 print_to_file: A dict with keys WHEN (same as forceprint) mapped to
196 a list of tuples with (template, filename)
197 forceurl: Force printing final URL. (Deprecated)
198 forcetitle: Force printing title. (Deprecated)
199 forceid: Force printing ID. (Deprecated)
200 forcethumbnail: Force printing thumbnail URL. (Deprecated)
201 forcedescription: Force printing description. (Deprecated)
202 forcefilename: Force printing final filename. (Deprecated)
203 forceduration: Force printing duration. (Deprecated)
204 forcejson: Force printing info_dict as JSON.
205 dump_single_json: Force printing the info_dict of the whole playlist
206 (or video) as a single JSON line.
207 force_write_download_archive: Force writing download archive regardless
208 of 'skip_download' or 'simulate'.
209 simulate: Do not download the video files. If unset (or None),
210 simulate only if listsubtitles, listformats or list_thumbnails is used
211 format: Video format code. see "FORMAT SELECTION" for more details.
212 You can also pass a function. The function takes 'ctx' as
213 argument and returns the formats to download.
214 See "build_format_selector" for an implementation
215 allow_unplayable_formats: Allow unplayable formats to be extracted and downloaded.
216 ignore_no_formats_error: Ignore "No video formats" error. Usefull for
217 extracting metadata even if the video is not actually
218 available for download (experimental)
219 format_sort: A list of fields by which to sort the video formats.
220 See "Sorting Formats" for more details.
221 format_sort_force: Force the given format_sort. see "Sorting Formats"
222 for more details.
223 prefer_free_formats: Whether to prefer video formats with free containers
224 over non-free ones of same quality.
225 allow_multiple_video_streams: Allow multiple video streams to be merged
226 into a single file
227 allow_multiple_audio_streams: Allow multiple audio streams to be merged
228 into a single file
229 check_formats Whether to test if the formats are downloadable.
230 Can be True (check all), False (check none),
231 'selected' (check selected formats),
232 or None (check only if requested by extractor)
233 paths: Dictionary of output paths. The allowed keys are 'home'
234 'temp' and the keys of OUTTMPL_TYPES (in utils.py)
235 outtmpl: Dictionary of templates for output names. Allowed keys
236 are 'default' and the keys of OUTTMPL_TYPES (in utils.py).
237 For compatibility with youtube-dl, a single string can also be used
238 outtmpl_na_placeholder: Placeholder for unavailable meta fields.
239 restrictfilenames: Do not allow "&" and spaces in file names
240 trim_file_name: Limit length of filename (extension excluded)
241 windowsfilenames: Force the filenames to be windows compatible
242 ignoreerrors: Do not stop on download/postprocessing errors.
243 Can be 'only_download' to ignore only download errors.
244 Default is 'only_download' for CLI, but False for API
245 skip_playlist_after_errors: Number of allowed failures until the rest of
246 the playlist is skipped
247 force_generic_extractor: Force downloader to use the generic extractor
248 overwrites: Overwrite all video and metadata files if True,
249 overwrite only non-video files if None
250 and don't overwrite any file if False
251 For compatibility with youtube-dl,
252 "nooverwrites" may also be used instead
253 playliststart: Playlist item to start at.
254 playlistend: Playlist item to end at.
255 playlist_items: Specific indices of playlist to download.
256 playlistreverse: Download playlist items in reverse order.
257 playlistrandom: Download playlist items in random order.
258 matchtitle: Download only matching titles.
259 rejecttitle: Reject downloads for matching titles.
260 logger: Log messages to a logging.Logger instance.
261 logtostderr: Log messages to stderr instead of stdout.
262 consoletitle: Display progress in console window's titlebar.
263 writedescription: Write the video description to a .description file
264 writeinfojson: Write the video description to a .info.json file
265 clean_infojson: Remove private fields from the infojson
266 getcomments: Extract video comments. This will not be written to disk
267 unless writeinfojson is also given
268 writeannotations: Write the video annotations to a .annotations.xml file
269 writethumbnail: Write the thumbnail image to a file
270 allow_playlist_files: Whether to write playlists' description, infojson etc
271 also to disk when using the 'write*' options
272 write_all_thumbnails: Write all thumbnail formats to files
273 writelink: Write an internet shortcut file, depending on the
274 current platform (.url/.webloc/.desktop)
275 writeurllink: Write a Windows internet shortcut file (.url)
276 writewebloclink: Write a macOS internet shortcut file (.webloc)
277 writedesktoplink: Write a Linux internet shortcut file (.desktop)
278 writesubtitles: Write the video subtitles to a file
279 writeautomaticsub: Write the automatically generated subtitles to a file
280 allsubtitles: Deprecated - Use subtitleslangs = ['all']
281 Downloads all the subtitles of the video
282 (requires writesubtitles or writeautomaticsub)
283 listsubtitles: Lists all available subtitles for the video
284 subtitlesformat: The format code for subtitles
285 subtitleslangs: List of languages of the subtitles to download (can be regex).
286 The list may contain "all" to refer to all the available
287 subtitles. The language can be prefixed with a "-" to
288 exclude it from the requested languages. Eg: ['all', '-live_chat']
289 keepvideo: Keep the video file after post-processing
290 daterange: A DateRange object, download only if the upload_date is in the range.
291 skip_download: Skip the actual download of the video file
292 cachedir: Location of the cache files in the filesystem.
293 False to disable filesystem cache.
294 noplaylist: Download single video instead of a playlist if in doubt.
295 age_limit: An integer representing the user's age in years.
296 Unsuitable videos for the given age are skipped.
297 min_views: An integer representing the minimum view count the video
298 must have in order to not be skipped.
299 Videos without view count information are always
300 downloaded. None for no limit.
301 max_views: An integer representing the maximum view count.
302 Videos that are more popular than that are not
303 downloaded.
304 Videos without view count information are always
305 downloaded. None for no limit.
306 download_archive: File name of a file where all downloads are recorded.
307 Videos already present in the file are not downloaded
308 again.
309 break_on_existing: Stop the download process after attempting to download a
310 file that is in the archive.
311 break_on_reject: Stop the download process when encountering a video that
312 has been filtered out.
313 break_per_url: Whether break_on_reject and break_on_existing
314 should act on each input URL as opposed to for the entire queue
315 cookiefile: File name or text stream from where cookies should be read and dumped to
316 cookiesfrombrowser: A tuple containing the name of the browser, the profile
317 name/pathfrom where cookies are loaded, and the name of the
318 keyring. Eg: ('chrome', ) or ('vivaldi', 'default', 'BASICTEXT')
319 legacyserverconnect: Explicitly allow HTTPS connection to servers that do not
320 support RFC 5746 secure renegotiation
321 nocheckcertificate: Do not verify SSL certificates
322 client_certificate: Path to client certificate file in PEM format. May include the private key
323 client_certificate_key: Path to private key file for client certificate
324 client_certificate_password: Password for client certificate private key, if encrypted.
325 If not provided and the key is encrypted, yt-dlp will ask interactively
326 prefer_insecure: Use HTTP instead of HTTPS to retrieve information.
327 At the moment, this is only supported by YouTube.
328 http_headers: A dictionary of custom headers to be used for all requests
329 proxy: URL of the proxy server to use
330 geo_verification_proxy: URL of the proxy to use for IP address verification
331 on geo-restricted sites.
332 socket_timeout: Time to wait for unresponsive hosts, in seconds
333 bidi_workaround: Work around buggy terminals without bidirectional text
334 support, using fridibi
335 debug_printtraffic:Print out sent and received HTTP traffic
336 include_ads: Download ads as well (deprecated)
337 default_search: Prepend this string if an input url is not valid.
338 'auto' for elaborate guessing
339 encoding: Use this encoding instead of the system-specified.
340 extract_flat: Do not resolve URLs, return the immediate result.
341 Pass in 'in_playlist' to only show this behavior for
342 playlist items.
343 wait_for_video: If given, wait for scheduled streams to become available.
344 The value should be a tuple containing the range
345 (min_secs, max_secs) to wait between retries
346 postprocessors: A list of dictionaries, each with an entry
347 * key: The name of the postprocessor. See
348 yt_dlp/postprocessor/__init__.py for a list.
349 * when: When to run the postprocessor. Allowed values are
350 the entries of utils.POSTPROCESS_WHEN
351 Assumed to be 'post_process' if not given
352 post_hooks: Deprecated - Register a custom postprocessor instead
353 A list of functions that get called as the final step
354 for each video file, after all postprocessors have been
355 called. The filename will be passed as the only argument.
356 progress_hooks: A list of functions that get called on download
357 progress, with a dictionary with the entries
358 * status: One of "downloading", "error", or "finished".
359 Check this first and ignore unknown values.
360 * info_dict: The extracted info_dict
361
362 If status is one of "downloading", or "finished", the
363 following properties may also be present:
364 * filename: The final filename (always present)
365 * tmpfilename: The filename we're currently writing to
366 * downloaded_bytes: Bytes on disk
367 * total_bytes: Size of the whole file, None if unknown
368 * total_bytes_estimate: Guess of the eventual file size,
369 None if unavailable.
370 * elapsed: The number of seconds since download started.
371 * eta: The estimated time in seconds, None if unknown
372 * speed: The download speed in bytes/second, None if
373 unknown
374 * fragment_index: The counter of the currently
375 downloaded video fragment.
376 * fragment_count: The number of fragments (= individual
377 files that will be merged)
378
379 Progress hooks are guaranteed to be called at least once
380 (with status "finished") if the download is successful.
381 postprocessor_hooks: A list of functions that get called on postprocessing
382 progress, with a dictionary with the entries
383 * status: One of "started", "processing", or "finished".
384 Check this first and ignore unknown values.
385 * postprocessor: Name of the postprocessor
386 * info_dict: The extracted info_dict
387
388 Progress hooks are guaranteed to be called at least twice
389 (with status "started" and "finished") if the processing is successful.
390 merge_output_format: Extension to use when merging formats.
391 final_ext: Expected final extension; used to detect when the file was
392 already downloaded and converted
393 fixup: Automatically correct known faults of the file.
394 One of:
395 - "never": do nothing
396 - "warn": only emit a warning
397 - "detect_or_warn": check whether we can do anything
398 about it, warn otherwise (default)
399 source_address: Client-side IP address to bind to.
400 call_home: Boolean, true iff we are allowed to contact the
401 yt-dlp servers for debugging. (BROKEN)
402 sleep_interval_requests: Number of seconds to sleep between requests
403 during extraction
404 sleep_interval: Number of seconds to sleep before each download when
405 used alone or a lower bound of a range for randomized
406 sleep before each download (minimum possible number
407 of seconds to sleep) when used along with
408 max_sleep_interval.
409 max_sleep_interval:Upper bound of a range for randomized sleep before each
410 download (maximum possible number of seconds to sleep).
411 Must only be used along with sleep_interval.
412 Actual sleep time will be a random float from range
413 [sleep_interval; max_sleep_interval].
414 sleep_interval_subtitles: Number of seconds to sleep before each subtitle download
415 listformats: Print an overview of available video formats and exit.
416 list_thumbnails: Print a table of all thumbnails and exit.
417 match_filter: A function that gets called for every video with the signature
418 (info_dict, *, incomplete: bool) -> Optional[str]
419 For backward compatibility with youtube-dl, the signature
420 (info_dict) -> Optional[str] is also allowed.
421 - If it returns a message, the video is ignored.
422 - If it returns None, the video is downloaded.
423 - If it returns utils.NO_DEFAULT, the user is interactively
424 asked whether to download the video.
425 match_filter_func in utils.py is one example for this.
426 no_color: Do not emit color codes in output.
427 geo_bypass: Bypass geographic restriction via faking X-Forwarded-For
428 HTTP header
429 geo_bypass_country:
430 Two-letter ISO 3166-2 country code that will be used for
431 explicit geographic restriction bypassing via faking
432 X-Forwarded-For HTTP header
433 geo_bypass_ip_block:
434 IP range in CIDR notation that will be used similarly to
435 geo_bypass_country
436
437 The following options determine which downloader is picked:
438 external_downloader: A dictionary of protocol keys and the executable of the
439 external downloader to use for it. The allowed protocols
440 are default|http|ftp|m3u8|dash|rtsp|rtmp|mms.
441 Set the value to 'native' to use the native downloader
442 hls_prefer_native: Deprecated - Use external_downloader = {'m3u8': 'native'}
443 or {'m3u8': 'ffmpeg'} instead.
444 Use the native HLS downloader instead of ffmpeg/avconv
445 if True, otherwise use ffmpeg/avconv if False, otherwise
446 use downloader suggested by extractor if None.
447 compat_opts: Compatibility options. See "Differences in default behavior".
448 The following options do not work when used through the API:
449 filename, abort-on-error, multistreams, no-live-chat, format-sort
450 no-clean-infojson, no-playlist-metafiles, no-keep-subs, no-attach-info-json.
451 Refer __init__.py for their implementation
452 progress_template: Dictionary of templates for progress outputs.
453 Allowed keys are 'download', 'postprocess',
454 'download-title' (console title) and 'postprocess-title'.
455 The template is mapped on a dictionary with keys 'progress' and 'info'
456 retry_sleep_functions: Dictionary of functions that takes the number of attempts
457 as argument and returns the time to sleep in seconds.
458 Allowed keys are 'http', 'fragment', 'file_access'
459
460 The following parameters are not used by YoutubeDL itself, they are used by
461 the downloader (see yt_dlp/downloader/common.py):
462 nopart, updatetime, buffersize, ratelimit, throttledratelimit, min_filesize,
463 max_filesize, test, noresizebuffer, retries, file_access_retries, fragment_retries,
464 continuedl, noprogress, xattr_set_filesize, hls_use_mpegts, http_chunk_size,
465 external_downloader_args, concurrent_fragment_downloads.
466
467 The following options are used by the post processors:
468 prefer_ffmpeg: If False, use avconv instead of ffmpeg if both are available,
469 otherwise prefer ffmpeg. (avconv support is deprecated)
470 ffmpeg_location: Location of the ffmpeg/avconv binary; either the path
471 to the binary or its containing directory.
472 postprocessor_args: A dictionary of postprocessor/executable keys (in lower case)
473 and a list of additional command-line arguments for the
474 postprocessor/executable. The dict can also have "PP+EXE" keys
475 which are used when the given exe is used by the given PP.
476 Use 'default' as the name for arguments to passed to all PP
477 For compatibility with youtube-dl, a single list of args
478 can also be used
479
480 The following options are used by the extractors:
481 extractor_retries: Number of times to retry for known errors
482 dynamic_mpd: Whether to process dynamic DASH manifests (default: True)
483 hls_split_discontinuity: Split HLS playlists to different formats at
484 discontinuities such as ad breaks (default: False)
485 extractor_args: A dictionary of arguments to be passed to the extractors.
486 See "EXTRACTOR ARGUMENTS" for details.
487 Eg: {'youtube': {'skip': ['dash', 'hls']}}
488 mark_watched: Mark videos watched (even with --simulate). Only for YouTube
489 youtube_include_dash_manifest: Deprecated - Use extractor_args instead.
490 If True (default), DASH manifests and related
491 data will be downloaded and processed by extractor.
492 You can reduce network I/O by disabling it if you don't
493 care about DASH. (only for youtube)
494 youtube_include_hls_manifest: Deprecated - Use extractor_args instead.
495 If True (default), HLS manifests and related
496 data will be downloaded and processed by extractor.
497 You can reduce network I/O by disabling it if you don't
498 care about HLS. (only for youtube)
499 """
500
501 _NUMERIC_FIELDS = {
502 'width', 'height', 'tbr', 'abr', 'asr', 'vbr', 'fps', 'filesize', 'filesize_approx',
503 'timestamp', 'release_timestamp',
504 'duration', 'view_count', 'like_count', 'dislike_count', 'repost_count',
505 'average_rating', 'comment_count', 'age_limit',
506 'start_time', 'end_time',
507 'chapter_number', 'season_number', 'episode_number',
508 'track_number', 'disc_number', 'release_year',
509 }
510
511 _format_fields = {
512 # NB: Keep in sync with the docstring of extractor/common.py
513 'url', 'manifest_url', 'manifest_stream_number', 'ext', 'format', 'format_id', 'format_note',
514 'width', 'height', 'resolution', 'dynamic_range', 'tbr', 'abr', 'acodec', 'asr',
515 'vbr', 'fps', 'vcodec', 'container', 'filesize', 'filesize_approx',
516 'player_url', 'protocol', 'fragment_base_url', 'fragments', 'is_from_start',
517 'preference', 'language', 'language_preference', 'quality', 'source_preference',
518 'http_headers', 'stretched_ratio', 'no_resume', 'has_drm', 'downloader_options',
519 'page_url', 'app', 'play_path', 'tc_url', 'flash_version', 'rtmp_live', 'rtmp_conn', 'rtmp_protocol', 'rtmp_real_time'
520 }
521 _format_selection_exts = {
522 'audio': {'m4a', 'mp3', 'ogg', 'aac'},
523 'video': {'mp4', 'flv', 'webm', '3gp'},
524 'storyboards': {'mhtml'},
525 }
526
527 def __init__(self, params=None, auto_init=True):
528 """Create a FileDownloader object with the given options.
529 @param auto_init Whether to load the default extractors and print header (if verbose).
530 Set to 'no_verbose_header' to not print the header
531 """
532 if params is None:
533 params = {}
534 self.params = params
535 self._ies = {}
536 self._ies_instances = {}
537 self._pps = {k: [] for k in POSTPROCESS_WHEN}
538 self._printed_messages = set()
539 self._first_webpage_request = True
540 self._post_hooks = []
541 self._progress_hooks = []
542 self._postprocessor_hooks = []
543 self._download_retcode = 0
544 self._num_downloads = 0
545 self._num_videos = 0
546 self._playlist_level = 0
547 self._playlist_urls = set()
548 self.cache = Cache(self)
549
550 windows_enable_vt_mode()
551 stdout = sys.stderr if self.params.get('logtostderr') else sys.stdout
552 self._out_files = Namespace(
553 out=stdout,
554 error=sys.stderr,
555 screen=sys.stderr if self.params.get('quiet') else stdout,
556 console=None if compat_os_name == 'nt' else next(
557 filter(supports_terminal_sequences, (sys.stderr, sys.stdout)), None)
558 )
559 self._allow_colors = Namespace(**{
560 type_: not self.params.get('no_color') and supports_terminal_sequences(stream)
561 for type_, stream in self._out_files if type_ != 'console'
562 })
563
564 if sys.version_info < (3, 6):
565 self.report_warning(
566 'Python version %d.%d is not supported! Please update to Python 3.6 or above' % sys.version_info[:2])
567
568 if self.params.get('allow_unplayable_formats'):
569 self.report_warning(
570 f'You have asked for {self._format_err("UNPLAYABLE", self.Styles.EMPHASIS)} formats to be listed/downloaded. '
571 'This is a developer option intended for debugging. \n'
572 ' If you experience any issues while using this option, '
573 f'{self._format_err("DO NOT", self.Styles.ERROR)} open a bug report')
574
575 def check_deprecated(param, option, suggestion):
576 if self.params.get(param) is not None:
577 self.report_warning(f'{option} is deprecated. Use {suggestion} instead')
578 return True
579 return False
580
581 if check_deprecated('cn_verification_proxy', '--cn-verification-proxy', '--geo-verification-proxy'):
582 if self.params.get('geo_verification_proxy') is None:
583 self.params['geo_verification_proxy'] = self.params['cn_verification_proxy']
584
585 check_deprecated('autonumber', '--auto-number', '-o "%(autonumber)s-%(title)s.%(ext)s"')
586 check_deprecated('usetitle', '--title', '-o "%(title)s-%(id)s.%(ext)s"')
587 check_deprecated('useid', '--id', '-o "%(id)s.%(ext)s"')
588
589 for msg in self.params.get('_warnings', []):
590 self.report_warning(msg)
591 for msg in self.params.get('_deprecation_warnings', []):
592 self.deprecation_warning(msg)
593
594 if 'list-formats' in self.params.get('compat_opts', []):
595 self.params['listformats_table'] = False
596
597 if 'overwrites' not in self.params and self.params.get('nooverwrites') is not None:
598 # nooverwrites was unnecessarily changed to overwrites
599 # in 0c3d0f51778b153f65c21906031c2e091fcfb641
600 # This ensures compatibility with both keys
601 self.params['overwrites'] = not self.params['nooverwrites']
602 elif self.params.get('overwrites') is None:
603 self.params.pop('overwrites', None)
604 else:
605 self.params['nooverwrites'] = not self.params['overwrites']
606
607 self.params.setdefault('forceprint', {})
608 self.params.setdefault('print_to_file', {})
609
610 # Compatibility with older syntax
611 if not isinstance(params['forceprint'], dict):
612 self.params['forceprint'] = {'video': params['forceprint']}
613
614 if self.params.get('bidi_workaround', False):
615 try:
616 import pty
617 master, slave = pty.openpty()
618 width = compat_get_terminal_size().columns
619 width_args = [] if width is None else ['-w', str(width)]
620 sp_kwargs = {'stdin': subprocess.PIPE, 'stdout': slave, 'stderr': self._out_files.error}
621 try:
622 self._output_process = Popen(['bidiv'] + width_args, **sp_kwargs)
623 except OSError:
624 self._output_process = Popen(['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs)
625 self._output_channel = os.fdopen(master, 'rb')
626 except OSError as ose:
627 if ose.errno == errno.ENOENT:
628 self.report_warning(
629 'Could not find fribidi executable, ignoring --bidi-workaround. '
630 'Make sure that fribidi is an executable file in one of the directories in your $PATH.')
631 else:
632 raise
633
634 if auto_init:
635 if auto_init != 'no_verbose_header':
636 self.print_debug_header()
637 self.add_default_info_extractors()
638
639 if (sys.platform != 'win32'
640 and sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968']
641 and not self.params.get('restrictfilenames', False)):
642 # Unicode filesystem API will throw errors (#1474, #13027)
643 self.report_warning(
644 'Assuming --restrict-filenames since file system encoding '
645 'cannot encode all characters. '
646 'Set the LC_ALL environment variable to fix this.')
647 self.params['restrictfilenames'] = True
648
649 self.outtmpl_dict = self.parse_outtmpl()
650
651 # Creating format selector here allows us to catch syntax errors before the extraction
652 self.format_selector = (
653 self.params.get('format') if self.params.get('format') in (None, '-')
654 else self.params['format'] if callable(self.params['format'])
655 else self.build_format_selector(self.params['format']))
656
657 # Set http_headers defaults according to std_headers
658 self.params['http_headers'] = merge_headers(std_headers, self.params.get('http_headers', {}))
659
660 hooks = {
661 'post_hooks': self.add_post_hook,
662 'progress_hooks': self.add_progress_hook,
663 'postprocessor_hooks': self.add_postprocessor_hook,
664 }
665 for opt, fn in hooks.items():
666 for ph in self.params.get(opt, []):
667 fn(ph)
668
669 for pp_def_raw in self.params.get('postprocessors', []):
670 pp_def = dict(pp_def_raw)
671 when = pp_def.pop('when', 'post_process')
672 self.add_post_processor(
673 get_postprocessor(pp_def.pop('key'))(self, **pp_def),
674 when=when)
675
676 self._setup_opener()
677 register_socks_protocols()
678
679 def preload_download_archive(fn):
680 """Preload the archive, if any is specified"""
681 if fn is None:
682 return False
683 self.write_debug(f'Loading archive file {fn!r}')
684 try:
685 with locked_file(fn, 'r', encoding='utf-8') as archive_file:
686 for line in archive_file:
687 self.archive.add(line.strip())
688 except OSError as ioe:
689 if ioe.errno != errno.ENOENT:
690 raise
691 return False
692 return True
693
694 self.archive = set()
695 preload_download_archive(self.params.get('download_archive'))
696
697 def warn_if_short_id(self, argv):
698 # short YouTube ID starting with dash?
699 idxs = [
700 i for i, a in enumerate(argv)
701 if re.match(r'^-[0-9A-Za-z_-]{10}$', a)]
702 if idxs:
703 correct_argv = (
704 ['yt-dlp']
705 + [a for i, a in enumerate(argv) if i not in idxs]
706 + ['--'] + [argv[i] for i in idxs]
707 )
708 self.report_warning(
709 'Long argument string detected. '
710 'Use -- to separate parameters and URLs, like this:\n%s' %
711 args_to_str(correct_argv))
712
713 def add_info_extractor(self, ie):
714 """Add an InfoExtractor object to the end of the list."""
715 ie_key = ie.ie_key()
716 self._ies[ie_key] = ie
717 if not isinstance(ie, type):
718 self._ies_instances[ie_key] = ie
719 ie.set_downloader(self)
720
721 def _get_info_extractor_class(self, ie_key):
722 ie = self._ies.get(ie_key)
723 if ie is None:
724 ie = get_info_extractor(ie_key)
725 self.add_info_extractor(ie)
726 return ie
727
728 def get_info_extractor(self, ie_key):
729 """
730 Get an instance of an IE with name ie_key, it will try to get one from
731 the _ies list, if there's no instance it will create a new one and add
732 it to the extractor list.
733 """
734 ie = self._ies_instances.get(ie_key)
735 if ie is None:
736 ie = get_info_extractor(ie_key)()
737 self.add_info_extractor(ie)
738 return ie
739
740 def add_default_info_extractors(self):
741 """
742 Add the InfoExtractors returned by gen_extractors to the end of the list
743 """
744 for ie in gen_extractor_classes():
745 self.add_info_extractor(ie)
746
747 def add_post_processor(self, pp, when='post_process'):
748 """Add a PostProcessor object to the end of the chain."""
749 self._pps[when].append(pp)
750 pp.set_downloader(self)
751
752 def add_post_hook(self, ph):
753 """Add the post hook"""
754 self._post_hooks.append(ph)
755
756 def add_progress_hook(self, ph):
757 """Add the download progress hook"""
758 self._progress_hooks.append(ph)
759
760 def add_postprocessor_hook(self, ph):
761 """Add the postprocessing progress hook"""
762 self._postprocessor_hooks.append(ph)
763 for pps in self._pps.values():
764 for pp in pps:
765 pp.add_progress_hook(ph)
766
767 def _bidi_workaround(self, message):
768 if not hasattr(self, '_output_channel'):
769 return message
770
771 assert hasattr(self, '_output_process')
772 assert isinstance(message, compat_str)
773 line_count = message.count('\n') + 1
774 self._output_process.stdin.write((message + '\n').encode())
775 self._output_process.stdin.flush()
776 res = ''.join(self._output_channel.readline().decode()
777 for _ in range(line_count))
778 return res[:-len('\n')]
779
780 def _write_string(self, message, out=None, only_once=False):
781 if only_once:
782 if message in self._printed_messages:
783 return
784 self._printed_messages.add(message)
785 write_string(message, out=out, encoding=self.params.get('encoding'))
786
787 def to_stdout(self, message, skip_eol=False, quiet=None):
788 """Print message to stdout"""
789 if quiet is not None:
790 self.deprecation_warning('"YoutubeDL.to_stdout" no longer accepts the argument quiet. Use "YoutubeDL.to_screen" instead')
791 self._write_string(
792 '%s%s' % (self._bidi_workaround(message), ('' if skip_eol else '\n')),
793 self._out_files.out)
794
795 def to_screen(self, message, skip_eol=False, quiet=None):
796 """Print message to screen if not in quiet mode"""
797 if self.params.get('logger'):
798 self.params['logger'].debug(message)
799 return
800 if (self.params.get('quiet') if quiet is None else quiet) and not self.params.get('verbose'):
801 return
802 self._write_string(
803 '%s%s' % (self._bidi_workaround(message), ('' if skip_eol else '\n')),
804 self._out_files.screen)
805
806 def to_stderr(self, message, only_once=False):
807 """Print message to stderr"""
808 assert isinstance(message, compat_str)
809 if self.params.get('logger'):
810 self.params['logger'].error(message)
811 else:
812 self._write_string(f'{self._bidi_workaround(message)}\n', self._out_files.error, only_once=only_once)
813
814 def _send_console_code(self, code):
815 if compat_os_name == 'nt' or not self._out_files.console:
816 return
817 self._write_string(code, self._out_files.console)
818
819 def to_console_title(self, message):
820 if not self.params.get('consoletitle', False):
821 return
822 message = remove_terminal_sequences(message)
823 if compat_os_name == 'nt':
824 if ctypes.windll.kernel32.GetConsoleWindow():
825 # c_wchar_p() might not be necessary if `message` is
826 # already of type unicode()
827 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
828 else:
829 self._send_console_code(f'\033]0;{message}\007')
830
831 def save_console_title(self):
832 if not self.params.get('consoletitle') or self.params.get('simulate'):
833 return
834 self._send_console_code('\033[22;0t') # Save the title on stack
835
836 def restore_console_title(self):
837 if not self.params.get('consoletitle') or self.params.get('simulate'):
838 return
839 self._send_console_code('\033[23;0t') # Restore the title from stack
840
841 def __enter__(self):
842 self.save_console_title()
843 return self
844
845 def __exit__(self, *args):
846 self.restore_console_title()
847
848 if self.params.get('cookiefile') is not None:
849 self.cookiejar.save(ignore_discard=True, ignore_expires=True)
850
851 def trouble(self, message=None, tb=None, is_error=True):
852 """Determine action to take when a download problem appears.
853
854 Depending on if the downloader has been configured to ignore
855 download errors or not, this method may throw an exception or
856 not when errors are found, after printing the message.
857
858 @param tb If given, is additional traceback information
859 @param is_error Whether to raise error according to ignorerrors
860 """
861 if message is not None:
862 self.to_stderr(message)
863 if self.params.get('verbose'):
864 if tb is None:
865 if sys.exc_info()[0]: # if .trouble has been called from an except block
866 tb = ''
867 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
868 tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
869 tb += encode_compat_str(traceback.format_exc())
870 else:
871 tb_data = traceback.format_list(traceback.extract_stack())
872 tb = ''.join(tb_data)
873 if tb:
874 self.to_stderr(tb)
875 if not is_error:
876 return
877 if not self.params.get('ignoreerrors'):
878 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
879 exc_info = sys.exc_info()[1].exc_info
880 else:
881 exc_info = sys.exc_info()
882 raise DownloadError(message, exc_info)
883 self._download_retcode = 1
884
885 Styles = Namespace(
886 HEADERS='yellow',
887 EMPHASIS='light blue',
888 FILENAME='green',
889 ID='green',
890 DELIM='blue',
891 ERROR='red',
892 WARNING='yellow',
893 SUPPRESS='light black',
894 )
895
896 def _format_text(self, handle, allow_colors, text, f, fallback=None, *, test_encoding=False):
897 text = str(text)
898 if test_encoding:
899 original_text = text
900 # handle.encoding can be None. See https://github.com/yt-dlp/yt-dlp/issues/2711
901 encoding = self.params.get('encoding') or getattr(handle, 'encoding', None) or 'ascii'
902 text = text.encode(encoding, 'ignore').decode(encoding)
903 if fallback is not None and text != original_text:
904 text = fallback
905 return format_text(text, f) if allow_colors else text if fallback is None else fallback
906
907 def _format_out(self, *args, **kwargs):
908 return self._format_text(self._out_files.out, self._allow_colors.out, *args, **kwargs)
909
910 def _format_screen(self, *args, **kwargs):
911 return self._format_text(self._out_files.screen, self._allow_colors.screen, *args, **kwargs)
912
913 def _format_err(self, *args, **kwargs):
914 return self._format_text(self._out_files.error, self._allow_colors.error, *args, **kwargs)
915
916 def report_warning(self, message, only_once=False):
917 '''
918 Print the message to stderr, it will be prefixed with 'WARNING:'
919 If stderr is a tty file the 'WARNING:' will be colored
920 '''
921 if self.params.get('logger') is not None:
922 self.params['logger'].warning(message)
923 else:
924 if self.params.get('no_warnings'):
925 return
926 self.to_stderr(f'{self._format_err("WARNING:", self.Styles.WARNING)} {message}', only_once)
927
928 def deprecation_warning(self, message):
929 if self.params.get('logger') is not None:
930 self.params['logger'].warning(f'DeprecationWarning: {message}')
931 else:
932 self.to_stderr(f'{self._format_err("DeprecationWarning:", self.Styles.ERROR)} {message}', True)
933
934 def report_error(self, message, *args, **kwargs):
935 '''
936 Do the same as trouble, but prefixes the message with 'ERROR:', colored
937 in red if stderr is a tty file.
938 '''
939 self.trouble(f'{self._format_err("ERROR:", self.Styles.ERROR)} {message}', *args, **kwargs)
940
941 def write_debug(self, message, only_once=False):
942 '''Log debug message or Print message to stderr'''
943 if not self.params.get('verbose', False):
944 return
945 message = '[debug] %s' % message
946 if self.params.get('logger'):
947 self.params['logger'].debug(message)
948 else:
949 self.to_stderr(message, only_once)
950
951 def report_file_already_downloaded(self, file_name):
952 """Report file has already been fully downloaded."""
953 try:
954 self.to_screen('[download] %s has already been downloaded' % file_name)
955 except UnicodeEncodeError:
956 self.to_screen('[download] The file has already been downloaded')
957
958 def report_file_delete(self, file_name):
959 """Report that existing file will be deleted."""
960 try:
961 self.to_screen('Deleting existing file %s' % file_name)
962 except UnicodeEncodeError:
963 self.to_screen('Deleting existing file')
964
965 def raise_no_formats(self, info, forced=False, *, msg=None):
966 has_drm = info.get('_has_drm')
967 ignored, expected = self.params.get('ignore_no_formats_error'), bool(msg)
968 msg = msg or has_drm and 'This video is DRM protected' or 'No video formats found!'
969 if forced or not ignored:
970 raise ExtractorError(msg, video_id=info['id'], ie=info['extractor'],
971 expected=has_drm or ignored or expected)
972 else:
973 self.report_warning(msg)
974
975 def parse_outtmpl(self):
976 outtmpl_dict = self.params.get('outtmpl', {})
977 if not isinstance(outtmpl_dict, dict):
978 outtmpl_dict = {'default': outtmpl_dict}
979 # Remove spaces in the default template
980 if self.params.get('restrictfilenames'):
981 sanitize = lambda x: x.replace(' - ', ' ').replace(' ', '-')
982 else:
983 sanitize = lambda x: x
984 outtmpl_dict.update({
985 k: sanitize(v) for k, v in DEFAULT_OUTTMPL.items()
986 if outtmpl_dict.get(k) is None})
987 for _, val in outtmpl_dict.items():
988 if isinstance(val, bytes):
989 self.report_warning('Parameter outtmpl is bytes, but should be a unicode string')
990 return outtmpl_dict
991
992 def get_output_path(self, dir_type='', filename=None):
993 paths = self.params.get('paths', {})
994 assert isinstance(paths, dict)
995 path = os.path.join(
996 expand_path(paths.get('home', '').strip()),
997 expand_path(paths.get(dir_type, '').strip()) if dir_type else '',
998 filename or '')
999 return sanitize_path(path, force=self.params.get('windowsfilenames'))
1000
1001 @staticmethod
1002 def _outtmpl_expandpath(outtmpl):
1003 # expand_path translates '%%' into '%' and '$$' into '$'
1004 # correspondingly that is not what we want since we need to keep
1005 # '%%' intact for template dict substitution step. Working around
1006 # with boundary-alike separator hack.
1007 sep = ''.join([random.choice(ascii_letters) for _ in range(32)])
1008 outtmpl = outtmpl.replace('%%', f'%{sep}%').replace('$$', f'${sep}$')
1009
1010 # outtmpl should be expand_path'ed before template dict substitution
1011 # because meta fields may contain env variables we don't want to
1012 # be expanded. For example, for outtmpl "%(title)s.%(ext)s" and
1013 # title "Hello $PATH", we don't want `$PATH` to be expanded.
1014 return expand_path(outtmpl).replace(sep, '')
1015
1016 @staticmethod
1017 def escape_outtmpl(outtmpl):
1018 ''' Escape any remaining strings like %s, %abc% etc. '''
1019 return re.sub(
1020 STR_FORMAT_RE_TMPL.format('', '(?![%(\0])'),
1021 lambda mobj: ('' if mobj.group('has_key') else '%') + mobj.group(0),
1022 outtmpl)
1023
1024 @classmethod
1025 def validate_outtmpl(cls, outtmpl):
1026 ''' @return None or Exception object '''
1027 outtmpl = re.sub(
1028 STR_FORMAT_RE_TMPL.format('[^)]*', '[ljqBUDS]'),
1029 lambda mobj: f'{mobj.group(0)[:-1]}s',
1030 cls._outtmpl_expandpath(outtmpl))
1031 try:
1032 cls.escape_outtmpl(outtmpl) % collections.defaultdict(int)
1033 return None
1034 except ValueError as err:
1035 return err
1036
1037 @staticmethod
1038 def _copy_infodict(info_dict):
1039 info_dict = dict(info_dict)
1040 info_dict.pop('__postprocessors', None)
1041 return info_dict
1042
1043 def prepare_outtmpl(self, outtmpl, info_dict, sanitize=False):
1044 """ Make the outtmpl and info_dict suitable for substitution: ydl.escape_outtmpl(outtmpl) % info_dict
1045 @param sanitize Whether to sanitize the output as a filename.
1046 For backward compatibility, a function can also be passed
1047 """
1048
1049 info_dict.setdefault('epoch', int(time.time())) # keep epoch consistent once set
1050
1051 info_dict = self._copy_infodict(info_dict)
1052 info_dict['duration_string'] = ( # %(duration>%H-%M-%S)s is wrong if duration > 24hrs
1053 formatSeconds(info_dict['duration'], '-' if sanitize else ':')
1054 if info_dict.get('duration', None) is not None
1055 else None)
1056 info_dict['autonumber'] = int(self.params.get('autonumber_start', 1) - 1 + self._num_downloads)
1057 info_dict['video_autonumber'] = self._num_videos
1058 if info_dict.get('resolution') is None:
1059 info_dict['resolution'] = self.format_resolution(info_dict, default=None)
1060
1061 # For fields playlist_index, playlist_autonumber and autonumber convert all occurrences
1062 # of %(field)s to %(field)0Nd for backward compatibility
1063 field_size_compat_map = {
1064 'playlist_index': number_of_digits(info_dict.get('__last_playlist_index') or 0),
1065 'playlist_autonumber': number_of_digits(info_dict.get('n_entries') or 0),
1066 'autonumber': self.params.get('autonumber_size') or 5,
1067 }
1068
1069 TMPL_DICT = {}
1070 EXTERNAL_FORMAT_RE = re.compile(STR_FORMAT_RE_TMPL.format('[^)]*', f'[{STR_FORMAT_TYPES}ljqBUDS]'))
1071 MATH_FUNCTIONS = {
1072 '+': float.__add__,
1073 '-': float.__sub__,
1074 }
1075 # Field is of the form key1.key2...
1076 # where keys (except first) can be string, int or slice
1077 FIELD_RE = r'\w*(?:\.(?:\w+|{num}|{num}?(?::{num}?){{1,2}}))*'.format(num=r'(?:-?\d+)')
1078 MATH_FIELD_RE = rf'(?:{FIELD_RE}|-?{NUMBER_RE})'
1079 MATH_OPERATORS_RE = r'(?:%s)' % '|'.join(map(re.escape, MATH_FUNCTIONS.keys()))
1080 INTERNAL_FORMAT_RE = re.compile(rf'''(?x)
1081 (?P<negate>-)?
1082 (?P<fields>{FIELD_RE})
1083 (?P<maths>(?:{MATH_OPERATORS_RE}{MATH_FIELD_RE})*)
1084 (?:>(?P<strf_format>.+?))?
1085 (?P<remaining>
1086 (?P<alternate>(?<!\\),[^|&)]+)?
1087 (?:&(?P<replacement>.*?))?
1088 (?:\|(?P<default>.*?))?
1089 )$''')
1090
1091 def _traverse_infodict(k):
1092 k = k.split('.')
1093 if k[0] == '':
1094 k.pop(0)
1095 return traverse_obj(info_dict, k, is_user_input=True, traverse_string=True)
1096
1097 def get_value(mdict):
1098 # Object traversal
1099 value = _traverse_infodict(mdict['fields'])
1100 # Negative
1101 if mdict['negate']:
1102 value = float_or_none(value)
1103 if value is not None:
1104 value *= -1
1105 # Do maths
1106 offset_key = mdict['maths']
1107 if offset_key:
1108 value = float_or_none(value)
1109 operator = None
1110 while offset_key:
1111 item = re.match(
1112 MATH_FIELD_RE if operator else MATH_OPERATORS_RE,
1113 offset_key).group(0)
1114 offset_key = offset_key[len(item):]
1115 if operator is None:
1116 operator = MATH_FUNCTIONS[item]
1117 continue
1118 item, multiplier = (item[1:], -1) if item[0] == '-' else (item, 1)
1119 offset = float_or_none(item)
1120 if offset is None:
1121 offset = float_or_none(_traverse_infodict(item))
1122 try:
1123 value = operator(value, multiplier * offset)
1124 except (TypeError, ZeroDivisionError):
1125 return None
1126 operator = None
1127 # Datetime formatting
1128 if mdict['strf_format']:
1129 value = strftime_or_none(value, mdict['strf_format'].replace('\\,', ','))
1130
1131 return value
1132
1133 na = self.params.get('outtmpl_na_placeholder', 'NA')
1134
1135 def filename_sanitizer(key, value, restricted=self.params.get('restrictfilenames')):
1136 return sanitize_filename(str(value), restricted=restricted, is_id=(
1137 bool(re.search(r'(^|[_.])id(\.|$)', key))
1138 if 'filename-sanitization' in self.params.get('compat_opts', [])
1139 else NO_DEFAULT))
1140
1141 sanitizer = sanitize if callable(sanitize) else filename_sanitizer
1142 sanitize = bool(sanitize)
1143
1144 def _dumpjson_default(obj):
1145 if isinstance(obj, (set, LazyList)):
1146 return list(obj)
1147 return repr(obj)
1148
1149 def create_key(outer_mobj):
1150 if not outer_mobj.group('has_key'):
1151 return outer_mobj.group(0)
1152 key = outer_mobj.group('key')
1153 mobj = re.match(INTERNAL_FORMAT_RE, key)
1154 initial_field = mobj.group('fields') if mobj else ''
1155 value, replacement, default = None, None, na
1156 while mobj:
1157 mobj = mobj.groupdict()
1158 default = mobj['default'] if mobj['default'] is not None else default
1159 value = get_value(mobj)
1160 replacement = mobj['replacement']
1161 if value is None and mobj['alternate']:
1162 mobj = re.match(INTERNAL_FORMAT_RE, mobj['remaining'][1:])
1163 else:
1164 break
1165
1166 fmt = outer_mobj.group('format')
1167 if fmt == 's' and value is not None and key in field_size_compat_map.keys():
1168 fmt = f'0{field_size_compat_map[key]:d}d'
1169
1170 value = default if value is None else value if replacement is None else replacement
1171
1172 flags = outer_mobj.group('conversion') or ''
1173 str_fmt = f'{fmt[:-1]}s'
1174 if fmt[-1] == 'l': # list
1175 delim = '\n' if '#' in flags else ', '
1176 value, fmt = delim.join(map(str, variadic(value, allowed_types=(str, bytes)))), str_fmt
1177 elif fmt[-1] == 'j': # json
1178 value, fmt = json.dumps(value, default=_dumpjson_default, indent=4 if '#' in flags else None), str_fmt
1179 elif fmt[-1] == 'q': # quoted
1180 value = map(str, variadic(value) if '#' in flags else [value])
1181 value, fmt = ' '.join(map(compat_shlex_quote, value)), str_fmt
1182 elif fmt[-1] == 'B': # bytes
1183 value = f'%{str_fmt}'.encode() % str(value).encode()
1184 value, fmt = value.decode('utf-8', 'ignore'), 's'
1185 elif fmt[-1] == 'U': # unicode normalized
1186 value, fmt = unicodedata.normalize(
1187 # "+" = compatibility equivalence, "#" = NFD
1188 'NF%s%s' % ('K' if '+' in flags else '', 'D' if '#' in flags else 'C'),
1189 value), str_fmt
1190 elif fmt[-1] == 'D': # decimal suffix
1191 num_fmt, fmt = fmt[:-1].replace('#', ''), 's'
1192 value = format_decimal_suffix(value, f'%{num_fmt}f%s' if num_fmt else '%d%s',
1193 factor=1024 if '#' in flags else 1000)
1194 elif fmt[-1] == 'S': # filename sanitization
1195 value, fmt = filename_sanitizer(initial_field, value, restricted='#' in flags), str_fmt
1196 elif fmt[-1] == 'c':
1197 if value:
1198 value = str(value)[0]
1199 else:
1200 fmt = str_fmt
1201 elif fmt[-1] not in 'rs': # numeric
1202 value = float_or_none(value)
1203 if value is None:
1204 value, fmt = default, 's'
1205
1206 if sanitize:
1207 if fmt[-1] == 'r':
1208 # If value is an object, sanitize might convert it to a string
1209 # So we convert it to repr first
1210 value, fmt = repr(value), str_fmt
1211 if fmt[-1] in 'csr':
1212 value = sanitizer(initial_field, value)
1213
1214 key = '%s\0%s' % (key.replace('%', '%\0'), outer_mobj.group('format'))
1215 TMPL_DICT[key] = value
1216 return '{prefix}%({key}){fmt}'.format(key=key, fmt=fmt, prefix=outer_mobj.group('prefix'))
1217
1218 return EXTERNAL_FORMAT_RE.sub(create_key, outtmpl), TMPL_DICT
1219
1220 def evaluate_outtmpl(self, outtmpl, info_dict, *args, **kwargs):
1221 outtmpl, info_dict = self.prepare_outtmpl(outtmpl, info_dict, *args, **kwargs)
1222 return self.escape_outtmpl(outtmpl) % info_dict
1223
1224 def _prepare_filename(self, info_dict, *, outtmpl=None, tmpl_type=None):
1225 assert None in (outtmpl, tmpl_type), 'outtmpl and tmpl_type are mutually exclusive'
1226 if outtmpl is None:
1227 outtmpl = self.outtmpl_dict.get(tmpl_type or 'default', self.outtmpl_dict['default'])
1228 try:
1229 outtmpl = self._outtmpl_expandpath(outtmpl)
1230 filename = self.evaluate_outtmpl(outtmpl, info_dict, True)
1231 if not filename:
1232 return None
1233
1234 if tmpl_type in ('', 'temp'):
1235 final_ext, ext = self.params.get('final_ext'), info_dict.get('ext')
1236 if final_ext and ext and final_ext != ext and filename.endswith(f'.{final_ext}'):
1237 filename = replace_extension(filename, ext, final_ext)
1238 elif tmpl_type:
1239 force_ext = OUTTMPL_TYPES[tmpl_type]
1240 if force_ext:
1241 filename = replace_extension(filename, force_ext, info_dict.get('ext'))
1242
1243 # https://github.com/blackjack4494/youtube-dlc/issues/85
1244 trim_file_name = self.params.get('trim_file_name', False)
1245 if trim_file_name:
1246 no_ext, *ext = filename.rsplit('.', 2)
1247 filename = join_nonempty(no_ext[:trim_file_name], *ext, delim='.')
1248
1249 return filename
1250 except ValueError as err:
1251 self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')')
1252 return None
1253
1254 def prepare_filename(self, info_dict, dir_type='', *, outtmpl=None, warn=False):
1255 """Generate the output filename"""
1256 if outtmpl:
1257 assert not dir_type, 'outtmpl and dir_type are mutually exclusive'
1258 dir_type = None
1259 filename = self._prepare_filename(info_dict, tmpl_type=dir_type, outtmpl=outtmpl)
1260 if not filename and dir_type not in ('', 'temp'):
1261 return ''
1262
1263 if warn:
1264 if not self.params.get('paths'):
1265 pass
1266 elif filename == '-':
1267 self.report_warning('--paths is ignored when an outputting to stdout', only_once=True)
1268 elif os.path.isabs(filename):
1269 self.report_warning('--paths is ignored since an absolute path is given in output template', only_once=True)
1270 if filename == '-' or not filename:
1271 return filename
1272
1273 return self.get_output_path(dir_type, filename)
1274
1275 def _match_entry(self, info_dict, incomplete=False, silent=False):
1276 """ Returns None if the file should be downloaded """
1277
1278 video_title = info_dict.get('title', info_dict.get('id', 'video'))
1279
1280 def check_filter():
1281 if 'title' in info_dict:
1282 # This can happen when we're just evaluating the playlist
1283 title = info_dict['title']
1284 matchtitle = self.params.get('matchtitle', False)
1285 if matchtitle:
1286 if not re.search(matchtitle, title, re.IGNORECASE):
1287 return '"' + title + '" title did not match pattern "' + matchtitle + '"'
1288 rejecttitle = self.params.get('rejecttitle', False)
1289 if rejecttitle:
1290 if re.search(rejecttitle, title, re.IGNORECASE):
1291 return '"' + title + '" title matched reject pattern "' + rejecttitle + '"'
1292 date = info_dict.get('upload_date')
1293 if date is not None:
1294 dateRange = self.params.get('daterange', DateRange())
1295 if date not in dateRange:
1296 return f'{date_from_str(date).isoformat()} upload date is not in range {dateRange}'
1297 view_count = info_dict.get('view_count')
1298 if view_count is not None:
1299 min_views = self.params.get('min_views')
1300 if min_views is not None and view_count < min_views:
1301 return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views)
1302 max_views = self.params.get('max_views')
1303 if max_views is not None and view_count > max_views:
1304 return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views)
1305 if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')):
1306 return 'Skipping "%s" because it is age restricted' % video_title
1307
1308 match_filter = self.params.get('match_filter')
1309 if match_filter is not None:
1310 try:
1311 ret = match_filter(info_dict, incomplete=incomplete)
1312 except TypeError:
1313 # For backward compatibility
1314 ret = None if incomplete else match_filter(info_dict)
1315 if ret is NO_DEFAULT:
1316 while True:
1317 filename = self._format_screen(self.prepare_filename(info_dict), self.Styles.FILENAME)
1318 reply = input(self._format_screen(
1319 f'Download "{filename}"? (Y/n): ', self.Styles.EMPHASIS)).lower().strip()
1320 if reply in {'y', ''}:
1321 return None
1322 elif reply == 'n':
1323 return f'Skipping {video_title}'
1324 elif ret is not None:
1325 return ret
1326 return None
1327
1328 if self.in_download_archive(info_dict):
1329 reason = '%s has already been recorded in the archive' % video_title
1330 break_opt, break_err = 'break_on_existing', ExistingVideoReached
1331 else:
1332 reason = check_filter()
1333 break_opt, break_err = 'break_on_reject', RejectedVideoReached
1334 if reason is not None:
1335 if not silent:
1336 self.to_screen('[download] ' + reason)
1337 if self.params.get(break_opt, False):
1338 raise break_err()
1339 return reason
1340
1341 @staticmethod
1342 def add_extra_info(info_dict, extra_info):
1343 '''Set the keys from extra_info in info dict if they are missing'''
1344 for key, value in extra_info.items():
1345 info_dict.setdefault(key, value)
1346
1347 def extract_info(self, url, download=True, ie_key=None, extra_info=None,
1348 process=True, force_generic_extractor=False):
1349 """
1350 Return a list with a dictionary for each video extracted.
1351
1352 Arguments:
1353 url -- URL to extract
1354
1355 Keyword arguments:
1356 download -- whether to download videos during extraction
1357 ie_key -- extractor key hint
1358 extra_info -- dictionary containing the extra values to add to each result
1359 process -- whether to resolve all unresolved references (URLs, playlist items),
1360 must be True for download to work.
1361 force_generic_extractor -- force using the generic extractor
1362 """
1363
1364 if extra_info is None:
1365 extra_info = {}
1366
1367 if not ie_key and force_generic_extractor:
1368 ie_key = 'Generic'
1369
1370 if ie_key:
1371 ies = {ie_key: self._get_info_extractor_class(ie_key)}
1372 else:
1373 ies = self._ies
1374
1375 for ie_key, ie in ies.items():
1376 if not ie.suitable(url):
1377 continue
1378
1379 if not ie.working():
1380 self.report_warning('The program functionality for this site has been marked as broken, '
1381 'and will probably not work.')
1382
1383 temp_id = ie.get_temp_id(url)
1384 if temp_id is not None and self.in_download_archive({'id': temp_id, 'ie_key': ie_key}):
1385 self.to_screen(f'[{ie_key}] {temp_id}: has already been recorded in the archive')
1386 if self.params.get('break_on_existing', False):
1387 raise ExistingVideoReached()
1388 break
1389 return self.__extract_info(url, self.get_info_extractor(ie_key), download, extra_info, process)
1390 else:
1391 self.report_error('no suitable InfoExtractor for URL %s' % url)
1392
1393 def __handle_extraction_exceptions(func):
1394 @functools.wraps(func)
1395 def wrapper(self, *args, **kwargs):
1396 while True:
1397 try:
1398 return func(self, *args, **kwargs)
1399 except (DownloadCancelled, LazyList.IndexError, PagedList.IndexError):
1400 raise
1401 except ReExtractInfo as e:
1402 if e.expected:
1403 self.to_screen(f'{e}; Re-extracting data')
1404 else:
1405 self.to_stderr('\r')
1406 self.report_warning(f'{e}; Re-extracting data')
1407 continue
1408 except GeoRestrictedError as e:
1409 msg = e.msg
1410 if e.countries:
1411 msg += '\nThis video is available in %s.' % ', '.join(
1412 map(ISO3166Utils.short2full, e.countries))
1413 msg += '\nYou might want to use a VPN or a proxy server (with --proxy) to workaround.'
1414 self.report_error(msg)
1415 except ExtractorError as e: # An error we somewhat expected
1416 self.report_error(str(e), e.format_traceback())
1417 except Exception as e:
1418 if self.params.get('ignoreerrors'):
1419 self.report_error(str(e), tb=encode_compat_str(traceback.format_exc()))
1420 else:
1421 raise
1422 break
1423 return wrapper
1424
1425 def _wait_for_video(self, ie_result):
1426 if (not self.params.get('wait_for_video')
1427 or ie_result.get('_type', 'video') != 'video'
1428 or ie_result.get('formats') or ie_result.get('url')):
1429 return
1430
1431 format_dur = lambda dur: '%02d:%02d:%02d' % timetuple_from_msec(dur * 1000)[:-1]
1432 last_msg = ''
1433
1434 def progress(msg):
1435 nonlocal last_msg
1436 self.to_screen(msg + ' ' * (len(last_msg) - len(msg)) + '\r', skip_eol=True)
1437 last_msg = msg
1438
1439 min_wait, max_wait = self.params.get('wait_for_video')
1440 diff = try_get(ie_result, lambda x: x['release_timestamp'] - time.time())
1441 if diff is None and ie_result.get('live_status') == 'is_upcoming':
1442 diff = round(random.uniform(min_wait, max_wait) if (max_wait and min_wait) else (max_wait or min_wait), 0)
1443 self.report_warning('Release time of video is not known')
1444 elif (diff or 0) <= 0:
1445 self.report_warning('Video should already be available according to extracted info')
1446 diff = min(max(diff or 0, min_wait or 0), max_wait or float('inf'))
1447 self.to_screen(f'[wait] Waiting for {format_dur(diff)} - Press Ctrl+C to try now')
1448
1449 wait_till = time.time() + diff
1450 try:
1451 while True:
1452 diff = wait_till - time.time()
1453 if diff <= 0:
1454 progress('')
1455 raise ReExtractInfo('[wait] Wait period ended', expected=True)
1456 progress(f'[wait] Remaining time until next attempt: {self._format_screen(format_dur(diff), self.Styles.EMPHASIS)}')
1457 time.sleep(1)
1458 except KeyboardInterrupt:
1459 progress('')
1460 raise ReExtractInfo('[wait] Interrupted by user', expected=True)
1461 except BaseException as e:
1462 if not isinstance(e, ReExtractInfo):
1463 self.to_screen('')
1464 raise
1465
1466 @__handle_extraction_exceptions
1467 def __extract_info(self, url, ie, download, extra_info, process):
1468 ie_result = ie.extract(url)
1469 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
1470 return
1471 if isinstance(ie_result, list):
1472 # Backwards compatibility: old IE result format
1473 ie_result = {
1474 '_type': 'compat_list',
1475 'entries': ie_result,
1476 }
1477 if extra_info.get('original_url'):
1478 ie_result.setdefault('original_url', extra_info['original_url'])
1479 self.add_default_extra_info(ie_result, ie, url)
1480 if process:
1481 self._wait_for_video(ie_result)
1482 return self.process_ie_result(ie_result, download, extra_info)
1483 else:
1484 return ie_result
1485
1486 def add_default_extra_info(self, ie_result, ie, url):
1487 if url is not None:
1488 self.add_extra_info(ie_result, {
1489 'webpage_url': url,
1490 'original_url': url,
1491 })
1492 webpage_url = ie_result.get('webpage_url')
1493 if webpage_url:
1494 self.add_extra_info(ie_result, {
1495 'webpage_url_basename': url_basename(webpage_url),
1496 'webpage_url_domain': get_domain(webpage_url),
1497 })
1498 if ie is not None:
1499 self.add_extra_info(ie_result, {
1500 'extractor': ie.IE_NAME,
1501 'extractor_key': ie.ie_key(),
1502 })
1503
1504 def process_ie_result(self, ie_result, download=True, extra_info=None):
1505 """
1506 Take the result of the ie(may be modified) and resolve all unresolved
1507 references (URLs, playlist items).
1508
1509 It will also download the videos if 'download'.
1510 Returns the resolved ie_result.
1511 """
1512 if extra_info is None:
1513 extra_info = {}
1514 result_type = ie_result.get('_type', 'video')
1515
1516 if result_type in ('url', 'url_transparent'):
1517 ie_result['url'] = sanitize_url(ie_result['url'])
1518 if ie_result.get('original_url'):
1519 extra_info.setdefault('original_url', ie_result['original_url'])
1520
1521 extract_flat = self.params.get('extract_flat', False)
1522 if ((extract_flat == 'in_playlist' and 'playlist' in extra_info)
1523 or extract_flat is True):
1524 info_copy = ie_result.copy()
1525 ie = try_get(ie_result.get('ie_key'), self.get_info_extractor)
1526 if ie and not ie_result.get('id'):
1527 info_copy['id'] = ie.get_temp_id(ie_result['url'])
1528 self.add_default_extra_info(info_copy, ie, ie_result['url'])
1529 self.add_extra_info(info_copy, extra_info)
1530 info_copy, _ = self.pre_process(info_copy)
1531 self.__forced_printings(info_copy, self.prepare_filename(info_copy), incomplete=True)
1532 if self.params.get('force_write_download_archive', False):
1533 self.record_download_archive(info_copy)
1534 return ie_result
1535
1536 if result_type == 'video':
1537 self.add_extra_info(ie_result, extra_info)
1538 ie_result = self.process_video_result(ie_result, download=download)
1539 additional_urls = (ie_result or {}).get('additional_urls')
1540 if additional_urls:
1541 # TODO: Improve MetadataParserPP to allow setting a list
1542 if isinstance(additional_urls, compat_str):
1543 additional_urls = [additional_urls]
1544 self.to_screen(
1545 '[info] %s: %d additional URL(s) requested' % (ie_result['id'], len(additional_urls)))
1546 self.write_debug('Additional URLs: "%s"' % '", "'.join(additional_urls))
1547 ie_result['additional_entries'] = [
1548 self.extract_info(
1549 url, download, extra_info=extra_info,
1550 force_generic_extractor=self.params.get('force_generic_extractor'))
1551 for url in additional_urls
1552 ]
1553 return ie_result
1554 elif result_type == 'url':
1555 # We have to add extra_info to the results because it may be
1556 # contained in a playlist
1557 return self.extract_info(
1558 ie_result['url'], download,
1559 ie_key=ie_result.get('ie_key'),
1560 extra_info=extra_info)
1561 elif result_type == 'url_transparent':
1562 # Use the information from the embedding page
1563 info = self.extract_info(
1564 ie_result['url'], ie_key=ie_result.get('ie_key'),
1565 extra_info=extra_info, download=False, process=False)
1566
1567 # extract_info may return None when ignoreerrors is enabled and
1568 # extraction failed with an error, don't crash and return early
1569 # in this case
1570 if not info:
1571 return info
1572
1573 new_result = info.copy()
1574 new_result.update(filter_dict(ie_result, lambda k, v: (
1575 v is not None and k not in {'_type', 'url', 'id', 'extractor', 'extractor_key', 'ie_key'})))
1576
1577 # Extracted info may not be a video result (i.e.
1578 # info.get('_type', 'video') != video) but rather an url or
1579 # url_transparent. In such cases outer metadata (from ie_result)
1580 # should be propagated to inner one (info). For this to happen
1581 # _type of info should be overridden with url_transparent. This
1582 # fixes issue from https://github.com/ytdl-org/youtube-dl/pull/11163.
1583 if new_result.get('_type') == 'url':
1584 new_result['_type'] = 'url_transparent'
1585
1586 return self.process_ie_result(
1587 new_result, download=download, extra_info=extra_info)
1588 elif result_type in ('playlist', 'multi_video'):
1589 # Protect from infinite recursion due to recursively nested playlists
1590 # (see https://github.com/ytdl-org/youtube-dl/issues/27833)
1591 webpage_url = ie_result['webpage_url']
1592 if webpage_url in self._playlist_urls:
1593 self.to_screen(
1594 '[download] Skipping already downloaded playlist: %s'
1595 % ie_result.get('title') or ie_result.get('id'))
1596 return
1597
1598 self._playlist_level += 1
1599 self._playlist_urls.add(webpage_url)
1600 self._fill_common_fields(ie_result, False)
1601 self._sanitize_thumbnails(ie_result)
1602 try:
1603 return self.__process_playlist(ie_result, download)
1604 finally:
1605 self._playlist_level -= 1
1606 if not self._playlist_level:
1607 self._playlist_urls.clear()
1608 elif result_type == 'compat_list':
1609 self.report_warning(
1610 'Extractor %s returned a compat_list result. '
1611 'It needs to be updated.' % ie_result.get('extractor'))
1612
1613 def _fixup(r):
1614 self.add_extra_info(r, {
1615 'extractor': ie_result['extractor'],
1616 'webpage_url': ie_result['webpage_url'],
1617 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1618 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1619 'extractor_key': ie_result['extractor_key'],
1620 })
1621 return r
1622 ie_result['entries'] = [
1623 self.process_ie_result(_fixup(r), download, extra_info)
1624 for r in ie_result['entries']
1625 ]
1626 return ie_result
1627 else:
1628 raise Exception('Invalid result type: %s' % result_type)
1629
1630 def _ensure_dir_exists(self, path):
1631 return make_dir(path, self.report_error)
1632
1633 @staticmethod
1634 def _playlist_infodict(ie_result, **kwargs):
1635 return {
1636 **ie_result,
1637 'playlist': ie_result.get('title') or ie_result.get('id'),
1638 'playlist_id': ie_result.get('id'),
1639 'playlist_title': ie_result.get('title'),
1640 'playlist_uploader': ie_result.get('uploader'),
1641 'playlist_uploader_id': ie_result.get('uploader_id'),
1642 'playlist_index': 0,
1643 **kwargs,
1644 }
1645
1646 def __process_playlist(self, ie_result, download):
1647 # We process each entry in the playlist
1648 playlist = ie_result.get('title') or ie_result.get('id')
1649 self.to_screen('[download] Downloading playlist: %s' % playlist)
1650
1651 if 'entries' not in ie_result:
1652 raise EntryNotInPlaylist('There are no entries')
1653
1654 MissingEntry = object()
1655 incomplete_entries = bool(ie_result.get('requested_entries'))
1656 if incomplete_entries:
1657 def fill_missing_entries(entries, indices):
1658 ret = [MissingEntry] * max(indices)
1659 for i, entry in zip(indices, entries):
1660 ret[i - 1] = entry
1661 return ret
1662 ie_result['entries'] = fill_missing_entries(ie_result['entries'], ie_result['requested_entries'])
1663
1664 playlist_results = []
1665
1666 playliststart = self.params.get('playliststart', 1)
1667 playlistend = self.params.get('playlistend')
1668 # For backwards compatibility, interpret -1 as whole list
1669 if playlistend == -1:
1670 playlistend = None
1671
1672 playlistitems_str = self.params.get('playlist_items')
1673 playlistitems = None
1674 if playlistitems_str is not None:
1675 def iter_playlistitems(format):
1676 for string_segment in format.split(','):
1677 if '-' in string_segment:
1678 start, end = string_segment.split('-')
1679 for item in range(int(start), int(end) + 1):
1680 yield int(item)
1681 else:
1682 yield int(string_segment)
1683 playlistitems = orderedSet(iter_playlistitems(playlistitems_str))
1684
1685 ie_entries = ie_result['entries']
1686 if isinstance(ie_entries, list):
1687 playlist_count = len(ie_entries)
1688 msg = f'Collected {playlist_count} videos; downloading %d of them'
1689 ie_result['playlist_count'] = ie_result.get('playlist_count') or playlist_count
1690
1691 def get_entry(i):
1692 return ie_entries[i - 1]
1693 else:
1694 msg = 'Downloading %d videos'
1695 if not isinstance(ie_entries, (PagedList, LazyList)):
1696 ie_entries = LazyList(ie_entries)
1697 elif isinstance(ie_entries, InAdvancePagedList):
1698 if ie_entries._pagesize == 1:
1699 playlist_count = ie_entries._pagecount
1700
1701 def get_entry(i):
1702 return YoutubeDL.__handle_extraction_exceptions(
1703 lambda self, i: ie_entries[i - 1]
1704 )(self, i)
1705
1706 entries, broken = [], False
1707 items = playlistitems if playlistitems is not None else itertools.count(playliststart)
1708 for i in items:
1709 if i == 0:
1710 continue
1711 if playlistitems is None and playlistend is not None and playlistend < i:
1712 break
1713 entry = None
1714 try:
1715 entry = get_entry(i)
1716 if entry is MissingEntry:
1717 raise EntryNotInPlaylist()
1718 except (IndexError, EntryNotInPlaylist):
1719 if incomplete_entries:
1720 raise EntryNotInPlaylist(f'Entry {i} cannot be found')
1721 elif not playlistitems:
1722 break
1723 entries.append(entry)
1724 try:
1725 if entry is not None:
1726 # TODO: Add auto-generated fields
1727 self._match_entry(entry, incomplete=True, silent=True)
1728 except (ExistingVideoReached, RejectedVideoReached):
1729 broken = True
1730 break
1731 ie_result['entries'] = entries
1732
1733 # Save playlist_index before re-ordering
1734 entries = [
1735 ((playlistitems[i - 1] if playlistitems else i + playliststart - 1), entry)
1736 for i, entry in enumerate(entries, 1)
1737 if entry is not None]
1738 n_entries = len(entries)
1739
1740 if not (ie_result.get('playlist_count') or broken or playlistitems or playlistend):
1741 ie_result['playlist_count'] = n_entries
1742
1743 if not playlistitems and (playliststart != 1 or playlistend):
1744 playlistitems = list(range(playliststart, playliststart + n_entries))
1745 ie_result['requested_entries'] = playlistitems
1746
1747 _infojson_written = False
1748 write_playlist_files = self.params.get('allow_playlist_files', True)
1749 if write_playlist_files and self.params.get('list_thumbnails'):
1750 self.list_thumbnails(ie_result)
1751 if write_playlist_files and not self.params.get('simulate'):
1752 ie_copy = self._playlist_infodict(ie_result, n_entries=n_entries)
1753 _infojson_written = self._write_info_json(
1754 'playlist', ie_result, self.prepare_filename(ie_copy, 'pl_infojson'))
1755 if _infojson_written is None:
1756 return
1757 if self._write_description('playlist', ie_result,
1758 self.prepare_filename(ie_copy, 'pl_description')) is None:
1759 return
1760 # TODO: This should be passed to ThumbnailsConvertor if necessary
1761 self._write_thumbnails('playlist', ie_copy, self.prepare_filename(ie_copy, 'pl_thumbnail'))
1762
1763 if self.params.get('playlistreverse', False):
1764 entries = entries[::-1]
1765 if self.params.get('playlistrandom', False):
1766 random.shuffle(entries)
1767
1768 x_forwarded_for = ie_result.get('__x_forwarded_for_ip')
1769
1770 self.to_screen(f'[{ie_result["extractor"]}] playlist {playlist}: {msg % n_entries}')
1771 failures = 0
1772 max_failures = self.params.get('skip_playlist_after_errors') or float('inf')
1773 for i, entry_tuple in enumerate(entries, 1):
1774 playlist_index, entry = entry_tuple
1775 if 'playlist-index' in self.params.get('compat_opts', []):
1776 playlist_index = playlistitems[i - 1] if playlistitems else i + playliststart - 1
1777 self.to_screen('[download] Downloading video %s of %s' % (
1778 self._format_screen(i, self.Styles.ID), self._format_screen(n_entries, self.Styles.EMPHASIS)))
1779 # This __x_forwarded_for_ip thing is a bit ugly but requires
1780 # minimal changes
1781 if x_forwarded_for:
1782 entry['__x_forwarded_for_ip'] = x_forwarded_for
1783 extra = {
1784 'n_entries': n_entries,
1785 '__last_playlist_index': max(playlistitems) if playlistitems else (playlistend or n_entries),
1786 'playlist_count': ie_result.get('playlist_count'),
1787 'playlist_index': playlist_index,
1788 'playlist_autonumber': i,
1789 'playlist': playlist,
1790 'playlist_id': ie_result.get('id'),
1791 'playlist_title': ie_result.get('title'),
1792 'playlist_uploader': ie_result.get('uploader'),
1793 'playlist_uploader_id': ie_result.get('uploader_id'),
1794 'extractor': ie_result['extractor'],
1795 'webpage_url': ie_result['webpage_url'],
1796 'webpage_url_basename': url_basename(ie_result['webpage_url']),
1797 'webpage_url_domain': get_domain(ie_result['webpage_url']),
1798 'extractor_key': ie_result['extractor_key'],
1799 }
1800
1801 if self._match_entry(entry, incomplete=True) is not None:
1802 continue
1803
1804 entry_result = self.__process_iterable_entry(entry, download, extra)
1805 if not entry_result:
1806 failures += 1
1807 if failures >= max_failures:
1808 self.report_error(
1809 'Skipping the remaining entries in playlist "%s" since %d items failed extraction' % (playlist, failures))
1810 break
1811 playlist_results.append(entry_result)
1812 ie_result['entries'] = playlist_results
1813
1814 # Write the updated info to json
1815 if _infojson_written is True and self._write_info_json(
1816 'updated playlist', ie_result,
1817 self.prepare_filename(ie_copy, 'pl_infojson'), overwrite=True) is None:
1818 return
1819
1820 ie_result = self.run_all_pps('playlist', ie_result)
1821 self.to_screen(f'[download] Finished downloading playlist: {playlist}')
1822 return ie_result
1823
1824 @__handle_extraction_exceptions
1825 def __process_iterable_entry(self, entry, download, extra_info):
1826 return self.process_ie_result(
1827 entry, download=download, extra_info=extra_info)
1828
1829 def _build_format_filter(self, filter_spec):
1830 " Returns a function to filter the formats according to the filter_spec "
1831
1832 OPERATORS = {
1833 '<': operator.lt,
1834 '<=': operator.le,
1835 '>': operator.gt,
1836 '>=': operator.ge,
1837 '=': operator.eq,
1838 '!=': operator.ne,
1839 }
1840 operator_rex = re.compile(r'''(?x)\s*
1841 (?P<key>width|height|tbr|abr|vbr|asr|filesize|filesize_approx|fps)\s*
1842 (?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
1843 (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)\s*
1844 ''' % '|'.join(map(re.escape, OPERATORS.keys())))
1845 m = operator_rex.fullmatch(filter_spec)
1846 if m:
1847 try:
1848 comparison_value = int(m.group('value'))
1849 except ValueError:
1850 comparison_value = parse_filesize(m.group('value'))
1851 if comparison_value is None:
1852 comparison_value = parse_filesize(m.group('value') + 'B')
1853 if comparison_value is None:
1854 raise ValueError(
1855 'Invalid value %r in format specification %r' % (
1856 m.group('value'), filter_spec))
1857 op = OPERATORS[m.group('op')]
1858
1859 if not m:
1860 STR_OPERATORS = {
1861 '=': operator.eq,
1862 '^=': lambda attr, value: attr.startswith(value),
1863 '$=': lambda attr, value: attr.endswith(value),
1864 '*=': lambda attr, value: value in attr,
1865 '~=': lambda attr, value: value.search(attr) is not None
1866 }
1867 str_operator_rex = re.compile(r'''(?x)\s*
1868 (?P<key>[a-zA-Z0-9._-]+)\s*
1869 (?P<negation>!\s*)?(?P<op>%s)\s*(?P<none_inclusive>\?\s*)?
1870 (?P<quote>["'])?
1871 (?P<value>(?(quote)(?:(?!(?P=quote))[^\\]|\\.)+|[\w.-]+))
1872 (?(quote)(?P=quote))\s*
1873 ''' % '|'.join(map(re.escape, STR_OPERATORS.keys())))
1874 m = str_operator_rex.fullmatch(filter_spec)
1875 if m:
1876 if m.group('op') == '~=':
1877 comparison_value = re.compile(m.group('value'))
1878 else:
1879 comparison_value = re.sub(r'''\\([\\"'])''', r'\1', m.group('value'))
1880 str_op = STR_OPERATORS[m.group('op')]
1881 if m.group('negation'):
1882 op = lambda attr, value: not str_op(attr, value)
1883 else:
1884 op = str_op
1885
1886 if not m:
1887 raise SyntaxError('Invalid filter specification %r' % filter_spec)
1888
1889 def _filter(f):
1890 actual_value = f.get(m.group('key'))
1891 if actual_value is None:
1892 return m.group('none_inclusive')
1893 return op(actual_value, comparison_value)
1894 return _filter
1895
1896 def _check_formats(self, formats):
1897 for f in formats:
1898 self.to_screen('[info] Testing format %s' % f['format_id'])
1899 path = self.get_output_path('temp')
1900 if not self._ensure_dir_exists(f'{path}/'):
1901 continue
1902 temp_file = tempfile.NamedTemporaryFile(suffix='.tmp', delete=False, dir=path or None)
1903 temp_file.close()
1904 try:
1905 success, _ = self.dl(temp_file.name, f, test=True)
1906 except (DownloadError, IOError, OSError, ValueError) + network_exceptions:
1907 success = False
1908 finally:
1909 if os.path.exists(temp_file.name):
1910 try:
1911 os.remove(temp_file.name)
1912 except OSError:
1913 self.report_warning('Unable to delete temporary file "%s"' % temp_file.name)
1914 if success:
1915 yield f
1916 else:
1917 self.to_screen('[info] Unable to download format %s. Skipping...' % f['format_id'])
1918
1919 def _default_format_spec(self, info_dict, download=True):
1920
1921 def can_merge():
1922 merger = FFmpegMergerPP(self)
1923 return merger.available and merger.can_merge()
1924
1925 prefer_best = (
1926 not self.params.get('simulate')
1927 and download
1928 and (
1929 not can_merge()
1930 or info_dict.get('is_live') and not self.params.get('live_from_start')
1931 or self.outtmpl_dict['default'] == '-'))
1932 compat = (
1933 prefer_best
1934 or self.params.get('allow_multiple_audio_streams', False)
1935 or 'format-spec' in self.params.get('compat_opts', []))
1936
1937 return (
1938 'best/bestvideo+bestaudio' if prefer_best
1939 else 'bestvideo*+bestaudio/best' if not compat
1940 else 'bestvideo+bestaudio/best')
1941
1942 def build_format_selector(self, format_spec):
1943 def syntax_error(note, start):
1944 message = (
1945 'Invalid format specification: '
1946 '{}\n\t{}\n\t{}^'.format(note, format_spec, ' ' * start[1]))
1947 return SyntaxError(message)
1948
1949 PICKFIRST = 'PICKFIRST'
1950 MERGE = 'MERGE'
1951 SINGLE = 'SINGLE'
1952 GROUP = 'GROUP'
1953 FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters'])
1954
1955 allow_multiple_streams = {'audio': self.params.get('allow_multiple_audio_streams', False),
1956 'video': self.params.get('allow_multiple_video_streams', False)}
1957
1958 check_formats = self.params.get('check_formats') == 'selected'
1959
1960 def _parse_filter(tokens):
1961 filter_parts = []
1962 for type, string, start, _, _ in tokens:
1963 if type == tokenize.OP and string == ']':
1964 return ''.join(filter_parts)
1965 else:
1966 filter_parts.append(string)
1967
1968 def _remove_unused_ops(tokens):
1969 # Remove operators that we don't use and join them with the surrounding strings
1970 # for example: 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9'
1971 ALLOWED_OPS = ('/', '+', ',', '(', ')')
1972 last_string, last_start, last_end, last_line = None, None, None, None
1973 for type, string, start, end, line in tokens:
1974 if type == tokenize.OP and string == '[':
1975 if last_string:
1976 yield tokenize.NAME, last_string, last_start, last_end, last_line
1977 last_string = None
1978 yield type, string, start, end, line
1979 # everything inside brackets will be handled by _parse_filter
1980 for type, string, start, end, line in tokens:
1981 yield type, string, start, end, line
1982 if type == tokenize.OP and string == ']':
1983 break
1984 elif type == tokenize.OP and string in ALLOWED_OPS:
1985 if last_string:
1986 yield tokenize.NAME, last_string, last_start, last_end, last_line
1987 last_string = None
1988 yield type, string, start, end, line
1989 elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]:
1990 if not last_string:
1991 last_string = string
1992 last_start = start
1993 last_end = end
1994 else:
1995 last_string += string
1996 if last_string:
1997 yield tokenize.NAME, last_string, last_start, last_end, last_line
1998
1999 def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False):
2000 selectors = []
2001 current_selector = None
2002 for type, string, start, _, _ in tokens:
2003 # ENCODING is only defined in python 3.x
2004 if type == getattr(tokenize, 'ENCODING', None):
2005 continue
2006 elif type in [tokenize.NAME, tokenize.NUMBER]:
2007 current_selector = FormatSelector(SINGLE, string, [])
2008 elif type == tokenize.OP:
2009 if string == ')':
2010 if not inside_group:
2011 # ')' will be handled by the parentheses group
2012 tokens.restore_last_token()
2013 break
2014 elif inside_merge and string in ['/', ',']:
2015 tokens.restore_last_token()
2016 break
2017 elif inside_choice and string == ',':
2018 tokens.restore_last_token()
2019 break
2020 elif string == ',':
2021 if not current_selector:
2022 raise syntax_error('"," must follow a format selector', start)
2023 selectors.append(current_selector)
2024 current_selector = None
2025 elif string == '/':
2026 if not current_selector:
2027 raise syntax_error('"/" must follow a format selector', start)
2028 first_choice = current_selector
2029 second_choice = _parse_format_selection(tokens, inside_choice=True)
2030 current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), [])
2031 elif string == '[':
2032 if not current_selector:
2033 current_selector = FormatSelector(SINGLE, 'best', [])
2034 format_filter = _parse_filter(tokens)
2035 current_selector.filters.append(format_filter)
2036 elif string == '(':
2037 if current_selector:
2038 raise syntax_error('Unexpected "("', start)
2039 group = _parse_format_selection(tokens, inside_group=True)
2040 current_selector = FormatSelector(GROUP, group, [])
2041 elif string == '+':
2042 if not current_selector:
2043 raise syntax_error('Unexpected "+"', start)
2044 selector_1 = current_selector
2045 selector_2 = _parse_format_selection(tokens, inside_merge=True)
2046 if not selector_2:
2047 raise syntax_error('Expected a selector', start)
2048 current_selector = FormatSelector(MERGE, (selector_1, selector_2), [])
2049 else:
2050 raise syntax_error(f'Operator not recognized: "{string}"', start)
2051 elif type == tokenize.ENDMARKER:
2052 break
2053 if current_selector:
2054 selectors.append(current_selector)
2055 return selectors
2056
2057 def _merge(formats_pair):
2058 format_1, format_2 = formats_pair
2059
2060 formats_info = []
2061 formats_info.extend(format_1.get('requested_formats', (format_1,)))
2062 formats_info.extend(format_2.get('requested_formats', (format_2,)))
2063
2064 if not allow_multiple_streams['video'] or not allow_multiple_streams['audio']:
2065 get_no_more = {'video': False, 'audio': False}
2066 for (i, fmt_info) in enumerate(formats_info):
2067 if fmt_info.get('acodec') == fmt_info.get('vcodec') == 'none':
2068 formats_info.pop(i)
2069 continue
2070 for aud_vid in ['audio', 'video']:
2071 if not allow_multiple_streams[aud_vid] and fmt_info.get(aud_vid[0] + 'codec') != 'none':
2072 if get_no_more[aud_vid]:
2073 formats_info.pop(i)
2074 break
2075 get_no_more[aud_vid] = True
2076
2077 if len(formats_info) == 1:
2078 return formats_info[0]
2079
2080 video_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('vcodec') != 'none']
2081 audio_fmts = [fmt_info for fmt_info in formats_info if fmt_info.get('acodec') != 'none']
2082
2083 the_only_video = video_fmts[0] if len(video_fmts) == 1 else None
2084 the_only_audio = audio_fmts[0] if len(audio_fmts) == 1 else None
2085
2086 output_ext = self.params.get('merge_output_format')
2087 if not output_ext:
2088 if the_only_video:
2089 output_ext = the_only_video['ext']
2090 elif the_only_audio and not video_fmts:
2091 output_ext = the_only_audio['ext']
2092 else:
2093 output_ext = 'mkv'
2094
2095 filtered = lambda *keys: filter(None, (traverse_obj(fmt, *keys) for fmt in formats_info))
2096
2097 new_dict = {
2098 'requested_formats': formats_info,
2099 'format': '+'.join(filtered('format')),
2100 'format_id': '+'.join(filtered('format_id')),
2101 'ext': output_ext,
2102 'protocol': '+'.join(map(determine_protocol, formats_info)),
2103 'language': '+'.join(orderedSet(filtered('language'))) or None,
2104 'format_note': '+'.join(orderedSet(filtered('format_note'))) or None,
2105 'filesize_approx': sum(filtered('filesize', 'filesize_approx')) or None,
2106 'tbr': sum(filtered('tbr', 'vbr', 'abr')),
2107 }
2108
2109 if the_only_video:
2110 new_dict.update({
2111 'width': the_only_video.get('width'),
2112 'height': the_only_video.get('height'),
2113 'resolution': the_only_video.get('resolution') or self.format_resolution(the_only_video),
2114 'fps': the_only_video.get('fps'),
2115 'dynamic_range': the_only_video.get('dynamic_range'),
2116 'vcodec': the_only_video.get('vcodec'),
2117 'vbr': the_only_video.get('vbr'),
2118 'stretched_ratio': the_only_video.get('stretched_ratio'),
2119 })
2120
2121 if the_only_audio:
2122 new_dict.update({
2123 'acodec': the_only_audio.get('acodec'),
2124 'abr': the_only_audio.get('abr'),
2125 'asr': the_only_audio.get('asr'),
2126 })
2127
2128 return new_dict
2129
2130 def _check_formats(formats):
2131 if not check_formats:
2132 yield from formats
2133 return
2134 yield from self._check_formats(formats)
2135
2136 def _build_selector_function(selector):
2137 if isinstance(selector, list): # ,
2138 fs = [_build_selector_function(s) for s in selector]
2139
2140 def selector_function(ctx):
2141 for f in fs:
2142 yield from f(ctx)
2143 return selector_function
2144
2145 elif selector.type == GROUP: # ()
2146 selector_function = _build_selector_function(selector.selector)
2147
2148 elif selector.type == PICKFIRST: # /
2149 fs = [_build_selector_function(s) for s in selector.selector]
2150
2151 def selector_function(ctx):
2152 for f in fs:
2153 picked_formats = list(f(ctx))
2154 if picked_formats:
2155 return picked_formats
2156 return []
2157
2158 elif selector.type == MERGE: # +
2159 selector_1, selector_2 = map(_build_selector_function, selector.selector)
2160
2161 def selector_function(ctx):
2162 for pair in itertools.product(selector_1(ctx), selector_2(ctx)):
2163 yield _merge(pair)
2164
2165 elif selector.type == SINGLE: # atom
2166 format_spec = selector.selector or 'best'
2167
2168 # TODO: Add allvideo, allaudio etc by generalizing the code with best/worst selector
2169 if format_spec == 'all':
2170 def selector_function(ctx):
2171 yield from _check_formats(ctx['formats'][::-1])
2172 elif format_spec == 'mergeall':
2173 def selector_function(ctx):
2174 formats = list(_check_formats(
2175 f for f in ctx['formats'] if f.get('vcodec') != 'none' or f.get('acodec') != 'none'))
2176 if not formats:
2177 return
2178 merged_format = formats[-1]
2179 for f in formats[-2::-1]:
2180 merged_format = _merge((merged_format, f))
2181 yield merged_format
2182
2183 else:
2184 format_fallback, seperate_fallback, format_reverse, format_idx = False, None, True, 1
2185 mobj = re.match(
2186 r'(?P<bw>best|worst|b|w)(?P<type>video|audio|v|a)?(?P<mod>\*)?(?:\.(?P<n>[1-9]\d*))?$',
2187 format_spec)
2188 if mobj is not None:
2189 format_idx = int_or_none(mobj.group('n'), default=1)
2190 format_reverse = mobj.group('bw')[0] == 'b'
2191 format_type = (mobj.group('type') or [None])[0]
2192 not_format_type = {'v': 'a', 'a': 'v'}.get(format_type)
2193 format_modified = mobj.group('mod') is not None
2194
2195 format_fallback = not format_type and not format_modified # for b, w
2196 _filter_f = (
2197 (lambda f: f.get('%scodec' % format_type) != 'none')
2198 if format_type and format_modified # bv*, ba*, wv*, wa*
2199 else (lambda f: f.get('%scodec' % not_format_type) == 'none')
2200 if format_type # bv, ba, wv, wa
2201 else (lambda f: f.get('vcodec') != 'none' and f.get('acodec') != 'none')
2202 if not format_modified # b, w
2203 else lambda f: True) # b*, w*
2204 filter_f = lambda f: _filter_f(f) and (
2205 f.get('vcodec') != 'none' or f.get('acodec') != 'none')
2206 else:
2207 if format_spec in self._format_selection_exts['audio']:
2208 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none'
2209 elif format_spec in self._format_selection_exts['video']:
2210 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') != 'none' and f.get('vcodec') != 'none'
2211 seperate_fallback = lambda f: f.get('ext') == format_spec and f.get('vcodec') != 'none'
2212 elif format_spec in self._format_selection_exts['storyboards']:
2213 filter_f = lambda f: f.get('ext') == format_spec and f.get('acodec') == 'none' and f.get('vcodec') == 'none'
2214 else:
2215 filter_f = lambda f: f.get('format_id') == format_spec # id
2216
2217 def selector_function(ctx):
2218 formats = list(ctx['formats'])
2219 matches = list(filter(filter_f, formats)) if filter_f is not None else formats
2220 if not matches:
2221 if format_fallback and ctx['incomplete_formats']:
2222 # for extractors with incomplete formats (audio only (soundcloud)
2223 # or video only (imgur)) best/worst will fallback to
2224 # best/worst {video,audio}-only format
2225 matches = formats
2226 elif seperate_fallback and not ctx['has_merged_format']:
2227 # for compatibility with youtube-dl when there is no pre-merged format
2228 matches = list(filter(seperate_fallback, formats))
2229 matches = LazyList(_check_formats(matches[::-1 if format_reverse else 1]))
2230 try:
2231 yield matches[format_idx - 1]
2232 except LazyList.IndexError:
2233 return
2234
2235 filters = [self._build_format_filter(f) for f in selector.filters]
2236
2237 def final_selector(ctx):
2238 ctx_copy = dict(ctx)
2239 for _filter in filters:
2240 ctx_copy['formats'] = list(filter(_filter, ctx_copy['formats']))
2241 return selector_function(ctx_copy)
2242 return final_selector
2243
2244 stream = io.BytesIO(format_spec.encode())
2245 try:
2246 tokens = list(_remove_unused_ops(tokenize.tokenize(stream.readline)))
2247 except tokenize.TokenError:
2248 raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
2249
2250 class TokenIterator:
2251 def __init__(self, tokens):
2252 self.tokens = tokens
2253 self.counter = 0
2254
2255 def __iter__(self):
2256 return self
2257
2258 def __next__(self):
2259 if self.counter >= len(self.tokens):
2260 raise StopIteration()
2261 value = self.tokens[self.counter]
2262 self.counter += 1
2263 return value
2264
2265 next = __next__
2266
2267 def restore_last_token(self):
2268 self.counter -= 1
2269
2270 parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
2271 return _build_selector_function(parsed_selector)
2272
2273 def _calc_headers(self, info_dict):
2274 res = merge_headers(self.params['http_headers'], info_dict.get('http_headers') or {})
2275
2276 cookies = self._calc_cookies(info_dict)
2277 if cookies:
2278 res['Cookie'] = cookies
2279
2280 if 'X-Forwarded-For' not in res:
2281 x_forwarded_for_ip = info_dict.get('__x_forwarded_for_ip')
2282 if x_forwarded_for_ip:
2283 res['X-Forwarded-For'] = x_forwarded_for_ip
2284
2285 return res
2286
2287 def _calc_cookies(self, info_dict):
2288 pr = sanitized_Request(info_dict['url'])
2289 self.cookiejar.add_cookie_header(pr)
2290 return pr.get_header('Cookie')
2291
2292 def _sort_thumbnails(self, thumbnails):
2293 thumbnails.sort(key=lambda t: (
2294 t.get('preference') if t.get('preference') is not None else -1,
2295 t.get('width') if t.get('width') is not None else -1,
2296 t.get('height') if t.get('height') is not None else -1,
2297 t.get('id') if t.get('id') is not None else '',
2298 t.get('url')))
2299
2300 def _sanitize_thumbnails(self, info_dict):
2301 thumbnails = info_dict.get('thumbnails')
2302 if thumbnails is None:
2303 thumbnail = info_dict.get('thumbnail')
2304 if thumbnail:
2305 info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}]
2306 if not thumbnails:
2307 return
2308
2309 def check_thumbnails(thumbnails):
2310 for t in thumbnails:
2311 self.to_screen(f'[info] Testing thumbnail {t["id"]}')
2312 try:
2313 self.urlopen(HEADRequest(t['url']))
2314 except network_exceptions as err:
2315 self.to_screen(f'[info] Unable to connect to thumbnail {t["id"]} URL {t["url"]!r} - {err}. Skipping...')
2316 continue
2317 yield t
2318
2319 self._sort_thumbnails(thumbnails)
2320 for i, t in enumerate(thumbnails):
2321 if t.get('id') is None:
2322 t['id'] = '%d' % i
2323 if t.get('width') and t.get('height'):
2324 t['resolution'] = '%dx%d' % (t['width'], t['height'])
2325 t['url'] = sanitize_url(t['url'])
2326
2327 if self.params.get('check_formats') is True:
2328 info_dict['thumbnails'] = LazyList(check_thumbnails(thumbnails[::-1]), reverse=True)
2329 else:
2330 info_dict['thumbnails'] = thumbnails
2331
2332 def _fill_common_fields(self, info_dict, is_video=True):
2333 # TODO: move sanitization here
2334 if is_video:
2335 # playlists are allowed to lack "title"
2336 title = info_dict.get('title', NO_DEFAULT)
2337 if title is NO_DEFAULT:
2338 raise ExtractorError('Missing "title" field in extractor result',
2339 video_id=info_dict['id'], ie=info_dict['extractor'])
2340 info_dict['fulltitle'] = title
2341 if not title:
2342 if title == '':
2343 self.write_debug('Extractor gave empty title. Creating a generic title')
2344 else:
2345 self.report_warning('Extractor failed to obtain "title". Creating a generic title instead')
2346 info_dict['title'] = f'{info_dict["extractor"].replace(":", "-")} video #{info_dict["id"]}'
2347
2348 if info_dict.get('duration') is not None:
2349 info_dict['duration_string'] = formatSeconds(info_dict['duration'])
2350
2351 for ts_key, date_key in (
2352 ('timestamp', 'upload_date'),
2353 ('release_timestamp', 'release_date'),
2354 ('modified_timestamp', 'modified_date'),
2355 ):
2356 if info_dict.get(date_key) is None and info_dict.get(ts_key) is not None:
2357 # Working around out-of-range timestamp values (e.g. negative ones on Windows,
2358 # see http://bugs.python.org/issue1646728)
2359 with contextlib.suppress(ValueError, OverflowError, OSError):
2360 upload_date = datetime.datetime.utcfromtimestamp(info_dict[ts_key])
2361 info_dict[date_key] = upload_date.strftime('%Y%m%d')
2362
2363 live_keys = ('is_live', 'was_live')
2364 live_status = info_dict.get('live_status')
2365 if live_status is None:
2366 for key in live_keys:
2367 if info_dict.get(key) is False:
2368 continue
2369 if info_dict.get(key):
2370 live_status = key
2371 break
2372 if all(info_dict.get(key) is False for key in live_keys):
2373 live_status = 'not_live'
2374 if live_status:
2375 info_dict['live_status'] = live_status
2376 for key in live_keys:
2377 if info_dict.get(key) is None:
2378 info_dict[key] = (live_status == key)
2379
2380 # Auto generate title fields corresponding to the *_number fields when missing
2381 # in order to always have clean titles. This is very common for TV series.
2382 for field in ('chapter', 'season', 'episode'):
2383 if info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
2384 info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
2385
2386 def process_video_result(self, info_dict, download=True):
2387 assert info_dict.get('_type', 'video') == 'video'
2388 self._num_videos += 1
2389
2390 if 'id' not in info_dict:
2391 raise ExtractorError('Missing "id" field in extractor result', ie=info_dict['extractor'])
2392 elif not info_dict.get('id'):
2393 raise ExtractorError('Extractor failed to obtain "id"', ie=info_dict['extractor'])
2394
2395 def report_force_conversion(field, field_not, conversion):
2396 self.report_warning(
2397 '"%s" field is not %s - forcing %s conversion, there is an error in extractor'
2398 % (field, field_not, conversion))
2399
2400 def sanitize_string_field(info, string_field):
2401 field = info.get(string_field)
2402 if field is None or isinstance(field, compat_str):
2403 return
2404 report_force_conversion(string_field, 'a string', 'string')
2405 info[string_field] = compat_str(field)
2406
2407 def sanitize_numeric_fields(info):
2408 for numeric_field in self._NUMERIC_FIELDS:
2409 field = info.get(numeric_field)
2410 if field is None or isinstance(field, (int, float)):
2411 continue
2412 report_force_conversion(numeric_field, 'numeric', 'int')
2413 info[numeric_field] = int_or_none(field)
2414
2415 sanitize_string_field(info_dict, 'id')
2416 sanitize_numeric_fields(info_dict)
2417 if (info_dict.get('duration') or 0) <= 0 and info_dict.pop('duration', None):
2418 self.report_warning('"duration" field is negative, there is an error in extractor')
2419
2420 if 'playlist' not in info_dict:
2421 # It isn't part of a playlist
2422 info_dict['playlist'] = None
2423 info_dict['playlist_index'] = None
2424
2425 self._sanitize_thumbnails(info_dict)
2426
2427 thumbnail = info_dict.get('thumbnail')
2428 thumbnails = info_dict.get('thumbnails')
2429 if thumbnail:
2430 info_dict['thumbnail'] = sanitize_url(thumbnail)
2431 elif thumbnails:
2432 info_dict['thumbnail'] = thumbnails[-1]['url']
2433
2434 if info_dict.get('display_id') is None and 'id' in info_dict:
2435 info_dict['display_id'] = info_dict['id']
2436
2437 self._fill_common_fields(info_dict)
2438
2439 for cc_kind in ('subtitles', 'automatic_captions'):
2440 cc = info_dict.get(cc_kind)
2441 if cc:
2442 for _, subtitle in cc.items():
2443 for subtitle_format in subtitle:
2444 if subtitle_format.get('url'):
2445 subtitle_format['url'] = sanitize_url(subtitle_format['url'])
2446 if subtitle_format.get('ext') is None:
2447 subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower()
2448
2449 automatic_captions = info_dict.get('automatic_captions')
2450 subtitles = info_dict.get('subtitles')
2451
2452 info_dict['requested_subtitles'] = self.process_subtitles(
2453 info_dict['id'], subtitles, automatic_captions)
2454
2455 if info_dict.get('formats') is None:
2456 # There's only one format available
2457 formats = [info_dict]
2458 else:
2459 formats = info_dict['formats']
2460
2461 # or None ensures --clean-infojson removes it
2462 info_dict['_has_drm'] = any(f.get('has_drm') for f in formats) or None
2463 if not self.params.get('allow_unplayable_formats'):
2464 formats = [f for f in formats if not f.get('has_drm')]
2465 if info_dict['_has_drm'] and all(
2466 f.get('acodec') == f.get('vcodec') == 'none' for f in formats):
2467 self.report_warning(
2468 'This video is DRM protected and only images are available for download. '
2469 'Use --list-formats to see them')
2470
2471 get_from_start = not info_dict.get('is_live') or bool(self.params.get('live_from_start'))
2472 if not get_from_start:
2473 info_dict['title'] += ' ' + datetime.datetime.now().strftime('%Y-%m-%d %H:%M')
2474 if info_dict.get('is_live') and formats:
2475 formats = [f for f in formats if bool(f.get('is_from_start')) == get_from_start]
2476 if get_from_start and not formats:
2477 self.raise_no_formats(info_dict, msg=(
2478 '--live-from-start is passed, but there are no formats that can be downloaded from the start. '
2479 'If you want to download from the current time, use --no-live-from-start'))
2480
2481 if not formats:
2482 self.raise_no_formats(info_dict)
2483
2484 def is_wellformed(f):
2485 url = f.get('url')
2486 if not url:
2487 self.report_warning(
2488 '"url" field is missing or empty - skipping format, '
2489 'there is an error in extractor')
2490 return False
2491 if isinstance(url, bytes):
2492 sanitize_string_field(f, 'url')
2493 return True
2494
2495 # Filter out malformed formats for better extraction robustness
2496 formats = list(filter(is_wellformed, formats))
2497
2498 formats_dict = {}
2499
2500 # We check that all the formats have the format and format_id fields
2501 for i, format in enumerate(formats):
2502 sanitize_string_field(format, 'format_id')
2503 sanitize_numeric_fields(format)
2504 format['url'] = sanitize_url(format['url'])
2505 if not format.get('format_id'):
2506 format['format_id'] = compat_str(i)
2507 else:
2508 # Sanitize format_id from characters used in format selector expression
2509 format['format_id'] = re.sub(r'[\s,/+\[\]()]', '_', format['format_id'])
2510 format_id = format['format_id']
2511 if format_id not in formats_dict:
2512 formats_dict[format_id] = []
2513 formats_dict[format_id].append(format)
2514
2515 # Make sure all formats have unique format_id
2516 common_exts = set(itertools.chain(*self._format_selection_exts.values()))
2517 for format_id, ambiguous_formats in formats_dict.items():
2518 ambigious_id = len(ambiguous_formats) > 1
2519 for i, format in enumerate(ambiguous_formats):
2520 if ambigious_id:
2521 format['format_id'] = '%s-%d' % (format_id, i)
2522 if format.get('ext') is None:
2523 format['ext'] = determine_ext(format['url']).lower()
2524 # Ensure there is no conflict between id and ext in format selection
2525 # See https://github.com/yt-dlp/yt-dlp/issues/1282
2526 if format['format_id'] != format['ext'] and format['format_id'] in common_exts:
2527 format['format_id'] = 'f%s' % format['format_id']
2528
2529 for i, format in enumerate(formats):
2530 if format.get('format') is None:
2531 format['format'] = '{id} - {res}{note}'.format(
2532 id=format['format_id'],
2533 res=self.format_resolution(format),
2534 note=format_field(format, 'format_note', ' (%s)'),
2535 )
2536 if format.get('protocol') is None:
2537 format['protocol'] = determine_protocol(format)
2538 if format.get('resolution') is None:
2539 format['resolution'] = self.format_resolution(format, default=None)
2540 if format.get('dynamic_range') is None and format.get('vcodec') != 'none':
2541 format['dynamic_range'] = 'SDR'
2542 if (info_dict.get('duration') and format.get('tbr')
2543 and not format.get('filesize') and not format.get('filesize_approx')):
2544 format['filesize_approx'] = info_dict['duration'] * format['tbr'] * (1024 / 8)
2545
2546 # Add HTTP headers, so that external programs can use them from the
2547 # json output
2548 full_format_info = info_dict.copy()
2549 full_format_info.update(format)
2550 format['http_headers'] = self._calc_headers(full_format_info)
2551 # Remove private housekeeping stuff
2552 if '__x_forwarded_for_ip' in info_dict:
2553 del info_dict['__x_forwarded_for_ip']
2554
2555 if self.params.get('check_formats') is True:
2556 formats = LazyList(self._check_formats(formats[::-1]), reverse=True)
2557
2558 if not formats or formats[0] is not info_dict:
2559 # only set the 'formats' fields if the original info_dict list them
2560 # otherwise we end up with a circular reference, the first (and unique)
2561 # element in the 'formats' field in info_dict is info_dict itself,
2562 # which can't be exported to json
2563 info_dict['formats'] = formats
2564
2565 info_dict, _ = self.pre_process(info_dict)
2566
2567 if self._match_entry(info_dict, incomplete=self._format_fields) is not None:
2568 return info_dict
2569
2570 self.post_extract(info_dict)
2571 info_dict, _ = self.pre_process(info_dict, 'after_filter')
2572
2573 # The pre-processors may have modified the formats
2574 formats = info_dict.get('formats', [info_dict])
2575
2576 list_only = self.params.get('simulate') is None and (
2577 self.params.get('list_thumbnails') or self.params.get('listformats') or self.params.get('listsubtitles'))
2578 interactive_format_selection = not list_only and self.format_selector == '-'
2579 if self.params.get('list_thumbnails'):
2580 self.list_thumbnails(info_dict)
2581 if self.params.get('listsubtitles'):
2582 if 'automatic_captions' in info_dict:
2583 self.list_subtitles(
2584 info_dict['id'], automatic_captions, 'automatic captions')
2585 self.list_subtitles(info_dict['id'], subtitles, 'subtitles')
2586 if self.params.get('listformats') or interactive_format_selection:
2587 self.list_formats(info_dict)
2588 if list_only:
2589 # Without this printing, -F --print-json will not work
2590 self.__forced_printings(info_dict, self.prepare_filename(info_dict), incomplete=True)
2591 return
2592
2593 format_selector = self.format_selector
2594 if format_selector is None:
2595 req_format = self._default_format_spec(info_dict, download=download)
2596 self.write_debug('Default format spec: %s' % req_format)
2597 format_selector = self.build_format_selector(req_format)
2598
2599 while True:
2600 if interactive_format_selection:
2601 req_format = input(
2602 self._format_screen('\nEnter format selector: ', self.Styles.EMPHASIS))
2603 try:
2604 format_selector = self.build_format_selector(req_format)
2605 except SyntaxError as err:
2606 self.report_error(err, tb=False, is_error=False)
2607 continue
2608
2609 formats_to_download = list(format_selector({
2610 'formats': formats,
2611 'has_merged_format': any('none' not in (f.get('acodec'), f.get('vcodec')) for f in formats),
2612 'incomplete_formats': (
2613 # All formats are video-only or
2614 all(f.get('vcodec') != 'none' and f.get('acodec') == 'none' for f in formats)
2615 # all formats are audio-only
2616 or all(f.get('vcodec') == 'none' and f.get('acodec') != 'none' for f in formats)),
2617 }))
2618 if interactive_format_selection and not formats_to_download:
2619 self.report_error('Requested format is not available', tb=False, is_error=False)
2620 continue
2621 break
2622
2623 if not formats_to_download:
2624 if not self.params.get('ignore_no_formats_error'):
2625 raise ExtractorError(
2626 'Requested format is not available. Use --list-formats for a list of available formats',
2627 expected=True, video_id=info_dict['id'], ie=info_dict['extractor'])
2628 self.report_warning('Requested format is not available')
2629 # Process what we can, even without any available formats.
2630 formats_to_download = [{}]
2631
2632 best_format = formats_to_download[-1]
2633 if download:
2634 if best_format:
2635 self.to_screen(
2636 f'[info] {info_dict["id"]}: Downloading {len(formats_to_download)} format(s): '
2637 + ', '.join([f['format_id'] for f in formats_to_download]))
2638 max_downloads_reached = False
2639 for i, fmt in enumerate(formats_to_download):
2640 formats_to_download[i] = new_info = self._copy_infodict(info_dict)
2641 new_info.update(fmt)
2642 try:
2643 self.process_info(new_info)
2644 except MaxDownloadsReached:
2645 max_downloads_reached = True
2646 # Remove copied info
2647 for key, val in tuple(new_info.items()):
2648 if info_dict.get(key) == val:
2649 new_info.pop(key)
2650 if max_downloads_reached:
2651 break
2652
2653 write_archive = {f.get('__write_download_archive', False) for f in formats_to_download}
2654 assert write_archive.issubset({True, False, 'ignore'})
2655 if True in write_archive and False not in write_archive:
2656 self.record_download_archive(info_dict)
2657
2658 info_dict['requested_downloads'] = formats_to_download
2659 info_dict = self.run_all_pps('after_video', info_dict)
2660 if max_downloads_reached:
2661 raise MaxDownloadsReached()
2662
2663 # We update the info dict with the selected best quality format (backwards compatibility)
2664 info_dict.update(best_format)
2665 return info_dict
2666
2667 def process_subtitles(self, video_id, normal_subtitles, automatic_captions):
2668 """Select the requested subtitles and their format"""
2669 available_subs, normal_sub_langs = {}, []
2670 if normal_subtitles and self.params.get('writesubtitles'):
2671 available_subs.update(normal_subtitles)
2672 normal_sub_langs = tuple(normal_subtitles.keys())
2673 if automatic_captions and self.params.get('writeautomaticsub'):
2674 for lang, cap_info in automatic_captions.items():
2675 if lang not in available_subs:
2676 available_subs[lang] = cap_info
2677
2678 if (not self.params.get('writesubtitles') and not
2679 self.params.get('writeautomaticsub') or not
2680 available_subs):
2681 return None
2682
2683 all_sub_langs = tuple(available_subs.keys())
2684 if self.params.get('allsubtitles', False):
2685 requested_langs = all_sub_langs
2686 elif self.params.get('subtitleslangs', False):
2687 # A list is used so that the order of languages will be the same as
2688 # given in subtitleslangs. See https://github.com/yt-dlp/yt-dlp/issues/1041
2689 requested_langs = []
2690 for lang_re in self.params.get('subtitleslangs'):
2691 discard = lang_re[0] == '-'
2692 if discard:
2693 lang_re = lang_re[1:]
2694 if lang_re == 'all':
2695 if discard:
2696 requested_langs = []
2697 else:
2698 requested_langs.extend(all_sub_langs)
2699 continue
2700 current_langs = filter(re.compile(lang_re + '$').match, all_sub_langs)
2701 if discard:
2702 for lang in current_langs:
2703 while lang in requested_langs:
2704 requested_langs.remove(lang)
2705 else:
2706 requested_langs.extend(current_langs)
2707 requested_langs = orderedSet(requested_langs)
2708 elif normal_sub_langs:
2709 requested_langs = ['en'] if 'en' in normal_sub_langs else normal_sub_langs[:1]
2710 else:
2711 requested_langs = ['en'] if 'en' in all_sub_langs else all_sub_langs[:1]
2712 if requested_langs:
2713 self.write_debug('Downloading subtitles: %s' % ', '.join(requested_langs))
2714
2715 formats_query = self.params.get('subtitlesformat', 'best')
2716 formats_preference = formats_query.split('/') if formats_query else []
2717 subs = {}
2718 for lang in requested_langs:
2719 formats = available_subs.get(lang)
2720 if formats is None:
2721 self.report_warning(f'{lang} subtitles not available for {video_id}')
2722 continue
2723 for ext in formats_preference:
2724 if ext == 'best':
2725 f = formats[-1]
2726 break
2727 matches = list(filter(lambda f: f['ext'] == ext, formats))
2728 if matches:
2729 f = matches[-1]
2730 break
2731 else:
2732 f = formats[-1]
2733 self.report_warning(
2734 'No subtitle format found matching "%s" for language %s, '
2735 'using %s' % (formats_query, lang, f['ext']))
2736 subs[lang] = f
2737 return subs
2738
2739 def _forceprint(self, key, info_dict):
2740 if info_dict is None:
2741 return
2742 info_copy = info_dict.copy()
2743 info_copy['formats_table'] = self.render_formats_table(info_dict)
2744 info_copy['thumbnails_table'] = self.render_thumbnails_table(info_dict)
2745 info_copy['subtitles_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('subtitles'))
2746 info_copy['automatic_captions_table'] = self.render_subtitles_table(info_dict.get('id'), info_dict.get('automatic_captions'))
2747
2748 def format_tmpl(tmpl):
2749 mobj = re.match(r'\w+(=?)$', tmpl)
2750 if mobj and mobj.group(1):
2751 return f'{tmpl[:-1]} = %({tmpl[:-1]})r'
2752 elif mobj:
2753 return f'%({tmpl})s'
2754 return tmpl
2755
2756 for tmpl in self.params['forceprint'].get(key, []):
2757 self.to_stdout(self.evaluate_outtmpl(format_tmpl(tmpl), info_copy))
2758
2759 for tmpl, file_tmpl in self.params['print_to_file'].get(key, []):
2760 filename = self.prepare_filename(info_dict, outtmpl=file_tmpl)
2761 tmpl = format_tmpl(tmpl)
2762 self.to_screen(f'[info] Writing {tmpl!r} to: {filename}')
2763 if self._ensure_dir_exists(filename):
2764 with open(filename, 'a', encoding='utf-8') as f:
2765 f.write(self.evaluate_outtmpl(tmpl, info_copy) + '\n')
2766
2767 def __forced_printings(self, info_dict, filename, incomplete):
2768 def print_mandatory(field, actual_field=None):
2769 if actual_field is None:
2770 actual_field = field
2771 if (self.params.get('force%s' % field, False)
2772 and (not incomplete or info_dict.get(actual_field) is not None)):
2773 self.to_stdout(info_dict[actual_field])
2774
2775 def print_optional(field):
2776 if (self.params.get('force%s' % field, False)
2777 and info_dict.get(field) is not None):
2778 self.to_stdout(info_dict[field])
2779
2780 info_dict = info_dict.copy()
2781 if filename is not None:
2782 info_dict['filename'] = filename
2783 if info_dict.get('requested_formats') is not None:
2784 # For RTMP URLs, also include the playpath
2785 info_dict['urls'] = '\n'.join(f['url'] + f.get('play_path', '') for f in info_dict['requested_formats'])
2786 elif info_dict.get('url'):
2787 info_dict['urls'] = info_dict['url'] + info_dict.get('play_path', '')
2788
2789 if (self.params.get('forcejson')
2790 or self.params['forceprint'].get('video')
2791 or self.params['print_to_file'].get('video')):
2792 self.post_extract(info_dict)
2793 self._forceprint('video', info_dict)
2794
2795 print_mandatory('title')
2796 print_mandatory('id')
2797 print_mandatory('url', 'urls')
2798 print_optional('thumbnail')
2799 print_optional('description')
2800 print_optional('filename')
2801 if self.params.get('forceduration') and info_dict.get('duration') is not None:
2802 self.to_stdout(formatSeconds(info_dict['duration']))
2803 print_mandatory('format')
2804
2805 if self.params.get('forcejson'):
2806 self.to_stdout(json.dumps(self.sanitize_info(info_dict)))
2807
2808 def dl(self, name, info, subtitle=False, test=False):
2809 if not info.get('url'):
2810 self.raise_no_formats(info, True)
2811
2812 if test:
2813 verbose = self.params.get('verbose')
2814 params = {
2815 'test': True,
2816 'quiet': self.params.get('quiet') or not verbose,
2817 'verbose': verbose,
2818 'noprogress': not verbose,
2819 'nopart': True,
2820 'skip_unavailable_fragments': False,
2821 'keep_fragments': False,
2822 'overwrites': True,
2823 '_no_ytdl_file': True,
2824 }
2825 else:
2826 params = self.params
2827 fd = get_suitable_downloader(info, params, to_stdout=(name == '-'))(self, params)
2828 if not test:
2829 for ph in self._progress_hooks:
2830 fd.add_progress_hook(ph)
2831 urls = '", "'.join(
2832 (f['url'].split(',')[0] + ',<data>' if f['url'].startswith('data:') else f['url'])
2833 for f in info.get('requested_formats', []) or [info])
2834 self.write_debug(f'Invoking {fd.FD_NAME} downloader on "{urls}"')
2835
2836 # Note: Ideally info should be a deep-copied so that hooks cannot modify it.
2837 # But it may contain objects that are not deep-copyable
2838 new_info = self._copy_infodict(info)
2839 if new_info.get('http_headers') is None:
2840 new_info['http_headers'] = self._calc_headers(new_info)
2841 return fd.download(name, new_info, subtitle)
2842
2843 def existing_file(self, filepaths, *, default_overwrite=True):
2844 existing_files = list(filter(os.path.exists, orderedSet(filepaths)))
2845 if existing_files and not self.params.get('overwrites', default_overwrite):
2846 return existing_files[0]
2847
2848 for file in existing_files:
2849 self.report_file_delete(file)
2850 os.remove(file)
2851 return None
2852
2853 def process_info(self, info_dict):
2854 """Process a single resolved IE result. (Modifies it in-place)"""
2855
2856 assert info_dict.get('_type', 'video') == 'video'
2857 original_infodict = info_dict
2858
2859 if 'format' not in info_dict and 'ext' in info_dict:
2860 info_dict['format'] = info_dict['ext']
2861
2862 # This is mostly just for backward compatibility of process_info
2863 # As a side-effect, this allows for format-specific filters
2864 if self._match_entry(info_dict) is not None:
2865 info_dict['__write_download_archive'] = 'ignore'
2866 return
2867
2868 # Does nothing under normal operation - for backward compatibility of process_info
2869 self.post_extract(info_dict)
2870 self._num_downloads += 1
2871
2872 # info_dict['_filename'] needs to be set for backward compatibility
2873 info_dict['_filename'] = full_filename = self.prepare_filename(info_dict, warn=True)
2874 temp_filename = self.prepare_filename(info_dict, 'temp')
2875 files_to_move = {}
2876
2877 # Forced printings
2878 self.__forced_printings(info_dict, full_filename, incomplete=('format' not in info_dict))
2879
2880 if self.params.get('simulate'):
2881 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
2882 return
2883
2884 if full_filename is None:
2885 return
2886 if not self._ensure_dir_exists(encodeFilename(full_filename)):
2887 return
2888 if not self._ensure_dir_exists(encodeFilename(temp_filename)):
2889 return
2890
2891 if self._write_description('video', info_dict,
2892 self.prepare_filename(info_dict, 'description')) is None:
2893 return
2894
2895 sub_files = self._write_subtitles(info_dict, temp_filename)
2896 if sub_files is None:
2897 return
2898 files_to_move.update(dict(sub_files))
2899
2900 thumb_files = self._write_thumbnails(
2901 'video', info_dict, temp_filename, self.prepare_filename(info_dict, 'thumbnail'))
2902 if thumb_files is None:
2903 return
2904 files_to_move.update(dict(thumb_files))
2905
2906 infofn = self.prepare_filename(info_dict, 'infojson')
2907 _infojson_written = self._write_info_json('video', info_dict, infofn)
2908 if _infojson_written:
2909 info_dict['infojson_filename'] = infofn
2910 # For backward compatibility, even though it was a private field
2911 info_dict['__infojson_filename'] = infofn
2912 elif _infojson_written is None:
2913 return
2914
2915 # Note: Annotations are deprecated
2916 annofn = None
2917 if self.params.get('writeannotations', False):
2918 annofn = self.prepare_filename(info_dict, 'annotation')
2919 if annofn:
2920 if not self._ensure_dir_exists(encodeFilename(annofn)):
2921 return
2922 if not self.params.get('overwrites', True) and os.path.exists(encodeFilename(annofn)):
2923 self.to_screen('[info] Video annotations are already present')
2924 elif not info_dict.get('annotations'):
2925 self.report_warning('There are no annotations to write.')
2926 else:
2927 try:
2928 self.to_screen('[info] Writing video annotations to: ' + annofn)
2929 with open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
2930 annofile.write(info_dict['annotations'])
2931 except (KeyError, TypeError):
2932 self.report_warning('There are no annotations to write.')
2933 except OSError:
2934 self.report_error('Cannot write annotations file: ' + annofn)
2935 return
2936
2937 # Write internet shortcut files
2938 def _write_link_file(link_type):
2939 url = try_get(info_dict['webpage_url'], iri_to_uri)
2940 if not url:
2941 self.report_warning(
2942 f'Cannot write internet shortcut file because the actual URL of "{info_dict["webpage_url"]}" is unknown')
2943 return True
2944 linkfn = replace_extension(self.prepare_filename(info_dict, 'link'), link_type, info_dict.get('ext'))
2945 if not self._ensure_dir_exists(encodeFilename(linkfn)):
2946 return False
2947 if self.params.get('overwrites', True) and os.path.exists(encodeFilename(linkfn)):
2948 self.to_screen(f'[info] Internet shortcut (.{link_type}) is already present')
2949 return True
2950 try:
2951 self.to_screen(f'[info] Writing internet shortcut (.{link_type}) to: {linkfn}')
2952 with open(encodeFilename(to_high_limit_path(linkfn)), 'w', encoding='utf-8',
2953 newline='\r\n' if link_type == 'url' else '\n') as linkfile:
2954 template_vars = {'url': url}
2955 if link_type == 'desktop':
2956 template_vars['filename'] = linkfn[:-(len(link_type) + 1)]
2957 linkfile.write(LINK_TEMPLATES[link_type] % template_vars)
2958 except OSError:
2959 self.report_error(f'Cannot write internet shortcut {linkfn}')
2960 return False
2961 return True
2962
2963 write_links = {
2964 'url': self.params.get('writeurllink'),
2965 'webloc': self.params.get('writewebloclink'),
2966 'desktop': self.params.get('writedesktoplink'),
2967 }
2968 if self.params.get('writelink'):
2969 link_type = ('webloc' if sys.platform == 'darwin'
2970 else 'desktop' if sys.platform.startswith('linux')
2971 else 'url')
2972 write_links[link_type] = True
2973
2974 if any(should_write and not _write_link_file(link_type)
2975 for link_type, should_write in write_links.items()):
2976 return
2977
2978 def replace_info_dict(new_info):
2979 nonlocal info_dict
2980 if new_info == info_dict:
2981 return
2982 info_dict.clear()
2983 info_dict.update(new_info)
2984
2985 try:
2986 new_info, files_to_move = self.pre_process(info_dict, 'before_dl', files_to_move)
2987 replace_info_dict(new_info)
2988 except PostProcessingError as err:
2989 self.report_error('Preprocessing: %s' % str(err))
2990 return
2991
2992 if self.params.get('skip_download'):
2993 info_dict['filepath'] = temp_filename
2994 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
2995 info_dict['__files_to_move'] = files_to_move
2996 replace_info_dict(self.run_pp(MoveFilesAfterDownloadPP(self, False), info_dict))
2997 info_dict['__write_download_archive'] = self.params.get('force_write_download_archive')
2998 else:
2999 # Download
3000 info_dict.setdefault('__postprocessors', [])
3001 try:
3002
3003 def existing_video_file(*filepaths):
3004 ext = info_dict.get('ext')
3005 converted = lambda file: replace_extension(file, self.params.get('final_ext') or ext, ext)
3006 file = self.existing_file(itertools.chain(*zip(map(converted, filepaths), filepaths)),
3007 default_overwrite=False)
3008 if file:
3009 info_dict['ext'] = os.path.splitext(file)[1][1:]
3010 return file
3011
3012 success = True
3013 if info_dict.get('requested_formats') is not None:
3014
3015 def compatible_formats(formats):
3016 # TODO: some formats actually allow this (mkv, webm, ogg, mp4), but not all of them.
3017 video_formats = [format for format in formats if format.get('vcodec') != 'none']
3018 audio_formats = [format for format in formats if format.get('acodec') != 'none']
3019 if len(video_formats) > 2 or len(audio_formats) > 2:
3020 return False
3021
3022 # Check extension
3023 exts = {format.get('ext') for format in formats}
3024 COMPATIBLE_EXTS = (
3025 {'mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'ismv', 'isma'},
3026 {'webm'},
3027 )
3028 for ext_sets in COMPATIBLE_EXTS:
3029 if ext_sets.issuperset(exts):
3030 return True
3031 # TODO: Check acodec/vcodec
3032 return False
3033
3034 requested_formats = info_dict['requested_formats']
3035 old_ext = info_dict['ext']
3036 if self.params.get('merge_output_format') is None:
3037 if not compatible_formats(requested_formats):
3038 info_dict['ext'] = 'mkv'
3039 self.report_warning(
3040 'Requested formats are incompatible for merge and will be merged into mkv')
3041 if (info_dict['ext'] == 'webm'
3042 and info_dict.get('thumbnails')
3043 # check with type instead of pp_key, __name__, or isinstance
3044 # since we dont want any custom PPs to trigger this
3045 and any(type(pp) == EmbedThumbnailPP for pp in self._pps['post_process'])):
3046 info_dict['ext'] = 'mkv'
3047 self.report_warning(
3048 'webm doesn\'t support embedding a thumbnail, mkv will be used')
3049 new_ext = info_dict['ext']
3050
3051 def correct_ext(filename, ext=new_ext):
3052 if filename == '-':
3053 return filename
3054 filename_real_ext = os.path.splitext(filename)[1][1:]
3055 filename_wo_ext = (
3056 os.path.splitext(filename)[0]
3057 if filename_real_ext in (old_ext, new_ext)
3058 else filename)
3059 return f'{filename_wo_ext}.{ext}'
3060
3061 # Ensure filename always has a correct extension for successful merge
3062 full_filename = correct_ext(full_filename)
3063 temp_filename = correct_ext(temp_filename)
3064 dl_filename = existing_video_file(full_filename, temp_filename)
3065 info_dict['__real_download'] = False
3066
3067 downloaded = []
3068 merger = FFmpegMergerPP(self)
3069
3070 fd = get_suitable_downloader(info_dict, self.params, to_stdout=temp_filename == '-')
3071 if dl_filename is not None:
3072 self.report_file_already_downloaded(dl_filename)
3073 elif fd:
3074 for f in requested_formats if fd != FFmpegFD else []:
3075 f['filepath'] = fname = prepend_extension(
3076 correct_ext(temp_filename, info_dict['ext']),
3077 'f%s' % f['format_id'], info_dict['ext'])
3078 downloaded.append(fname)
3079 info_dict['url'] = '\n'.join(f['url'] for f in requested_formats)
3080 success, real_download = self.dl(temp_filename, info_dict)
3081 info_dict['__real_download'] = real_download
3082 else:
3083 if self.params.get('allow_unplayable_formats'):
3084 self.report_warning(
3085 'You have requested merging of multiple formats '
3086 'while also allowing unplayable formats to be downloaded. '
3087 'The formats won\'t be merged to prevent data corruption.')
3088 elif not merger.available:
3089 msg = 'You have requested merging of multiple formats but ffmpeg is not installed'
3090 if not self.params.get('ignoreerrors'):
3091 self.report_error(f'{msg}. Aborting due to --abort-on-error')
3092 return
3093 self.report_warning(f'{msg}. The formats won\'t be merged')
3094
3095 if temp_filename == '-':
3096 reason = ('using a downloader other than ffmpeg' if FFmpegFD.can_merge_formats(info_dict, self.params)
3097 else 'but the formats are incompatible for simultaneous download' if merger.available
3098 else 'but ffmpeg is not installed')
3099 self.report_warning(
3100 f'You have requested downloading multiple formats to stdout {reason}. '
3101 'The formats will be streamed one after the other')
3102 fname = temp_filename
3103 for f in requested_formats:
3104 new_info = dict(info_dict)
3105 del new_info['requested_formats']
3106 new_info.update(f)
3107 if temp_filename != '-':
3108 fname = prepend_extension(
3109 correct_ext(temp_filename, new_info['ext']),
3110 'f%s' % f['format_id'], new_info['ext'])
3111 if not self._ensure_dir_exists(fname):
3112 return
3113 f['filepath'] = fname
3114 downloaded.append(fname)
3115 partial_success, real_download = self.dl(fname, new_info)
3116 info_dict['__real_download'] = info_dict['__real_download'] or real_download
3117 success = success and partial_success
3118
3119 if downloaded and merger.available and not self.params.get('allow_unplayable_formats'):
3120 info_dict['__postprocessors'].append(merger)
3121 info_dict['__files_to_merge'] = downloaded
3122 # Even if there were no downloads, it is being merged only now
3123 info_dict['__real_download'] = True
3124 else:
3125 for file in downloaded:
3126 files_to_move[file] = None
3127 else:
3128 # Just a single file
3129 dl_filename = existing_video_file(full_filename, temp_filename)
3130 if dl_filename is None or dl_filename == temp_filename:
3131 # dl_filename == temp_filename could mean that the file was partially downloaded with --no-part.
3132 # So we should try to resume the download
3133 success, real_download = self.dl(temp_filename, info_dict)
3134 info_dict['__real_download'] = real_download
3135 else:
3136 self.report_file_already_downloaded(dl_filename)
3137
3138 dl_filename = dl_filename or temp_filename
3139 info_dict['__finaldir'] = os.path.dirname(os.path.abspath(encodeFilename(full_filename)))
3140
3141 except network_exceptions as err:
3142 self.report_error('unable to download video data: %s' % error_to_compat_str(err))
3143 return
3144 except OSError as err:
3145 raise UnavailableVideoError(err)
3146 except (ContentTooShortError, ) as err:
3147 self.report_error(f'content too short (expected {err.expected} bytes and served {err.downloaded})')
3148 return
3149
3150 if success and full_filename != '-':
3151
3152 def fixup():
3153 do_fixup = True
3154 fixup_policy = self.params.get('fixup')
3155 vid = info_dict['id']
3156
3157 if fixup_policy in ('ignore', 'never'):
3158 return
3159 elif fixup_policy == 'warn':
3160 do_fixup = 'warn'
3161 elif fixup_policy != 'force':
3162 assert fixup_policy in ('detect_or_warn', None)
3163 if not info_dict.get('__real_download'):
3164 do_fixup = False
3165
3166 def ffmpeg_fixup(cndn, msg, cls):
3167 if not (do_fixup and cndn):
3168 return
3169 elif do_fixup == 'warn':
3170 self.report_warning(f'{vid}: {msg}')
3171 return
3172 pp = cls(self)
3173 if pp.available:
3174 info_dict['__postprocessors'].append(pp)
3175 else:
3176 self.report_warning(f'{vid}: {msg}. Install ffmpeg to fix this automatically')
3177
3178 stretched_ratio = info_dict.get('stretched_ratio')
3179 ffmpeg_fixup(
3180 stretched_ratio not in (1, None),
3181 f'Non-uniform pixel ratio {stretched_ratio}',
3182 FFmpegFixupStretchedPP)
3183
3184 ffmpeg_fixup(
3185 (info_dict.get('requested_formats') is None
3186 and info_dict.get('container') == 'm4a_dash'
3187 and info_dict.get('ext') == 'm4a'),
3188 'writing DASH m4a. Only some players support this container',
3189 FFmpegFixupM4aPP)
3190
3191 downloader = get_suitable_downloader(info_dict, self.params) if 'protocol' in info_dict else None
3192 downloader = downloader.FD_NAME if downloader else None
3193
3194 if info_dict.get('requested_formats') is None: # Not necessary if doing merger
3195 ffmpeg_fixup(downloader == 'hlsnative' and not self.params.get('hls_use_mpegts')
3196 or info_dict.get('is_live') and self.params.get('hls_use_mpegts') is None,
3197 'Possible MPEG-TS in MP4 container or malformed AAC timestamps',
3198 FFmpegFixupM3u8PP)
3199 ffmpeg_fixup(info_dict.get('is_live') and downloader == 'DashSegmentsFD',
3200 'Possible duplicate MOOV atoms', FFmpegFixupDuplicateMoovPP)
3201
3202 ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed timestamps detected', FFmpegFixupTimestampPP)
3203 ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed duration detected', FFmpegFixupDurationPP)
3204
3205 fixup()
3206 try:
3207 replace_info_dict(self.post_process(dl_filename, info_dict, files_to_move))
3208 except PostProcessingError as err:
3209 self.report_error('Postprocessing: %s' % str(err))
3210 return
3211 try:
3212 for ph in self._post_hooks:
3213 ph(info_dict['filepath'])
3214 except Exception as err:
3215 self.report_error('post hooks: %s' % str(err))
3216 return
3217 info_dict['__write_download_archive'] = True
3218
3219 if self.params.get('force_write_download_archive'):
3220 info_dict['__write_download_archive'] = True
3221
3222 # Make sure the info_dict was modified in-place
3223 assert info_dict is original_infodict
3224
3225 max_downloads = self.params.get('max_downloads')
3226 if max_downloads is not None and self._num_downloads >= int(max_downloads):
3227 raise MaxDownloadsReached()
3228
3229 def __download_wrapper(self, func):
3230 @functools.wraps(func)
3231 def wrapper(*args, **kwargs):
3232 try:
3233 res = func(*args, **kwargs)
3234 except UnavailableVideoError as e:
3235 self.report_error(e)
3236 except DownloadCancelled as e:
3237 self.to_screen(f'[info] {e}')
3238 if not self.params.get('break_per_url'):
3239 raise
3240 else:
3241 if self.params.get('dump_single_json', False):
3242 self.post_extract(res)
3243 self.to_stdout(json.dumps(self.sanitize_info(res)))
3244 return wrapper
3245
3246 def download(self, url_list):
3247 """Download a given list of URLs."""
3248 url_list = variadic(url_list) # Passing a single URL is a common mistake
3249 outtmpl = self.outtmpl_dict['default']
3250 if (len(url_list) > 1
3251 and outtmpl != '-'
3252 and '%' not in outtmpl
3253 and self.params.get('max_downloads') != 1):
3254 raise SameFileError(outtmpl)
3255
3256 for url in url_list:
3257 self.__download_wrapper(self.extract_info)(
3258 url, force_generic_extractor=self.params.get('force_generic_extractor', False))
3259
3260 return self._download_retcode
3261
3262 def download_with_info_file(self, info_filename):
3263 with contextlib.closing(fileinput.FileInput(
3264 [info_filename], mode='r',
3265 openhook=fileinput.hook_encoded('utf-8'))) as f:
3266 # FileInput doesn't have a read method, we can't call json.load
3267 info = self.sanitize_info(json.loads('\n'.join(f)), self.params.get('clean_infojson', True))
3268 try:
3269 self.__download_wrapper(self.process_ie_result)(info, download=True)
3270 except (DownloadError, EntryNotInPlaylist, ReExtractInfo) as e:
3271 if not isinstance(e, EntryNotInPlaylist):
3272 self.to_stderr('\r')
3273 webpage_url = info.get('webpage_url')
3274 if webpage_url is not None:
3275 self.report_warning(f'The info failed to download: {e}; trying with URL {webpage_url}')
3276 return self.download([webpage_url])
3277 else:
3278 raise
3279 return self._download_retcode
3280
3281 @staticmethod
3282 def sanitize_info(info_dict, remove_private_keys=False):
3283 ''' Sanitize the infodict for converting to json '''
3284 if info_dict is None:
3285 return info_dict
3286 info_dict.setdefault('epoch', int(time.time()))
3287 info_dict.setdefault('_type', 'video')
3288
3289 if remove_private_keys:
3290 reject = lambda k, v: v is None or k.startswith('__') or k in {
3291 'requested_downloads', 'requested_formats', 'requested_subtitles', 'requested_entries',
3292 'entries', 'filepath', '_filename', 'infojson_filename', 'original_url', 'playlist_autonumber',
3293 }
3294 else:
3295 reject = lambda k, v: False
3296
3297 def filter_fn(obj):
3298 if isinstance(obj, dict):
3299 return {k: filter_fn(v) for k, v in obj.items() if not reject(k, v)}
3300 elif isinstance(obj, (list, tuple, set, LazyList)):
3301 return list(map(filter_fn, obj))
3302 elif obj is None or isinstance(obj, (str, int, float, bool)):
3303 return obj
3304 else:
3305 return repr(obj)
3306
3307 return filter_fn(info_dict)
3308
3309 @staticmethod
3310 def filter_requested_info(info_dict, actually_filter=True):
3311 ''' Alias of sanitize_info for backward compatibility '''
3312 return YoutubeDL.sanitize_info(info_dict, actually_filter)
3313
3314 def _delete_downloaded_files(self, *files_to_delete, info={}, msg=None):
3315 for filename in set(filter(None, files_to_delete)):
3316 if msg:
3317 self.to_screen(msg % filename)
3318 try:
3319 os.remove(filename)
3320 except OSError:
3321 self.report_warning(f'Unable to delete file {filename}')
3322 if filename in info.get('__files_to_move', []): # NB: Delete even if None
3323 del info['__files_to_move'][filename]
3324
3325 @staticmethod
3326 def post_extract(info_dict):
3327 def actual_post_extract(info_dict):
3328 if info_dict.get('_type') in ('playlist', 'multi_video'):
3329 for video_dict in info_dict.get('entries', {}):
3330 actual_post_extract(video_dict or {})
3331 return
3332
3333 post_extractor = info_dict.pop('__post_extractor', None) or (lambda: {})
3334 info_dict.update(post_extractor())
3335
3336 actual_post_extract(info_dict or {})
3337
3338 def run_pp(self, pp, infodict):
3339 files_to_delete = []
3340 if '__files_to_move' not in infodict:
3341 infodict['__files_to_move'] = {}
3342 try:
3343 files_to_delete, infodict = pp.run(infodict)
3344 except PostProcessingError as e:
3345 # Must be True and not 'only_download'
3346 if self.params.get('ignoreerrors') is True:
3347 self.report_error(e)
3348 return infodict
3349 raise
3350
3351 if not files_to_delete:
3352 return infodict
3353 if self.params.get('keepvideo', False):
3354 for f in files_to_delete:
3355 infodict['__files_to_move'].setdefault(f, '')
3356 else:
3357 self._delete_downloaded_files(
3358 *files_to_delete, info=infodict, msg='Deleting original file %s (pass -k to keep)')
3359 return infodict
3360
3361 def run_all_pps(self, key, info, *, additional_pps=None):
3362 self._forceprint(key, info)
3363 for pp in (additional_pps or []) + self._pps[key]:
3364 info = self.run_pp(pp, info)
3365 return info
3366
3367 def pre_process(self, ie_info, key='pre_process', files_to_move=None):
3368 info = dict(ie_info)
3369 info['__files_to_move'] = files_to_move or {}
3370 info = self.run_all_pps(key, info)
3371 return info, info.pop('__files_to_move', None)
3372
3373 def post_process(self, filename, info, files_to_move=None):
3374 """Run all the postprocessors on the given file."""
3375 info['filepath'] = filename
3376 info['__files_to_move'] = files_to_move or {}
3377 info = self.run_all_pps('post_process', info, additional_pps=info.get('__postprocessors'))
3378 info = self.run_pp(MoveFilesAfterDownloadPP(self), info)
3379 del info['__files_to_move']
3380 return self.run_all_pps('after_move', info)
3381
3382 def _make_archive_id(self, info_dict):
3383 video_id = info_dict.get('id')
3384 if not video_id:
3385 return
3386 # Future-proof against any change in case
3387 # and backwards compatibility with prior versions
3388 extractor = info_dict.get('extractor_key') or info_dict.get('ie_key') # key in a playlist
3389 if extractor is None:
3390 url = str_or_none(info_dict.get('url'))
3391 if not url:
3392 return
3393 # Try to find matching extractor for the URL and take its ie_key
3394 for ie_key, ie in self._ies.items():
3395 if ie.suitable(url):
3396 extractor = ie_key
3397 break
3398 else:
3399 return
3400 return f'{extractor.lower()} {video_id}'
3401
3402 def in_download_archive(self, info_dict):
3403 fn = self.params.get('download_archive')
3404 if fn is None:
3405 return False
3406
3407 vid_id = self._make_archive_id(info_dict)
3408 if not vid_id:
3409 return False # Incomplete video information
3410
3411 return vid_id in self.archive
3412
3413 def record_download_archive(self, info_dict):
3414 fn = self.params.get('download_archive')
3415 if fn is None:
3416 return
3417 vid_id = self._make_archive_id(info_dict)
3418 assert vid_id
3419 self.write_debug(f'Adding to archive: {vid_id}')
3420 with locked_file(fn, 'a', encoding='utf-8') as archive_file:
3421 archive_file.write(vid_id + '\n')
3422 self.archive.add(vid_id)
3423
3424 @staticmethod
3425 def format_resolution(format, default='unknown'):
3426 if format.get('vcodec') == 'none' and format.get('acodec') != 'none':
3427 return 'audio only'
3428 if format.get('resolution') is not None:
3429 return format['resolution']
3430 if format.get('width') and format.get('height'):
3431 return '%dx%d' % (format['width'], format['height'])
3432 elif format.get('height'):
3433 return '%sp' % format['height']
3434 elif format.get('width'):
3435 return '%dx?' % format['width']
3436 return default
3437
3438 def _list_format_headers(self, *headers):
3439 if self.params.get('listformats_table', True) is not False:
3440 return [self._format_out(header, self.Styles.HEADERS) for header in headers]
3441 return headers
3442
3443 def _format_note(self, fdict):
3444 res = ''
3445 if fdict.get('ext') in ['f4f', 'f4m']:
3446 res += '(unsupported)'
3447 if fdict.get('language'):
3448 if res:
3449 res += ' '
3450 res += '[%s]' % fdict['language']
3451 if fdict.get('format_note') is not None:
3452 if res:
3453 res += ' '
3454 res += fdict['format_note']
3455 if fdict.get('tbr') is not None:
3456 if res:
3457 res += ', '
3458 res += '%4dk' % fdict['tbr']
3459 if fdict.get('container') is not None:
3460 if res:
3461 res += ', '
3462 res += '%s container' % fdict['container']
3463 if (fdict.get('vcodec') is not None
3464 and fdict.get('vcodec') != 'none'):
3465 if res:
3466 res += ', '
3467 res += fdict['vcodec']
3468 if fdict.get('vbr') is not None:
3469 res += '@'
3470 elif fdict.get('vbr') is not None and fdict.get('abr') is not None:
3471 res += 'video@'
3472 if fdict.get('vbr') is not None:
3473 res += '%4dk' % fdict['vbr']
3474 if fdict.get('fps') is not None:
3475 if res:
3476 res += ', '
3477 res += '%sfps' % fdict['fps']
3478 if fdict.get('acodec') is not None:
3479 if res:
3480 res += ', '
3481 if fdict['acodec'] == 'none':
3482 res += 'video only'
3483 else:
3484 res += '%-5s' % fdict['acodec']
3485 elif fdict.get('abr') is not None:
3486 if res:
3487 res += ', '
3488 res += 'audio'
3489 if fdict.get('abr') is not None:
3490 res += '@%3dk' % fdict['abr']
3491 if fdict.get('asr') is not None:
3492 res += ' (%5dHz)' % fdict['asr']
3493 if fdict.get('filesize') is not None:
3494 if res:
3495 res += ', '
3496 res += format_bytes(fdict['filesize'])
3497 elif fdict.get('filesize_approx') is not None:
3498 if res:
3499 res += ', '
3500 res += '~' + format_bytes(fdict['filesize_approx'])
3501 return res
3502
3503 def render_formats_table(self, info_dict):
3504 if not info_dict.get('formats') and not info_dict.get('url'):
3505 return None
3506
3507 formats = info_dict.get('formats', [info_dict])
3508 if not self.params.get('listformats_table', True) is not False:
3509 table = [
3510 [
3511 format_field(f, 'format_id'),
3512 format_field(f, 'ext'),
3513 self.format_resolution(f),
3514 self._format_note(f)
3515 ] for f in formats if f.get('preference') is None or f['preference'] >= -1000]
3516 return render_table(['format code', 'extension', 'resolution', 'note'], table, extra_gap=1)
3517
3518 delim = self._format_out('\u2502', self.Styles.DELIM, '|', test_encoding=True)
3519 table = [
3520 [
3521 self._format_out(format_field(f, 'format_id'), self.Styles.ID),
3522 format_field(f, 'ext'),
3523 format_field(f, func=self.format_resolution, ignore=('audio only', 'images')),
3524 format_field(f, 'fps', '\t%d'),
3525 format_field(f, 'dynamic_range', '%s', ignore=(None, 'SDR')).replace('HDR', ''),
3526 delim,
3527 format_field(f, 'filesize', ' \t%s', func=format_bytes) + format_field(f, 'filesize_approx', '~\t%s', func=format_bytes),
3528 format_field(f, 'tbr', '\t%dk'),
3529 shorten_protocol_name(f.get('protocol', '')),
3530 delim,
3531 format_field(f, 'vcodec', default='unknown').replace(
3532 'none', 'images' if f.get('acodec') == 'none'
3533 else self._format_out('audio only', self.Styles.SUPPRESS)),
3534 format_field(f, 'vbr', '\t%dk'),
3535 format_field(f, 'acodec', default='unknown').replace(
3536 'none', '' if f.get('vcodec') == 'none'
3537 else self._format_out('video only', self.Styles.SUPPRESS)),
3538 format_field(f, 'abr', '\t%dk'),
3539 format_field(f, 'asr', '\t%dHz'),
3540 join_nonempty(
3541 self._format_out('UNSUPPORTED', 'light red') if f.get('ext') in ('f4f', 'f4m') else None,
3542 format_field(f, 'language', '[%s]'),
3543 join_nonempty(format_field(f, 'format_note'),
3544 format_field(f, 'container', ignore=(None, f.get('ext'))),
3545 delim=', '),
3546 delim=' '),
3547 ] for f in formats if f.get('preference') is None or f['preference'] >= -1000]
3548 header_line = self._list_format_headers(
3549 'ID', 'EXT', 'RESOLUTION', '\tFPS', 'HDR', delim, '\tFILESIZE', '\tTBR', 'PROTO',
3550 delim, 'VCODEC', '\tVBR', 'ACODEC', '\tABR', '\tASR', 'MORE INFO')
3551
3552 return render_table(
3553 header_line, table, hide_empty=True,
3554 delim=self._format_out('\u2500', self.Styles.DELIM, '-', test_encoding=True))
3555
3556 def render_thumbnails_table(self, info_dict):
3557 thumbnails = list(info_dict.get('thumbnails') or [])
3558 if not thumbnails:
3559 return None
3560 return render_table(
3561 self._list_format_headers('ID', 'Width', 'Height', 'URL'),
3562 [[t.get('id'), t.get('width', 'unknown'), t.get('height', 'unknown'), t['url']] for t in thumbnails])
3563
3564 def render_subtitles_table(self, video_id, subtitles):
3565 def _row(lang, formats):
3566 exts, names = zip(*((f['ext'], f.get('name') or 'unknown') for f in reversed(formats)))
3567 if len(set(names)) == 1:
3568 names = [] if names[0] == 'unknown' else names[:1]
3569 return [lang, ', '.join(names), ', '.join(exts)]
3570
3571 if not subtitles:
3572 return None
3573 return render_table(
3574 self._list_format_headers('Language', 'Name', 'Formats'),
3575 [_row(lang, formats) for lang, formats in subtitles.items()],
3576 hide_empty=True)
3577
3578 def __list_table(self, video_id, name, func, *args):
3579 table = func(*args)
3580 if not table:
3581 self.to_screen(f'{video_id} has no {name}')
3582 return
3583 self.to_screen(f'[info] Available {name} for {video_id}:')
3584 self.to_stdout(table)
3585
3586 def list_formats(self, info_dict):
3587 self.__list_table(info_dict['id'], 'formats', self.render_formats_table, info_dict)
3588
3589 def list_thumbnails(self, info_dict):
3590 self.__list_table(info_dict['id'], 'thumbnails', self.render_thumbnails_table, info_dict)
3591
3592 def list_subtitles(self, video_id, subtitles, name='subtitles'):
3593 self.__list_table(video_id, name, self.render_subtitles_table, video_id, subtitles)
3594
3595 def urlopen(self, req):
3596 """ Start an HTTP download """
3597 if isinstance(req, str):
3598 req = sanitized_Request(req)
3599 return self._opener.open(req, timeout=self._socket_timeout)
3600
3601 def print_debug_header(self):
3602 if not self.params.get('verbose'):
3603 return
3604
3605 def get_encoding(stream):
3606 ret = str(getattr(stream, 'encoding', 'missing (%s)' % type(stream).__name__))
3607 if not supports_terminal_sequences(stream):
3608 from .utils import WINDOWS_VT_MODE # Must be imported locally
3609 ret += ' (No VT)' if WINDOWS_VT_MODE is False else ' (No ANSI)'
3610 return ret
3611
3612 encoding_str = 'Encodings: locale %s, fs %s, pref %s, %s' % (
3613 locale.getpreferredencoding(),
3614 sys.getfilesystemencoding(),
3615 self.get_encoding(),
3616 ', '.join(
3617 f'{key} {get_encoding(stream)}' for key, stream in self._out_files
3618 if stream is not None and key != 'console')
3619 )
3620
3621 logger = self.params.get('logger')
3622 if logger:
3623 write_debug = lambda msg: logger.debug(f'[debug] {msg}')
3624 write_debug(encoding_str)
3625 else:
3626 write_string(f'[debug] {encoding_str}\n', encoding=None)
3627 write_debug = lambda msg: self._write_string(f'[debug] {msg}\n')
3628
3629 source = detect_variant()
3630 write_debug(join_nonempty(
3631 'yt-dlp version', __version__,
3632 f'[{RELEASE_GIT_HEAD}]' if RELEASE_GIT_HEAD else '',
3633 '' if source == 'unknown' else f'({source})',
3634 delim=' '))
3635 if not _LAZY_LOADER:
3636 if os.environ.get('YTDLP_NO_LAZY_EXTRACTORS'):
3637 write_debug('Lazy loading extractors is forcibly disabled')
3638 else:
3639 write_debug('Lazy loading extractors is disabled')
3640 if plugin_extractors or plugin_postprocessors:
3641 write_debug('Plugins: %s' % [
3642 '%s%s' % (klass.__name__, '' if klass.__name__ == name else f' as {name}')
3643 for name, klass in itertools.chain(plugin_extractors.items(), plugin_postprocessors.items())])
3644 if self.params.get('compat_opts'):
3645 write_debug('Compatibility options: %s' % ', '.join(self.params.get('compat_opts')))
3646
3647 if source == 'source':
3648 try:
3649 sp = Popen(
3650 ['git', 'rev-parse', '--short', 'HEAD'],
3651 stdout=subprocess.PIPE, stderr=subprocess.PIPE,
3652 cwd=os.path.dirname(os.path.abspath(__file__)))
3653 out, err = sp.communicate_or_kill()
3654 out = out.decode().strip()
3655 if re.match('[0-9a-f]+', out):
3656 write_debug('Git HEAD: %s' % out)
3657 except Exception:
3658 with contextlib.suppress(Exception):
3659 sys.exc_clear()
3660
3661 def python_implementation():
3662 impl_name = platform.python_implementation()
3663 if impl_name == 'PyPy' and hasattr(sys, 'pypy_version_info'):
3664 return impl_name + ' version %d.%d.%d' % sys.pypy_version_info[:3]
3665 return impl_name
3666
3667 write_debug('Python version %s (%s %s) - %s' % (
3668 platform.python_version(),
3669 python_implementation(),
3670 platform.architecture()[0],
3671 platform_name()))
3672
3673 exe_versions, ffmpeg_features = FFmpegPostProcessor.get_versions_and_features(self)
3674 ffmpeg_features = {key for key, val in ffmpeg_features.items() if val}
3675 if ffmpeg_features:
3676 exe_versions['ffmpeg'] += ' (%s)' % ','.join(sorted(ffmpeg_features))
3677
3678 exe_versions['rtmpdump'] = rtmpdump_version()
3679 exe_versions['phantomjs'] = PhantomJSwrapper._version()
3680 exe_str = ', '.join(
3681 f'{exe} {v}' for exe, v in sorted(exe_versions.items()) if v
3682 ) or 'none'
3683 write_debug('exe versions: %s' % exe_str)
3684
3685 from .compat.compat_utils import get_package_info
3686 from .dependencies import available_dependencies
3687
3688 write_debug('Optional libraries: %s' % (', '.join(sorted({
3689 join_nonempty(*get_package_info(m)) for m in available_dependencies.values()
3690 })) or 'none'))
3691
3692 self._setup_opener()
3693 proxy_map = {}
3694 for handler in self._opener.handlers:
3695 if hasattr(handler, 'proxies'):
3696 proxy_map.update(handler.proxies)
3697 write_debug(f'Proxy map: {proxy_map}')
3698
3699 # Not implemented
3700 if False and self.params.get('call_home'):
3701 ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode()
3702 write_debug('Public IP address: %s' % ipaddr)
3703 latest_version = self.urlopen(
3704 'https://yt-dl.org/latest/version').read().decode()
3705 if version_tuple(latest_version) > version_tuple(__version__):
3706 self.report_warning(
3707 'You are using an outdated version (newest version: %s)! '
3708 'See https://yt-dl.org/update if you need help updating.' %
3709 latest_version)
3710
3711 def _setup_opener(self):
3712 if hasattr(self, '_opener'):
3713 return
3714 timeout_val = self.params.get('socket_timeout')
3715 self._socket_timeout = 20 if timeout_val is None else float(timeout_val)
3716
3717 opts_cookiesfrombrowser = self.params.get('cookiesfrombrowser')
3718 opts_cookiefile = self.params.get('cookiefile')
3719 opts_proxy = self.params.get('proxy')
3720
3721 self.cookiejar = load_cookies(opts_cookiefile, opts_cookiesfrombrowser, self)
3722
3723 cookie_processor = YoutubeDLCookieProcessor(self.cookiejar)
3724 if opts_proxy is not None:
3725 if opts_proxy == '':
3726 proxies = {}
3727 else:
3728 proxies = {'http': opts_proxy, 'https': opts_proxy}
3729 else:
3730 proxies = compat_urllib_request.getproxies()
3731 # Set HTTPS proxy to HTTP one if given (https://github.com/ytdl-org/youtube-dl/issues/805)
3732 if 'http' in proxies and 'https' not in proxies:
3733 proxies['https'] = proxies['http']
3734 proxy_handler = PerRequestProxyHandler(proxies)
3735
3736 debuglevel = 1 if self.params.get('debug_printtraffic') else 0
3737 https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
3738 ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
3739 redirect_handler = YoutubeDLRedirectHandler()
3740 data_handler = urllib.request.DataHandler()
3741
3742 # When passing our own FileHandler instance, build_opener won't add the
3743 # default FileHandler and allows us to disable the file protocol, which
3744 # can be used for malicious purposes (see
3745 # https://github.com/ytdl-org/youtube-dl/issues/8227)
3746 file_handler = compat_urllib_request.FileHandler()
3747
3748 def file_open(*args, **kwargs):
3749 raise compat_urllib_error.URLError('file:// scheme is explicitly disabled in yt-dlp for security reasons')
3750 file_handler.file_open = file_open
3751
3752 opener = compat_urllib_request.build_opener(
3753 proxy_handler, https_handler, cookie_processor, ydlh, redirect_handler, data_handler, file_handler)
3754
3755 # Delete the default user-agent header, which would otherwise apply in
3756 # cases where our custom HTTP handler doesn't come into play
3757 # (See https://github.com/ytdl-org/youtube-dl/issues/1309 for details)
3758 opener.addheaders = []
3759 self._opener = opener
3760
3761 def encode(self, s):
3762 if isinstance(s, bytes):
3763 return s # Already encoded
3764
3765 try:
3766 return s.encode(self.get_encoding())
3767 except UnicodeEncodeError as err:
3768 err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.'
3769 raise
3770
3771 def get_encoding(self):
3772 encoding = self.params.get('encoding')
3773 if encoding is None:
3774 encoding = preferredencoding()
3775 return encoding
3776
3777 def _write_info_json(self, label, ie_result, infofn, overwrite=None):
3778 ''' Write infojson and returns True = written, 'exists' = Already exists, False = skip, None = error '''
3779 if overwrite is None:
3780 overwrite = self.params.get('overwrites', True)
3781 if not self.params.get('writeinfojson'):
3782 return False
3783 elif not infofn:
3784 self.write_debug(f'Skipping writing {label} infojson')
3785 return False
3786 elif not self._ensure_dir_exists(infofn):
3787 return None
3788 elif not overwrite and os.path.exists(infofn):
3789 self.to_screen(f'[info] {label.title()} metadata is already present')
3790 return 'exists'
3791
3792 self.to_screen(f'[info] Writing {label} metadata as JSON to: {infofn}')
3793 try:
3794 write_json_file(self.sanitize_info(ie_result, self.params.get('clean_infojson', True)), infofn)
3795 return True
3796 except OSError:
3797 self.report_error(f'Cannot write {label} metadata to JSON file {infofn}')
3798 return None
3799
3800 def _write_description(self, label, ie_result, descfn):
3801 ''' Write description and returns True = written, False = skip, None = error '''
3802 if not self.params.get('writedescription'):
3803 return False
3804 elif not descfn:
3805 self.write_debug(f'Skipping writing {label} description')
3806 return False
3807 elif not self._ensure_dir_exists(descfn):
3808 return None
3809 elif not self.params.get('overwrites', True) and os.path.exists(descfn):
3810 self.to_screen(f'[info] {label.title()} description is already present')
3811 elif ie_result.get('description') is None:
3812 self.report_warning(f'There\'s no {label} description to write')
3813 return False
3814 else:
3815 try:
3816 self.to_screen(f'[info] Writing {label} description to: {descfn}')
3817 with open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
3818 descfile.write(ie_result['description'])
3819 except OSError:
3820 self.report_error(f'Cannot write {label} description file {descfn}')
3821 return None
3822 return True
3823
3824 def _write_subtitles(self, info_dict, filename):
3825 ''' Write subtitles to file and return list of (sub_filename, final_sub_filename); or None if error'''
3826 ret = []
3827 subtitles = info_dict.get('requested_subtitles')
3828 if not subtitles or not (self.params.get('writesubtitles') or self.params.get('writeautomaticsub')):
3829 # subtitles download errors are already managed as troubles in relevant IE
3830 # that way it will silently go on when used with unsupporting IE
3831 return ret
3832
3833 sub_filename_base = self.prepare_filename(info_dict, 'subtitle')
3834 if not sub_filename_base:
3835 self.to_screen('[info] Skipping writing video subtitles')
3836 return ret
3837 for sub_lang, sub_info in subtitles.items():
3838 sub_format = sub_info['ext']
3839 sub_filename = subtitles_filename(filename, sub_lang, sub_format, info_dict.get('ext'))
3840 sub_filename_final = subtitles_filename(sub_filename_base, sub_lang, sub_format, info_dict.get('ext'))
3841 existing_sub = self.existing_file((sub_filename_final, sub_filename))
3842 if existing_sub:
3843 self.to_screen(f'[info] Video subtitle {sub_lang}.{sub_format} is already present')
3844 sub_info['filepath'] = existing_sub
3845 ret.append((existing_sub, sub_filename_final))
3846 continue
3847
3848 self.to_screen(f'[info] Writing video subtitles to: {sub_filename}')
3849 if sub_info.get('data') is not None:
3850 try:
3851 # Use newline='' to prevent conversion of newline characters
3852 # See https://github.com/ytdl-org/youtube-dl/issues/10268
3853 with open(sub_filename, 'w', encoding='utf-8', newline='') as subfile:
3854 subfile.write(sub_info['data'])
3855 sub_info['filepath'] = sub_filename
3856 ret.append((sub_filename, sub_filename_final))
3857 continue
3858 except OSError:
3859 self.report_error(f'Cannot write video subtitles file {sub_filename}')
3860 return None
3861
3862 try:
3863 sub_copy = sub_info.copy()
3864 sub_copy.setdefault('http_headers', info_dict.get('http_headers'))
3865 self.dl(sub_filename, sub_copy, subtitle=True)
3866 sub_info['filepath'] = sub_filename
3867 ret.append((sub_filename, sub_filename_final))
3868 except (DownloadError, ExtractorError, IOError, OSError, ValueError) + network_exceptions as err:
3869 msg = f'Unable to download video subtitles for {sub_lang!r}: {err}'
3870 if self.params.get('ignoreerrors') is not True: # False or 'only_download'
3871 if not self.params.get('ignoreerrors'):
3872 self.report_error(msg)
3873 raise DownloadError(msg)
3874 self.report_warning(msg)
3875 return ret
3876
3877 def _write_thumbnails(self, label, info_dict, filename, thumb_filename_base=None):
3878 ''' Write thumbnails to file and return list of (thumb_filename, final_thumb_filename) '''
3879 write_all = self.params.get('write_all_thumbnails', False)
3880 thumbnails, ret = [], []
3881 if write_all or self.params.get('writethumbnail', False):
3882 thumbnails = info_dict.get('thumbnails') or []
3883 multiple = write_all and len(thumbnails) > 1
3884
3885 if thumb_filename_base is None:
3886 thumb_filename_base = filename
3887 if thumbnails and not thumb_filename_base:
3888 self.write_debug(f'Skipping writing {label} thumbnail')
3889 return ret
3890
3891 for idx, t in list(enumerate(thumbnails))[::-1]:
3892 thumb_ext = (f'{t["id"]}.' if multiple else '') + determine_ext(t['url'], 'jpg')
3893 thumb_display_id = f'{label} thumbnail {t["id"]}'
3894 thumb_filename = replace_extension(filename, thumb_ext, info_dict.get('ext'))
3895 thumb_filename_final = replace_extension(thumb_filename_base, thumb_ext, info_dict.get('ext'))
3896
3897 existing_thumb = self.existing_file((thumb_filename_final, thumb_filename))
3898 if existing_thumb:
3899 self.to_screen('[info] %s is already present' % (
3900 thumb_display_id if multiple else f'{label} thumbnail').capitalize())
3901 t['filepath'] = existing_thumb
3902 ret.append((existing_thumb, thumb_filename_final))
3903 else:
3904 self.to_screen(f'[info] Downloading {thumb_display_id} ...')
3905 try:
3906 uf = self.urlopen(sanitized_Request(t['url'], headers=t.get('http_headers', {})))
3907 self.to_screen(f'[info] Writing {thumb_display_id} to: {thumb_filename}')
3908 with open(encodeFilename(thumb_filename), 'wb') as thumbf:
3909 shutil.copyfileobj(uf, thumbf)
3910 ret.append((thumb_filename, thumb_filename_final))
3911 t['filepath'] = thumb_filename
3912 except network_exceptions as err:
3913 thumbnails.pop(idx)
3914 self.report_warning(f'Unable to download {thumb_display_id}: {err}')
3915 if ret and not write_all:
3916 break
3917 return ret